Memoirs of the California Academy of Sciences

n Endowment Symposium
iber 23,

Pleistocene Image and Symbol

: Edited by
" Margaret W. Conkey, Olga Soffer,

1 & Nina G. Jablonski

Wattis Symposium Series in Anthropology

MEMOIRS OF THE CALIFORNIA ACADEMY
OF SCIENCES, NUMBER 23

Swit HSON/A4D

NOV 0 4 1997

LIBRARIES

Beyond Art: Pleistocene
Image and Symbol

Editors

Margaret W. Conkey
University of California, Berkeley, California
Olga Soffer
University of Illinois, Urbana, Illinois
Deborah Stratmann
University of California, Stanford, California
Nina G. Jablonski

California Academy of Sciences, San Francisco, California

Wattis Symposium Series in Anthropology

MEMOIRS OF THE CALIFORNIA ACADEMY OF SCIENCES
Number 23

San Francisco, California

October 3, 1997

SCIENTIFIC PUBLICATION COMMITTEE:

Alan E. Leviton, Editor

Katie Martin, Managing Editor
Thomas F. Daniel

Robert C. Drewes

Michael T. Ghiselin

Nina G. Jablonski

Woyciech J. Pulawski

Adam Schitt

Gary C. Williams

(1997 by the California Academy of Sciences
Golden Gate Park
San Francisco, Calitornia 94118
All rights reserved. No part of this publication may be reproduced or transmitted in any
form or by any means, electronical or mechanical, including photocopying, recording, or any
Information storage or retrieval system, without permission in writing from the publisher.
Library of Congress Catalog Card Number: 97-68235
ISBN 0-940228-38-6 (cloth)
ISBN 0-940228-37-8 (paper)

Printed in the United States of America by the Allen Press

Distributed by the University of California Press

Beyond Art: Pleistocene
Image and Symbol

TABLE OF CONTENTS

Page

Pretace a suis adem tacks Se 2 Gas Hees Steen Sa ee oe 1X
AcknowledsomentS: «2-67 ota St 2 ee SS we we Xl
Studying Ancient Visual Cultures,

by Olga Soffer and Margaret W. Conkev... 1... 2.2... 1
Section I. Analytical Methods: From Dating to

Technologies of Image Making ................ 17
Paleolithic Marks: Archaeometric Perspectives

by Alan WGIChMman: g. 4a Sw Gis och ite a, 2 Meteo 1
New Laboratory Techniques and Their Impact

on Paleolithic Cave Art, by Jean Clottes .......... 37
Paleolithic Image Making and Symboling in Europe

and the Middle East: A Comparative Review,

By lexander Marsngck We slan, 6 soda: aad Rw 2 OD
Substantial Acts: From Materials to Meaning in Upper

Paleolithic Representation, by Randall White... ..... oe:
Section II. Approaches to the ‘“‘Why’s” of

Presence and ADSEMCE ners ites. oe we A 'cah Ge ee ee 123
The Power of Pictures, by Jain Davidson ........... 125
Symbolic Expressions in Later Prehistory of the Levant:

Why are They So Few? by Ofer Bar-Yosef......... 161

When the Beasts Go Marchin’ Out! The End of
Pleistocene Art in Cantabrian Spain,
by Manuel R. Gonzalez Morales ...........044. 189

TABLE OF CONTENTS

Section III. Interpreting Regional Imagery in the
European Upper Paleolithic

Art of the Light and Art of the Depths, by Jean Clottes . . .

Carving, Painting, Engraving: Problems with the Earliest
Italian Design, by Margherita Mussi and
Daniela Zampetti . .

The Mutability of Upper Paleolithic Art in Central and
Eastern Europe: Patterning and Significance,
by Olga Soffer.

Section IV. The Interpretive Process

Places of Art: Art and Archaeology in Context,
by Silvia Tomaskova .

Archaeological Signatures of the Social Context of
Rock Art Production, by Andrée Rosenfeld .

Processions and Groups: Human Figures, Ritual Occasions
and Social Categories in the Rock Paintings of the
Western Cape, South Africa, by John Parkington
and Anthony Manhire ... 2... .

Harnessing the Brain: Vision and Shamanism in Upper
Paleolithic Western Europe, by J.D. Lewis-Williams

Beyond Art and Between the Caves: Thinking About
Context in the Interpretive Process,

by Margaret W. Conkeyv .

Index

. 201

203

301

4

Preface

This volume takes its name from the title of the Second Paul L. and Phyllis Wattis
Foundation Endowment Symposium, which took place at the California Academy of
Sciences on March 11, 1995. The primary inspiration for the symposium came from
Deborah Stratmann, then of the Academy’s Department of Anthropology. Her keen
interest in the imagery created by humans of the latest Pleistocene and Holocene was
coupled with her recognition that the last 20 years had witnessed major changes in our
views of this imagery. New theoretical and technological developments had made
possible new investigations of the ages, raw materials and environmental contexts of
these images, and had brought into question our conception of them as “art.” Debo-
rah’s interests were matched by those of Phyllis Wattis, who had herself spent many
enjoyable hours in years past crawling through caves and lying on her back examin-
ing cave art in France. Together, they made Beyond Art happen.

After the symposium, it was decided that it would be fitting and most fruitful to
combine into one volume the proceedings of the Wattis symposium with those of the
First Oregon Archaeological Retreat, which was organized by Meg Conkey and Olga
Soffer and which took place in 1993. The result is a volume which epitomizes the phi-
losophy of the Wattis Foundation Symposia in Anthropology, namely, one that brings
diverse theoretical and methodological approaches to bear on a significant issue in
anthropology.

Nina G. Jablonski

Irvine Chair of Anthropology
California Academy of Sciences
June 12, 1997

Xi

Acknowledgments

It is next to impossible to appropriately acknowledge all individuals who have as-
sisted, aided, and abetted in the production of an edited volume such as this one — if
for no other reason than because there have been so many. Some, however, clearly
stand out in their contribution. Amongst these, first and foremost are Jean and Ray
Auel without whose generous hospitality there would never have been the First Ore-
gon Archaeological Retreat where this volume began. It was amemorable experience
that the participants will never forget — including the nouveau Oregonian combina-
tion of pizza and vintage Dom Perignon. Next, we wish to thank the California Acad-
emy of Sciences for independently confirming that the recent research on this topic of
late Pleistocene image and symbol had reached a critical point for a lively and in-
formed symposium. The Wattis Endowment, the many staff people and, most re-
cently, the active and supportive role played in bringing this volume to fruition by
Anthropology Chair Nina Jablonski and the Scientific Publications Committee must
be heralded. This volume owes its physical existence to the diligent and highly
skilled efforts of Nancy Gee, who performed the typesetting, Katie Martin, who as-
sisted in copy editing, and Vivian Young, who designed the cover and dust jacket.
We also thank Charlene Woodcock of the University of California Press for arrang-
ing that the Press would market the volume.

Finally, we extend our heartfelt thanks to the outside reviewers whose insightful
comments and detailed commentaries aided us enormously in sharpening our think-
ing and in improving the final product before you. Fora project to have four coeditors
and more than a dozen different contributors on four different continents has been a
challenge, and we thank everyone for their willingness to put together a combined
volume that we believe can begin to do justice to an enormously rich and potentially
revealing inquiry into the symbolic and material worlds of ancient humans.

Studying Ancient Visual
Cultures

Olga Soffer

Department of Anthropology
University of Illinois

109 Davenport Hall

Urbana, IL 61801

Margaret W. Conkey

Department of Anthropology
University of California

232 Kroeber Hall

Berkeley, CA 94720

The Background to the Volume

The volume before you addresses the appearance and distribution of images and
symbols encoded in material culture in prehistory — something collectively termed
prehistoric or Paleolithic “art.” This topic has, certainly, received a great deal of at-
tention and been a subject of much research with ensuing publications. A century of
study, however, has not produced a definitive theory about this “art,” but rather has
brought forth a number of conflicting claims. We address some of the factors respon-
sible for this state of affairs below, offer our own suggestions on how to advance our
understanding of prehistoric imagery, and begin doing so by decoupling this body of
archaeological evidence about past lifeways from its categorization as “art.”

We do so because, as we argue below, such initial categorization was both based
on unwarranted assumptions and constrained our understanding of the subject matter.
Because of this, we term this corpus of data as prehistoric imagery and, when using
the term, put “art” in quotation marks.

Our growing dissatisfaction with past approaches to prehistoric imagery, raised
against both insufficient attention directed to concrete times and places when the im-
ages were produced and used, against unwarranted uniformitarian assumptions, and
against a broadly ahistoric and often decontextualized frame of reference, led us to
organize two meetings to examine what we call, broadly speaking, prehistoric im-
agery froma more archaeological perspective. Specifically, we argued that there was

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

ty

SOFFER AND CONKEY

a need to examine this category of archaeological artifacts using an archaeological
frame of reference. After all, we do have evidence — usually very material evidence —
with which to construct context, and we do have an increasing repertoire of both
method and theory that is applicable to this enterprise. To begin this, thanks to the
generosity of Jean and Ray Auel, we gathered a group of specialists at the first Ore-
gon Archaeological Retreat in April, 1993 for two days of presenting data and arguing
over their significance. This was followed directly by a session at the 58th Annual
Meeting of the Society for American Archaeology to examine “What Means this Art?
~ Late Pleistocene ‘Creative Explosion’ in the Old World.”

Independently, a symposium was being planned by the California Academy of
Sciences, held in March 1995, which involved many of the same participants. As de-
scribed in the Preface the core of common issues and the range of approaches and re-
gions naturally led to the idea of a more robust publication that would combine the
participants from all events. This 1s the result of that endeavor.

Prehistoric Imagery vs. Prehistoric “Art”

The discovery, beginning in 1879, of painted caves in Spain and France, depicting
monochrome and polychrome images of the real and of the imagined Pleistocene bes-
tiary, together with enigmatic signs, was followed by much debate about their antiq-
uity (Bahn & Vertut 1988). Once the official scientific recognition was proclaimed in
1902 that these images could be accepted as the works of Upper Paleolithic peoples,
the remainder of the century saw not only the discovery of large numbers of examples
of Paleolithic images, both parietal and portable in form, but also a myriad of ap-
proaches to studying them (Conkey 1987, 1993). Most of these approaches began by
classifying these images as “art.” Yet were they?

Although most of us would agree that many of these images are indeed aestheti-
cally pleasing, to call them “art” is both misleading and limiting. There are two as-
sumptions behind the term “art” that are particularly troublesome. First, as defined
in the past century, art is a cultural phenomenon that is assumed to function in what
we recognize and even carve off separately as the aesthetic sphere. There are long-
standing debates, even within the western intellectual and cultural traditions about the
definition(s) of “art” (e.g., Kristeller 1980; see also Conkey 1997:174-176). This aes-
thetic function is something that we cannot assume to have been the case in prehis-
tory. In fact, ethnographic data from nonwestern cultures clearly show us otherwise;
most such cultures do not make the kinds of distinctions that we do, they do not have
an equivalent term for “art” nor do they often differentiate the aesthetic from the sym-
bolic from the sacred from the utilitarian, and soon. What we can say about this early
imagery is that it was created right from within the mundane world (after Alpers in
Schoch 1988). The so-called “artists” of 30,000 years ago did not discover some-
thing that is true for all humans at all times in all places — as in transcendent imagery
and art. Thus, as Tomaskova suggests, many of the traditional approaches of art his-
tory are not appropriate to the interpretation of prehistoric remains because we can-
not assume that their primary function was aesthetic.

The equally troubling second assumption behind the term “art” as applied to these
very early visual cultures is this idea that it is somehow transcendent and therefore
provides us with transcendent values. To claim Lascaux, for example, as the “first
masterpiece” of human art history reveals more about our cultural categories and
constructions than about the contexts of prehistoric social action within which the

STUDYING ANCIENT VISUAL CULTURES 3

production of such imagery as that at Lascaux took place and had meanings. While it
may be a significant contribution to contemporary life that we can find and draw upon
some assumed transcendent values from this imagery, nonetheless, as the ensuing
chapters clearly reaffirm, there is more to be gained from inquiry into this prehistoric
and early imagery. If anything, the view that it is transcendent has limited not pro-
moted deeper and more contextualized understandings (fora discussion, see Conkey
1995) .

By classifying prehistoric imagery as “art,” our predecessors brought with them
much unwarranted baggage. Before we had the global perspectives presented in this
volume, interpretation had been dominated by what researchers thought about the
Paleolithic imagery of southwestern Europe. For decades the richly preserved im-
ages on the cave walls there were taken as “the” origins of art. This Euro-focused
view of human evolution — with the pinnacle of symbolic expression manifest there
—has long been tangled up in the self-congratulatory history of the achievements of
“the West.”

From the very beginning of this interpretation, what was implicated was the ap-
pearance of modern humans in Europe and the appearance of early art, symbolism,
culture and modernity — a modernity marked by painting (itself a well-known hall-
mark of modernism and of Art in the 18th and 19th centuries). This created a nice ex-
ample of what Bakhtin (1981) has called an epic chronotrope: an entire period of
human life, in this case, was evoked through the trope of the origins of art.

The idea that we can explain human prehistory as part of a progressive
evolutionary trajectory is an idea that has now been weakened considerably. We
have begun to recognize and accept that the ground on which we stand to represent
others in the past is not firm, but more tectonic and shifting (after Clifford 1986:22).
The archaeological record tells us that the past has a mosaic rather thana lineal trajec-
tory: we may have anatomically modern humans in Australia before they were in
France; we appear to have cohabiting Neanderthals and modern humans, using simi-
lar tool kits in the Middle East and moving into the same sites and valleys, and we
have early image making in many places (for a discussion, see Klein 1995 with refer-
ences).

Today, we have refocused our attentions to include other locales, other contexts,
other situations and, most importantly, withdrawn from seeing this body of data as
representing “art.” In doing so we argue that such a catholic view is absolutely cru-
cial in reorienting the entire field of inquiry and interpretation.

The Significance of Image Making

Inthe contemporary world, regardless of our geographic and/or cultural contexts,
we find ourselves surrounded with symbol laden images. They are ubiquitous, but
always differentially valued and understood. These images may be three-
dimensional objects or two dimensional shapes and designs superimposed on flat sur-
faces: murals, sculptures, paintings, photography, and the like. Some are located in
more public places, while others are sequestered in private spaces. Some of these im-
ages are held to be aesthetically pleasing and exhibited as such. Others do not obvi-
ously function in the aesthetic domain but nonetheless signify and/or evoke values
and beliefs. All of them, however, carry emotionally charged meanings that are more
readily apparent to specific individuals and groups of people, and opaque, at best, to
others. What is universal to all of us, however, is that we all mark durable objects

4 CLOTTES

with symbols —a behavior strongly associated with anatomically modern humans and
a modern human way of life that is embedded in and expressed through culture.

As Geertz (1973) noted, humans traffic in symbols. We make sense of our world
by producing a symbolic one that includes real and imaginary creatures, myths, and
beliefs. Itis,in part, this association of modern humans and material symbolization
that makes this phenomenon a historical one — something that came into existence at
some point in time and appears not to have existed prior to that.

An Evolutionary Perspective

All people alive today are classified as belonging to a single subspecies, Homo sa-
piens sapiens —anatomically modern beings, whose ancestry as such, in various parts
of the world, dates back to between some 125,000 and 40,000 years ago (see Klein
1995, with references). It is this species, and this species alone, which is firmly and
unequivocally associated with the habitual and patterned symbolic marking of ob-
jects — a practice made manifest in material culture recovered from archaeological
sites dating from some 40,000 years ago onward (e.g., Klein 1995; Stringer & Mellars
1989, 1990, with references).

The tenure of our ancestors on Earth, however, goes back much further into the
past. Beginning with the deepest time, we know that the last common ancestor of AF
rican apes and humans lived some 5-7 million years ago. The oldest known hominids
— the australopithecines — were around in A frica from some 4.4 million years ago, but
their lifeways probably were only weakly, if at all, mediated by what we know as cul-
ture (Klein 1995 with references). The making and using of tools, the first harbinger
of technologically assisted existence, something which is essential but not sufficient
for a fully human way of life, is in evidence by some 2.3 million years ago and coinci-
dent with the appearance of the genus Homo. It is the later representatives of this ge-
nus who, by about | million years ago, expanded the hominid geographic range out of
Africa to inhabit as well the southern latitudes of Eurasia.

Tracing their descendants through time down to some 50,000 years ago, we have
evidence that they further expanded into the middle latitudes of Europe and Asia, re-
fining their technologies. What we do not see, however, except ina small handful of
highly contested enigmatic nonutilitarian objects, sometimes overmarked with lines
(see Bendarik 1996 with references), is the systematic and habitual production of
symbolically marked items of material culture, i.c., images. The reasons for their ab-
sence is a matter of heated debate, but generally interpreted to mean that life was or-
ganized differently than it is today or in the historic past (for this debate see the
Davidson and Marshack chapters, with their bibliographies, this volume). While the
absence of preserved material remains of accepted symbolic significance is not suffi-
cient to negate that our distant ancestors had symbolic capacities and activities, it 1s
the appearance, forms, and contexts of those symbolic manifestations that have pre-
served what provides a crucial baseline.

It is this practice of symbolic organization and mediation of life, of arbitrarily
ordering matter and energy, which signals the advent of fully modern way of life —
one embedded in and shaped by culture, dependent upon the making of meaning.
When we look back into the millennia over which anatomically modern humans ex-
isted, we see evidence for these practices in material form all across the then —
occupied world at about 40,000 BP (before present) from ostrich eggshell beads in
Kenya dating to some 39,000 years ago (Ambrose 1994), painting on plaquettes in

STUDYING ANCIENT VISUAL CULTURES 5

southern A frica at 26,000 years ago (Wendt 1976), parietal art in Australia that is pos-
sibly considerably more than 30,000 years old (but see Davidson, Rosenfeld, &
Watchman, this volume) to sculptured figurines made out of ivory in Austria and Ger-
many dated to some 32,000 BP (Davidson, this volume), to, finally, possibly the best
known examples, cave paintings in the Franco-Cantabrian region of Europe, some of
which are now dated to 30,000 years ago (Clottes ef al. 1995).

There is much debate in the literature about the proliferation of image-making ina
great variety of media as documented in the archaeological record from some 40,000
BP onward. Most of the attention has been directed toward the origin of this prolif
eration, and to what has often been termed “The Creative Explosion.” The production
and use of personal adornments, of small animal and human figurines, of decoration
ona myriad of implements made of antler, bone, ivory, and stone, and of painting , en-
graving, and sculpting on the walls of rock shelters and caves.

While some have argued that image making may well have gone on prior to this
marked appearance, although executed in perishable media that did not survive
through the ages, this is a problematic argument for archaeologists, and even an un-
acceptable one, in that it is empirically untenable. Without new lines of evidence, the
idea of a prior and equally rich material culture cannot be supported nor disproved.
Rather than focusing ona hypothetical scenario that, at least for the time being, can-
not be addressed, what we urge here is that we do more with the material remains that
we do have.

People have tried to answer the question of “why art at some 40,000 years ago and
not before?” in a number of ways, most of which come down to hypotheses about the
advent of some modern capacity for symboling among anatomically modern people,
from the invention of symbol-based language to proposed neurological changes or
some sort of “point mutation.” These explanations are intriguing but ultimately un-
satisfactory fora number of reasons. By equating the proliferation of image making
with the advent of the capacity to do so, such explanations disregard at least three
things: 1) the sparse and ambiguous presence of some images considerably prior to
40,000 BP: 2) the spatial discontinuity in image making across regions occupied by
equally modern humans (e.g., the abundance of Upper Paleolithic images in the
Franco-Cantabrian regions vs. their sparseness in the Levant, Africa and Asia); and
3) the temporal and even spatial discontinuities in the production of images within the
different regions (e.g., the virtual disappearance of image making in the Franco-
Cantabrian region with the advent of the Holocene vs. its proliferation in the Levant at
the same time). Furthermore, it is now clear that this image making cannot be equated
with the mere biological phenomenon of Homo sapiens sapiens, because apparently
our anatomically modern ancestors were present in Africa and the Middle East for as
many as 60,000 years before the notable material repertoire appears in archaeological
sites.

We suggest that there is a poor fit between explanations arguing that image mak-
ing appears when it does because of some new capacity for this behavior and the data
on hand because such hypotheses conflate two components of behavior. The two are
capacity and performance , and while the first is necessary for the second, the second
need not necessarily follow the first— it may, or may not, depending on contextual cir-
cumstances (Soffer 1994, 1995). Making this distinction, after Vygotsky (1986), and
evaluating each component against the archaeological record appears to indicate that
the capacity for image making may have been present among the ancestors of fully
modern people. The presence of admittedly few unique pieces from Middle Paleo-
lithic and possibly earlier contexts suggests that the capacity for making and using

6 SOFFER AND CONKEY

such objects could well have been in place. What was different prior to the Upper Pa-
leolithic or its African equivalent, the Late Stone Age, was that this capacity was not
regularly or routinely exercised. What was absent, then, was what we have termed
performance, the habitual use of the capacity in patterned behavior. What was absent
were the social and cultural contexts within which such material practice would have
been efficacious and would have had appropriate meanings. Sucha perspective re-
quires us to begin examining the circumstances under which image making becomes
habitual, anapproach that can help us understand the temporal and regional disconti-
nuities in the global record.

Although the debate over origins will not readily go away, this volume focuses on
the dynamic and variable nature of early image making for two reasons. First, it is
only now that researchers are beginning to document and interpret the evidence ona
more global scale, with an eye towards understanding the contexts and histories of re-
gional and local manifestations. Second, to probe and understand the range and vari-
ability of the earliest accepted image making is crucial to the task of inquiring into
how and under what circumstances such material symbolization came into habitual
use.

The Historical Context

As we noted in the beginning of this chapter, the acceptance of the antiquity of pre-
historic imagery was followed by numerous approaches to its study. While each re-
gion of the world with prehistoric images has experienced its own traditions of
research, here we focus on the research conducted on the classic region of Franco-
Cantabrian cave art and do so to provide a glimpse into the efforts and problems to
date, and how the approaches brought together in this volume might begin to redress
these and to reconceptualize what we can do.

The first broad approach, and the one that has dominated most of the discussion to
date, was directed towards deciphering the meaning of the imagery. We direct read-
ers interested in the history of this research to such works (in English) as Bahn & Ver-
tut (I988), Sieveking (1979), Leroi-Gourhan (1965, 1982) or Ucko & Rosenfeld
(1967) (see also Delporte 1990, but in French). The net result of this avenue of re-
search has provided nearly as many explanations for the meaning of this category of
material culture as scholars addressing the subject, e.g., totemism and sympathetic
hunting magic, initiation rites, fertility cults, the gendered structuring of culture, aes-
thetic sensibilities, erotic fantasies or achievements, self-representations at different
stages of life, efc., each with claims to being the correct one. There are many prob-
lems with an approach that seeks “the” meaning(s), given that an archaeology of
meaning is understandably fraught with challenges that come from many sources: un-
developed theoretical approaches, partial evidence, methods that are stil] experimen-
tal, and, most importantly, the extreme difficulty of eliciting meaning for “deep
prehistory” for whom we have no ethnographic, ethnohistoric or textual anchors.

These past searches for meaning began with unjustified uniformitarian
assumptions that some meanings are universal or transcendent. This, however, is ne-
gated by the very essence of symbols, namely, their meaning(s) 1s arbitrary and cul-
turally determined. Furthermore, researchers tried to explain the entire corpus of
imagery — made over some 25,000 years ina variety of geographic locales and ina
multiplicity of media — by reference to a single overarching account, such as sympa-
thetic hunting magic (fora discussion, see Conkey 1987, 1993, 1996). We know now

STUDYING ANCIENT VISUAL CULTURES 7

that cultural meanings do not work this way: They are not an eternal static construct,
but bounded in time and to context.

A second approach, heavily influenced by the writings of structuralists, also
looked for meaning by scrutinizing the patterning of images (see, for example,
Leroi-Gourhan 1965, 1982). It assumed that the patterning had its sources in the
structure of human cognition, which was seen as universal. Leaving aside the contro-
versial issues dealing with the nature of cognition, the main methodological weak-
nesses in this approach, which analyzed the structure of a presumed composition,
were that it took as a given that the images indeed are part ofa composition that: a) has
integrity, and b) was created instantly or over a very short period of time. This, as
Clottes (this volume), among others, demonstrates, is clearly something that we can
neither assume nor demonstrate (also see Conkey 1989).

The third broad approach to the Franco-Cantabrian Paleolithic art, one which saw
it as a medium for the transmission of information, also paid little attention to chro-
nology. Using information theory as its basis, some scholars argued that all images
carried messages between the creators of the images and their viewers and users. This
perspective fit well with archaeological approaches that saw human adaptations as
being about obtaining matter, energy, and information (see Gamble 1991 with refer-
ences). While many new insights have been gained from focusing on Paleolithic im-
ages as signifiers — part of the processes by which humans use images and the
material world to situate themselves in their social and symbolic worlds — many of
these approaches failed to contextualize the imagery in time and place. Specifically,
scholars rarely asked either about the “referential context(s) of social action” (after
Hodder 1982:16) or about what might have made these specific images meaningful
to their makers and users. If we accept the idea that what imagery does Is to “galva-
nize the production of meanings” (after Bryson 1983:x1ii), the contextual features
must be closely considered in specific relation to attributes and aspects of the imagery
itself.

Our predecessors in the interpretive process of Paleolithic imagery reached their
own solutions, but one generation’s solutions often become the next generation’s
problems or challenges. The history of the study of early human imagery is one that
we are inevitably caught up in. We can neither go back nor extricate ourselves from
the histories of interpretation that have become a part of interpretation itself. With
each new generation, or even with each new study, we are adding a layer to what is a
veritable palimpsest of understandings and of interpretations, and we are enmeshed in
that palimpsest.

The Perspective of this Volume

As archaeologists, we saw that our past approaches were not using the images in
their local and historical contexts to understand the wider phenomena of human mate-
rial symbolization and of image making as material practice in specific settings. The
approach we advocate and the one that is reflected in this volume, is to examine pre-
historic images as we would any category of archaeological artifacts. This involves a
dual look that includes a study of the artifacts themselves as well as the context in
which they were made and used.

We asked our authors to examine one of the components of the emerging, more
complex interpretive equation. When focusing on the images themselves and exam-
ining anumber of possible variables (e.g., the aesthetics involved in their production,

8 SOFFER AND CONKEY

the technology used and technological choices made, the possible meanings that the
images might have had for their users or the relationship between the creation of an
image and its meaning) we asked for rigorous patterning and, if possible, converging
lines of evidence. In examining the context(s) of production and use of varied sym-
bolic repertoires, we encouraged studies to take up the chronology involved in the
production of the record on hand, to examine the patterning, ifany, in the distribution
of the images, and finally, to investigate any associations between the image and the
medium used for its expression.

We also sought explicitly regional approaches, so that the closer focus of under-
standing the discontinuous distribution of imagery and the disparate forms and con-
texts used could be addressed. Within regions we encouraged, where possible, the
investigation into variation: temporal, spatial, technological, in materials, and so
forth.

Scholars whose works are assembled here see themselves as students of material
culture whose work is simultaneously empirical and interpretive. To help with the in-
terpretive part of the endeavor, we borrow widely from social theory to extremely
technical scientific methods (such as the x-ray diffraction of pigment composition).
In spite of this, we do not offer a definitive scenario about prehistoric imagery. Many
of us insist on both the challenges and possibilities of what ts a more realistic situa-
tion, namely, that archaeological data and the issues we explore are full of ambiguity,
and that this is to be exploited (for examples of this in other archaeological subjects
see Tringham 1991, Schrire 1995 ).

Inhis assessment of a related field, Bryson (1983, especially xiii-xiv) has argued
persuasively that art history is a provisional or necessarily incomplete enterprise and
that to accept this is a situation that should enable growth, more than signal the limita-
tions of the field. We suggest that the same holds true for the study of early “art.”
Nothing can arrest the continual transformation of the imagery into meanings, “codes
of recognition” circulate constantly through imagery, the viewer is an interpreter, in-
terpretation changes, and we'll never lay claim to absolute knowledge of our object.
The archaeology and the interpretation of early “art” is necessarily always provi-
sional.

When we interpret, even the simplest cultural notions that we put forth are inten-
tional creations. Furthermore, as interpreters, we constantly construct ourselves
through the others we study. Few people in the audience of early “art” research are
there purely to find out about the past, separate and distinct from the present, separate
from an inquiry into ourselves and our places in our own cosmological schemes.

We all tell stories and this is the process of interpretation, the process of making
inferences from archaeological remains. All of science does this (Terrell 1990).
Some tell better stories than others and the agreed-upon stories often change. This
volume attests to how some agreed-upon stories are changing and are going to
change. Certainly this happens because there are new finds, new data, new facts: but
our stories also change because of new assumptions, because new interpreters have
taken up the topic, or because of the shifts in the geopolitical frameworks of re-
search. The results of this approach are before you. The ensuing chapters do not pur-
port to be the final word on the subject at hand. To the contrary, we suggest that the
approaches to the study of prehistoric images offered in this volume represent prom-
ising new directions that have already and will continue to yield robust insights into
the question of why and when images are made immortal in permanent media. We
hope that the very term, Paleolithic “art,” will be redefined and, more importantly,
re-conceptualized by these new directions.

STUDYING ANCIENT VISUAL CULTURES 9

The Organization of the Volume

We have divided this volume thematically into four sections, which examine cen-
tral issues in understanding prehistoric image making. Given that many of the issues
considered crosscut these heuristic categories, it is inevitable that the reader will find
some overlap between the sections. Thus, for example, questions about the regional
distribution of certain categories of images are addressed in more than one section.
Similarly, issues about the possible significance and meaning of some of the images
necessarily surface in discussions other than those found in the last section of the vol-
ume. We have attempted to keep such overlap to a minimum and are certain that our
readers will understand that some is unavoidable.

The first section, Analytical Methods: from Dating to Technologies of Image-
Making discusses recent insights gained from the use of diverse dating methods to es-
tablish the chronology of image making, presents results from the application of
methods of material science to understanding of how the images were made, as well
as insights obtained from a consideration of the production sequences involved in the
making of these images.

The past decade has been particularly exciting in terms of the new ways in
which images have been studied which have moved our studies beyond questions of
identification (“What is it? A horse? A ‘sign’?”) into the realms of “when” and
“how.” While many new analytical techniques are both available and seem to work
today, we should not expect that they have solved all of our problems. For example,
the control of chronology has been a significant interpretative dilemma given that so
few of the images on cave walls have been or can be dated directly (see Clottes and
Watchman, this volume); in fact, many have suggested that the inclusion of imagery
and “art” in mainstream archaeology has been inhibited due to this problem of dating.
While new methods of dating and the ability to now date images quite directly has
been or enormous boost in both the credibility and kinds of interpretations now possi-
ble, nonetheless problems remain. The plethora of chronometric dates available now
for some of the Upper Paleolithic sites have shown us that the fundamental issue of
chronological resolution is unlikely to be simple and straightforward, and we must
continue to complement our increasingly refined techniques for dating with other
lines of evidence and inference.

The first two chapters in this section, by Alan Watchman and Jean Clottes
respectively, discuss dating techniques and other laboratory techniques that can be
brought to bear on the cultural questions that we want to address. Watchman’s review
of the issues of dating imagery on rock surfaces stems explicitly from archaeometry
and focuses on potential dating methods for images that lack carbon-bearing context
and thus cannot be dated by Carbon-14. Although focusing on his own multiple ana-
lytical techniques applied to imagery in Australia, he also discusses issues related to
the dating of the European Paleolithic images (including the most recent heated de-
bate about the rock engravings and the Foz Céa dam project in Portugal). The devel-
opment of these new dating techniques and the accumulation of comparative absolute
dates (see Davidson, Chapter 6, Table 1) are clearly providing the stimulus and
knowledge for evaluating previous dating methods, including archaeological and sty-
listic approaches, and for challenging existing views about the antiquity of particular
rock art styles.

In the next chapter, Jean Clottes focuses on some laboratory techniques that
have recently contributed to the analysis and interpretation of Paleolithic cave images

10 SOFFER AND CONKEY

in France. He considers the recent analyses of the “paints” used to create some of the
imagery. These results have provided evidence for deliberate mixes or “recipes” that
were used both on wall images and on pieces of mobile art on bone/antler. These pio-
neering pigment and paint studies have brought us much closer to the individual acts
of image makers, with hints at the dynamics of image-making itself. They document
much variation both in how cave painting was done and in the use-life of caves for im-
age making.

Clottes also addresses the issue of dating the imagery. The recent dates obtained
for four of the painted caves in France, which challenge standard stylistic notions
about chronology, indicate that what we need for meaningful interpretations are not
only suites of dates but as many other lines of corroborating, and preferably independ-
ent, evidence as we can muster.

Alexander Marshack, in the next chapter, shows how a very close analysis of
en-graved marks and of individual motifs, usually by means ofa microscope, can be a
potentially powerful mode of analysis of the materials themselves. In the context ofa
brief review of some of his research along these lines for more than 25 years, he con-
siders some of the notably less abundant Paleolithic as well as early Holocene objects
from the Middle East. His findings lead him to suggest that both the early farmers of
the Middle East as well as the Upper Paleolithic peoples of Eurasia had a comparable,
although contextually different, conceptual framework of a “time-factoring” and
“time-factored” cultural matrix.

In the last chapter in this section, Randall White uses Leroi-Gourhan’s (1964,
1965) anthropological concept of the chaine opératoire to analyze some Upper Paleo-
lithic images from Europe. This term, best translated as “the operational sequence,”
refers to studies of the conventionalized sequence of technical operations, which are
inherently cultural, that brings an object or artifact into existence. Considering per-
sonal ornaments from the Aurignacian period (dating between some 40,000-28 ,000
years ago) and anthropomorphic figures from the Gravettian period (dating between
some 29,000 and 22,000 years ago), he documents the complex relationships between
the raw materials selected, the application of differing techniques/steps/stages for
manufacture, and what we know about the contexts of production, use, and disposal,
and how we might use these lines of inquiry to not only dismantle the monolithic ap-
proach to interpreting Paleolithic “art” but also to infer more situated understandings
of what must have been culturally appropriate but also very individualized practices
by Paleolithic peoples.

Like other new analytic techniques, the chaine opératoire is simultaneously an
enlightening as well as a potentially problematic approach. There are assumptions be-
hind the concept that need to be both identified and addressed. For example, we need
to understand better the relationship between specific technological behaviors and
conscious human choice, which we know is always culturally negotiated.

The close, thoughtful, and detailed analyses of the materials we have to work with
will provide us with more contextualized, nuanced and local understandings of the
imagery. Since image making is nota vehicle for the mere perception of meaning, e1-
ther by us or by past peoples and because image making and the imagery itself pro-
duces meanings, it is in the scrutiny of the production of the images themselves that
we may find a crucial link into comprehending what made the imagery meaningful
for prehistoric peoples. This will come from the conjoining of technical analyses with
theoretically-framed studies of particular objects, in particular media, made by speci-
fiable productive techniques, and through culturally and socially meaningful techni-
cal acts.

STUDYING ANCIENT VISUAL CULTURES 1]

Section IH, Approaches to the Presence and Absence, consists of three chapters
that focus on the possible reasons for the appearance and disappearance of image
making in prehistory, and reconceptualizes this early image making as a more local
and historic phenomenon, linked to specific cultural ways of being in the world.

In the section’s first chapter, lain Davidson, taking a comparative approach that
juxtaposes data from Europe to those from Australia, argues that the similarities in the
development of early image making shows the universal use of art to reflect the nego-
tiation of environmental relationships: both of the natural and of the social environ-
ments involved. He sees the burst of image making after some 40,000 BP as reflecting
the exploration of the limits of a recently discovered power of symbolically based
communication. In his scenario, the discovery of symbolism first leads to the mark-
ing of social identity through symbolic means and then develops into the marking of
place.

Ofer Bar-Yosef restricts himself geographically to the Levant in the next chapter,
but examines its record over a wide span of time — from the Upper Paleolithic through
the early Neolithic. His detailed look at this regional record through time, one which
shows the virtual absence of image making and use before the Natufian period, per-
mits him not only to identify a similar trajectory in the development of image making
as outlined by Davidson, but to argue that this development is associated with particu-
lar changes in the demographic and social histories of the region. This leads Bar-
Y osef to conclude that we can only understand the presence or absence of image mak-
ing through regional and historical perspectives — ones that see modern humans mak-
ing and using images to address social problems that are grounded ina specific time,
place, and social realities.

Manuel Gonzalez-Morales, in the next chapter, takes up a unique perspective on
the “art” of the European Paleolithic by asking why did it disappear? Given the predi-
lection for identifying the origins of image making, Gonzalez-Morales draws our at-
tention to the other end of a question that probes the contexts. From this vantage
point, he considers the subject matter of Upper Paleolithic cave art in Cantabrian
Spain and begins witha detailed critique of the shortcomings of previous interpreta-
tions which, as he argues, were based on highly subjective selective favoring of one
variable from amongst many. In addressing why this record of spectacular cave
painting disappears so suddenly at the end of the Paleolithic, Gonzalez-Morales iden-
tifies a developmental trajectory in the subject matter of this art. He points to the con-
cordance between the animals hunted and those painted during the Early Upper
Paleolithic (before 20,000 years ago) and charts the divergence of the two through
time. He argues that this divergence signals the decoupling of mythology from the
perceived real world and how this is derived from changes in economic behavior and
social organization.

Although dealing with disparate times and places, the three chapters in this
section suggest that we can identify some universal, albeit broad, trajectories in the
development of image making and in its use. On the one hand, these chapters rein-
force the potential of undertaking the more local approach that does not try to account
for specific sets of imagery in terms of overarching and monolithic accounts. Further,
they show that the symbolic marking of material culture is not a universal phenome-
non of human behavior but something that is sometimes done and, at other times, not
done. On the other hand, once this kind of material practice does come into use, how-
ever, itmay often first be used to mark social identity and , later, to signal corporate
entities. Although one type of marking need not necessarily follow the other, the or-
der of their appearance is likely significant and may well reflect increasing complexi-

12 SOFFER AND CONKEY

ties of social worlds occupied by larger numbers of people who remained together on
a more permanent basis.

Section III, Contextualizing and Interpreting Regional Imagery in the European
Upper Paleolithic, contains three chapters that examine parts of the European record
and document significant variability in time and across space both in the abundance
and in the composition of image making. We have chosen to focus on the European
record here precisely because this record has been studied the longest by the greatest
number of specialists. Because of this, it is the most appropriate one with which to
document intraregional variability.

Jean Clottes, in the next chapter, examines the binary trajectory outlined by pre-
vious scholars for the location of Paleolithic images: Those placed “in the light,” at or
near the entrances of caves and rock-shelters, and those found “in the dark,” deep in-
side caverns. His survey concludes that similar frequencies for imagery in both lo-
cales can be observed throughout the Upper Paleolithic time and in all of the
Franco-Cantabrian region, but with the one and notable exception of the later part of
the Magdalenian_ period (after ca. 14,000 years ago) when the frequency of image
making and use in deep caves appears to increase significantly. His documentation of
this change indicates that the places where the images were placed — in or near the
light and next to occupied sites, or in the dark in the rarely and ephemerally visited lo-
cations — may have been specially selected for, and that by the Middle Magdalenian
this choice favored the “dark.”

Margherita Mussi and Daniella Zampetti, in the following chapter, discuss in de-
tail the Italian record throughout the Upper Paleolithic. This 1s itself an important
contribution, since the region is not often included in the classic syntheses of Upper
Paleolithic “art,” which have focused on the Franco-Cantabrian region. Furthermore,
from their assessment they conclude, first and foremost, that this region did not, as
previously postulated, form a thematically separate Mediterranean artistic province.
Noting the general similarities in the art and items of personal adornment from this re-
gion to those from Franco-Cantabria, they document also local micropatterning in the
nature and volume of symbolically marked items of material culture. Thus, for exam-
ple, in spite of the presence of caves, the Italian record has almost no parietal paint-
ings and few engravings. In addition, they also document the waxing and waning in
the volume of the pieces recovered, and apparently produced — from abundant during
the Gravettian period to sparse during the Epigravettian (18,000-14.000 BP) to volu-
minous again during the Final Epigravettian (14,000-10,000 BP). They superimpose
this record on the demographic one documenting both the sequence of colonization
and abandonment of the different regions by people through time, as well as the in-
crease in the numbers of people inhabiting the region through time. They argue that
this concurrence is causally related and that, in general, we can best understand the
production and use of this category of material culture as material expressions of par-
ticular, but as yet unspecified, solutions to issues of biological and social reproduc-
tion.

In the next chapter, Olga Soffer reaches similar conclusions about the re-
lation-ship between the intensity of artistic production and demographic factors.
Working with the record from Central and Eastern Europe, she documents regional
variability in subject matter and medium as well as in context of use and volume of
production. Underscoring the heterogeneity evident in the depictions of females, for
example, she argues for the inadequacy of any one single explanation for their signifi-
cance and advocates instead semantic polysemy. Examining the changes evident in
the manner of depicting females through time, which are progressively schematized,

STUDYING ANCIENT VISUAL CULTURES e}

she argues for a progressive restriction of access to the semantic significance or
meaning of these images and sees this as a reflection of the rise of decoding special-
ists, that is, of people who explained their meaning to others.

The three chapters in this section also come up with some similar conclusions.
First, the documented diversity in artistic production clearly argues for a diversity in
meanings from place to place, through time, and likely between not only groups but
individuals. Second, the Italian and Central and East European records, just like those
from the Levant and Cantabria discussed in Section III, show variability through time
in the intensity of artistic output and use. Third, they infer that there was a restriction
of access to at least some symbolically marked items through time that is manifest in
a variety of ways, from increased use of deep caves to schematization and geometri-
zation of subject matter previously depicted more naturalistically. These conclusions
resonate with those advanced by Gonzalez- Morales, who argues that the Cantabrian
Upper Paleolithic record suggests a separation of mythology from the natural world
and its transfer to the supernatural one. This patterning, if it can be sustained, adds
more insights to the hypothesized developmental scenario for the uses of symboli-
cally marked artifacts through time in those areas where this material practice was in
use.

Section I, The Interpretive Process, includes five chapters that assess the more
theoretical aspects of what interpretation has involved and offer richly grounded in-
terpretive scenarios that draw on ethnography, ethnohistory, and other cultural mod-
els that enable the authors to get closer to the meanings and illustrate the complexity
of what we are grappling with in a study of early imagery.

In preceding sections we learned from closely-honed studies that early image
makers knew the properties and possibilities of their materials. This section points
out that materials might have been selectedas much for their symbolic properties as
for their physical ones: For the shine they make, for the feel they produce, for the ef-
fect of transparency that may result, or for the resonance they may have with experi-
ences and visions.

Chapters in this section are concerned with the basic idea of image making as a
social practice and with more nuanced understandings of what archaeologists often
take for granted as context. Taken together, they underscore that the imagery studied
is not transcendent and does not provide us with transcendent values. Realizing this
directs us to deal with the context of their production, a concept that is also complex
and mutable.

This is dealt with by Sylvia Tomaskova in the next chapter, who illustrates the
point that context is historically contingent. She also points out that context means
different things in archaeology and inart history. Furthermore, within archaeology it-
self, different kinds of artifacts are approached differently, depending on what we
have inferred their contexts to be. Tomaskova discusses the impact that national and
intellectual traditions have had on our research by focusing on the disparate treatment
of materials from the Gravettian Willendorf (in present-day Austria) and from the
coeval Pavlov sites (in the Czech Republic). Her discussion also underscores that the
comparative method itself, a mainstay in the interpretive tool kit of anthropology and
archaeology, is by its very nature a highly selective tool because what you select to
compare influences your conclusions.

Andrée Rosenfeld, in the following chapter, uses generalized ethnographic and
ethnohistoric insights into the making of rock art to establish some parameters within
which interpretation can proceed. First, she documents that Australian rock art is nota
monolithic entity but that the images and their production articulate with aspects of

14 SOFFER AND CONKEY

individual identity, supraindividual identity, and territoriality. Rosenfeld conveys
how rock art is both a vehicle for the expression and determination of a social land-
scape as well as continually open to constant negotiation for varied social purposes.

In the next chapter, John Parkington and Tony Manhire deal with Holocene im-
agery in rock painting, even if such practices do have documented Pleistocene antiq-
uity in southern Africa. Their contribution, together with those of Rosenfeld,
Marshack, and Bar- Yosef, remind us that the demarcation of a research boundary be-
tween the Pleistocene and Holocene ts an artificial heuristic device.

While examining the same subject matter — South African rock art — as does
Lewis-Williams in the following chapter, they come up with considerably different
conclusions. We have included both approaches in this volume because they reflect
an ongoing dialogue among researchers in southern A frica about the centrality of sha-
mans and trance in interpreting the imagery. For Parkington and Manhire the imagery
as part of how people came to understand their social and sexual relations, often, but
not always, through the vehicle of trance. Specifically, they suggest that the imagery
is about being a hunter or a gatherer and about the assumption of adult roles rather
than about hunting and gathering per se.

In the next chapter, David Lewis-Williams works with South African and Upper
Paleolithic European imagery to imply that what image making does is to actually
“galvanize a production of meaning” (Bryson 1983: x11); that is, to propel, mobilize,
or bring forth what may be multiple and cascading meanings, with multiple refer-
ents. Using neuropsychological evidence about the optical images produced during
altered states of consciousness, together with the evidence on the pervasiveness of
shamanism in hunting-gathering societies, he further develops his hypothesis that the
European Upper Paleolithic cave imagery can best be understood as shamanistic (see
Lewis-Williams and Clottes 1996),

The implications we can draw from this chapter are that at least some of the images
in Upper Paleolithic caves are not likely to be just animals, “signs,” females or males
that somehow can be read directly as a list of species, nor can they be treated only as
ethological or gynecological specimens. Rather, these are representations, these are
“sites,” as Davis (1984) has suggested, where material practices and representation
come together to form and re-form the experiences of some people in various times,
places, contexts, and circumstances.

In the final chapter, Conkey reflects on some of the major concerns that the con-
tributors have broached in this volume, such as the need to take up a more contextual
approach and the very challenging and prickly subject of interpretation, especially
since the very spirit of the volume is to open new avenues for interpretation in the
study of early and ancient image making. It is a moment at which the study is now
global in scope and is addressing issues that are “beyond art” and beyond a search for
an overarching account. The term, “context” itself has varied meanings and refer-
ences; she explores as well the “constituting contexts” of our own research. Lastly,
she briefly describes the early stages of a research project that seeks to better under-
stand the “social geographies" within which at least one distinctive regional visual
culture of late Upper Paleolithic France (the Magdalenian of the Midi-Pyrénées) was
practiced and produced.

laken together, these last chapters illustrate well, with both theory and empirical
observations, the point we made earlier, namely, that the images themselves are poly-
semic and that therefore there not only can be, but must be polysemy in our interpreta-
tions of them. This concluding section confronts how we explicitly or implicitly
engage in the interpretation of prehistoric imagery. These are not just peripheral is-

STUDYING ANCIENT VISUAL CULTURES 15

sues specific to the individual personalities or national preferences of researchers, or
to asmall segment of the field. They are central issues in all of archaeology, in the in-
terpretation of human culture, and in the human sciences, namely, how do we know
what we know? What can we know and is there such a thing as absolute or secure
knowledge? Are we in the business of producing knowledge to control humans or are
we engaged with our materials in order to produce humanly-compelling knowledge
and ideas that matter, ideas and knowledge that can themselves be used to transform
and invigorate our communities and lives? The appeal of the study of early human im-
agery and symbol is that it provides us with these kinds of humanly compelling ideas.

Coda

We conclude with a word about the chronometric dates presented in the volume.
In all cases the dates given, presented as assays BP (Before Present), represent un-
calibrated dates, usually obtained through radiocarbon dating of organic remains as-
sociated, wherever possible, with the images in question. For readers who prefer to
begin with some chronological grounding, we suggest first a review of Table 1 in the
Davidson chapter, where he provides a list of dates for a set of European cave paint-
ings.

Literature Cited

Ambrose, E. S. H. 1994. Technological change, volcanic winter, and genetic spread of modern
humans from Africa. Paper presented at the Third Meeting of the Paleoanthropology Soci-
ety, Anaheim, CA., April 19-20.

Bahn, P. & J. Vertut. 1988. /mages of the Ice Age. Bellew Publishing Ltd., London, U.K.

Bakhtin, M. 1981. The Dialogic Imagination. University of Texas Press, Austin.

Bendarik, R. G. n.d. The most ancient known manifestations of art. Pages 133-142 in Beltran
and A. Vigliardi, eds., Art in the Palaeolithic and Mesolithic Colloquium, XV1, XIII Inter-
national Congress of Prehistoric and Protohistoric Sciences. Forli A.B.A.C.O, Italy.

Bryson, N. 1983. Vision and Painting: The Logic of the Gaze. Yale University Press,

New Haven.

Clottes, J., J-M. Chauvet, E. Brunel-Deschamps, C. Hillaire, J-P. Daugas, M. Arnold, H.
Cachier, J. Eving, P. Fortin,C. Oberlin, N. Tisart, & H. Valladas. 1995. Radiocarbon dates
for the Chauvet-Pont-d'Are Cave. /ntl. Nwsltr. Rock Art 11:1-2.

Conkey, M. W. 1987. New approaches in the search for meaning? A review of research in Pa-
leolithic art. J. Field Arch. 14:413-430.

_ 1989. The structural analysis of Paleolithic art. Pages 135-154 in C.C.Lamberg-
Karlovsky, ed., Archaeological Approaches in the Americas. Cambridge University Press,
Cambridge, U.K.

_1993. Humans as materialists and symbolists: Image making in the Upper Paleolithic.
Pages 95-118 in D. T. Rasmussen, ed., The Origin and Evolution of Humans and Human -
ness. Jones & Bartlet, London and Boston.

. 1996. Paleovisions: Intepreting the imagery of Ice Age Europe. Pages 401-419 in S.C.

Scott, ed., The Art of Interpreting. Pennsylvania State University Papers in Art History, X

Pennsylvania State University, University Park, PA.

16 SOFFER AND CONKEY

—_. 1997. Mobilizing ideologies: Paleolithic “art”, gender trouble, and thinking about al -
ternatives. Pages 172-207 in Lori D. Hager, ed., Women in Human Evolution. Routledge,
London, U.K..

Delporte, H. 1990. L'image des Animaux dans L'art Préhistorique. Picard, Paris, France.

Gamble, C. 1991. The social context for European Paleolithic art. Pages 3-15 in Proceedings of
the Prehistoric Society 57, part 1.

Geertz, C. 1973. The Interpretation of Cultures. Basic Books, New York.

Hodder, I. 1982. Theoretical archaeology: A reactionary view. Pages 1-16 in Il. Hodder, ed.,
Structural and Symbolic Archaeology. Cambridge University Press, Cambridge, U.K..

Klein, R.G. 1995, Anatomy, behavior, and modern human origins. J. World Prehist. 9:167-
198.

Leroi-Gourhan, A. 1965. Treasures of Prehistoric Art. Abrams, New Y ork.

—. 1982. The Dawn of European Art. Cambridge University Press. Cambridge, U. K.
Lewis-Willliams, D. & J. Clottes. 1996. Le Chaminisme dans le Préhistoire. Le Seuil, Paris.
Sieveking, A. 1979. The Cave Artists. Thames and Hudson, London, U.K.

Soffer, O. 1994. Ancestral lifeways in Eurasia — The Middle and Upper Paleolithic records.
Pages 101-119 in M. H. Nitecki and D. V. Nitecki, eds., Origins of Anatomically Modern
Humans. Plenum, New York.

1995. Artistic apogées and biological nadirs: Upper Paleolithic cultural complex ity re-
considered. Pages 615-627 in M. Otte, ed., Nature et Culture, E.R.A.U.L. 68.

Stringer, C. & P. Mellars, eds. 1989, 1990. The Human Revolution, 2 vols. Edinburgh Univer-
sity Press, Edinburgh, Scotland, U.K.

Ucko, P. & A. Rosenfeld. 1972. Paleolithic Cave Art. McGraw-Hill, New York.

Vygotsky, L.S. 1986. Thought and Language. Translation revised and edited by A. Kozulin.
MIT Press, Cambridge, USA.

Wendt,W. E. 1976. ‘Art Mobilier’ from the Apollo 1] Cave, South West Africa's oldest dated
works of art. S. Afr. Archaeol. Bull. 35:5-11.

Section I

Analytical Methods: From
Dating to Technologies of
Image Making

Paleolithic Marks:
Archaeometric Perspectives

Alan Watchman

Department of Archaeology
James Cook University of North Queensland
Townsville, Australia

In the last five years technological developments have enabled the extraction of
micrograms of carbon in paint and laminae under and over paintings making it possi-
ble to date small samples of Upper Paleolithic rock art. Plant fiber binders, blood,
charcoal and bees wax resins have been used in the radiocarbon dating of organic
components intimately associated with relatively recent and Upper Paleolithic rock
paintings. However, new techniques are needed to date ancient paintings and engrav-
ings where organic substances are not part of the human rock marking process itself.
Carbon-bearing components in the ‘canvas’ and overlying ‘varnish’ have been radio-
carbon dated using accelerator mass spectrometry to constrain paintings and engrav-
ings between maximum and minimum time limits. Bacterial, fungal and algae
remains, calcium oxalate minerals and charcoal particles in rock surface accretions
covering rock art have been identified as potential sources of datable carbon. In the
process of examining and dating these substances, so that their associated rock art can
be dated, we have also come to appreciate the complex association between carbon
directly related to a painting event and background (contaminating) sources of envi-
ronmental carbon. Development of new scientific dating techniques and the accumu-
lation of comparative absolute dates are providing the stimulus and knowledge for
evaluating previous dating methods, including archaeological and stylistic ap-
proaches, and for challenging existing views about the antiquity of particular rock art
styles. This paper presents information about the methods and problems of dating
rock surface accretions associated with Paleolithic marks on rocks, particularly paint-
ings that do not contain organic binders.

Beyond Art: Pleistocene Image and Symbol Memoirs of the Califommia Academy of Sciences
Editors, M, Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

20 WATCHMAN

Dating Paleolithic Rock Art
Stylistic Perspective

Fundamental to discussions about Paleolithic rock art is the question of antiquity
and how a painting or engraving ona limestone cave wall fits into the sequence of A.
Leroi-Gourhan (1965), whose chronology has formed the basis of stylistic dating in
Europe. The key to allocating a motif to a Paleolithic period is the identification of
stylistic traits in the art, and this is where dating by style runs into difficulty. Firstly,
because the perception of a particular style is very subjective and variable through
time, with the stylistic specialists being unable to recognize consistently those mark-
ers that can be used irrefutably to place a figure precisely in time (Conkey & Hastorf
1990). Secondly, the chronology of ‘type’ paintings and the acceptance of the linear
evolution of Paleolithic rock art at Niaux (Clottes et a/. 1990:21), Quercy (Lor-
blanchet et a/. 1990, 1995), and at other Rhone Valley and Quercy sites (Lorblanchet
etal. 1990) has changed over the last 80 years even though the specialists at the time
have used essentially the same stylistic criteria to support their opinions. Thirdly, the
use of style as a chronological tool in rock art studies has waned as exceptional cases
have been found, such as the Grottes Cosquer and Chauvet, and although the use of
style has been strongly criticized and rejected by many scholars (Bahn 1993; Bed-
narik 1995), some prehistorians prefer to maintain the stylistic dating tradition de-
spite the recent advances in dating technology (Clottes, this volume). The
breathtaking naturalistic depictions of groups of horses, bison, fighting rhinoceros,
bears, and other animals in the Grotte Chauvet, that fit the typical Paleolithic conven-
tion, contain properties that to an art historian’s eye might appear too advanced for
their age. This recent discovery therefore not only challenges the prevailing notions
about the evolution of Paleolithic style through time (refer to the radiocarbon deter-
minations in Table 1, Davidson, this volume), but it also makes us rethink our ideas on
the development of human culture. Fourthly, the extent to which reliable scientific
methods have been used to develop stylistic chronologies is seldom reviewed criti-
cally by prehistorians, and the timing may now be right to do so because the new dat-
ing methods have revealed some surprising results. As we will see later in this volume
even when separate archaeometric dating methods independently arrive at scientific
evidence contrary to prevailing opinions about the age of rock art (determined stylis-
tically), considerable argument and publicity still cannot persuade prehistorians to re-
consider their entrenched chronological perspectives (for example at Foz Coa,
Portugal).

In trying to date Paleolithic rock art today we therefore find ourselves in a contro-
versial dichotomy, between the stylists and the archaeometrists (see below). The styl
ists, on the one hand, support the comparison of artistic conventions and traits seen in
mobiliary art (engraved bone, amber and plaques), and those apparently similar fea-
tures recognized in the paintings and engravings on cave walls and in open air set-
tings. On the other hand, the archaeometrists measure physical and chemical
attributes and compare their results with subjective descriptions of earth surface
weathering processes and landforms associated with the environment surrounding
the panels of art. As described in the following chapters the various specialists do not
always agree on the antiquity of particular figures thought to be Paleolithic, possibly

ARCHAEOMETRY AND PALEOLITHIC IMAGE 21

because an aura, ‘mystique,’ envelops Paleolithic art, allowing emotions to influence
analysis.

The Archaeometric Perspective

One of the most frustrating aspects for archaeologists studying Paleolithic rock
pictures has been the lack ofa reliable scientific method to establish their absolute an-
tiquities, rather than their relative ages according to inferred archaeological and sty-
listic evidence. During the last decade, and especially in the last four years, major
progress has been made to reduce much of that frustration by the development of new
dating methods. However, in place of frustration has come anxiety, and even hostil-
ity, about the reliability of these new dating methods. For the archaeometrist the chal-
lenges of absolute dating rock art are enormous, exciting and hopefully rewarding,
but there are also substantial limitations, pressures and conflicts.

Unfortunately, in the rush to be the first to produce dates for an established rock art
style using a new dating method, the archaeometrist often does not have the opportu-
nity, funds and resources to examine fully all possible causes of potential errors. Ini-
tially this does not seem to disturb the professional archaeologists, prehistorians and
enthusiastic amateurs because a new date for another rock art style is exciting news.
However, the archaeometrist had better produce a date that satisfies existing theories
about the antiquity of the art or else the new method will soon become obsolete, as the
potential errors rapidly become the noose and gallows for the academic lynch mob.

At the time of the Wattis Foundation symposium (March 1995), these remarks
were probably obtuse to many in the audience, but the circumstances rapidly became
reality during the controversy that occurred several months later when petroglyphs in
the Céa Valley, Portugal, were threatened by flooding. Four independent archaeome-
trists, using separate techniques on different samples determined that the engravings
were most probably of recent antiquity (Bednarik 1995; Watchman 1995), and their
results contradicted the stylistic specialists who had earlier stated that the engravings
showed characteristics typical of Paleolithic works of art (Bahn 1995; Clottes 1995).
The major implication inferred by many Portuguese archaeologists from the archaeo-
metrists’ measurements was that the results decreased the value of the engravings,
heightening fear that the dam project would go ahead. Anger, frustration, high fervor
and great anxiety of the dam opponents led to vitriolic personal attacks and almost li-
belous accusations of corruption and complicity against the archaeometrists. By not
obtaining results compatible with the stylists’ views the archaeometrists felt the full
passion and force of the traditional prehistorians, a situation which potentially threat-
ened further refinement and development of these new dating methods.

Admittedly, the history of dating rock art using archaeometric methods is only
short compared with the use of style and archaeological methods, and because of this
it seems that some prehistorians regard the methods as experimental (Clottes et al.
1995), and therefore defective. Archaeometrists have their own concerns about the
assumptions and techniques of the new dating methods, as illustrated in the following
example. In 1983 a potential approach for dating varnished rock engravings, the cat-
ion ratio method, was realized (Dorn 1983). Archaeologists who had been deprived of
a scientific dating method for establishing the antiquity of their rock engravings ea-
gerly welcomed this magic formula. It was the chronological elixir, the salvation and
inspiration for bored archaeologists and geomorphologists because now they could
accurately fix their interpretations in time and relate rock art motifs to other well dated
cultural relics and place landform evolution in established paleoenvironmental peri-

22 WATCHMAN

ods. However, many questions arose about the method (Bednarik 1988; Clarke 1989;
Clegg 1988: Lanteigne 1991; Reneau & Harrington 1988; Watchman 1989), and few
of them were answered to the satisfaction of geochemical critics. The method was re-
viewed by its founder (Dorn 1989), and a few critics wavered in their opposition.
Some archaeologists became true believers and promoted the method so they could
use it to date their rock engravings, espousing support when subsequent dates didn’t
destroy their hypotheses.

However, skeptics did not believe the hyperbole and conducted their own inde-
pendent research of rock varnishes, to examine the assumptions of the new dating
method and the analytical procedures used. Serious doubts arose (Bierman &
Gillespie 1994; Dragovich 1988; Krinsley & Anderson 1989; Raymond e¢ al. 1991;
Reneau & Raymond 1991; Reneau 1993; Watchman 1992a, b), and the cation ratio
dating method has now lost support among geomorphologists and geochemists. A
few archaeologists remain cation ratio dating disciples despite the increasing weight
of independent analytical evidence against its universal application for dating all rock
varnishes. Cation ratio dating is still used when no other method seems possible
(Francis etal, 1993). Results obtained have been startling (Dorn et al. 1992; Nobbs &
Dorn 1988), and perhaps for this reason alone Dr. Dorn has succeeded in drawing at-
tention to the possibility of dating very thin coatings on rock surfaces. Seemingly out-
rageous Claims of great antiquity for rock engravings in North America (Whitley &
Dorn 1994), equally the Magdalenian in Europe, have stirred archaeological and an-
thropological interest in rock art of the New World. The significance and value of
rock art has now taken on new meaning for a variety of specialists, not only for deci-
phering when metaphorical symbols were pecked into rocks (for example Martineau
1990; Patterson-Rudolph 1993), or when shamanistic compositions were painted
(Lewis-Williams, this volume), but also to improve our understanding of human cog-
nitive development through the Paleolithic, because it now seems possible to obtain
absolute dates for individual motifs.

The challenge for geochronologists and archaeometrists is, and probably always
will be, to satisfy the archaeological and anthropological experts that new dating re-
sults are scientifically reliable despite results that contradict current theories about the
antiquity of certain painting and engraving styles. Comparisons are inevitably made
with existing chronologies whose bases either rely on personal interpretations of slim
stylistic and archaeological evidence, or on invalid assumptions. For instance, com-
paring the antiquity of charcoal fragments found on the floor of caves today with simi-
lar radiocarbon measurements obtained from charcoal particles scraped from
paintings on nearby cave walls (e.g., Grottes Chauvet [Clottes et a/, 1995], Cosquer
[Clottes et al. 1993] and Téte du Lion [Combier 1984]), should not be used by prehis-
torians to prove the antiquity of the paintings. Following Bednarik's reasoning (Bed-
narik 1989:10), coincidence between such pairs of dates only suggests that the
charcoal pieces on the floors were probably used to make the paintings, it does not
prove that the paintings were done at the time determined by the amount of "'C re-
maining in the charcoal samples. The results do, however, reveal that depositional
and erosional forces in caves are extremely slow, and that pieces of charcoal can re-
main oncave floors for long periods, making them readily available for the creation of
post-Paleolithic works of art.

The archaeometrist, with little money available for substantive testing, must rely
on one or sometimes a handful of dates to demonstrate potential in a new dating
method, whereas academic prominence may only be necessary to fix a rock art style
in antiquity. An archaeometrist can expect praise and glory following good correla-

ARCHAEOMETRY AND PALEOLITHIC IMAGE 23

tions between the first few scientific dates and the expected archaeological age of a
painting or engraving. If the age is older than expected, then that proves human devel-
opment was even earlier than previously believed and so the work is commended. The
inventor of the dating method is acclaimed, funds flow for more research and a re-
spectable academic career is guaranteed. Few critics appear. However, dates that
challenge existing archaeological opinions about the great antiquity of a particular
rock art style provoke strong opposition from academics and rock art enthusiasts. The
new dating method is denounced and the archaeometrist’s credibility is censured (as
typified by the Céa Valley controversy), but few prehistorians seem willing to chal-
lenge the archaeological evidence and the assumptions used in the stylistic arguments
that support the prevailing opinion.

On a personal level, my approach to the dating of rock art comes from conserva-
tion science and the use of various analytical tools to authenticate works of art on can-
vas. In studying easel paintings the conservation scientist examines the canvas, the
ground layer beneath the paint, the pigment colorant, binder and extender, and the
overlying varnish, and determines antiquity using the known availability of various
manufactured components and by radiocarbon dating the organic components. When
this approach is adapted to prehistoric rock art the canvas ground layer and varnish
become natural rock surface coatings such as oxalate crusts, secondary calcite layers,
rock varnishes and silica skins. Often these deposits contain carbon-bearing sub-
stances and by dating them an estimate can be obtained for sedimentary layers di-
rectly associated with paintings and engravings, therefore making it possible to date
rock art. Charcoal particles and organic binders in paints can also be radiocarbon
dated, but this is where the problems start for the archaeometrist.

Constraints on Sampling and Data

What scope is there for adequately testing a new dating method? The archaeome-
trist who aims to date rock art is restricted in many ways, but primarily by three major
factors. Firstly, rock paintings and engravings are regarded by most people as cultural
artifacts worthy of preserving either for aesthetic pleasure, for their archaeological
and scientific values or for their unique links with existing or previous cultural tradi-
tions. Indigenous peoples regard rock pictures as direct expressions of their culture.
They are often highly significant images placed in specific locations for special, se-
cret or sacred activities. It is inappropriate for a archaeometrist to take large pieces of
a painting or stone slabs bearing rock pictures because the aesthetic damage would be
obvious, and the spiritual impact immense. Consequently, after obtaining permission
to sample from traditional owners and modern managers, the archaeometrist is lim-
ited to taking only small samples of painted or engraved rock.

Secondly, sample size for dating carbon extracted from components associated
with paintings and engravings is constrained by the minimum weight of graphite ion-
ized by the source of an accelerator mass spectrometer (AMS). For most spectrome-
ters this is ideally 100 ug of graphite; equivalent to 2.5 ml of CO; gas, 5 mg of wood
(cellulose) or 3 mg of charcoal. To meet this specification the piece of painted or en-
graved rock must be sampled so that enough graphite is produced for dating after the
sample has been pretreated. Under highly stable accelerator operating conditions and
very low background count rates measurements can be made on as little as 50 wg of
graphite. Therefore, in order to obtain more than 50 wg graphite the paint or mineral
coating must have more than 1.5 mg charcoal or 2.5 mg cellulose fibers.

24 WATCHMAN

This size limit fora graphite target is not always possible to achieve, especially in
the first few samples collected at a new painting or engraving site where the organic
content is unknown. Importantly, each new sampling site has different paint and asso-
ciated mineral deposits and until analyses are done to establish the carbon content in
these materials the precise amount of graphite that is possible for dating is uncertain.
For example, in the recent dating of Bradshaw style paintings in Western Australia
that are thought to be at least 15,000 years old (Walsh 1994), powder scraped from sil-
ica covered paint across 10 cm? of rock face and weighing 8.9 mg produced insuffi-
cient graphite for CAMS dating (only 2 wg). Similar situations have been found
elsewhere because paint does not contain an organic binder and insufficient carbon is
contained in mineral skins associated with art (Watchman & Lessard 1993). Deciding
how much paint can be removed from paintings without causing discernible damage
is a difficult situation for archaeometrists, rock art custodians and for the traditional
owners.

Thirdly, the archaeometrist must be able to collect in one way or another datable
substances that are directly associated with recognizable rock art. To obtain an
archaeologically acceptable date the association between the carbon bearing sub-
stance that is dated and a painting or engraving must be clearly demonstrated (Batten,
et al. 1986). Such an objective is not always easy to achieve (for example, Loy et al.
1990: Loy 1994: Nelson 1993), because organic binders were not used in every paint
recipe (Pepe et al. 1991; Wainwright 1985; Watchman 1993a, Watchman ef al.
1993), many years may have passed between the time of paint and pigment applica-
tion or engraving and its masking by mineral salts or rock varnish, and it may not be
physically possible to remove a flake of rock from where the archaeometrist and pre-
historian would like to take a sample. Alternative means for dating paintings must
therefore be used. If we cannot date a painting precisely because carbon-bearing com-
ponents are not present in the paint itself, then our next best option is to date carbon
bearing substances above and below a paint layer to define the approximate painting
time limits. For an engraving, we must find and date carbon bearing substances 1m-
mediately above it to determine its minimum limiting age.

We cannot date precisely the surface of an engraving or the surface under a paint
layer because the boundary itself has no contact or extractable substance (Bourgeois
1990). Furthermore, the accuracy of boundary definition at any one locality can be
highly variable, especially at the nanometer scale of rock surface accretions, thin
paint layers and engraved surfaces. Irregular surface features such as vermiform tex-
tures (worm-like network deposits), microstromatolitic mounds and spot concentra-
tions (pimples) produce rough, microtopographically undulating surfaces, and these
are difficult to sample accurately. Physical and biological disturbances decrease the
accuracy of delineating the boundaries between paint and the ‘sandwiching’ layers of
minerals because ingredients mix along the upper contact in particular, resulting in
mixtures that may have formed over a considerable period of time. Diagenetic pro-
cesses, such as postdepositional chemical alteration also affect the boundary between
paint and superimposed mineral layers, or between an engraving and its overlying
rock varnish.

Bright red, pale yellow and dull black laminations are occasionally observed by
microscope when some rock surface mineral layers are cut at right angles to the layer-
ing, and these colored layers are thought to represent past painting episodes. While
these traces of paint in sections, and similar ones occurring as apparent splotches on
rock walls, signify the presence of humans rather than natural depositional processes
they do not necessarily make a recognizable painting. They could be smudges or acci-

ARCHAEOMETRY AND PALEOLITHIC IMAGE

Ne
n

dental spills and therefore not of cultural significance. On the other hand, rock art con-
servators may be able to use microexcavation techniques in the future to reveal what
some of these encrusted paintings look like.

Bracketing a paint layer by dating substances in layers on both sides of an ‘artistic’
boundary at least gives an approximate time frame for its production, with maximum
and minimum limits. The amount of time represented by dating carbon bearing sub-
stances in a ‘layer’ removed by physical excavation (in a micrometer scale excava-
tion) is unknown because the measured age is an average of all carbon substances in
that excavated sample. Determining the average sedimentation rate ina series of adja-
cent laminations through a well laminated succession of accreted dust and salt in-
creases the degree of confidence in the age determination for a layer adjacent to paint.
Measuring an instantaneous sedimentation rate is generally not possible, but the mod-
ern annual deposition rate can be recorded, and this will indicate approximately how
long it takes for components to accumulate in a lamination. Very slow sedimentation
rates, periods of virtually no deposition or intervals of erosion complicate estimates
of accumulation rate, making it even more important to carry out predating examina-
tions and analyses.

Detailed petrographic and chemical compositional studies combined with pa-
leoenvironmental information help the archaeometrist to understand accretion for-
mation processes and changes in the accumulation rate over time, and through
variable climatic conditions. For reliable dating of laminated coatings the archaeome-
trist must know that depositional processes have been regular, and that erosional epi-
sodes have been of significantly smaller scale than the time span in question.

An additional sampling problem is caused by the hardness of the underlying rock
because natural surface weaknesses that result in partly detached exfoliation flaking
may not occur precisely where the archaeometrist wants to take a sample of painted
stone. Chiseling or drilling into the surface with hollow diamond bits makes unsightly
and unacceptable scars, so paint on hard quartzite, limestone or sandstone may not
easily be removed. This means that designated figures representing ‘type’ examples
of a particular style may not be datable.

The following sections look at the possibilities for dating crusts and skins, and at
two recently developed techniques for extracting carbon bearing substances from
rock art. I then conclude this paper by reviewing scientific dates obtained for charcoal
(particularly for Paleolithic art), organic binders, oxalate salts, and organics in silica
skins and rock varnishes.

Bracketing Rock Art in Time

In examining rock art surfaces as part of condition assessment for conservation
projects I recognized that many Australian rock paintings were naturally covered by
finely laminated, dusty gypsum and oxalate deposits (crusts), or lustrous amorphous
siliceous coatings (skins; Watchman 1987, 1990, 1992c). I found datable carbon-
bearing substances in these rock surface accretions; the calcium oxalate minerals
whewellite (CaC,O4.2H,O) and weddellite (CaC2O,.H2O) in crusts, phytoplankton
(diatoms), fungi, bacteria and algae in silica skins, and fine charcoal particles in both,
These datable carbon-bearing substances exist in stratified microsedimentary layers
directly beneath and over paint. Although the paint itself cannot be dated directly be-
cause it does not contain carbon bearing substances the potential exists for indirectly
dating the period between deposition of the ‘canvas’ under, and the ‘varnish’ over a

26 WATCHMAN

painting. Establishing a time bracket for many paintings is therefore possible pro-
vided the carbon bearing substances and mineral deposits are coeval and the carbon
can be extracted from its microstratigraphic position without contamination.

Oxalate in Layered Crusts

Many Australian rock paintings are on the relatively smooth, dusty back and side
walls of sandstone, limestone and quartzite rock shelters. Although these surfaces are
superficially protected from direct rain they slowly accumulate evaporitic salts, such
as gypsum, polyhalite, and other hydrated sulphate and nitrate minerals, that crystal-
lize from splashed and sprayed rain water droplets that moisten shelter walls during
torrential storms. In dry periods dust particles from a variety of local and regional
sources also settle on these surfaces, eventually becoming cemented by salts into hard
dark crusts.

Over time these salt and dust deposits gradually increase in thickness, reaching 4
mm on stable low angle ledges. In cross section the crusts usually consist of many fine
laminae representing changes in the compositions of airborne components that have
settled and crystallized on the surface reflecting environmental fluctuations. Increase
in gypsum is tentatively correlated with warm and dry, conditions whereas oxalate
salt is thought to form when the rocks were frequently warm and damp. Laminae
dominated by quartz and clay imply hot and dry conditions, and abundant particulate
charcoal associated with quartz and clay represent periods of drought, regular re-
gional fires, dust storms and human occupation.

One of the simplest ways to date a painting whose paint does not have an organic
binder is to determine the AMS radiocarbon ages of charcoal particles in laminae un-
der and over the paint (Watchman 1993b). However, not all crusts have enough par-
ticulate charcoal associated with layers directly associated with paintings for the
masked paintings to be dated. The other difficulty is to extract the charcoal particles
from each layer without introducing modern contaminants, a problem that is poten-
tially solved by the laser extraction method which is described later.

The back walls of many rock shelters are ideal habitats for insects, fungi and algae,
especially when conditions are frequently damp and warm. Insect exoskeletons, com-
posed mainly of chitin, are incorporated into crusts with salt and dust particles. Fungi
metabolize dead organic material that litter the dusty salt encrusted surfaces, and are
themselves integrated into the deposit with other crust components. Under warm and
humid conditions algal colonies develop, making use of moisture, indirect sunlight
and trace elements in dust to establish themselves on a stable nutrient rich substrate.
Oxalic acid is often produced as a consequence of these microorganic activities and
by the degradation of organic substances. Reactions between oxalic acid and dust par-
ticles, particularly calcium, lead to crystallization of calcium oxalate salts such as
whewellite. As carbon in whewellite can be dated by AMS '!C the layer containing
that salt, and other layers with particulate charcoal can be used to date past episodes of
painting (Watchman 1993b; Watchman & Campbell in press).

Fossilized Microorganisms in Laminated Silica Skins
Translucent to white opaque coatings composed of hydrous amorphous silica

cover many painted rock faces throughout the world. They are formed by deposition
of silica from runoff and seepage water that flow regularly across rock surfaces.

ARCHAEOMETRY AND PALEOLITHIC IMAGE Pag

When silica precipitates rapidly from evaporating water the bacteria, fungi and algae
that inhabit these moist rock faces become encapsulated in the resulting silica skin.
Complete diatom frustules are often preserved in the coatings, and bacterial and algal
fatty acids are identifiable (Watchman 1992c), so these synsedimentary organic and
inorganic deposits can be radiocarbon dated (Watchman 1994). The approach is simi-
lar to removing carbon-bearing organic binders from paint and radiocarbon dating the
age of the painting, but in the case of silica skins it is the age of silica deposition that ts
determined.

It is assumed in this dating method that chasmolithic and endolithic microorgan-
isms do not invade the siliceous coatings, and that only carbon from fossilized micro-
organisms is dated, otherwise their presence would lead to a false age for silica
deposition. Detailed microscopic observations and organic chemical tests are used to
establish the validity of these assumptions. The additional assumption that no subse-
quent microorganic or inorganic process has introduced carbon-bearing substances
into the silica seems very reasonable as the fossilized carbon is enclosed by encapsu-
lating silica and a chemical process would be necessary to penetrate the silica film in
order to contaminate the fossil carbon. If these assumptions are valid, and obvious
physical contamination from rootlets is not a problem, then the age of the organic
matter in the silica skin will be the same as the age of silica deposition. Control over
the reliability of this method is achieved by the internal consistency of results for an
accretion. This is achieved by dating a sequence of stratigraphic layers in these thin
coatings and establishing conformability; a process that overcomes the problem of a
direct date alluded to by Clottes (this volume). Dating organic matter in laminated sil-
ica under and over a paint layer therefore provides a very good estimate of the age of
that painting. Dating silica deposited in an engraving gives a minimum age for that
engraving, whereas dating the fossilized organics in silica on an adjacent surface
through which the engraving was made establishes a maximum age (contrary to the
views of some prehistorians; Clottes ef a/. 1995:20).

One example where organics trapped in silica have been used to provide a maxi-
mum age for paintings (effectively the canvas age, to use the conservation science
analogy, but unfortunately not Paleolithic), is at the Nisula site, Québec, where an-
thropomorphic figures, vertical bars and abstract designs have been painted in hema-
tite ona glacially polished cliff face. The paint rests directly ona thin (0.08 mm thick)
film of hydrated amorphous silica that has formed by deposition out of seepage water
that runs infrequently across the gneiss. Fossilized organic matter in the silica skin
that was combusted and dated by AMS '4C gave 2,500 + 275 radiocarbon years BP
(OZA403); consistent with archaeological expectations for paintings approximately
2,000 years old (Arsenault 1995:27). A second sample from approximately one meter
away gave a radiocarbon measurement of 2,440 + 610 years BP (OZB350), substanti-
ating the first result for the maximum painting age. Although these outcomes and
other dating results for silica skins (Watchman 1994) demonstrate the method's po-
tential for dating rock art the contradictory results for the allegedly Paleolithic en-
gravings at Foz Céa, Portugal (Watchman 1995), and the hostile response that this
aroused in stylistic prehistorians does not encourage application of the method for
dating other European rock engravings.

Secondary Calcite Laminae

One research area that potentially offers a way of validating the radiocarbon meas-
urements on charcoal from Paleolithic figures on the cave walls of France, and else-

28 WATCHMAN

where, is the dating of secondary calcite (calcium carbonate) deposits formed over
painted surfaces. Using the conservation science approach, one should be able to de-
termine microstratigraphic conformability between the age of secondary calcite pre-
cipitation and evidence of paint in a layer covered by a carbonate film. Combining
uranium series dating (Schwarez 1982) and radiocarbon methods to establish the age
of secondary calcite formation, with the radiocarbon dating of charcoal from a paint-
ing surrounded by secondary calcite should give independent control over the age of
the painting. Unfortunately, while this approach has only been attempted for engrav-
ings in Australia (Bednarik 1984) and China (Huashan er a/. 1987) on mid-Holocene
surfaces it has not been attempted on any Paleolithic painting sites in Europe despite
my repeated urgings for scientists to do so.

Organic Matter in Layered Rock Varnish over Engravings

Rock varnish, consisting essentially of iron and manganese oxides, occasionally
covers engravings on exposed rock surfaces. Inorganic varnish components have
been dated by the cation ratio method (Dorn 1983, 1989), and organic substances
trapped in basal varnish layers have been dated by AMS '"C (for example, Dorn er al.
1992: Francis eral. 1993; Whitley & Dorn 1994). The fundamental premise of the cat-
ion ratio dating method is the semilogarithmic relationship between the ratio of (Ca +
K) / Ti (the cation ratio) and the age of the same varnish. Although an attempt was
made to indicate cation leaching sites in varnishes to explain the gradual loss of mo-
bile calcium and potassium (Dorn & Krinsley 1991, Watchman 1991), no valid scien-
tific explanation exists for the experimentally derived decreasing ratio with
increasing varnish age.

A calibration curve is constructed for rock varnishes in an area by dating organic
matter encapsulated in the basal parts of varnishes using AMS "'C and by determining
the cation ratio for its accompanying varnish (Dorn 1989). The basic assumptions
necessary to establish a cation ratio calibration curve are: (1) that varnish continually
forms by accretion of dust particles and there is no physical, chemical or biological re-
working of the components; (2) varnishes at the calibration site are developed in the
same way as varnishes at engraving sites selected for cation ratio dating: (3) the less
than 2 um fraction of dust in rock crevices provides a measure of the regional ‘initial
ratio’ for the onset of varnish formation; (4) rock varnish probably starts to form about
100 years after a rock surface is exposed; and (5) organic matter in varnishes is con-
temporaneous with its accompanying inorganic components.

The cation ratio dating method has been strongly challenged on theoretical, statis-
tical (Lanteigne 1989, 1991), textural, and geochemical bases (Fleming & Barnes
1994; Reneau & Harrington 1988: Reneau & Raymond 1991; Watchman 1989;
1993c), and substantial doubts about its integrity exist among many archaeometrists,
geomorphologists and geochemists. One example of a potential problem with the
method is that subvarnish organic matter from a well-layered varnish may have been
eroded from older varnish and deposited over younger material as demonstrated at a
calibration site in Australia where plant fibers in basal varnish (1,375 + 60 years BP,
AA7722). were found to be younger than similar fibers in overlying layers (6,970 +
100 years BP, NZA2570; Watchman 1993c).

Applying the assumptions of the cation ratio dating method to every varnished
rock surface without undertaking compositional and microstratigraphic studies 1g-
nores the highly complex nature of rock varnish formation, varnish inhomogeneity
and compositional variability (Bierman & Gillespie 1994; Raymond et al. 1991; Re-

ARCHAEOMETRY AND PALEOLITHIC IMAGE 20

neau & Raymond 1991; Reneau 1993). Calibration curve reproducibility is question-
able, even using varnishes from one site (Watchman 1992a), and verifying cation
ratio dates with AMS !4C ages of subvarnish organic matter (Whitley & Dorn 1994),
does not provide a truly independent antiquity check because a cation ratio calibration
curve is constructed using varnish organic matter ages.

Recent Carbon Extraction Procedures

Existing methods for collecting organic substances associated with rock paintings
include the physical extraction of charcoal (David 1992; Valladas et al. 1990, 1992),
plant fibers (Watchman & Cole 1993) and beeswax (Nelson et a/. 1992), and the
chemical dissolution of blood (Loy et a/. 1990). Removing calcium oxalate minerals
and fatty acids derived from algae, fungi and bacteria is generally not practicable by
either physical or dissolution methods, especially the latter as all carbon is generally
removed from the sample. Two methods, oxygen plasma and laser induced oxidation,
have recently been developed for removing trace amounts of carbon from rock sur-
face accretions and paints. The oxygen plasma method essentially removes molecular
organic carbon rather than oxalate and carbonate carbon, whereas the laser method
combusts charcoal and molecular carbon, and decomposes oxalate minerals. Selec-
tive use of laser power and wavelength enables the preferential extraction of separate
carbon bearing substances in a laminated rock surface accretion thereby maintaining
archaeological sampling integrity while minimizing contamination. These two dis-
tinct methods, combined with AMS "C dating, now provide two of the best means for
extracting carbon from binders and accretions associated with Paleolithic rock art.

Oxygen Plasma

Oxygen plasma oxidation is used where organic binders or charcoal are known to
exist in paints (Chaffee er a/. 1993; Russ et al. 1990). In this process a painted rock
sample is placed in a high vacuum reaction chamber made of borosilicate glass, and
subjected to a low temperature oxygen plasma produced by two external electrodes
connected to a radio frequency generator. Reaction of oxygen radicals, formed when
the generator is switched on, with the organic phase produces carbon dioxide and wa-
ter which are collected in liquid nitrogen traps attached to the reaction vessel. A fter
sufficient gas is liberated it is frozen ina glass tube, sealed offand sent toan AMS fa-
cility for determining its '*C content (Chaffee et a/. 1993). Application of the method
for dating charcoal scraped from two horses at Le Portel has yielded radiocarbon de-
terminations of Recent Magdalenian age (Igler ef a/. 1994).

Laser-Induced Photooxidation

The laser extraction method is used to convert organics, particulate charcoal and
oxalate minerals, observed in laminae revealed in thin cross sectional slices of a
painted skin or crust, into carbon dioxide (Watchman & Lessard 1993). Small slices
of encrusted rock are mounted in a vacuum tight combustion chamber, the air is re-
placed by pure oxygen at low positive pressure, and light from a Krypton ion continu-
ous laser (413 nm) is focused through a glass window in the chamber to a spot
diameter of 0.003 mm directly onto a lamination (Watchman & Lessard 1993). Car-
bon dioxide that is produced by photooxidation of carbon bearing substances under

30 WATCHMAN

the beam is collected ina glass tube immersed ina liquid nitrogen trap. When the pro-
cedure is complete the CO; is converted to graphite for AMS "C dating.

For some inexplicable reason there appears to be considerable confusion amongst
some geomorphologists and prehistorians about the use of this laser technique for dat-
ing rock art (Clottes, this volume; Werlhof et a/. 1995). While the method was devel-
oped to demonstrate the potential for extracting small quantities of carbon-bearing
substances from laminated mineral accretions and paint layers not a single radiocar-
bon determination has been obtained for a rock painting or engraving.

Paleolithic Dating Results

Results for Charcoal

Accepting at face value the radiocarbon dates for charcoal in supposedly Paleo-
lithic paintings may not be wise, as the following late Holocene example demon-
strates. At the Giant Horse site, near Laura in North Queensland, Australia, a small
fragment of carbonized plant stem (0.41 mg), was removed from the front leg of a
very large polychrome painted animal, interpreted as a horse (Flood 1990:122:
Trezise 1993:97). The charcoal, believed to be an accidental inclusion during prepa-
ration or application of white paint, was considered to be highly suitable for dating,
but its age was unbelievably too old (435 + 65 years BP, AA9685; AD 1440 [AD 1331
to AD 1639, 2 sigma], University of Washington CALIB program, 1987), because
horses were not introduced by Europeans into northern Australia until after the Ed-
mund Kennedy and William Hann expeditions in 1848 and 1872, respectively.

Even though charcoal may look perfectly suitable for dating the implications for
precisely determining the age of rock art using only charcoal are considerable. If large
pieces of charcoal have ‘floated’ around a rock shelter or cave floor for thousands of
years and are still present today, then the age of carbon in the charcoal is probably un-
related to the time of the painting event. We have no way of determining the length of
time between the burning of wood and the use of that charcoal as a paint ingredient.
Should we therefore not question the antiquity of charcoal paintings throughout the
world (Table 1, Davidson, this volume)? How reliable are the direct radiocarbon
dates for Paleolithic charcoal paintings in Cougnac (Valladas ef a/. 1990), Cosquer
(Clottes et al. 1992), Chauvet (Clottes et al. 1995), Pech-Merle (Lorblanchet et al.
1995), and in Niaux, El Castillo and Altamira (Valladas et a/. 1992)? Several of these
results are linked to dated pieces of charcoal found on the floor of the caves today, and
this evidence is used to support the age of the paintings (Clottes, this volume). How-
ever, if we look objectively at these correlations we can appreciate that while loose
charcoal from the floor and scrapings from a hand stencil or drawing may have identi-
cal ages this only proves that the charcoal samples have the same age. There is no way
of knowing with absolute certainty from that charcoal when the painting was actually
done because the charcoal has been on the present floor and available for use from the
time of its creation until today. Taking this example a step further, if we assume that
the age of the charcoal on the floor of a cave matches the age of charcoal in a wall
painting and the AMS determinations represent the time of painting, then the ar-
chaeological deposit immediately beneath the charcoal on the floor should be margin-
ally older than the age of the charcoal. Such a situation should be easily tested by
dating the floor deposits, but this has yet to be done convincingly (although an unsuc-
cessful attempt was recently made at the Grotte d’ Arcy-sur-Cure; Girardet al. 1995).

ARCHAEOMETRY AND PALEOLITHIC IMAGE 31

Results from Organics

No Paleolithic age determinations have been obtained by using organic binders in
paints or fossilized microorganic remains in rock surface accretions, but some ancient
minimum-limiting ages have been produced from organic matter trapped at the base
of rock varnish in some Australian engravings (Dorn et al. 1992).

Plasma oxidation not only oxidizes paint organic binder, but it also removes car-
bon from all other similar organic substances on and in the rock. Although it does not
decompose oxalate or carbonate minerals the ozone generated in the reaction cham-
ber attacks bacteria, fungi and algae that inhabit pores and crevices in the substrate.
Dating graphite produced by oxidizing carbon from unpainted rock samples can pro-
duce ages older than pigment alone (Chaffee e¢ a/. 1993:71), implying that consider-
able levels of background carbon exist in mineral accretions and the rocks beneath
paintings. Additionally, filamentous saprophytic fungi can temporarily invade paints
containing organic binders, use the carbon for metabolism and leave crystalline ox-
alate and other chemical byproducts. Therefore, although dates can be obtained for
organic substances In paints detailed observations, and organic and microbiological
analyses are essential prior to extracting the carbon for dating, and multiple dates are
necessary to guard against spurious single measurements (“one date is no date”:
Clottes, this volume).

Fossil carbon from carbonate is not generally a problem in the dating of organics
encapsulated in silica skins because X-ray diffraction and infrared spectra analyses
have so far not revealed carbonate. A more substantial problem with some silica skins
is that they can exfoliate, exposing old silica, organics and fragments of trapped char-
coal which can be painted over. Sampling and dating charcoal in drawings immedi-
ately above these eroded surfaces without due consideration to the microstratigraphic
conformability of the layers will lead to spurious radiocarbon dates (McDonald e al.
1990). A radiocarbon date (1,490 + 50 years BP, CAMS-16755) for algae and dia-
toms trapped ina layer of silica skin immediately over a red hematite Bradshaw Fig-
ure in the Kimberley area of Western Australia strongly suggests that these paintings,
allegedly equivalent in age to the Paleolithic (Walsh 1994), were most likely painted
during mid-Holocene times.

Laminated Sequences of Oxalate

A conformable sequence of dates for carbon in the oxalate mineral whewellite has
been obtained fora thick, laminated gypsum and oxalate crust at the Laura South site
in north Queensland (Watchman 1993b). Dates for four oxalate layers are stra-
tigraphically conformable and also consistent with dates for two charcoal layers,
positive evidence of regular sedimentation. Lateral concordance and vertical congru-
ency between dates for two carbon bearing substances across 4 m of laminated crust
demonstrate that contamination by fossil carbonate carbon and isotopic exchange
have not occurred. A date of 24,600 + 220 years BP (NZA2570, Watchman 1993b)
for carbon in oxalate in a stratigraphically equivalent position to bright red hematite
paint provides substantive evidence of human painting activity on the rock face at that
time. Traces of paintings at about 16,000 and 7,000 years BP in the same crust indi-
cate continuing periodic use of rock paints over a long time span.

Since the presentation of this paper additional dating results have been obtained
for a finely laminated sequence of salts, dust and red and yellow paint layers at

BS)
to

WATCHMAN

Walkunder Arch Cave, Chillagoe, Australia. Ten stratigraphically conformable ra-
diocarbon determinations spanning almost 30,000 years have been measured in a
crust only 2.2 mm thick (Watchman & Campbell in press). The oldest evidence of
hematite painting in that crust occurs at 28,100 + 400 radiocarbon years BP
(OZA391), witha yellow goethite painting lying at 25,800 + 280 years BP (OZA392),
and two red paint layers at about 16,100 + 130 (OZA395) and 6,790 + 70 (OZA398)
years BP, respectively. Archaeological evidence so far obtained from the incom-
pletely excavated deposit beneath the painted walls and boulders suggests that people
periodically visited the site from at least 16,000 years ago (the base of the present ex-
cavation), until the present (Campbell & Mardaga-Campbell 1993).

While the form and style of the paintings represented by these traces of paint are
unknownat this time, to suggest that these paint remnants do not necessarily “make” a
painting ignores the fact that microexcavation laterally across the surface will reveal
the form of the paintings in the same way that widening a trench allows an archaeolo-
gist to examine successive sequences of debris in a cave or rock shelter excavation.
Prehistorians therefore should not dismiss these results out of hand because the impli-
cations are enormous. How many paintings are obscured beneath rock surface accre-
tions at other sites?

Implications and Conclusions

Paleolithic rock art can be scientifically dated despite the restrictions imposed on
the archaeometrists, but only after thorough observations of microstratigraphic and
microbiological relationships, and by completing organic and inorganic analyses of
paint, and its underlying and overlying mineral layers. The archaeometrist needs to
find out whether an organic binder is present in paint or if it has been replaced by deg-
radational products, because dating is simplified when a binder exists. When an or-
ganic binder is not present datable components must be found associated with mineral
layers under and over the painting, but this complicates the dating process and in-
creases the risk of contamination,

Contamination of small samples is a major radiocarbon dating problem, but to
date rock paintings and engravings the archaeometrist is forced to take only small
samples to avoid damaging the artifacts. Various sources of potential contamination
exist. but with care and attention to cleanliness most causes of contamination by
younger substances can be avoided. Carbonate minerals can be identified by X-ray
diffraction and infrared spectroscopy, and can be removed by acid, but the identity
and history of subvarnish organic matter is often uncertain, making interpretation of
AMS "C dates questionable. Humic and fulvic acids, identified by chromatographic
means, are eliminated by acid and alkali treatments prior to dating, but the effects of
photosynthetic algae, that can preferentially absorb heavy carbon isotopes from fossil
sources (carbonate dissolved in groundwater and not in equilibrium with the atmos-
phere of the time), produce a potentially greater threat to the accuracy of a radiocar-
bon measurement (Fowler et al. 1986:44 1; Long et al. 1992:557). Considerable stress
is often placed on the need for pretreating samples prior to radiocarbon dating (Werl-
hoferal. 1995:267), but it is the amount and the nature of the contaminating agent that
really matters (Watchman 1995). Knowing the chemistry of the carbon bearing sub-
stances that are to be dated, and whether mixtures exist are important because differ-
ent degrees of pretreatment will produce different results (correctly pointed out by
Clottes, this volume). Dating single samples (“One date is no date,” Clottes, this vol-

ARCHAEOMETRY AND PALEOLITHIC IMAGE

ive)
ivs)

ume), cannot produce reliable results, and even multiple dates from the same figure,
especially by scraping several parts of the same figure, may lead to ambiguous out-
comes (McDonald et al. 1990). “One style is no style” either as we have seen over the
years (Bednarik 1995; Clottes et al. 1990; Lorblanchet er al. 1990), so what should be
done?

To increase the reliability of dating Paleolithic rock art the archaeometrist should
ideally measure replicates from the same sample, repeat measurements on different
samples from the same figure, establish a conformable sequence across a painting or
engraving and measure background and maximum ages of adjacent unpainted and
unengraved rock surface accretions. I believe the only way to establish reliable ages
for Paleolithic rock art is through the systematic and scientific use of a combination of
dating methods carried out by separate research teams working independently on
finely laminated rock surface accretions (canvas and varnish), on stratified floor de-
posits, and on paint components. We should not be relying on stylistic estimates or on
measuring the age of single charcoal scrapings (as has happened until now). How-
ever, funding limitations, politics, conservation aspects and practical limitations may
not make these exemplary procedures practicable.

New methods for dating rock surfaces and rock art are being developed and re-
fined, and eventually it should be possible to date a figure or stylistic sequence using
two or more independent methods. There are no simple solutions in the dating of rock
art, and no single archaeometric method is superior to the others because each sam-
pling situation is different, and analytical and technical adjustments must be made ac-
cording to the circumstances. Prehistorians wishing to use archaeometric methods for
dating rock art need to be aware ofall the assumptions and other influences (including
their own biases) that are likely to affect the dating results and their interpretations.
We now know the limitations of different sampling, preparation, pretreatment and
analytical procedures used by archaeometrists and we understand more about Paleo-
lithic paint recipes, sources of contamination and ancient rock surface accretions, so
we are in a good position to date rock art absolutely. Prehistorians may now be at a
watershed in the way they view their existing stylistic chronologies and theories
about the evolution of Paleolithic rock art, and hopefully, reliable scientific dating
evidence will play a significant part during that reevaluation.

Acknowledgments

lam grateful to the Wattis Foundation for giving me the opportunity to express my
views on the dating of Paleolithic rock art. [also wish to thank Deborah Stratmann for
organizing my splendid visit to San Francisco.

Literature Cited

Arsenault, D. 1995. Le project Nisula: recherche pluridisciplinaire autour d'un site a picto-
grammes (DeEh-1) en Haute-Cote-Nord. Paléo-Québeéc 23:17-57.

Bahn, P.-G. 1993. “The ‘dead wood stage’ of prehistoric art studies: style 1s not enough,” Pages
51-59 in P.-G. Bahn & M. Lorblanchet, eds., Rock Art Studies: The Post-Stylistic Era or
Where Do We Go from Here? Oxbow Monograph 35, Australia.

. 1995. Paleolithic engravings endangered in Céa Valley, Portugal. La Pintura 21:1-3.

Batten, R. J.,R. Gillespie, J. A. J. Gowlett & R. E.M. Hedges. 1986. The AMS dating of sepa-
rate fractions in archaeology. Radiocarbon 28:698-701.

34 WATCHMAN

Bednarik, R.G. 1984. Die Bedeutung der palaolithischen Fingerlinientradition. Anthropolo-
gie 23:73-79.

. 1988. Comments on Nobbs and Dorn 1988. Rock Art Res.5:80-82.

_. 1989. Perspectives of Koongine Cave and scientistic archaeology. Australian Archae-

ology 29:9-16.

. 1995. Taking the style out of Panaramitee style. 4URA Nwsltr. 12:1-5.

Bierman, P.R. & A.R. Gillespie. 1994. Evidence suggesting that methods of rock varnish
cation-ratio dating are neither comparable nor consistently reliable. Quart. Res. 41:82-90.

Bourgeois, J. 1990, Boundaries: A stratigraphic and sedimentological perspective. Geol. Soc
Amer. Spec. Pap. 247:411-416

Campbell, J. B. & M. Mardaga-Campbell. 1993. From micro- to nano-stratigraphy. Pages 57-
63 in J. Steinbring and A. Watchman,eds., The Dating of Rock Art. Australian Rock Art
Research Association, Melbourne, Australia.

Chatfee, S.D.,M. Hyman & M. Rowe. 1993. AMS ve dating of rock paintings. Pages 67-73
in J. Steinbring and A. Watchman, eds., The Dating of Rock Art. Australian Rock Art Re-
search Association, Melbourne, Australia

Clarke, J. 1989. Further comments on Nobbs and Dorn 1988. Rock Art Res. 6:63-65.

Clegg, J. 1988. Comments on Nobbs and Dorn 1988. Rock Art Res. 5:139-141

Clottes, J. 1995. Paleolithic petroglyphs at Foz Céa, Portugal. /ntl. Nwsltr. Rock Art 10:2.

Clottes, J., M. Menu & P. Walter. 1990. New Light on the Niaux paintings. Rock Art Res
7:21-26.

Clottes, J., J. Courtin & H. Valladas. 1992. A well-dated Paleolithic cave: the Cosquer Cave at
Marseille. Rock Art Res. 9:122-129.

. 1993. Some very old hand stencils. /ntl. Nwsltr. on Rock Art 4:3-4.

Clottes, J., M. Lorblanchet & A. Beltran. 1995. Are the Foz Coa engravings actually Holo-
cene? Intl, Nwsltr. Rock Art 12:19-21.

Clottes, J. , J.-M. Chauvet, E. Brunel-Deschamps, C. Hillaire, J.-P. Daugas, M. Arnold, H.
Cachier, J. Evin, P. Fortin, C. Oberlin, N. Tisnerat & H. Valladas. 1995. Les peintures
paléolithiques de la Grotte Chauvet-Pont-d'Are, a Vallon-Pont-d'Are (Ardeche, France):
datations directes et indirectes par la méthode du radiocarbone. Comptes-rendus de I'A-
cadémie des Sciences, Paris, série 11a, tome 320, 11:1133-1140.

Combier, J. 1984. Grotte de la Téte-du-Lion. Pages 595-599 in L'Art des Cavernes. Im-
primerie nationale, Paris, France.

Conkey, M. & C. Hastorf, ed. 1990. The Uses of Stvle in Archaeology. Cambridge University
Press, Cambridge, U.K.

David, B. 1992. An AMS date for north Queensland rock art. Rock Art Res. 9:139-140

Dorn, R.I. 1983. Cation-ratio dating: anew rock varnish age-determination technique. Quarter
Res. 20:49-73.

—_. 1989. Cation-ratio dating of rock varnish: a geographic assessment. Prog. Phys. Geogr
13:559-596

Dorn, R.I. & D.H. Krinsley. 1991, Cation leaching sites in rock varnish. Geology 19:1077-
1080.

Dorn, R.I., P.B. Clarkson, M.F. Nobbs, L.L. Loendorf & D.S. Whitley. 1992. Radiocarbon
dating inclusions of organic matter in rock varnish, with examples from Drylands. Ann. As-
soc. Amer. Geogr. 82:136-151

Dragovich, D. 1988. A preliminary electron probe study of microchemical variations in desert
varnish in western New South Wales, Australia. Earth Surface Processes and Landforms
13:259-270.

Fleming, D. & M. Barnes. 1994. Are the Puquios of Nazca prehistoric? Soc. Amer. Archeol.,
Paper General Session S74, 59th Annual Meeting, Anaheim, California, April 1994.

ARCHAEOMETRY AND PALEOLITHIC IMAGE

oe)
nN

Flood, J. 1990. The Riches of Ancient Australia. University of Queensland Press, St. Lucia,
Australia.

Fowler, A.J., R. Gillespie & R.E.M. Hedges. 1986. Radiocarbon dating of sediments. Radio-
carbon 28:441-450.

Francis, J.E., L.L. Loendorf & R.I. Dorn. 1993. AMS radiocarbon and cation-ratio dating of
rock art in the Bighorn Basin of Wyoming and Montana. Amer. Antig. 58:71 1-737.

Girard, M., D. Baffier, H. Valladas & R. Hedges. 1995. He dates at the Grande Grotte at
Arcy-sur-Cure (Yonne, France). /ntl. Nwsltr. on Rock Art 12:1-2.

Huashan, Q.S., Q. Tsailuan, L. Minfei & Y. Juyu, ed. 1987. The Investigation and Research of
the Cliff and Mural Paintings of the Zuojiang River Valley in Guangxi. Guangxi National
Printing House, Nanning, China.

Krinsley D.H. & S. Anderson. 1989. Desert Varnish: a new look at chemical and textural
variations. Geological Society America Abstracts with Program 21:103.

Lanteigne, M.P. 1989. Further comments on Nobbs and Dorn 1988. Rock Art Res. 6:145-149.

. 1991. Cation-ratio dating of rock engravings: a critical appraisal. Antiquity 65:292-
295.

Long, A., O.K. Davis & J. De Lanois. 1992. Separation and Be dating of pure pollen from lake
sediments: nanofossil AMS dating. Radiocarbon 34:557-560.

Lorblanchet, M., M. Labeau, J.L. Vernet, P. Fitte, H. Valladas, H. Cachier & M. Arnold. 1990.
Palaeolithic pigments in the Quercy, France. Rock Art Res. 7:4-12

Lorblanchet, M., H. Cachier & H. Valladas. 1992. Direct date for one of the Pech-Merle spotted
horses. /ntl. Nwsitr. on Rock Art 12:2-3.

Loy, T.H. 1994. Direct dating of rock art at Laurie Creek (Northern Territory), Australia: a re-
ply to Nelson. Antiquity 68:147-148.

Loy, T.H., R. Jones, D.E. Nelson, B. Meehan, J. Vogel, J. Southon & R. Cosgrove. 1990. Ac-
celerator radiocarbon dating of human blood proteins in pigments from Late Pleistocene art
sites in Australia. Antiquity 64:1 10-116.

Martineau, L. 1990. The Rocks Begin to Speak. KC Publications, Las Vegas.

McDonald, J., K. Officer, T. Jull, D. Donahue, J. Head & B. Ford. 1990. Investigating ee
AMS: dating prehistoric rock art in the Sydney sandstone basin, Australia. Rock Art Res
7:83-92.

Nelson, D.E. 1993. Second thoughts on a rock-art date. Antiquity 67:893-895.

Nelson, D.E., C. Chippindale, G. Chaloupka & J. Southon. 1992. AMS dating: Possibilities,
and some results. 2nd AURA Congress, Cairns, Abstract, 21.

Nobbs, M. & R.I. Dorn. 1988. Age determinations for rock varnish formation within petro-
glyphs: cation-ratio dating of 24 motifs from the Olary region, South Australia. Rock Art
Res. 5:1 08-146.

Patterson-Rudolph, C. 1993. Petroglyphs and Pueblo Myths of the Rio Grande. Avanyu Pub-
lishing Inc, New Mexico.

Pepe, C., J. Clottes, M. Menu & P. Walter. 1991. Le liant des peintures paléolithiques are-
geoises. Comptes-rendus de l'Académie des Sciences de Paris (Préhistoire) 312:929-934,

Raymond, J.R.,R.,S.L. Reneau & C.D. Harrington. 1991. Elemental relationships in rock var-
nish determined by scanning electron microscopy and energy dispersive X-ray elemental
line profiling. Scanning Microscopy 5:37-46.

Reneau, S.L. 1993. Manganese accumulation in rock varnish on a desert piedmont, Mojave
Desert, California, and application to evaluating varnish development. Quat. Res. 40:309-
37

Reneau, S.L. & C.D. Harrington. 1988. Comments on Nobbs and Dorn 1988. Rock Art Res
5:141-142.

36 WATCHMAN

Reneau, S.L. & R. Raymond Jr. 1991. Cation-ratio dating of rock varnish: why does it work?
Geology 19:937-940.

Russ, J..M. Hyman, H.J. Shafer & M.W. Rowe. 1990. Radiocarbon dating of prehistoric rock
paintings by selective oxidation of organic carbon. Nature 348:710-711.

Schwarez, H.P. 1982. Applications of U-series dating to Archaeometry. Pages 303-325 in M.
Ivanovich and R.S. Harmon, eds., Uranium Series Disequilibrium Applications to Environ-
mental Problems. Oxford University Press, Oxford, U.K.

Trezise, P. 1993. Dream Road. Allen & Unwin, St. Leonards, U.K.

Valladas, H., H. Cachier & M. Arnold. 1990. AMS C14 dates for the prehistoric Cougnac cave
paintings and related bone remains. Rock Art Res. 7:18-19.

Valladas, H., H. Cachier, P. Maurice, F. Bernaldo de Quiros, J. Clottes, V. Cabrera Valdes, P.
Uzquiano & M. Arnold. 1992. Direct radiocarbon dates for prehistoric paintings at the
Altamira, El Castillo and Niaux caves. Nature 357:68-70.

Wainwright, ILN.M. 1985. Rock art conservation research in Canada. Bulletino des Centros
Camuno di Studi Preistorici 22:\5-46.

Walsh, G. L. 1994. Bradshaws.: Ancient Rock Paintings of North-west Australia. Edition
Limitée, Carouge-Geneva, Switzerland.

Watchman, A. 1987. Preliminary determinations of the age and composition of mineral salts
on rock art surfaces in the Kakadu National Park. Pages 36-42 in W.R. Ambrose & J.M.J.
Mummery, eds., Archaeometry, Further Australasian Studies. Department of Prehistory,
Australian National University, Canberra, Australia.

. 1989. Further comments on Nobbs and Dorn 1988. Rock Art Res. 6:65-66.

1990, What are silica skins and how are they important in rock art conservation? Aus-
tral. Aborig. Stud. \:21-29.

. 1992a,. Doubtful dates for Karoita engravings. Austral, Aborig. Stud. 1:51-55.
.1992b. Comment on ‘Cation leaching sites in rock varnish.” Geology 20:1050.

_. 1992¢, Composition, formation and age of some Australian silica skins. Austral Abo-
rig. Stud. 1:61-66.

__. 1993a. Perspectives and potentials for absolute dating prehistoric rock engravings. An-
tiquity 67:58-65.

. 1993b. Evidence of a 25,000 year old pictograph in northern Australia. Geoarchae-
ology. 8:465-473.
—. 1993c. More information about South Australian cation-ratio dates. Rock Art Res
10:40.

—_. 1995. Recent petroglyphs, Foz Coa, Portugal. Rock Art Res. 12: 104-108

Watchman, A. & J.B. Campbell. in press. Micro-stratigraphic analyses of laminated oxalate
films in northern Australia. in The Oxalate Films in the Conservation of Works of Art. Pre-
print of the International Symposium. Centro CNR Gino Bozza, Milan, Italy.

Watchman, A. & N. Cole. 1993. Accelerator radiocarbon dating of plant fibre binders in rock

Watchman, A. & R. Lessard. 1993. Focused laser extraction of carbonaceous substances for
AMS radiocarbon dating (FLECS-AMS). Pages 74-77 in J. Steinbring & A. Watch-
man,eds., The Dating of Rock Art. Australian Rock Art Research Association, Melbourne.

Watchman, A., J. Sirois & N. Cole. 1993. Mineralogical examination of Aboriginal rock
painting pigments near Laura, north Queensland. Pages 141-150 in B.L. Fankhauser & R.
Bird, eds., Archaeometry: Current Australasian Research. Department of Prehistory, Aus-
tralian National University, Canberra, Australia.

Whitley, DS. & RI. Dorn. 1994. New perspectives on the Clovis versus pre-Clovis
controversy. Amer. Antig. 58:626-647.

New Laboratory Techniques and
Their Impact on Paleolithic Cave
Art

Jean Clottes

Conservateur Général du Patrimoine
UMR 9933

11, rue du Fourcat

09000 Foix, France

A Brief History of Paint Analysis

Early in the twentieth century, as soon as Paleolithic cave art had been accepted as
genuine by academics, archaeologists turned to laboratory techniques to determine
the paint constituents used. Henri Moissan, a Nobel Prize winning chemist, worked
on pigments from two caves in the Dordogne, Font-de-Gaume and La Mouthe (Mois-
san 1902, 1903). His analysis determined that the paints used consisted of hematite or
manganese oxide mixed with calcite and small grains of quartz. He could find no trace
of any organic matter that could have been used as a binder. At about the same time
(1902), another researcher, Courty, obtained similar results for red and black pig-
ments from Laugerie-Haute, also in the Dordogne (Clottes e¢ a/.1990a: 171).

From then on the problem of what had been used for paint was largely considered
solved. For decades the Abbe Breuil characterized the paintings by visual observation
and when he did refer to the sources of the paint, he mentioned only discoveries of
paint materials in contemporary archaeological levels (Breuil 1952:45). It was not
until the seventies that a renewal of interest in the subject led to a number of physico-
chemical analyses of samples lifted from Paleolithic wall paintings (Vandiver 1983),
as well as to analysis of fragments of coloring material such as those found inside
shells full of paint in Altamira (Cabrera-Garrido 1978) or in the form of lumps of
hematite bearing traces of use, discovered in Lascaux (Couraud & Laming-
Emperaire 1979). Claude Couraud also experimented with various types of paint and
binders in order to reproduce Paleolithic paintings (Couraud 1982, 1988).

A flurry of analyses followed in the eighties, both for portable art and for wall art
(see Clottes et a/. 1990a: 171-172). In particular, four samples from the Salon Noir in
Niaux (Ariége) were studied and determined to contain burnt wood, probably juniper.
The conclusion drawn was that some animals represented in the Salon Noir had been

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

38 CLOUIES

done with charcoal rather than painted with a brush as Breuil and everyone since had
assumed (Brunet ef a/. 1982). Michel Lorblanchet began a series of analyses on paint-
ings in the Lot, at Marcenac and especially at Cougnac (Lorblanchet et a/. 1988). The
paint on some objects from the Upper Magdalenian levels in the cave of La Vache
(Alliat), next to Niaux, was found to have been made with a mixture of pigments and
minerals, the latter used as extenders (Buisson ef a/. 1989). This discovery inspired
the paint analysis at Niaux which will be discussed in greater detail.

Present Day Techniques

Several advances have been made in methods of paint analysis. Most importantly,
it has now become possible to lift samples so small as to minimize radically any sur-
face damage to the paintings. Photographs taken before and after show sampling
hardly visible to the naked eye. This means that the paintings to be studied can be cho-
sen as a function of testable hypotheses derived from a coherent research plan,
whereas earlier the selection of paintings had depended solely on the presence of
paint thick enough to preclude damage to their appearance. Now the restrictive con-
cerns of conservation can readily accommodate the rigorous demands of sound re-
search methods: what is desirable has become feasible as well (Clottes
1993:223-224).

Another significant development has been that paint samples need no longer be
destroyed in preparation for analysis. Instead they can be studied in their original state
and then preserved. Since the verification of results no longer requires the removal of
more paint, the analyses of samples carried out at one laboratory can now easily be
sent for replication to another. Once again, the requirements of conservation and of
good research design can both be met (op. cit.:224).

Even though samples are minute, it is of course advisable to confine what ts lifted
to that strictly necessary for providing answers to well-defined problems. Optical in-
spection, nowadays augmented by macrophotography and direct examination by
means of a specially adapted binocular microscope, helps determine how paint was
applied to the cave walls and detect superimpositions. Above all, such inspection en-
ables researchers to check for homogeneity in the composition of each figure and to
determine whether or not further laboratory analyses are needed. Direct examination
can thus eliminate the systematic taking of multiple samples or it can show that new
samples are indispensable. Recently, direct optical analysis of the paint constituents
without any sampling has been carried out in the Quercy caves (Lorblanchet
1994a:167).

The analyses done at the Laboratoire de Recherche des Musées de France in Paris
are of particular interest in that they not only provide information about chemical
composition but also about quantities and proportions of the paint constituents, as
well as about texture. In Niaux, for example, some paint components had been finely
ground while others had been left more coarse. These analyses can even detect trace
elements, which is particularly useful in establishing where the original paint matert-
als were collected.

Among paint constituents, one of the most essential 1s a binder to ensure cohesion
and fluidity. After much speculation anda few experiments (Couraud 1988), analyses
have been carried out on samples taken in four caves only, all in the Ariege (Fontanet,
Les Trois-Fréres, Enléne, Le Réseau Clastres [Figure 1]). These analyses are so few
because determining the exact nature of the binder is a highly delicate process involv-

NEW ANALYTICAL TECHNIQUES 39

Ficure |. Black horse from Le Réseau Clastres (Niaux, Ariege). No evidence of a binder was found for the
animals in Le Réseau Clastres, probably because they were quickly painted with whatever materials were
at hand. Photograph by R. Simonnet

ing long and costly laboratory work, and also because samples can very easily be con-
taminated, for example, by someone touching a painting with bare hands. For this
reason sampling can only be done when there is reasonable certainty this has not been
the case (Pepe ef al. 1991).

The purpose of this paper is not to describe the laboratory procedures and their
technical details, but rather to show how they have been applied to Paleolithic cave art
and how their results have been interpreted. Concerning the techniques employed, see
Menu & Walter 1991; Clottes et a/. 1990a, b; Clottes 1993; Lorblanchet ef al. 1990a,
b; Menu et al. 1993; Pepe eral. 1991.

Another important advance in laboratory techniques has been the use of an accel-
erator (Accelerator Mass Spectrometry, or AMS for short) for direct dating of prehis-
toric paintings. After initial attempts in South A frica (Van der Merwe et al. 1987), a
series of datings soon followed in Australia (Loy et a/. 1990), North America (Russ ef
al. 1990), and also in Europe (Lorblanchet 1990a, b; Valladas et a/. 1992). In all, 26
dates have been published for 19 different figures in six caves in France and in Spain
(Clottes 1994b; Clottes in press; Lorblanchet 1994a, Clottes et al. 1995).

On other continents, where the location of the art, the nature of the rock and the cli-
mate differs widely from what is known in Europe, other dating methods have been
developed in an attempt to obtain direct dates (Franklin 1993). A number of sub-
stances, among them blood, have been identified as binders in Australia (Bednarik
1994:161). Most can be dated by AMS. Dating of the varnish on petroglyphs has been
done in the Americas and in Australia both by the cation-ratio method (Dorn 1992)
and by radiocarbon dating when small pieces of organic matter are trapped beneath
the varnish and can be extracted by laser and AMS dated (Nobbs & Dorn 1988;
Watchman & Lessard 1992). This provides a minimum date for the varnished petro-
glyphs. In Europe no petroglyphs have yet been dated by laboratory methods, except
for some very recent attempts, as yet unpublished, at Foz Coa in Portugal. Sample
analyses may also bring unexpected information, such as pollen of forty plant species
from a Chinese rock painting at Cangyuan, in Yunnan province, or the discovery by

40 CLOTTES

Alan Watchman of up to ten layers of paint at sites that appeared totally undecorated
to the naked eye (Bednarik 1994:161-162).

One fascinating possibility as yet unexplored in Europe is obtaining ancient DNA
from paintings or from finger tracings. This might give information as to which spe-
cies of animal had provided grease or blood as binders. This technique might even de-
termine the individual person(s) who left finger tracings or hand stencils: Was it the
same person for several figures, or persons belonging to the same family? This type of
approach was considered for the Cosquer Cave (Marseille) in the fall of 1994, but was
given up because of the overly large samples the physicists required.

Paint Analysis: Its Uses and Limitations

The latest research in wall art paint analysis for European caves has been carried
out in several laboratories: by Michel Menu and Philippe Walter at the Laboratoire de
Recherche des Musees de France, by Pamela Vandiver at the Smithsonian Institution,
and by Michel Labeau at the Ecole Nationale Supérieure de Physique de Grenoble.

Exposing Forgeries

Itis widely believed that “one of the most crucial advantages of the current revolu-
tion in pigment analysis and in dating techniques is the growing ability to expose or
authenticate fakes” (Bahn 1993:58), and Paul Bahn, for example, wishes for analyses
of some of the Rouffignac paintings about which he still entertains doubts. What he
Writes 1s true for direct analysis, but his confidence ts less well founded with respect to
pigment analysis.

Physicochemical analyses have recently been used in Spain in an attempt to deter-
mine whether or not the Zubialde cave paintings, discovered in 1990 in the province
of Alava, were genuinely Paleolithic. As soonas the cave of Zubialde was discovered
anda few photographs published in the press, two English specialists (Peter Ucko, Jill
Cook) said that the paintings were fakes as they looked strange in several ways be-
cause of proportions and style, the presence of unlikely animals and signs, and the di-
versity of the fauna represented. The Spanish prehistorians in charge (Ignacio
Barandiaran, Jesus Altuna and Jose Maria Apellaniz) wrote in their detailed prelimi-
nary report (6 March 1991) that they had not found anything questionable about the
works of art but they advocated further study and analysis. On 10 June 1991, I was
fortunate enough to be allowed inside the cave for a long close look at all the paint-
ings. The figures were spectacular and impressively well done, but the complete ab-
sence of weathering on any of them made me entertain serious doubt as to their
authenticity. The next day, I wrote to Barandiaran, “if, as is alas possible, this 1s a
fake, I doubt that the pigment analyses will bring any certainty about it, as the quality
of the art is such that a would-be faker would certainly have used natural pigment.”
(11 June 1991). A similar opinion had been expressed by Jill Cook in an interview
with The European (22-24 March 1991:3).

Michel Menu and Philippe Walter went to the cave five times to lift eighty samples
which they then studied with their usual techniques. They found that manganese ox-
ide had been used for the black figures and iron oxide for the red. The similarity of the
results to what they had seen in Pyrenees caves such as Niaux convinced them that the
Zubialde paintings were genuine. This sustained the hopes of our Spanish colleagues
for months, until Manuel Hoyos of the Instituto de Geologia in Madrid found out that

NEW ANALYTICAL TECHNIQUES 4]

some strange remnants removed from the walls were in fact synthetic fibers identical
to those in moder plastic sponges and that they held traces of pigment. A sponge had
been used to spread some of the paint in an attempt to weather it. In addition, various
fragments of animal hair had been found near or even in the paint. Hoyos’s discovery
clinched the matter. The Zubialde paintings were fakes (Altuna et a/. 1993:80).

This recent experience once again highlights the fact that the results of analyses do
not necessarily speak for themselves, except in the simplest cases, such as the discov-
ery of the plastic sponge fragments. Even the presence of animal hairs and spider legs
was not enough to betray the forgery: it only posed the problem of how long tiny ani-
mal remains could survive on the cave walls. As for the pigment analyses, they re-
vealed nothing more than that natural pigments had been used: No other inference
should have been drawn.

In the cave of Niaux, the pigment ina fissure colored bright red both inside and out
(Figure 2) was analyzed and found to consist of pure hematite with no quartz grains or
extender. This set it apart from the rest of the Niaux paint analyzed, and so we doubted
its authenticity, all the more so as it had never been mentioned by those who first stud-
ied the cave and also because two initials, ‘TD,’ until then considered a later addition,
could be seen to the right of the fissure (Clottes er a/. 1990b:22). Our conclusion ofa
recent forgery, then, depends as much on external elements as on the analysis itself.

Information on the Techniques Used by Paleolithic Artists

Several attempts have been made to determine what kind of binder was used for
the paint. The analyses dealt with lipids such as triglycerides, the main constituent of
oil or grease, and with sterols, the molecular structure of which makes it possible to
say whether they are of animal or vegetable origin. The technique used by Claude
Pepe was gas chromatography, verified by mass spectrometry and fragmentometry, a
particularly sensitive method (Pepe ef a/. 1991:933).

In the Réseau Clastres (Figure 1) the analysis was negative, probably because the
binder used was water. Water could be readily obtained on the spot, and its use would
support the idea that the Magdalenians made only quick expeditions into the exten-
sive Réseau Clastres, as they had to the deeper galleries of Niaux. There, far under-
ground, they did not take as much time and care doing the paintings as they had in the
Salon Noir of Niaux, a place nearer the entrance and easier of access where they re-
turned often and stayed for longer periods.

At Fontanet, Enléne and Les Trois-Fréres, Pepe’s analyses revealed the presence
of organic fatty elements in the paint used. In the last two caves, which are very close
to each other and which were used by the same people (Begouen & Clottes 1991), the
fatty elements are identical. They were derived from plants, whereas at Fontanet their
origin was animal. This means that although the source of the binders might vary,
Magdalenian artists used genuine “oil paint” to realize their work in the caves of the
Ariége (Pepe et a/. 1991; Clottes 1993:226, 229). The identity of the binders used in
Enléne and Les Trois-Fréres reinforces the idea of links between the two caves.

The possibility of complex recipes for the paint was first contemplated for Las-
caux, after minerals — generally quartz — had been systematically found in the pig-
ments used (Ballet et a/. 1979). Quartz would thus have been used as an extender.
From her own analyses, Pamela Vandiver (1983) also concluded that mineral extend-
ers had been added to ensure better control over the color and fluidity of the paint. Af-
ter extenders had been found in the paint on some objects of the late Magdalenian in
the cave of La Vache, it was tempting to check if the artists of Niaux had used similar

42 CLOTTES

FiGuRE 2. Niaux (Ariege). In one of the deeper galleries, a fissure has been colored in red. Paint analysis re -
vealed that this was done with pure hematite and not with a complex paint, contrary to all the other figures
in the same cave. This is probably a modern fake. Photograph by J. Clottes

or different recipes, since Niaux is right across the valley from La Vache, a bare half-
hour’s walk.

The first analyses were a big surprise. Not only had the La Vache recipe been
used, but two others had as well. At La Vache and for most of the figures in the Salon
Noir as well as for the Fontanet paintings, as we learned later, an extender consisting
of potassium feldspar had been added to manganese oxide or to charcoal for the
blacks and to hematite for the reds. In Niaux, in addition, a number of figures had been
painted with two other extenders: One consisted of talcum and the other, more widely
used, of a large amount of biotite with potassium feldspar. As at Lascaux, a highly
variable number of quartz grains were also found in the paint. We think that the quartz
found in the samples is a consequence of the grinding and that other minerals consti-
tuted the actual extenders (Clottes et al. 1990b:22). This all shows how sophisticated
the techniques of the Magdalenian painters were.

Niaux

When we chose the figures for paint samples, we had several questions in mind.
Was the same paint used for both animals and signs? Among the geometric signs,
claviforms are typical of Magdalenian art in the Pyrenees and are supposed to belong
to one short period, the Middle Magdalenian; in Niaux, were they all done with the
same paint? Are the various panels in the Salon Noir homogeneous? Could any differ-
ences be detected between the Salon Noir and the most remote galleries, or between
horses and bison and other animals?

NEW ANALYTICAL TECHNIQUES 43

We discovered that the composition of the paint did not vary with particular sub-
jects or locations. Both animals and signs, including claviforms, had been painted at
times with recipe B (biotite) and at others with recipe F (potassium feldspar) or with
recipe T (talcum). The two main recipes, B and F, were found in the farthest galleries
as well as in the Salon Noir, where they sometimes appear on the same panel. This
suggested that the choice of recipes was not determined by either theme or location,
and so might have a chronological value, with one recipe having replaced the other
over time. This could be checked. We already knew from the ten samples analyzed at
La Vache that recipe B had been used during the Upper Magdalenian. Eight more
analyses were done from samples well-dated to the period immediately before, the
Middle Magdalenian, in the caves of Enléne, Les Trois-Fréres and Le Mas d’ Azil, all
in the Ariége. They were all found to include the F recipe, which seems then to have
preceded the B recipe (Clottes et a/.1990a, b).

In Niaux, however, three direct radiocarbon datings from painted figures provided
ambiguous results. Those figures had all been painted with recipe B. The paint on one
bison was dated to 12,890 BP + 160 (Gif A 91.319) and an isolated black line on the
same panel to 13,060 BP + 200 (Gif A 92.499). These dates tally perfectly with the
known dates at La Vache, while the paint on another bison gave a date of 13,850 BP +
150 (Gif A 92.501) (Clottes et a/. 1992). The paint from that bison consisted mainly of
charcoal with only faint traces of biotite. Does this mean that the B recipe originated
during the Middle Magdalenian, far earlier than was thought, (a possibility which had
already been addressed [Clottes 1993:232] ), or that the two main recipes had been
used simultaneously, or that the gap between the two radiocarbon dates might not be
as wide as it first seemed (see below about direct AMS dating)? Unfortunately, no
more direct AMS dating could be done at Niaux for conservation reasons, so that
there is no date for any figure painted with the F recipe.

The Niaux paint analyses provided other unexpected information. For example,
all the animals and signs on an entire panel in the Salon Noir (Panel 6) (Figure 3), the
one with the later radiocarbon dates, were painted with the B recipe, probably during
the same period. However, trace elements were found to vary and two types of min-
eral associations were detected: rutile with aluminum silicates, or albite alone. It
would therefore appear that while the same recipe was used, the source of basic raw
materials changed during the production of this panel; either it was not painted in a
single session or it was the work of different artists using the same recipe but not the
same mix, or both (Clottes ef a/. 1990b:22-23).

In the Salon Noir, sample analyses showed that paint had been applied over a
sketch done first with pure charcoal. This was later confirmed by optical examination
and macrophotographs (Menu et a/. 1993). Not all the Niaux animals were sketched
before being painted, however. In the Salon Noir, sketches had been done only when
the B recipe had been used, never when the paintings were done with the F recipe. In
the remote galleries and in the Réseau Clastres, no sketching was done, even for the
B-recipe animals. A preliminary sketch suggests premeditation, advance planning
and a great deal of care which tallies with what we know of the Salon Noir, where the
animals are large, numerous and very detailed. It was a sort of sanctuary where people
went repeatedly, especially during the Upper Magdalenian, and where they took time
to create the paintings, whereas the remote galleries may have been explored very few
times by artists who did not linger and who drew animals here and there ina much less
sophisticated way (Clottes et al. 1990b:23-24; Clottes 1974, 1995).

44 CLOTTES
Cougnac

At Cougnac the first researchers to work in the cave noticed that several red fig-
ures had been repainted in black and they expressed the view that this happened
probably not long after the animals had first been painted in red, even if several artis-
tic phases could be distinguished in the composition of the art (Méroc & Mazet
1956:33-34). Recent studies confirmed a series of applications on the walls (Lor-
blanchet 1994a:177-178), and paint analyses show that the early group of red figures
had all been painted with the same pigment which included the same minerals. That
pigment had been taken from a heap of ocher still visible on the floor of the chamber.
Small brown figures were added later. Their color varied and so did their composi-
tion: some had been done with hematite with | or 2% manganese while in others man-
ganese could reach up to 30%, which made a much darker color. Black animals were
then added and a few red ones were repainted (op. cit.:178).

Michel Lorblanchet rejected the idea that the story of the Cougnac animal paint-
ings could be reduced to three distinct “phases” because of the heterogeneity of the
brown figures and especially because of a series of widely spaced radiocarbon dat-
ings. According to him, the Cougnac art includes real compositions with animals

Figure 3. The Salon Noir, in Niaux (Ariége). All the animals of Panel 6 were painted with the same recipe
(B). The small vertical line just above the horns of the bison in the middle was dated by AMS to 13,060 BP +
200. Photograph by N. Aujoulat

NEW ANALYTICAL TECHNIQUES 45

done all at one time, but also partial retouching of some of them, the addition of new
figures and of numerous traces testifying to a ritual use of the place (op. cit.: 179). The
most fascinating example of such traces is the way Paleolithic people at unspecified
periods dipped the tips of their fingers in the pile of ocher on the ground to leave their
marks on the walls. As to the model proposed by Lorblanchet ofa cave frequented and
reused for millennia, it is rather close to the Abbé Breuil’s views which had all but
been discarded since the structuralist work of André Leroi-Gourhan in the sixties.
Lorblanchet’s model relies heavily on the interpretation of the AMS direct dates ob-
tained in Cougnac.

Direct Dating by AMS

The main advantage of the AMS technique is that it has now become possible to
obtain dates from very small amounts of organic material, whether lifted from the
paintings themselves or from objects closely associated with the paintings. Being
able to date a figure directly is certainly a huge advance over earlier methods and
when the first results were published, this was hailed as an important breakthrough in
the field of rock art studies. It was felt that the dates were “well-founded data” (Lor-
blanchet 1992:XVII) which allowed for “more objective dating” (Lorblanchet
1990a:10) that would “in due course revolutionize rock art studies” (Bednarik
1992:148). Dating figures directly would enable specialists to construct chronologi-
cal frameworks, first on a cave-by-cave, region-by-region basis (Franklin 1993:8).
The exact duration of styles, techniques, particular conventions and motifs would at
last be determined (Lorblanchet 1993:69). Finally, when enough absolute dates be-
came available, they would provide an overall framework into which undated paint-
ings and engravings could be fitted (Bednarik 1992:148; see also Valladas et al.
1992:69). The routine application of this method to rock paintings would then make
its study more scientific (Bednarik 1994:163).

It could be argued that the new dating methods will dramatically change our out-
look in a short time, as happened in the fifties and sixties when radiocarbon dating
was systematically applied to archaeological sites of all periods, providing surprises,
upheavals in current hypotheses, and a renewal of theories and methods (Clottes
1993:21). This is debatable (see below) even if the score or so of direct AMS dates so
far published from five Spanish and French caves have already provided a number of
interesting findings.

Direct Dating in the Caves of Niaux, Altamira, Castillo,
Cosquer and Cougnac

At Niaux, as we have seen, two dates corroborated the contemporaneity of some
of the Salon Noir paintings with the living site of La Vache, while another date — ear-
lier by one millennium — seemed to indicate long, repeated use of the sanctuary. Un-
fortunately, some of the problems raised by the discrepancy between the dates could
not be solved due to the practical impossibility of lifting and dating more samples.

In Spain, three dates ranging from 13,570 BP to 14,330 BP + 190 were obtained at
Altamira, and two others around 13,000 BP (+ 170 and + 200) at El Castillo, all for
very similar bison. This confirmed that the bichrome animals at Altamira were Mid-
dle Magdalenian, as had long been assumed for stylistic reasons (Leroi-Gourhan
1965), and also that the technique might have lasted until the Upper Magdalenian

46 CLOTTES

(Valladas er al. 1990). Lorblanchet challenged the interpretation of the Altamira
panel as an Early Style 1V composition because, according to him, the dates were in-
terpreted through the bias of Leroi-Gourhan’s chronological and stylistic system,
which led our Spanish colleagues “to neglect a gap of more than seven centuries be-
tween the dated bison” (Lorblanchet 1994a:173). In fact, it is obvious that with a
two-sigma interval of + 380 (a 95 % chance of being within the right time span), the
three dates are either very close to each other or even overlap for long periods, which
they do in any case with a three-sigma interval (a 99 % chance). This overconfidence
in the precision of the radiocarbon dating method may lead one to rather rash conclu-
sions, as we shall see with Cougnac.

At Cosquer, seven dates have so far been obtained from three animals and one
hand stencil with incomplete fingers. The samples from two of the animals and from
the stencil were divided into two parts which were then dated separately. The main re-
sults (Clottes & Courtin 1994) were:

1) The paint from the hand stencil was both times dated to 27,110 (+ 350 and +
390). This was first corroborated by two other dates for charcoal from torches (27,870
BP + 430 and 26,360 BP + 400). Then, at Gargas in the Pyrenees, a fragment of bone
found with many others in a crack next to hand stencils with incomplete fingers 1den-
tical to the ones in Cosquer was dated to 26,860 BP + 460 (Clottes er a/. 1992). Those
converging dates set the beginning of hand stenciling at an earlier date than expected.

2) We could establish that there had been at least two different phases in the art of
the Cosquer cave and that the animals were later by several millennia than the stencils
and finger tracings. We had already established these two phases from evidence of su-
perimpositions. The AMS dates from the animals plus another one from charcoal on
the ground brought a welcome corroboration and set Phase Two between 18,500 and
19,200.

3) A well-known stylistic convention, supposedly characteristic of the Middle
Magdalenian, was moved back about 6,000 years. That convention consisted in de-
picting the difference in color between the belly and the upper part of the body of
horses in the form of a flattened M (Figure 4).

4) Our attention was also drawn to some of the limits of the AMS method. Samples
must be purified before analysis. As they are minute, a balance must be struck be-
tween an excess of purification which might destroy most of the sample and preclude
analysis, and insufficient purification. The first dating of a painted bison gave 18,010
+190. As the dates for the other animals in the cave were older by halfa millennium or
more, it was then decided to do another dating on the second halfof the sample, which
had been preserved, and this time to push the purification further. The new date was
18,530+180, which tallied more closely with the others. This means that if two condi-
tions had not been met, that of other dates available for the same art context and of a
relatively large amount of charcoal, a date too recent by more than 500 years would
have been accepted since there would have been no reason to question It.

At Cougnac, two main results were achieved by AMS dating. First, two dots were
dated to around 14,000 BP, anda bone on the ground to about 15,000 BP (Lorblanchet
1994a, b). This was a surprise, since the art had always been considered stylistically
homogeneous and attributable to Leroi-Gourhan’s Style III, about 17 to 18,000 BP.
The three “late” dates, however, need not contradict the postulated stylistic homoge-
neity, as some people might well have gone into the cave several millennia after it was
painted and in addition to leaving a few bones, might have made a few drawings non-
descript enough to fit in with the older ones.

NEW ANALYTICAL TECHNIQUES 47

Ficure 4. In the Cosquer Cave (Marseille), the horse on the lower left of this panel was dated to 18,840 BP +

240 and 18,820 BP + 310. On the one at top right, the difference in color between the belly and the upper
part of the body has been depicted by leaving the belly blank in a sort of flattened M, a convention so far
supposed to be much later. Photograph by A. Chéné

Then, still at Cougnac, four dates were obtained from two Megaloceros, a male
and a female close to one another (Figure 5). This time, they were much earlier than
expected: 19,500 BP + 270 and 25,120 BP + 390 for the female, while the dates for the
male were somewhat more consistent with one another: 23,610 BP + 350 and 22,750
BP + 390 (Lorblanchet 1994a, b). At Cougnac, contrary to what had been done at
Cosquer, it was not the same sample that was split and dated separately. Because of
the lack of pigment, several extremely minute samples were in each case scraped and
mixed,

Those results could be interpreted as evidence that part of the art of Cougnac was
much earlier than had been thought, which was no mean achievement. However, Lor-
blanchet went much further. He argued that the two Megaloceros were “no doubt not
contemporary” (Lorblanchet 1994c:7). Besides the discrepancy in the dates, he based
his opinion on the female having been done with charcoal while the male had been
“drawn with burnt bone” (1994c:7), an unwarranted assumption given that the spe-
cialist who analyzed the pigments mentioned ‘tan undetermined kind of pigment
(male Megaloceros)” since neither the hypothesis of charcoal nor that of burnt bone
could be confirmed (Labeau 1993:72, 73). In order to explain the wide difference in
dates for the female Megaloceros, Lorblanchet assumed that this was most probably
due to repainting and retouching rather than contamination, and that the same was
likely to be true for the male (1994c: 7; 1994a: 171). If he was right this would be of
the utmost importance as it would reveal long-lasting rites of rejuvenation of the
paintings, similar to those known all over Australia.

48 CLOTTES

Unfortunately, alternative hypotheses can be proposed. The minute size of the
samples as well as the fact that they did not consist of one solid lump of charcoal but
were scraped from different locations makes the possibility of contamination likely
for the female despite all the precautions taken. As for the male, the two dates do re-
main compatible with one another within a two-sigma range, with an overlap of more
than 600 years. Therefore there is no need to postulate that the two black Megaloceros
Were painted and repainted at widely spaced intervals.

In other words, at Cougnac there is definite evidence: 1) that a few animal figures
are more than 22,000 years old and were probably painted during the Late Gravettian
or the Early Solutrean; 2) that between 14,000 and 15,000 years BP the cave was fre-
quented once again by people who left bones, dots and perhaps other traces. As to the
repainting of certain animals, this is visible from close examination of the figures and
it certainly did happen (Meéroc & Mazet 1956; Lorblanchet 1994a:178), but present-
day methods are not fine enough to pinpoint the exact moment(s) when this occurred.
For all we know, it might have been over a more or less extended period of time and

23610 + 350 25120 + 390
22750 + 390 + 270

1M

_

FIGURES. Male (in front) and female black Megaloceros at Cougnac (Payrignac, Lot), with arrows showing
where the samples for AMS dating were lifted. From M. Lorblanchet (1994a.)

NEW ANALYTICAL TECHNIQUES 49

even at intervals of hundreds or thousands of years’ time as Lorblanchet supposed,
but there is an equal possibility that immediately after the earlier outlines had been
sketched, people might have retouched some figures in the course of ritual use which
Lorblanchet himself thinks the cave was put to (Lorblanchet 1994a:179), or for any
other reason.

The Limits of AMS Direct Dating

From the few examples quoted above, some limits of AMS direct dating of Paleo-
lithic paintings are apparent. It has often been said that the method is statistical, i.e., it
is only reliable within wide enough margins and, in addition, provided a number of
dates corroborate one another. This is true for all archaeological samples analyzed, ir-
respective of their nature. Analyzing rock paintings, however, is far more difficult
than dating an archaeological living site.

First, because of the necessity of not damaging the art, only minute amounts of
paint can be lifted. This means that in most cases, physical scientists in their laborato-
ries are obliged to work on the fringes of what the method allows. The risk of error due
to contamination, which can produce dates either too early or too late, is far greater
with tiny amounts of material than with the relatively large quantities of charcoal or
bones that can be retrieved from many archaeological sites. Second, if laboratory er-
ror or initial contamination occurs when dating a layer from a habitat, it is immediate-
ly apparent because of the presence of artifacts that enable one to assign a tentative
date to the site: it can be checked against other analyses. Not so with rock art, where a
direct date rests alone, with no possibility of other validation. In the best of cases it
means that we have dated one figure or even one portion of a figure, if there was re-
painting. It does not mean that the entire panel or the whole cave is dated to the same
period. If for some reason or other the date is wrong, and this has been known to occur
in a number of cases in traditional archaeological contexts for which thousands of
dates are now published, we have no way of checking it at present. Only when a date
can be confirmed by different methods should it be considered valid.

When no other corroborating method is available, we must keep in mind the old
saying “one date is no date.” It is just one piece of evidence that must be kept in mind
but should be considered tentative until other dates have been obtained and a picture
emerges, as in the case of well-dated archaeological cultures. Will this happen one
day with rock art ? It is far from certain, as conservation constraints make it impossi-
ble to liftas many samples as are needed. Even once techniques have improved, there
will always be a difficult balance to strike between the size of the sample and the trust
worthiness of the analysis. As we have seen, the smaller the sample, the greater the
possibility of error due to contamination, whatever the method used (Clottes 1994a:
21).

Conclusion

In 1992, one of the symposia at the Cairns International Conference on Rock Art
was called: “The Post-Stylistic Era: Where Do We Go From Here?,” a provocative ti-
tle which implied that with the recent advances new technologies permitted, espe-
cially paint analyses and AMS dating, stylistic studies had lost their importance and
had been or would soon be replaced by more scientific methods. Such optimistic con-
fidence in the “hard” sciences is certainly understandable. However, after several

50 CLOTTES

years of practice, the state of rock art studies for the Upper Paleolithic of Europe is far
from ideal. Two dozen dates are now available for 19 figures in six caves, when nearly
300 caves and many thousands of paintings and engravings are known. In order to
achieve some of the aims of modern research, for example, fixing the exact duration
of particular conventions and techniques, the different phases in the art of a cave, and
the relationship between various caves and among different regions, a satisfactory
statistical probability must be reached. To do so, it would obviously be necessary to
have sufficient dates and paint analyses. This would hopefully avoid making unwar-
ranted assumptions or straying too far in one’s interpretations, as is easy to do when
the results are too few and do not conform to expectations.

Two difficulties stand in the way. One is the funding and the involvement of spe-
cialized laboratories: Will they be willing to make scores of analyses per cave? And if
they are, how will the cost be met? The second difficulty, even more serious, is that of
sampling possibilities and the necessity of preserving the art. In Niaux and in Coug-
nac it is no longer possible to find enough charcoal, halfa milligram, to meet the re-
quirements for new AMS dating. This is also the case in Lascaux and in many other
caves, which explains why results are still so few.

AMS dating, paint analyses and other laboratory techniques have already brought
a number of fascinating results and raised both hopes and problems. They will no
doubt assume an even greater place in rock art research, especially ifeconomic obsta-
cles do not stand in the way and if technological advances allow the lifting of stll
smaller samples. Many new discoveries, both welcome refinements and a few up-
heavals, may be expected during the years to come. However, it us doubtful that a sig-
nificantly different picture of Paleolithic art will emerge before a great many more
years,

Acknowledgment

I am grateful to Marilyn Garner for her help in smoothing my English.

Literature Cited

Altuna, J., J. M. Apellaniz & I. Barandiaran. 1993. Estudio de las Pinturas de Zubialde
(Alava). Resumen de los Resultados. Vitoria-Gasteiz, Diputacion foral de Alava.

Bahn, P.-G. 1993. The ‘dead wood stage’ of prehistoric art studies: style is not enough. Pages
51-59 in P.-G. Bahn & M. Lorblanchet, ed., Rock Art Studies: The Post-Stvlistic Era or
Where Do We Go from Here? Oxbow Monograph 35, Australia.

Ballet, F., A. Bocquet, R. Bouchez, J.-M.-D. Coey & A. Cornu. 1979. Etude technique de pou-
dres colorées de Lascaux. Pages 171-174 in A. Leroi-Gourhan & J. Allain, eds., Lascaux In-
connu. XII° Supplement to Gallia Préhistoire, Editions du CNRS, Paris, France.

Bednarik, R. 1992. Developments in rock art dating. Acta Archaeol. 63:141-155.

. 1993. The direct dating of rock art. Rock Art Res. 10, 1:48-51
1994 . On the scientific study of palaeoart. Semiotica 100, 2/4:141-168.

Begouen, R. & J. Clottes. 1991. Portable and Wall Art in the Volp Caves, Montesquieu-
Avanteés (Ariége). Proceed. of the Prehist. Soc. 57, 1:65-79.

Breuil, H. 1952. Quatre Cents Siécles d'Art Pariétal. Montignac: Centre d’ Etudes et de Docu-
mentation préhistoriques, Dordogne, France.

NEW ANALYTICAL TECHNIQUES 5]

Brunet, J., B. Callede & G. Orial. 1982. Tarascon/Ariége (Ariege). Grotte de Niaux: mise en
évidence de charbon de bois dans les traces préhistoriques du Salon Noir. Studies in Con-
servation 27:173-179.

Buisson, D., M. Menu, G. Pincon & P. Walter. 1989. Les objets colorés du Paléolithique supe -
rieur. Cas de la Grotte de La Vache (Ariége). Bull. de la Soc. Préhist. Frangaise 86, 6:183-
191.

Cabrera-Garrido, J. M. 1978. Les matériaux des peintures de la Grotte d’ Altamira. Pages I-9 in
Actes de la 5° Réunion Triennale de L’ICOM. Zagreb, Yugoslavia.

Chaffee, S.-D.,M. Hyman & M. Rowe. 1993. AMS NC dating of rock paintings. Pages 67-77
in J. Steinbring, A. Watchman, P. Faulstich & S.-C. Tagon, eds., 7ime and Space. Dating
and Spatial Considerations in Rock Art Research. Occasional AURA Publication 8, Aus-
tralia.

Clottes, J. 1974. A propos d’un ouvrage récent sur la Caverne de Niaux. Bulletin de la Société
Préhistorique Frangaise 71, 5:137-139.

Clottes, J., 1990a. La préparation des peintures magdaléniennes des cavernes ariegeoises. Bull
de la Soc. Préhist. Francaise 87:1 70-192.

Clottes, J. 1993. Paint analyses from several Magdalenian caves in the Ariége region of France.
J. Archaeol. Sci. 20:223-235.

Clottes, J. 1994a. Post-Stylistic? Pages 19-25 in P. G. Bahn & M. Lorblanchet, eds., The Post-
Stylistic Era or Where Do We Go from Here? Second AURA Congress, Cairns, Australia.

Clottes, J. 1994b. Dates directes pour les peintures paléolithiques. Bulletin de la Société
Préhistorique Ariege-Pyrénées XLIX:51-70.

Clottes, J. 1995. Les Cavernes de Niaux. Art Magdalénien en Ariége. Seuil, Paris, France.

Clottes, J. in press. Investigating AMS dating for Paleolithic cave paintings. Archaeometry
Symposium, Armidale, Australia.

Clottes, J., J.-M. Chauvet, E. Brunel-Deschamps, C. Hillaire, J.-P. Daugas, M. Arnold, H.
Cachier, J. Evin, P. Fortin, C. Oberlin, N. Tisnerat & H. Valladas. 1995. Les peintures
paléolithiques de la Grotte Chauvet-Pont-d’ Arc, a Vallon-Pont-d’ Are (Ardéche, France):
datations directes et indirectes par la méthode du radiocarbone. Comptes-rendus de 1'A-
cadémie des Sciences de Paris, 320:1 133-1140.

Clottes, J. & J. Courtin. 1994 La Grotte Cosquer. Peintures et Gravures de la Caverne
Engloutie. Seuil, Paris.

Clottes, J., M. Menu & P. Walter. 1990b. New light on the Niaux paintings. Rock Art Res
7:21-26.

Clottes, J., H. Valladas, M. Cachier & M. Arnold. 1992. Des dates pour Niaux et Gargas. Bu//.
de la Soc. Préhist. Francaise 89:270-274.

Couraud, C. 1982. Techniques de peintures préhistoriques. Expériences. Couleur, Revue du
Centre Frangais de la Couleur 19:3-6.

Couraud, C. 1988. Pigments utilisés en préhistoire, provenance, préparation, mode d*utilisa-
tion. L ‘Anthropologie 92:17-28.

Couraud C. & A. Laming-Emperaire. 1979. Les colorants. Pages 153-170 in A. Leroi-Gourhan
and J. Allain, eds., Lascaux Inconnu. XII° Supplement to Gallia Prehistoire, Editions du
CNRS, Paris, France.

Dorn, R. 1992. A review of rock varnish dating of rock engravings. /ni/. Nws/tr. Rock Art 2:10-
14.

Franklin, N. 1993. Style and dating in rock art studies: The post-stylistic era in Australia and
Europe question. Pages 1-13 in P.G. Bahn & M. Lorblanchet, eds., The Post-Stylistic Era
or Where Do We Go from Here? Oxbow Monograph 35, Australia.

an
to

CLOTTES

Labeau, M. 1993. New analysis of the Cougnac cave pigments. Pages 72-73 in P.G. Bahn & M.
Lorblanchet, eds. The Post-Stvlistic Era or Where Do We Go from Here? Oxbow Mono-
graph 35, Australia.

Leroi-Gourhan, A. 1965. Préhistoire de 1'’Art Occidental. Mazenod, Paris, France.

Lorblanchet, M. 1990a. L’art parietal. De nouvelles méthodes de datation de Part prehistori-
que. Pour la Science 186:10-12.

Lorblanchet, M. 1990b. Etude des pigments de grottes ornées paléolithiques du Quercy. Bulle-
tinde la Société des Etudes Littéraires, Scientifiques et Artistiques du Lot CX1, 2:93-113.
Lorblanchet, M. 1992 Introduction. Pages XV-XXXII in M. Lorblanchet, ed., Rock Art in the

Old World. New-Delhi, India.

Lorblanchet, M. 1993. From style to dates. Pages 61-72 in P.-G. Bahn & M. Lorblanchet, eds.,
The Post-Stvlistic Era or Where Do We Go from Here ? Oxbow Monograph 35, Australia.

Lorblanchet, M. 1994a. La datation de art pariétal paléolithique. Bi/letin de la Société des
Etudes Littéraires, Scientifiques et Artistiques du Lot CXV, 3:161-182.

Lorblanchet, M. 1994b. Le mode d’utilisation des sanctuaires paléolithiques. Pages 235-251 in
Homenaye al Dr. Joaquin Gonzalez Echegaray. Museo y Centro de Investigacion de Alta-
mura, Spain,

Lorblanchet, M. 1994¢. Cougnac. /ntl. Nwsltr. Rock Art 7:6-7.

Lorblanchet, M., M. Labeau, J.-L. Vernet, R. Fitte, H. Valladas, H. Cachier & M. Arnold.
1990a. Etude des pigments des grottes ornées paléolithiques du Quercy. Bulletin de la Soci-
été des Etudes Littéraires, Scientifiques et Artistiques du Lot CX1, 2:93-143.

Lorblanchet, M., M. Labeau, J.L. Vernet, R. Fitte, H. Valladas, H. Cachier & M. Arnold.
1990b. Paleolithic pigments in the Quercy, France. Rock Art Res. 7, 1:4-20.

Loy, T.-H., R. Jones, D.-E. Nelson, B. Meehan, J. Vogel, J. Southon & R. Cosgrove. 1990, Ac-
celerator radiocarbon dating of human blood proteins in pigments from Late Pleistocene art
sites in Australia. Antiquity 64:1 10-116.

Menu, M. and P. Walter. 1991. Les premiers artistes peintres. La Recherche 235, 22:1086-
1089.

Menu, M., P. Walter, D. Vigears & J. Clottes. 1993, Fagons de peindre au Magdalenien. Bull.
de la Soc. Préhist. Frangaise 90, 6:426-432.

Meroe, L. & J. Mazet. 1956. Cougnac, Grotte Peinte. Stuttgart: W. Kohlammer Verlag, Ger-
many.

Moissan, H. 1902. Compte-rendu a |’ Académie des Sciences. Comptes-rendus de |'Académie
des Sciences |34:1536-1540.

Moissan, H. 1903. Compte-rendu a |’ Académie des Sciences. Comptes-rendus de |'Academie
des Sciences 136:144-146.

Nobbs, M. & R. Dorn. 1988. Age determinations for rock varnish formation within petro -
glyphs: Cation-ratio dating of 24 motifs from the Olary region, South Australia. Rock Art
Res.5:1 08-146

Pepe, C., J. Clottes, M. Menu & P. Walter. 1991. Le lant des peintures paleolithiques arie-
geoises. Comptes-rendus de l'Académie des Sciences de Paris 312, 11:929-934.

Russ, J.,M. Hyman, H.J. Shater & M.-W. Rowe. 1990. Radiocarbon dating of prehistoric rock
paintings by selective oxidation of organic carbon. Nature 348, 6303:710-711.

Valladas, H., H. Cachier, M. Arnold, F. Bernaldo de Quiros, F., J. Clottes & P. Uzquiano.
1992. Direct radiocarbon dates for prehistoric paintings at the Altamira, El Castillo and Ni-
aux caves. Nature 357:68-70.

Van der Merwe, N.-J., J. Seely & R. Yates. 1987, First accelerator carbon-14 date for pigment
from a rock painting. S. Afr, J. Set, 83:56-57

Vandiver, P. 1983. Paleolithic Pigments and Processing. Unpublished M. Sc. Thesis, Depart-
ment of Material Science and Engineering, Massachusetts Institute of Technology.

Watchman, A. & D. Lessard. 1992. Dating prehistoric rock art by laser: a new method for ex-
tracting trace organic matter. /ntl, Nwsltr, Rock Art 2:14-15.

Paleolithic Image Making and
Symboling in Europe and the

Middle East: A Comparative
Review

Alexander Marshack

4 Washington Square Village
New York, NY 10012-1908

The long held belief that “art,” or image making, began in Europe as a “species”
event during the Aurignacian period at the time that anatomically modern humans
displaced the Neanderthals has begun to dissipate as the still rare archeological evi-
dence for early image making has begun to be found in the Middle East, Africa, Asia,
and Australia. The evidence that Neanderthals may have produced beads in France
during the Chatelperronian, perhaps as a tradition derived from contemporaneous
“modern” human, has also validated an argument that I have been making for a quar-
ter of acentury, that the capacity for diverse forms of symboling was not the result ofa
sudden, stochastic, neurological ‘“‘re-wiring” of the anatomically modern human
brain, a change that supposedly made complex modes of symboling, and even mod-
ern language, “suddenly” possible. I had also long argued that the “creative explo-
sion” of the European Upper Paleolithic was not the “origin” or “beginning” of “art”
(Marshack 1972,1991a). Both arguments were derived from direct, microscopic
analysis of the complex and variable archeological symbolic materials.

The European Aurignacian

In the 1960's I conducted the first study of the Aurignacian (ca. 32,000 BP) ivory
figures from Vogelherd, Germany. Microscopic study not only revealed a remarkable
skill and capacity for animal representation, but the presence also of diverse and com-
plex modes of image use “over time,” involving both long-term curation and different
modes of periodic over-marking. At the same time, my microscopic study of other
classes of Aurignacian artifacts documented the presence of a tradition involving the
non-decorative accumulation of incised sets and subsets, often on fragments of waste
or scrap bone or broken tools. These latter artifacts suggested the presence of a vari-
able tradition of Upper Paleolithic “notation.” The unexpected recent acknowledg-

Beyond Art: Pleistocene Image and Symbol Memorrs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

54 MARSHACK

ment of Upper Paleolithic “notation” by the one researcher who had been its most
methodological and systematic critic, has reopened for discussion the importance of
this abstract, specialized form of early symboling (cf. D'Errico and Cacho 1994; D'Er-
rico 1995; but see also Marshack 1995a, 1996 a,b).

These early findings suggested that Upper Paleolithic symboling modes were not
always intended as “art.” but were often aspects of what I had termed “time- factored”
and “time-factoring” symbol systems, systems that marked and differentiated modes
of periodic symbolic behavior within these early cultures. The presence of carved
vulvas in Aurignacian home sites of the Dordogne, France, (cf. Marshack 199 1b). of
vulvas in the Gravettian of east and Central Europe (Marshack 1996c), and the pres-
ence of ritual burials indicated other modes of periodic symboling and reference.

These studies of early Upper Paleolithic imagery initiated a wide-ranging micro-
scopic investigation of possible still earlier, pre-Upper Paleolithic symboling in
Europe. They documented a production of red ocher in the Acheulian of Becov,
Czechoslovakia, ca. 250,000 BP (Marshack 1981) and the carving and curation of a
Mousterian “nonrepresentational” oval plaque at Tata, Hungary, ca. 100,000 BP
(Marshack 1976), traditions that suggested the production and use of symbolic mate-
rials at particular times and for particular purposes. If symboling systems helped to
mark and structure the periodic, ongoing, cultural tapestries of these early human cul-
tures, then the later Upper Paleolithic “creative explosion” may merely have repre-
sented the development of unique “time-factored” and ‘“time-factoring” cultural
systems under particular conditions and by use of particular technologies. When the
conditions and the context changed, the “time-factored” symbolic tapestry disinte-
grated and “Ice Age Art” ended.

The Levant
A Middle Paleolithic Engraving

In the 1980's an engraved plate of flint cortex (Figure la, b) was excavated froma
Middle Paleolithic level near the village of Quneitra, on the Golan Heights (Goren-
Inbar 1990). Itcame from a level dated at ca. 54,000 BP, some 15 to 20,000 years ear-
lier than the first images to appear at the European “Middle Paleolithic/Upper Paleo-
lithic transition.” During this period anatomically modern humans were living in the
Middle East with Neanderthals. There is no evidence that these two groups inter-
acted; but they both used a similar Middle Paleolithic lithic technology, worked skins
and wood, hafted tools, had hearths, used red ocher, and buried their dead. Since they
were inhabiting the same territory, | assume that there may have been some level of
cultural contact or recognition between the two populations, as there was apparently
20,000 years later in France during the brief Chatelperronian period.

Who, therefore, engraved this composition 20,000 years before the Upper Paleo-
lithic began in Europe — modern humans or the Neanderthals?

The archaeological chronology of Israel (Figure 2) places Quneitra near the top of
the Middle Paleolithic sequence, shortly before the Middle Paleolithic/Upper Paleo-
lithic transition that began in Israel about 47,000 BP, thousands of years before the
MP/UP transition began in Europe (Marks 1983; Bar-Y osefand Belfer-Cohen 1988).
If the cortex was incised by an anatomically modern human, as part of a tradition of
image making, did descendants of these modern humans eventually move north into
Europe carrying their culture and symboling traditions? The Levantine Neanderthals,

w

Uy
(‘Adoososoiut Aq paurutiajap sy) “sayois popuadde “yoys ‘ysis Jo pasodwios ase sapodlotwWias sole paysau ayf, “Auo}s ay} JO
Wa] ayy 1e syieul [B9TLIOA 3] pue asayy surpunouns Buryieur JULe} OU] “Sa[OITOILuas INOF [LUA ayy Burjeoipul “xaLIO9S ay} uo pasioul uoltsoduios ay} JO UOHTpual St eUaAYOS
(q : oi) d A]Pplyy aunuvaay ‘sau JOA BUIPUNOLNS pure SapostolWas pajsau Ino} YM pastour (ud 7'7) awed xayO9 1e]F ‘eIauNd (v “] AY
Y)
a
VY
<
=
1S)
=
a
e
fe)
ica)
<
a
i
(eS)
>
=
=)
Re
Nn
w
—
=
<
[a4
<
a
=
[e)

C

56 MARSHACK

ka ARCHAEOLOGICAL ENTITIES : HUMAN ISOTOPE
AND SITES i FOSSILS STAGES
os UPPER PALAEOLITHIC
30 LEVANTINE AURIGNACIAN E yeerar eee
AHMARIAN TRADITION } EGGBERT (KSAR AKIL)
40 Transitional Industries po) es

:

=a * QUNEITRA are TOR SABIHA AMUD
BARA ;

= ae pi le KEBARA

70 "“TABUN - B type” H

i SHANIDAR 7
80 i

i TABUN Woman ?
100 QAFZEH i QAFZEH
110

SKHUL i SKHUL
120
130 } “TABUN - C type”

; TABUN Woman ?
me i TABUN C jaw
150 AIN DIFLA ? i

TERE REN. “AIN ACE i

160 DOUARA IV
. , . ROSH EIN MOR }

170 |. "TABUN- D type :
?

180 i
190
200 ACHEULO-YABRUDIAN
210 ?
555 (MUGHARAN TRADITION) |

i
230 YABRUD 1, 11-23
240

ZUTTIYEH ?
250

hiGURE 2. Chronological chart of the Acheulian, Middle Paleolithic and Upper Paleolitihic sequence in Is -
rael, Quneitra, occurring in the late Middle Paleolithic, is near the top of the sequence, not far from the be-
ginning of the Levantione Upper Paleolithic. A proposed chronology for Levantine sites and human
fossils. Question marks indicate the uncertain position of fossils or those which are not yet dated. The ap-
parent overlapping of the Mousterian industries reflects the standard deviations on the a ge Measurements
and conflicting dates. The industries are stratified and therefore cannot be contemporary with each other.”
(After Bar-Y osef 1992:195)

COMPARATIVE STUDY OF PALEOLITHIC IMAGES

nn
a

represented by skeletons at Kebara and Amud, were apparently still in Israel when the
Quneitra engraving was made, but they would shortly disappear. According to the
“Out-of-A frica” hypothesis for modern human origins and dispersal, it was within the
period near the bottom of the Israeli sequence (Figure 2), i.e., between ca. 200,000
and 100,000 BP, that the “moderns” began to move north, first into the Middle East

and then into Eurasia. I shall discuss artifacts from each period in this evolutionary
and cultural sequence.

The Quneitra Engraving

Microscopic study of the Quneitra engraving reveals an extraordinary skill and so-
phistication. Holding the small stone (7.2 cm) in one hand, presumably the left, a set
of four concentric semicircles was incised by an engraving tool held in the other hand,
presumably the right. Each arc was made by appending small straight strokes linearly
to one another (Figures |b, 3a, 3b), while the secondary hand turned the stone and ori-
ented it so that each stroke could be properly incised (cf Marshack 1996d). This se-
quence was performed four times, during which the engraver had to evaluate the
different, coordinated actions of the two hands and the developing, preconceived im-
age. After the arcs were engraved, long strokes were incised both above and around
the arcs.

Who, then, incised this image? The capacity for planning and for sustained, com-
plex production, using a visually mediated Ovo-handed skill, was of the type that was
involved in the making and use of Middle Paleolithic stone tools. These capacities
were, therefore, shared by the Neanderthals and anatomically modern humans.

It is perhaps significant that this early image 1s not of an animal, a human, or the
decoration of a tool. It represents, instead, an abstract, schematic depiction of some
sort. Whatever its meaning, the composition contradicts the view that imagemaking
began in Europe at the Middle Paleolithic/Upper Paleolithic transition. There is no
evidence during this period foran advanced bone technology in Israel and there are no
bone beads. I nevertheless assume that the makers of this image wore personal deco-
rations, probably of perishable materials such as skins, leather, or twined thongs, that
they used red ocher, and perhaps tied their hair and made beads or amulets of perish-
able animal parts or wood. Wood carving has been documented, for instance, even in
the Middle Paleolithic of Spain (Carbonell & Castro-Curel 1992).

Since the Quneitra image was incised shortly before the Middle Paleolithic/Upper
Paleolithic transition began in the Middle East, it is probable that it was a part of in-
cipient cultural developments that led to the Upper Paleolithic transition in Israel and
that the composition was, therefore, made by anatomically modern humans, some of
whose descendants would disperse into Eurasia.

An Acheulian Figurine

A few kilometers from Quneitra on the Golan Heights, at the site of Berekhat Ram,
Israeli archeologists excavated a shaped piece of volcanic tuffina level dated, mini-
mally, at ca. 233,000 + 3,000 BP (Figures 4a, b) (Feraud et al. 1983; Goren-Inbar
1986). This places the artifact at the lower end, or just outside the lower range, of the
theoretical mtDNA “Out of Africa” scenario for the origins and dispersal of modern
humans (but see Ayala, 1995, fora possible increase in the early range of these dates).
Because of the early date it had been assumed by commentators who had not studied

58 MARSHACK

Ficure 3. A close-up of the marking at the right of the Quneitra cortex, indicating a portion of the concen -
tric ares and the straight line incising of the short strokes; a) An extreme close-up of the stroke incising a
double track at the far right of the stone. The stroke tails out with a single long track, made at an angle; b)
The short stroke incising a double track in the outer, fourth semicircle. The top of this stroke over-crosses
and extends beyond the upper stroke to which itis appended and at bottom itextends beyond the lower ver-
ical to which it 1s appended. There is a faint indication of striation within the double tracks of strokes “a.”

and “b.”

the artifact that the shape might be natural, since “scoria” can assume odd shapes and
even grooving when volcanic material is hurled through the air as molten lava and im-
pacts the earth (Pelcin 1994). Dr. Sergiu Pelz, a volcanologist and pyroclastic special-
ist with the Geological Survey of Israel, after studying my microscopic analyses of
the figure and then studying the artifact, declared that it is nota scoria, but a piece of
compacted, agglomerate volcanic tuff, shaped by use of a tool (Marshack 1995b;
Goren-Inbar & Peltz 1995). A microscopic analysis of the carving and a discussion of
the data's relevance 1s being prepared for publication. Here it may be sufficient to in-
dicate that the grooves of the neck, i.e., those coming from the front and the back, do
not meet but instead cross over each other on the left side, a process that does not oc-
cur in the grooving noted for scoria but is common when one is attempting to incise a
groove around a three dimensional form. In addition, the grooves of the neck present
flat beveled planes that are made at a wide angle, with a deep groove incised between
the planes.

COMPARATIVE STUDY OF PALEOLITHIC IMAGES

‘durms 3a] ayy Jo Suruayey ev pue syoonng ayi ie Furpunos ys w sysadans reas ayy Jo ajtjord ay] “Iwao st aury Yoou ayy Fuge ysayo ayy Jo aurjd ayy “apts ws ayy
Jip Wapeur sey vase Jappnoys ayy ur uorsnut apqqod v Jo aouasaid ay] “WOT ay} spreMo} aAooIs Yoau ay fo adojs premuMop ayy Sunes
-Iput ‘apts Yo] (q. {Woy UL aAOIT Yoau aty WoL spuadsap ysayo ayy fo aueyd ydnage ayL, “opts ayy ye we pasterzdn uv Jo WHIOJ ay} st ay] “Jappnoys ayy Sunsasans aaooss Yyoau

ayy wrory aueyd ydnaqe ay) pue yoou ay} ye Futpanaq ary Buneorput ‘apts YAY (ev ‘ypry ayesowoysuoo ve ut padeys amnsy uewmy syewuayag ‘uelNayoy ‘wey wYyyaag “p IuNO1J

uo savadde se yons aueyd ve ayeaia 0} YN

60 MARSHACK

My microscopic analysis indicates that a natural form had been altered to enhance
the image ofa “figurine.” Slight modifications of natural forms are, of course, well-
known in the Upper Paleolithic. It must be noted that the capacity to envision a poten-
tial form in unworked natural material is a crucial perceptual, cognitive aspect of both
Acheulian and later lithic technology.

The Levantine Aurignacian

Twenty thousand years after the Quneitra engraving, a complex and varied Aurig-
nacian bone technology, with an associated bone “art,” developed in Western Europe.
Some two or three thousand years after the more sophisticated Aurignacian bone im-
ages (Vogelherd, Hohlenstein-Stadel) had been produced in Europe, hunter-
gatherers carrying the Aurignacian culture made a foray into the Middle East (Tixier
1974; Tixier & Inizan 1981; Belfer-Cohen & Bar-Yosef 1981, Bar-Yosef & Belfer-
Cohen 1988), reaching as far south as Lebanon and Israel.

Levantine Aurignacian Artifacts

Some of these hunter-gatherers, carrying their European bone and lithic technol-
ogy, camped at the Israeli cave of Hayonim (Figures 5, 6). They also brought much of
their symbolic culture (Figure 7). The Aurignacian animal tooth beads from Hay-
onim, for instance, are precisely like Aurignacian beads that I studied in Germany and
France.

It has often been argued that Aurignacian beads mark a supposed “beginning” for
human self-awareness, social complexity, art, the standardization of human produc-
tion (but see Belfer-Cohen & Goren-Inbar 1994) and even the beginnings of fully
modern human language (White 1989 a, b; 1992, 1993). Did these migrating hunter-
gatherers, then, bring supposedly “new” human capacities for symboling into the
Middle East? Or was it primarily their stone and bone fec/ino/ogy and the symboling
traditions and skills that they had developed in the use of these materials, that they
brought into the Middle East?) The migrants were composed of small, scattered
groups of mobile hunter-gatherers strung along the fertile coastal plain of the eastern
Mediterranean. The demographic and networking complexities that were at this time
“exploding” within the diversified ecologies and vastness of Europe did not exist in
the Middle East. Lassume that it was Aurignacian technology and the symboling con-
cepts and traditions that had developed in use of that technology that were brought
into the Middle East, rather than a “new,” anatomically modern, biological, species
capacity, or the purported “first” traditions of human art (cf; Marshack 1994). After
some generations the “Aurignacian” population and culture were apparently ab-
sorbed into the indigenous Middle East hunting-gathering cultures and populations,
populations and cultures that had their own language, symboling traditions, social
complexity and forms of personal decoration.

The “Aurignacians” brought with them their European mode of symbolic engrav-
ing, including a tradition of accumulating sets of marks such as those I had termed
“notation” in Europe (c/. Tixier 1974). A small engraved pebble (See Bar-Y osef this
volume, Figure 5, p. 174) was excavated in Aurignacian level D at Hayonim (Figures
8 a-d). It contains a fossil intrusion that runs through the limestone at one side and it
may have been chosen for that reason. The initial Israeli publications (Belfer-Cohen
& Bar-Yosef 1981:35; Bar-Yosef & Belfer-Cohen 1988) depict a rather crude me-

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 61

FiGure 5. Bone points from the Aurignacian of Hayonim D, Israel, of the type found during this period in
Europe.

FiuRE 6. Dufour points from the Aurignacian of Hayonim D, Israel comparable to those found in the
Aurignacian of Europe.

62 MARSHACK

Ficure 7. Animal tooth beads from the Aurignacian of Hayonim D, Israel, of the type found during this pe -
riod in Europe, ca. 29-28,000 BC

lange of lines within which one can make out the simple outline ofa horse. This horse
is the earliest known example of animal engraving in the Levant. The stone is covered
with red ocher in its central area and it may therefore have been used as a rubbing
stone to make ocher powder.

Complexity of the Horse

The seemingly “random” linear marking found in and around the horse also oc-
curs on symbolic artifacts found in Ice Age Europe. It is a type of marking not only
found with animal images but also alone on mobiliary artifacts and it is found often in
the Franco-Cantabrian caves (cf: Marshack 1977, 1992, 1993). Microscopic study of
the pebble reveals that it consists of different types of imagery or “motifs” compara-
ble to the variability found in Europe.

The first image to be incised on the pebble was the outline ofa running horse made
in what Leroi-Gourhan, in mid-century, had labeled as “Styles I-II” of the French
Aurignacian (Leroi-Gourhan 1965). The horse has no hooves, no pelage, no mouth or
nostrils and no underbelly; the back is merely an undifferentiated arc; the eye is a sim-
ple gash. The image could easily pass as a typical Aurignacian engraving.

It is not the horse, however, that is most interesting. After the horse was engraved
it was “killed” by darts, spears or wounds incised in and across the animal in different
“styles,” with different points and, therefore, presumably at different times (Figure
8b). This variation in the style of depicting schematic weapons is typically European.
But the horse was not only symbolically “killed”; it was associated with motifs, also
made in different “styles,” at different times and with different points (Figure 8c).
These motifs include two schematic “streams” or “bands,” a hatch design that sur-
rounds and over crosses the fossil intrusion, and arcs that may represent the back ofan
animal. As noted, the stone had also been overmarked with red ocher, suggesting the
presence of symboling traditions of still a different type. The pebble (Figure 8d),
therefore, provides us with evidence for various modes of image and symbol use, a

63

COMPARATIVE STUDY OF PALEOLITHIC IMAGES

punoy uryzeuraao Jo addy ayn squasarg (p swuodey] wrogy ayqgad ayy uo spout pue SaAVUUT JUDIALIP BUIAL

aypoayeg Jaddy uradoing ayy noysnomp
Bua Jo aouanbas ayy JO UONIpPuas NNeUIAYIS *g INNA

64 MARSHACK

range and variability that is also found in all periods of the West European Upper Pa-
leolithic, from the Aurignacian to the Magdalenian.

Atasimple level we can probably categorize the horse as “art” made in a temporal,
regional “style” and indicate that it was associated with diverse motifs, apparently in-
cised over a period of time. But why was the horse engraved? As “art”? For public
display? As a form of information encoding? As an animal that was hunted and was
therefore involved in “hunting magic’’? Or was it a symbol and metaphor that could
be used in different ways at different times? But if the horse was a symbol and meta-
phor, who made the engraving? A hunter, ora ritual specialist such as a “shaman” for
whom each type or instance of marking had a different meaning? If it was a “sha-
man,” over how longa period was the pebble kept and in what contexts or seasons was
it incised and used? For what purposes? Was it engraved when the horse was season-
ally present, or was it marked when the horse was vot present but was expected and
had not yet arrived? Or was the horse merely “killed” symbolically for ritual pur-
poses such as an initiation, a birth, a death, or ina prayer or invocation? Does the ana-
lytical complexity suggest the presence ofa “ritual specialist,” the member ofan elite
class that may have carried aspects of Aurignacian culture into the Levant (c/. Bar-
Yosef, this volume, p.165)? Whatever the interpretation, it is clear that we are dealing
with more than mere “art” or depiction. The association ofa horse with different types
of symbolic marking seems, in fact, to have been as important as representation of the
horse itself,

The French Magdalenian
The “Type” Image

I began to study these variable Upper Paleolithic modes of using animal images
nearly thirty years ago (cf, Marshack 1969a, 1972, 199 1a). One of the earliest of these
studies was of the woolly mammoth incised on a piece of ivory that was excavated at
La Madeleine, France, in 1864 (cf. Lister & Bahn, 1994:96; Marshack 1995:33). This
was the image that validated the presence of art in the “Reindeer Age” and it has, as a
result, been examined and published many times. My interest had been originally
roused by Breuil’s rendition (Maska er a/. 1912:277) which suggested that the mam-
moth had been “used” and “reused” in different ways by a series of later additions.

[he first image to be engraved was a magnificent woolly mammoth (Figure 9a).
The outline of the body and head and the hairy coat are realistic, the small eye is beau-
tifully rendered, even the anal flap and upraised tail are depicted. The mammoth is
leaning forward with the rear leg thrust back as though in defense or attack. Modern
Eurasian elephants, when attacked or attacking, raise their tail in this manner and we
may therefore have the image of a mammoth under attack.

The microscope reveals that after the mammoth was engraved it was “killed” with
spears or darts made at different times, in different styles and to different lengths (Fig-
ure 9b). Some of the weapons or “signs” are typical of the European Upper Paleo-
lithic. But the mammoth had not only been “killed,” it had also been “renewed” or
“reused” by adding schematic parts of the animal. There are three or four extra backs
within the mammoth; an extra tail, also upright, was appended at the rear and a
number of extra tusks were added at the front (Figure 9c). The entire composition
(Figure 9d) was apparently accumulated over a period of time, ina manner similar to,

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 6

nn

but more highly evolved than, that found at Hayonim in Israel more than 15,000 years
earlier.

How does one interpret such a composition? We can discuss the realistic “style”
as a chronological marker for the Magdalenian; or the fact that this level of realism de-
veloped only in the West European Upper Paleolithic, but not in Central Europe or on
the Russian plain and we can ask why. We can discuss it as an example of the ways in
which animal images were used in the Upper Paleolithic; or discuss the fact that when
this mammoth was incised the species had begun to diminish in Europe and would
soon be extinct. We can discuss the changes in climate occurring during the Mag-
dalenian and the fact that the mammoth was probably incised during a cold period.
We can discuss the fact that the mammoth was never a major item of diet in the Mag-
dalenian but that it was nevertheless, like the “killed” Aurignacian lions of Vogelherd
and Chauvet, a significant symbolic animal. Or we can discuss the image in terms of
“art” and “style”; as a form of “information encoding” or as an indication of group
membership and differentiation. Or we can note that in the Middle East there were no
mammoths or the European range of midlatitude and subarctic fauna, or migratory
herds that roamed great distances (reindeer, bison, efc.). Or we can discuss the ex-
traordinary variability of image use found in Ice Age Europe and compare it to the
variability that is beginning to be found in the Middle East (see Bar-Y osef, this vol-
ume, p.165)

In the Magdalenian cave of Les Trois Freres, in the French Pyrenees, there are
hundreds of engraved animal images, but only one mammoth. When I conducted re-
search in this cave, I was struck by the fact that the images of bison, horse, reindeer,
ibex and lion were made and used or reused in many ways; occasionally animals were
represented by merely a part of the animal, a head, a tail, a leg, a rump, etc.

A beautiful bison at Les Trois Freres, incised in the realistic Magdalenian style,
had been repeatedly “killed” by numerous spears or darts (Figure 10). The initial en-
graving depicted a bison in summer moult, with a bare upper body but a scraped indi-
cation of pelage in the lower half, a typical Magdalenian mode of rendering the bison
in its summer coat. Added to the upper body are two zigzag motifs made by different
engraving points and presumably at different times. The zigzag is a motif that I have
argued is the schematic rendering of the summer molt when it is incised within this
upper area of the body in certain species. Molting, which occurs among bison in the
late spring, begins as jagged zigzag clusters or clumps of hair that fall from the upper
half of the body (cf Altuna & Appaleniz 1976:169, Figures 109, 110).

In the original engraving the tail of the bison hung downward. Over this tail a sec-
ond tail, standing vertically upright, was incised by a different point, as on the La
Madeleine mammoth. The “summer” bison had been “killed” — but it had also been
reused or renewed by adding different motifs or images, having different meanings.
We can therefore discuss this bison at numerous levels; in terms of Magdalenian
style, the meaning of the motifs, the varying modes of use, the probable season in
which bison appeared in the Pyrennean foothills, the profusion of bison in the cave in
comparison to the one mammoth, etc. It may, however, be sufficient to indicate that
similar modes of animal image use occur in France, Spain, Germany and Italy (c/.
Marshack 1969).

MARSHACK

66

Runpretu jo oun pue Aurteawy Uo sit ped Ajuaiedde yoeq ‘ash adeu pue Ruryeus JOU! JO sapout WaaTIP JO aouanb
-as p syuatunsop Aypeniae ing asurjatut WlopurTe JO urorssazduut ayy saat uontsodiuos peur ayy (p sotyooted raddq uvadosng ayy Ur MOLI) are ash amu peuutue Jo sapouu
asayy ‘syuna pauontsod <pood pur ]e7 pasted BNXS uv Jo VON Ippe ayy “ToLutTeeTt ayy UTA SHV pirxa om] Jo LoNIppe ayy Aq pasha, 10 pomatay,, os]P se Ay YJOLUTURLUL
ayy (9 ‘Aaameurr peuurue styiypoayed saddp uvadoing yay ay) UE WOLULUOS ada ap jo ,stodvam,, 10 sauty Jo Jaquunt e Uy ysnomp yonas seay yyoummew ay.L (4 ‘pasreidn
[ta] TIM pure prwavigg Furuvay “YPOUUUBLUEL FO agetut anstpeas ary sem anbyyd ayy to pastout aFeut sip YL (v Buravrdua ourpapeyn eT UNIO suonipuaronnuMayss °6 IWATA

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 67

ID

LE
Ke
Rul MW 4
fom,

of

tay

“

FIGURE 10. Schematic rendition of the bison indicating the added tail, the added zigzags and the multiple
spears or darts.

Anatolia

The Trois Fréres and Pyrenean images have been dated to the Middle Mag-
dalenian (Clottes 1988-89, 1990, 1991, 1993), that is, toward the approaching end of
the European Ice Age, around 13,500-14,000 BP, when the climate had begun to os-
cillate. During one warm phase the Middle Magdalenian culture had dispersed out of
Western Europe, reaching Czechoslovakia and Poland (Marshack 1994b). Its influ-
ence may have extended even to the Russian plain.

During this period the modes of animal image use that we have been describing
were present also in the Middle East.

The Cave of Okiizini

A series of incised pebbles were excavated in mid-twentieth century at the cave
shelters of Okiizini and Karain in Anatolia, Turkey, situated in the foothills of the
Taurus Mountains at the edge of the Mediterranean alluvial plain (KGkten 1958,
1961). The pebbles come from approximately the same period as the Middle Mag-
dalenian images found in Europe, ca. 14-13,000 + BP (Yalinkaya eral. 1995), though
in the Middle East this period is called the ““Epipaleolithic.”

They were initially described by the Turkish excavator, Kékten, and then by
Anati, but they were rarely, thereafter, discussed or studied. This may have been due
to the presumption that a relevant tradition of Paleolithic image and symbol did not
exist in the Middle East, in part because Paleolithic excavations in the Middle East
have been sparse when compared to their profusion in Europe. I discuss two of the
Anatolian pebbles; these and others are also discussed elsewhere (cf. Marshack
199Sc, e).

68 MARSHACK

One of the pebbles had been illustrated by Anati (Figure 11) (Anati 1968, 1972).
My study of the pebble indicates that it has a slightly different shape than was origi-
nally depicted (Figure 12), that the aurochs was accurately drawn (Figure 13) and
looks, in fact, like bovids incised on pebbles in the Romanellian culture of Italy (Vi-
gliardi 1972, 1976).

A microscopic study indicates that the aurochs was the first image engraved on the
pebble; over the bovid a male was incised; he has a bent arm that is thrusting a spear
into the animal's chest. The seeming “buttocks” image depicted in Figure 11 was
found to be acrack in the pebble. Apparently after the hunter was engraved, a series of
spears and darts were incised into and around the animal.

K6kten reported that an incised aurochs had been found ona wall, yet no aurochs
bones were found at the site J-M. Leotard, personal communication; Yalginkaya et al.
1995; Otte et al. 1995: Marshack 1995c,1995e). The site yielded primarily ibex and
wild sheep, though remains of gazelle, deer and hare were also present. The aurochs,
therefore, reminds one of animal images found in the Franco-Cantabrian caves; the
species that were depicted were not necessarily major items of diet. Yet despite a lack
of aurochs bones, the morphology of the animal was known; it was accurately de-
picted, and it was probably hunted, though perhaps ritually in certain seasons. What-
ever its role, the mage of the aurochs could be “killed” and renewed.

My surprise came in study of the aurochs head. The original muzzle was morpho-
logically accurate, but three or four rapidly made, crude muzzles were then incised
over it. A close-up of the muzzle indicates a perfectly formed round eye, to which has
been added a later, second, oval eye: from the first eye there flows a stream of “tears”
(Figure 14). Images of the tear duct in bovid and cervid eyes occur in Magdalenian en-

FIGURE | 1. Okuzini, Anatolia. Epipaleolithic period. The rendition of an incised pebble with a crude bovid
The rendition suggests a female figure in the Magdalenian buttocks style at the right (afte r Anati 1972).

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 69

FiGuRE 12. The Okiizini pebble indicating a different shape for the pebble and the bovid than originally
rendered

Figure 13. Okiizini. Schematic rendition of the engraving indicating the human male thrusting a spear into
the chest of the bovid. A series of schematic weapons were incised into and around the bovid

70 MARSHACK

gravings. The Okiizini composition, apparently accumulated over a period, seems to
represent an evolved “Upper Paleolithic” style of animal rendition and image use.
Given the “complexity” of this engraving, the absence of other examples in Anatolia
may be due to the sparcity of excavation. There are indications that the underlying
modes and traditions found at Okiizini and Karain survived and developed beyond the
Epipaleolithic. In the early Neolithic village of Catal Htiytik, located on the Anatolian
plateau, there are panels containing male hunters painted at different times and with
different paints; the hunters are ritually “killing” or dancing around a huge bull
aurochs in one panel and an antlered stag in another (Mellaart 1967:Figures 61-64.).
The isolated, horned aurochs head was a major cult image at Catal Hityiik. If the ant-
lered stag was “killed” in a seasonal ritual “hunt,” probably near the rutting season,
the aurochs bull may have been ritually hunted or killed ina different season, perhaps
in the spring or summer. On the coastal plain of Israel, while the aurochs was present
and perhaps hunted in small numbers during the Epipaleolithic, it did not become a
symbolic image until the prepottery Neolithic (PPNB) (Bar-Yosef, this volume).

Abstraction and Notation

The most important composition to come from Okiizini occurs on another pebble.
It provides a different kind of comparison with symboling modes in Europe. The peb-
ble contains no animal image but is incised with two different types of abstract, sche-
matic imagery (Figure 15 a,b). Like the aurochs pebble, this pebble was never used as
atool. Yet it is heavily hand-worn and polished along all of its edges, suggesting cura-
tion and handling over a considerable period of time. On one face (Figure I 5a) there
are twelve small circles (slightly triangular) engraved inside a large, encompassing,
circular, containing line. At its lower end the containing line is extended to meet a
“Jadder-like” motif at the right which consists of an accumulation of sets and subsets
of marks made by different points, to different lengths, in different rhythms, with
varying pressures and containing occasional “signs.” A profusion of such “ladder-
like” motifs is accumulated on the other face (Figure 15b).

This second face contains three blocks of tiny marks arranged in horizontal rows:
each block is incised ona slightly different plane of the stone's surface and at 90° to its
neighbor. Analysis of the “ladder-like” accumulations on this face indicates that they,
too, are composed of sets and subsets incised by different points, pressures and
rhythms and they are also interspersed with “signs” and cueing marks (Marshack
1995c). The compositions on the two faces of the pebble were apparently accumu-
lated over a considerable period of time, a length of time perhaps indicated by the
hand polish along the edges of the pebble.

On the first face, within the area containing the twelve circles, other small circles
were later incised, one at a time (Figure 15a). This accumulation of circles and of sets
and subsets on a single artifact represents a mode of symboling different from that
found on the bovid pebble. Yet both pebbles come from the same level, Level IV at
Okiizini, dated at ca. 13,000-12,000 + BP, the period of the Middle Magdalenian in
Western Europe and the early Romanellian in Italy, cultures within which variants of
comparable traditions are found (Vigliardi 1972, 1976). These two pebbles document
a few of the highly evolved, complex symboling' traditions found in the Epipaleo-
lithic of the Middle East. Recent excavations at Okiizini and Karain have begun to en-
large both the symbolic corpus and its variability.

A microscopic analyses of the sets and subsets on the Okiizini pebble has been
published (Marshack 1995c). That analysis documents the presence on the pebble of

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 71

FIGURE 14. Schematic rendition of the bovid head indicating a second eye, tears flowing from the tear duct
in the original eye, and an overmarking of the original finely incised muzzle with a series of crude later
muzzles

a cumulative, nonarithmetical “notation” comparable to nonarithmetical “notations”
found during this period in the late Upper Paleolithic of Europe (Marshack 199 1c;
D'Errico & Cacho 1994). Analysis of the first block (Block A) on the second face, for
instance, documents an accumulation of sets and subsets, incised at different angles,
with different tools and pressures, interspersed with “cueing marks” and “signs”
(Figures 16a, b). Some of this engraving on limestone would have been as visible as
ink on paper.

Nondepictive Symboling in Israel

Comparable accumulations of sets and subsets occur in the Epipaleolithic of Is-
rael. They are found on bone in the early Kebaran at Ohalo II (Rabinovich & Nadel
1994-5), dated to ca. 19,000 BP: on limestone in the late Geometric Kebaran at Urkan
E- Rub II, dated to ca. 15,000 BP (Hovers 1990): and on plaquettes and blocks of
limestone in the Natufian at Hayonim (Belfer-Cohen 1991) at ca. 13,000 to 10,700
BP. They document a tradition that was maintained and developed in this area for al-
most 10,000 years through major changes in climate and hunting-gathering adapta-
tions (Bar-Yosef & Valla 1990; Bar-Yosef & Belfer-Cohen 1991, 1992; Lieberman
1993). It was at the end of this period, during the cold, dry spell of the Younger Dryas
(Dryas II, ca.11-10,000 BP), that farming began (Aurenche & Cauvin 1989; Moore
1989; Bar-Yosef & Belfer-Cohen 1989, 1991; Gebauer & Price 1992). It is my sug-
gestion that a developing tradition of “time-factored” notation may have played a
preparatory role in the shift to farming.

An early Natufian (ca. 12,300 + BP, uncalibrated) fragment of incised limestone
(8.8 cm), was excavated in the 1980's by F. Valla on the Hayonim terrace (Figure
17a). The central engraving consists of two parallel “ladder-like” containing lines

d2 MARSHACK

within which an accumulation of sets and subsets of marks were incised by different
points, at different angles, and with different pressures. It contains 34 unit marks; 30
marks within the central accumulation, divided into subsets, with four tiny marks
added at the end, at far left. Subsidiary sets, engraved at different angles, to different
lengths, and by different points and pressures, were engraved around the central
marking. There are 46 units in this later marking, divided into eight sets of from three
to seven strokes (Figure 17b). A tradition of adding subsidiary sets to an initial, pri-

Ficure 15, Okiizini. Two faces of an incised limestone pebble (1 2.22 cm): a) Face one has a large contain -
ing circle incised with twelve smaller circles within. A number of later small circles were ritually added

within the area at left. The large circle is attached to a row of marks that consists of sets and subsets accu -
mulated in the manner of the rows on Face 2, with internal “cueing signs” and other differentiations among
the marks and sets; b) Face two has three blocks (A-B-C) of horizontal rows that contain a com plex accu-
mulation of sets and subsets of marks. Each block (A-B-C) ends with a “cueing sign of closure” and each
row contains a variable range of cueing signs and differentiations among the sets

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 73

(© aoa)

Arti
HUE LC

AW SumuNQHANou

A 8 c 12) E F G H ' J K

Ficure 16. Okiizini, Anatolia. Epipaleolithic, ca. 14-13,000 BP; a) Linear rendition of all the incised marks
within four rows of block A on the pebble, indicating the subsets made by different points, with different
pressures, to different lengths, at different angles, as well as the “cueing” marks or “signs” on particular unit

strokes. Block A is terminated with an inverted “Y” sign. Blocks B andC have different signs of closure; b)
An exploded rendition of the first 18 marks of row 3 in Block A, indicating the differences in incising

“Cueing” marks are appended to strokes C and K. The subsets are made in different directions and with dif -
ferent tools and pressures. Some strokes are arced to the right, others to the left. This type of variability ex -
ists among all the subsets and rows of Block A, B and C.

mary accumulation is quite common in the early notations (cf, Marshack 198Sb;
199 1a, b).

The Hayonim composition is neither “art” nor “decoration” but a mode of inten-
tional, cumulative marking. The stone was nota practical tool, but was used only for
this engraving. As on the Okiizini pebble, much of the engraving is today too faint to
be seen by the naked eye. For many reasons, then, the engraving would not have been
intended for public “display.” It seems to have been a record kept by one person. A

74 MARSHACK

FIGURE 17, a) Small limestone pebble from the Hayonim Terrace. Early Natufian. It is incised witha cen-
tral “ladder-like” motif with later subsidiary sets added around the central accumulation, b) Schematic line
rendition of the sets of marks incised on a fragment of stone from the Natufian site of Hayonim. The ini -
tial “ladder-like” accumulation of marks is incised by different tools, at different ang les, and with different
pressures. The accumulation of subsidiary of marks were made by different tools Unpublished arti -

fact, excavated by F. Valla

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 75

study of the engraving indicates that the central motif and, apparently, the subsidiary
sets began at the right and were accumulated towards the left; the accumulation could
probably, therefore, be “read” by the maker as a “continuous” sequence.

Natufian “Art”

This pebble had been incised in the Middle East during the period of high artistic,
symboling creativity in the Late Magdalenian of France, /.e., about 13 to 12,000 BP. It
is thus a thousand or so years earlier than the far more complex Grotte du Tai notation
excavated in France (Marshack 1991b). To indicate that this engraving represents a
specialized mode of symboling, rather than a lack of artistic skill, I present some of
the true “art” that was being produced in this period by the Natufians. The stone carv-
ing of asmall human head from EI Wad in Israel has over large eyes and a nose, mouth
and ears (Figure 18). A more abstracted, schematic human head in stone, perhaps
a''spirit" head in which over large eyes are again the dominant feature, comes from
the site of Eynan (Mallaha), in Israel (Figure 19). An ungulate, perhaps a gazelle, is
carved ona bone sickle handle from El Wad (Figure 20a); it is as carefully carved as
ibex kids made during this same period on spear-throwers made of reindeer antler
from the French Magdalenian sites of Bedeilhac and Mas d'Azil (Figure 20b). In each
case an animal decorates a working “bone” tool: in Europe hunting tools, in Israel ce-
real harvesting sickles. There are also phallic images and female images in the Na-
tufian, comparable to those found in the Magdalenian of France. A phallic image is
incised upon a piece of natural flint at El Wad (Figures 21 a, b) (Weinstein-Evron &
Belfer-Cohen 1993): a more realistic, unpublished phallic image, carved of lime-
stone, comes from the site of Fazael; it has no date but it is clearly made in this re-
gional tradition (Figure 22). Comparable phallic i::ages were carved of reindeer
antler during this period in the Magdalenian of France and Spain (Figure 23) (Mar-
shack 1972:330:1991:330) A piece of natural flint from El Wad has the form of fe-
male images made in the so-called “buttocks” style found during the Magdalenian of
Western Europe (Figure 24) (Weinstein-Evron & Belfer-Cohen 1993). A female
identity is suggested by the hip belt incised around the waist. Such hip belts are de-
picted on female images throughout much of Ice Age Europe (Marshack 1991c). The
hip belt would be retained as a primary attribute of female images in the Neolithic of
both Europe and the Middle East.

What is significant is that so many types of image and symboling were shared in
this core area during the same period, despite regional differences in “style” and prob-
able variations in the use and “meaning” of the imagery. Of particular interest is the
fact that the variability of the mid-latitude European Pleistocene fauna and ecology
had provided much of Upper Paleolithic iconographic reference; in the Middle East,
despite symboling traditions of underlying similarity and great complexity, the refer-
ential variability was somewhat different. In the Middle East, for instance, in addition
to a less variable fauna, there was often a tendency to a different order or type of ab-
straction.

The Hayonim Block

In 1994 a large rectangular limestone block (ca. 58 cm long), some 35 to 40
pounds in weight, was excavated by Bar-Yosef from Locus 8, an early Natufian level
(Level B), in the cave of Hayonim (Figure 25). It has not yet been published and
comes from approximately the same period as the pebble found on the Terrace, ca.

76 MARSHACK

FIGURE 18. El Wad. Natufian. Carved head in stone

| 2-13,000 BP. The stone 1s too heavy to have been a mobiliary artifact and so was a
relatively stationary object in the cave. One flat plane of the block is incised with a
very faint, long “ladder-like” motif consisting of sets and subsets of marks made by
different points, in different rhythms, and with strokes of varying length (Figure 26).
[he sets are interspersed with “cueing marks” and “signs,” recalling a type of mark-
ing variability found on the pebble from Okiizini. The scale of marking is, as on the
Okiizini pebble, roughly that of a yardstick or meter rule.

Again, the incising ts so light that it could not be easily seen at midday in the ambi-
ent light within Hayonim cave. It is today mainly visible in strong side light. When
originally incised, however, the engraving would have been visible as stark white,
powdery lines that would have persisted for months. Because of the difficulty in pho-

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 77

FIGURE 19. Eynan (Mallaha). Natufian. Schematic head carved of stone

tographing this faint marking, Figure 29a presents a schematic overlay of the accu-
mulation in black ink, as determined by microscopic study. Given the scale, delicacy,
and structure of the engraving, it is unlikely that this accumulation was intended as
“art,” “decoration,” or for public viewing. Belfer-Cohen (1991) has reported other
“Jadder-like” markings Hayonim cave. Many of these have not yet been published.

Microscopic analysis indicates that the engraving consists of an accumulation of
sets and subsets of from one to seven or eight unit marks in length, similar to the sets
and subsets found on the Hayonim terrace pebble and the pebble from Okiizini in
Anatolia. These sets were made by different points, pressures, rhythms and, in this in-
stance, with strokes of dramatically varying length. The variability and complexity of
the engraving suggests that it could not have been incised rhythmically, that is, “at one
sitting.”

The profusion of accumulated “cueing marks” and “signs” strengthened the prob-
ability that this was a form of record-keeping maintained inside the cave. For reasons

78 MARSHACK

Ficure 20a. El Wad. Natufian Carved gazelle on bone sickle handle

of scale, and because of the linear structure, the accumulation could not have been
“read” arithmetically; but it could probably have been “read” as a developing se-
quence, aided by the position of cueing marks and signs within the developing se-
quence. | assume, therefore, that the cueing signs and subsets served as abstractions

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 79

FiGuRE 20b. Bedeilhac, France. Middle Magdalenian, An ibex kid with inset eye carved as the head of a
reindeer antler spear-thrower.

and references to a visually differentiated sequence of real world, phenomenological
observations. During my study of the marking, I came to sense a possible division or
“chunking” of the accumulation into perceptual “thirds” — one third at the right, one in
the middle, and one at left. This approximate division is presented to indicate the per-
ceptual difference within each (Figures 27a-d). Each presumed “third” notates a “two
month” period and the three groups notate a six month period (60+60+60 =180). A
seventh, extremely faint group of 30+ marks was appended at far left, apparently as a
subsidiary, terminal accumulation (Figure 27d).

What sort of questions can we ask of this composition? Levantine archaeology
documents an increasingly sedentary occupation of major sites or “base-camps”’ by
hunting-gathering groups of the Natufian, accompanied by an increasing collection
and storage of wild cereals (Henry 1981; Bar-Yosef & Belfer-Cohen 1992). This ten-
dency toward sedentism had begun by the early Kebaran ca. 19,000 BP, when there is
evidence for periods of relatively long-term habitation at certain sites and a collection
of wild cereals (Kislev et a/. 1992; Nadel & Hershkovitz 1991;Nadel 1991; Nadel er
al. 1994). Gathering and sedentism increased during the Natufian, and these are gen-
erally recognized as part of the process that led to the beginning of farming, a shift that
occurred when the climate abruptly changed at the last cold snap of the Pleistocene
(Dryas II) about 11 to 10,000 years ago. That change in climate initiated drier condi-
tions, and it apparently placed subsistence pressure on the Natufians, pressures that
induced changes in the planning and scheduling of seasonal activities and the more
intense gathering and storage of cereals.

80 MARSHACK

Ficure 21. a) A piece of natural flint in the shape ifa phallus, from the Natufian of El Wad. It has been in -
cised with an encircling groove to indicate the “foreskin” and a slitat the top; b) close-up of the head of the
flint “phallus” indicating the incised slit

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 81

FIGURE 22. Carved limestone phallus from the site of Fazael in Israel, perhaps post-Natufian, indicating
both the “foreskin” groove and the slit. The tradition is similar to that found in the Natufian

FiGure 23. Montastruc. France. Middle Magdalenian. Carved phallus of reindeer antler from approxi -
mately the same period as the El Wad phallus

Time and Seasonality in the Natufian

In the Natufian, sickles, mortars and pestles were being manufactured for use with
wild cereals and plants. Baskets were being woven (at the proper time) for the gather-
ing, portage, and storage of an estimated, required quantity of wild cereals and plants
to accommodate the needs of increasing sedentism and an increasing population. A

82 MARSHACK

FiGure 24. El Wad. Natufian. Small flint nodule in the shape of a schematic buttocks figure with an incised

“belt”

dry cave such as Hayonim was an excellent location for cereal storage, and there is
evidence for such storage in the presence of domesticated commensals such as mice.
At other Natufian sites elaborate stone structures and storage pits were built. There is
also evidence for skin working, hide polishing, and the manufacture of symbolic arti-
facts of stone and bone. The materials to make beads and amulets were often acquired
in long distance trade, and this suggests that there were also proper seasons or times
for aggregation and exchange. Each of these periodic activities suggests a significant
degree of on-site planning and scheduling.

It is probable that these schedules involved increasingly differentiated, special-
ized roles and activities for women and men at each point in the seasonal sequence.
Activities scheduled at a cave such as Hayonim may have included rituals (perhaps a
ritual request for, or thanks for, the winter rains; perhaps rituals at times of group ag-
gregation following the gathering and storage of cereals, and so forth). With or with-
out notation, there would have developed an increasingly pragmatic, diversely
specialized time-and-space “calendar.” The presence of notation, however, suggests

iES

IMAC

COMPARATIVE STUDY OF PALEOLITHIC

84 MARSHACK

a more formal structuring of the cultural tapes-
try and the probable presence of a “time-
factoring” specialist — a leader, shaman, elder,
or head of a kinship group.

The incised limestone block at Hayonim
may have been maintained by such a specialist,
someone who kept track of the economic and
ritual sequence. Among the probable observa-
tions that would have been made and have been
incorporated within the notation were those re-
cording the variability of the winter rains, a vari-
ability that would have been keenly monitored
with increasing sedentism and an increase in the
gathering of wild plants and cereals. The vari-
ability in the arrival and intensity of the winter
rains, particularly during periods of climatic and
weather oscillation, would also have encour-
aged the development of alternative subsistence
strategies and schedules. The same variability
would probably have invited ritual interven-
tions, whether to encourage rain and plant
growth or to address failures and successes in
the year round.

The limestone block at Hayonim may, there-
fore, document the “sedentary” presence at the
cave of a specialist who marked the stone for
seven months in a particular year. “Sedentary,”
that is, within the probable compass of a few
days walk from the cave, perhaps for gathering
or hunting, or for visits to neighboring groups or
the coast. The presence of small subsets in the
composition, apparently incised at different
times, may record such periods. There would at
such times be an engraving of the number of
days that the engraver was absent (c/. Marshack
1974). Other subsets might record periods of
specialized on-site activity. Though the notation
does not indicate lunar periods, an indication of
the months or “moons” would probably also
have underlain the accumulation. The increase
in sedentism, the need for specialized schedul-
ing, and the presence of notation may, therefore,
allow us to assume that cultural and phenome-
nological “time” was being monitored and read
at a number of levels and in different
ways.

Observations of seasonal changes in the
natural realm, accompanied by adaptive
changes in subsistence behavior, are profusely
documented in the Upper Paleolithic/ Meso-

=o pot 7
LHL |
LEI S

Ya f

j
comer
/

FiGuRE 26. The marking on the limestone
block as determined by microscopic study

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 85

FIGURE 27a-d. Schematic rendition of all the incised marks on the large Hayonim block, divided into four
sections on the basis of the apparent perceptual difference in their marking. The drawings indicate the ir -
regular breakdown of the marking into sets and subsets of long and short lines with occassional “cueing”
marks.

So MARSHACK

lithic of both Europe and the Middle East (Soffer 1985; Price 1978; Price & Brown
1985; Pike-Tay & Bricker 1993; Pike-Tay 1993; Pike-Tay & White 1989; Lieberman
1993). The relevance of such “time-factored” observations and changes in behavior
for our understanding of the referential aspects found in Upper Paleolithic/Mesolithic
image and symbol has been argued by this researcher (Marshack 1972, 1975, 1995d).
The presence of notation suggests, however, that there were probably structural and
conceptual differences in cultural developments that occurred during this period
within this core area of Europe and the Middle East from those occurring elsewhere,
or that were later present among certain groups of remnant hunter-gatherers that sur-
vived into the ethnographic present at the Earth's peripheries ( Lee & deVore 1968;
but see Smith 1991).

“Reading” the Hayonim Notations

How, then, lacking the cultural referents, do we today “read” or interpret the early
Middle East notations? We can, today, analytically study the mode of accumulation,
but a Natufian “reading” of the Hayonim block would probably have involved knowl-
edge of the “time-factored” cultural sequence in the economic-social and the
phenomenological-natural realms. The notation would presumably have recorded
relevant “blocks” or periods of time, as well as happenstance events and aberrant sea-
sonal variations. On the other hand, the small pebble found on the Hayonim terrace
may have recorded a smaller relevant period, perhaps a journey from the cave, or ex-
pectation of a particular event (c/. Marshack 199 1b).

Conclusion

The first two great “revolutions” in complex modern human culture were a) the
widespread explosion of image and symbol in the Upper Paleolithic of Europe and b)
the beginnings of agriculture in a small area of the Middle East immediately after the
Pleistocene. | suggest that both revolutions may have been conceptually related and
that they were probably developmentally and referentially different from processes
that occurred among the hunter-gatherer cultures that persisted into the historic pres-
ent. These suggestions may help explain why the European Upper Paleolithic sym-
bolic cultures ended when the Pleistocene “calendar frame” and its accompanying
conceptual tapestry disintegrated, and help explain the seemingly “sudden” appear-
ance and spread of agriculture in the Middle East on the basis of an equally long, but
contextually different, “time-factored” and “time-factoring” cultural preparation.
These conjoint developments may help explain both the rapid spread of farming in the
Middle East and the relatively rapid adoption of farming among the indigenous
hunter-gatherers of Europe who retained aspects of their own, indigenous “time-
factoring” cultural preparation.

The beginnings of human image making, and the development and “explosion” of
complex symboling traditions in the core area of Europe and the Middle East, present
us with processes of extraordinary analytical and theoretical complexity. A proper
study, it is argued, will have to go beyond mere ethnographic and archeological com-
parison, descriptions of imagery as “art,” a comparison of “styles,” discussions of
Upper Paleolithic imagery as “symbol” or “metaphor,” and a “gendering” of the cor-
pus. It will have to be based on methodological, systematic, direct analyses of the di-
verse symbolic materials and traditions and their development. It would involve, as

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 87

well, an interdisciplinary attempt to determine the nature and range of the “potentially
variable” hominid-to-human problem-solving and symboling capacity, and the range
and variability of that capacity as evident in its early uses.

Acknowledgments

The research in this paper was funded by the American School of Prehistoric Re-
search, of the Peabody Museum of Archaeology and Ethnology, Harvard University,
with special grants from Jean Auel. Thanks are given the Ministry of Monuments and
Museums in Ankara for the permission to conduct research in Turkey, and to I.
Yalcinkaya of the University of Ankara and Marcel Otte of the University of Liege for
their assistance in that research. Thanks are given to the Israel Antiquities Authority
for permission to work with different collections in that country. Particular thanks are
given to O. Bar-Yosef, A. Belfer-Cohen, and N. Goren-Inbar for making published
and unpublished Israeli materials and artifacts available for study and for offering
necessary suggestions and corrections to the manuscript, without altering the author's
analyses and interpretations.

Literature Cited

Altuna, J. & J. M. Appellaniz. 1976. Las Figuras Rupestres Paleoliticas de la Cueva de Altx-
erri (Guipuzcoa). MUNIBE Ano XXVIII, fase. 1-3. Sociedad de Ciencias Aranzadi, San
Sebastian, Spain.

Anati, E. 1968. Anatolia's earliest art. Archaeology 21:21-35.

Anati, E. 1972. Arte Preistorica in Anatolia. Studi Camuni, Vol. 4. Edizioni del Centro, Valca-
monica, Italy.

Aurenche, O. & J. Cauvin. Editors. 1989. Neolithisations: Proche et Moyen Orient, Méditer-
ranée Orientale, Nord de l'Afrique, Europe méridonale. Chine, Amérique du Sud. BAR In-
ternational Series 516. Archaological Series No. 5. Maison de l'Orient Meéditerrancen
(CNRS — Université Lumiére — Lyon 2) Lyon, France.

Ayala, F. 1995. The myth of Eve: Molecular biology and human origins. Science 270:1930-
1936,

Bar-Y osef, O. 1992. The role of western Asia in modern human origins. Phil. Trans. Roy. Soc
London. B 337:193-200.

Bar-Yosef, O. & A. Belfer-Cohen. 1988. The early Upper Paleolithic in Levantine caves
Pages 23-41 in J.F. Hoffecker & C.A. Wolf,ed., Early Upper Paleolithic: Evidence from
Europe and the Near East. British Archaeological Reports International Series 437, Ox-
ford, U.K.

Bar-Yosef, O. & A. Belfer-Cohen. 1989. The origins of sedantism and farming communities in
the Levant. J. World Prehist. 3:447-498

Bar-Y osef, O. & A. Belfer-Cohen. 1991. From sedentary hunter-gathers to territorial farmers
in the Levant. Pages 181-202 in S.A. Gregg, ed., Between Bands and States. Center for Ar-
chaeological Investigations, Occasional Paper No. 9. Carbondale.

Bar-Yosef, O. & A. Belfer-Cohen. 1992. From foraging to farming in the Mediterranean Le-
vant. Pages 21-48 in A.B.Gebauer & T. D.Price, ed., Transition to Agriculture in Prehis-
tory Monographs in World Archaeology No. 4..Prehistory Press, Madison.

Bar-Yosef, O. & F. R. Valla, eds. 1991. The Natufian Culture in the Levant International
Monographs in Prehistory. Archaeological Series I. Ann Arbor.

88 MARSHACK

Belfer-Cohen, A. 1988. The Appearance of Symbolic Expression in the Upper Pleistocene of
the Levant as Compared to Western Europe. L ‘Homme de Neandertal, Vol. 5, La Pensée,
Liége. pp.25-29.

Belfer-Cohen, A. 1991. Art items from layer B, Hayonim Cave: A case study of art in a Na-
tufian context. Pages 569-588 in O. Bar-Yosef & F. R. Valla, ed., The Natufian Culture in
the Levant. International Monographs in Prehistory, Ann Arbor.

Belfer-Cohen, A. & Bar-Yosef, O. 1981. The Aurignacian at Hayonim Cave. Paléorient 7:19-
42.

Belfer-Cohen, A. & N. Goren-Inbar. 1994. Cognition and communication in the Levantine
Lower Paleolithic. World Archaeol. 3:144-157

Bietti, A. 1981. The Mesolithic cultures in Italy: New activities in connection with Upper Pa-
leolithic cultural traditions. Pages 33-50 in Bernard Gramsch, ed., Mesolithikium in Eu-
ropa. Proceedings of an International Symposium, 1978. Verdffentlichungen 14/15. Des
Museums fiir Ur-Und Friihgeschichte:Potsdam, Germany.

Carbonell, E. & Z. Castro-Curel. 1992. Paleolithic wooden artefacts from Abric Romani
(Capellades, Barcelona, Spain). J. Archaeol. Sci, 19:707-719

Clottes, J. 1988-1989, Niaux: Magdalénien moyen ou recent? Ars Praehistorica 7-8:53-67.

1990, The parietal art of the Late Magdalenian. Antiquity 64:527-548.

—. 199]. The Cave of Niaux. Castelet, Boulogne, France.

—__. 1993, Des mains tres anciennes. /ndl. Nwsltr. Rock Art. Vol. 1.

D’ernico, F. 1995. A new model and its implications for the origins of writing: the La Marche
antler revisited. Cambridge Archaeol. J. 5:163-206.

D’ernrico, F_ & C. Cacho. 1994, Notation versus decoration in the Upper Palaeolithic. A case-
study from Tossal de la Roca (Alicante, Spain). J. Archaeol. Sci. 21:185-200.

Feraud, G., D. York, C. Hall and H.P. Schwarex. 1983. War Ar age limit for an Acheulan
site in Israel. Nature 304:263-265.

Gebauer, A.B. & T.D. Price. 1992. Transitions to Agriculture in Prehistory. Monographs in
World Archaeology No 4. Prehistory Press, Madison, Wisconsin.

Goren-Inbar, N. 1986. A figurine from the Acheulian site of Berekhat Ram. Mi 'tekufat
Ha'even 19:7-12.

Goren-Inbar, N. & S. Peltz. 1995. Additional remarks on the Berekhat Ram figurine. Rock Art
Res. 12:131-132

Goren-Inbar, N., ed. 1990. Quneitra: A Mousterian Site on the Golan Heights. Institute of Ar-
chaeology, Jerusalem.

Henry, D. O. 1981. Ananalysis of settlement patterns and adaptive strategies of the Natufian,
Pages 421-431 in J. Cauvin, ed., Préhistoire du Levant: Chronologie et Organization de
l'Espace Depuis les Origines jusqu'au Vle Millenaire. Maison de !'Orient, Lyon, 1980.
Colloques Internationax du C.N.R.S. No. 598. CNRS, Paris, France.

Hovers, E. 1990. Art in the Levantine Epi-Palaeolithic: An engraved pebble from a Kebaran
Site in the Lower Jordan Valley. Curr. Anthropol. 31:317-322.

Hovers, E., L.K. Horwitz, D.E. Bar-Yosef & C. Copoe-Miyashiro. 1988. Subsistence and mo-
bility patterns at the Kebaran site of Urkan-e-Rub Ha, Lower Jordan Valley. Mi 'tekufat
Ha‘even 21:20-48.

Kislev, M-E., D. Nadel & I. Carmi. 1992. Epipalaeolithic (19,000 BP) cereal and fruit diet at
Ohalo II, Sea of Galilee, Israel. Rev. Paleobot. Palvnol. 73:161-166

Kékten, I.K. 1958. Tarsus-Antalya arasi sahil serti tizerinde ve Antalya Bélgesinde yapilan
Tarih6ces: Arastirmalan. 7itirk Arkeoloji Dergist 8:10 ff.

_. 1961. Maras ve Antalya Vilayetleinde siireli Dip Tarih Arastirmalart hakkinda kisa bir
Rapor. Titirk Arkeoloji Dergisi 11:40 ff.
Lee, R.B. & I. Devore. Editors. 1968. Man the Hunter. Aldine, Chicago.

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 89

Leroi-Gourhan, A. 1965. Préhistoire de l'Art Occidental. Mazenod, Paris, France.

Lieberman, D.E. 1993. Variability in hunter-gatherer seasonal mobility in the southern Levant:
From Mousterian to the Natufian. Pages 207-219 in G.L. Peterkin, H.M. Bricker and P.
Mellars,eds., Hunting and Animal Exploitation in the Later Palaeolithic and Mesolithic of
Eurasia. Archeological Papers of the American Anthropological Association # 4.

Lister, A. & P. Bahn. 1994. Mammoths. MacMillan, New York and Marshall, London.

Marks, A.E. 1983. The Middle to Upper Paleolithic transition in the Levant. Pages 51-98 in F.
Wendort & A.E. Close, eds., Advances in World Archaeology, Vol. 2.

Marshack, A. 1969a. New techniques in the analysis and interpretation of Mesolithic notation
and symbolic art. Pages 479-494 in E. Anati, ed., Valcamonica Symposium, 1968, Actes du
Symposium International d'Art Préhistorique. Capo di Ponte, Italy.

. 1969b. Polesini, a reexamination of the engraved Upper Paleolithic mobiliary materi-
als of Italy by a new methodology. Rivista di Scienze Preistoriche 24:219-281.

. 1972. The Roots of Civilization. McGraw Hill, New York.

. 1974. The Chamula calendar board: an internal and comparative analysis. Pages 255-
270 in N. Hammond, ed., Mesoamerica Archaeology. New Approaches. University of
Texas Press, Austin.

. 1975. Exploring the mind of Ice Age man. Natl. Geogr. 147:62-89.

. 1976. Some implications of the Paleolithic symbolic evidence for the origins of lan -
guage. Curr. Anthropol. 17:274-282

. 1977. The meander as a system: The analysis and recognition of iconographic units in
Upper Paleolithic compositions. Pages 286-317 in P. Ucko, ed., Form in Indigenous Art
Schematization in the Art of Aboriginal Australia and Prehistoric Europe. Duckworth,
London, U.K.

. 1981. On Paleolithic ochre and the the early uses of color and symbol. Curr. Anthropol
22:188-191

. 1985. Hierarchical Evolution of the Human Capacity: The Paleolithic Evidence.
Fifty-fourth James Arthur Lecture on “The Evolution of the Human Brain,” 1984. Ameri-
can Museum of Natural History, New York.

. 1989. Evolution of the human capacity: The symbolic evidence. Yrbk. Phys. Anthropol
32:1-34.

.1991a The Roots of Civilization 2nd Ed. Mt. Kisco, Moyer Bell, New York.

. 1991b. The Tai plaque and calendrical notation in the Upper Palaeolithic. Cambridge
Archaeol. J. 1:25-61.

.1991c. The female image: A “time-factored” symbol. A study in style and modes of im-
age use in the European Upper Palaeolithic. Proc. Prehist. Soc.57:17-31.

. 1992. The analytical problems of subjectivity in the maker and user. Pages 181-210 in
T. Shay & J. Clottes, ed., The Limitations of Archaeological Knowledge. Etudes et Recher-

ches Archéologique de |'Université de Liege 49. Université de Liege, Liége, France.

. 1993. Maszycka iconography: A study of the dispersed symbol systems of the Mag -
dalenien anavettes. Pages 205-216 in K. Kozlowski, ed., Maszycka Cave: A Magdalenian
site in Southern Poland. Jahrbuch des Rémisch-Germanischen Zentralmuseums, Vol 40.
Rémischchen-Germanischen Zentralmuseums, Mainz, Germany.

. 1994. Commentary on A.M. Byers “Symboling and the Middle-Upper Palaeolithic
transition.” Curr. Anthropol. 35:386-387.

. 1995a. Methodology and the search for notation. A Review Article on F. d'Errico, 1994,
“|'Art gravé azilien: de la technique a la signification.” Gallia Prehistoire XXX]. Antiquity
69:1049-1051.

. 1995b. Concerning the “geological” explanation of the Berekhat Ram figurine. A re-
sponse to Pelcin. Curr. Anthropol. 35:495.

90 MARSHACK

—__. 1995¢. Variabilité de catégorie dans I'imagerie symbolique d'Okiizini et de Karain
(Turquie). L'Anthropologie 99:586-594.

—__. 1995d. Images of The Ice Age. 1995. Archaeology 40:28-39.

—_. 1995e. The Epi-Palaeolithic of Okiizini cave (SW Anatolia) and its mobiliary art. An-
tiquity 69:93] -944.

—_. 1996a. On the difficulties of “close reading.” A commentary and response to James
Elkins' “On the impossibility of close reading.” Curr. Anthropol. 37(2): np. (In press)
—_. 1996b. La Marche revisited: Good technology, poor science. Cambridge Archaeol. J.

6:np. (In press).

_. 1996c. Figuring the figurines. A commentary and response to Leroy McDermott's

“Self-Representation ... during the European Upper Paleolithic.” Curr. Anthropol. 37:np.

(In press).

—. 1996d. A Middle Paleolithic symbolic composition from the Golan Heights. Curr. An-

thropol. 37:np.(In press).

Maska, Ch., H. Obermaier & H. Breuil. 1912. La statuette de mammouth de Predmost. L‘An-
thropologie 23:273-285.

Mellaart, J. 1967. Catal Hiivtik: A Neolithic Town in Anatolia. McGraw-Hill, New York.

Moore, A.M.T. 1989. The transition from foraging to farming in Southwest Asia: present prob-
lems and future directions. Pages 632-641 in D.R. Harris & G.C. Gillman, eds., Foraging
and Farming: The Evolution of Plant Domestication. One World Archaeology. Routledge,
London, U.K.

Moore, A.M.T. 1994. On seasonal mobility and agriculture in the Levant. Curr. Anthropol
35:48-49.

Nadel, D. 1991. Ohalo I] — The Third Season (1991). J. Isr. Prehist. Soc. 24:158-163

Nadel, D., A. Danin, E. Werner, T. Schick, M.E. Kislev & K. Stewart. 1994. 19,000-Y ear-Old
Twisted Fibers from Ohalo I. Curr. Anthropol. 35:451-458.

Nadel, D. & I. Hershkovitz. 1991 New subsistence data and human remains from the earliest
Levantine Epipalaeolithic. Curr. Anthropol. 32:631-635.

Pelcin, A. 1994. A geological explanation for the Berekhat Ram figurine. Curr. Anthropol.
35:674-675.

Pike-Tay, A. 1993. Hunting in the Upper Périgordian: A matter of strategy or expedi-
ence? Pages 85-99 in H. Knecht, A. Pike-Tay & R. White, ed., Before Lascaux’ The Com-
plex Record of the Early Upper Paleolithic. CRC Press, Boca Raton.

Pike-Tay, A. & H. M. Bricker. 1993. Hunting in the Gravettian: An examination of evidence
from Southwestern France. Pages 127-143 in G.L. Peterkin, G.L., H.M. Bricker & P. Mel-
lars, eds., Hunting and Animal Exploitation in the Later Palaeolithic and Mesolithic of
Eurasia. Archeological Papers of the American Anthropological Association. Number 4.

Pike-Tay, A. & R. White. 1989. Le temps saisonnier au Paleolithique Supérieur. Pages 52-53
in Jean-Pierre Mohen, ed., Le Temps de la Préhistoire, proceedings of the XXIII Congres
Préhistorique de France, Société Préhistorique Frangaise, Edition Archeologia, Paris,
France.

Price, T. D. 1978. Mesolithic settlement systems in the Netherlands. Pages 81-113 in P. Mel-
lars, ed., The Early Postglacial Settlement of Northern Europe. University of Pittsburgh
Press, Pittsburgh.

Price, T. D. & J. A. Brown, eds. 1985. Prehistoric Hunter-gatherers: The Emergence of Cul-
tural Complexity. Academic Press, Orlando.

Rabinovich, R. & D. Nadel. 1994-5. Bone tools from Ohalo Il — A morphological and func -
tional Study. J. /sr. Prehist. Soc.26:32-63.

Smith, E. A. 1991. The current state of hunter-gatherer studies. Curr. Anthropol. 32:72-75.

COMPARATIVE STUDY OF PALEOLITHIC IMAGES 91

Soffer, O. 1985. The Upper Paleolithic of the Central Russian Plain. Academic Press, Or-
lando.

Tixier, J. 1974. Poingon decor du Paleolithique Supérieur a Ksar' Aqil (Liban). Pa/éorient
2:187-192.

Tixier, J. & M. Inizan. 1981. Ksar 'Akil, stratigraphie et ensemble lithiques dans le Paléo-
lithique supérieur. Fouilles 1971-1975. In Préhistoire du Levant. Chronologie et Organiza-
tion de l'Espace depuis les Origines jusqu'au Vle Millénaire. Colloque international du
CNRS, 1980, Paris, France.

Vighardi, A. 1972. Le incisioni su petra romanelliane della Grotta del Cavallo (Uluzzo, Lecce).
Rivista di Scienze Preistoriche 28:57-115.

Vighiardi, A. 1976. L'art mobilier Paléolithique de la provence Méditerranéene. Preétirage, Col-
loque XIV, pp.54-73. UISPP 1996, Nice, France.

Weinstein-Evron, M. & A. Belfer-Cohen. 1993. Natufian figurines from the new excavations
of the El-Wad Cave, Mt. Carmel, Israel. Rock Art Res. 10:102-106.

White, R. 1989a. Visual thinking in the Ice Age. Scient. Amer. 261:92-99

. 1989b. Production complexity and standardization in early Aurignacian bead and pen -
dant manufacture: Evolutionary implications. Pages 360-399 in I. P.Mellars and C.
Stringer, ed., The Human Revolution: Behavioral and Biological Perspectives in the Ori-
gins of Modern Humans. Edinburgh University Press, Edinburgh, U.K.

. 1992. Beyond art: towards an understanding of the origins of material representation in
Europe. Ann. Rev. Anthropol. 31:537-64.

. 1993. A social and technological view of Aurignacian and Castelperronian personal or-
naments in Southwest Europe. Pages 327-357 in V. Cabrera Valdés, ed., E/ Origen del
Hombre Moderno en el Sudoeste de Europa. Ministerio de Educacion y Cienca, Madrid,
Spain.

Yalginkaya, |. ef a/. 1995. Les occupations Tardiglaciares du site d'Oktizini (Sud-oouest de la
Turquie): Résultats préliminaires. L'Anthropologie.99: 562-583

Yalginkaya, |., J-M. Leotard, M. Kartal, M. Otte, O. Bar-Y osef, I. Caimi, A. Gautier, E. Gilot,
J. Kozlowski, D. Lieberman, I. Lopez-Bayon, M. Palikowski, St. Thiebault, V. Ancion, A.
Marshack, M. Patou & A. Barbier, eds.. 1996. Fouilles 4 Okiizini un site Paléolithique Final
au Sud de la Turquie. L'Anthropologie (In press).

i}

93

Substantial Acts: From Materials

to Meaning in Upper Paleolithic
Representation

Randall White

Department of Anthropology
New York University

25 Waverly Place

New York, NY 10003

“What does it mean’? This is at one and the same time the most frequently asked
and the most naively conceived question asked of Upper Paleolithic representations.
No art historian trained in the past two decades would ask this question of a Van Gogh
ora Picasso. Why then have generations of prehistorians treated entire corpora of Pa-
leolithic representations as if a single meaning and motivation lay behind them? Ex-
amples are numerous: Female sculptures as fertility figures: painted and engraved
animals as instruments of hunting magic; geometric signs as markers of ethnic iden-
tity and/or as the painted equivalents of gendered articles of speech.

The worst and most pervasive misconception is not even formulated as a question
but as an assumption embedded in the very term “Paleolithic art” Conkey and I have
loudly and frequently decried the use of the concept “art,” because of its status as an
historical artifact of the later stages of the so-called Western tradition. Indeed, 20th
century usage of the term “art” bears almost no relation to the Latin concept of “ars,”
which integrated the domains that we distinguish as “art” and “savorr-faire.”

Any thorough treatment of the anthropological literature on cultural esthetics
makes clear the wide diversity of cosmologies, philosophies and social contexts that
underpin what we generalize as “art.” I prefer the term “representation,” which has a
wide and theoretically complex usage in anthropology. Thus representations can take
many forms, can have widely differing underlying logics, can be diversely motivated;
and, importantly, many representational media do not even operate in the visual/for-
mal channel.

This understood and accepted, we can forsake a focus on the origins of “art” or
“the arts,” which have enormous, but highly ethnocentric, cultural value. Rather we
can redefine what occurs at the beginning of the Upper Paleolithic as the invention of
material forms of representation (White 1992). Such an approach has the additional
advantage of allowing us to expand outward from graphic imagery to include, for ex-

Beyond Art: Pleistocene nage and Symbol

Memoirs of the Califomia Academy of Sciences
Editors, M. Conkey. O. Soffer, D. Stratmann, & N.G. Jablonski

Number 23, Copyright ©1997

O4 WHITE

ample, the representational domain of personal adornment, a critical area of study in
anthropological analyses of meaning, value and social identity.

It is my position here that by asking, “What does it mean?” or even the somewhat
less anthropologically objectionable, “What purpose does it serve.?” and by conceiv-
ing Upper Paleolithic representations as “art” in our sense, we have prevented a seri-
ous treatment of meaning(s). How then do we reconceptualize our subject matter, our
notion of meaning and, more importantly, how do we operationalize these new con-
ceptions in real archaeological research? My answer to this latter question is to focus,
literally, on how meaningful representations are/were constructed, what I have de-
scribed in the title of this paper as “substantial acts.”

Technological Orientation

The term “technology” has hada very limited and theoretically uninteresting con-
notation in American anthropology for decades (cf, Dobres & Hoffman 1994). Mate-
rialist frameworks such as those of White (1959) and Steward (1955) left relatively
little room for considering tools and techniques as anything more than culturally pre-
scribed means of production. Materialist theory in American anthropology immedi-
ately before and after World War II seems to have operated in ignorance of seminal
theoretical developments in what became European structuralism (Mauss 1936;
Leroi-Gourhan 1943, 1945). My own orientation with respect to technology is de-
scended from that of the late André Leroi-Gourhan, who himself was clearly inspired
by Marcel Mauss’s Techniques du Corps, and whose major theoretical and methodo-
logical works on technology have only begun to be translated into English (Leroi-
Gourhan 1993). Critical to Leroi-Gourhan’s research design was the notion of the
chdines opératoire, which he saw as a conventionalized, learned sequence of techni-
cal operations implicated in all cultural production from the manufacture of stone
tools to the painting of underground cavities to the modern assembly line. These
chains were constituents of culture rather than byproducts.

At the level of individual material production episodes, these c/dines operatoires
are constituted of sequences of applied techniques, which Leroi-Gourhan viewed as
the manipulation of conventional tools by means of habitual, learned gestures.
Viewed more broadly, chdines opératoires can be seen as organized into more en-
compassing technical systems, driven by a limited number of technical principles.
Leroi-Gourhan clearly recognized that these underlying technical systems were/are
never the only ones possible and argued that they emerged from profound cultural
contexts and historical/evolutionary trajectories.

For him, regional variation in underlying technical systems manifested itself in
what archaeologists recognize as style, which he saw as a new and revolutionary kind
of material behavior that replaced the need for regional biological diversification in
humans. A new generation of scholars of technology and material culture (Lechtman
1971: Lemonnier 1983: Schlanger 1994) have built on the foundation constructed by
Leroi-Gourhan (cf, Dobres & Hoffman 1994).

My goal here is to illustrate how detailed observation and experimentation
aimed at understanding the chaine opératoire underlying the construction of material
representations can lead to new insights into the social, economic and ideational con-
texts of the representations themselves.

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 95

Case Study I: Aurignacian I Personal Ornaments

The Sample

European sites attributed to Aurignacian I (roughly 35-30,000 BP) have yielded
an abundance of personal ornaments in the form of beads, pendants, pierced animal
teeth and pierced marine shells (White 1989; Taborin 1993; Hahn 1972, 1986). Else-
where, I have explored in detail raw material choice and fabrication techniques and
the ways in which these varied across the European landscape at the very beginning of
the Upper Paleolithic. It has been and it remains my position that this large corpus of
personal ornaments, following a near absence in the preceding Mousterian (White
1995a), is reflective of significant transformations in human society with the onset of
the Aurignacian sometime prior to 35,000 years ago. In light of what we know gener-
ally about the social context of modern human bodily adornment, it seems reasonable
to hypothesize that the first appearance in the archaeological record of large and var-
ied assemblages of personal ornaments implies the material construction and repre-
sentation of a diversity of individual and social identities.

A Schematic Chdine Opératoire for Aurignacian I Personal Ornaments

A schematic outline of the constituents of a chdine opératoire for Aurignacian (or
any other) system of personal ornamentation might look something like the follow-
ing:

1) Cultural assumptions about personhood.

2) Beliefs about the relationship between materials, representational acts, repre-

sentational constructs and social/supernatural efficacy.

3) Choice and acquisition (by direct extraction or by social mechanisms of ex-

change) of raw material based on above.

4) Choice of forms, textures, colors or subject matters.

5) Organization of production, (social, temporal and spatial).

6) Combination of gestures and tools into techniques for ornament production that

are coherent with the encompassing, regional technical systems and that enable

7) Representation of desired signifiers of social identity, age, reproductive

status supernatural associations, etc.

8) Use of the ornamental representations in (socially, esthetically and cosmologi-

cally) meaningful acts.

9) Purposeful (based on intentions for future retrieval/use or ideas about their re-

sidual power and efficacy) or accidental (as a byproduct of human activity) dis-

posal of the ornaments.

It is important to emphasize that, procedurally, archaeologists tend to work in re-
verse through these chains; beginning with basic pattern recognition concerning the
distribution of objects in the ground and moving backward through multifaceted,
higher order analyses and inferences.

Formal Attributes of Beads

By far the majority of preserved personal ornaments in European Aurignacian-
age sites (White 1995b) were fabricated of ivory. Indeed, many of the famous Vogel-
herd statuettes were pierced for suspension. In the Western European Aurignacian,

96 WHITE

facsimiles of red deer canines and marine shells (Figure |) were executed in ivory (ex-
amples are also known in tale and limestone). In addition, there 1s a variety of more-
or-less idiosyncratic “pendants” manufactured of ivory and, less often, of bone, antler
or tale. For reasons of space, I shall focus here on operational sequences for the mass
production of ivory “beads,” giving less detailed treatment to rarer ivory ornaments.

Two formal attributes of Aurignacian | beads and pendants are particularly note-
worthy: surface luster and facsimile. First, texturally, almost all Aurignacian I beads
and pendants exhibit remarkable surface luster that is not a product of postdeposi-
tional processes. The luster on ivory beads (Figure 2) was intentionally produced by
techniques (discussed later in this paper) of grinding and polishing, especially the use
of an effective metallic abrasive, powdered ocher (hematite).

The potential of ivory to take on lustrous polishes can be seen as having been ex-
ploited through the creation and application of appropriate techniques for replicating
the naturally occurring tactile characteristics of other ornamental media, such as
mother of pearl, talc/steatite and dental enamel. In other words, the polishing of ivory
was itself a representation of textures experienced elsewhere in the natural world.
Our own predominant cultural medium of representation is visual: as a result we have
tended to treat tactile characteristics of objects as nonrepresentational or minimally
representational. However, a moment of reflection will reveal that, even in the so-
called “Western Tradition,” polished ivory’s tactile qualities have been so sought-
after that elephants have been brought to the verge of extinction.

The hypothesis that the polishing of ivory was an attempt to represent the surface
qualities of other substances of tactile interest is supported by the existence in Aurig-
nacian I sites of ivory facsimiles of seashells (mother-of-pearl), and animal teeth
(dental enamel and naturally lustrous dentine). Indeed, the basket-shaped ivory and

BOO RRGERORRNEROOREORRHG ARGH ARREE BRU
INES % CSRS ew ARCS

\

Figure |. Three of six facsimiles of seashells sculpted in ivory from the Aurignacian | siteofLa Souquette,

France

MATERIALS AND MEANING IN PALEOLITHIC IMAGES o7

talc beads so frequent in SW French Aurignacian I sites (see below) show a striking
resemblance in form and size (size standardization is discussed below and has been
quantified in White 1989) to a species of Mediterranean seashell found in some
Aurignacian I sites in SE France: Cyclope neritea (Figure 3). In light of this hypothe-
sized tactile constituent of Aurignacian I representation, it is perhaps not surprising to
observe that more than 95% of Aurignacian I personal ornaments are constructed of
ivory, talc, shell or animal tooth.

Ficure 2. Ivory basket-shaped bead from Abri Blanchard, France, showing characteristic luster. Scale in

mm

98 WHITE
Regional Perspectives

The fabrication sequence (a mere fragment of the much broader and inclusive
chaine opératoire outlined above) for Aurignacian ivory and stone beads varies both
intra- and interregionally. In France, the most common form, represented by more
than 1000 specimens, is what has been called basket-shaped beads. Found in large
quantities early in this century at Abri Blanchard, Abri Castanet, Abri de la Souquette,
Isturitz and Saint Jean de Verges, these have now been radiocarbon dated by Delporte
at Brassempouy (Delporte & Buisson 1990) to between 33- and 32,000 years BP.
They were created (Figure 4) from pencil-like rods of ivory or talc that were then cir-
cuminscribed and snapped into cylindrical blanks one to two centimeters long. These
were then bilaterally thinned at one end to form a sort of stem. A perforation was then
created at the junction of the stem and the unaltered end. This was done by gouging
from each side, rather than by rotational drilling. These rough-outs were then ground
and polished into their final basket-shaped form using, first coarse abrasives and then
fine metallic abrasives (powdered hematite) as abrasive. My experiments indicate
that a mean of one to two hours of labor per ivory bead and 30 minutes per talc bead
are required by this process.

Ivory beads in south German Aurignacian sites, also radiocarbon dated to be-
tween 32- and 33,000 years BP, are substantially different, although the basic princi-
ple of reducing an ivory baton was the same (Figure 5). In the case of GeiBenklésterle
for example (Hahn 1986), a baton elliptical in section was circumincised and
snapped. The blank was then thinned and perforated by gouging. In this case, how-
ever. two holes separated by a bulge were dug into the blank. This type of bead is as
unknown in France as the basket-shaped form is in Germany.

Raw Material Choice and Acquisition

The full range of ornamental raw materials preserved in the record includes vari-
ous mineral and animal substances including limestone, schist, tale-schist, tale, mam-
malian teeth, bone, antler and ivory, fossil and contemporary species of marine and
freshwater shells, fossil coral, fossil belemnite, jet, lignite, hematite, and pyrite. How-
ever, this relatively extensive list should not be taken to suggest a kind of random use
of materials encountered in the environment. A number of pronounced choices were
made.

If one examines the species of animals whose teeth were chosen for Aurignacian
objects of suspension, there are clear choices that vary somewhat regionally. In
France, Belgium, Germany and Russia fox canine teeth predominate followed in
much smaller quantities by the vestigial canines of cervids, almost always red deer. In
Spain and Italy, almost all pierced teeth are vestigial canines of red deer, with fox ca-
nines being absent. At Mladec in Czechoslovakia, beaver incisors dominate, closely
followed by moose and bovid incisors. In nearly all instances the species sample for
objects of suspension is fundamentally different from that found in the food debris,
suggesting choices based on ideas exclusive of dietary preferences.

Pierced marine, freshwater and fossil shells (Taborin 1993) constitute perhaps a
third of the objects of suspension found in French Aurignacian sites. Outside of
France, with the possible exception of Northern Italy (Bartolomei ef a/. 1992), they
are so rare as to be considered virtually absent in Aurignacian-age sites in Spain, Bel-
gium, Germany, Czechoslovakia, Hungary and Russia. Local marine fossils (belem-

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 99

FIGURE 3. A specimen of Cyclope neritea, closely resembling many of the Aurignacian | basket-shaped

beads. Length 6.5 mm. (Courtesy of Y. Taborin)

nite, coral) served as raw materials for objects of suspension in some
Aurignacian-age sites in Russia (White 1993).

Occasional pierced objects in limestone are found throughout Europe. On the
other hand, rarer soft stones, such as talc, lignite and hematite were turned into
Aurignacian-age items of suspension most often in southern France, but also occa-
sionally in Spain, Italy, and Germany.

Personal ornaments are frequently manufactured of materials exotic to the
re-gions in which they are found. This is especially true of shells and rare minerals,
but may also be true of ivory, since mammoth remains are virtually absent from
Aurignacian I sites (Delpech 1983) in SW France (although the collection by Aurig-
nacian people of subfossil ivory from geological sources is a distinct possibility; see
below). In general, rare minerals in France fall off with distance from point of natural
origin. For example, talc attenuates as one proceeds north from the Pyrenees. Like-

100 WHITE

wise Atlantic and Mediterranean shell species attenuate as one proceeds into the
French interior.

Contribution of Raw Materials to Representational Form:
Mammoth-Ivory as a Raw Material

Proboscidean tusks are merely specialized teeth (upper incisors), and are
composed predominately of dentine. In contrast to many other ivories, mammoth and
elephant tusks have a complex structure (Figure 6) formed by specialized cells known
as odontoblasts. These odontoblasts produce new dentine along the lining of the pulp
cavity. As new dentine is produced, the odontoblasts migrate to the new surface of the

680 4a Gas
. hy
| O40 —

Randal) White 1993

Figure 4. The production sequence for Aurignacian | basket-shaped beads

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 10]

pulp cavity, leaving behind processes known as dentinal tubules (MacGregor 1985)
or so-called Hunter-Schreger bands. These “radiate outward from the pulp cavity and
incline obliquely towards the tip” of the tusk. The result, in transverse section, is a
complex three-dimensional structure (Figure 7) that takes the form of arcs running
counter to each other across the width of the tusk, and producing by their intersection
what are frequently referred to as “engine-turnings” of the Hunter-Schreger pattern.

Finished Product
Vv

FiGure 5. The production sequence for south German Aurignacian | two-holed beads (after Hahn 1986)

102 WHITE

In transverse section, this Hunter-Schreger pattern crosscuts broad concentric
rings often referred to as laminae (Figure 8). These are actually growth interfaces be-
tween superimposed cones of ivory. In living tusk, the boundaries between these
laminae are not areas of structural weakness, being bound together by the complex in-
tersecting dentinal tubules composing the Hunter-Schreger lines and by collagen fi-
bers that fill the interstices between the dental tubules. However. upon desiccation
and accompanying deterioration of dentinal collagen, tusks tend to split or spall along
these concentric boundaries, which are far more developed in the outer zones of the
tusk’s diameter. The inner core of the tusk, in the area surrounding the central nerve
canal, is highly compact, homogeneous and virtually immune to such spalling. As a
result, such inner ivory 1s exceedingly difficult to work with stone tools. It is perhaps
no surprise then that in ivory-rich Aurignacian sites in SW France, this “core ivory” is
abundantly represented in the waste products of tusk reduction.

Ivory has properties (color, luster, softness or warmth of touch) that set it apart
from such media as bone and antler. However, these qualities can only be realized by
means of polishing with fine abrasives. According to Ritchie (1969), modern ivory
workers prefer fine metallic abrasives including jeweler’s rouge, which is nothing
more than hematite/red ocher. Indeed, SEM analysis of Aurignacian I ivory beads
(White 1995b) revealed particles of red ocher embedded in the fine polishing striae on
their surfaces. Moreover, large caches of red ocher have been recovered from two of
the richest ivory-bearing Aurignacian sites in SW France: Abri Blanchard and Abri
Castanet (Didon 1911; Peyrony 1935).

Tools, Techniques and Gestures: Experimental Perspectives

My experimental research into the working of ivory falsified several naive
misconceptions about proboscidean tusks as raw material. My first mistake was to
presume that modern A frican elephant tusk was a suitable experimental analogue for
woolly mammoth tusk. Although, superficially, the differences do not appear great,
the two forms of ivory exhibit quite different fracture patterns that are the product of
significant differences in the angle of intersection of Hunter-Schreger bands (known
as Hunter-Schreger angles).

Contrary to my initial presumption, fresh tusks proved extremely difficult to break
into, and to reduce to usable parts. Fresh ivory does not fracture along the large- scale
concentric laminae observable in prepared sections, as we had so naively imagined.
Nor can a fresh elephant tusk be sectioned longitudinally by beginning a split at the
thin, hollow proximal end (pulp cavity) and propagating it toward the distal end of the
tusk.

Neither Alaskan permafrost ivory collected in the 1920's, nor fresh African
elephant ivory could be worked effectively by direct percussion. Attempts at direct
percussion produced the same results that one might expect of hitting a stout piece of
hardwood with a hammerstone! Indeed, there are distinct similarities in the response
of hardwood and ivory to percussion and wedging, based upon similarities in struc-
ture and grain. Sizable flakes were removed by percussion only where the tusk had al-
ready been partially split by desiccation.

Neither fresh nor artificially desiccated (kiln-dried) African elephant ivory could
be worked by splitting-and-wedging. Upon desiccation, elephant ivory developed in-
cipient concentric fractures that conformed to the broad internal laminae of the tusk.
However, these were too poorly developed to serve as points of access for wedges.

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 103

Mammoth Tusk Morphology

Pulp Cavity

Dentine

Cementum

Enamel

FIGURE 6. The morphology of proboscidean tusks

104 WHITE

FiGure 7. The intersecting Hunter-Schreger bands seen here in transverse section in a late Pleistocene
mammoth tusk. A single desiccation fracture is seen to crosscut the bands

FIGURE 8. Concentric desiccation laminae, seen in transverse section ina late Pleistocene mammoth tusk
Note the radial fractures crosscutting the concentric desiccation fractures.

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 105

Overheating in a kiln made modern elephant ivory so brittle that it was breakable by
hand, and structurally weaker than many 35,000 year-old Aurignacian artifacts.

Atthe time I received it, the medial segment of Alaskan permafrost mammoth tusk
showed a quite developed concentric fracture at a lamellar boundary near its external
surface. Attempts to exploit this desiccation fracture by percussion-driven wedging
from the distal extremity of this medial segment produced broad flakes that hinged
outward part way down the segment’s length. However, a split could not be propa-
gated through the entire length of the segment, presumably because the lines of force
were directed outward when they encountered the interfaces between “superimposed
cones” of ivory. With enough labor however, it is certainly possible to scrape, grind
and polish such large flakes of fresh ivory into desired forms. Thus, theiremployment
by Aurignacian ivory-workers cannot be entirely excluded.

In addition, the slightly desiccated Alaskan permafrost tusk showed radial
fractures that crosscut the concentric laminar fractures (Figure 8). While these radial
fractures did not provide an effective purchase for percussion-driven wedges to split
the tusk longitudinally, this 1s probably attributable to the extremely “fresh” state of
the permafrost tusk. Similar attempts with more highly desiccated lamellar fragments
of mammoth tusk led to considerable success in creating long, workable splinters by
longitudinal splitting and wedging.

Splitting-and-wedging constituted a fundamental Aurignacian strategy (Knecht
1993) for working organic materials, but I conclude that they are/were not especially
appropriate to the structure of fresh or even slightly desiccated proboscidean tusks.
The use of subfossil tusks as raw material has been hypothesized previously by Hahn
(1986) for the German Aurignacian and by Phillipov (1983) for the Upper Paleolithic
of the Russian Plain. I think it nearly certain that, apart from some form of artificial
drying or purposeful curing of fresh tusks, “subfossil” or at least mammoth tusks
some years old were sought as raw material by Aurignacian ivory workers.

Even then, the production of long rods of ivory by splitting and wedging of tusks
is not suggested by the structure of ivory, and Is difficult to achieve. These rods seem
to have been a preexisting goal of Aurignacian ivory workers even if they required the
solution of significant mechanical problems. My best explanation for this determina-
tion by Aurignacian I ivory workers to obtain pencil-like rods of ivory is that it is a
technique highly conducive to standardization. Cylindrical rods of roughly the same
diameter lead easily to the production of standardized bead-blanks and ultimately
standardized beads. I have shown elsewhere (White 1989) the high degree of stan-
dardization in Aurignacian beads, a standardization that I believe to be motivated by
the anticipated arrangement of uniform, sewn-on beads on garments. In other words,
the visual impact of Aurignacian beads was not as individual objects, but as arrange-
ments constituted of uniform elements.

Laboratory experiments using Aurignacian blades of Bergerac flint, a fine
limestone grinding stone, powdered red ocher and liberal applications of water (es-
sential to softening and lubricating the surface of the ivory), produced stigmata and
highly polished surfaces similar to those observed on Aurignacian ivory beads. Sig-
nificantly, while most Aurignacian beads show traces of hematite, few if any of them
are profoundly stained. In my experience working permafrost mammoth tusk, indeli-
ble staining only occurred when powdered hematite was mixed with fat or oil. The su-
perficial nature of hematite deposits on Aurignacian beads supports the use of water
rather than fat as a softening/lubricating agent.

In my experience, there is no archaeological evidence before Sungir (dating to ca.
28-25,000 years BP or earlier) for the preparation of whole tusks by softening (heat-

106 WHITE

ing, boiling), which would have allowed the ivory to have been more easily worked.
Even at Sungir however, the operational chain for bead production (White 1993)
seems not to have employed thoroughly softened ivory. Simple soaking of tusks in
water has only superficial effects. However, once the tusk is reduced to much thinner
fragments, such soaking can penetrate the entire thickness, making drilling, scraping
and gouging much easier. This soaking also works well on subfossil tusk fragments.
Comparison of our experimental sample with actual Aurignacian production debris
indicates clearly the use of water in the final stages of bead production.

In my experience with extant archaeological collections, there is no evidence in
the Aurignacian for the oft-cited “groove-and-splinter” technique, so common in
later periods. While grooving and splintering of fresh tusks is a feasible (if enor-
mously tedious and time-consuming) approach, the splinters that existin Aurignacian
assemblages show no traces of having been incised out of the surface of the tusk. The
French sites that have produced the greatest quantities of 1vory ornaments and pro-
duction debris have yielded few significant tusk segments that might yield insight
into Aurignacian approaches to tusk reduction.

Contexts of Production, Use and Disposal

Lacking direct association between Aurignacian I beads and human skeletons, we
are obliged to design research strategies to demonstrate first, that these were indeed
objects of suspension and second, how precisely, they were suspended. A combined
SEM/experimental replication program has already yielded to us significant insight
into the attachment of Aurignacian basket-shaped beads. Sewn-on clothing beads are
implied, but there is substantial variation in patterns of wear (Figure 9) on the surfaces
of bead-holes.

Elsewhere, I have proposed that sites where beads were manufactured in quantity
were special places on the landscape, perhaps loci of aggregations of otherwise dis-
tant groups. This would explain the hyperabundance of exotic raw materials in these
sites. | have also noted that adjacent sites show abundant but differing frequencies of
bead production stages, perhaps indicating more complicated divisions of labor than
previously imagined. There is substantial evidence that beads were worked down toa
penultimate stage (e.g., perforated but not shaped by grinding/polishing, stored as
such, and only finished at the moment at which they were to be sewn or strung). This
may have allowed the ivory worker the flexibility to create a final product tailored to
the size and form requirements of the moment.

More than 25% of Aurignacian I basket-shaped beads show prehistoric breaks,
usually across the neck of the hole. This could easily be explained by beads broken off
a garment during activity, or by the mending of garments at which time worn or bro-
ken beads may have been removed and discarded. However, if use and maintenance
of beaded garments are suggested, so is intensive production of beads manifest in
large quantities of fabrication debris. Caching of large numbers of beads, or even of
whole, decorated garments, cannot be excluded and might explain the high percent-
age of intact, unbroken beads. Unfortunately, excavation techniques early in this cen-
tury did not involve spatial proveniencing of artifacts; thus, spatial analysis within
sites 1s impossible for the moment.

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 107

FIGURE 9. Wear facets on the inside of the holes of tale basket-shaped beads trom Abri Blanchard, France
(Scale in mm)

Case Study II: Gravettian Anthropomorphic Figurines

My goal in this case study is to illustrate how detailed observation and experi-
men-tation aimed at understanding Upper Paleolithic female image production can
lead to new insights into the social, economic and ideational contexts of the objects
themselves. In so doing, I seek to reemphasize (see White 1996b) and avoid some of
the dangers and misunderstandings evident in recent writing on the subject.

The Sample

I am concerned here with anthropomorphic representations, especially sculpted
figurines, traditionally attributed to the Gravettian culture 28-22,000 years BP. In
contrast to recent American writers on the subject (Rice 1981; Nelson 1990), Ido not
consider it acceptable to treat all of Upper Paleolithic female imagery as if it were a
coherent whole. It is not! The Magdalenian and the Gravettian are worlds apart in
terms of the diversity, form and context of female representations. However, Bisson
and others (1996) indicated that so-called Gravettian figurines from Grimaldi cover a
much longer chronological range. Dating problems notwithstanding, the proportion
of decidedly pregnant female images differs dramatically between these two cultures
(Duhard 1993).

My geographic scope over the past few years has been from the Atlantic sea-
board of France to the Don Valley of European Russia. I have examined, with varying
degrees of precision, approximately 100 of these figurines, including those from Gri-
maldi, Savignano, Sireuil, Tursac, Abri Pataud, Trou Magrite, Willendorf, Brassem-
pouy, Lespugue, Dolni Vestonice, Predmosti, Avdeevo, Kostienki 1,and Gagarino.

108 WHITE

Lying behind the previous lack of recognition of differences between the Gravet-
tian and the Magdalenian are serious terminological problems that mask significant
differences in the technology of representation between the two cultures. In particu-
lar, the term “Venus,” which is interpretive rather than descriptive, has been used in
such an all encompassing way as to give the illusion that the repertoire of female im-
ages in the Magdalenian is simply a continuation of that of the Gravettian. Nothing
could be further from the truth.

My position is that descriptive terminology should be based in representational
techniques, rather than in presumed but undemonstrated function, meaning or view-
ing context. Even the terms statuette and figurine probably presume too much. Terms
such as female bas-relief in limestone, or sculpted female in ivory or woman molded
and fired in loess are preferable, in that they provide an understanding of raw mate-
rial, technique and subject, without embedding meaning in description.

Female Representations as an Analytical Category

A question never asked 1s whether female representations constitute a discrete,
natural category of analysis. I find it troubling that associated imagery, and its rela-
tionship to female representations is seldom examined. It is noteworthy that in every
region in which female figurines exist in the Gravettian there are accompanying rep-
resentations of other subjects, from animal to geometric; and these vary greatly by re-
gion. To my knowledge, the only scholar to have addressed this question with respect
to female figurines is Henri Delporte, who notes an association in Siberia between 1m-
ages of women and birds: and in France between women and bison.

Chadine Operatoire for Gravettian Female Sculptures

For our purposes, the general constituents of a chaine opératoire for female figu-
rines might include the following, keeping in mind as previously stated that archae-
ologists usually address the components in reverse order:

1) Cultural assumptions about women.

2) Beliefs about the relationship between materials, representational acts, re-

pre-sentational constructs and social/supernatural efficacy.

3) Choice and acquisition of raw material.

4) Choice of subject matter.

5) Organization of production, (social, temporal and spatial).

6) Combination of gestures and tools into techniques for figurine production that

are coherent with the encompassing, regional technical system and that enable

7) Representation of desired signifiers of social identity, age, reproductive status,

supernatural associations, efc.

8) Use of the female representations in (socially, esthetically and cosmologically)

meaningful acts.

9) Disposition of the figurines based on ideas about their residual power and

efficacy.
The Grimaldi Subsample

I wish now to illustrate the insight that can be gained by analyzing in detail a group
of figurines with the above operational scheme in mind. I will focus in particular on

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 109

15 anthropomorphic figurines from Grimaldi (Liguria), Italy, seven of which had
never been analyzed or published (Bisson & Bolduc 1994; Bisson & White 1996). In
concluding, I will attempt to place these in a broader European context with respect to
the technology of figurine production, use and disposition.

The 15 objects (Figure 10) in question were recovered from the sites of Barma
Grande and the Grotte du Prince between 1883 and 1895 by Louis Jullien. Presently,
seven of them are in the Piette Collection at the Musée des Antiquités Nationales at
Saint Germain-en-Lave outside Paris, one of them is in the Peabody Museum at Har-
vard, and the remaining seven are in private hands in Montreal, where Jullien emi-
grated in 1898.

Raw Materials

The sample of 15 figurines breaks down as in Table 1:
Table |. The raw material breakdown for the 15 Grimaldi sculptures.

—_ Raw Material _ “Number of Figurines

Fossil ivory

Bone or subfossil ivory

Light green fibrous serpentine
Dark green fibrous serpentine
Dark green chlorite

Yellow talc

—fUNN—

All of the soft stone materials are from relatively local alpine sources, that is,
within 100 km from the Grimaldi sites. Hardnesses vary from | to 4 on the Mohs
scale. Several other objects from the Grimaldi sites, especially beads and pendants,
are manufactured in similar soft stones. Ivory and soft stone are the same materials
that dominate the sample of Gravettian female images across Europe. Although the
use of these silicates, such as talc, steatite and serpentine, can be explained by their
being relatively easy to work, they also have remarkable tactile qualities. Especially
when polished, they are indistinguishable to the touch from the warm lustrous sensa-
tion of polished ivory. I propose that, as in the case of Aurignacian beads discussed
above, in our pursuit of the meaning of Gravettian female representations, we have
focused too little attention on carefully chosen and constructed tactile qualities that
had/have enormous evocative potential.

The ivory, constituted of unmineralized to partially mineralized mammoth tusk,
was probably obtained from known geological exposures (Giraudi 1981, 1983).
Mammoths were not present in Italy during the Late Pleistocene, although a small
number of ivory objects, notably pendants, have been recovered from Upper Paleo-
lithic sites in Northern Italy. Exchange with groups farther north is certainly a possi-
bility.

Certain raw materials are conspicuously absent, although they are key constitu-
ents of organic and lithic tools such as spear points, awls, lamps etc. Notably, bone,
antler and limestone are absent. This severe choice may have nothing to do with
workability, as some of the substances ignored are more plastic than those chosen for
figurine production. Raw material choice may have responded to the pursuit of visual
and tactile qualities, and to the cosmology surrounding certain substances.

110 WHITE

IGURE |0a-d. Fourteen of the anthropomorphic figurines from Grimaldi, Italy

1]

MATERIALS AND MEANING IN PALEOLITHIC IMAGES

FIGURE 10e-g

112 WHITE

FIGURE 1Oh-k

MATERIALS AND MEANING IN PALEOLITHIC IMAGES

Figure 10 |-n

114 WHITE
Contribution of Raw Materials to Object Form

The form in which raw material was obtained seems to have influenced in some
cases at least the size and form of the finished figurines. Octobon describes an un-
worked serpentine pebble, oval in form and flattened in section like a skipping stone.
This shape of support accounts for the flattened shape of the Femme au cou perfore, le
buste, la figurine aplatie, le figure and la double figurine. For example, close exami-
nation of la double figurine (Figure 10a) reveals that the fissure separating the human
and the animal figures was a natural flaw in the stone that conditioned the design of
the piece. The minuscule scale of the Grimaldi specimens would be explained 1f sup-
ports for the remaining figurines were as small as the pebble described by Octobon.

Formal Attributes of the Grimaldi Figures

All of the 15 sculpted representations from Grimaldi appear anthropomorphic.
The overall inventory with sexual/reproductive attribution shows an obvious pre-
dominance of pregnant women among complete figurines (Table 2). Three of the 15
have two faces sharing the same body. In two cases these are Human heads back to
back, while a third juxtaposes a human head with that ofan animal. The three cases of
two-faced figurines 1s, so far as Lam aware, unique to Grimaldi. It is also worth noting
that the several cases of blank visages are very explicit; that 1s, faces are not simply
absent, empty faces are very carefully constructed.

lable 2. Sex and reproductive status attributions for the 15 Grimaldi sculptures

Specimen Sex Status
La Tete negroide (broken) 9 9
Le Polichinelle I

pregnant
La Figurine non decrite (broken) a 4
1 ‘Hermaphrodite I pregnant
Le Losange I pregnant
La Femme au goure I pregnant
La Statuette en steatite jaune I pregnant
La Femme au cou perfore ft
Le buste (broken) |
Le Figure (possibly animal) 9
La Figurine en Wwoire brun
La Figurine aplatte

pregnant
»
pregnant
9

La Figurine doubk

pregnant
La Femelle a deux tetes

|
I

La Figurine en ivoire a locre rouge I pregnant
I

I pregnant

Tools, Techniques and Gestures

Microscopic and experimental analysis reveal characteristic stigmata of tech-
niques of gouging, scraping, abrading and incising in the production of the Grimaldi
figurines (Figure 11). The redundancy with which these techniques are applied con-
tributes to a remarkable consistency in breast and abdominal form, as well as hair
treatment. lechnical differences between ivory and soft stone (chlorite/steatite/ser-
pentine) sculpting at Grimaldi and more-or-less contemporaneous Russian sites
(where ivory and marl predominate) could not be clearer. The Russian female figures
show a predominance of gouging and especially abrasion allowing for much
smoother contours than the highly incised and much more angular Grimaldi women.

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 115

However, the forming of the figurines is only part of the story. There is also clear
evidence for burnishing, polishing and, seemingly, glazing (Figure 12). These latter
techniques produce visual and, once again, tactile qualities not producable by stone
tools alone. Both polished ivory and polished/glazed serpentine have very special tac-
tile qualities that may speak to one of the sensory means by which the figurines were
experienced (White 1996b). Moreover, the process of heating talc dramatically deep-
ens its raw color throughout its thickness.

Technological Perspectives on Hair and Clothing

What technologies are codified in representation (e.g., hairstyles, clothing, etc.:
cf. Dobres 1992) and how can careful technological analysis inform such interpreta-
tions (e.g., string aprons, hair nets)? The Grimaldi figurines are essentially silent on
this point, in contrast to many of the Russian figurines, which make explicit reference
to clothing and adornment. There is nothing in any of the Grimaldi objects to contra-
dict the sense that nude women were intended.

Other Subjects of Representation

Although there are provenience problems, a small number of nonhuman repre-
sentations occur at the Grimaldi sites. These are essentially pierced talc objects with
patterned incisions similar to those found on at least two of the female figurines.

Contexts of Use and Disposal

Eight of the 12 unbroken figurines from Grimaldi are perforated for suspension
(Figure 13) and others such as /a figurine en ivoire brun, have carved furrows sug ges-
tive of suspension. Of course, suspension can take many forms in addition to the
wearing of these objects on the body. For example, they may have been suspended in-
side dwellings attached to articles such as skin bags and baskets.

Like many of the Gravettian figurines recovered to date, the Grimaldi speci-mens
were found (to the best of our knowledge) carefully placed in an area peripheral to in-
tense human occupation. They come from two sites, the Grotte du Prince and Barma
Grande. While those from the Barma Grande were recovered from occupational hori-
zons, those from the Grotte du Prince were found in a small niche adjacent to the main
cave. Michael Bisson and I are working on comparisons of sediment on the figurines
with that still adhering to artifacts of known stratigraphic provenience. More interpre-
tively, we have proposed (Bisson & White 1996) particular use contexts for the Gri-
maldi figurines, to which I now turn.

The Grimaldi sculptures, which are small and designed for suspension, fit the eth-
nographic pattern of amulets or fetishes. However, the majority of human figurines
made by living circumpolar peoples are significantly different from the Grimaldi
figurines in gender ratio, a much higher frequency of facial and extremity detail, anda
much lower incidence of genital and abdominal prominence. Since the circumpolar
ethnographic record is clear that recent human sculptures were used to promote fertil-
ity, these contrasts strengthen our doubts about a fertility magic explanation. Never-
theless, given the tendency of depictive amulets to represent unambiguously the
intended goal of the user (i.e., many hunting amulets were naturalistic sculptures of

116 WHITE

FIGURE | 1, Distinct stone tool marks on the surface of one of the Grimaldi figurines

desired prey animals), it is most likely that the characteristics of the Grimaldi figu-
rines refer to a reproductive context, and that this context was childbirth itself.

Use of the Grimaldi, and at least some other Gravettian female sculptures in the
context of childbirth is consistent with an archaeological context in which they are of-
ten found in clusters as if cached away for future use; childbirth being an occasional
occurrence in small human groups. Moreover, the idea that the sculptures themselves
were perceived as having power is supported by recent finds from Avdeevo on the
Russian Plain (Grigoriev 1996 and personal communication). There, in addition to
purposeful pit-burial of whole sculptures, sometimes more than one to a pit, Go-
vozdover and Grigoriev have found fragments of the same broken figure buried me-
ters apart in meticulously dug pits of a special, cone-like form. If the sculptures were
perceived as inherently powerful, it is easy to imagine that the disposal of broken ex-
amples would have been attended by great care and ritual.

Childbirth is both an emotionally charged and potentially dangerous event. It is
predictable in its general timing (/.e., the average length of gestation), but unpredict-
able as to the timing of the onset of labor, the sex of the offspring, and the survival of
the mother and/or child. We hypothesize that the Grimaldi figurines are best inter-
preted as individually owned amulets meant to ensure the safe completion of preg-
nancy. Amulets employ the principle of similarity to influence the outcome of
uncertain events. They are often made by their owners, although they may also be ob-
tained from shamans. Since the ethnographic record shows that in many societies
amulets are thought to gain power with age, the sculptures may have been passed
from mother to daughter over a number of generations.

[his scenario also satisfies many of the legitimate demands of the feminist
critique. It does not require the figurines to represent a generalized concept of wom-
anhood, but instead recognizes that they may be produced by and for individual

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 117

Ficure 12. Heat induced glaze on the abdominal surface of “/a double figurine.”

women, with no necessary inclusive or monolithic meaning that derives from gender
alone. Individual production probably accounts for the great variability of the figu-
rines. Our interpretation also does not imply the subordination or commoditization of
women as do the fertility goddess (Gimbutas 1989), paleopornography (Guthrie
1979), and mating alliance (Gamble 1982) scenarios. Instead, we recognize the im-
portance of women in themselves, not just as sources of babies, since we suspect the
motivation behind these amulets was the survival of the mother rather than the baby.
From this perspective, women are envisioned as taking active control of an important
part of their lives using magical means that would have been entirely rational within
their cultural context.

We conclude with the observation that the pregnancy symbolism of the Grimaldi
figurines need not be their only symbolic meaning, although it would seem to be a pri-
mary one. Clan or guardian spirits may also be invoked, particularly by parts of the
body suchas heads, hair, duplicate faces, associated animals, etc., that are not directly
involved in our notions of reproduction. In each case these additional referents can be

118 WHITE

Ficure 13. The gouged perforation (top) and additional detail on the tront of “/a Figurine aplatie.”

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 119

seen as statements by the makers of the figurines, perhaps with respect to the actual
spiritual source from which the amulet draws its power.

Conclusions

I believe that the above approach of close observation and analysis, conceived in
terms ofa version of Leroi-Gourhan’s chaine opératoire expanded far beyond simple
production events to include broader social and cultural contexts in which those
events occur, serves as a substantial foundation for accessing domains of thought and
action in the distant past. In particular, a broad technological perspective, when com-
bined with indications of archaeological context and the prudent and restricted use of
relevant ethnographic analogues leads us to consider new interpretive frameworks.
This framework, I believe, has the merit of situating Aurignacian beads and Gravet-
tian/Epigravettian female sculptures, not in some broad, culturally mandated system
of bodily adornment, erotic or reproductive iconography, but in culturally appropri-
ate but very personal and individual practices by which real people negotiated the
normal challenges (physical, social and spiritual) of everyday life.

To return to the original issue of meaning, I wish to reemphasize the multidimen-
sional nature of constructed and construed meanings, meaning viewed here not so
much as ideas but as highly contextualized productions, performances, choices and
skills. | hope that [have made obvious the futility of maintaining “art” and “technol-
ogy” as distinct categories of analysis.

Acknowledgments

The author gratefully thanks M. Pierre Bolduc and Madame Laurence Jullien
Lavigne for permission to study the Grimaldi sculptures in their possession. He also
thanks Dani and Ofer Bar- Yosef of Harvard University, and the directors and staff of
the Peabody Museum, Harvard University for permission and assistance in the study
of collections in their care. At various times assistance was provided by Michael Bis-
son, Henri Delporte, Marie-Hélene Marino, Dominique Buisson, Martin Oliva,
Maria Govozdover, Natalia Leonova, Nicolai Praslov and Gennadi Grigoriev. He
wishes to thank the Wattis Foundation, Jean and Ray Auel, and the editors, Meg Con-
key, Olga Soffer, Deborah Stratmann, and Nina Jablonski for making this volume
possible.

Funding for this research was provided by New York University, the Institute for
Ice Age Studies, the American Council of Learned Societies, and the National Sci-
ence Foundation. Any errors or omissions are the sole responsibility of the author.

Literature Cited

Bartolomei, G., A. Broglio, P. Cassoli, M. Cremaschi, G. Giacobini, G. Malerba, A. Maspero,
M. Peresani & A. Taghiacozzo. 1992. Risultati preliminari delle nuove ricerche al riparo di
Fumane. L 'Annuario storicco della Valpolicella 1991-92.

Bisson, M. & P. Bolduc. 1994. Previously Undescribed Figurines from the Grimaldi Caves.
Curr. Anthropol. 35:458-68

Bisson, M., Tisnerat, N. & R. White. 1996. Radiocarbon dates from the Upper Paleolithic of the
Barma Grande. Curr. Anthropol. 37:156-62.

120 WHITE

Bisson, M. & R. White. 1996. Female imagery from the Paleolithic: The case of Grimaldi. Cul-
ture 16:5-64 .

Delpech, F. 1983. Les Faunes du Paléolithique Supérieur dans le Sud-Ouest de la France. Ca-
hiers du Quaternaire, 6. CNRS, Bordeaux, France.

Delporte, H. & D. Buisson. 1991. Brassempouy: Les fouilles de 1988 a 1990. Bulletin de la So-
ciété de Borda.\16:143-157.

Didon, L. 1911. L’Abri Blanchard des Roches (commune de Sergeac).: Gisement aurignacien
moyen. Bull. Soc. Hist. et Archaeol. du Périgord 87:246-61 and 321-45.

Dobres, M.-A. (1992). Re-considering venus figurines: A feminist-inspired re-analysis. 4n-
cient Images, Ancient Thought: The Archaelogy of Ideology. Calgary: Proceedings of the
1990 Chacmool Conterence, pp. 245-262.

Dobres, M. & C. Hoffman.1994. Social agency and the dynamics of prehistoric technology. /
Arch. Method and Theory |:211-58.

Duhard, J.-P.1993. Réalisme de |'Image Féminine Paléolithique. Cahiers du Quaternaire, No.
19. CNRS, Bordeaux, France.

Gamble, C.1982. Interaction and alliance in Palaeolithic society. Man 17:92-107.

Gimbutas, M.1989. The Language of the Goddess. Thames and Hudson, London, U.K.

Giraudi, C. 198]. Presenza di depositi medio-pleistocenici intensamente deformati in Val
Cerrina (Monferrato settentrionale). Geogr. Fis. Dinam. Quat. 4:69-74.

1983. Provincia di Alessandria: |. Pontestura. Quaderni della Soprintendenza Archeo-
logica del Piemonte 2:143-144.

Grigorieyv, G. 1996. Le travail de Pivoire sur la plaine Russe: le cas d’ Avdeevo. in M. Menu, P.
Walter & F. Widemann, eds. Le Travail et |'Usage de |'Ivoire au Paléolithique Supérieur,
Centre Universitaire européen pour les Biens Culturels, Ravello, Italy.

Guthrie, D. 1979. Ethological observations from Palaeolithic art. Pages 35-74 in Bandi, H.-G.
etal, ed., La Contribution de la Zoologie et de l'Ethologie a I'Interprétation de l'Art des
Ppeuples Chasseurs Préhistoriques. Editions Universitaires Fribourg, Fribourg, France.

Hahn, J. 1972. Aurignacian signs, pendants and art objects in Central and Eastern Europe.
World Archaeol. 3:252-56

_ 1986. Krafi und Aggression: Die Botschafi der Eiszeitkunst im Aurignacien Sud-
deutschlands. Archaeol. Venatoria, Tubingen, Germany.

Knecht, H. 1993. Splits and wedges: The techniques and technology of early Aurignacian ant -
ler working. Pages 137-62 in H. Knecht, A. Pike-Tay & R. White, eds., Before Lascaux:
The Complex Record of the Early Upper Paleolithic. CRC Press, Boca Raton,

Lechtman, H. 1977. Style in technology: Some early thoughts, Pages 3-20 in H. Lechtman & R.
Merril, eds., Material culture: Styles, Organization and Dynamics of Technology Ameri-
can Ethnological Society, St. Paul.

Lemonnier, P. 1983. L'Etude des systémes techniques, une urgence en technologie culturelle.
Techniques et Culture \:11-34.

Leroi-Gourhan, A. 1943. Evolution et Techniques: |'Homme et la Matiére. Albin Michel,

Paris, France
A. 1945. Evolution et Techniques: Milieu et Technique. Albin Michel, Paris, France.

_ A. 1993. Gesture and Speech. MIT Press, Cambridge.

MacGregor, A. 1985, Bone, Antler, Ivory and Horn: The Technology of Sskeletal Materials
Since the Roman Period. Croon Helm, London, U.K.

Mauss, M. 1936. Les techniques du corps. J. de Psych. Vol. 32.

Nelson, S. 1990. Diversity of the Upper Palaeolithic “Venus” figurines and archaeological my-
thology. Pages I1-22 in S. Nelson & A. Kehoe, eds., Powers of Observation: Alternative
Views in Archaeology Archaeological Papers of the American Anthropological Associa-
tion, Washington, DC.

MATERIALS AND MEANING IN PALEOLITHIC IMAGES 12]

Peyrony, D. 1935. Le gisement Castanet, vallon de Castelmerle, commune de Sergeac (Dor-
dogne). Aurignacien I et If. Bull. Soc. Préhist. Frang. 32:418-43.

Phillipov, A. 1983. Antropomorfniye statuetki is Mezina. Pages 34-38 in R.S. Vasil evskii, ed.
Plastika i risunki drevnikh kul ‘tur. Novosibirsk, Nauka, Russia.

Ritchie, C. 1969. /vory Carving. Barker, London, U.K.

Rice, P. 1981. Prehistoric venuses: symbols of motherhood or womanhood? J. Field Archaeol
37:402-414.

Schlanger, N. 1994. Mindful technology: unleashing the chdine opératoire for an archaeology
of mind. Pages 143-151 in C. Renfrew & E. Zubrow, eds., The Ancient Mind: Elements of
Cognitive Archaeology. Cambridge University Press, Cambridge, U.K.

Steward, J. 1955. Theory of Culture Change. University of Illinois Press, Urbana.

Taborin, Y. 1993. La Parure en Coquillage Paléolithique. XX1Xth supplement to Gallia
Préhistoire. CNRS, Paris, France.

White, L. 1959. The Evolution of Culture. McGraw Hill, New York.

White, R. 1989. Production complexity and standardization in Eraly Aurignacian bead and
pendant manufacture: evolutionary implications. Pages 366-90 in P. Mellars & C. Stringer,
eds., The Human Revolution Edinburgh University Press, Edinburgh, U.K.

. 1992. Beyond art: Toward an understanding of the origins of material representation in
Europe. Ann. Rev. Anthropol. 21:537-64.

1993. Technological and social dimensions of “Aurignacian-Age” body ornaments
across Europe. Pages 277-99 in H. Knecht, A. Pike-Tay & R. White, eds., Before Lascaux
The Complex Record of the Early Upper Paleolithic. CRC Press, Boca Raton.

. 1995a. Comment on symbol mediated marking in the Lower Paleolithic by R. Bed-
narik. Curr. Anthropol. 37:623-625 .

___. 1995b. Ivory personal ornaments of Aurignacian age: Technological, social and sym -
bolic perspectives. Pages 264-267 in M. Menu, P. Walter & F. Widemann, eds., Le Travail
et 1'Usage de I'Ivoire au Paléolithique Supérieur. Centre Universitaire européen pour les
Biens Culturels, Ravello, Italy.

. 1996a. Les images féminines paléolithiques: Un coup d'oeil sur quelques perspectives
ameéricaines. in H. Delporte, ed., Centenaire de la Découverte de la Dame de Brassempouy
Ses Ancétres, ses Contemporaines, ses Héritiéres. Sur Quelques Perspectives Americaines
Etudes et Recherches Archéologique de |' Université de Liege. Liege, France.

. 1996b. Comment on L. McDermott’s “Self-representation in Pavlovian, Kostenkian
and Gravettian female figurines during the European Upper Paleolithic.” Curr. Anthropol
37:264-267.

Section II

Approaches to the
“Why’s” of Presence and
Absence

i)
an

The Power of Pictures

lain Davidson

Department of Archaeology
University of New England
Armidale NSW 2351
Australia

Europe is a small and not very well-defined peninsula of Asia which has had a sig-
nificant influence on the cultural development of modern humans. Arguably, the
most fundamental influence is not the sculpture of the Classical Greeks, the paintings
of the Italian Renaissance, the plays of Shakespeare, or the musical tradition of Mo-
zart and Beethoven, but the paintings and engravings of the Upper Paleolithic be-
tween 30 thousand years ago (kyr) and 10 thousand years ago (kyr) in a small region
of (mostly) western Europe (Aujoulat ef a/. 1993). Twill argue in this paper that early
humans worked out their relationships with land and with each other through this
“art.” The similarities between the sequences of changes in different regions suggests
this was not just the product of individual historical particularities, but part of a single
process by which humans explored the limits of recently-discovered symbolism. So
important was this in establishing what it was to be a human in these regions that it is
probably misleading to call the paintings, engravings, drawings and stencils a7? at all,
given the baggage that that word carries with it.

I have attended to the problem of definition elsewhere (Davidson 1992). I note
here that “art” is that making or marking of surfaces people talk about when they refer
to paintings, engravings (including sculptures), drawings and stencils of the Pleisto-
cene. It is sometimes convenient to refer to this making or marking of surfaces as
PEDS. Whether this “art” meets anyone’s criteria for art is entirely up to them. I do not
believe it is useful or necessary to attempt to define art (without the inverted com-
mas), because as every controversy about art, from Braque to Warhol to spray-can
graffiti, indicates whether something 1s art depends on whether someone thinks itis. It
is not analytically possible to explore the origins or significance of such a concept. |
set out below our argument about language origins. It is impossible for creatures with-
out language to hold the opinions about making or marking of surfaces that make
them art. It was only after the processes described in this paper that we might consider
this as art, though, even then, its functions were probably as complex as those of art to-
day.

Duff, Clark and Chadderton (1992) take the Upper Paleolithic as the standard for
recognizing symbolism but admit that it is also anomalous, because symbolism did
not emerge elsewhere in the world populated by modern Homo sapiens. Schepartz

Beyond Art: Pleistocene Image and Symbol

Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski

Number 23, Copyright ©1997

126 DAVIDSON

(1993) suggests the “blasphemy” that “most of the spectacular art of the Upper Paleo-
lithic is irrelevant to the origin of modern Homo sapiens and complex language.”
Schepartz goes on: “its late occurrence after 35,000 BP in the region where unques-
tionably modern Homo sapiens is last identified makes it an afterthought or adden-
dum to modern human origins.”

Itis my purpose in this paper to take up the challenges of these statements. First, is
it useful to consider the Upper Paleolithic as the standard for recognizing symbolism
in the archeological record? Second, 1s it true that there is no comparable patterning in
the rest of the world of the Late Pleistocene? Third, is there any common thread in the
sequence of changes in Upper Paleolithic art and in Australian prehistoric art that en-
ables us to understand the significance of both in the evolution of human behavior,
and should we expect that this pattern would be repeated in other regions of the
world? I will suggest that, insignificant though Europe may be, the art of the Upper
Paleolithic was far from insignificant, not irrelevant and not an afterthought. It gives
us unique insights into evolutionary processes that were far from unique for they
probably defined the very nature of one way of being a modern human.

Language and Symbols

I shall be taking it as a fundamental assumption throughout that language emerged
late in human evolution. It is a position that we have expended much effort in arguing
(e.g., Davidson 199 1a; Davidson & Noble 1989, 1992a, 1993a; Noble & Davidson
1989, 199 1a, 1993a, 1996) and should not need to argue here. We recognize that there
is an argument that cranial expansion is first identifiable two million years ago (e.g.
Falk 1987; Tobias 1987) and that there is some argument that the regions of the brain
relevant to language production emerged at the same time (Falk 1983; Holloway
1983). We recognize too that the vocal apparatus evolved gradually over the course of
human evolution (Lieberman 1991; Lieberman er al. 1992) so that individual discov-
eries of relevant portions of the anatomy from different periods may give the appear-
ance of suitability for modern language production through speech (Arensburg ef al.
1990; Arensburg etal. 1989). But language must be distinguished from other forms of
communication in two fundamental characteristics: its production and comprehen-
sion inany ofa number of different modes, such as speech, sign, or writing; and its use
of symbols in any of these modes. It is crucial to our argument that the transition be-
tween language absence (though surely with vocal and other communication) and
language use involved the recognition of symbols. We see no evidence that the argu-
ment which associates an early language origin with the early encephalization step Is
supported by a credible case for symbol use.

Like Chase and Dibble (1987; Chase 1991) we see symbolism as a phenomenon of
the Late Pleistocene, though details of our argument are rather different from theirs.
Our speculative scenario about the transition involves the production of signs which
we would recognize as iconic. This has been represented as meaning that we believe
that the origin of language can be identified in the origin of art. I do not wish to get
bogged down in refuting that argument (again), rather I shall be exploring the impli-
cations of a late language origin for our understanding of aspects of the archaeology
(including “art”) of the Late Pleistocene. For the record, we have argued that the earli-
est evidence which requires an interpretation that there were people who had lan-
guage is the fact of colonization of Australia (Davidson & Noble 1992a) around 60
thousand years ago, though this necessarily entails that there was an origin of lan-

THE POWER OF PICTURES 127

guage earlier, in Asia, for which we have no evidence. This seems to be 20 thousand
years earlier than the Upper Paleolithic of Europe.

We have developed one other argument that is fundamental to the framework I
shall discuss here (Davidson & Noble 1992a; Noble & Davidson 1993a): The arbi-
trariness that is a distinctive feature of the symbols which characterize language and
no other communication system results in the possibility of ambiguity. Cheney and
Seyfarth (1990) describe how vervet monkeys attending to the calls of their conspe-
cifics have different behavioral responses to calls uttered in the presence of a leopard,
an eagle or a python as if the calls were alarms with “meanings” such as “Watch out!
There is a leopard (eagle, python)!” This communication system has no possibility of
ambiguity because it is honed by natural selection. There can be no more perfect com-
munication than one with unvarying signals produced in response to precisely de-
fined stimuli such that they elicit consistent behavioral responses from the receiver.
But language, characterized by conventionally defined arbitrariness has an inherent
potential for “untrustworthy communication by accident or design” (Davidson & No-
ble 1992a). As we wrote previously:

Successful communication using language, thus requires means of identification, by the
partners in the exchange, that the utterances are trustworthy. Through this, members of
human groups can confer the advantage of information and knowledge on an ingroup
and restrict access for an outgroup....

In tracing the evolutionary pattern of behavior emerging after the first emergence
of language, we would expect that the archeological record would show signs of the
ways in which people coped with this phenomenon. Dunbar (1993) has recently al-
luded to the use of emblems among early language users. We suggest that there
should be signs of these in the archeological record. Moreover, we see itas fundamen-
tally significant in this argument that Barbujani & Sokal (1990) have demonstrated
that, in modern Europe, language variability has been responsible for creating genetic
diversity.

The Meaning of “Meaning”

Many of the innovations in the study of Late Pleistocene paintings, engravings,
drawings and stencils (PEDS) have come froma recognition, which I trace to the pub-
lications of Conkey (e.g., 1978) in the late 1970s, perhaps influenced by Ucko and
Rosenfeld’s (1967) demolition of Leroi-Gourhan’s more extreme fantasies, to the ef-
fect that we can never recover the “meaning” that prehistoric pictures may have had
for people, either makers or observers, at the time of their making. As Conkey (1987)
put it: “We cannot expect to capture ‘meaning’ as a single, inclusive, empirical entity
or category of our inquiry.”

Macintosh’s (1977) simple demonstration of the difficulty of corroborating inter-
pretations of meanings brought this home to Australian scholars. In this case, plausi-
ble identifications of the figures in rock paintings at a site in northern Australia were
generally different from the identifications made by an Aboriginal informant. If this
was true even for one case, the significance was clear for a// studies of prehistoric
paintings, engravings, drawings and stencils. Identification of the “meaning” of “art”
even at this simplest of levels was enmeshed in the social context of making the iden-
tifications. For many of us, this was an observation that did not need to be made since
the proposition was obvious. In saying that, I have to acknowledge that we had not im-
mediately recognized the liberating effect that Conkey’s work (much less that of

128 DAVIDSON

Macintosh) would have for the study of late Pleistocene archaeology. The proposition
is now made abundantly clear by the publication of Morphy’s (1991) most detailed
study of the production of meaning ina society that lives by producing pictures. Here
we see that the original meaning can only be said to exist through the contexts of pro-
duction and consumption of either individual paintings or parts of paintings. Al-
though this 1s an example from Australia, 1am confining my reading of it to the most
general interpretation that can be translated into other cultural situations.

It follows from Morphy’s argument and evidence that discovering the “original”
meaning is more likely to be an accident than a result of the subtlety of our analysis,
since we are unable to become partners in the contexts of production or consumption.
Even ethnoarchaeology of “art” is vitiated by this observation: Unless and until writ-
ing about “art” analytically becomes part of the contexts of production and consump-
tion of the “art,” it will always be true that the analysis stands outside that original
meaning.

There may be an element of truth in the proposition that for the participants in the
context “the meaningfully constituted material record is not an ‘expression’ or ‘refle-
ction,’ nor even a ‘record,’ but an active, constructing, constituting agency, which
does not express meaning, but produces it” (Conkey 1989), but not for non-
participants. The material record, being an object and not a person, cannot ever be an
agent in the process of producing meaning. People may, through material things, ac-
tively construct and constitute something, especially in the sort of context of modern
people described by Morphy. They may even produce meaning by their interaction
with material that is not modern, usually in terms of stories about the past. Sometimes,
we believe that those stories tell us about similar interactions of construction and con-
stitution in the past, by different people with the same materials. Usually we look fora
way to determine whether our stories are more or less plausible than others, for with-
out that we have only three, not wholly satisfying, characterizations of the stories: 1)
they are “intellectual games for the meritocracy”; 2) “the materials speak to me”; and
3) the stories (about the past) are on/y about the production and consumption of mean-
ing in the present, and, as with “art,” they are concerned with the construction and
constitution of social relationships between the story teller and the audience (ifany).

In this view, meaning is nota property of the “art” or any other material, but of the
interaction between the human agents and the material. The meaning | will find in the
earliest “art” produced by people is more a product of my thought than it is of their
view of the “art.” Whether anyone agrees with me about this interpretation will be a
matter of their opinions of me or my argument (though they will probably couch it in
terms of their view of some part of the same material). It is inevitable that the perspec-
tive I will take, given that I can see material from a period of 25 thousand years, differs
from any perspective available to any participants in the process of production.

Investigating Contexts

Freed from the search for meanings of individual images or groups of them, the
way was open for attempts to integrate studies of Late Pleistocene “art” with other
studies of the archaeology contemporary with the pictures. The problem has always
been the lack of chronological resolution for the paintings and engravings on cave and
shelter walls, but now that we may be poised on the brink of a revolution in that area
(Clottes et al. 1992, Lorblanchet et a/. 1990, Loy et al. 1990, McDonald et al. 1990,
Valladas eral, 1992, Van de Merwe eral.1987), the full potential of the study of paint-

THE POWER OF PICTURES 129

ings and engravings in the context of other aspects of prehistoric societies may be re-
alized. My caution about whether direct radiocarbon dating of paintings by
Accelerator Mass Spectrometry (AMS) will provide the hoped for chronology de-
rives from recent research which shows how complex is the organic environment of
prehistoric paints (Ridges 1995), suggesting that most dates published hitherto
should be considered with caution (see also Watchman, this volume).

My own first attempt at integration occurred in 1980 at the completion of my
Ph.D. thesis on Late Pleistocene fauna and economy in eastern Spain (Davidson
1989c). Discovering that Clan of the Cave Bear had just been published, making fool-
ish my desire to turn the thesis into a novel, I turned my attention to a subject deliber-
ately neglected, though not ignored, in the thesis: the painted and/or engraved
plaquettes from Parpallo. In trepidation, I ventured into the world of “art” specialists
from the relative security of faunal identification and found that Parpallé was even
more significant than we already knew it was.

Parpallo is a cave site in the hills south of Valencia about 10 km inland from the
present shore. It was discovered as an archeological site in the nineteenth century, and
visited by Breuil in 1913. He found an engraved plaquette on the surface of the site.
Pericot (1942) completely excavated the site in three seasons from 1929-31, finding,
for the first time, a sequence of stone and bone industries from the region which could
convincingly be compared to the sequences in France. The faunal remains were
mostly of red deer and ibex, with some horse and cattle, species which exist in some
form or other in Spain today. Thus, as the fauna provided no clues to the chronology,
one of the important aspects of the easy comparison with French stone and bone in-
dustries was that the site could be tied to a chronology worked out in the more envi-
ronmentally sensitive regions to the North. In my thesis I studied the bulk of the
surviving faunal remains as well as the context of the site in the region. I obtained ra-
diocarbon dates from the site from the faunal remains which established beyond
doubt the accuracy of the earlier comparison between the industries of France and
Spain (Davidson 1974), with a Solutrean date 20, 490 +900/-800 years BP effectively
identical to the date from Laugerie Haute, and a coherent series of dates that has al-
lowed detailed comparison with the nearby dated site of Les Mallaetes (Bofinger &
Davidson 1977; Fortea & Jorda 1976) for the period between 21 kyr and 14 kyrat Par-
pallo and 30 kyrand 10 kyrat Les Mallaetes (see Davidson 1989c for full discussion).

Breuil’s discovery of an engraved plaquette proved not to be a lucky chance:
rather it was a sign of what was to come. Pericot found large numbers of plaquettes
stratified in the deposits at Parpallé of which more than 6000 surfaces were engraved
or painted or painted and engraved. Some of these had 486 animal figures (of which
250 are identifiable [Villaverde Bonilla 1992]) which allowed acceptable identifica-
tion of species.

On the basis of these identifications I was able, in the thesis (see Davidson 1989c:
59-60, 229), to make a comparison between the fauna in the site and that in the pic-
tures. As with almost all other studies of prehistoric pictures, of course, the pictures
were “neither a check list nora menu” (Vinnicombe 1972) but this was the first study
where a large number of pictures of known date could be compared with the bones
from the same date and this over a period of several thousand years.

The plaquettes have recently been restudied (Villaverde Bonilla 1990, 1992; Vil-
laverde Bonilla & Marti Oliver 1984), providing some new identifications and de-
tailed analyses of style changes beyond what Pericot (1942) could attempt in the
Spain of the late 1930s. The pattern of relative frequency of representation of animals
in the pictures is different from their representation in the faunal remains in general.

130 DAVIDSON

Horse and cattle are relatively more frequent in the representations than in the bones
(though there are taphonomic problems which prevent quantification of this). For the
relative abundance of ibex and deer, ibex are underrepresented in the paintings and
engravings in the early layers, though there is some correspondence between pictures
and bones after the Last Glacial maximum. Whatever the primary reason for the accu-
mulation of the faunal remains, human or nonhuman sampling (see Davidson &
Estévez 1985; Lindly 1988), the result tends to confirm Vinnicombe’s neat judgment.

The task I set myself was to consider the uniqueness of Parpallo. For all its similar-
ity in stone artifacts to the sequences of France, Parpallo has proved to be outstanding
in eastern Spain, and to some extent throughout the Upper Paleolithic world, in the
abundance and distribution of the painted and engraved plaquettes. In eastern Spain
there are still less than a couple of dozen other plaquettes with paintings and engrav-
ings, and very few of these have figures of animals on them. One of these is from Les
Mallaetes and is from the layers earlier than the occupation of Parpall6, while two
others are from the site of San(t) Gregori, 250 km to the north in Tarragona province,
and is almost certainly later in date than the Parpallo tradition. It is not an exaggera-
tion to describe Parpalldé as unique. The problem is that it is impossible to find an ex-
planation of unique events or phenomena; this can only come from finding a context
in which the event or phenomenon is not unique. Even the more recent finds from
Tossal de la Roca (D’errico & Cacho 1994; Ripoll Lopez & Cacho Quesada 1990)
seem to belong to a later tradition, probably contemporary with the San(t) Gregori
piece (Figure 1).

I chose, for reasons I have elaborated elsewhere (Davidson 1989a), to emphasize
the context of sites with paintings and engravings on stone plaquettes, and in doing so
discovered that, although through western Europe there are many sites with one or
two plaquettes, there are very few which have large numbers and these are rather sur-
prisingly spread out. The list [came up with was Labastide, Gonnersdorf, Limeuil, La
Marche, Isturitz, Badegoule, Enléne, Tito Bustillo and possibly Fontales (Figure 2).

I suggested (Davidson 1989a: 450) that:

If we treat the “art” painted and/or engraved on plaquettes as a separate category of “art.”
with its own special information content, then we must conclude, either that “art” sites
were aggregation sites (cf Conkey 1980), or that information was not only imparted
through the “art” but also restricted, or both. The evidence from the study of the prehis-
toric economy suggests that it was both.

The argument that followed was that the plaquette sites might all be considered as
similar centers for the dissemination and restriction of access to meaning. I note that
since my publication, Bosinski (1991) has assembled an argument that suggests that
Gonnersdorf could have functioned in this way. Such restriction of access to informa-
tion, | argue, indicates differential access to power in the Upper Paleolithic societies
of Western Europe beginning about 21 thousand years ago. I believe this is the earliest
reliable evidence for such social structure in the archeological record. Moreover,
looking at the distribution of these sites suggests that there may have been a dynamic
aspect to this tradition, with centers of power changing in number and distribution
over a period of ten thousand years.

THE POWER OF PICTURES 131

Sant Gregori Sant Gregori

Tossal de la Roca Tossal de la Roca

Tossal de la Roca

Mallaetes

Ambrosio Cova del Barranc

FiGure |. Plaquette representations from sites in eastern Spain, other than Parpallo. From Fortea Perez
(1978), Fullola i Pericot, Viitas i Vallverdu, & Garcia Arguelles i Andreu (1990), Olaria e¢ al. (1981),
Ripoll Lopez & Cacho Quesada (1990)

Limeuil

Istuntz

FiGURE 2, Sites with many plaquettes (including Parpallo),

plaquettes

DAVIDSON

La Marche Gonnersdort

ceed

cease ot 10.5

La Marche 15
e

Badegoule (1/4)

of 9@ — @Fontales
Limeuil (12) Ww i¢,
Be a
“} Bustillo is. ee ©, Leoesiite 14 , ¢
20-10) eniéne 137 Sate
3 \
ne
STS
Les Mallaetes mt fs
oy) ‘
Parpaild 22- rs #0 ee ‘ j

pee
eee

@ site with plaquettes

22 radiocarbon age Gs

(12) age estimated trom
artefact association

Labastide Enléne

es Mallaeles Parpailo

showing approximate dates of presence of

THE POWER OF PICTURES lle

ws)
(vs)

Categories of “Art”

In my discussion of the plaquettes (Davidson 1989a) it was absolutely necessary
to consider that different categories of “art” might have different contexts of meaning
and that in any society such different contexts of meaning might not be excluding.
That is to say that the presence of one category of “art” with its contexts and meanings
does not exclude the possibility of another category in the same society having en-
tirely separate contexts and meanings. There may have beena multiplicity of contexts
for the use of “art.” The example I gave was the historical use in Australian Aborigi-
nal society of designs on shields and on message sticks. Although there are some
similarities between the designs on these two categories of objects, the contexts of
their use were very different. Designs on shields are large and visible, and while the
designs could be used to signal a friendly presence, more often they would have indi-
cated that the bearer of the shield had some expectation that it might be needed for de-
fense. (Note, in passing, that the advertisement of difference must imply some
convention of understanding of meaning across a boundary.) Message-sticks on the
other hand were used in contexts of interpersonal communication to facilitate friendly
interaction (Figure 3).

Analysis of the designs on shields need not necessarily take into account that there
were similar designs on any other category of object, but ina system where both cate-
gories were known to exist there might be different stories to tell about groups which
used both categories and groups which used only one or other of them. At Parpallo I
found that I needed to satisfy myself that the restriction of painting an engraving to
stone plaquettes was a real choice and not one determined by a lack of other raw mate-
rials. Fortunately, I found among the bones I was studying one and only one bone with
an engraved design (Aparicio 1981) demonstrating that the people of Parpallo had a
choice of context for production of their information, and exercised it ina very biased
way (Figure 4). The original argument was produced in 1980, but, being wary, then as
now, of entering the world of people who knew so much more than I about Upper Pa-
leolithic art, I did not submit it for publication until 1986.

What I did not dare to do at that time, but do now, was to consider the implications
of separating the distinct categories of “art” for the possibility of looking at a sequen-
tial history of such categories. I think that at the time I preferred to think of the unity of
the Upper Paleolithic tradition in terms of an artistic tradition that lasted twice as long
as the period since it was last produced. As a result of my wider reading, I am, now,
absolutely sympathetic to the need to consider it as several traditions as Conkey has
repeatedly entrained us to do. But I think it is also valuable to think of it as an evolving
tradition or traditions that has or have the capacity to reveal regularities, not just west-
ern European particularities, about the evolution of human behavior. It is not the first
time such a claim has been made, of course, though usually it has been made from a
Eurocentric point of view. I am arguing, rather, from a point of view that regards
Europe as a rather isolated little peninsula far away from most of the scenes of most of
the major events of the prehistoric period. That does not mean that people in western
Europe were not subject to the same sorts of evolutionary pressures as other evolving
humans elsewhere in the world. It is my argument that the fine detail of the story can
be told from the massively revealed archeological record of western Europe because
it is confirmed by the record of Australian prehistory, as described by Rosenfeld
(1993), together with a perspective on the early Australian evidence recently pub-
lished by Noble and myself (Davidson & Noble 1992a).

34 DAVIDSON

For detail see enlargement |

FiGuRE 3. Shields and message sticks from Queensland

THE POWER OF PICTURES

Bune] JO sis Ayeue $ UOSplAed suunp punoy ‘oypedaeg Wwol} 9u0q pa AvIBUY p NOI

L
\ N

. Kw.

is

136 DAVIDSON

The Australian Sequence

Reviewing the evidence for the emergence of rock art in Australia, Rosenfeld dis-
cusses the sites which have been radiocarbon dated, indirectly or directly, those
“dated” by measurement of cation ratios and claims for dating on the grounds of con-
tent, particularly through the claim that there are representations of the extinct mega-
fauna. Rock engravings from the “Early Man” site, also in Cape York (Figure 5), were
covered by deposits certainly dated at 13 kyr and possibly 15 kyr (Rosenfeld 1981):
carbonate within a complex varnish covering engravings at Sturt’s Meadow, in west-
ern New South Wales (Dragovich 1986), dates to around 10 kyr; pigment from hand
stencils in Judd’s Cavern, Tasmania has been dated by AMS to just less than 11 kyr
(Loy etal. 1990); in addition, Morwood (Morwood et al. 1995, 79) found an engraved
fragment of the rock of the shelter in deposits at Sandy Creek 1, in Cape York, dated at
14 kyr. The claimed early dates by cation ratios depend on a local calibration which
Watchman (1992; this volume) has found to be problematical. Accelerator Mass
Spectrometer dating at Gnatalia Creek and Laurie Creek may both involve inclusion
of older oxalates into the dated pigment samples (McDonald et a/. 1990). There is un-
certainty about the meaning of the date of 20 kyr obtained from a rock fragment with
an apparently artificial red coloring at Laurie Creek (Loy 1994, Loy et al. 1990, Nel-
son 1993). Dating by apparent depiction of megafauna (e.g., Murray & Chaloupka
1984) involves a series of unwarrantable assumptions about the “accuracy” of repre-
sentation, the form of animals which are now extinct, and the age of that extinction.
Rosenfeld concludes that the “earliest solid evidence for structured visual systems on
rock date to the terminal Pleistocene — early Holocene.” Possibly earlier are the
“finger-flutings” at Koonalda Cave, but the date of 20 kyr here comes from charcoal
found on the floor of the cave below the engravings, which could result from a visit
earlier than the engravings, especially given the later dates attributed to apparently
similar markings elsewhere in southern Australia (Maynard & Edwards 1971: Rosen-
feld 1993). Rosenfeld also raises the possibility that these markings should be “con-
sidered to belong to a system of meanings distinct from referential symbols” just as
has been argued for hand stencils. Watchman (Cole et a/. 1995:156) has recently an-
nounced AMS dates for gypsum crusts at Sandy Creek 2, in Cape York.

The lowest layer of these crusts contains hematite and in one of the crusts a date of
24 kyr was obtained. These crusts do not cover any recognizable painting, and there-
fore are not direct dates of painting. There 1s a possibility that the hematite was not in-
volved in symbolic marking of the rocks.

This skeptical and rigorous assessment of the chronology of Australian PEDS
suggests that the phenomenon emerged rather late (perhaps no earlier than 13 kyr) by
comparison with our expectations from the standard story of the emergence of “art”
elsewhere in the world. Symbolically complex statuettes are dated to 32 kyr in Swabia
(Hahn 1986): there are painted plaquettes from Namibia at 26 kyr (Lewis-Williams
1983; Wendt 1974); the Parpallo plaquettes are at least 20 kyr (Davidson 1974), and
there are direct AMS dates of cave walls with applied paint discussed below (Figure
5). Yet Australia was first colonized earlier than 50 kyr (Roberts er a/. 1990; but see
Davidson & Noble 1992b) and, we have argued, the abilities which make possible the
production of “art” were necessary to build the boat and conceive the voyage that
brought people to Australia (Davidson & Noble 1992a). One conclusion might be that
we are being too rigorous in the standards of acceptable dating of Australian rock
paintings and engravings especially in terms of a view that the necessary abilities

POWER OF PICTURES

E

TH

udye} oie xo] =)08) passnosip }

sandy pure sajyep Surmoys “(1 661) UyeR U

pajd

OF 4 X8819
eyyeyeuy,

/

gi uew Ape is a6 :
/ {dvb olody
/ (
| )
—& \ ( SE
vx 3 ane Japiog
( )
Oz \dnjew 3 )
4 \ l
7 \ |
E a (
\ Se aujeg :
\ wiuoAeHé Fhe) 02
\ PL Qny-3 ueyin 3/ rJeyseue |
\ ly zesy ae 29 f
\ \ J Al —
2 prayjabon gg 3
axa f\ | SV SS

ae z Ze alsseuay ey

asoyl) pom ayy punoae says ue, Ava fo dey “¢ 3
aeq 92
jwelag 2

pajuied

aneg sanbso9

138 DAVIDSON

were present among the first Australians. But arguments cannot be based on wishful
thinking.

Davidson and Noble (1992a) sought to assemble the evidence from Australia that
artifacts with “imposed form” were widespread early, among artifact assemblages
that are not characterized by large numbers of “types.” Ground edged axes and
ground bone points appear very early in the Australian sequence. In addition there are
other traces of evidence which imply symbolic marking, but not marking of places.
The earliest evidence in Australia, dated by thermoluminescence at more than 50 tho-
usand years ago, is accompanied at two sites by lumps of ocher with ground facets.
This repetitive patterning in sites only a few kilometers apart distinguishes these
ocher discoveries from most earlier claims in the world. Yet there is no plausible case
for rock painting at this time. Added to this is the existence in Mandu Mandu ofa se-
ries of pierced shells from a layer with radiocarbon dates of 34 kyr and 30 kyr (Figure
6). These give every sign of having been strung, and should probably be considered as
personal ornaments (Morse 1993). Once we have opened up the possibility of per-
sonal ornaments, the ground ocher from Malakunanja and Nauwalabila seems plausi-
bly interpretable as a sign of body painting. Some confirmation that such attention
was paid to bodies relatively early in the Lake Mungo III burial, estimated to be dated
between 30 kyr and 28 kyr (radiocarbon ages) (but actually undated), which ts sur-
rounded by ocher staining (Bowler & Thorne 1976). In addition there is a bead from
Devil's Lair dated between 19 kyr and 17 kyr (radiocarbon ages) (see references In
Davidson & Noble 1992a) (Figure 6).

Rosenfeld (1993) concludes:

[ T]he use of areferential system of symbols to mark the landscape according to a corpo-
rate system of meanings appears to have its origins during the period of rapidly changing
environmental conditions of the terminal Pleistocene. ... [T]he integration of features of
the landscape into a corporate symbolic system is entirely consistent with a model of
tightened social and territorial organization at the end of the Pleistocene. Corporate ter-
ritorial expression through the indelible marking of place with a stylistic graphic system
may have been a powerful means of asserting corporate rights and relationships.

If we accept that an argument can be sustained for such corporate territoriality in
the Late Pleistocene, it must be pointed out that such a corporate social organization is
an identity that needs to be established and which could only exist with some form of
marker of the identity of one rather than the other corporation, Such marking may re-
quire the personal marking or decoration of individuals. There would, therefore, be an
evolutionary scenario accounting for the apparent sequence from personal ornamen-
tation to the marking of places.

The Upper Paleolithic Standard?

The proposition I am putting here is that the sequence of events in the Upper Pa-
leolithic of Europe was similar to that in Australia, though replete with additional de-
tail. We can see that personal ornamentation is earlier in the surviving evidence than
the marking of places. I suggest that personal ornamentation is the key to understand-
ing the process of change in the “art” of the Upper Paleolithic and in Australia. Hence,
we identify a process which may be best explained as an evolutionary one. That is not
to say that there was an inevitability of the “progress” from personal decoration to the
painting of caves. We are considering, rather, that in both continents the newly ar-
rived humans had only recently discovered language and its symbolic properties. In
this context, they then discovered personal decoration. The options that opened up

139

THE POWER OF PICTURES

(C661 Aatpe ay) ROLY YINOS pue ([ 6H] YOvysieyy) PINsny WoO] sajdurexa juadal aIOU aTe UMOYS OS|TY (C661 FU “6861

AMY AA) 219294 ayi “(ZL6] UYeH) BIqeas (C66 2SIOJ) pure (pR6] YOHO) elpeNsny ul JAY OE URY) JalpIea WoI] UONLIOIAp [eUOSIad Jo sjoafqo JayjO puke speag °9 TUAYI4

ei 11e7 SIAAG

Ze NpuewW npuew

s (2)
Q a
a ©
S &
é s 5 —

Ze pueyouelg gy ; 7 |

@®@GS OER
|@ @ & © &

ZE aya\sojyuassiay

140 DAVIDSON

led. in one direction, to symbolic marking of places. We do not have time, here, to do
more than foreshadow the options that opened up in the other direction. Evolutionary
processes of natural selection apply to the emergence of “art,” and it is fundamental to
such an argument involving selection, that there existed options other than those we
are considering here. The consequences of pursuing those options were different, and,
in the end, those who chose them overwhelmed those who did not (Davidson 1989b).

Sequential Categories in the Upper Paleolithic

I will consider the sequential evidence from the Upper Paleolithic for personal
decoration and conventionalized arbitrariness; the visibility of early symbolism; the
use of symbolism in social networks; plaquettes, power and restriction; and cave art. I
regard the decorated objects of the end of the Pleistocene as possibly involving
“mere” decoration, but I will not develop the argument here.

Personal Decoration and Conventionalized Arbitrariness. The Upper Paleo-
lithic evidence is well known, and now that AMS dating of pigments is becoming an
acceptable technique the details of the chronology are becoming clearer, though cau-
tion 1s clearly necessary (Clottes in press).

I have previously dismissed much of the evidence, except the Tata tooth, claimed
to show symbolic material from earlier than the Upper Paleolithic (Davidson 1990,
199 1a). Ifwe accept this, then the earliest evidence for something symbolic in Europe
is represented by the beads and pierced canines of both Swabia and South western
France (e.g., Hahn 1972, 1986; White 1989) with dates between 36 kyr and 30 kyr, but
conventionally assigned a date of about 32 kyr (Figure 6).

The evidence that personal decoration was an important feature of the earliest
symbolic marking gains intriguing confirmation from examination of the bone and
ivory sculptures from Swabia (Hahn 1986). Here, the statuettes from Vogelherd,
Geissenklosterle and Hohlenstein Stadel all have distinctive marking on them which
suggests a repeated patterning involving something other than representation (Figure
7). The important point here ts that the “meaning” of this marking, or of the patterning
in this marking, requires a convention of understanding. It requires it for us, but cru-
cially we can infer that the makers had a convention, though we cannot say what the
meaning of it was. Noble and I have frequently referred to the significance of this ob-
servation as a demonstration that a quite different order of communication had
emerged by 32 kyr.

The situation in southwest France is not quite so neat, for, although many claims
have been made about the most distinctively marked objects of the Early Upper Pa-
leolithic there (Bahn 1986; Delluc & Delluc 1978), no one has claimed that the so-
called “vulvas” engraved on stone blocks from seven sites in the Vezere region are or
represent personal decoration. But for our purposes we note that there was plenty of
evidence that personal decoration was going on, with the manufacture of beads
through the piercing of canines and other means (e.g., White 1989; this volume). At
the same time, the engraved blocks from La Ferrassie and the other six Vézere sites
are marked in ways that, like the Swabian statuettes, implies convention. The marks
which have caused most discussion, the so-called “vulvas,” might better be called
“enclosed grooves” to avoid the suggestion that they are or were representational. The
essence of the issue ts that there are blocks from seven sites roughly contemporary
with each other marked by “enclosed grooves” and that there is no agreement on what,
if anything, is represented by the “enclosed grooves” (Figure 8). Just as with the

THE POWER OF PICTURES 141

markings on the Swabian statuettes, there is an element of conventionalized arbitrari-
ness here which is strongly reminiscent of the distinctive difference between lan-
guage and other forms of communication. This is unsurprising given the conceptual
complexity of the Hohlenstein Stadel figure, which seems to combine a feline head, a
human body (that lacks sexual characteristics) with coded decoration on the arms.
This complexity seems to suggest the possibility that we must acknowledge the pres-
ence of language somewhere unknown earlier than 32 kyr. The Swabian and Vézere
evidence is nevertheless the earliest unequivocal evidence so far discovered for the
conventionalized packaging of arbitrariness in this way.

The Visibility of Early Symbolism? Two arguments follow from these observa-
tions. First, language seems likely to have emerged earlier than the first signs of it in
Europe, a position confirmed by the Australian evidence as we have argued at length

Geissenklosterle

Geissenklosterle

Vogelherd

Hohlenstein-Stadel Vogelherd

FiGuRE 7. Statuettes from Swabia showing “personal” markings. From Albrecht ev al. (1989)

142 DAVIDSON

elsewhere (Davidson & Noble 1992a). Secondly, the earliest language and hence
symbolism may have left no material trace. This second is something we must deal
with, as it involves a point of view that is distressingly widespread and is in the order
ofa religious statement: It is a statement of belief about the way the prehistoric world
was for which there is no evidence. Acceptance of this belief institutionalizes igno-
rance about the world because it discourages attempts to understand the world. The
early symbolism belief depends on taphonomic argument that all evidence earlier
than that we have found has now been destroyed. I may be an old fashioned empiri-
cist, but I believe that the best way to approach the world is through the assertion that
“the best evidence we've got is the evidence we’ve got.”

In the study of evolutionary processes, particularly those concerned with pro-
cesses of first emergence or “origins,” the very nature of an origin presents problems
of interpretation. Before the origin in question some feature was absent, and after that
origin, the feature was present. An emergence or origin necessarily involves the first
appearance of a behavior which previously was not manifest. Thus the change from
an absence toa presence of evidence can be crucial to the understanding of the process
we are studying. The problem is how to interpret absence of evidence, since the
sources of knowledge do not preserve all of the relevant evidence. Evidence may be
absent for no other reason than that the archaeological record is imperfect, what we
may call taphonomic imperfection; that is, an imperfection due to the processes of de-
struction resulting from the burial of material evidence of behavior. Evidence may
also be absent because the behavior, and hence its material manifestations, was ab-
sent. If we accept that taphonomic imperfection accounts for the absence of evidence,
then we are unable to distinguish between the time before emergence and the time af-
ter it. And no new evidence can make any difference to this state of affairs. Suppose
we postulated an origin of symbolism as the behavior which made the difference be-
tween the earliest ancestors of humans and those of chimpanzees. We would have to
choose to account for the lack of evidence for this hypothesis as due to the lack of ma-
terial consequences of symbolism combined with the taphonomic effects on such evi-
dence as was originally produced. If, however, we accept that, properly considered,
absence of evidence is evidence of absence, then new evidence can allow modifica-
tion of our hypothesis. Chase and Dibble (1987:47) conclude that the appropriate sci-
entific approach can be to accept the evidence of absence. This is the approach I have
just advocated for the consideration of the Australian evidence.

Duff, Clark and Chadderton (1992) adopt a seductive position in suggesting that
such phenomena as symbolism can have no material expression, therefore the ar-
chaeological record is an inappropriate source of evidence for addressing the prob-
lem. We might be attracted to an argument which says, rather, that there are problems
such as the origins of symbolism that should not be approached through the archaeo-
logical record, but Noble and I have been attempting to develop a theoretical under-
standing of the problems which allows us to understand the material consequences of
the emergence of symbolism. One of the elements of this argument is the coherence of
the interpretation of the sequence of events of Upper Paleolithic symbolism provided
by the perspective that language was newly emerged in Europe at the beginning of
that sequence.

Social Networks. Gamble has been one of the architects of our new vision of the
significance of the symbolism of the Upper Paleolithic. Using an extended ethno-
graphic analogy with the use of art and ceremony in harsh environments of Central
Australia in recent times, Gamble (1982) argued that the stylistic similarity of many

THE POWER OF PICTURES 143

Poisson 3 Ferrassie

FIGURE 8. “Enclosed groove” figures from the Vézére region, showing convention (courtesy of J.J. Clayet-

Merle, Les Eyzies Museum)

of the so-called “Venus figurines” of northern, central and eastern Europe implied the
existence of “open social networks” in the harsh periglacial environments of northern
Europe around 26 thousand years ago (Figure 9).

Soffer (1987) has criticized this argument, claiming that female figures were more
diverse and their chronology of occurrence less restricted than required by Gamble’s

144 DAVIDSON

argument. Gamble (1991) has replied with a defense of the basic thrusts of his argu-
ment, and I am in general sympathy with it. As Soffer (1987) has shown, there are
many prehistoric representations of female humans. Some of these have features in
common, though there is diversity even then, while others have little in common other
than that they represent human females. This wide diversity of figurines can be subdi-
vided according to the objectives of the analysis, and the criteria of recognition of
membership in one class or another.

What seems to me still unchallenged is that within the corpus of female figurines
there is still a group depicting naked women in three dimensions with no faces and no
feet. [tno longer seems tenable that there is a single restricted time period, which may
be as short as a couple of thousand years around 26 kyr, when there were few other
ways of depicting women. The figurines from Central Europe seem to be dated about
28 kyr to 24 kyr, while those from the Russian Plain are later than this date. While in-
terpretations of these dates are still open to discussion, there is some agreement that
there was a shift of people from Central Europe to the Russian plain (Grigor’ev 1993;
Soffer 1993).

Gamble’s (1982) argument was couched in terms of the value of such symbols in
constituting the open social networks that, by analogy with modern Australian Abo-
rigines dwelling in hot deserts, would have been necessary for humans learning sur-
vival in the harsh arctic desert environments of Europe. | do not want to enter into the
argument about whether such symbols do play a role in constituting social relations in
this way, rather I want to point to the veracity of Gamble’s claim for open social net-
works on the basis that there must have been a sufficient flow of information for the
conventions of depiction to be held in common over such large regions.

Female figurines before 20 kyr had conventions about representation in common.
These conventions may be, as Soffer (1987) claims, “superficial” and “not hold up to
closer scrutiny,” but they can be identified, and I would argue that the most important
ones are not those that Soffer refers to as “emotionally charged,” by which she pre-
sumably means the breasts and buttocks, so often commented on by (mostly male) ar-
chaeologists. While Gvozdover’s (1989) analysis does show that the stylistic
conventions of the Russian figurines is susceptible to much more detailed analysis,
this does not deny the existence of a set of general conventions within which these
styles were developed: The lack of faces and feet and the restriction of the depiction of
sex to females. Gamble is not alone in recognizing the similarity in such conventions
across Europe (Grigor’ev 1993), and, while the possibility of demographic shifts
makes sense for eastern Europe, it is still far from clear how the conventions might
have been shared between Central/Eastern Europe and the western region without
some of the openness that Iam supporting here.

There have been other claims for convention, most notably in the form of stone ar-
tifacts, especially Acheulean handaxes. We (Davidson & Noble 1993a) have argued,
however, that these conventions derive from learned means of manufacture in a con-
text where learning was primarily through observational learning and imitation rather
than through the language of instruction. Indeed, it is arguable that in such learning re-
petitive patterning is an advantage. Moreover, even if we do not accept that argument,
Dibble (1989) has shown that some of the regularity in Acheulean handaxe morphol-
ogy is either an illusion created by typology, through the division of a continuous
range of morphology into apparently discrete types, or an artifact of the method of
metrical analysis, through the definition that the longest dimension of handaxes
should always be named as the length. Such illusions and methodological fallacies do
not define the conventions of the “Venus figurines.” | would argue that the distribu-

THE POWER OF PICTURES 145

Es

Eliseevitchi Avdeevo Gagarino Kostenki

Willendort Vestonice Moravany

Lespugue Laussel Grimaldi Savignago

Ficure 9. “Venus” figurines from Europe. From Delporte (1979)

146 DAVIDSON

tion of the common features of handaxes is so wide in space and time that we are deal-
ing not with a conventional limitation on arbitrariness, but a situation where
arbitrariness hardly existed. We might be dealing with species-specific inability to
choose rather than culturally confined choices.

In the female figurines the general conventions are so precise that there is no doubt
for most of the figurines that we can say that there were no faces. This is not the parti-
tioning of continuous variation, but the straightforward identification of a conven-
tion, however superficial. Feet are a bit more of a problem, as some are broken, but
where the ends of the legs survive, the feet are either absent or rudimentary. Bear in
mind that the earlier statuette from Hohlenstein Stadel had both facial features and
feet, while some of the other Swabian figures have faces and others do not. Moreover,
the vast majority of the figurines are female. As Conroy (1993) has pointed out the
significant opposition is between female figurines and figurines with no representa-
tion of sex. These conventions were held in common over the region from the Atlantic
to the Don, north of the Mediterranean and the Alps, and rather restricted to a short pe-
riod before 20 kyr. The wide, but not unlimited, distribution of these conventions con-
trasts with the rather restricted distribution of the conventions of the Swabian
statuettes and the Dordogne enclosed grooves, though not for the items of personal
decoration.

In the matter of personal decoration, the female figurines in question are also re-
vealing. The most famous ofall, the Willendorf figure, has highly distinctive pattern-
ing on the head, usually interpreted as decoration. But the figurine also has clear,
though indistinct, markings on the wrists which I take to be indications of bracelets.
Marshack’s (1991) photograph shows that this ts in the form ofa zigzag (Figure 6).

And she is not alone. One of the essences of Soffer’s argument against Gamble is
that Gvozdover’s analysis of the Russian figurines shows “patterning in body decora-
tion” which is “duplicated in other media” (Gvozdover 1989). This sounds very much
like the situation described by Wiessner (1989) for the uses of style in modern socie-
ties. People, she asserts, use style to show their membership within a group associated
with the style, but seek to signal to other group members their separate identities
within the group through variations in the style. Wobst (1977) also wrote about the
way in which different aspects of style could be used to convey messages about group
affiliation and individuality. Wobst used as his examples the now tragic circum-
stances of the uses of hats for identification in Sarajevo (Wobst 1977: 333): “the Ba-
zaar of Sarajevo sports a large section of hatmakers in residence. The city is known for
the intense competition among its Serbian, Croatian and Muslim inhabitants.” His-
torical irony aside, the point is that personal decoration can serve as a marker of group
identification.

We can, of course, only guess what sorts of divisions of society and types of iden-
tity might have been signaled by the varieties of decoration on the figurines. But we
should be cautious about the types of division or identity that might have been present
at this time. Conroy’s (1993) argument suggests that the representation of female sex
of figurines might best be interpreted as the marker of the existence of gender in these
societies. We can all acknowledge that both male and female persons existed, the re-
striction of the sex of the image to one of these classes might seem to be an indication
that sex roles were now socially defined — that is to say gendered. The argument is
strengthened by the observation that in addition to the female figurines there are oth-
ers in which sex is not depicted at all. What seems to be being observed in some part of
the convention is a gendered opposition between female and other rather than one
which might be mere representation of the sexes female and male, and in the same

THE POWER OF PICTURES 147

way as other artistic systems routinely represent humans as either male, female or
without sex (see Davidson 1991b). This, itseems to me, destroys Soffer’s dismissal of
this convention as “superficial.” On the contrary, the convention turns out to reveal a
subtlety of social classification never observed in earlier material. I cannot say
whether this meets Gamble’s (1991) “challenge of explaining why the figurines were
female,” though I suspect it does.

At this stage, then, we can summarize the two parts of the sequence we have ana-
lyzed as showing that personal decoration was an early part of the complex of sym-
bolic representation in the Upper Paleolithic of Europe. This was the case in at least
two regions where different symbolic expressions occurred. Later evidence suggests
that decoration formed a means of identifying partitions within a widespread tradition
of a segmented social organization. But we could perhaps argue that identification is
of people rather than places. This is quite consistent with the argument I mentioned at
the start, that among people who have acquired language relatively recently (on an
evolutionary time-scale), there would be selective advantages to the development of
means of identifying group membership.

Plaquettes, Power and Restriction. | have already indicated the nature of the ar-
gument I produced on a previous occasion (Davidson 1989a) about the production of
pictures by painting and/or engraving on stone plaquettes. As opposed to the conven-
tions indicated by the female figurines before 20 kyr, the plaquettes (mostly after 20
kyr) indicate, by the restriction of information to one particular context at a restricted
number of sites, that access to information may have been controlled. The practice
can be dated at Parpallo to 22 kyr, and at probably the same date at Isturitz. By the time
of the glacial maximum the practice had spread to the Dordogne-Vézere region.
About 15-13 kyr it had spread to several other centers, the Central Pyrenees, Cantab-
ria, and northern France. At the end of the glacial period perhaps only Isturitz, and
G6nnersdorf, in western Germany, remained, though the upper levels at Parpall6
were badly disturbed, and Enlene might also belong to this final phase.

Unlike the earliest phase of European symbolism, this cannot be interpreted as re-
sulting from independent local traditions. The argument that the practices involving
plaquettes entailed restricted access to information requires common practices in re-
gions with separate centers of access to information. The pattern of segmentation of
social organization and the structuring of networks was quite different from that asso-
ciated with the earlier female figurines. The situation corresponds, in fact, to the pro-
cess of “intensive negotiation” that Gamble (1991) attributes to the phase of “Closed
Paleolithic Refugia” between 21 kyr and 13 kyr. But it is a situation that perhaps re-
quired an earlier stage of open social networks (as well as some “talking across the
fence,” as with all territorial boundary maintenance) for it to be achieved as a wide-
spread phenomenon.

There is a further implication of this interpretation. If there is any force at all to the
argument that painting and/or engraving on plaquettes was used in a context of dis-
semination and restriction of access to information, there 1s a necessary corollary that
what was involved concerned association of people from a wide area with a “center”
such as Parpallo. This argument is rather akin to Conkey’s discussion of aggregation
sites (Conkey 1980), in which people from different parts of Cantabria had a common
focus at Altamira. The most important implication of these arguments 1s that there
was some level of common ideology and identification among the people using quite
extensive regions. Within this extensive region, there were smaller regions with rela-
tively closed communication networks. Moreover, Soffer’s description (this volume)

148 DAVIDSON

of the regional elaboration of symbolic expression arising out of the “Venus” figurine
tradition in the East but not the West suggests that there were limits to the region of
common conventions.

From the perspective that is provided by viewing the changes as working out the
evolutionary consequences ofa late language emergence in Europe, these three stages
can be seen as a three part sequence.

1) The first language users in Europe discovered (or indeed brought with them)

means of defining personal identity, as the earliest Australians did.

2) As open social networks emerged, for whatever reason and by whatever means,

so too did gender identification and possibly other social partitioning which was

accompanied by distinctive visual marking through personal decorations.

3) This situation gave way to one where social structure involved more control of

access to information, involving a limitation of the use of space through control at

the center, integrated by a collective ideology at some level.

Cave Art. Thus far my argument has hardly mentioned the most famous manifes-
tation of Upper Paleolithic “art,” the painted and engraved caves. The reason is
straightforwardly chronological. The major part of this phenomenon may be later in
date than the Swabian statuettes, the Vézére blocks, the female figurines and the
emergence of plaquette power. This claim needs some justification, and we can pre-
dict that it will be tested to the full in the next few years as AMS dates for pigments be-
come more routine and more reliable. While this paper was in press, Grotte Chauvet
was discovered, published (Chauvet et a/. 1995) and direct and indirect AMS dates
were obtained (Clottes et al. 1995). The AMS dates for these paintings may be older
than any previously reported for painted depictions. As no details of the samples have
been published (yet), we cannot assess them, but note that other recent work has
shown the complex organic environment of prehistoric paints which may lead to inac-
curacies even in direct AMS dates (Ridges 1995).

The difficulties of dating Upper Paleolithic paintings have been well summarized
by Valladas and colleagues (1992). In particular the results reported in that paper con-
firm what most of us have long known: that stylistic dating is too imprecise to be of
much value. Dates from paintings of the “same” style were produced thousands of
years apart. The dates for Grotte Chauvet estimated from stylistic considerations may
be as much as 10 thousand years later than the AMS dates. In addition, indirect dates
of materials derived from human activity in the cave but not from the painting itself
are also imprecise. Painted caves were used and visited before and after the date of the
paintings. Asa result, the dates for paintings in several painted caves are rather later in
date than might have been expected (see Table 1) while others were earlier.

The implications of these are too many to discuss here. I limit myself to two obser-
vations. First, it would be tempting to relate the pattern of distinct use of regions that I
have suggested can be derived from the understanding of plaquette power to the very
closely related divisions of regions revealed by Jochim’s (1983) analysis of the
chronological and spatial distribution of painted caves. But part of the patterning re-
vealed in Jochim’s analysis derived from the use of Leroi-Gourhan’s (1965) defini-
tions of four styles of the paintings. I have always preferred to avoid arguments based
on this stylistic classification. The new evidence suggests that that caution is well
placed, and that it would be premature to change that preference.

Secondly, it has long been an argument that the earliest figures in Upper Paleo-
lithic Cave Art are the hand stencils, including those from Gargas. This seems to be
confirmed by the early dates from Cosquer Cave (Clottes & Courtin 1994). The pro-

THE POWER OF PICTURES 149

Table 1. AMS dates for Upper Paleolithic figures

Site _ Date -Labno | Figure Reference si
Cougnac —-14. 300 + 180 Gif A 89250 Dot Lorblanchet ef al. 1990
13 810+ 210 Gif A 92500 Dot Clottes in press
2361514351 Gif A 91183 Male Megaceros Clottes in press
22 750 + 390 Gif A 92426 Male Megaceros Clottes in press
19 498 + 267 Gif A 91324 Female Megaceros Clottes in press
25 120 + 390 Gif A 92425 Female Megaceros Clottes in press
Altamira 13 570+ 190 Gif A 91 178 Bison Valladas et al. 1992
14 410 + 200 Gif A 91 249 Valladas et al. 1992
13 940 + 170 Gif A 91 179 Bison Valladas etal. 1992
4710+ 200 Gif A 91 254 Valladas etal. 1992
4 330 + 190 Gif A 91 181 Bison Valladas etal. 1992
4 250 + 180 Gif A 91 330 Valladas etal. 1992
13 060 + 200 Gif A 91 004 Valladas et al. 1992
2 190 + 180 Gif A 91 172 Valladas etal. 1992
Castillo 3 060 + 200 Gif A 91004 Bison Valladas etal. 1992
2910+ 180 Gif A 91172 Bison Valladas etal. 1992
Niaux 2 890 + 160 Gif A 91 319 Bison Valladas etal. 1992
2 440 + 190 Gif A 91 173 Valladas etal. 1992
3 060 + 200 Gif A 92499 Line Valladas etal. 1992
3 850 + 150 Gif A 92501 Bison Valladas eral. 1992
Cosquer 27 110 + 390 Gif A 92 409 Hand stencil Clottes ef al. 1992
27 110 +350 Gif A 92 49] Hand stencil Clottes ef al. 1992
9 200 + 220 Gif A 92418 Feline head Clottes et al. 1992
8 820 + 310 Gif A 92 417 Horse Clottes et al. 1992
8 840 + 240 Gif A 92 416 Horse Clottes et al. 1992
8 010+ 190 Gif A 92419 Bison Clottes et al. 1992
8 530+ 180 Gif A 92 492 Bison Clottes ef al. 1992
Le Portel 2,180 + 125 AA-9465 Horse Igler et al. 1994
11,600 + 150 AA-9766 Horse Igler et al. 1994
24,640 + 390 Gif A 95 357 Horse Lorblanchet ef al. 1995
30,940 + 610 Gif A 95 126 Rhino | Clottes et al. 1995
30,790 + 600 Gif A 95 133 Rhino 2 Clottes et al. 1995
32,410 + 720 Gif A 95 132 Rhino 2 Clottes et al. 1995
30,340 + 570 Gif A 95 128 Bison Clottes ef al. 1995

duction of hand stencils is one of the truly universal features of human behavior, pro-
duced all over the world and at almost all times since there have been humans. Yet, I
venture to suggest that none of us knows why people put stencils of hands on surfaces.
And I do not believe that we can ignore the significance of our ignorance. For those
marking the walls of Gargas and Cosquer Cave with hand stencils a “meaning” may
have been clear, but it would not have been so to any other person who was not a wit-
ness to that act, any more than it is to us. To be sure, marking the walls inside caves in-
dicates to another person venturing into the darkness of the cave equipped with
adequate lighting and paying attention to the walls of the cave rather than the floor,
that someone else has been there, but I do not think that it can be said to do much more.
It may have been enough.

But these two examples, together with my own work with the plaquettes suggests
a, not highly original, view that cave paintings may have been involved in “territorial
adjustments.” It has always been a mystery to me how these may have worked. A re-
cent review (Gosling 1982) of the use of scent marking in the territories of nonhuman
animals provides a valuable synthesis of the sorts of argument about the function of
territorial markers. This is not the place to summarize these arguments. Suffice it to
say that ethologists have now gone a long way beyond the view that territorial mark-
ers are concemed primarily with defense or advertisement. This can only be liberating
in the study of Paleolithic Cave Art, as it has always seemed implausible that paint-

150 DAVIDSON

ings hidden out of sight deep in caves could possibly be concerned with defense or ad-
vertisement of territory.

Itseems more likely that such paintings could only be involved with territory if, in
some way, the important point was that the paintings enabled people to identify some
aspect of the painters. Gosling (1982) argues that scent marking allows owners to be
identified. On this evidence, it should not be necessary for people to use cultural
markers of their identity, since other animals are adapted to identify sufficient infor-
mation about each other by means of the distinctive emissions from their bodies. At
the beginning of this paper, | referred to an argument that Noble and I have developed
(Davidson & Noble 1992a: Noble & Davidson 1993a), in which the arbitrariness in-
volved in language as distinct from other communication systems results in the acci-
dental or deliberate utterance of untrustworthy communications. In this context,
given that there are physical limits on the numbers of individuals with whom one can
communicate ona regular basis, identification of individuals with whom conventions
on the limits to arbitrariness are shared is an essential requirement for efficient com-
munication through language. It may be that visual cues provide a more rapid identifi-
cation than cues based on distinctive patterns of speech. Thus, we might expect that
individual identification through personal decoration might be a feature of human
groups early in the history of emergence of language. The significance of this can
only be identified by a full understanding of the process of Upper Paleolithic “tart.”

Summary of the Process of Emergence of Upper Paleolithic Art

The evidence I have discussed seems to show that, within a time period which was
short relative to the period of distinctiveness of hominids, in two very separate parts
of the world humans invented means of marking individuals. In the European case,
the sequence of events that followed seems to suggest that identification by this
means played a role in identifying segments of society when open social networks
presumably made mobility between groups more frequent. Subsequently there was
an emergence and spread of more closed social networks, which, from the manner in
which we have identified them, seem to have involved the establishment of associa-
tion between people and some form of center of their world. We might, to use the ex-
pression used by Rosenfeld about Australia, envisage this as the beginnings of some
form of corporate identity of groups. Following from this, the marking of places
through symbols painted and engraved on cave walls might be involved with the iden-
tification of the manner of corporate ownership of territories.

The evidence is more fragmentary in Australia than in Europe, but the two ends of
the sequence are clear. At the start there is personal marking, at the end there 1s mark-
ing of places. The European evidence suggests that the two ends may have been con-
nected by a complex process of working out the corporate identity of groups and their
relations to places. Is the pattern repeated elsewhere?

Pattern of Process in Other Regions?

I would have hoped to compare the sequence of events with that in other regions,
beginning with Southern Africa. I was disappointed that Lewis-Williams’ argument
(Lewis-Williams & Dowson 1988) about altered states of consciousness and the “orl
gins of art” was not framed in terms of the earliest South A frican evidence. My limited
understanding of South Africa suggests that the evidence for the earliest “art” 1s cur-

THE POWER OF PICTURES 151

rently even more fragmented than that of Australia, but certainly involved the early
appearance of artifacts with imposed form, and an early appearance of paintings on
plaquettes. Painting and engraving of places such as caves and rock shelters seems to
be later. Wadley (1993) has summarized the evidence from the Later Stone Age of
South Africa, and shows that there is indeed a case for the early appearance of per-
sonal ornamentation and the much later appearance of marking of places (Figure 6).
There 1s even a hint that plaquettes were used in between, though the evidence 1s so
fragmentary that it cannot be regarded as sufficient confirmation of the hypothesis.

There is some hint that, appropriately addressed, the evidence from southern Af-
rica might yield a sequence compatible with the scenario I am suggesting developed
in Europe and Australia as people, newly endowed with language, gradually evolved
means of dealing with the creativity it provided. The same cannot be said for many
other regions of the world, but some of this may be due to more fragmentary attention
to the relevant archeological record.

The region where it cannot be said that attention to the archeological record has
been fragmentary is the East Mediterranean. Yet here there is little sign that the out-
lined sequence can be identified, largely due to the lack of evidence of marking
places. There may be some sign of decoration, and one or two hints of something akin
to plaquettes, but no cave or shelter art or similar marking of place until artificially
constructed places begin to be so marked in the Neolithic. Bar-Y osef has summarized
the evidence (or lack of it) in this volume.

I have suggested previously that we should take seriously the possibility that it 1s
no coincidence that in the east Mediterranean the Late Pleistocene archaeology is be-
reft of signs of symbolism and at the end of the Pleistocene agriculture emerged,
while in the west the Late Pleistocene archaeology is full of signs of symbolism and
agriculture was only introduced after it emerged in the east (Davidson 1989b). This is
not the time or the place to elaborate that argument (Figure 10). Nevertheless, this
view of the sequence in Australia and Europe involves the working out by the first
people in those continents of what may be interpreted as a collective appropriation of
nature (cf. Ingold 1980). In failing to work out this ideological position or choosing
not to, people in other parts of the world may have left themselves free to develop
more individualized attitudes to appropriation, leading to the ideologies of agricul-
ture.

One difficulty for this argument is the independent invention of agriculture in the
Americas. Whilst not ruling out some similar argument, the hypothesis that the
Americas were colonized late might suggest that the relationships being described
here were already worked out before the colonization. Thus the nature of the pro-
cesses involved in the differentiation of fisher-gatherer-hunter and agricultural ways
of life might have been quite different from those of the Old World.

Conclusion

I suggest that there 1s a case fora common sequence of Late Pleistocene evolution-
ary change in the manner of use of symbols in two separate continents, Australia and
Europe, and a possibility ina third, southern Africa. There are some regions, notably
southwest Asia, where, despite the detail of the archaeological record this sequence
does not seem to have occurred. The Upper Paleolithic cannot, therefore, be taken as
the standard but there is comparable patterning in some other parts of the world of the
Late Pleistocene I have identified a common thread in the sequence of changes in Up-

(L961) uepeH
Aq pauljap UIAHO [eINyNIURe Jo , stayUAgUOU,, puke $19}U99 ay} 0} UONLIAL UI (166 Uyeg] wo uayey are yxa) a} UL passnosip jou asoy}) sas jue,, AJIva yo dey “QT TWADTY
DIWeIID

EDIeye)
pajdinos 10 peaesBug

awoo

payed

O€4 49D
eeyeuy

61 aaeg
juua
z HI » 92
5 G1 URW Aue3 d\ tt o1lody
n ‘
a) SI 7c O€s || elueunyejew
= yaai9 Apues Ooz eliqejemen
<
a)
0¢
m—~_pJEWeWP |
20
f... _6{'8¢
aaeg sanbso9
ZS
\
Ze aisseuay e7

N

THE POWER OF PICTURES 153

per Paleolithic art and in Australian prehistoric art that enables us to understand the
significance of both in the evolution of human behavior. I suggest it tends to confirm
the case Noble and I have made for a late language origin.

There is no inevitability to the sequence of evolutionary change in artistic systems.
Symbolism was part and parcel of turning communication into language, but the use
of symbols separate from language could only have been a product of language. Since
the essence of symbols and language is conventionally defined arbitrariness, the pro-
ductivity and creativity we associate with humans 1s very much a product of language.
Because of this the pattern of symbol production need not have patterning to it from
one region to another. And yet, I suggest, it does. The limit may turn out (I do not
claim that the argument is complete yet) to have something to do with the limited op-
tions in the appropriation of nature that distinguish fisher-gatherer-hunters with col-
lective appropriation from agriculturalists and pastoralists with individual
appropriation (Ingold 1980). If 1 am right, then we should expect that the Upper Pa-
leolithic pattern would be repeated in other regions of the world populated by fisher-
gatherer-hunters with collective appropriation of nature, and different patterns would
be found in those regions where agriculture emerged independently.

On this view we see the grandest vision of what we can make this art mean. In such
an evolutionary argument it signifies the way humans first worked out their relations
with land and with each other once they had discovered language. It means the oppo-
sition between fisher-gatherer-hunters and agriculturalists and thus it means the larg-
est themes of human history: the conflict between different ways of life and
dispossession through colonization. The influence of this, | suggest, was more funda-
mental than that of Pheidias, Leonardo, Shakespeare or Beethoven.

Acknowledgments

I would like to thank the organizers of the symposium for which this paper was
prepared, Meg Conkey and Olga Soffer, for their generosity in the invitation and Jean
Auel for her generosity in providing assistance with fares and for her hospitality in the
First Oregon Archeological Retreat. As usual I also owe a debt of gratitude to Bill No-
ble without whom my thinking on these matters would be very much less well devel-
oped. In the development of these ideas, I also gratefully acknowledge assistance in
various ways and at various times from Helen Arthurson, Peter Jarman, Clive Gam-
ble, Chris Lovell-Jones, Alexander Marshack, Andrée Rosenfeld, Mike Morwood,
Alan Watchman, Linda Conroy, Heather Burke and Doug Hobbs. I should also ac-
knowledge Peter Jim without whom this paper would have been written sooner and
later in the day. Otherwise it is all my own work, and I take the brickbats.

Literature Cited

Albrecht, G., G. Bosinski, R. Feustel, J.Hahn, B. Klima & H. Muller-Beck. 1989. Los Comien-
zos del Arte en Europa Central. Museo Arqueologico Nacional, Madrid, Spain.

Aparicio, J. 1981. Nueva pieza de arte mobiliar parpallense. Archivo de Prehistoria Levantina
16:39-58.

Arensburg, B., L. A. Shepartz, A. M. Tillier, B. Vandermeersch & Y. Rak. 1990. A reappraisal
of the anatomical basis for speech in Middle Palaeolithic hominids. Amer. J. Phys. An-
thropol.83:137-146.

154 DAVIDSON

Arensburg, B., A. M. Tillier, B. Vandermeersch, H. Duday, L. A. Schepartz & Y. Rak. 1989. A
Middle Palaeolithic human hyoid bone. Nate 388:758-760.

Aujoulat, N., C. Barriere, R. Begouen, J. M. Bouvier, J. Clottes, M. Cremades, M. Dauvois, V.
Ferughio, J. Gaussen, M. Lorblanchet, Y. Martin, F. Rouzaud, D. Sacchi & G. Sauvet. 1993.
L’Art Pariétal Paléolithique. Techniques et Méthodes d'Etude. Ministére de |’ Enseigne-
ment Superieur et de la Recherche, Paris, France.

Bahn, P. G. 1986. No sex please, we’re Aurignacians. Rock Art Res. 3:99-120

___. 1991. Pleistocene images outside Europe. Proc. Prehist. Soc. 57:91-102.

Bahn, P.G. & J. Vertut. 1988. /mages of the Ice Age. Thames and Hudson, London, U.K.

Balbin Behrmann, R. D. & J. A. Moure Romanillo. 1981. La “Galeria de los Caballos” de la
Cueva de Tito Bustillo Pages 85-118 in Alramira Symposium. Mimsterio de Cultura, Ma-
drid, Spain.

Barbujani, G. & R. R. Sokal. 1990. Zones of sharp genetic change in Europe are also linguistic
boundaries. Proc. Natl. Acad. Sct. U.S.A, 87:1816-9

Begouen, R. & J. Clottes. 1991. Portable and wall art in the Volp Caves, Montesquieu-A vantes
(Ariege). Proc. Prehist. Soc. 57:65-79.

Bofinger, E. & I. Davidson. 1977. Radiocarbon age and depth: A statistical treatment of two se-
quences of dates from Spain. J. Arch. Sci. 4: 231-243.

Bosinski, G. 1991. The representation of female figures in the Rhineland Magdalemian. Proc.
Prehist. Soc. 57:51-65.

Bowler, J. M. & A. Thorne. 1976. Human remains from Lake Mungo: discovery and excava-
tion of Lake Mungo IIL. Pages 127-38 in R. L. Kirk & A. G. Thorne, eds., The Origin of the
Australians. Australian Institute of Aboriginal Studies, Canberra, Australia.

Chase, P.G. 1991. Symbols and paleolithic artefacts: style, standardization, and the imposition
of arbitrary form. J. Anthropol. Arch..10:193-214

Chase, P. G. & H. L. Dibble. 1987. Middle Palaeolithic symbolism: A review of current evi-
dence and interpretations. J. Anthropol. Arch. 6:263-296.

Chauvet, J.-M., E. Brunet-Deschamps & C. Hillaire. 1995, La grotte Chauvet a vallon Pont-
d'Arc. Seuil, Paris, Paris.

Cheney, D. L. & R. M. Seyfarth. 1990. How Monkeys See the World. The University of Chi-
cago Press, Chicago.

Clottes, J., J.-M. Chauvet, E. Brunel-Deschamps, C. Hillaire, J.-P. Daugas, M. Arnold, H.
Cachier, J. Evin, P. Fortin, C. Oberlin, N. Tisnerat & H, Valladas. 1995. Radiocarbon dates
for the Chauvet-Pont-d’ Are Cave. Intl. Nwsltr. Rock Art 11:1-2.

Clottes, J. & J. Courtin. 1994. La Grotte Cosquer. Peintures et Gravures de la Caverne
Engloutie. Seuil, Paris, France.

Clottes, J., J. Courtin & H. Valladas. 1992. A well-dated palaeolithic cave: the Cosquer Cave at
Marseille. Rock Art Res. 9:122-129.

ClottesJ., J-M. Chauvet, E. Brunez-Deschamps, C. Hillaire, J-P. Daugas, M. Arnold, H. Cach-
ier, J.Evin, P. Fortin, C.Oberlin, N. Tisnerat & H. Valladas 1995. Les peintres paléo-
lithique de la grotte Cahuvet-Pont d'Are, a Vallon-Pont-d'Are (Ardéche, France): datations
directes et indirectes pas le méthode du radiocarbone. Comptes Rendus de l'Académie Sci-
ences, serie lla, 320:1133-1140.

Cole, N.A., A. Watchman & M.J. Morwood. 1995. Chronology of Laura rock art. Pages 147-
160 in M.J. Morwood & D.R. Hobbs, eds., Quinkan Prehistory. The Archaeology of Abo -
riginal Artin SE. Cape York Peninsula, Australia. Tempus Vol.3, Anthropology Museum,
University of Queensland, St. Lucia, Australia.

Conkey, M. W. 1978. Style and information in cultural evolution: toward a predictive model
for the Paleolithic. Pages 61-85 in C. L. Redman, M. J. Berman, E. V. Curtin, J. William T.

nN

THE POWER OF PICTURES i)

Langhorne, N. M. Versaggi & J.C. Wanser, eds., Social Archaeology Beyond Subsistence
and Dating. Academic Press, New Y ork.

. 1980. The identification of prehistoric hunter-gatherer aggregation sites: The case of
Altamira. Curr. Anthropol. 21:609-630.

. 1987. New approaches in the search for meaning? A review of research in “Paleolithic
Art”. J. Field Arch. 14:413-430.

. 1989. The structural analysis of Paleolithic art. Pages 135-154 in C. C. Lamberg-
Karlovsky, ed., Archaeological Thought in America. Cambridge University Press, Cam-
bridge, U.K.

Conroy, L. 1993. Female figurines of the Upper Palaeolithic and the emergence of gender.
Pages 153-160 in H. du Cros & L. Smith, eds., Women in Archaeology. A Feminist Critique.
Australian National University, Canberra, Australia.

D’errico, F. & C. Cacho. 1994. Notation versus decoration in the Upper Palaeolithic: A case
study from Tossal de la Roca. J. Arch. Sci. 21:185-200.

Davidson, I. 1974. Radiocarbon dates for the Spanish Solutrean. Antiquity 48:63-65.

. 1989a. Freedom of information: aspects of art and society in western Europe during the
last ice age. Pages 440-456 in H. Morphy, ed., Animals into Art. Unwin Hyman, London,
ULK.

. 1989b. Is intensification a condition of the fisher-hunter-gatherer way of life? Arch
Oceania 24:75-78.

. 1989c. La Economia del Final del Paleolitico en la Espana Oriental. Diputacion Pro -
vincial de Valencia, Valencia, Spain.

. 1990. Bilzingsleben and early marking. Rock Art Res. 7:52-56.

.1991a. The archaeology of language origins. Antiquity 65:39-48.

. 1991b. Marginal notes: a male view from the Ist Women in Archaeology conference,
Albury, February 1991. Austr. Arch. 32:41-43.

Davidson, I. & J. Estevez. 1985. Problemas de arqueotafonomia. Formacion de yacimientos
con fauna. Quaderns 1:67-84.

Davidson, I. & W. Noble. 1989. The archaeology of perception: Traces of depiction and lan-
guage. Curr. Anthropol. 30:125-155.

. 1992a. Why the first colonization of the Australian region is the earliest evidence of
modern human behaviour. Arch. Oceania 2:135-142.

. 1992b. Language gap. Nature 355:403.

. 1993a. Tools and language in human evolution. in K. Gibson & T. Ingold, eds., Tools,
Language and Cognition in Human Evolution. Cambridge University Press, Cambridge,
U.K.

Delluc, B. & G. Delluc. 1978. Les manifestations graphiques aurignaciennes sur support
rocheux des environs des Eyzies (Dordogne). Gallia Préhistoire 21:213-438.

Delporte, H. 1979. L ‘Image de la Femme dans |'Art Préhistorique. Picard, Paris, France.

Dibble, H. L. 1989. The implications of stone tool types for the presence of language during the
Lower and Middle Paleolithic. Pages 415-431 in P. Mellars & C. Stringer, eds., The Human
Revolution: Behavioral and Biological Perspectives on the Origins of Modern Humans.
Edinburgh University Press, Edinburgh, U.K.

Dortch, C. 1984. Devil's Lair, A Study in Prehistory. Western Australian Museum, Perth,
Australia,

Dragovich, D. 1986. Minimum age of some desert varnish near Broken Hill, New South Wales.
Search 17:149-151.

Duff, A. I.,G. A. Clark & T. J. Chadderton. 1992. Symbolism in the Early Palaeolithic: Acon-
ceptual odyssey. Cambridge Arch. J. 2:21-229.

156 DAVIDSON

Dunbar, R. 1. M. 1993. Coevolution of neocortical size, group size and language in humans. Be-
hav. Brain Sci. 16:681-735.

Falk, D. 1983. Cerebral cortices of East African early hominids. Science 221:1072-1074.

__. 1987. Hominid paleoneurology. Ann. Rev. Anthropol. 16:13-30.

Fortea, J. & F. Jorda. 1976. La Cueva de Les Mallaetes y los problemas del Paleolitico Supe-
rior del Mediterraneo espanol. Zephyrus 26-27: 129-166.

Fortea Perez, J. 1978. Arte Paleolitico del Mediterraneo espanol. Trabajos de Prehistoria 35:
99-149.

Fullola I Pericot, J. M., R. Vinas I Vallverdu & P Garcia Arguellas | Andreu. 1990. La nouvelle
plaquette gravee de Sant Gregori (Catalogne, Espagne). Pages 279-285 in J. Clottes, ed.,
L'Art des Obyets au Paléolithique. Tome 1: L'Art Mobilier et son Contexte. Ministére de la
Culture, Paris, France.

Gamble, C. 1982. Interaction and alliance in Palaeolithic society. Man: J. Roy. Anthropol. In-
stit. 17:92-107.

___. 1991. The social context for European Palaeolithic art. Proc. Prehist. Soc. 57:3-15.

Gosling, L.M. 1982. A reassessment of the function of scent marking in territories. Zeitschrift
fiir Tierpsychologie 6:89-118

Grigorev, G. P. 1993. The Kostenki-Avdeevo archaeological culture and the Willendorf-
Pavlov-Kostenki-Avdeevo cultural unity. Pages 51-65 in O. Soffer & N. D. Praslov, eds.,
From Kostenki to Clovis: Upper Paleolithic-Paleo-Indian Adaptations. Plenum Press,
New York.

Gvozdover, M. 1989. The typology of female figurines of the Kostenki Palaeolithic culture.
Soviet Anthropology and Archaeology 27:32-95.

Hadingham, E. 1980. Secrets of the Ice Age. The World of the Cave Artists. Heinemann, Lon-
don, U.K.

Hahn, J. 1972. Aurignacian signs, pendants and art objects in Central and Eastern Europe.

World Arch. 3:252-266.

___. 1986. Kraft und Aggression. Die Botschaft der Eiszeitkunst im Aurignacien Stid-

deutschlands? Institut Fiir Urgeschichte der Universita, Tiibingen, Germany.

Harlan, J. 1967. A wild wheat harvest in Turkey. Archaeology 20:197-201.

Holloway, R. L. 1983. Human paleontological evidence relevant to language behaviour. Hum.
Neurobiol. 2:105-114.

Igler, W. M.Dauvois, M. Hyman,M.Menu,M.Rowe, J.Vezian & P. Walter, 1994. Datation ra-
diocarbone de deux figures pariétales de la grotte du Portel (Commune de Loubens,
Ariege). Bull. de la Soc. Préhist. de l'Arieége, XLIX:231-236.

Ingold, T. 1980. Hunters, Pastoralists and Ranchers. Cambridge University Press, Cambridge,
U.K.

Jochim, M. 1983. Palaeolithic cave art in ecological perspective. Pages 212-219 in G.N. Bai-
ley, ed., Hunter-gatherer Economy in Prehistory. A European Perspective. Cambridge
University Press, Cambridge, U.K

Leroi-Gourhan, A. 1965. Treaures of Prehistoric Art. Abrams, New York.

Lewis-Williams, D. J. 1983. The Rock Art of Southern Africa. Cambridge University Press,
Cambridge, U.K.

Lewis-Williams, D. J., & T. A. Dowson. 1988. The signs of all times: Entoptic phenomena in
Upper Palaeolithic art. Curr. Anthropol. 29:201-245.

Lieberman, P. 1991. Uniquely human. The Evolution of Speech, Thought, and Selfless Behav -
vor. Harvard University Press, Cambridge.

Lieberman, P., J. T. Laitman, J. S. Reidenberg & P. J Gannon. 1992. The anatomy, physiology,
acoustics and perception of speech: Essential elements in analysis of the evolution of hu-
man speech. J. Hum. Evol. 23:447-467.

THE POWER OF PICTURES 157

Lindly, J. 1988. Hominid and carnivore activity in Middle and Upper Palaeolithic cave sites in
eastern Spain. Munibe 40:45-70.

Lorblanchet, M., M. Labeau, J. L. Vernet, P. Fitte, H. Valladas, H. Cachier & M Arnold. 1990.
Palaeolithic pigments in the Quércy, France. Rock Art Res. 7:4-20.

Lorblanchet, M., H. Cachier & H. Valadas 1995. Datation des cheveaux ponctués du Pech-
Merle. /ntl. Nwsltr. Rock Art 12:2-3

Loy, T. 1994. Direct dating of rock art at Laurie Creek (NT), Australia: A reply to Nelson. An-
tiquity 68:147-148.

Loy, T., R. Jones, D. E. Nelson, B. Meehan & J. Vogel. 1990. Accelerator radiocarbon dating of
human blood proteins in pigments from Late Pleistocene art sites in Australia. Antiquity
64:110-116.

Macintosh, N. W. G. 1977. Beswick Creek cave two decades later: a reappraisal. Pages 191-
197 in P. J. Ucko, ed., Form in Indigenous Art. Schematisation in the Art of Aboriginal Aus -
tralia and Prehistoric Europe. Australian Institute of Aboriginal Studies, Canberra, Aus-
tralia.

Marshack, A. 1991. The female image: a “time-factored” symbol. A study in style and aspects
of image use in the Upper Palaeolithic. Proc. Prehist. Soc. 57:17-31.

Maynard, L., & R. Edwards. 1971. Wall markings. Pages 61-80 in R. V.S. Wright, ed., Archae-
ology of the Gallus Site, Koonalda Cave. Australian Institute of Aboriginal Studies, Can-
berra, Australia.

Mcdonald, J., K. Officer, T. Jull, D. Donahue, J. Head & B. Ford. 1990. Investigating 8
AMS: dating prehistoric rock art in the Sydney Sandstone basin, Australia. Rock Art Res. 7:
83-92.

Morphy, H. 1991. Ancestral Connections. Art and an Aboriginal System of Knowledge. The
University of Chicago Press, Chicago.

Morse, K. 1993. Shell beads from Mandu Mandu Creek rockshelter, Cape Range peninsula,
Western Australia, dated before 30,000 b.p. Antiquity 67: 877-83.

Morwood, M.J., D.R. Hobbs & D.M. Price. 1995. Excavations at Sandy Creek | and 2. Pages
71-91 in M.J. Morwood and D.R. Hobbs, eds., Quinkan Prehistory. The Archaeology of
Aboriginal Artin S.E. Cape York Peninsula, Australia. Tempus Vol. 3, Anthropology Mu-
seum, University of Queensland, St. Lucia, Australia.

Muller-Karpe, H. 1966. Handbuch de Vorgeschichte. Tome |. Altsteinzeit. Beck, Munchen,
Germany.

Murray, P. & G. Chaloupka. 1984. The dreamtime animals: Extinct megafauna in Arnhem
Land rock art. Arch. Oceania 19:105-116.

Nelson, D. E. 1993. Second thoughts on a rock-art date. Antiquity 67:893-895.

Noble, W. & I. Davidson. 1989. On depiction and language. Curr. Anthropol. 30:337-342.

. 199 1a. The evolutionary emergence of modern human behaviour: Language and its ar-
chaeology. Man 26:223-253.

. 1993a. Tracing the emergence of modern human behaviour: Methodological pitfalls,
and a theoretical path. J. Anthrop. Arch.12:121-149.

. 1996. Human Evolution, Language and Mind. Cambridge University Press, Cam-
bridge, U.K.

Olaria, C., F. Gusi, J. Estevez, J. Casabo & M.-L Rovira. 1981. El yacimiento magdaleniense
superior de Cova Matutano (Villafames, Castell6n): Estudio del sondeo estratigrafico
1979. Cuadernos de prehistoria y arqueologia Castellonenses 8:21-100.

Pericot Garcia, L. 1942. La Cueva del Parpallo. Consejo Superior de Investigaciones Cientifi-
cas, Madrid, Spain.

Ridges, M. 1995. An investigation into the Aboriginal rock art paints of the Sewlyn Ranges re -
gion in North West Queensland. B.A..Thesis, University of New England, Australia.

158 DAVIDSON

Ripoll Lopez, S. & C. Cacho Quesada. 1990. Art mobiher du Paleolithique méditerranéen es-
pagnok quelques nouvelles decouvertes. Pages 287-293 in J. Clottes, ed., L Art des Objets
au Paleolithique. Tome 1: L’Art Mobilter et son Contexte. Ministere de la Culture, Paris,
France.

Roberts, RG. R. Jones & M.A Smith. 1990. Thermoluminescence dating of a 50,000-year-
old human occupation site in northern Australia. Nature 345: 153-156.

Rosenfeld, A. 19ST. Rock engravings in the Laura area. Pages 50-89 in Rosenteld, D. Horton &
J. Winter, eds., Early Man in North Queensland. A Terra Australis, Australian National
University, Canberra, Australia.

—__. 1993. A review of the evidence for the emergence of rock art in Australia. Pages 71-80
mM. A. Smith, M. Spriggs & B. Fankhauser, eds., Safu/ in Review: Pleistocene Archae-
ology in Australia, New Guinea and Island Melanesia. Australian National University,
Canberra, Australia.

Schepartz, L. A. 1993. Language and modern human origins. Yrbk. Phys, Anthropol. 36:91-
126.

Simonnet, G., L. Simonnet & R. Simonnet. 1990. Art mobilier et art parietal a Labastide. i J.
Clottes, ed., L Art des Ohjets au Paléolithique. Tome 1; L’Art Mobilier et son Contexte
Ministere de la Culture, Paris, France.

Softer, O. 1987. Upper Paleolithic connubia, refugia, and the archaeological record from East-
ern Europe. Pages 333-348 iO. Soffer, ed., The Pleistocene Old World. Regional Perspec-
fives, Plenum Press, New York.

1993, Upper Paleolithic adaptations in Central and Eastern Europe and Man-Mammoth
interactions. Pages 31-49 in O. Sotfer & N. D. Praslov, eds., From Kostenki to Clovis. Up-
per Paleolithic-Paleo-Indian Adaptations. Plenum Press, New York.

Tobias, P. V. 1987. The brain of Homo habilis: A new level of organization in cerebral evolu-
tion. J. Hum. Evol. 16:741-761

Ucko, P. J. & A. Rosenfeld. 1967, Palaeolithic Cave Art. Weidenteld and Nicolson, London.
U.K

Valladas, H., H. Cachier, P. Maurice, F. Bernaldo de Quiros, J. Clottes, V. Cabrera Valdes
Uzquiano, P. & M. Arnold. 1992. Direct radiocarbon dates for prehistoric paintings at the
Altamira, El Castillo and Niaux caves. Nature 357:08-70.

Van de Merwe, N. J., J. Sealy & R. Yates. 1987. First accelerator carbon-14 date tor pigment
from a rock painting. S Afr J) Sci. 83:56-57,

Villaverde Bonilla, V. 1990. Animation et scenes sur les plaquettes du Parpallo (Gandia, Es-
pagne): quelques considerations sur la pictographie dans art mobilier. Pages 227-241 in J.
Clottesed., L drt des Objets au Paléolithique. Tome 2: Les Votes de la Recherche. Min-
istere de la Culture, Paris, France.

1992. Principaux traits evolutifs de la collection dart mobilier de la grotte del Parpallo,
L Anthropologie (Parts) 96:375-396

Villaverde Bonilla, V. & B. Marti Oliver. 1984. Paleolitic ¢ Epipaleolitic. Les Societats Caga-
dores de la Prehistoria Valenciana. Server d Investigacio Prehistorica. Diputacio de Va-
lencia, Valencia, Spain.

Vinmcombe, P. 1972. Myth, motive and selection in southern African rock art. Africa 42:192-
204.

Wadley, L. 1993. The Pleistocene later Stone Age south of the Limpopo River. J. World Pre-
hist. 7:243-296

Watchman, A. 1992. Doubtful dates for Karolta engravings. Austr. Aborig. Stud. 1992:51-55.

Wendt, W. bE. 1974. Artmobilier” aus der Apollo | 1-Grotte in Sudwest-A frika. Acta Praehist
et Arch. 5:\-42.

THE POWER OF PICTURES 159

White, R. 1989. Production complexity and standardisation in early Aurignacian bead and pen -
dant manufacture: evolutionary implications. Pages 366-390 in P. Mellars & C. Stringer,
eds., The Human Revolution. Behavioural and Biological Perspectives on the Origins of
Modern Humans. Princeton University Press, New Jersey.

. 1993. Technological and social dimensions of “Aurignacian-Age” body ornaments
across Europe. Pages 277-299 in H. Knecht, A. Pike-Tay & R. White, eds., Before Lascaux.
The Complex Record of the Early Upper Palaeolithic. CRC Press, Boca Raton.

Wiessner, P. 1989. Style and changing relations between the individual and society. Pages 56-
63 in |. Hodder, eds., The Meanings of Things. Material Culture and Symbolic Expression.
Harper Collins Academic, London, U.K.

Wobst, H. M. 1977. Stylistic behavior and information exchange. Pages 317-342 in C. E. Cle-
land, ed., For the Director: Reseach Essays in Honor of James B. Griffin. Museum of An-
thropology, University of Michigan, Ann Arbor.

161

Symbolic Expressions in Later
Prehistory of the Levant: Why
are They So Few?

Ofer Bar-Y osef

Department of Anthropology
Peabody Museum

Harvard University
Cambridge, MA 02138

This chapter brings together evidence for symbolic behavior from the Levant, a
well defined region within the Near East that encompasses the area from southeast
Turkey to the southern tip of the Sinai peninsula (Egypt) (Figure 1). Occasionally, I
refer to the broader region and include some information from Anatolia. The chrono-
logical sequence (Figure 2) described and discussed in the following pages lasts from
ca. 45,000 to ca. 8,000 BP (all radiocarbon dates cited in this chapter are in uncali-
brated years BP). While discussing some of the potential interpretations of these artis-
tic expressions, I focus on the question of why, in spite of the continuous occupation
of this area during the Upper Paleolithic and early Epipaleolithic periods, the number
of art objects is relatively so small. They became more abundant during the late
Epipaleolithic and especially the Early Neolithic.

Before delving into the records from the Near East, it is worth reiterating the
notion that many meaningful, informative human activities leave no archaeological
record. This is true for public ceremonies and private rituals. Nor are there oral tradi-
tions in the Near East that go beyond the mid-Holocene. Thus, nothing is left in the ar-
chaeological record from cultures in which activities such as dancing, for example,
were a major way of individual and/or communal expression. Public areas that may
have been used for communal activities, however, were identified in Neolithic sites
such as Cay6nii where an open space situated between the largest domestic dwellings
and next to public houses was called the ‘plaza’ (Ozdogan & Ozdogan 1989).

Decomposition is the main enemy of prehistoric art. Human expressions on per-
ishables such as hides, early twined cloths, and basketry are not preserved. Even fres-
coes painted on the walls of houses and sacred buildings in later periods such as the
Neolithic or the Chalcolithic (10,000- 5,500 BP), are rarely preserved. The rare exam-
ples from Neolithic Catal Hiiyiik (in Anatolia) or Chalcolithic Tuleilat el Ghassul (in

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

162 BAR-YOSEF

the Jordan Valley) indicate the amount of artistic expression lost forever in all other
excavated sites (Mellaart 1967, 1975).

There are only isolated reports and partial reviews of symbolic expressions of Pa-
leolithic and Epipaleolithic items in the Near East (e.g., Garrod and Bate 1937; Neu-
ville L951; Anati 1968, 1972; Cauvin 1972, 1994; Solecki 1980; Belfer-Cohen &
Bar-Y osef 1981; Bar-Yosef & Belfer-Cohen 1988; Goren-Inbar 1986, 1990; Hovers
1990; Bahn 1991; Belfer-Cohen 1988, 1991a, b; Belfer-Cohen & Hovers 1992;

CAFER
+“. GAYONU

* +) HALLAN CEMI

GRITILLE , ; ° :

NEVALLI CORI _
sy

Sia +GATAL HUYUK

~e
Hasakan

MUREYBIT
a
ABU HUREIRA
a “A
Dayr az Zawr “37 BOUQRAS

EL-KOWM

Al Hacithan &

Mediterranean Sea KSAR AKI! + YABRUD Iii
Beirut > *
JEBEL Saige a
BEISAMOUN—__/ *RAMAD
) ~~ AIN MALLAHA
OHALO Ii ~_HAYONIM «QUNEITRA
KFAR HAHORESH y NAHAL EIN GEV I!

NAHAL | SCQRFZEH

Damascus

—

EL-WAD — + WADPHAMMEH 27
KEBARA
a NETIV HAGDUD. a URKAN E-RUB IIA
SALIBIYA IX * + AIN GHAZAL
Jerusalem & Amman
AIN SAKHAI + ae
= ~JERICHO
LS ~ AD
hand oN Pon Said
J “> 4
e IN _ ra
4 GEBEL LAGAMA ~

> NAHAL HEMAR®
ROSH ZIN
« BEIDHA
+ BASTA

Eilatge Aqaba

« =Town/City

+ = Site

ee

0 Kilometers 100
NSN
Ss
ty
~

\
Hurghada

Red Sea 5

Ficure |. A map indicating the location of sites mentioned in the text

PREHISTORIC IMAGES OF THE LEVANT 163

Radiocarbon Levantine Levantine and Anatolian | West European

(uncalibrated years) | Archaeological Sites with Art Objects | Entities & Periods
BP. Entities & Periods

Pottery Neolithic Sha'ar‘HaGolan

PPNC Catal Huyuk Bouqras

Ramad Neolithic
Nahal Hemar
Cayonu Nevali Cort
Ain Ghazal
Beisamoun Munhatta
Netiv Hagdud Mureybet
Saliba Ix, _-Gilgat |
LATE Harifian Rosh Zin
Nahai Oren Hallan Gemi

Um e-Zuweitina
NATUFIAN Ain Sakhi —Okuzin’ Late Magdalenian

_ _ Hayonim

Pre-Pottery Neolithic B

$ Sultanian Mureybetian

om Mesolithic
Q = Khiamian

Mushabian _

Geometric
Kebaran Urkan e-Rub lla

Kebaran Complex
Early Magdalenian

EPIPALEOLITHIC NEOLITHIC

Late Anmarian Ohalo II

Ksar Akil I-III Solutrean

|
|

Hayonim D Gravettian

Levantine Aurignacian

= = Chatelperonnian

Early Ahmarian

UPPER PALE OEIMATC

Auriganican

Transitional Industry = Emiran

Mousterian

Quneitra

Moustenan

Figure 2. Chronological chart indicating the tentative correlation between the Levantine entities and
Western Europe

Weinstein-Evron & Belfer-Cohen 1993; Marshack, this volume; Yal¢inkaya et a/
1995). In the absence of an overall review paper on symbolic expressions in Le-
vantine prehistory, I describe in this chapter the flimsy evidence from the various ar-
chaeological periods and cultures (Figure 2). While I begin with a few comments on
the Middle Paleolithic mortuary practices, the use of red ocher and marine shells, I
mainly concentrate on Upper Paleolithic and Epipaleolithic (or Late Paleolithic in the

164 BAR-YOSEF

European terminology) remains, and briefly discuss finds from the Early Neolithic,
focusing on the Levant.

I include a discussion of the Neolithic period in order to stress the proliferation of
art objects and rock engravings during the Holocene in contrast to their paucity in the
Late Pleistocene, with the exception of the Natufian culture. I feel that the inclusion of
the Early Neolithic in this discussion helps us answer the question of why prehistoric
‘art’ objects became relatively abundant only during the very late prehistoric times in
the Near East. Furthermore, an overview that brings together Late Pleistocene and
Holocene symbolic expressions is more comparable to what has been recorded in
South A frica and Australia (see Parkington & Manhire, Lewis-Williams, and Rosen-
feld, this volume).

The ensuing discussion considers the implications of several interpretations of
prehistoric art to the Near Eastern records and ends with tentative interpretations of
Levantine symbolic expressions.

Mortuary Practices

Mortuary practices are known from the Middle Paleolithic of the Near East where
humans buried some of their dead in the very cave sites they inhabited (e.g., Shanidar,
Dederiyeh, Amud, Qafzeh, Skhul, Tabun and Kebara). Two nearly complete graves
(Skhul V and Qafzeh 11) provide direct evidence for the intentional inclusion of
grave offerings (a jaw of wild boar in Skhul and deer antlers in Qafzeh; Garrod and
Bate 1937; Vandermeersch 1969b).

During the Upper Paleolithic, the main trend seems to have been to bury the dead
outside the occupied zone. Because of this, the number of known burials is small.
They include the skeleton of ‘Egbert’ from Ksar *Akil (Bergman & Stringer 1989),
the individual flexed skeleton in Nahal Ein Gev I (ca. 25-20,000 BP; Arensburg
1977), and the shallow supine burial in Ohallo I (19,000 BP; Nadel 1994). The pres-
ence ofa few isolated human bones at various cave sites may indicate shallow burials
or the use of selected bones for other purposes (e.g., rituals). If graves were located
away from the main occupational horizons in caves, rock shelters, and open-air sites,
the chances of archaeologists recovering them are haphazard unless the excavation
strategies incorporate ‘outside’ areas. This hypothesis is supported by the isolated
burial from Ohallo II where the skeleton was uncovered outside the living zone. That
this practice may reflect a separation between the ‘interior’ and the ‘exterior,’ the
‘mundane’ and the ‘sacred,’ or the ‘pure’ and ‘impure’ is a plausible proposition.

Late Paleolithic sites (or Early Epipaleolithic, ca. 17,000-13,000 BP, with the ex-
ception of the Natufian culture) have yielded isolated, semiflexed burials at Ein Gev I
(Kebaran; Arensburg & Bar-Yosef 1973), Neve David (Geometric Kebaran; Kauf-
man & Ronen 1987), where the skeleton was covered with a few mortars and stone
bowls. and Kharaneh IV (Kebaran) in Jordan (Muheisen 1985). In spite of the small
number of excavated sites from this period, on-site burials seem to have become again
a common practice, with the more elaborate ones found at the subsequent Natufian
settlements (ca. 13,000-10,300 BP).

At Natufian sites, graves occur in open areas, beyond dwelling structures and
abandoned houses. This practice is generally interpreted as symbolizing ownership of
the locale. Burials vary in position (flexed, semiflexed, and extended), number of in-
dividuals per grave (ranging from one to five or more), grave structure, body decora-
tions (found only in Early Natufian graves) and accompanying objects. About

PREHISTORIC IMAGES OF THE LEVANT 165

one-third of the burials are of children. Most of the Natufian cemeteries are generally
well recorded (e.g., Garrod & Bate 1937; Perrot & Ladiray 1988; Valla 1991; Bar-
Yosef & Goren 1973; Noy 1991; D. E. Bar-Yosef 1991; Belfer-Cohen 1988, in press:
Belfer-Cohen & Hovers 1992).

The new custom of selective skull removal was first observed in the Late Na-
tufian burials from Hayonim cave and from the terrace at Nahal Oren (e.g., Belfer-
Cohen 1988; Stekelis & Yizraeli 1963). Special graves, in which dogs and humans
were buried together, were found at Ain Mallaha (Davis & Valla 1978) and on the
Hayonim terrace (Valla 1990). These animal/human burials testify to a changing
view of human relationships with nature during the Natufian (Valla 1990).

Efforts to decipher social hierarchy from the Natufian burials have not been suc-
cessful (Wright 1978; Belfer-Cohen in press; Byrd & Monahan 1995). However, the
presence or absence of body decorations (Figure 3) and their particular combinations
within the graves have enabled us to differentiate between various base camps such as
El-Wad cave and terrace, Hayonim cave and Terrace, Ain Mallaha, and Nahal Oren
(Belfer-Cohen 1991a; Belfer-Cohen & Bar-Yosef n.d.; Belfer-Cohen et al. 1991).

During the Early Neolithic, the same burial practices continued. In the Pre-Pottery
Neolithic A and B periods (abbreviated as PPNA and PPNB; ca. 10,300 to 7,700 or
7,500 BP), burials were flexed and semiflexed and placed in open areas and aban-
doned houses. Adults were clearly treated differently than children. Adult skulls were
removed (without the jaw) while juvenile burials were not touched.

During the PPNB, selected adult skulls were covered with plaster and painted par-
ticularly on their frontal part (Ramad, Beisamoun, Jericho, Kefar Hahoresh, Ain
Ghazal). In one case from the cave of Nahal Hemar, the top and back of the skulls
were covered with a mixture of collagen and asphalt in a net pattern (Bar-Yosef &
Alon 1988). These practices of skull modification are often interpreted as ‘the cult of
the ancestors’ (e.g., Cauvin 1972, 1994; Bar-Yosef & Belfer-Cohen 1992 and refer-
ences therein). Since the number of skull-less skeletons is larger than the number of
known treated skulls, either the missing skulls were placed elsewhere in the site or
only some of the skulls (perhaps those belonging to an elite group) were treated. Writ-
ten Ugaritic mythical stories support the interpretation of special treatment (Margalit
1983).

Regional differences within the Near East evolved during the PPNB and led to the
formation of at least three major interaction spheres (the Levantine, the Zagrosian and
the Anatolian). This development of a more specific social geography may have been
caused by the advent of agriculture and animal husbandry (Bar-Yosef & Meadow
1995). Of the mortuary practices from this period, one of the most notable is the
‘house of the skeletons’ from the village site of Cay6nii in southeast Turkey (Cambel
1980; Ozdogan & Ozdogan 1989). In marginal areas of the Levantine interaction
sphere, among PPNB desertic groups in southern Sinai who descended directly from
earlier hunter-gatherers, secondary burials were practiced and skulls were kept with
several postcranial elements in underground storage facilities (Bar-Yosef 1984;
Hershkovitz & Gopher 1990; Hershkovitz et al. 1994).

Certain changes took place during the final phase of the PPNB (also knownas Late
PPNB; 8,500-8,000 BP) and during what is currently termed the PPNC (8,000-
7,700/500 BP; Rollefson & Kéhler-Rollefson 1993). The practice of removing skulls
from adult skeletons ceased and secondary burials became more common than in the
previous period.

Major changes occurred during the Late Neolithic (also known as the Pottery
Neolithic; ca. 7,500-6,500 BP) and the Chalcolithic (ca. 6,500-5,500/300 BP) peri-

166 BAR-YOSEF

ods. Graveyards were removed from the ‘interior’ (dwelling areas) to the ‘exterior’
(the outskirts of the sites).

Red Ocher, Marine Shells, Beads and Pendants

Both red ocher and marine shells, on clothing, headgear, belts, etc., are consid-
ered elements of body decoration (e.g., Wreschner 1976; D. E. Bar-Yosef 1989). The
earliest presence of ocher is noted in Mousterian sites. A fine example of a large,
scraped lump of red ocher was found at Qafzeh (Vandermeersch 1969a). Several
Mousterian artifacts from Hayonim cave retained red ocher on the retouched edge.

Red ocher was also used during the Upper Paleolithic and the Epipaleolithic. A
limestone slab and a hand stone both smeared with red ocher were uncovered in the
Upper Paleolithic layers in Qafzeh cave (Ronen & Vandermeersch 1972) in layers
that contain early Ahmarian industry (Figure 2). Several limestone slabs bearing
markings of red ocher were found in Hayonim layer D (Levantine Aurignacian) and
flint artifacts stained with red ocher were uncovered in the Lagaman sites in northern
Sinai (Figure 2; for references see Belfer-Cohen & Bar-Yosef 1981; Bar-Yosef &
Belfer 1977; Gilead 1991). Red ocher is also recorded from almost every site dating
between 30,000 BP and 8,000 BP.

A few marine shells of nonedible species, mostly Glycvmeris, occur for the first
time in Skhul and Qafzeh caves in Middle Paleolithic contexts (Bar-Yosef 1989;
Belfer-Cohen &Hovers 1992). Upper Paleolithic assemblages contain several spe-
cies of marine shells collected along the shores of the Mediterranean Sea. Often the
apex or the body of the shell was abraded in order to facilitate their use as beads, pen-
dants or buttons (D. E. Bar-Yosef 1989),

Atearly Natufian sites, we can note a significant shift in the suite of collected shell
species and their frequencies. While most were collected along the Mediterranean
coast. a few originated from the Red Sea. The most common species is Dentalium sp.
which was used in great numbers for decorating headgear, belts, bracelets, and per-
haps also sewn onto clothing (Figure 3; D. E. Bar-Yosef 1991; Reese 1991). In sites
from the Harifian culture (for chronological placement see Figure 2), which is consid-
ered to be a Late Natufian adaptation to the semiarid zone of the Negev and northern
Sinai (Bar-Yosef 1987: Goring-Morris 1991), a different geographic orientation in
shell collection has been noted (D. E. Bar-Yosef 1989). These sites contain a signifi-
cantly greater number of species from the Red Sea and the Indian Ocean than do Na-
tufian sites within the Mediterranean coastal ranges, although the frequencies of
Mediterranean shells are still the highest.

Natufian jewelry also includes other types of beads and pendants that were made
of limestone, basalt, greenstone, malachite, bone and teeth, mostly wolf canines
(Figure 3). Greenstone beads were found in many sites but it is not clear whether they
were brought as finished products or as raw materials from their unknown sources in
Syria, Jordan, or Sinai. Their presence does indicate local exchange over distances of
100-200 km. Even greater distances are attested to by the first use of obsidian from
Anatolia and of the fresh water shell Aspatharia from the Nile at Ain Mallaha (Valla
1987).

During the Neolithic, the variability of shell species used for making beads de-
creased considerably. While this holds true for the collections retrieved from farming
communities, the situation differs considerably for sites occupied by hunter-gatherers

PREHISTORIC IMAGES OF THE LEVANT 167

FiGure 3. Aurignacian and Natufian beads and pendants: (1, 2) Levantine Aurignacian beads from Hay -
onim Cave, (3,4) Dentalium shell beads from erg el-Ahmar, (5) perforated wolf canine, (6, 7, 10, 12) bone
pendant from Hayonim Cave, (8, 9) bone beads from erg el-Ahmar, (11) double twin pendants from El
Wad. After Belfer-Cohen & Bar-Yosef (1981), original drawings of the author

in the semiarid and arid zones such as southern Sinai where many species are repre-
sented among the large collections of shells (D. E. Bar-Yosef in press).

Beads made of hard minerals and a smaller number shaped from animal bone con-
tinue to be present in most Neolithic contexts. Anatolian obsidian found in many Le-
vantine Neolithic sites indicates exchange over a distance of 500-700 km.

Incised, Engraved Objects and Figurines

Upper Paleolithic and Early Epipaleolithic

The number of incised or engraved items found in Upper Paleolithic contexts, in-
cluding the early Epipaleolithic, is small. The following descriptions list most of
these objects.

Materials dating to the Upper Paleolithic include the Levantine Aurignacian as-
semblage of Hayonim layer D, radiocarbon dated to 29,000-27,000 BP, where two
engraved limestone slabs were found, one of which has been published (Belfer-
Cohen & Bar-Y osef 1981; Bar-Yosef & Belfer-Cohen 1988; Figure 4). The cluster of

168 BAR-YOSEF

finely incised lines on this small slab, interpreted as depicting a ‘horse’ is supported
by Marshack’s (this volume) detailed analysis. In addition, a few small limestone
slabs, not yet published, bear stains of red and black paint.

A bone point with a series of vertical incisions came from the Late Ahmarian as-
semblages in Ksar *Akil (Tixier 1974; Marshack, this volume). A somewhat similar
bone object with a series of vertically incised lines was recovered at Ohallo II, and
dated to 19,000 BP (Nadel 1994).

A more elaborate pattern created by finely incised lines on both faces of a flat
limestone pebble came from a Kebaran site of Urkan er-Rubb Ila (Hovers 1990), and
was dated to 15,000/14,500 BP. One face contains fine, semicircular, parallel lines
with short vertical ones and two sets of parallel lines with a ‘ladder’ pattern (see Hov-
ers 1990 and Marshack, this volume). Five series of lines connect the previous ones to
the edges of the pebble. The other side bears two clear ‘ladder’ patterns and a fill of
cross- hatched pattern. Incised pebbles were also discovered in the Okiizini cave in
southwest Turkey (Marshack, this volume). The date of these objects is unknown but
recent excavations there (Yal¢inkaya et al. 1995) indicate that they may have been re-
trieved by KGkten from layers dated to 13,200— 11,900 BP and are therefore contem-
porary with the Early Natufian culture.

It is worth mentioning that Kékten named Okiizini cave after a depiction of an
aurochs which he apparently found on the cave wall (Yalcinkaya et al. 1995). While
this figure is not visible today, a clear engraving of an aurochs on a pebble has been
identified. A simplified human figure at the lower part of the aurochs is interpreted as
thrusting a spear into the animal’s chest (Marshack, this volume).

A second pebble bears three ‘blocks’ of parallel incised lines in the ‘ladder’ pat-
tern on one surface. Two groupings of ‘ladders’ are ‘horizontal’ and the group in be-
tween Is ‘vertical’ in relation to the other two. The other surface of this pebble has an
open circle with smaller circles in it, two branching lines forming a ‘corridor’ witha
few additional incomplete circles and an additional branching ‘ladder.’

In the Levantine context, it is interesting to note that this large open circle and
‘corridor’ resemble depictions of ‘desert kites’ used as drives for gazelle hunting
(Meshel 1974: Betts & Helms 1986). This type of drive with two ‘arms’ built either as
two rows of stones or as wooden pegs is known also from various geographic regions
such as Scandinavia (e.g., Barth 1982) and the North American high plains (e.g., Fri-
son 1968) where they were used to hunt other herd animals. In southwest Turkey, wild
goats and/or wild sheep, the bones of which are common in the faunal assemblages at
Okiizini (Yalginkaya et al. 1995), may have been hunted by a similar technique.
While aurochs bones are rare in this site, they do occur at other Late Paleolithic sites
such as Hallan Cemi (Rosenberg & Davis 1992). This taxon 1s an especially common
figure at Catal Hiiyiik, and could have played an important role in prehistoric Anato-
lian rituals (Cauvin 1994),

The Natufian Culture (13,000/12,800-10,500/300 BP)

Natufian ‘art objects’ include both figurines and incised and engraved objects
(Figure 5). The relative richness of this Levantine inventory, in part contemporary
with the European Magdalenian (see Figure 2), was overlooked in a recently pub-
lished overview of non-European symbolic expressions (Bahn 1991). Both the
mobiliary objects and the incised limestone slabs found at these sites emphasize the
uniqueness of the Natufian culture among contemporary Near Eastern Epipaleolithic
entities.

PREHISTORIC IMAGES OF THE LEVANT 169

e Aurignacian). Photograph courtesy of Israel Museum

r D (Levan

>

‘horse’

The

FIGURE 4

170 BAR-YOSEF

Figure 5S. Natufian objects: (1,4) decorated sickle hafts, (2) gazelle head, (3) human head, (5) hu man torso,
(6) couple engaged in intercourse, (7) schematic human head, (8) schematic human (?), (9) d ouble figurine,
ungulate and human head. Proveniences: (1) Kebara cave, (2, 9) Nahal Oren, (3, 4, 8) El Wad, (5,7) Ain
Mallaha. (6) Ain Sakhri. After Perrot (1966), Noy (1991), and Weinstein-Evron & Belfer-Cohen (1993)

Recovered incised objects include, among others, several large, heavy limestone
slabs with incised geometric forms and one large ‘fish’ (?) unearthed in Hayonim
(Belfer-Cohen 1991a). The common pattern on these slabs consists of numerous ver-
tical strokes enclosed within two long parallel lines forming the ‘ladder pattern’. One
of these large slabs is described by Marshack (this volume). It is worth mentioning
that the ‘ladder’ pattern was also incised on several sides of a large, squarish, elon-
gated block (Figure 8), referred to as the ‘stele,’ which apparently stood in one of the
rounded ‘rooms’ at Hayonim cave (Belfer-Cohen 199 1a).

PREHISTORIC IMAGES OF THE LEVANT 171

Other large carved limestone slabs exhibit a ‘meander’ pattern. These were un-
covered at Wadi Hammeh 27 (Figure 6:2). The same pattern is found on broken large
bowls at Ain Mallaha and Shukbah (Edwards 1991; Noy 1991, and see Figure 6).

No human or animal figurines dating prior to 13,000 BP have been found in the
Near East to date. With the emergence of the Natufian culture, however, came an in-
creasing number of figurines, a trend that continued into the Early and Late Neolithic.
Two of the Natufian figurines, both found by Neuville in the collections of Bethlehem
monasteries in the late 1920’s, attracted much attention. One, a small item which rep-
resents a couple engaged in sexual intercourse (Figure 5:6), has recently been re-
examined in detail (Boyd & Cook 1993). The general form of this figurine resembles
a schematic one uncovered at Nahal Oren from a PPNA context (see Stekelis & Noy
1963). The second figurine is the small limestone ‘kneeling gazelle’ in which the head
is missing. Amiran (pers. comm.) considers this item to be of mid-Holocene Egyptian
origin.

Other Natufian animal figurines were found at Kebara cave, El-Wad cave and ter-
race, Ain Mallaha and Nahal Oren. Some, carved on bone, were previously inter-
preted as depicting young gazelles (e.g., Garrod & Bate 1937; Stekelis & Yizraeli
1963; Perrot 1968; Cauvin 1972, 1994; Bar-Y osef 1983; see Figures 5:1, 2,4). This is
probably an overstatement, since all that can be said with certainty about them is that
they are generalized images of young ungulates. They appear as decorations of sickle
hafts or as individual mobile items. Two animal figurines carved and incised on lime-
stone, uncovered at Nahal Oren (Noy 1991), represent a double image of a dog and
owl (Figure 6:1) anda ‘baboon’ (Figure 6:3). Another double image, carved on a horn
core from the same site, depicts a human face and an ungulate head (Figure 5:9).

Rare human representations from El-Wad and Ain Mallaha (Garrod & Bate 1937;
Weinstein-Evron & Belfer-Cohen 1993; Marshack, this volume), are often rather
schematic in shape, with no specific indications of gender (Figure 5:3). They herald
what later became a major subject in Neolithic iconography.

Other ‘decorated objects’ include several domestic tools. For example, shaft
straighteners made of basalt from Nahal Oren have a special shape (Figure 6:4) and
are decorated with a carved meander pattern. Other objects include pestles, which
were often made of basalt and brought to the coastal sites from distances of 60-100 km
(Weinstein-Evronet al. 1995). Some of these bear engraved designs while others have
a phallic shape. The latter can also be found in Harifian contexts (Goring-Morris
1991) and in the Southern Sinai site of Abu Madi I which dates to the PPNA period.

The Early Neolithic (ca. 10,300-7,800/500 BP)

Art objects became numerous during the Early Neolithic. Sites from the earliest
phase, the PPNA, however, have not revealed as many items as those from the follow-
ing phase, the PPNB. PPNA figurines are often small, made of clay or limestone and
are generally broken. They have been uncovered at Mureybet HI (Cauvin 1977;
1994), Jericho (Kenyon & Holland 1983), Gilgal I (Noy 1986), Salibiya LX (Bar-
Yosef 1980) and Netiv Hagdud (Bar-Y osef et a/. 1991; Bar-Y osef & Gopher in press:
Noy 1986). Most figurines represent females, as indicated by the presence of breasts.

The female figure seems to occupy a particular place in PPNA imagery. A few of
the eight female figurines, made of limestone and baked clay, are shown in Figure 7.
They originate from Mureybet III (dated to the early PPNA, contemporary with the
Khiamian). The other figurines from Salibiya IX, (Figure 7:1) Gilgal land Netiv Hag-

72 BAR-YOSEF

FiGuRe 6. Natutfian art objects: (1) limestone double figurine, dog and owl, (2) large slab with engraved me -
ander and square patterns, (3) limestone ‘baboon’ figurine, (4) basalt shaft straightener, (5,6) decorated ba -
salt mortars. Proveniences: (1, 3,4) Nahal Oren, (2) Wadi Hammeh 27, (5) Shukbah, (6) Ain Ma Ilaha. After
Noy (1991)

dud (Figure 7:4) depict either a ‘seated woman,’ a motif that was repeated in later as-
semblages, or standing females.

The seated female figure (Figure 7:4, Figure 8:2, 3,5) represents an image com-
mon to the entire Near East particularly during the Early Neolithic. In the PPNB and
the Pottery Neolithic periods, female figurines occur at Anatolian sites such as
Cayénti, Cafer Héyiik, and Catal Hiiyiik (e.g, Mellaart 1975; Cauvin 1972, 1985,
1994). Often, these are interpreted as being in the posture of giving birth (Noy 1986).
Only the later seated female from Catal Htiytik, who 1s shown resting her hands on the
two standing panthers at her sides, heralds the different concept of an enthroned ‘go-
ddess’ or ‘mother-goddess’ (Mellaart 1967; Cauvin 1994: Noy 1986).

Human statues made of reeds and plaster depicting either complete forms or busts
alone seem to have been a Levantine innovation (Figure 8). Despite having been
painted, they show no clear indication of gender, and are generally interpreted as rep-
resenting deities. The best preserved of these were uncovered at Ain Ghazal (Rollef-
son 1983, 1986; Rollefson ef a/.1992), and in fragmentary states at Jericho, Tel
Ramad and Nahal Hemar (Garstang & Garstang 1940; Contenson 1967). Their depo-
sition in pits 1s considered to be the interment of used paraphernalia, a custom which
is well recorded in the Near East from Early Bronze age contexts (Garfinkel 1994).

If Amiran’s (1962) and Margalit’s (1983) proposal is valid, then the human stat-
ues and plastered skulls are the direct expressions of a Neolithic cosmology that was
written down three or four millennia later in the Mesopotamian and Ugaritic texts on
human creation and ancestor veneration. Oral traditions lasted in the pre-literate
world fora very long time and therefore the proposed relations between Holocene ar-
chaeological remains and documented mythologies in the same region are plausible.

PREHISTORIC IMAGES OF THE LEVANT 173

The small female figurines that are present in PPNB contexts, even if not numer-
ous, continue from the PPNA and into the Yarmukian culture (Stekelis 1972), which
is a Pottery Neolithic entity (see Figure 2). A selection from the Yarmukian is pre-
sented in Figure 9, demonstrating the depiction of figures wearing a special garment
with a high headgear. Incised pebbles where only one line is present were interpreted
as representations of female genitalia.

Animal representations continue from the Natufian onward with depictions of
birds of prey found at Mureybet III, Gilgal Tand PPNA Jericho. Cauvin (1994) inter-
prets the combination of humans and animal figurines as the dichotomy in the world
of deities. Later, during the PPNB, the aurochs takes the place as a major element, a
counterpart to the female figure, giving rise to what Cauvin calls ‘the new religion,’
namely that of the ‘Woman and the Ox.’ The presence of complete aurochs’ horns at
several sites such as Hallan Cemi (of Late Natufian age; Rosenberg & Davis 1992),
and at Mureybet III, together with more direct evidence of bucranes and frescoes from
other sites, most notably Catal Hiiyiik, indicates the prominence of this ritual (Cauvin
1994) in the northern Levant and Anatolia which was probably where it emerged.

Finally, it is worth mentioning that incised pieces with specific designs are found
in small numbers through the entire sequence of the Early Neolithic. Among these,
the incised flat pebble from Netiv Hagdud is most notable (see Bar-Y osefet al. 1991).
PPNB shaft straighteners decorated on their back side with zigzag incision or deeper
engravings found in Mureybet II, Cafer Hiiyiik and Cheikh Hassan (Cauvin 1994)
may indicate familial ties between these northern Levantine and southeastern Anato-
lian sites. On the whole, most of the artistic expressions on mundane tools, as well as
the appearance of stamps after 9,000 BP (e.g., Cauvin 1994), can be interpreted as a
further development in the materialistic expression of individual ownership.

FiGuRE 7. PPNA human figurines: (1) Salibiya IX, (2, 3, 5, 6) Mureybet, (4) Netiv Hagdud. After Bar-
Yosef (1980), Noy (1986), and Cauvin (1994)

174 BAR-YOSEF

Bore ai “A
Ti:

oY ON NG

)
4
ma. 10
1°) sem
—=
Figure 8. Early Neolithic (PPNB) female and male figurines and plaster statues: (1, 6) males, (2,3,5.7,9)

females, (4) human, (8, 10) plaster human statues. Proveniences: (1,5) Cater, (2, 3) Cayonu, (4) Ramad, (6,

7) Munhatta, (8, 9) Ain Ghazal. (10) Jericho. After Cauvin (1994)

In sum. most of the art objects in the Levant, except for the few large Natufian
slabs, generally seem to be objects of sacred value for personal, domestic use. There-
fore they may be similar to Central or Eastern Europe mobiliary art. The Neolithic and
Chaleolithic frescoes in the Near East seem to resemble the European and Australian

PREHISTORIC IMAGES OF THE LEVANT Les

Ficure 9. Selected Yarmukian female figurines: (1) Munhata, (2, 4) Sha’ar HaGolan, (3) Byblos. After
Noy (1991)

cave paintings and large rock art panels which were probably executed by shamans or
holy men.

Discussion

In discussing the potential answers to the issue of what this art means, we need to
respond to the following questions: (1) Why are Near Eastern Upper Paleolithic and
Epipaleolithic symbolic expressions so few in number; (2) Why are these predomi-
nantly mobiliary and notational instead of full fledged rock art; and (3) What are the
possible interpretations of the various classes of Epipaleolithic and Early Neolithic
symbolic expressions?

Art, Environment, and Demography
In responding to the first two questions we can ascertain that not all contemporary

populations of Homo sapiens sapiens across the Old World left behind a large corpus
of artistic or symbolic expressions. That the artistic expressions are not the result of

176 BAR-YOSEF

the availability of proper sites (rock shelters and deep caves) or raw materials (e.g.,
bone and antler) can be demonstrated by even a cursory comparison between two
similar regions. For example, during the Late Pleistocene, western Georgia, which
lies next to the Black Sea, witnessed climatic conditions that were generally similar to
those of the Franco-Cantabrian region (Velichko 1984). Nor did the rock shelters and
limestone caves in the deep gorges of the Caucasus differ much in morphological
configuration and types of sediment from those of the Dordogne. Nevertheless, no
Upper Paleolithic parietal or mobiliary art has ever been found in western Georgia de-
spite numerous excavations (e¢.g., Liubin 1989).

The archaeological evidence presented above from the small area of the Mediter-
ranean Levant tells a similar story even though the number of excavated Upper Pa-
laeolithic sites is limited (e.g., Bar-Y osef & Belfer-Cohen 1988; Gilead 1991). This
difference between the Franco-Cantabrian region and the circum-Mediterranean area
is not accidental — a fact amply demonstrated by the new discoveries at Grotte
Cosquer and Grotte Chauvet in France (Clottes, this volume) — but rather a reflection
of a unique cultural history. The uniqueness of this region forces us to reverse the
question: instead of identifying Franco-Cantabrian art with the emergence of modern
human behavior and language (e.g., Mellars 1989; Eccles 1989; Davidson & Noble
1989, White 1989), we should ask why and under what kind of social conditions was
art produced in this particular area. We could also delve into the meaning of this art,
but that aspect is dealt with by other contributors to this volume (see Davidson,
Clottes, Gonzales-Morales, this volume).

Studies of pan-European Pleistocene art may enhance our insight into the unique-
ness of western Europe. Furthermore, understanding the particular circumstances as-
sociated with the appearance of mobiliary art in central and eastern Europe, including
Siberia, may shed light on specific human behavior in these regions and, by inference,
enable us to interpret the paucity of art objects in other regions of the Old World (Sof-
fer, this volume). Why this may be so is considered below.

Most scholars agree that imagery is related to ritual, an essential part of social ac-
tivities shared by all hunting and gathering societies (e.g., Leroi-Gourhan 1965;
Lewis-Williams & Dowson 1988; Eliade 1989; Marshack 1990; with references
therein). If Near Eastern peoples did not produce rock art and produced only very few
mobiliary objects, they may have differed from their Eurasian and Australian con-
temporaries in their social structure (see also Davidson, this volume). They were not,
however, alone. Other Late Pleistocene hunter-gatherers, for example most of the
groups in North and South A frica, did not produce imagery items in great quantity el-
ther. This stands in sharp contrast to the Holocene florescence of symbolic expres-
sions across the entire world. Why was this so?

The search for a common denominator behind the creation of rock art led Lewis-
Williams (e.g, Lewis-Williams 1984, 1991; Lewis-Williams & Dowson 1988,
1989). to look fora neuropsychological model that would explain why shamans paint.
The model is based on the way the human brain reacts to hallucinatory states. The
subjects of the depicted signs and figures are interpreted to be entoptic (within the op-
tic system) visions. Lewis-Williams proposed using the same model to explain Upper
Paleolithic European art (Lewis-Williams 1984, 1991, this volume). What remains
unanswered in the course of this systematic effort is why shamans in groups living in
other regions did not express their altered state of consciousness in a similar way.

If we assume that all prehistoric bands of hunter-gatherers, regardless of the fluid-
ity of their social structure, had some sort of a holy person, then according to the
Lewis-Williams model, we should expect to find similar artistic expressions, espe-

PREHISTORIC IMAGES OF THE LEVANT 177

cially in regions where rock shelters and the appropriate raw materials were available.
However, we have seen that this is not the case. Therefore, it seems that there may be
another social and ecological dimension to this issue (Mithen 1991). Differences in
patterns of symbolic behavior could have been related to the particular social struc-
ture of the groups, i.e., the intensity and frequency of social interaction on all levels
(intra- and intergroup, within macrobands or ‘tribes’).

Concerning the Near East, the following considerations could constitute a basis
for such an explanation. It is common to argue that the distribution, reliability, acces-
sibility, and predictability of resources within a given area dictated the foraging
strategies of hunter-gatherer populations and resulted in the formation of a given so-
cial structure. Most Near Eastern environments, especially those close to the Mediter-
ranean Sea, provided predictable resources. These include animal protein (over 30
species of mammals, reptiles and birds) and especially abundant plant food with over
150 species of exploitable seeds, fruits, leaves, tubers and roots. All resources could
be exploited over short distances (1-20 km). Even when seasonal differences between
the cold and rainy winters and the generally warm, often dry summers are taken into
account, this region was more lush than many other parts of the Old World and well
suited for continuous habitation of human groups. We have argued elsewhere (Bar-
Yosef & Belfer-Cohen 1992 with references) that a minimal territory could be cau-
tiously estimated as 250-500 square kilometers per band. Within the steppic belt,
which characterizes either the semiarid plains or the high plateaus, larger territories
(500-2500 square kilometers) would have been necessary to support a band. Biologi-
cally viable populations (ca. 400-500 individuals) in the Levant would have required
a territory of 4,000-8,000 square kilometers or less. The coastal ranges of the eastern
Mediterranean thus provided sufficient resources to accommodate numerous groups
of neighboring hunter-gatherer bands with circumscribed territories, small geo-
graphic ranges of mating systems, where communication and information exchange
was easy and frequent.

Intensity of social interaction, whether within a family, extended family or larger
social group (band and macroband) level differs among different cultures. Among
modern humans, this intensity is related, at least in part, to the physical proximity of
family members, extended families and bands. Absolute proximity is socially con-
structed in terms ofa relative scale of human density. So, for example, what is consid-
ered in Belmont, Massachusetts to be a ‘thickly settled’ area (as noted on a road sign),
in any Levantine country would be considered a spacious suburb. Social, economic
and psychological requirements, based on the history of a local population are re-
sponsible for creating perceived population pressures. Ina situation of a dense physt-
cal neighborhood, a heightened emphasis on discussion, teaching, and the passing
down of oral traditions would serve as the mechanism for resolving intra- and inter-
group conflicts.

Social agendas are very different, however, among groups that are geographi-
cally widespread and scheduled meetings are required. Under conditions where cohe-
sion needs to be guarded across large distances, ritual is important in order to increase
social and biological fitness. On a continental scale, we may see competition among
groups (macrobands?) for resources resulting in group selection (Wilson & Sober
1994). In this situation, we would expect to find evidence of nonperishable ‘artistic’
activities, whether executed by shamans, medicine man or other members of the
group (e.g., Lewis-Williams 1984, 1991, this volume).

It can be speculated that certain kinds of social stresses caused in western Europe
by climatic fluctuations during the Upper Paleolithic (Gamble 1991) were similar to

178 BAR-YOSEF

those felt by the San during the time when the paintings of the South African rock
shelters were done. The Levantine region, even under the impact of climatic changes
during most of the Paleolithic, did not witness major shifts in the resource base. It was
only in the closing millennia of the Pleistocene that population growth resulted in ter-
ritorial packing (Binford 1983) and sedentism (Bar-Yosef & Belfer-Cohen 1991,
1992: Henry 1989) and therefore in the production of the Natufian art (Belfer-Cohen
1988).

The apparent contradiction between this model (high population densities lead-
ing to art production) and that of the earlier period (geographically widespread popu-
lations leading to art production) might be explained by the differences in the kind of
art produced in these two situations. Group oriented rock art might have been created
by shamans in order to create social cohesion while the smaller scale domestic, per-
sonal and notational art objects might have been used to resolve small scale intra- and
intergroup conflicts.

Near Eastern Art and Meaning

I now turn to the third question posed above, namely, the interpretation of the Le-
vantine artistic expressions. I do so ina chronological order from the Upper Paleo-
lithic through Early Neolithic (for the time scale see Figure 2).

Levantine Upper Paleolithic sites, from 45,000 through 20,000 BP, are not numer-
ous but suggest that groups or families were more mobile than they had been in the
preceding Middle Paleolithic. This can be seen at Kebara cave where there are more
scavenged and gnawed bones in the Upper Paleolithic faunal assemblages than in the
Middle Paleolithic layers (Speth in Bar-Yosef et a/. 1992) indicating longer periods
of abandonment. The same conclusion was reached for Upper Paleolithic sites in the
Negev and Sinai (e.g., Marks & Friedel 1977; Marks 1993; Marks & Ferring 1988;
Phillips 1988). This makes the site of Ksar ‘Akil, where 18 meters of Upper Paleo-
lithic layers were exposed, a unique occurrence (e.g., Ohnuma 1988; Bergman 1987;
Mellars & Tixier 1989). This site, however, is found in a prime ecological area and
could have served as a major settlement, perhaps even as an aggregation site. In spite
of this, it has produced only one incised bone object.

Evidence for symbolic expressions have been recovered from Levantine Aurig-
nacian deposits (ca. 36,000-27,000 BP) with relatively rich bone and antler industries
(including split-base points). These assemblages, as discussed above, also yielded
personal decorations such as beads made from deer, bear and horse teeth, as well as
painted and ocher smeared slabs. The incised ‘horse’ from Hayonim may reflect the
same tradition in animal engraving as found in Europe (see Marshack, this volume).
Whether the producers of these assemblages originated in Anatolia or even farther
away in southeast Europe remains a matter for speculation. In spite of the claims for
the presence of Aurignacian assemblages in the Zagros (Olszewski & Dibble 1994)
there is no valid evidence for such an assertion either there or in the Caucasus region
(e.g, Liubin 1989).

The rare ‘art objects’ from the Late Ahamrian (Ksar ‘Akil, Ohallo II; ca. 27,000-
18,000 BP) and the Kebaran (Urkan er-Rubb Ia) complexes indicate, as suggested by
Hovers (1990) following Conkey (1984), the use of special objects within a small so-
cial circle of an extended family or a macroband. The simple motif composed of se-
ries of strokes is interpreted as some sort of notation, but we have no clues as to what
the nature of this was. These prehistoric groups made minimal use of a few Mediterra-
nean species of marine shells for body decoration (D. E. Bar-Yosef 1989). Whether

PREHISTORIC IMAGES OF THE LEVANT 179

these shells were sewn on clothing or decorated handles of hafted tools and thus were
stylistic markers of group membership, or served solely as elements in exchange, is
unknown.

It is worth noting that there are only rare ‘artistic expressions’ which could indi-
cate the increase of social pressures, among Levantine groups during the Late Glacial
Maximum. This is probably due to a lesser need among the Kebarans for major social
aggregations alleviating scalar stress (Johnson 1982; Hovers 1990). Gatherings of
shorter duration and perhaps on a smaller scale are proposed as the Levantine recipe
for achieving social cohesion, thus decreasing the need for symbolic expressions
(Hovers 1990). It should be noted, however, that we have only a small number of ex-
cavated Kebaran sites on hand and that future fieldwork may enlarge the small collec-
tion of art objects recovered to date.

The major cultural change occurred in the Levant with the establishment of the
Natufian culture (Henry 1989; Bar-Yosef & Belfer-Cohen 1991, 1992), known for its
semisedentary or sedentary sites (e.g., Tchernov 1991; Lieberman 1993). The sudden
proliferation of ‘art objects’ at these sites is attributed to social stress (Belfer-Cohen
1988, 199 1b, in press; Belfer-Cohen er a/. nd)— one supposedly caused by a high de-
gree of sedentism accompanied by anticipated seasonal mobility. Under such circum-
stances the family heads and/or ‘shamans’ had to find more than one way of
alleviating social stress and resolving conflicts whether on intra-and/or intergroup
level. Means used for identifying individual groups could have included body decora-
tion, ornaments sewn on clothing, headgear, or the decorations of utilitarian artifacts
such as quivers. The scenario adopted in this paper (for details see Henry 1989; Bar-
Yosef & Belfer-Cohen 1991, 1992) is that the Early Natufian was formed through the
coalescence of Geometric Kebaran groups into larger bands. In the process of cluster-
ing, it is possible that additional enhancement of former identity of extended families
led to increasing individual differentiation, expressed in those body decorations
found in graves. At the same time, the membership in newly formed groups gave rise
to what can be identified as emblematic style (as defined by Wiessner 1984, 1989) re-
flected in body decorations and not in the observable changes in mundane artifact ty-
pology.

Under such circumstances, especially since the use of sickles for harvesting in-
stead of beaters and baskets indicates the maximizing of yields from small plots (Hill-
man & Davies 1990), the marking of territories became a necessity. This could have
led to some simple record keeping, perhaps reflected in the notational marks on lime-
stone slabs recovered from Hayonim (Belfer-Cohen 199 1a; Marshack, this volume).

The need for territorial markers continued in the Levant for the ensuing millen-
nia. Territorial markers are known from these later periods, when they occur at iso-
lated, well delineated sites which lack habitations in their immediate area. Examples
of this can be seen at the PPNB Nahal Hemar cave in the Judean desert, which con-
tained a major cache of discarded paraphernalia (Bar-Y osef & Alon 1988; Bar-Y osef
& Schick 1989; Garfinkel 1994), and at the Chalcolithic temple in En Gedi, which
presumably was the source for the collection of copper artifacts uncovered at Treas-
ure Cave (Bar-Adon 1971). These two desert localities have their historical analog in
the Sheihk tombs, such as those recorded in Sinai, which serve as boundary markers
as well as places of annual public gatherings (Marx 1977).

The emergence of cultivating communities during the PPNA (10,300-9,500/300
BP), marked major changes in the subject matter, raw materials and manufacturing
techniques in prehistoric art objects. Clearly shaped female figurines appear for the
first time, signaling, according to Cauvin (1985, 1994), a major shift in the role of

180 BAR-YOSEF

gender within Neolithic societies. Female images are interpreted as the symbols of fe-
cundity — thus expressing biological ‘survival,’ both in terms of human reproduction
and soil fertility. This image is also inseparable from the status of ‘enthroned god-
dess’ or of the ‘mother goddess’ which were perpetuated through Near Eastern cos-
mology into the later pantheons (Jacobsen 1976; Eliade 1989).

The human statues from such PPNB sites as Ain Ghazal, Tell Ramad, Jericho and
Nahal Hemar cave described above, may represent the creation of humans and/or dei-
ties as narrated in the Gilgamesh (Amiran 1962). These statues were probably placed
in specific (perhaps public?) houses and indicate the emergence of communal shrines
or even temples. The burial of used and broken divine representations heralds a pat-
tern of behavior that marks Near Eastern religions dating to later times (Garfinkel
1994). A similar interpretation is suggested for the large human features carved in
stone at Nevali Cori (Mellink 1992) in the Euphrates Valley (Turkey).

Plastered skulls are known from the deposits of PPNB. Stone masks have been
found at Nahal Hemar cave, Basta and a few plundered sites. It seems that the devel-
opment of social hierarchies from PPNA times onward motivated the creation of ‘art
objects’ for various levels of interaction — the family, the village and the ‘tribe’. Plas-
tered skulls could have been kept in the main household within quarters shared by
several families belonging to the same clan or lineage. In addition, specific collec-
tions of paraphernalia were cached in particular locales, such as at Nahal Hemar cave.

I assume, given the different Neolithic contexts described by Voigt (1983, 1991),
that the small human and animal figurines were instrumental on the family level,
whether within the household or as public donations. Ethnographic examples indicate
that figurines prepared especially for rituals lose their value when the feast is over and
are given to children (Siiger ef a/. 1991).

Animal figurines seem to proliferate during the PPNB probably due to the advent
of domestication. In general, they depict taxa in the bovid family and are not always
shaped with enough detail to enable identification to a particular species. A few ex-
amples found pierced by flint blades have been interpreted as representing hunted
animals (Rollefson 1983, 1986).

Any discussion of ‘prehistoric art’ in the Mediterranean region should not omit
data from the periphery, namely the high mountains and the deserts (Bar- Yosef &
Belfer-Cohen 1992). Research on the emergence of pastoral societies (see Bar-Y osef
and Khazanov 1992) may enable us to understand the numerous rock engravings
found in the Sinai and the Arabian peninsula (Rhotert 1938; Anati 1993). These
mostly represent herded and hunted animals and are simply engraved and pecked.
Sometimes entire scenes are present, such as hunting with the aid of drives (the “de-
sert kites’).

Conclusions

In closing, I return to the question of why there are only few artistic expressions in
the Upper and Epipaleolithic in the Near East.

First, in comparison to Western Europe, we have few sites on hand. Additional
excavations will undoubtedly uncover more art pieces. In spite of this, we can hardly
expect to find such concentrations as those found at Magdalenian sites. In the Levant,
where because of less stable annual precipitation, the formation of stalagmitic crusts
was slower than in Western Europe, parietal art may not have survived. Intensive and
extensive surveys of underground cavities in Israel and the speleological investiga-
tions in Turkey, however, have produced no positive evidence for rock paintings,
carvings or incised designs.

PREHISTORIC IMAGES OF THE LEVANT 181

Southwestern Turkey presents a special case. A few rock paintings have been
recorded at the entrances to the Karain and the Okiizini caves (Anati 1968) but their
age is unknown. The painted pebbles from Beldibi (Bostanci 1964) and the incised
ones from Okiizini are of greater importance because they may indicate a higher inci-
dence of art objects in this area than in the Levant. It should be remembered, however,
that all these items are essentially of Epipaleolithic age, contemporary with the Na-
tufian, even though they come from small hunter-gatherer campsites.

Other explanations for the paucity of art objects may posit that the artistic
manifestations were made of perishables. The survival of a fiber fragment at the
water-logged site of Ohallo II (Nadel 1994) suggests taphonomy may be involved.

All this suggests that we need to seek answers in the social realm. The
Epipaleolithic social context may not have required large aggregation sites. Inten-
sive, frequent social contacts and geographically small mating systems may well ac-
count for the paucity of mobiliary items in the Levant (Hovers 1990).

The major socioeconomic change in the Natufian culture heralded an
organizational restructuring of social groups. The need for alleviating social stress
and for identifying various groups probably resulted in an increase in the use of body
decorations, notational records, and in caching of special objects. This tendency con-
tinued through the Early Neolithic. Increasing sizes of social units which reached the
level of entire biologically viable entities living in one village, such as Catal Htiyiik,
led to the emergence of complex symbolic behaviors for enhancing group cohesion,
and resolving intragroup conflicts.

Finally, as mentioned above, the fundamental lesson to be learned from the
Levantine prehistoric art record is that we need to reverse our questions and ask why
artistic/symbolic manifestations proliferated in Upper Paleolithic Eurasia and Aus-
tralia while in other parts of the world, it emerged in the terminal Pleistocene and pro-
liferated during the Holocene.

Acknowledgments

lam grateful to Meg Conkey and Olga Soffer who invited me to the ‘archaeolog-
ical retreat’ in Oregon where prehistoric art expressions and their potential meanings
were intensively discussed, and to Jean and Ray Auel for their hospitality. The nu-
merous comments and suggestions by Anna Belfer-Cohen and Erella Hovers (Insti-
tute of Archaeology, Hebrew University) on earlier versions of this manuscript were
extremely helpful. The final version greatly benefited from the editorial skills of Olga
Soffer and Margot Fleischman. This paper was written while I was a Varon Scholar in
the Department of Environmental Studies and Energy Research at the Weizmann In-
stitute of Science, Rehovot, Israel. Needless to say, | am solely responsible for all the
shortcomings of this paper.

Literature Cited

Amiran, R. 1962. Myths of the Creation of Man and the Jericho Statues. Bull. Amer. Sch. Ori-
ent. Res, 167:23-25.
Anati, E. 1968. Anatolia's earliest art. Archaeology 211:21-35.
. 1972. Arte Prehistorica in Anatolia. Edizioni del Centro, Valcamonica, Italy.
. 1993. Har KarKom in the Light of New Discoveries. Studi camuni 11. Edizion del Cen-
tro, Capo di Ponte, Italy.

182 BAR-YOSEF

Arensburg, B. 1977. New Upper Palaeolithic human remains from Israel. Pages 208-215 in B.
Arensburg & O. Bar-Y osef, eds., Eretz Israel 13 (Moshe Stekelis Memorial Volume). \srael
Exploration Society, Jerusalem, Israel.

Arensburg, B. & O. Bar-Yosef. 1973. Human remains from Ein Gev I, Jordan Valley, Israel.
Paléorient 1:201-206.

Bahn, P. G, 199], Pleistocene images outside Europe. Proc. Prehist. Soc, 571:91-102.

Bar-Adon, P. 1971. The Cave of the Treasure: The Finds from the Caves of Nahal Mishmar. \s-
rael Exploration Society, Jerusalem, Israel.

Barth, E. K. 1982. Ancient methods for trapping wild reindeer in south Norway. Pages 21-45 in
A. Hultkrantz & O. Vorren, ed., The Hunters: Their Culture and Way of Life. Universitets-
forlaget, Tromso, Norway.

Bar-Y osef, D. E. 1989. Late Paleolithic and Neolithic marine shells in the southern Levant as
cultural Markers, Paper presented at the Shell Bead Conference, Rochester, New York.

. 1991. Changes in the selection of marine shells from the Natutian to the Neolithic.
Pages 629-636 in O. Bar-Yoset & F. R. Vallayed., The Natufian Culture in the Levant. \n-
ternational Monographs in Prehistory, Ann Arbor.

—_. Inpress. “Neolithic Shell Bead Production in Sinai.”

Bar-Yosef, O. 1980. A figurine from a Khiamian site in the lower Jordan Valley. Paleorient 6:
193-200.

—. 1983. The Natufian of the Southern Levant. Pages | 1-42 mC. T. Young, Jr., P. E.L.
Smith & P. Mortensen, eds., The Hilly Flanks and Beyond. The Oriental Institute, Chicago.

. 1984. Seasonality among Neolithic hunter-gatherers in southern Sinai. Pages 145-160

in J. Clutton-Brock & C. Grigson, eds., Animals and Archaeology, 3. Herders and Their

Flocks. British Archaeological Reports International Series 202, Oxford, U.K.

_. 1987. The Late Pleistocene in the Levant. Pages 219-236 in O. Soffer, ed., 7he Pleisto-
cene Old World: Regional Perspectives. Plenum Press, New York.

—__. 1989. Upper Pleistocene Cultural Stratigraphy in Southwest Asia. Pages 154-179 in E.
Trinkaus, ed., Patterns and Processes in Later Pleistocene Human Emergence. Cambridge
University Press, Cambridge, U.K.

Bar-Yosef, O. & D. Alon. 1988. Excavations in the Nahal Hemar Cave. Atigot 18: 1-30,

Bar-Yosef, O. & A. Belfer. 1977. The Lagaman Industry. Pages 42-84 in O. Bar-Yosef and J.
L. Phillips.eds., Prehistoric Investigations in Gebel Maghara, Northern Sinai Hebrew
University, Jerusalem, Israel.

Bar-Yosef, O. & A. Belfer-Cohen. 1988. The Early Upper Palaeolithic in Levantine caves.
Pages 23-41 in J. H. Hoffecker & C. A. Wolf, eds., The Early Upper Paleolithic in Europe
and the Near East. British Archaeological Reports International Series 437, Oxford, U.K.

__. 1991. From sedentary hunter-gatherers to territorial farmers in the Levant. Pages 181-
202 in S. A. Gregg, ed., Benveen Bands and States. Center for Archaeological Investiga-
tions, Carbondale.

1992. From foraging to farming in the Mediterranean Levant. Pages 21-48 i Gebauer,

A.B. & T.D. Price, eds., Transitions to Agriculture in Prehistory. Prehistory Press, Madi-

son.

Bar-Yosef, O. & A. Gopher. In press. An Early Neolithic Village in the Jordan Valley. Ameri-
can School of Prehistoric Research, Cambridge, U.K.

Bar-Y osef, O. A. Gopher, E. Tchernov, & M. E. Kislev. 1991. Netiv Hagdud: an early Neo-
lithic village site in the Jordan Valley. J. Field Arch. 184:405-424

Bar-Yosef, O. & N. Goren. 1973. Natufian remains in Hayonim Cave. Paleorient 11-49-68.

Bar-Y osef, O. & A. Khazanov, eds. 1992. Pastoralism in the Levant: Archaeological Materi-
als in Anthropological Perspectives. Prehistory Press, Madison.

PREHISTORIC IMAGES OF THE LEVANT 183

Bar-Y osef, O. & R. H. Meadow. 1995. The origins of agriculture in the Near East. Pages 39-94
in T. D. Price and A. B. Gebauer, eds., Last Hunters, First Farmers: New Perspectives on
the Prehistoric Transition to Agriculture. School of American Research Press, Santa Fe.

Bar-Yosef, O. & T. Schick. 1989. Early Neolithic organic remains from Nahal Hemar cave.
Natl. Geog. Res.5:176-190.

Bar-Y osef, O., B. Vandermeersch, B. Arensburg, A. Belfer-Cohen, P. Goldberg, H. Laville, L.
Meignen, Y. Rak, J.D. Speth, E. T. Chernov, A.-M. Tiller, & S. Weiner. 1992. The excava-
tions in Kebara Cave, Mt. Carmel. Curr. Anthropol. 335:497-550.

Belfer-Cohen, A. 1988. The appearance of symbolic expression in the Upper Pleistocene of the
Levant as compared to western Europe. Pages 25-29 in O. Bar-Y osef, ed., La Pensée. Uni-
versité de Liege, Liege, France.

. 1991a. Art items from Layer B, Hayonim Cave: A case study of art in a Natufian con-
text. Pages 569-588 in O. Bar-Y osef & F. R. Valla, eds., The Natufian Culture in the Levant.
International Monographs in Prehistory, Ann Arbor.

. 1991b. The Natufian in the Levant. Ann. Rev. Anthropol. 20:167-86.

_ In press. Rethinking social stratification in the Natufian culture: The evidence from
burials in Archaeology of Death.

Belfer-Cohen, A. & O. Bar-Yosef. 1981. The Aurignacian at Hayonim Cave. Paléorient 7:
19-42.

_n.d. Early sedentism in the Near East: A bumpy ride to village life (13,000-8,000 BP).
Paper presented at the 1993 Meeting of the Society for American Archaeology, Anaheim,
CA.

Belfer-Cohen, A. & E. Hovers. 1992. In the eye of the beholder: Mousterian and Natufian burt -
als in the Levant. Curr. Anthropol. 334:463-471.

Belfer-Cohen, A. A. Schepartz & B. Arensburg. 1991. New biological data for the Natufian
populations in Israel. Pages 411-424 in O. Bar-Yosef & F. R. Valla, eds., The Natufian Cul-
ture in the Levant. International Monographs in Prehistory, Ann Arbor.

Bergman, C. A. 1987. Ksar Akil, Lebanon: A technological and Typological Analysis of the
Later Palaeolithic Levels of Ksar Akil. BAR International Series 329, Oxford, U.K.

Bergman, C. A. & C. B. Stringer. 1989. Fifty years after: Egbert, an Early Upper Paleolithic ju-
venile from Ksar Akil, Lebanon. Paléorient 15:99-112.

Betts, A. & S. Helms. 1986. Rock Artin eastern Jordan: “Kite” carvings? Paléorient |2:67-72.

Binford, L. R. 1983. /n Pursuit of the Past. Thames and Hudson, London, U.K.

Bostanci, E. Y. 1964. Important artistic objects from the Beldibi excavations. Antropoloji |:
21-31.

Boyd, B. & J. Cook. 1993. A reconsideration of the 'Ain Sakhri' figurine. Proc. Prehist. Soc
59:399-405.

Byrd, B. F. & C.M. Monahan. 1995, Death, mortuary ritual, and Natufian social structure. J
Anthrop. Arch. 14:251-287.

Cambel, H. 1980. Chronologie et organisation de l'espace a Cayénii. Pages 531-553 in J. Cau-
vin & P. Sanlaville, eds., Préhistoire du Levant. CNRS, Paris, France.

Cauvin, J. C. 1972. Les Religions Néolithiques de Syro-Palestine. Maisonneuve, Paris, France.

. 1977. Les fouilles de Mureybet (1971-1974) et leur signification pour les origines de la
sédentarisation au Proche-Orient. Ann. Amer. Sch. of Orient. Res. 44:19-48.

. 1985. La question du 'Matriarcat Préhistorique' et le rdle de la femme dans la prehis-
toire. Pages 7-18 in A. M. Vérilhac, ed., La Femme Dans le Monde Mediterranéen. Maison
de l'Orient, Lyon, France.

. 1994. Naissance des Divinités, Naissance de l'Agriculture. CNRS, Paris, France.

Clarke, D. L. 1978. Analytical Archaeology. Metheun, London, U.K.

184 BAR-YOSEF

Conkey, M. W. 1984. To find ourselves: Art and social geography of prehistoric hunter-
gatherers. Pages 253-276 in C. Schrire, ed., Past and Present in Hunter-gatherer Studies.
Academic Press, New York.

Contenson, H. De 1967. Troisieme campagne de fouilles a Tell Ramad, 1966: Rapport prélimi-
naire. Annales Archeologiques Arabes Syriennes 117:17-24.

Davidson, I. & W. Noble. 1989. The archaeology of perception: traces of depiction and lan-
guage. Curr. Anthrop, 30:12-55.

Davis, S.J.M.& F. Valla. 1978. Evidence tor domestication of the dog, 12.000 years ago in the
Natufian of Israel. Nature 276:608-610.

Eccles, J. 1989. Evolution of the Brain: Creation of the Self. Routledge, London, U.K.

Edwards, P. 1991. Wadi Hammeh 27: An Early Natufian site at Pella, Jordan. Pages 123-148 in
O. Bar-Y oset & F. R. Valla, eds., The Natufian Culture in the Levant. International Mono-
graphs in Prehistory, Ann Arbor.

Eliade, M. 1989. Shamanism: Archaic Techniques of Ecstasy. Arkana, London, U.K.

Frison, G. C. 1978. Prehistoric Hunters of the High Plains. Academic Press, New Y ork.

Gamble, C. 1991. Social context for European Palaeolithic art. Proc. Prehist. Soc. 57:3-15.

Garfinkel, Y. 1994. Ritual burial of cultic objects: The earliest evidence. Camb. Arch. J.
42:159-188.

Garrod, D. A. E. & D. M. Bate. 1937. The Stone Age of Mount Carmel. Clarendon Press, Ox-
ford, U.K.

Garstang, J. & J.B. E. Garstang. 1940. The Story of Jericho. Hodder and Stroughton, London,
U.K.

Gilead, I. 1991. The Upper Paleolithic in the Levant. J. World Prehist. 52:105-154.

Goren-Inbar, N. 1986. A figurine from the Acheulian site of Berekhat Ram. J. /srael Prehist
Soe. 19:7-12.

. 1990. Quneitra: A Mousterian site on the Golan Heights. Hebrew University, Jerusa-

lem, Israel.

Goring-Morris, A. N. 1991. The Harifian of the southern Levant. Pages 173-216 in O. Bar-
Yosef & F.R. Valla, eds., The Natufian Culture in the Levant. International Monographs in
Prehistory, Ann Arbor.

Henry, D. O. 1989. From Foraging to Agriculture: The Levant at the End of the Ice Age. Uni-
versity of Pennsylvania Press, Philadelphia.

Hershkovitz, I.,O. Bar-Yoset & B. Arensburg. 1994. The pre-pottery Neolithic populations of
south Sinai and their relations to other circum-Mediterranean groups: Anthropological
study. Paleorient 20:59-84.

Hershkovitz, |. & A. Gopher. 1990. Paleodemography, burial customs, and food-producing
economy at the beginning of the Holocene: A perspective from the southern Levant. J. /s-
rael Prehist. Soc. 23:9-47.

Hillman, G. C. & M.S. Davies. 1990. Measured domestication rates in wild wheats and barley
under primitive implications. J. World Prehist. 4:157-222.

Hovers, E. 1990. Artin the Levantine Epi-Paleolithic: An engraved pebble from a Kebaran Site
in the Lower Jordan Valley. Curr. Anthropol. 31;317-322.

Jacobsen, T. 1976. The Treasures of Darkness: A History of Mesopotamian Religion. Yale
University Press, New Haven.

Johnson, G. A. 1982. Organizational structure and scalar stress. Pages 389-421 in A. C. Ren-
frew, M. J. Rowlands, & B. A. Segraves, eds., Theory and Explanation in Archaeology.
Academic Press, London, U.K.

Kaufman, D. & A. Ronen. 1987. La Sépulture kébarienne géométrique de Néve-David Haifa,
Israel. L'Anthropologie 911:335-342.

PREHISTORIC IMAGES OF THE LEVANT 185

Kenyon, K.M. & T. A. Holland. 1983. Excavations at Jericho, Vol. V: The Pottery Phases of
the Tell and Other Finds. British School of Archaeology, London, U.K.

Leroi-Gourhan, A. 1965. Treasures of Prehistoric Art. Abrams, New York.

Lewis-Williams, J. D. 1984. Ideological continuities in prehistoric southern Africa: The evi-
dence of rock art. Pages 225-252 in C. Schrire, ed., Past and Present in Hunter-gatherer
Studies. Academic Press, New Y ork.

. 1991. Wrestling with analogy: A methodological dilemma in Upper Palaeolithic art re-
search. Proc. Prehist. Soc. 57:149-162.

Lewis-Wilhiams, J. D. & T. A. Dowson. 1988. The signs of all times: entoptic phenomena in
Upper Palaeolithic art. Curr. Anthropol. 24: 201-245.

. 1989. Theory and data: A brief critique of A. Marshack's research methods and posi-
tions on Upper Paleolithic shamanism. Rock Art Res. 6:38-53.

Lieberman, D. E. 1993. The rise and fall of seasonal mobility among hunter-gatherers. Curr
Anthropol. 345:599-631.

Liubin, V. P. 1989. Palaeolithic of Caucasus. Nauka, Leningrad, Russia.

Margalit, B. 1983. The “Neolithic Connexion” of the Ugarit poet of Aght. Pa/eorient 92:93-98.

Marks, A. E. 1983. The Middle to Upper Paleolithic transition in the Levant. Pages 51-98 in F.
Wendort & A. E. Close, eds., Advances in World Archaeology. Academic Press, New Y ork.

. 1993. The Early Upper Paleolithic: The view from the Levant. Pages 5-22 in H. Knecht,
A. Pike-Tay & R. White, eds., Before Lascaux: The Complete Record of the Early Upper
Paleolithic. CRC Press, Boca Raton.

Marks, A. E. & C.R. Ferring. 1988. The Early Upper Paleolithic of the Levant. Pages 43-72 in
J. F. Hoffecker & C. A. Wolf, eds., The Early Upper Paleolithic. BAR International Series
437, Oxford, U.K..

Marks, A. E. & D. A. Freidel. 1977. Prehistoric settlement patterns in the Avdat/Aqev area.
Pages 131-158 in A. E. Marks, ed., Prehistory and Paleoenvironments in the Central Ne-
gev, Israel, Volume II, The Avdat/Aqev area, part 2 and the Har Harif. Southern Methodist
University, Dallas.

Marshack, A. 1990. Early hominid symbol and evolution of the human capacity. Pages 457-
498 in P. Mellars, ed., The Emergence of Modern Humans. Edinburgh University Press, Ed-
inburgh, U.K.

Marx, E. 1977. Communal and individual pilgrimage: The region of saints' tombs in South Si -
nai. Pages 29-51 in R. P. Werbner, ed., Regional Cults. Mouton, The Hague, The
Netherlands,

Mellaart, J. 1967. Catal Hiiytik, A Neolithic Town in Anatolia. Thames and Hudson, London,
U.K.

. 1975. The Neolithic of the Near East. Thames and Hudson, London, U.K.

Mellars, P. 1989. Major issues in the emergence of modern humans. Curr, Anthropol. 30:349-

385.

Mellars, P. & J. Tixier. 1989. Radiocarbon accelerator dating of Ksar Aqil Lebanon and the

chronology of the Upper Paleolithic sequence in the Middle East. Antiquity 63:761-768.

Mellink, M. J. 1992. Archaeology in Anatolia. Amer. J. Arch. 96:119-127,

Meshel, Z. 1974. New data about the “Desert Kites”. Te/ Aviv 1:129-143.

Mithen, S. J. 1991. Ecological interpretations of Palaeolithic art. Proc. Prehist. Soc. 57:103-

114.

Muheisen, M. 1985. L'Epipaleolithique dans le gisement de Khoraneh lV. Pa/éorient 11:149-
160.

Nadel, D. 1994. Levantine Upper Palaeolithic-Early Epipalaeolithic burial customs: Ohalo II
as a case study. Paléorient 20:1 13-122.

186 BAR-YOSEF

Neuville, R. 1951. Le Paléolithique et le Mésolithique de Désert de Judée. Masson et Cie,
Paris, France .

Noy, T. 1986. Seated clay figurines from the Neolithic Period, Israel. Pages 63-351 in A. Bo-
nanno, ed., Archaeology and fertility cult in the ancient Mediterranean. B.R. Gruner, Am-
sterdam, The Netherlands.

—_. 1990, Additional aspects of the human figure among the clay figurines from the Yar-
mukian. Eretz Israel 2|:226-232.

—_. 199]. Artand decoration of the Natufian at Nahal Oren. Pages 557-568 in O. Bar-Y osef
& F.R. Valla, eds., The Nanifian Culture in the Levant. International Monographs in Pre-
history, Ann Arbor.

Noy, T. & B. Brimer 1980. Adornment of early Natufian burials. The /srael Mus. News 16:
55-64.

Ohnuma, K. 1988. Ksar Akil, Lebanon: A Technological Study of the Earlier Upper Palaeo -
lithic Levels at Ksar Akil: Vol. Ill: Levels XXV-XIV. BAR International Series 426, Oxford,
U.K.

Olszewski, D. 1. & H.L. Dibble. 1994. The Zagros Aurignacian. Curr. Anthropol. 351:68-75.

Ozdogan, M. & A. Ozdogan. 1989. Cayénii, a Conspectus of Recent Work. Paléorient 151:
65-74.

Perrot, J. 1966. Le gisement natoufien de Mallaha (Eynan), Israel. L'Anthropologie 70:437-
484.

—_. 1968. La Préhistoire palestinienne. Pages 286-446 in L. Pirot, ed., Supplement au dic-
tionnaire de la Bible 8. Letouzey and Ane, Paris, France.

Perrot, J. & D. Ladiray. 1988. Les Hommes de Mallaha Evnan Israel. Association Palcorient,
Paris, France.

Phillips, J. L. 1988. The Upper Paleolithic of the Wadi Feiran, Southern Sinai. Paleéorient 142:
183-200.

Reese, D. R. 1991. The trade of Indo-Pacific shells into the Mediterranean basin and Europe.
Oxford J. Arch. 10; 159-196

Rhotert, H. 1938. Transjordanien. Verlag Strecker und Schréder, Stuttgart, Germany.

Rollefson, G. O. 1983. Ritual and ceremony at Neolithic Ain Ghazal Jordan. Pa/eorient
92:29-38.

7 _ 1986. Neolithic Ain Ghazal Jordan: Ritual and ceremony II. Paléorient 12:45-52.

Rollefson, G. O. & I. Kohler-Rollefson. 1993. PPNC adaptations in the first half of the 6th mil-
lennium B.C. Paléorient 19:33-42.

Rollefson, G. O., A. H. Simmons & Z. Kafafi. 1992. Neolithic cultures at ‘Ain Ghazal, Jordan.
J. Field Arch. \9:443-470.

Ronen, A. & B. Vandermeersch. 1972. The Upper Paleolithic sequence in the cave of Qatzeh
(Israel). Quaternaria 16: 189-202.

Rosenberg, M. & Davis, M. 1992. Hallam Cemi Tepesi, an early aceramic Neolithic site in
eastern Anatolia: Some preliminary observations concerning material culture. Anatolica
T8: 1-18.

Solecki, R. S. 1980. Art motifs and prehistory in the Middle East. Pages 59-77 in S. Diamond,
eds., Theory and Practice: Essays Presented to Gene Weltfish. Mouton, The Hague, The
Netherlands.

Stekelis, M. 1972. The Yarmukuian Culture. The Hebrew University, Jerusalem, Israel,

Stekelis,M. & T. Yizraeli. 1963, Excavations at Nahal Oren, preliminary report. /srae! Explor
J V34:1-12

Suger, H., S. Castenfeldt, & M. Fentz. 1991. Small functional items and regeneration of soci-
ety, dough figurines from the Kalash people of Chitral, northern Pakistan. Folk 33:37-66.

PREHISTORIC IMAGES OF THE LEVANT 187

Tchernov, E. 1991. Biological evidence for human sedentism in southwest Asia during the Na-
tufian. Pages 315-340 in O. Bar-Y osef & F. R. Valla, eds., The Natufian Culture in the Le-
vant. International Monographs in Prehistory, Ann Arbor.

Tixier, J. 1974. Poingon décore du Paléolithique Supérieur a Ksar Aqil Liban. Pa/éorient 2:
187-192.

Valla, F. R. 1987. Chronologie absolue et chronologies relatives dans le Natoufien. Pages
267-294 in O. Aurenche, F. Evin & F. Hours, eds., Chronologies in the Near East: Relative
Chronologies and Absolute Chronology 16,000-4,000 B.P. British Archaeological Reports
International Series 3791, Oxford, U.K.

. 1990. Le Natoufien: Une autre fagon de comprendre le monde? J. /srael Prehist. Soc

23:171-175.

. 1991. Les Natoufiens de Mallaha et l'espace. Pages 111-122 in O. Bar-Yoset & F. R.
Valla, eds., The Natufian Culture in the Levant. International Monographs in Prehistory,
Ann Arbor.

Vandermeersch, B. 1969a, Découverte d'un objet en ocre avec traces d'utilisation dans le
Mousteérien de Qafzeh Israél. Bu//. de la Soc. Préhist. de France 66:157-158.

. 1969b. Paléontologie humaine: Les nouveaux squelettes moustériens découverts a
Qatzeh (Israel) et leur signification. Comptes Rendus des Séances de l'Académie des Sci-
ences 268: 2562-2565.

Velichko, A. A. 1984. Late Quaternary Environments of the Soviet Union. University of Min-
nesota Press, Minneapolis.

Voight, M. M. 1983. Hajji Firuz Tepe, Iran: The Neolithic Settlement. The University Mu-
seum, Philadelphia.

. 1991. The goddess from Anatolia: An archaeological perspective. Orient. Rug Rev
112732239)

Weinstein-Evron, M. & A. Belfer-Cohen. 1993. Natufian figurines from the new excavations
of the el-Wad cave, Mt. Carmel, Israel. Rock Art Res. 102:102-106.

Weinstein-Evron, M., B. Lang, S. Hani, G. Steinitz, & D. Kaufman. 1995. K/Ar dating as a
means of sourcing Levantine Epi-Palaeolithic basalt implements. Archaeometry 371:37-
40.

White, R. 1989. Production complexity and standardization in early Aurignacian bead and pe n-
dant manufacture: Evolutionary implications. Pages 366-390 in P. Mellars & C. Stringer,
eds., The Human Revolution: Behavioural and Biological Perspectives on the Origins of
Modern Humans. Edinburgh University Press, Edinburgh, U.K.

Wiessner, P. 1984. Reconsidering the behavioral basis for style: A case study among the Kala-
hari San. J. Anthropol. Arch. 3:190-234.

. 1989. Style and changing relations between the individual and society. in I. Hodder, ed.,
The Meaning of Things: Material Culture and Symbolic Expression. Unwin Hyman, Lon-
don, U.K.

Wilson, D. S. & E. Sober. 1994. Reintroducing group selection to the human behavioral sci -
ences. Behav. Brain Sci. 17:585-609,

Wreschner, E. 1976. The red hunters: Further thoughts on the evolution of speech. Curr, An-
thropol. 17:717-719.

Wright, G. A. 1978. Social differentiation in the Early Natufian. Pages 201-233 in C. L. Red-
man, M. J. Berman, E. V. Curint, W. T. J. Langhorne, N. M. Versaggi & J.C. Wanser, eds.
Social Archaeology: Beyond Subsistence and Dating. Academic Press, New York.

Yalcinkaya, I., J.-M. Leotard, M. Katal, M. Otte, O. Bar-Yosef, I. Carmi, A. Gautier, E. Gilot P.
Goldberg, J. Kolzowlski, D. Lieberman, I. Lopez-Bayon, M. Pawliokowski S. Thiebault,
V. Ancion, M. Patou, A. Barbier & D. Bonjean. 1995, Les occupations tardiglaciaires du
site d'Okuzini (sud-ouest de la Turquie): Résultats préliminaires. L'Anthropologie
100:562-S85.

189

When the Beasts Go Marchin’
Out! The End of Pleistocene Art
in Cantabrian Spain

Manuel R. Gonzalez Morales

Universidad de Cantabria
Avenida de los Castros s/n
39005 Santauder, Spain

Recently, there has been a good deal of discussion about the beginning of sym-
bolic expression in human societies, i.e., the beginning of art among Paleolithic
hunter-gatherers linking it toa variety of causal variables. An important part of the de-
bate has been focused on the chronological coincidence, or lack of it, between the first
symbolic expressions preserved in the archaeological record and the appearance of
anatomically modern humans in Europe or Australia, and on the definition of what
can be named as unequivocal symbolic expressions, and their earliest chronology
(Conkey 1983, Chase & Dibble 1987, Davis 1986, Duff et a/. 1992).

Much less interest has focused on the other end of the question — that of the end of
Upper Paleolithic art and symbolism. Any attempt at global interpretation must deal
necessarily with both sides of the problem: The origins and the ends, because under-
standing one can help understand the other.

Elsewhere (Gonzalez Morales 1991), I have examined the subject matter depicted
in Upper Paleolithic Cantabrian parietal art, focusing on two issues: Temporal vari-
ability and the possible correspondence between the subjects depicted and the avail-
able resource base. This study revealed many problems extant in studying art,
including defining and describing the subject matter represented, and dealing with
long-standing biases in the study of this subject matter. In this chapter I focus on these
methodological problems because their resolution must precede any attempt to reach
higher orders of inference.

The study of the subject matter depicted in Paleolithic parietal art has often been
dealt with in a sketchy and descriptive manner, glossing over many issues that could,
and should, be given more detailed discussion. This glossing has led us away from
verifiable observations to subjective concepts and descriptions. Because of methodo-
logical conventions, some of these biases are repeated in the literature. In hopes to
critically challenge some of them, I introduce an alternative view below. I restrict my
discussion to parietal art only for several reasons: There are differences between the

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

190 GONZALEZ MORALES

representation of subject matter of rock art and mobiliary art which reflect differences
in their immediate contexts of production and use. In addition, we control better the
chronology of mobiliary art presents than of parietal art. While I focus on Cantabrian
art, | suggest that many of my observations are applicable to Paleolithic parietal art as
a whole.

Temporal Variability in the Subject Matter of Parietal Art

Some authors study the subject matter in Paleolithic art synchronically, without
taking into account the chronology of the entire tradition. This denies a historical di-
mension to Paleolithic art and negates the dynamics of change throughout the thou-
sands of years of its development. Many critics, among them Ucko & Rosenfeld
(1967), Layton (Layton 1977, 1991), Conkey (Conkey 1985, 1987), agree that vari-
ous “artistic traditions” could have appeared and disappeared during the long course
of the Upper Paleolithic. This suggests that general conclusions about all of Paleo-
lithic art without considering its temporality are gross oversimplifications. If we con-
sider artistic representations to be extensions of human activities, we cannot discuss
any change through time while assuming that the art remained invariable.

If we divorce the study of art, or of its themes from the study of other human activi-
ties, we assume that “artistic forms” can be analyzed apart from economic and social
domains of human existence (and vice versa ). Insucha case, the “artistic forms” take
ona phenomenological, not substantive, status. Specifically we think we can under-
stand their function and changes in those functions independently of the behavioral
context. From such a perspective, the disappearance of parietal art at the end of the
Upper Paleolithic becomes a simple anecdote or a coincidence in time, one of no im-
port for understanding the system as a whole.

I have argued elsewhere that, in studying prehistoric societies, we cannot decou-
ple the economic system from the social structure or from material manifestations of
ideology (7.e., art). All are integral to the structural web that defines those societies.
Because of this | have argued that the lack of concern with these communities during
the transition from the Upper Paleolithic to the Mesolithic avoids the very important
problem of the disappearance of art, specifically the disappearance of the material
evidence of a long-lived ideology (Gonzalez Morales 1991:212). Temporal changes
in the subject matter of this rock art have particular relevance to this debate because:

“The contradictions that emerge in the Cantabrian Magdalenian between the parietal
and the mobile art indicate how supposedly ritualized art forms (those found in sanctu-
ary contexts) retain a more conservative expression of content, while others, related to
mundane (e¢.g., objects in daily use), seem to reflect much more directly their thematic
(in good part, productive) content.” (Gonzalez Morales 1991:215)

In this paragraph I am arguing that, when the faunal records from archaeological
sites (Altuna 1972: Straus 1977) are given a temporal dimension and then correlated
with the parietal figures, the difference between the fauna depicted in the art and that
present at sites tends to vary over time. This suggests that those species which were
the core subjects (bison/aurochs and horses) were the most important to the subsis-
tence base at the beginning of the Upper Paleolithic, and perhaps reflected myths
which lay behind them. This core persists, with variations and additions of comple-
mentary elements, until the very end of the Paleolithic, when hunting became special-
ized around other species — red deer in Cantabria, reindeer in the Pyrénées and the

PLEISTOCENE ART OF SPAIN 19]

Perigord, and ibex in certain mountainous areas. This shift in fauna depicted vs. ex-
ploited shows us that the subsistence base diverged further and further from the spe-
cies at the center of the Late Paleolithic ideological universe, as represented in the art.
The divergence between taxa exploited and those possibly important to the traditional
myths, which were increasingly ritualized and divorced from the “real world,” seems
to be magnified through time. I suggest that this divergence between and ritual must
have figured centrally in the final crisis which caused the abandonment of the great
tradition of Paleolithic parietal art.

As noted, the correlation between what was hunted and what was depicted de-
pends on time. To study it we first need to establish a firm chronology for figures and
compositions and to correlate this chronology with the faunal record from the sites.
Furthermore, we should also consider the social requirements associated with differ-
ent hunting strategies.

In order for this study to be objective, however, we first need to specify the types
of subjects depicted, their chronology, and their relation to subsistence evidence at
each point in time. This task is far more complex than it appears. First, there are par-
ticular problems in documenting Paleolithic parietal art. In the Cantabrian region, this
documentation is often based on research dating to the beginning of this century,
which relied on a methodology very different from contemporary research. I dis-
cussed this problem several years ago (Gonzalez Morales 1989), and showed that ear-
lier techniques of observation and recording, the criteria of identification, and the
inventory of subjects depicted in the parietal art are less than reliable when compared
to those used today. In spite of this, many of the conclusions and observations pre-
sented in the literature are based on these earlier observations.

Bases of the Traditional Systematizations

Most of the studies dealing with the subject matter depicted in Cantabrian Paleo-
lithic art tend to follow the same kind of analytic scheme: 1) first, presenting a count
of animals represented which are considered to be the most fundamental aspect of the
art; 2) then presenting a separate count of anthropomorphic figures or images of ques-
tionable attribution; and finally 3) offering a separate count of signs. This organiza-
tional strategy cuts across other analytic techniques, such as those dealing with
chronology or with the methods of production. The result of this approach leads to a
model where the actual representations seem to be produced by the accidental inter-
section of the other suites of variables, which are thus seen as totally independent of
each other.

For example, a deer figure is considered primarily a deer figure, and the fact that it
is a deer and not a horse or a quadrangular sign is what is considered most relevant.
Other variables, such as whether this deer is painted in black or red, with filled spaces
or in outline, or is executed in one technique or another, or corresponds to some
chronological period or “stylistic phase,” or is painted in one physical position or an-
other, are all seen as less relevant. Such a hierarchical organization of elements
crosses the tenuous line separating “subject” from “meaning,” especially when done
in order to determine “meaningful compositions” and treat them statistically (which,
which few exceptions, means simple percentages). In such an approach we clearly
assume that we are dealing with “meaningful sets” of figures, and not with figures as
isolated entities.

192 GONZALEZ MORALES

Conclusions about the subject matter of all Paleolithic art, which fail to take into
account all of these other variables — especially time — are problematic and foster even
more erroneous generalizations, if only because of the accumulation of errors.

While in this chapter I do not address all problematic assumptions used in the
study of the subject matter of Paleolithic art, I next turn to some of the more seminal
ones. First, we have a lack of consensus in the definitions of some terms, especially
“subject,” “motif,” etc. Loosely speaking, we understand “subject” to be that which
the figures represent. This connotation, however, requires further clarification.
Subject-as-representation makes an immediate connection between the materiality of
the figure and some element of the “real” or “imaginary” world which we take to be
the “model” or “inspiration,” in the sense of Laming-Emperaire (1962) and Leroi-
Gourhan (1963). Here we encounter one of the most common pitfalls of evaluating a
set of "subjects" at a particular Paleolithic art site: Assuming that the subjects-as-
representations themselves maintain a symmetrical relation with the “real world”
models they depict, and assuming, at the same time, that they are organized according
to their own logical scheme, independent of that reality. This happens, for example,
when we draw inferences about the natural environment from a study of associations
among animal figures, and in doing so assume that the associations are a direct reflec-
tion of the environment, while at the same time we consider that the relative spatial or-
ganization, or the presence or absence of certain specific subjects has an essentially
symbolic significance.

Furthermore, when researchers identify parietal art “subjects,” their very act of
identification is a source of conflict. Ucko (1989) has underscored the inaccuracies
and inconsistencies that typically accompany the attempts to identify specific animal
figures and the high degree of subjectivity inherent in this process. Often, the strict
identification of an “art subject” depends on the observer and not on that which is rep-
resented (Clottes 1989; Lorbranchet 1989). I address this point below.

Zoological Subjects and the “Real World”

One of the most common topics in the study of Paleolithic art deals with the rela-
tionship between that which was painted and that which was eaten, often done by re-
searchers not familiar with the general study of rock art. These studies begin with
subjective identifications of the species and with tracking these through time. This is
done in spite of the fact that the assumed association between “hunted fauna” and
“painted fauna,” between the art and the faunal assemblages in the caves, is a tricky
and a poorly understood one. Some of these scholars argue that Paleolithic artists
generally painted things other than what they hunted and see this as an essential qual-
ity of their art (e.g., Altuna 1983, 1994; Criado Boado & Penedo Romero 1989; Rice
& Paterson 1985, 1986). These studies ignore variability through time. Since we have
considerable changes in environment and in parietal art throughout the Upper Paleo-
lithic, it is impossible to generalize about the total time span from a single moment,
and vice versa.

Second, when looking at the association between art and food, we need to define
precisely what unit of measurement we are using. To do so using the proportions in
the number of animal bones 1s as different from looking at the minimum number of in-
dividuals (MNI), as it is from considering the meat weight represented by those indi-
viduals. Even when looking at meat weights as reflecting the relative economic

PLEISTOCENE ART OF SPAIN 193

importance of the taxon, we still don't know how this can be associated with the quan-
titative identification of the animals depicted.

Next, the sites on hand represent only a fraction of the landscape occupied by hu-
man groups. Faunal remains found ina specific level of a specific cave reflect only a
segment of the groups' activities. Logic dictates, therefore, that the scale of analysis
of the associations between art and food needs to be done on a broader geographic
scale that a single site.

The same logic also dictates that chronological scales used in such research must
be more finite. Although many scholars recognize the problems with considering all
Paleolithic art as a homogeneous global phenomenon, they nonetheless do so in most
of their works (Criado Boado & Penedo Romero 1989; Rice & Paterson 1985, 1986).
In discussing the degree of similarity or difference between “real world” species and
the art, bisons are cheerfully lumped with horses, red deer with hinds, reindeer with
mammoths — all without taking into account the differential temporal distribution of
these different representations. Such a loose quantitative view, one approaching Pa-
leolithic art from a synchronic perspective, emphasizes “structure” instead of change.
The effect of this is to neglect or dismiss the fact that Paleolithic art is a phenomenon
of great temporal depth where the very changes in the art and their relationship to so-
cial transformations experienced by the artists are central — not peripheral — to our un-
derstanding of the art.

The lack of congruence between food and art can be seen at Altamira. A quantita-
tive analysis of the faunal remains recovered differs considerably from the animal fig-
ures represented in the art (primarily bison). The Magdalenians occupying the site,
however, left behind a faunal inventory dominated by red deer (proportions of re-
mains). Using MNI-based meat weights rather than proportions, however, shows the
dominance of bison and horse (Gonzalez Morales 1994). This exercise shows that
you can reach different conclusions depending on how you count.

Furthermore, we bring subjectively to such “calculations” by our inherent
assumptions that the “importance” or cultural significance is synonymous with the
number or numeric relevance of the things being counted. When we look for propor-
tional similarities or differences between hunted fauna and painted fauna, we adopt
modern criteria of quantification, that is, that the importance of a species is always
measured by the number of individuals represented, not by its position, its size, or its
possible symbolic relationship to other decorative or auxiliary elements. Yet ethno-
graphical data clearly show that there is a high degree of variability between quantita-
tive representation and cultural importance (see Rosenfeld, this volume).

In this discussion my intent is not to argue for absolute relativity. I believe, for ex-
ample, that the cooccurrence of particular depictions of painted animals throughout
the entire sequence of Upper Paleolithic art — like bison/aurochs and horses — is cen-
tral to its interpretation. Rather, what I question is the possibility of making sound as-
sociations based on solely quantified correlations — ones that assume that they convey
equivalent symbolic values.

In sum, my response to the question of whether or not animals good to eat were
also good to paint is a qualified yes: The art primarily portrays terrestrial ungulates
that were actually consumed by humans. I would qualify this yes, however, with two
points: 1) this generalization ts valid for all Paleolithic cave art, but without any strict
numeric equivalencies; and 2) this relationship between food and art seems to have
been continually modified throughout the Upper Paleolithic. As I and others have ar-
gued elsewhere, this was more true during the earlier phases than the later ones

194 GONZALEZ MORALES

(Gonzalez Morales, 1991; Gonzalez Sainz, 1989), even with individual exceptions
like Grotte Chauvet and its very early dates, according to preliminary results .

Animals and Signs: “Mundane” or “Ritual”?

Our ideas about the subject matter of Paleolithic art often presume its “non-
mundane” character, the scarcity of representation of human behavior, of tools and of
weapons. Paleolithic art is described as essentially “animalist.” This aspect is taken
for granted by many and warrants little discussion. As noted above, in discussing art
motifs we inevitably fall into the trap of beginning with the animal figures them-
selves, and center our discussion around them. Doing so ignores another category of
subject matter which, at least in quantity, equals that of animal representations —
namely signs. In recent years, there has been a tendency to consider abstract and geo-
metric signs as nonfigurative elements ina strict sense, and to see them as opposed to
the world of animal figures. From this we infer that they are “nonrepresentational” in
character and infer a lack of association between the signs and objects or beings in the
real world.

If, however, we reexamine analytic work prior to Laming-Emperaire (1962) and
Leroi-Gourhan (1971), we find abundant discussion of a variety of hunting tools as
well as animals killed with different types of spears and traps. It is these two scholars
who initiated a change in theoretical perspective, one which ignored a consideration
of artifacts used in daily life and overemphasized the variety of figures supposedly
derived from the schematization of the masculine sex (“open” or linear motifs) and of
the feminine sex (“closed” or “full” motifs).

There is clearly some circularity of reasoning pervading some of our common atti-
tudes about the relationship between “figurative subjects” and “abstract subjects” in
Paleolithic art. The rare number of explicit human representations argues for a sys-
tem where the “natural world” was the focus of attention. The absence of human fig-
ures suggests that human activities are not represented in parietal art. This leap of
faith leads us to search for alternative, symbolic meanings for the figures that, without
this prejudice, may have explicit and straightforward meaning, which, in fact, it fre-
quently does in the eyes of the “uninitiated” observer. For example, in the case of
Spanish Levantine art, the lines over animals can been seen as a “naturalist” represen-
tations of projectile points, and in Paleolithic art as “abstract signs.” Which interpre-
tation one prefers clearly is determined by one's theoretical viewpoint and not by the
actual representations.

Thus, the idea that the Paleolithic “signs” do not represent commonly used objects
results from preexisting theoretical biases about the meaning of the art. Such an as-
sumption depends on our ability to adapt our contemporary Western visual reading
codes to interpret representations executed within a very different context. It also de-
pends on our limited ability to recognize possible artifact forms in the art which, due
toa variety of postdepositional processes, no longer exist in the archeological record.

Bahn (1986) and more recently Mithen (1988, 1990, 1991) see an association be-
tween the signs in Paleolithic art and daily hunting activities. In this they reject the in-
terpretation of the signs as abstract representations. Their “naturalist” identification
of the signs, however, is different. In their works we find an extension of the idea that
Paleolithic art is essentially linked to representations of animal figures, now broad-
ened to include parts of those animals, such as their footprints. The reinterpretation of
“vulvae” as animal footprints, however, is not novel but an extension of research of

PLEISTOCENE ART OF SPAIN 195

Bourdier (1967:272) and Gomez-Tabanera (1971) using ethnographic data. Bahn
(1986) challenged the equivocal identification ofa series of Paleolithic signs as repre-
sentations of the female sex. For example, citing different contexts in which similar-
looking forms, ones systematically considered “vulvas” in Paleolithic art, are under-
stood as, among other things, animal footprints across various kinds of ground sur-
faces. In doing so he called attention to the subjectivity inherent in one identification
of signs in Paleolithic art.

Using the idea that much of the abstract signs mark calendaric events or note sea-
sonal changes, Marshack (1972) has identified certain signs as probable representa-
tions of plants or other mundane objects such as traps. Recently, Mithen (1990,
1991) went one step further. Within the context of a general interpretation of Paleo-
lithic art as a system for collecting and transmitting strictly practical information
about hunting, he interprets many signs as animal footprints, as strings of footprints,
prints surrounding vegetation and so forth. My primary objection to Mithen's work is
his neglect of time: He considers all parietal art to have been endowed with the same
meaning and function for more than 20,000 years.

Although many of these hypotheses that signs as related to objects or practical sea-
sonal activities are problematic, if not impossible to verify, they serve to illustrate the
lack of agreement among researchers about the abstract nature of the signs and their
exclusive association, as per Leroi-Gourhan (1971) with male and female sexual or-
gans.

The earlier tradition also diverges from contemporary research in the classifica-
tion systems used. Layton (1987:24-34) has shown how, in the European tradition,
one can distinguish two different ways of approaching the problem of signs and of the
figures which do not strongly suggest a natural model. Some researchers assume that
the signs had some meaning and assume that the artist intended to convey it. Others
tend to group series of signs together because they appear to have the same meaning to
the researcher, but not necessarily the artist. The first is an “essentialist” perspective,
the second, explicitly “formalist.” The two typically are not distinguished by many
researchers and one is left with an impression that vacillates between both perspec-
tives without ever achieving a critical analysis of the change in viewpoint.

Signs Without Animals

The above discussion also raises another basic issue — that of defining Paleolithic
art as a naturalistic art, essentially comprised of animal subjects accompanied by
signs. In actuality, there are whole sets of depictions in which only signs appear with-
out any associated animals, and many others where it is not possible to know if the
signs really accompany the animal figures or were painted as isolated, independent
groups. This is especially apparent on the Cantabrian coast of Spain, where some-
times we find only signs depicted in painted caves. These signs are not just isolated
representations but sometimes appear in complex compositions — as Entrecueves, La
Herreria, Balmori, El Tebellin and Mazaculos in Asturias or as Fuente del Salin and
Santian in Cantabria. In other cases, panels of signs alternate with animal figures,
showing differences in the methods of execution as well as composite superpositions.

Such a variability in the associations between animal figures and abstract signs
makes one consider the situations where signs were exclusive, or numerically domi-
nant in the composition. When we assume, a priori, that the signs accompany the ani-
mal figures and are limited to a secondary role, rather than the animal figures

196 GONZALEZ MORALES

accompany the signs, it is just one more example of oversimplification. Our finite
ability to identify the signs and our biases that the “predominance” of animal figures
represents the very core of Paleolithic art, make the cases of compositions without
signs aberrant if not incomprehensible. This, again, reflects the subjectivity of our
viewpoint. Layton (1987:23-38) has noted that many archaeologists working with
Paleolithic art believe that the “interpretational quality” of the figures is transparent
and that they are easily classified into “motifs” which signify bison, horse or deer.
Some, however, depending on the amount of information in the figures, are more
transparent than others. The problem with signs, on the other hand, is that many of
them had to convey enough information so that the artists and their contemporaries
could identify them, but convey little information to us. This needs to be coupled with
the possibility that ambiguity was intended to begin with.

Thus we have today a situation where signs are formally defined by exclusion: A
type to which we assign everything that does not seem to be “animal,” “human,” or
“anthropomorphic.” Beyond this there is no unanimous view about the figurative or
abstract nature of many of signs, and thus of the role they play in the compositions —
whether if they accompanied other figures or were entities with their own organiza-
tion.

Temporal Variability of the Signs

Leroi-Gourhan's (1971) stylistic chronology rests, in part, on the temporal distri-
bution of the parietal signs. This has been widely accepted by the majority of conti-
nental researchers and recent chronometric dates for parietal art seem to confirm the
essential correctness of that chronology. At the same time, Leroi-Gourhan (1980)
emphasized that the great signs, at times, tend to have a limited geographic distribu-
tion. This led him to suggest that such signs were “ethnic markers,” linked to territo-
rial groups. This hypothesis is a significant one, which I turn to below.

The first salient point is that the regionalization of the signs does not seem to be
stable in time but shows clear diachronic variability. In the Early Upper Paleolithic
the signs do not show precision or regularity in form. At the same time, certain
mobihiary art “subjects,” such as the so-called “Venuses,” show a wide distribution
throughout all of Europe, from the Pyrénées to Central Europe. Gamble (1982) has in-
terpreted this as a widespread transmission of information regarding these artistic
forms and, possibly, the ritual or symbolic practices associated with them.

As Jochim (1983, 1987) noted, the climatic deterioration, which culminated in the
glacial maximum around 20,000- 18,000 BP, must have caused a concentration of hu-
man groups (or persistence vs. extinction, as pointed by Gamble [1991 ]) in the south-
ern latitudes of Europe. Jochim linked this phenomenon with the development of
Paleolithic parietal art. The increase in Solutrean sites in the Cantabrian region, com-
pared to those dated to the Early Upper Paleolithic, seems to parallel significant in-
crease in the number of caves containing parietal art, as well as a greater complexity
in the painted compositions. In addition, there is also an increase in the use (see
Clottes, this volume) of interior spaces of the caves isolated from the external areas of
daily use. Furthermore, the most complex signs become formalized at this time and
show a clearly regionalized distribution.

Neither of these phenomena can simply be attributed to an increase in the number
of habitation sites, as would occur, for example, if there were an increase in the
number of painted caves. Instead, they may be related to changes in the behavior of
human groups, such as changes in the use of resources and territories due to competi-

PLEISTOCENE ART OF SPAIN 197

tion and negotiation with other human groups. Such complex social negotiations
could stimulate the use of parietal art as a medium of differentiation, giving it an ever
more ritualized character. This ritualization could have been expressed 1) by the re-
moval and separation of the painted and engraved walls from habitation areas, rele-
gating the art to the dark interior of the caves; 2) in the tendency to focus on an
iconography centered around a subject matter inherited from the earlier Upper Paleo-
lithic; and 3) in the numeric and positional importance that signs gradually acquire,
leading to their being represented independently of the animals.

In the Cantabrian region the progressive “concealment” of the parietal sanctuaries
is manifest in compositions ascribed to Style HI— within a general Solutrean chronol-
ogy — and not to the beginning of Style IV, which Leroi-Gourhan (1971) proposes on
the basis of the chronology of the great exterior bas-relief friezes (absent in Canta-
bria). If the above hypothesis is correct, the early occurrence of subterranean sanctu-
aries on the northern Spanish coast, together with the lack of development of later ex-
terior compositions may be related to competition for resources and social
interactions in a more circumscribed topographic and ecological environment.

Later in the Upper Paleolithic, during Leroi-Gourhan's style IV, the Cantabrian
signs seem to yield their central role to the great animal figures, which are again found
near the inhabited mouths of the caves (La Pasiega B, Castillo and Altamira are exam-
ples of differential distribution within a single site). At the same time we see an in-
creasing divergence between mobiliary and parietal subject matter and between
parietal subject matter and hunted fauna. The latter is primarily due to the occurrence
of ibex in some areas. At this time there is also an increase in the number of mobiliary
art works compared to the number of parietal art works. In spite of the discussed dan-
gers of associating art with food, the data do seem to indicate that artistic behavior re-
flects changes in the forms of social organization and the systems of resource
exploitation. The changes are archaeologically reflected in: 1) the increase in impor-
tance of food gathering; 2) the reduction in the mobility of human groups at the end of
the Magdalenian, evident from raw materials used and from the use of other resources
(Gonzalez Sainz & Gonzalez Morales 1986). On acontinental scale they also include
the progressive colonization of new territories spreading from the Pyrénées to north-
ern Europe and Great Britain, and the disappearance of animals which formed the
subsistence basis, such as reindeer, and their replacement with nonmigratory species
whose procurement implied reduced mobility. All these changes suggest either a de-
crease in competition ora decrease in periodic aggregations as a mechanism of large-
scale social organization. The culmination of this process seems to lead to the rapid
disappearance of parietal art and of all the abundant animal imagery represented in
the caves. This is shortly followed by almost total disappearance of figurative mobili-
ary art.

Conclusion

I have discussed elsewhere my hypothesis about the end of Paleolithic art
(Gonzalez Morales 1991). These findings are at the basis of my discussion here. My
current brief review of some issues pertinent in the study of parietal art represents a
series of probes into the problems inherent in the very definition of the content of pa-
rietal representations and their relationship to the spheres of economic behavior and
social organization. I conclude by noting that traditional analyses allow us to extract
only a few secure generalizations and that we need to develop new methodological

198 GONZALEZ MORALES

approaches in order to deal with these questions. These need to be developed, how-
ever, with the cognizance of the limitations inherent in earlier analyses and of the
level of subjectivity inherent in them.

Acknowledgments

I want to express my gratitude to several persons involved in the final making of
this paper. Geoffrey A. Clark (Arizona State University) arranged the translation of
the first manuscript by Elisabeth S. Dinsmore. Marcia-Anne Dobres and Heather
Price (University of California, Berkeley) made many corrections and important
comments to this original version, as well as Meg Conkey. The final version has been
reviewed and corrected again by Olga Soffer. Their efforts have improved in many
ways my first text.

This paper has been part of my research during a sabbatical leave at the Depart-
ment of Anthropology, University of California, Berkeley, in 1995, made possible by
a grant (reference PR94-058) from the DGICYT, an agency of the Spanish Ministry
of Education and Science.

Literature Cited

Altuna, J. 1972. Fauna de mamiteros de los yacimientos prehistoricos de Guipuzcoa. Munibe
24.

Altuna, J. 1983. On the relationship between archaofaunas and parietal art in the caves of the
Cantabrian region. Pages 227-238 in J. Clutton-Brock & C. Grigson, eds., Animals and Ar-
chaeology. 1. Hunters and Their Prey, vol. 1. BAR, Oxford, U.K.

Altuna, J. 1994, La relacion fauna consumida-fauna representada en el Paleolitico Superior
Cantabrico. Complutum 5:303-311.

Bahn, P. G. 1986. No sex, please, we're Aurignacians. Rock Art Res. 3:99-105.

Bourdier, F. 1967. Préhistoire de France. Flammarion, Paris, France.

Clottes, J. 1989. The identification of human and animal figures in European Paleolithic art.
Pages 21-56 in H. Morphy, ed., Animals into Art, One World Archaeology. Unwin Hyman,
London, U.K.

Conkey, M. W. 1983. On the origins of Palaeolithic art: A review and some critical thoughts.
Pages 201-227 in E. Trinkaus, ed., The Mousterian Legacy: Human Biocultural Change in
the Upper Pleistocene, vol. 164, BAR nternational Series. British Archaeological Reports,
Oxford, U.K.

1985. Ritual communication, social elaboration, and the variable trajectories of paleo-
lithic material culture. in T. D. P. Price & J. A. Brown, eds., Prehistoric Hunter Gatherers
The Emergence of Cultural Complexity, Academic Press, Orlando.

__. 1987. Interpretive problems in hunter-gatherer studies. Pages 63-77 i O. Softer, ed.
The Pleistocene Old World. Regional Perspectives. Plenum, New York/London.

Criado Boado, F. & R. Penedo Romero. 1989. Cazadores y salvajes: Una contraposicion entre
el arte Paleolitico y el arte post-glaciar Levantino. Munibe 41:3-22.

Chase, P. G. & H. D. Dibble. 1987. Middle Palaeolithic symbolism: A review of current evi-
dence and Interpretations. J. Anthropol. Arch. 6:263-296,

Davis, W. 1986. The origins of image making. Curr. Anthropol. 27:193-215.

Duff, A. 1.,G. A. Clark & T. J. Chadderdon. 1992. Symbolism in the Early Palaeolithic: A Con-
ceptual Odyssey. Cambr. Arch. J. 2:211-229.

Gamble, C. 1982. Interaction and alliance in Palaeolithic society. Man 17:92-107,

PLEISTOCENE ART OF SPAIN 199

. 1991. The social context for European Palaeolithic Art. PPS 57:3-15.

Gomez-Tabanera, J. M. 1971. Simbolismo y ritual en el arte rupestre paleolitico de la isoida
caucasoide. Zephvrus 2|-22:73-87.

Gonzalez Morales, M. R. 1989. Problemas de ocumentacion del arte rupestre cantabrico: Algu-
nos resultados. A/mansor 7:271-288.

. 1991. From hunter-gatherers to food producers in northern Spain: Smooth adaptive
shifts or revolutionary change in the Mesolithic? Pages 204-216 in G. A. Clark, ed., Per-
spectives on the Past. Theoretical Biases in Mediterranean Hunter-gatherer Research.
University of Pennsylvania Press, Philadelphia.

. 1994. Pero {hubo alguna vez once mil bisontes? Los temas del arte parietal paleolitico
de la Region Cantabrica. Complutum 5:291-301.

Gonzalez Sainz, C. 1989. Algunas reflexiones sobre el hecho artistico al final del Paleolitico
Superior. Pages 229-261 in M. R. Gonzalez Morales, ed. Cien Afios Después de Sautuola.
Diputacion Regional de Cantabria, Santander, Spain.

Gonzalez Sainz, C. & M. R. Gonzalez Morales. 1986. La Prehistoria en Cantabria. Vol. 1. His-
toria General de Cantabria. Santander: Ed. Tantin, Spain.

Jochim, M. 1983. Palaeolithic art in ecological perspective. Pages 212-219 in G. N. Bailey, ed.,
Hunter-gatherer Economy in Prehistory. Cambridge University Press, Cambridge, U.K.

. 1987. Late Pleistocene refugia in Europe. Pages 317-331 in O. Soffer, ed. The Pleisto-
cene Old World. Regional Perspectives. Plenum, New York/London.

Laming-Emperaire, A. 1962. La Signification de I'Art Rupestre Paleolithique. Methodes et
Applications. Picard, Paris, France.

Layton, R. 1977. Naturalism and cultural relativity in art. Pages 33-43 in P. J. Ucko, ed., Form
in Indigenous Art. Australian Institute of Aboriginal Studies, Canberra, Australia.

. 1987. The use of ethnographic paralells in interpreting Uper Palaeolithic rock art. in L.
Holy & R. Fardon, eds., Comparative Encounters. Blackwell, Oxford, U.K.

. 199]. The Anthropology of Art, 2° edition. Cambridge University Press, Cambridge,
U.K.

Leroi-Gourhan, A. 1971. Préhistoire de l'art occidental, 2nd edition. Mazenod, Paris, France.

. 1980. Les signes pariétaux comme marqueurs ethniques. Pages 289-294 in Altamira
Symposium. Ministerio de Cultura, Madrid, Spain.

Lorbranchet, M. 1989. From man to animal and sign in Palaeolithic art. Pages 109-143 in H.
Morphy, ed., Animals into Art, One World Archeology. Unwin Hyman, London, U.K.
Marschack, A. 1972. The Roots of Civilization. The Cognitive Beginnings of Man's First Art,

Symbol and Notation. McGraw-Hill, New York.

Mithen, S. J. 1988. Looking and learning: Upper Palaeolithic art and information gathering.
World Arch. \9:297-327.

. 1990. Thoughtful Foragers . Cambridge University Press, Cambridge, U.K.

. 1991. Ecological interpretations of Palaeolithic Art. PPS 57:103-114.

Rice, P.C. & A. L. Paterson. 1985. Cave art and bones: Exploring the interrelationships. Amer
Anthropol. 87:94-100.

. 1986. Validating the cave art-archeofaunal relationship in Cantabrian Spain. Amer.
Anthropol. 88:658-667.

Straus, L. G. 1977. Of deerslayers and mountain men: Palzolothic faunal exploitation in Cant-
abrian Spain. Pages 41-76 in L. Binford, ed., For Theory Building in Archaeology. Aca-
demic Press, New York.

Ucko, P. J. 1989. La subjetividad y el estudio del arte parietal paleolitico in M. R. Gonzalez
Morales, ed., Cien Avios Después de Sautuola. Diputacion Regional de Cantabria, San-
tander, Spain.

Ucko, P. J. & A. Rosenfeld. 1967. Palaeolithic Cave Art. Weindenfeld, London, U.K.

Section III

Interpreting Regional
Imagery in the European
Upper Paleolithic

203

Art of the Light and Art of the
Depths

Jean Clottes

Conservateur Général du Patrimoine, UMR 9933
11, rue du Fourcat

09000 Foix

France

Paleolithic rock art was first discovered at the end of the nineteenth century and
the beginning of the twentieth in deep caves such as Altamira, Marsoulas, La Mouthe,
Font-de-Gaume, Combarelles, Bernifal, Gargas and Niaux. Because of these first dis-
coveries, Upper Paleolithic rock art was associated with the dark and is still com-
monly called cave art (see L Art des Cavernes 1984). However, a number of caves
with engravings in the entrance, where they are more or less exposed to the light, had
been found in roughly the same period: Chabot in 1889, Pair-non-Pair in 1896, Teyjat
in 1903, Le Figuier in 1906, Oulen in 1907 and Huchard in 1908. In subsequent years,
bas-reliefs were discovered not in caves but in shelters: Cap-Blanc in 1909, Laussel in
1911 and Le Poisson in 1912. Despite these early finds, the image of this art as be-
longing to deep caves has persisted, and it has taken a long time for specialists to begin
considering possible reasons for the choice of such different locations.

Breuil just mentions the question: “Other sites such as Cap Blanc, the Roc de Sers
and Angles are open and inhabited places. We cannot say if these were less sacred
than the others, or if there were not many more. The open sites may have been more
numerous than we think: usually they have perished, while the deep galleries lie pro-
tected” (Breuil 1952:23-4). In these few words, in addition to acknowledging the ex-
istence of art in the open, Breuil introduces three problems, each of major importance:
1) the shelters or cave entrances were often inhabited; 2) different values may have
been attached to the art according to the type of location; and 3) the question of what
we would nowadays call the taphonomy of Paleolithic rock art.

However, it was not until 1962 that these problems were dealt with in depth.
Laming-Emperaire divided Paleolithic rock art into two groups, which she thought
distinct as to location, geography, techniques and themes. According to her, rock art
in shelters or in the entrances to caves consisted largely of sculptures (Figure 1) and
deep engravings, whereas fine engravings and paintings were the techniques most of-
ten represented in deep caves — with the very few exceptions noted in a footnote
(Laming-Emperaire 1962:187, note 1). At the very end of the Paleolithic, some fine
engravings were done in entrances, such as at Teyjat or at Sainte-Eulalie. The “out-

Beyond Art: Pleistocene Image and Symbol Memoirs of the Califomia Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

204 CLOTTES

side” group consisted mainly of sites to the north of southwestern France, while the
“deep cave” group consisted of sites to the South. Despite drawing on the same tradi-
tional sources and showing a deep unity in their inspiration (p. 293), themes differed
between the groups: Unlike the deep caves, a fair number of humans had been de-
picted outside, and the animals there were never accompanied by arrow-like signs.
Geometrical signs were in fact absent, as were composite beings. Laming-Emperaire
hypothesized that in the deep caves, the artists had wanted to portray the relation of
animals to one another and in the entrances, the relation of men to animals. She also
stressed the frequent presence of habitats next to the works of art in shelters and en-
trances (Laming-Emperaire 1962:196-7).

Leroi-Gourhan also devoted much thought to what he considered as two distinct
categories of sanctuaries, those under shelters or in the entrances to caves and those in
the dark (Leroi-Gourhan 1965:114). He wrote that the use of deep caves was one of
the characteristics of his Early Style 1V during the Middle Magdalenian. Then, with
the Late Magdalenian and the Late Style IV, “the great period of deep sanctuaries
comes to an end and wall art can again be found next to the entrances” (Leroi-
Gourhan 1965:42; see also 1965:73, 77, 156, 316, 319). He thus saw differences in
choice of location as being characterized both culturally and chronologically. The
first period, from the Gravettian to Magdalenian III, and the last were ones of art in the
open. As for sculpture, he thought it was for practical and technical reasons, such as
the long time needed to carve large figures out of rock, that they were only produced
outside (Leroi-Gourhan 1965:114). Contrary to Laming-Emperaire, in his opinion
there did not exist much thematic difference between the art in the dark and that in the
light (1962:115),.

Thirty years after the pioneering work of Laming-Emperaire and Leroi-Gourhan,
many more discoveries have been made and the art is far better known with regard to
chronology and archaeological context. As a result, it has become possible to test
their remarks on this important question against examples from all of France.

Ficure |. Woman in bas-relief in the La Magdeleine shelter (Penne, Tarn). Photograph by A. Serres

ART OF THE LIGHT 205

Rock Art in Shelters and in Caves, According to the
Different Areas and Periods

Since L ‘Art des Cavernes was published in 1984, a number of new sites have been
found or newly studied. To that list can be added: North of Périgord: La Grande
Grotte d’Arcy/Cure, Le Réseau Guy Martin, Le Placard; in Périgord: Fronsac, Font-
Bargeix, La Croix, Le Charretou, Vielmouly H, La Cave Pataud, La Cavaille, Le Trou
du Tai, Le Trou de la Marmite, Puy-Martin; in Quercy: Mazet; in the Pyrenees: Cam-
pome, Gourdan, Montconfort, Enlene; in the southeast: Cosquer, Deux-Ouvertures,
Aigueze. The total has now reached 149 sites.

In the following table, these sites have been classified according to the different
regions and periods. Such a table entails the oversimplification of a complex picture
with regard to the distinction between art in the dark and art in the light, as well as geo-
graphical locations and chronological attributions.

When dealing with a cave like Niaux or a rock shelter like Angles-sur-Anglin,
everything is simple. However, there do exist caves which are covered with engrav-
ings right from the lighted entrance to far inside the deepest galleries. In such cases,
which occur three times in Perigord (Comarque, Bernifal, Saint-Cirq), twice in the
Pyrenees (Marsoulas, Gargas) and once (Oulen) in the southeast, the caves have been
listed in both the “Light” and the “Dark” columns. In other caves, most of the art is in
the dark, and so the site has been listed as “dark,” even though there may exist one or
two petroglyphs in an area still reached by dim light (Croze a Gontran, Roc de Veézac,
Cassegros, Bayol, Marcenac).

A possible source of error lies in the changes that entrances to caves may have un-
dergone in the course of time. At present, thick vegetation may grow outside a cave
and obscure the gallery, especially if it happens to be low, while in Paleolithic times
the light may have penetrated much farther in. In other cases, it is simply not known
where the original entrance to the site was, since it became blocked after the cave was
decorated (Gravettian Gallery in Les Trois-Fréres). Even when such a blocked en-
trance can be determined (Fontanet), there is no way to know whether, at the time the
cave was frequented, the opening was large enough to let in the light, in which case
the nearest paintings would have been done in the light; or whether, as we now as-
sume, the scree that eventually closed the entrance entirely had already started and
just a small passage for crawling remained, in which case all the paintings were done
in the dark.

Light Dark I-Il I-I] Hl Hl IV IV Final Final

ee Light Dark Light Dark Light Dark Light Dark
Nord 9 5 0 | 2} | 4 2 | |
Perigord 33 27 12 10 8 10 5 15 2 4
Quercy 8 18 0 4 2 10 4 7 | |
Pyrénées 7 25 | 3 0 | 6 22: 0 |
Sud-Est 8 13 0 | 6 fl l 3 | 0
Total 65 88 13 12 21 27 20 49 5 7

It has been expedient at times to lump together sites which are very far apart. In-
cluded in the “North” group are all the sites north, west and east of Perigord, which
may not have much in common but are too few and scattered to allow for separate
categories. Included in the Quercy group are the two known caves in Limousin. On
the other hand, the caves of the Southeast include the Aude-Hérault sites which show
distinct cultural affinities with the Pyrenees.

206 CLOTTES

The problems are even greater when it comes to chronology, as most caves and
shelters have been assigned to a particular time or culture only through arguments
based on stylistic criteria, mainly in reference to Leroi-Gourhan’s scheme (Styles I-II,
III, IV) to which was added a column for the Final Magdalenian. Such “dating” is
largely hypothetical and the figures in the table are nothing more than a rough esti-
mate. When a cave gives evidence of more than one period in its wall art, it has then
been counted more than once.

The first result comes as a surprise. Art in the light is just slightly less frequent
(42.5 %) than art inside deep caves (57.5 %). This runs counter to the prevailing im-
pression that art in the open is far more rare than cave art because most of it has been
destroyed by natural and other causes, and that what is left constitutes “exceptions,”
however “notable” (Bednarik 1993:4). In this same article about European Ice Age
rock art, Bednarik mentions “its restriction to caves” and goes on to challenge the idea
of “cave art” “because it generally survived only there” (1993:4).

The taphonomy of European Upper Paleolithic rock art is not a new area of inter-
est. It has been considered by researchers working on the subject ever since Breuil
(Breuil 1952:24: Laming-Emperaire 1962:186), although nowadays more attention
is being paid to the question (Bahn & Vertut 1988:111-3: Delporte 1990:56-7).

In France (Campome), in Portugal (Mazouco, Foz Céa) and in Spain (Domingo
Garcia, Carbonero Mayor, Siega Verde) (Ripoll Lopez & Muncio Gonzalez 1994),
recent discoveries of engravings not only in shelters or in caves but also on rocks out
in the open, show that it is possible for Paleolithic petroglyphs to have survived in
fairly exposed locations. Obviously, paintings are a different proposition. They can
only endure in sheltered locations, making them much more vulnerable than rock art
of any other technique and so many were probably destroyed.

On the other hand, caves are certainly not museum-like places where wall art was
always perfectly preserved. Wall surfaces can be ravaged by natural causes such as
flooding (about two-thirds of the Cosquer Cave near Marseille), or by water seeping
through cracks and dissolving pigments (Niaux in 1978-1980), or by calcite covering
huge surfaces, or even by drafts. For example, it will never be known whether the ab-
sence of art in the first 400 meters of Niaux (Clottes 1995) and in the front part of Pin-
dal is due to cultural choice or to natural phenomena. The latter could very well be the
case, at least for Pindal, since it is directly open to sea winds. It should also be kept in
mind that many cave entrances have been blocked by scree or by collapses and so re-
main undiscovered. In the Tarascon-sur-Ariége area, two (Fontanet and Réseau Clas-
tres) out of six known caves have blocked entrances and were explored through an
access that could not have been used by the Magdalenians. In the case of Fontanet, the
prehistoric entrance is concealed by an extensive scree. Similar screes can be seen at
the bases of dozens of cliffs in various valleys around Tarascon-sur-Ariege, perhaps
concealing a surprising number of caves.

To postulate then that rock art in the open must have been the norm, since most of
it has disappeared, and that compared to this worldwide phenomenon, cave art was a
sort of side event, would mean taking the stance that the destruction of outside sites
was far more frequent than that of caves. This is possible but it remains a supposition
until in-depth geological and taphonomical studies are undertaken, assuming they are
feasible.

The finding that rock art sites in the light are nearly as numerous as those in caves
can therefore be interpreted in two different ways. It can be argued that it provides
grist for the mill of those who are persuaded of the prime importance of open-air rock
art. Then it would just be a closer look at the remains of a widespread phenomenon.

ART OF THE LIGHT 207

On the other hand, it might be said to show that petroglyphs can be preserved what-
ever their location, not just bas-reliefs and deeply-etched works, but even very fine
engravings, such as at Campome or at Domingo Garcia.

If the rock art sites known today represent a hopelessly biased sample because of
the destruction of a large number of open-air sites as compared to the relative preser-
vation of caves, one should find roughly the same proportion in each type of location
whatever the period. From the table above it is clear that this is not the case. The num-
bers of sites are about equal (13 and 12) for the earlier art (Leroi-Gourhan’s Styles
I-II, corresponding to the Aurignacian and Gravettian), for Style III (mostly So-
lutrean: 21 and 27) and for the Final Magdalenian (5 and 7), but change to a ratio of
nearly one to three for Style [V (Middle and Upper Magdalenian), with 20 sites in the
light (30%) and 49 in the dark (70%).

The phenomenon is also space-related, since the largest difference occurs in the
Pyrenees (7 sites in the light and 25 in the dark). Because in the Pyrenees the earliest
and latest sites are few, it is impossible to see whether the larger number of deep caves
with rock art is due only to the availability of caves atall periods or toa cultural choice
made during the Middle Magdalenian. However, this is something that can be con-
firmed in the other groups, by removing the Pyrenean sites from the total. In that case,
the results hardly change. The overall relation of light to dark approaches one to one,
with still a slight advantage to the caves (48% and 52%), but more importantly, a
strong difference persists for sites attributed to Leroi-Gourhan’s Style IV: 14 in the
light (34%) and 27 in the dark (66%). This is not that far from the overall total with the
Pyrenees included. The inescapable conclusion is that, just as Leroi-Gourhan had ar-
gued, people during the Middle and perhaps the Final Magdalenian made a cultural
choice to paint or engrave in deep caves more often than ever before or after. That
choice is more clear in the Pyrenees where a larger number of deep caves were avail-
able.

Even during that period, however, the Magdalenians never stopped using rock
shelters and cave entrances for their art. Conversely, dark caves had always been used
in earlier times: the oldest radiocarbon date obtained so far (27,110 BP + 350; GifA
92.491) is from a stenciled hand in the Cosquer Cave, very far back from the entrance
(Clottes et al. 1993). Recently, still older dates, ranging between 30,340 and 32,410
BP, were obtained for three paintings very far inside the Chauvet Cave (Clottes er al.
1995).

Some Aspects of Art in the Light

Bas-reliefs are not very numerous in French wall art: 18 are known at present.
They range through the Aurignacian (Bernous, Blanchard, Castanet), the Gravettian
(Poisson, Laussel, Pair-non-Pair), the Solutrean (Cave Pataud, Fourneau du Diable,
Roc de Sers) and the Magdalenian (La Magdeleine, Angles-sur-l’ Anglin, Cap Blanc)
(Delluc 1989, Roussot, 1989). They are to be found in all the geographical groups, ex-
cept the southeast: Angles-sur-l’ Anglin, Chaire a Calvin, Roc de Sers are to the north
and west of Dordogne; Cap-Blanc, Cave Pataud and others are in Perigord; Pair-non-
Pair is in Gironde, La Magdeleine is in the south of Quercy and Isturitz is in the Pyren-
ees. Beginning with the earliest wall art, then this technique was used in all periods
and in most locations.

Both Leroi-Gourhan and Laming-Emperaire stressed the fact that bas-reliefs were
carved only in shelters and never inside deep caves. According to Leroi-Gourhan

208 CLOTTES

(1965:114), this was probably because of the technical difficulties the artists faced
(1965) and the long hours needed to finisha sculpture. This practical type of explana-
tion may or may not apply to people who must have had a concept of time very differ-
ent from ours. We do know, however, that they had mastered lighting, using torches
and grease lamps, and so could remain inside a deep cave for as long as they chose
(Roussot 1989:68). Moreover, in Comarque, a few bas-reliefs were done in complete
darkness, even if near the entrance, so this was in fact feasible.

In certain cases, sculptures are held to have been done in the light even when this
is far from certain. For example, in Isturitz the sculptures are 35 meters from the en-
trance and are at present in the dark. However, it has been suggested that they still
might have been reached by daylight if the entrance at the time was wide enough
(Roussot 1989:52). What is certain is that bas-reliefs on walls were carved near en-
trances, Whether or not they were in the dark at the time. Practical considerations pro-
vide one kind of explanation but are not the only possibility. There may have been
religious taboos or even long-standing habits (“We do it this way because this is how
it has always been done, and so this is how it must be done”). The clay bas-reliefs are
something else again, since they have been found in only a very few caves (Monte-
span, Bedeilhac, Tuc d’Audoubert) deep underground and restricted to one area (the
Ariege Pyrenees) and one period (the Middle Magdalenian).

Fine engravings in entrances and shelters are no longer as unusual as they were
thirty years ago. Important sites such as Le Placard in the Solutrean or Gourdan in the
Magdalenian had entire walls covered with them (see also La Roque, Roc d’Allas,
Montconfort, Marsoulas, Sainte-Eulalie, Colombier II). The only art on a rock com-
pletely out in the open (Campome) ts engraved with very fine lines. Neither this tech-
nique nor art in the open is restricted to the final phase of the Magdalenian, as
Leroi-Gourhan mistakenly assumed: Sainte-Eulalie, in Leroi-Gourhan’s time be-
lieved to be Magdalenian VI, has since been shown to be Magdalenian III (Lor-
blanchet 1973). Colombier H, a shelter with fine engravings, has now also been
assigned to an earlier date (Middle Magdalenian) than had been thought (Onoratin1,
Combier, Ayrolles 1992).

Artin the open was carried out, then, at various periods using different techniques,
including painting (paintings or very faint traces of paint have been discovered in sev-
eral shelters or cave entrances, such as Laussel or Le Placard). If bas-reliefs are con-
sistently found either in the light or not far inside caves, which somewhat sets them
apart, as Laming-Emperaire had rightly argued, her other remarks about the unique
nature of art in the light compared to cave art from the point of view of themes and
meaning, already criticized by Leroi-Gourhan (1965), have not been borne out by the
findings of the last thirty years. The themes (animals, humans, signs) chosen by the
artists do not vary that much whatever the location of the art or the technique used
(Roussot 1989:65-67), and so it is unlikely that the concepts that informed them
should have been fundamentally different.

Since Breuil and especially since Laming-Emperaire, all authors have mentioned
the fact that rock art sites in the light almost always occur next to habitats. Whenever
properly excavated, archaeological layers in the immediate vicinity of the works of
art are exactly the same as those elsewhere. Whatever the significance of the art, this
means that whole groups lived everyday lives with their usual activities next to the
bas-reliefs and/or engravings and paintings (for example, Le Placard [Figure 2].
Gourdan, Gargas), whereas in deep caves such evidence ts scarce (e.g., Bédeilhac,
Labastide) and often points to short visits by small groups of people selected on some
basis we can only speculate about (see Niaux, Le Tuc d’Audoubert, Fontanet.

ART OF THE LIGHT 209

Cosquer). This indicates a significant difference in the way art in the light and art in
the dark was thought of at the time.

Another significant difference has gone largely unnoticed until now. In deep
caves, the art is almost never defaced, which is why Leroi-Gourhan mentioned “the
respect that people of the Paleolithic bore for the images made by their ancestors”
(Leroi-Gourhan 1965:140). At times, new engravings or paintings have been super-
imposed on earlier ones but without any destruction. The inference is that the images
on the walls retained their power and that the surfaces on which they had been drawn
also remained “sacred.” One might even argue that superimpositions show that the
importance attached to the particular cave walls where they occur did not wane over
long periods of time (Combarelles, Trois-Freéres). In inhabited shelters and entrances,
on the other hand, the images on the walls were eventually covered over by sediment
and accumulated refuse under which they completely disappeared. In Le Placard, for
example, this happened to the main panel of engravings. The radiocarbon dates ob-

Ficure 2. Le Placard (Vilhonneur, Charente) is ahuge shelter. The walls have been engraved and painted
over vast surfaces. The scaffolding is slightly above the lowest engravings, covered with black plastic
sheets. Archaeological layers have accumulated against the walls. They have entirely covered the petro -
glyphs and nearly reached the roof of the cave. Photograph by L. Duport

210 CLOTTES

tained for some breccia covering the engravings and for the archaeological levels per-
taining to the period when the art was done are the same (Clottes et a/. 1991), which
means that the art lost its interest and value not very long after it was created. The ex-
act time interval is impossible to ascertain because our dating methods are not precise
enough: It could be a few years, a few dozen years or even a century or two. However,
it is reminiscent of what the Magdalenians did with engraved plaquettes, which were
mostly broken up, reused, or thrown away among the refuse. It has even been argued
that their being broken up and dispersed was deliberate (Saint-Perier 1936:126, 127),
a strong possibility given that this observation has been reported again and again, at
Le Mas d’Azil, Lortet, Gourdan, Isturitz, Limeuil, Enval, Enlene and others (Clottes
1989). Engraved plaquettes were of temporary interest for their makers/users. It may
well have been the same for art in the open.

Deep Inside the Caves

The preference for deep caves in the Middle Magdalenian is best exemplified by
sites that had been lived in and painted or engraved at periods separated by many mil-
lennia, such as Enlene and Les Trois-Freres in the Ariege. Those two caves are next to
each other on the same hillside and Les Trois-Freres can be reached through Enlene.
The entrance to Enlene was inhabited fora long period during the Gravettian and fora
shorter period in the Middle Magdalenian. Beyond the first forty meters or so, all the
Paleolithic remains discovered in the depths of the cave are Middle Magdalenian
only, right to the end, some 200 meters away from the entrance, where red paint has
been smudged on the walls and where one big red dot and five vertical lines have also
been painted. This means that the uses to which the same cave was put are very differ-
ent. The Gravettians stayed in the entrance while the Magdalenians both lived and
painted far inside, in complete darkness.

In Les Trois-Fréres, some fires were built in the Chapelle de la Lionne and the
Salle du Foyer during the Magdalenian, but there is no evidence for a long occupation
at that time or before. The location of the wall art differs from the Gravettian to the
Magdalenian. Gravettian engravings were all made in one gallery (Galerie des Chou-
ettes), whereas Magdalenian engravings are found everywhere in the cave, even in
far-off side galleries and in out-of-the-way recesses. An entrance now blocked may
have existed to provide the Perigordians with direct access to the Galerie des Chou-
ettes (Breuil 1952; Bégouén & Clottes 1984:401); in that case, the layout of Gravet-
tian and Magdalenian engravings in Les Trois-Fréres followed the same pattern as in
their habitats in Enleéne.

Before the Magdalenian, however, deep caves were occasionally used, such as
Chauvet in the Aurignacian. Gravettians left stenciled hand prints, engravings and
finger tracings throughout fairly extensive caves such as Gargas and Cosquer, which
attests to a different frame of mind than the one glimpsed in the preceding examples.
[he Solutreans painted and engraved deep inside caves as well, for example at Coug-
nac, Pech-Merle, Cosquer and Oulen.

With the Middle Magdalenian, the main difference is that in a number of cases the
creators of the art consistently explored mile-long caves, such as Niaux, Montespan
and Rouffignac. They also crawled along very narrow passages (Massat, Le Cheval at
Arey/Cure), climbed avens (Tuc d’Audoubert) and precipitously narrow ledges
(Trois-Fréres, Etcheberri-ko-karbia), and even went down shafts several meters deep
(Fontanet). These speleological feats (Leroi-Gourhan 1992:367) were accomplished

ART OF THE LIGHT 211

again and again during this one period in southwestern Europe. Even if the Paleolithic
artists of Western Europe were not the only humans to have dared to go far into the
depths, as Leroi-Gourhan had supposed (1992:379) (see the Paleolithic caves in Aus-
tralia, such as Koonalda), they are unusual in that they did it habitually, whatever the
difficulties, and in this they are unique in world history until much more recent times.

It has often been noted that young children sometimes accompanied adults on
such expeditions. Whenever Paleolithic footprints have been found in deep caves,
they include children’s footprints. Several authors, such as Bégouen, Breuil, and
lately Pfeiffer, have argued that this was for initiation ceremonies. This idea is contra-
dicted by the fact that some of these children were far too young: The footprints of a
three-year old are preserved in Le Tuc d’Audoubert and the hand prints of a five or
six-year old in Fontanet (Figure 3). The only conclusion to be drawn ts that children
were not barred from the deep sanctuaries, even when very few people seem to have
made their way into them (Breuil 1952:23; Leroi-Gourhan 1965:123).

From their way of using the walls and from the choices they made far inside the
caves, we can get some idea of the artists’ considerations. For example, even if they
did not use all the possibilities in a very extensive cave (Barriere 1982:187, about
Rouffignac), most often they looked carefully for suitable panels with surfaces vast
and smooth enough on which to draw and without too many cracks or too much cal-
cite coating. In Niaux, at the entrance to the Salon Noir, red signs were painted at both
ends. A similar structure made up of two panels of signs, only more complex with
more numerous and varied symbols, exists at the entrance to the gallery leading to the
Salon Noir (Clottes 1988-9, 1995). The panel on the right is as near as possible to the
beginning of the gallery while the panel on the opposite side is a few dozen meters
away from the entrance because closer to it no other surface was suitable. In this par-
ticular case, even though the Magdalenians wanted to create a sort of symbolic frame
to the sanctuary, their choice ofa suitable place came before their wish for symmetry.
Lorblanchet (1981) cites another example. When he wanted to duplicate the Pech-
Merle Frise Noire in another cave in order to study its composition and the methods
used, he had a difficult time finding a suitable cave, despite being in an area (Quercy)
where caves are numerous and their locations well-known. Did the Pech-Merle artists
look as hard for a proper cave in which to work, or did they find it by chance and then
turn it into a major sanctuary?

The question is made even more relevant by the fact that Paleolithic artists so often
made use of natural outlines and reliefs in the caves. That they systematically did this
runs counter to the structuralist theories of Laming-Emperaire (1962) and Leroi-
Gourhan (1965), who thought that the artists had brought an existing schema to the
caves and had drawn animals and signs accordingly. Closely and constantly search-
ing for natural outlines stems from a very different outlook, since then the cave itself
imposes the representation of a particular animal, such as a bird at Altxerri or an up-
right bison and the head of a cervid at Niaux.

The lighting techniques used at the time — torches or grease lamps that would casta
dim and fluctuating light — help to explain this practice. When such conditions are
replicated, or even when visiting a cave with a candle, the walls seem to come alive
with the moving shadows cast by the flickering flame. It becomes very easy then to
see animals in the shapes of the rocks. As with some modern hunter-gatherers, per-
haps shapes seen in this way were taken for real and the artists then drew animals ex-
actly where they happened to see them, possibly bringing them to life or gaining their
good will (Clottes 1995).

to
to

CLOTTES

es

Ficure 3. Hand print of a six-year old preserved in the cave of Fontanet (Ornolac-Ussat-les-Bains,
Ariege). Photograph by R. Simonnet.

This may in part explain the apparently random scattering of figures that can be
seen in some of the largest caves. In extensive caves such as Niaux, Réseau Clastres,
Rouffignac or Tuc d’Audoubert, the artists had a nearly unlimited set of options.
However, their choices seem to follow two different trains of thought. Certain loca-
tions were selected as sanctuaries, such as the Salon Noir in Niaux, the Sanctuaire and
the Chapelle de la Lionne in Trois-Fréres, and the Grand Plafond in Rouffignac. In
these locations numerous animals and signs were drawn, sometimes superimposed
upon one another. The places chosen might be gallery endings (Niaux, Trois-Freres),
large impressive chambers (Réseau Clastres, Lascaux) or small side chambers
(Chapelle de la Lionne). Sometimes they were chosen because of their particular to-
pography: In the Trois-Fréres Sanctuaire, in addition to suitable wall surfaces, there
existed the possibility of crawling behind the main wall and up a sloping shaft to reach
anarrow ledge several meters above ground level in order to draw the impressive Sor-
cerer there. In Rouffignac, the Grand Plafond, which is both the main sanctuary and
the center of the whole subterranean ensemble (Barriére 1982:194), was no doubt
reached in such a way (Barriere 1982:187).

Scatterings of figures are no less enlightening about the artists’ motives and ac-
tions. In the deeper galleries of Niaux, for example, the paintings are dispersed all
along and were all drawn much more quickly and with far less detail than in the Salon
Noir (Clottes 1988-89, 1995). From this it can be inferred that few people went to the
ends of the cave and then did not stay long, unlike what happened in the main sanctu-
ary where they seem to have gone again and again, taking their time, painting a collec-
tion of large detailed animals. In the faraway galleries they did not leave figures just

ART OF THE LIGHT 218

vs)

anywhere. Locations were chosen in relation to topographical peculiarities, such as
jutting rocks, natural outlines, clefts in the walls, or else an unavoidable narrow
crawling passage (Lac Terminal) around which they drew an accumulation of signs
and an ibex, with a horse nearby.

It can be argued that these figures were not meant to be seen by many people, that
what counted was the action of creating them, that their value lay in themselves and
not in relation to a possible audience. This seems all the more likely when the loca-
tions chosen for paintings and engravings are out of the way, difficult of access, or
when only one or two people can see them at a time because of the narrowness of the
recess in which they happen to be, as in the Camarin of Le Portel or, in the same cave,
in the Jeannel Gallery (Figure 4). In Niaux, a solitary mysterious figure, perhaps hu-
man but always ambiguous (Figure 5), the only one of its kind in the whole cave, was
painted in the most remote part of the Salon Noir which can only be reached after first
going on all fours into the Cul-de-Four and then crawling through a narrow passage.
Whether or not this is a “holy of holies” can be debated (Clottes 1995). What is not de-
batable is that the thinking behind this figure is unlike that behind the spectacular,
easy-access animals that crowd the walls of the Salon Noir. For the Trois-Freres Sor
cerer the thinking is reversed: He was painted and engraved in the most conspicuous
spot possible while the animals engraved below are often in places where only one
person can squeeze in at a time.

Ficure 4. Several black horses were painted right at the end ofa very narrow gallery in Le Portel (Loubens,
Ariége). Photograph by J. Clottes

Conclusion

Laming-Emperaire and Leroi-Gourhan wrote their major books in the early six-
ties, when they dealt with the question of the various locations of the art and expressed
their theories about it. Since 1960, 51 new caves or shelters containing rock art have
been discovered, representing more than one third of the total now known (149). Sev-

214 CLOTIES

eral more have been discovered since this article was written (1993), including the
now famous Chauvet Cave; they are not included in the statistics. The area with the
most discoveries 1s the Southeast (8/20, i.e. 40%) and the one with the fewest new
caves is Quercy (7/26, i.e. 27%). The others are not far from the average: North, 4/14
or 28.5% ; Perigord, 22/59 or 37%; Pyrenees, 10/30 or 33%. This great increase in the
available evidence was bound to change at least some aspects of the overall picture
that Laming-Emperaire and Leroi-Gourhan had drawn.

Contrary to what they thought, art in the light cannot be characterized either cul-
turally, geographically or chronologically. It is to be found in all regional groups and
at all periods, with greater frequency than was previously believed. The same ts true
for art in the dark, with one proviso: As Leroi-Gourhan had rightly stressed, with the
Middle Magdalenian, deep and/or difficult caves were much more often frequented
and decorated than ever before. However, the same can be said of the Final Mag-
dalenian as well.

Art in the light and art in the dark do represent two divergent trains of thought,
even if the themes used are not as unlike as Laming-Emperaire had assumed. One of
the major differences is the regular occurrence of often important living sites next to
the art in shelters or entrances and what can be inferred from that fact. Inside the deep
caves, too, different ideas were at work in the choice of spectacular (Figure 6) or of se-
cluded locations (Figure 5) and in the grouping or scattering of figures.

A better understanding of these questions in order to determine why, when and
how the choices were made demands in-depth archaeological studies whenever pos-
sible, both in deep caves and in shelters. This has been accomplished with surprising
results in the Réseau Clastres (Clottes & Simonnet 1990). In France, such studies are
currently going on in the Volp Caves (Enléne, Trois-Fréres, Tuc d’Audoubert),
Bédeilhac, Le Portel and Montespan in the Pyrenees; in Les Deux-Ouvertures and Le
Colombier I in the Rhone Valley (southeast area); in the Grande Grotte at Arcy-sur-
Cure in the northeast; in Pergouset, Pech-Merle and Cougnac in Quercy and in Las-

Ficure 5. This mysterious black figure, painted in the innermost recess of the Salon Noir in Niaux
(Ariege), might represent a human. Photograph by J. Clottes

ART OF THE LIGHT 21

nN

FIGURE 6. The famous clay bisons in Le Tuc d’ Audoubert (Montesquieu-A vantes, Ari¢ge) were placed in a

conspicuous location at the end of the cave. Photograph by H. Beéranger (collection Begouen)

caux in Périgord. As a result, the next few years will probably bring as many changes
to our outlook as have occurred since the sixties.

Literature Cited

L’art des Cavernes \984, Atlas des Grottes ornées paléolithiques frangaises. Ministere de la
Culture — Imprimerie nationale, Paris, France.

Bahn, P.-G. & J. Vertut. 1988. /mages of the Ice Age. Smith & Son Ltd., Windward, London,
U.K

Barriere, C. 1982. L'Art Pariétal de Rouffignac. Picard, Fondation Singer-Polignac, Paris,
France.

Bednarik, R.-G. 1993. European Paleolithic Art — typical or exceptional? Oxford J. Arch
12:1-8.

Bégoéun, R. & J. Clottes. 1984. Grotte des Trois-Freres. Pages 400-409 in L ‘Art des Cavernes,
Ministére de la Culture — Imprimerie nationale, Paris, France.

Breuil, H. 1952. Quatre Cents Siécles d'Art Pariétal. Centre d’ Etudes et de Documentation
prehistoriques, Montignac, France.

Clottes, J. 1988-1989. Niaux: Magdalénien moyen ou récent? Ars Praehist. 7-8:53-67

___. 1989. Le Magdalénien des Pyrénées in Le Magdaleénien en Europe, Actes du Colloque
de Mayence 1987, Liege E.R.A.U.L. 38:28 1-360.

1995. Les Cavernes de Niaux. Art Préhistorique en Ariége. Seuil, Paris, France

216 CLOTTES

Clotttes, J., J-M.Chauvet, E. Brunel-Deschamps, C. Hillaire, J-P. Daugas, M. Arnold, H. Cach-
ier, J. Evin, P. Fortin, C. Oberlin, N. Tisnerat & H. Valladas 1995. Les peintures paléo-
lithiques de la Grotte Chauvet-Pont-d’Arc, a Vallon-Pont-d’Are (Ardéche, France):
Datations directes et indirectes par la méthode du radiocarbone. Comptes-rendus Acad. des
Sci. de Paris, 320:1133-1140.

Clottes, J., J. Courtin & H. Valladas. 1992. A well-dated paleolithic cave: The Cosquer Cave at
Marseille. Rock Art Res. 9:122-129.

Clottes, J., L. Duport & V. Feruglio. 1991. Derniers éléments sur les signes du Placard. Bull.
Soc. Prehist. Ariege-Pyrénées 46:119-132.

Clottes, J. & R. Simonnet. 1990. Retour au Réseau Clastres (Niaux, Ariége). Bull. Soc. Préhist
Ariege-Pyrénées 45:51-139.

Delluc, B.& G. Delluc. 1989. La sculpture rupestre avant Lascaux. Pages 29-43 in La Sculp-
ture Rupestre en France (de la Prehistoire a nos Jours) Actes du Colloque de Brant6me
1988, Péerigueux, France.

Delporte, H. 1990. L'/mage des Animaux dans |’Art Préhistorique. Picard, Paris, France.

Laming-Emperaire, A. 1962. La Signification de |'Art Rupestre Paléolithique. Picard, Paris,
France.

Leroi-Gourhan, A. 1965. Prehistoire de l’Art Occidental. Mazenod, Paris, France.

Leroi-Gourhan, A. 1992. L ‘Art Pariétal, Langage de la Préhistoire. Jerome Millon, Paris,
France,

Lorblanchet, M. 1973. La grotte de Sainte-Eulalie a Espagnac (Lot). Gallia-Préhistoire
16:3-62 and 2:233-325.

ed 1981. Les dessins noirs du Pech-Merle. Congres Préhistorique de France XX1,
Montauban-Cahors 1979, 1:178-207.

Onoratini, G., J. Combier & P. Ayroles. 1992. Datation 4c Pune gravure pariétale de bou-
quetin de Pabri magdalénien du Colombier (Vallon-Pont-d’Are, Ardeche). Comptes-
rendus de l’Acad. des Sci. de Paris 314:405-410.

Roussot, A. 1989. La sculpture rupestre magdalénienne en Aquitaine Pages 45-72 in La Sculp-
ture Rupestre en France (de la Préhistoire a nos Jours), Actes du Colloque de Brantome
1988, Perigueux, France.

Saint-Perier, R. de 1936. La Grotte d'Isturitz. Il, Le Magdalénien de la Grande Salle. Masson,
Archives de l'Institut de Paléontologie Humaine, Memoire 17, Paris, France.

Carving, Painting, Engraving:
Problems with the Earliest Italian
Design

Margherita Muss1

Dipartimento di Scienze dell’ Antichita
Universita di Roma “La Sapienza”
Rome, Italy

Daniela Zampetti

Dipartimento di Scienze dell’ AntichitB
UniversitB di Roma “La Sapienza”
Rome, Italy

There is a long archaeological record in Italy, starting with the Lower and continu-
ing through the Middle and Upper Paleolithic. Definite evidence for aesthetic sensi-
bility and symbolic capacity, however, as in the rest of Europe, dates only from the
early Upper Paleolithic onwards. Since this record is discontinuous, we briefly de-
scribe its characteristics during the different time periods, and relate them to the vary-
ing cultural and demographic contexts. In our conclusions, we discuss traditional
interpretations and put forward alternative explanations.

The Artistic and Cultural Records

The Early Upper Paleolithic

The beginning of the Upper Paleolithic is characterized by archaeological assem-
blages dating between ca. 35,000 and 30,000 BP (uncalibrated years), which are clas-
sified as the Uluzzian and the Aurignacian. The Uluzzian, a local development often
compared to the French Castelperronian, is limited to a few sites found in Southern
and Central Italy (Gioia 1990). The Italian Aurignacian closely resembles the Aurig-
nacian found in the rest of Europe, and is known from some 40 sites located over the
entire peninsula as well as in Sicily (Gioia 1984-1987; Mussi 1992).

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copynght ©1997

218 MUSSI AND ZAMPETTI

Pigments have been recovered at the Uluzzian site of Gr. del Cavallo, but there is
no evidence for how they were used (Palma di Cesnola 1965; Palma di Cesnola 1966).
A few notched bones and stones were found at three Aurignacian sites: Riparo Bom-
brini, Riparo di Fumane, and Riparo di Fontana Nuova (Bartolomei etal. 1991-1993;
Gioia 1984-1987; Vicino 1984). Perforated teeth and shells, as well as steatite pen-
dants, are known from a few of Uluzzian and Aurignacian sites (Mussi 1991). Items
which can be interpreted as figurative art are completely lacking from these sites.

The Gravettian

In the subsequent millennia, after 30,000 BP and before 26,000 to 25,000 BP,
there was either no human settlement in Italy ora very limited one. Then, with the ar-
rival of Gravettian groups, possibly coming from Western Europe, the peninsula and,
later on, Sicily and other islands, were settled once again (Mussi 1990).

Fewer than thirty sites from the peninsula date to this period, 7.e., from ca. 25,000
to 18,000 BP. At these sites the lithic industries belong to both the Gravettian proper
and to the subsequent Early Epigravettian, which is seen as a local development.
Some of these sites are large and multilayered, while others contain only very limited
archaeological remains. One out of four also contain works of art and other aesthetic
remains (Figure 1). As shown by many burial inventories, ornaments found are quite
elaborate (Mussi et a/. 1989: Mussi 1990). Numerous pendants have also been recov-
ered at residential sites, both in caves and in the open air. These were made of marine
shells, canines of red deer, steatite and ivory.

As is usually the case in Italy, bone tools are rare, and ivory items nearly nonexist-
ent. This is not surprising given that the fauna was generally temperate, with mam-
moths and reindeer very sparse in numbers, if extant at all, during the Late
Pleistocene. The shaft straighteners or batons a trous found in the lower, or “Prince,”
burial at Grotta (Gr.) delle Arene Candide are truly unique. They were made of elk
antler and three of the four similar implements deposited next to the shoulder and
waist of the deceased adolescent male were decorated with radial and transverse inci-
sions (Cardini 1942; Graziosi 1956). A radiocarbon date of 18,360 + 210 BP (R-745),
obtained fora level a couple of meters higher in the stratigraphic sequence, provides a
minimum age for the burial and related items.

The chronology of the fifteen female figurines recovered so far is poorly defined,
as none were recovered during regular excavations. They mostly come from the Gri-
maldi or Balzi Rossi sites, namely, from Barma Grande and, most probably, Gr. du
Prince (Mussi 1991; Mussi 1995). Seven of them have long been known in the litera-
ture, while eight have recently been located in old collections stored in North America
(Bisson & Bolduc 1994; Bolduc er al. 1996; Marshack 1986; Mussi, unpublished
data). They include a few small carved heads, including the one nicknamed Tete ne-
groide, which have been traditionally interpreted as female, but which also could be
depicting males. The raw materials used to fashion them were steatite and other soft
stones, as well as antler and ivory.

Iwo more bone figurines were found in Southern Italy, at Gr. delle Veneri, by
amateur archaeologists (Radmilli 1966), while two soft stone figurines were col-
lected as surface finds at Savignano and near the Trasimeno Lake (Mussi 1995; Zam-
petti 1993, 1996).

Stylistically, the Italian figurines are very close to those found in the rest of
Europe. When they depict the whole body, heads, hands and feet are poorly character-
ized, while the breasts, abdomen, and buttocks are accented. Some figurines are more

PALEOLITHIC IMAGES FROM ITALY 219

K

BAYS Re
i™N
sae

ve

Se
f
eee
re)

fv
a aps

Figure |. Sites with Gravettian and Early Epigravettian art, ca.25,000-18,000 BP. Legend; gray areas are
above 500 m asl. |) Gr. di Grimaldi (Gr, dei Balzi Rossi); Gr. delle Arene Candide; 3) Savignano; 4)Tra -
simeno; 5) Gr. Paglicci; 6) Gr. delle Vener.

realistic, while others more schematized. Both realistic and schematic traits are usu-
ally found on the same pieces.
Little other artistic material dates to this period. An ibex engraved on a bone was

360 (F-48) and 22,1 10+ 330 BP (F-49). The associated lithic industry 1s an “evolved”
Gravettian with truncated backed tools (Mezzena & Palma di Cesnola 1972). The
ibex is quite rigid, with only a few details of the head and tail, and is covered by an in-
cised chevron motif. Elsewhere we compared this piece to a similarly engraved ibex
from the Gravettian assemblage of Gr. De Gargas, in Southern France (Zampetti
1987).

Grotta Paglicci has also yielded the only example of parietal art in Italy. The paint-
ings were found in the rear of the cave, and include a minimum of five positive and
negative hand prints, with the thumbs lacking in two or more cases (Zori 1962). A

220 MUSSI AND ZAMPETTI

minimum of 17 caves with painted hand prints are known from Western Europe, con-
centrated in the Franco-Cantabrian region (Verbrugge 1969). A few more are located
in the Lower Rhone Valley (Figure 2). The recently discovered Grotte Cosquer, near
Marseille, (Clottes etal. 1992a; Clottes et a/. 1992b) and Grotte Chauvet, in the Ardeé-
che, help associate the solitary Gr. Paglicci examples with a much larger group in the
West. The associated industries recovered from these caves, and their chronology,
range from the Gravettian to the middle Magdalenian, with most being in the earlier
part of this sequence. A hand stencil at Gr. Cosquer has been directly dated to 27,110
£ 390 (Gif-A 92.409) and 27,110 + 350 BP (Gif-A 92.491) (Clottes er al. 1992b).
Dates in excess of 30,000 BP have been reported for the art at Gr. Chauvet (see
Clottes, this volume).

Grotta Paglicci also contained three red horses, one ina vertical position, in close
association with the hand prints. The animals are rigid, with a sinuous cervicodorsal
line, bulging bellies, thin muzzles, and legs finishing without proper hoofs. Overall,
they are similar to some French examples, such as the punctuated horses of Pech-
Merle, currently attributed to early Style III, in Leroi-Gourhan's classification (Gra-
ziosi 1973). In France, Style HI ranges in age from 17,000 to 13,000 BP (Leroi-
Gourhan 1986). Interestingly, the horses at Pech-Merle are also associated with nega-
tive hand prints. The three horses from Gr. Paglicci also have traits in common with
the painted horses at Cosquer, radiocarbon dated to 18,820 + 310 BP (Gif-A 92.417)
and 18,840 + 240 BP (Gif-A 92.416). These include bulging bellies, left devoid of
color in horse 3 at Cosquer just as at Gr. Paglicci, and the rigid legs which do not end
in hooves.

The vertical position of one of the Gr. Paglicci horses is uncommon for Franco-
Cantabrian art. A minimum of 11 vertical horses, however, are known from the region
(Ricol 1973). These date no earlier than the final Solutrean.

A clue to the chronology of the Gr. Paglicci horses can be obtained from the ex fo-
liated painted wall slab found at the entrance of the cave in level 14a, radiocarbon
dated to 15,750 + 160 BP (GrN-14323). This is clearly a minimum age for the paint-
ings. The slab depicts only hindquarters painted in red. The depiction is dynamic, and
the hoof is detailed. This figure is quite similar to the so-called “Chinese horses” of
Lascaux. The radiocarbon dates from Lascaux, ranging from 17,000 to 15,000 BP
(Leroi-Gourhan & Allain 1979), correspond well with the date for the painted slab.
The slab, thus, possibly dates later than the horses of the rear of the cave. Since parie-
tal art is so rare in Italy, however, we cannot be certain that animals depicted in the
cave, even though stylistically different, necessarily date to different time periods.

Finally, G. Vicino first noticed an engraved little equid at Gr. del Caviglione. one
of the Balzi Rossi or Grimaldi Caves, on the Italian side of the frontier with France,
and assigned it to Style II, 7.e., to the Gravettian (Vicino and Simone 1976). Since,
however, the engraving was found some 6 to 7 m above the Gravettian burial in this
cave, we suggest that the engraving is associated with a higher level of the living
floor, and accordingly with a later deposit (Zampetti 1987). Late Upper Paleolithic
occupations are documented at several of the Balzi Rossi caves (Muss 1992).

The Early Epigravettian and the Evolved Epigravettian

There is a marked increase in the number of known archaeological sites in Italy af-
ter the Last Glacial Maximum (Mussi 1992). These sites are associated with late Early
Epigravettian and Evolved Epigravettian industries, dating from ca. 18,000 to 14,500
BP, which parallel the final Solutrean and the first phases of the Magdalenian else-

221

PALEOLITHIC IMAGES FROM ITALY

‘ooya

“Bd (L| ‘sanbso_ (9] ‘JoAeg (¢| ‘auuoney-auneg eT (p] ‘oyonay Jap aquang (¢] SoypHse fq (Z| ‘ew (| | ‘Utes Jap aiuany (OT ‘Teputg [J (6 SoyJUsng omy, (g ‘Jauog

aT(L:

ueIIqly (9 ‘seaiey (

mie,

g:a

Hew

yoad (pf ‘Nopeurwsoy (¢ taumNey-ap-juo J (7 {peyruag (| ‘syuud puey aayesou Jo/pue aatisod yWM sayts Jo UoNNGISIp ay, °Z aUNDIA

222 MUSSI AND ZAMPETTI

where tn Western Europe. These sites contain a dearth of aesthetic remains: Orna-
ments have rarely been found, and works of art are limited to a few engravings on
bone and stone recovered from Gr. Paglicci. The later came from levels 9 and 8,
which radiocarbon date to 15,270 + 220 (F-67) and 15,460 + 220 BP (F-66).

The engravings depict horses, aurochs, red deer and birds including a possible
Alca impennis, the now extinct Great Auk (Mezzena & Palma di Cesnola 1972,
1992). This species, also depicted at Cosquer, survived to the very end of the Pleisto-
cene in the Western Mediterranean (for a different identification of the specimen
from Paglicci see D’Errico [1994]). Some of these engravings are quite complex. For
example, there is a grouping of horses where several of the quadrupeds are hit by ar-
rows or other projectiles. This scenic arrangement is also found in the French caves
and dated to the Magdalenian (see the engraved pebble of Grotte de la Colombiere or
the stone slab of Abri de Montastruc [Movius & Judson 1956; Sieveking 1987]). In
another instance, a bone shaft is decorated with a nesting bird, eggs, and a probable
snake. This bird, as others at Gr. Paglicci, is quite similar to the one recovered from
the Magdalenian VI level at Caune de Belvis in southern France (illustrated in Sacchi
1972-1973).

The Final Epigravettian

After 14,500 BP, during the so-called Final Epigravettian, which is contempora-
neous with the later Magdalenian, the Azilian, and other Epipaleolithic industries,
there is a sharp increase both in the number of archaeological sites, and in the aes-
thetic remains found in them. Over 70 different sized sites are documented in Italy,
with art found at nearly 30 of these (Figure 3 and 8). We subdivide this inventory into
naturalistic art, schematic art, and geometric art, and treat each category In sequence.
We do acknowledge, however, that the first two categories do sometimes grade into
each other.

Naturalistic art is found in 16 sites. It predominantly, but not exclusively, con-
sists of portable art in peninsular Italy and rock art in Sicily. The art is represented pri-
marily by engravings. Levanzo (Gr. dei Cervi), on the other hand, contains a human
being painted in red. With the exception of Levanzo, this art, found at Gr. del Cavi-
glione, Riparo (Rip.) Tagliente, Gr. Polesini, Gr. Giovanna, efc., consists of zoomor-
phic depictions which are often stylistically similar to Magdalenian depictions. The
figures are lively, the body proportions generally accurate, the hair is sometimes de-
picted, the head is detailed, with a conventional stylization of muzzles and horns simi-
lar to those found in Franco-Cantabria. Depictions of humans are less frequent, and
there is often a contrast between their realistic postures and their masked or animal
heads (Figure 4).

At Rip. Tagliente, in Venetia, a dozen of animals engraved in bone, on pebbles,
and slabs or blocks have been recovered (Guerreschi & Leonardi 1984; Zampetti
1987). Several were found out of context, and are generically assigned to the Final
Epigravettian. The animals depicted include bisons, lions, a single ibex, an aurochs,
and several undetermined herbivores. Some are fragmentary or incomplete, while
others are represented accurately. A beautiful lion or lioness, with detailing of the
muzzle, mouth, eye, ear, and hair decorated one of the blocks found covering the male
burial, radiocarbon dated to about 12,000 years ago (Bartolomei et al. 1974). We con-
sider this depiction similar to that found at La Vache in France ( Buisson and Delporte
1988: Figure 3). It should be noted that skeletal remains of lions are exceedingly rare
at Italian archaeological sites dating to the end of the Pleistocene.

PALEOLITHIC IMAGES FROM ITALY 223

| |

FiGuRE 3. The location of sites with zoomorphic engravings. Legend: Circles: naturalistic art; squares:
schematic art. Grey areas are above 500m asl. 1) Gr. di Grimaldi (Gr. dei Balzi Rossi); 2) Riparo Taghente;
3) Lustignano; 4) Vado all'Arancio; 5) Gr. Polesini; 6) Riparo Salvini; 7) Gr. Paglicci; 8) Gr. delle Mura; 9)
Gr. del Romito; 10) Gr. del Cavallo; 11) Gr. Romanelli; 12) Gr. dei Cervi (Levanzo); 13) Gr. Racchio; 14)
Gr. dei Puntali; 15) Gr. della Za Minica; 16) Gr. dell'Addaura; 17) Gr. Niscemi; and 18) Gr. Giovanna.

The very rich deposits at Grotta Polesini, near Rome, yielded halfa million lithic
implements, and 2 tons of faunal remains (Radmilli 1974). The site’s stratigraphy,
disturbed by the high water table of the nearby Aniene River, is poorly understood.
There is a single C' date for the site of 10,090 + 80 BP (R-1265). Grotta Polesini has
produced bones, stone slabs, and pebbles engraved with naturalistic and geometric
patterns. The most frequent species depicted include horses and aurochs (Figure 5), as
well as deer and leporids. The depictions are similar to those found elsewhere in asso-
ciation with Magdalenian industries (Mussi 1990-1991; Zampetti 1987; Zampetti
1990-1991).

An engraving of a wolf ona limestone pebble from this site was heretofore inter-
preted as an example of hunting magic: The little holes on the surface of the stone
were said to be the result of projectile points thrown at the depiction to magically kill
the animal. Such holes or pits are found on many other pebbles, both decorated and

224 MUSSI AND ZAMPETTI

be
Figure 4. The location of sites with anthropomorphic engravings. Legend: Grey areas are above 500 mas!

1) Vado all'Arancio; 2) Tolentino; 3) Gr. dei Cervi (Levanzo); 4) Gr. dell'Addaura; 5) Gr. Romanelli; and 6)
Gr. del Cavallo

plain. We suggest that the holes, which are not artificial in origin, were, most likely,
products of carbonate solution in water-logged sediments (Mussi & Zampetti 1993).

Grotta Giovanna, one of the earliest Sicilian sites, radiocarbon dating to 12,840 +
100 BP (R-484), has yielded a stone slab engraved with a bovid, most probably an
aurochs. The head of the animal is missing, the depiction is a dynamic one, with over-
developed forequarters. The piece was found near the bottom of the archaeological
deposit (Cardini 1971).

Grotta dei Cervi, containing 33 human and animal figures engraved on the rear
walls of the cave, is located on the islet of Levanzo which, during low beach levels,
was connected to nearby Sicily (Graziosi 1956: Tav. 286-294; Graziosi 1962). The
engraved animals, consisting of aurochs, wild asses, and deer, are depicted in an ele-
gant naturalistic style. Human representations have dynamic postures, but their heads
have odd mushroom shapes or have beaks. A single human figure, also with a mush-
room head, is painted in red. The topographic location of the human and animal fig-
ures here are in perfect accord with the spatial organization ofa “sanctuary” proposed
by Leroi-Gourhan (Leroi-Gourhan 1972; Zampetti 1987).

PALEOLITHIC IMAGES FROM ITALY

i)
i)
wn

a SS "
er
} \ - ‘
a (e) 3
ee)
f X Ny
é YX
f Se
i SS
| S
\ a AT,
ns ,
iS eo Ses
\ Ee (1, \
y J. \ 8 /

FiGure 5. Examples of naturalistic art. Engraved pebbles from Gr.Polesini (a, after Zampetti 1990-1991; b,
after Mussi 1990-1991)

At Gr. dell’Addaura, in Palermo, a unique scene is engraved on a rock wall. Sev-
eral men and one or more women have pointed and almost beak-like heads, as does
one individual at Gr. dei Cervi. The males also have voluminous hair or are wearing
hoods. They are moving in different directions, and are involved in activities difficult
to decipher: Some have raised arms; others carry poles or other objects. Two of the
males are in odd horizontal positions (Graziosi 1956: Tav. 295-297; Graziosi 1973:
Tav. 58-68). Two large horned deer, and several probable wild asses are depicted
alongside the humans. In one instance, the human figure is superimposed on the deer.

Representations of humans are found outside Sicily as well (Figure 4). A notable
example comes from Vado all’ Arancio in Tuscany (Minellono 1985-1986). The sin-
gle archaeological level here is radiocarbon dated to 11,330 + 50 BP (R-1333). The
bearded male engraved on a stone slab here has a realistic profile. We suggest that this
depiction is quite similar to the human profiles engraved on the many plaquettes at the
French Magdalenian sites of La Marche and Angles-sur-I’ Anglin (Zampetti 1987).
The cap or hairdo on the Vado all’Arancio male recalls those found at Gr. dell’ Ad-
daura. A female engraved ona bone recovered at Vado all’ Arancio is much more styl-
ized. Her unusual posture, with open arms and distinct legs, recalls depictions at both

226 MUSSI AND ZAMPETTI

Gr. dell’ Addaura and Levanzo. Similar depictions are also found outside Italy, for ex-
ample, at Mas d'Azil and at La Marche (Capitan et a/. 1928: Figure 97; Pales 1976:
Figure 60),

The engraved pebble recovered at Tolentino, which so far has been poorly de-
scribed in the literature (Rellini 1912; Lollini and Silvestrini n.d.), depicts a female
with an animal head. This piece is similar to specimens recovered from some French
Magdalenian levels, such as the pebble from La Madeleine (Capitan and Peyrony
1928).

Schematic depictions of animals and of humans are found engraved at eight sites,
most of which are located in Apulia and Sicily (Figures 3, 4 and 6). Their shapes are
poorly detailed, their postures are rigid, and, as can be seen at Gr. Romanelli, several
are characterized by a hatched filling (Figure 7a). At Gr. Romanelli, engraved rocks
were found in level C while others were recovered out of context (Acanfora 1967).
The radiocarbon dates for this site are somewhat contradictory (Taschini & Bietti
1972), and we assume that most of its Upper Paleolithic layers date to about | 1 ,000-
10,000 BP. Animals depicted here include a wild cat, two deer, a wild boar and sev-
eral undetermined mammals. Similar hatched and schematic animals are known from
other Late Pleistocene to early Holocene sites in Western Europe: For example, at
Trou du Frontal in Belgium, Gouy in France,and at Sant Gregori in Spain (Aparicio
Pérez 1990; Lejeune 1986; Martin 1989) (Figure 6).

Schematic and naturalistic depictions of animals are not allopatric. At Gr. Roman-
elli, for example, a block had a deer with a hatched filling on one side, and a bovid in
much more naturalistic style on the other side (Acanfora 1967: Figure 2-4).

Other schematic engraved figures have even more rigid, and almost rectangular
shapes. They are found at 7 sites in Apulia and in Sicily, executed both on cave walls
and on small blocks. At Gr. dei Cervi in Levanzo, for example, a stone block with an
engraved rectangular bovid was found at the entrance of the cave in level 3 (Figure
7b). Radiocarbon assays date it to 11,000 to 10,000 years ago (Graziosi 1962: Vi-
gliardi 1982). Itis assumed that the wall engravings found at the rear cave (see above)
are significantly older.

Similarly, at Gr. dell’ Addaura, two rectangular-shaped static bovids are superim-
posed ona very complex scene, described above, which depicts humans and natural-
istic animals (Graziosi 1956: Tav. 298; Graziosi 1973: Figures 66-67). At Gr. del
Cavallo in Apulia there is a “rectangular” male engraved on a block (Palma di Ces-
nola 1972; Vigliardi 1972) (Figure 4). While this object was not recovered im situ, it
can, nonetheless, be associated with local Romanellian and Epiromanellian assem-
blages, which date slightly younger that the levels at Gr. Romanelli, and are, possibly,
early Holocene in age.

Geometric engraved or painted decorations are found all over Italy (Figure 8).
Some specific patterns are geographically limited: Ribbon-like ones done on little
stone blocks are known from Southern Apulia, from Gr. Romanelli (Acanfora 1967:
Figures 23-27), and from Gr. del Cavallo (Vighiardi 1972: Tav. IL), The ones from Gr.
Romanelli are somewhat younger, and possibly date to the early Holocene. Similar
patterns have been found on stone slabs at Cueva del Parpallo in Eastern Spain, were
they came from the Magdalenian ITI-IV levels (Fortea Perez 1978). The geographical
and chronological gap between the two Is of sufficient magnitude, however, to pre-
clude direct filiation of the two (contra Graziosi 1956: see below).

Other patterns are more widely distributed both inside and outside the peninsula.
Riparo Villabruna, in northeastern Italy, where there was rapid sedimentation around
12,000 BP, contains remnants of red painting on the wall (Broglio 1992). This site

PALEOLITHIC IMAGES FROM ITALY

(g ‘sandnyjdsg (/ Shay Jap auog(-g yenyy (¢ UNO] (f SUOquIY,p jog (¢

Lt
b oe

or

Hfaucuoy (T] pur ‘oyeacy (TT ‘tursayog (OT ‘opjedaeg (6 {o8aID jus
‘AnOD (J ‘[eo1y P Nor (1 “Burpy payorey v yyim speuurue pa AvIsua JO MONNQISIp ayy “9 AWAD

>

cS)
a

°6

228 MUSSI AND ZAMPETTI

b

FiGuReE 7. Examples of schematic art. a) from Gr. Romanelli (length = 7.5 em) (after Acanfora 1967); b)
from Gr. dei Cervi, level 3 (Levanzo) (length =17.5 cm) (after Graziosi 1973)

produced a burial in which the skeleton ofa male was covered by large stones, one of
which was painted with a red herringbone geometric pattern — possibly depicting
some sort ofa plant. A similarly painted block was found in the lower part of the ar-
chaeological deposit.

Another block witha geometric painting in red was discovered in level C-B at Gr.
Romanelli, and is slightly younger in age. The design here consists of rows of little
arched patterns placed next to each other (Cardini 1972).

Different styles of depiction are not spatially discrete. As we have discussed
above, a male burial dated to about 12,000 BP, found at Riparo Tagliente in northeast-
ern Italy, was also covered with stone blocks, one of which was decorated with a natu-
ralistic engraving of a lion.

Other geometric designs are found on pebbles. Painted or engraved Azilian type
pebbles have been found at several Late Pleistocene and early Holocene sites (Figure
9). A painted “M” motif is on a piece from Gr. delle Arene Candide, dating to ca.
1 1,000-10,000 BP, as well as on a piece from Gr. della Madonna, dating to the early
Holocene (ca. 9,000 BP) (Cardini 1972) (Figure 10c). A similar motif is known from
Mas d’Azil (Couraud 1985: Figure 29.14).

Other either painted or engraved pebbles are similar to Azilian pebbles found at
French or Spanish sites. Azilian pebbles with red points, common at the type site.
have also been found in Italy, out of context at Gr. delle Prazziche (Borzatti von
Léwenstern 1965) and in situ at Gr. della Serratura. At the latter they are associated
with Final Epigravettian industry, radiocarbon dated to 10,000 + 130 (UtC-754),
10,270 + 140 (UtC-755), and 10,220 + 60 (BIn-3571) BP (Martini 1992).

A pebble from Riparo Tagliente and one from Terlago (Figure 10a), both found
out of context, can be dated to the generic late Upper Paleolithic. Both are decorated

PALEOLITHIC IMAGES FROM ITALY 229

23

UL

FiGure 8. The location of sites with geometric engravings or paintings, including Azilian pebbles, ca
13,000 — 10,000 BP. Legend: gray areas are above 500 m. asl. 1) Gr. di Grimaldi (Gr. dei Balzi Rossi); 2)
Gr. delle Arene Candide; 3) Terlago; 4) Riparo Tagliente; 5) Riparo di Villabruna; 6) Gr. della Ferrovia; 7)
Gr. Polesini; 8-9) Gr. Maritza, Gr. di Ortucchio; 10) Gr. Paglicci; 11) Gr. della Serratura; | 2) Gr. della Ma -
donna; 13) Riparo del Romito; 14-16) Gr. dei Pipistrelli, Cozzica, S. Martino; 17) Gr. delle Mura; 18-19)
Gr. del Cavallo, Gr. delle Vener1; 20) Gr.Romanelli; 21) Gr. delle Prazziche; 22) Gr. dei Cervi (Levanzo);
and 23) Gr. Giovanna.

with groups of transversal incisions in a striped pattern (Bagolini & Dalmeri 1983;
Leonardi 1972). They are quite similar to those recovered from Mas d’Azil in France
illustrated in Couraud (1985).

The painted pebbles from Gr. dei Cervi on the islet of Levanzo closely match the
incised ones from the Azilian site of Rochedane (France) (Graziosi 1973: Tav. IX;
Thévenin 1989: Figure 7). In both inventories groups of transverse lines are parted in
two and symmetrically arranged.

Fusiform engraved patterns, filled with transverse incisions which follow the
maximum diameter of the pebble, are found on a specimen recovered from level 2 at
Gr. Paglicci (Figure 10b). This level dates to the very Late Pleistocene (Mezzena &
Palma di Cesnola 1992). The same pattern also occurs ona specimen from Gr. Poles-

230 MUSSI AND ZAMPETTI

Ficure. 9. The location of sites with Azilian pebbles (after Mussi& Zampetti_ 1988). 1) Gr. di Grimaldi (Gr
dei Balzi Rossi); 2) Gr, della Arene Candide; 3) Gr. Polesini; 4) Gr. della Madonna; 5) Gr. dei Cervi (Le-
vanzo), 6) Gr. delle Prazziche; 7) Gr. Romanelli; 8) Gr. Paglicei; 9) Gr. della Ferrovia; and 10) Terlago

FiGure 10. Examples of Azilian pebbles. a) fragment of an engraved pebble from Terlago (after Bagolini &
Dalmeri 1983), b) engraved pebble from Gr. Paglicci, level 2 (redrawn after Palma di Cesnola 1988); and c)
red-painted pebble from Gr. delle Arene Candide (redrawn after Graziosi 1974)

PALEOLITHIC IMAGES FROM ITALY 231

ini (see above for a discussion of the chronology of this site) (Radmilli 1974), on
specimens from Abri Dufaure in France (D’Errico 1989: Figures 353 and 355), as
well as ona piece from the Belgian site of Leduc (Lejeune 1986: Figure 18). The last
two sites date to the late Upper Paleolithic.

The incised barbed wire pattern found on a pebble discovered out of context at the
late Upper Paleolithic or Mesolithic site of Gr. della Ferrovia (Broghio & Lollini
1982) is also found on Azilian bone pendants from Rascano and La Chora in Cantab-
rian Spain (Utrilla 1990: Figure 6).

Recognizable engraved bone tools are exceedingly rare in Italy. Discovered frag-
ments of engraved bones may well represent remnants of bone tools. Quite similar
ones have been found at sites distant from each other both inside and outside the pen-
insula. For example, the quadrangular pattern on the specimens from Gr. Polesini
(Radmilli 1974: Figure 28), those from Riparo Tagliente (Guerreschi & Leonardi
1984:272), Gr. Maritza (Grifoni & Radmilli 1964: Figure 16.1), as well as Gr. du Tai
in France (Marshack 1973).

Conclusions

The study of Paleolithic art in Italy has been strongly influenced by the approach
advanced by Paolo Graziosi. He began his research on Paleolithic art in the thirties,
when he compared the schematic and geometric engravings found at Gr. Romanelli to
similar ones recovered at Gr. del Parpallo in Spain. This eventually led to his well-
known hypothesis about the existence of a “Mediterranean Province” in art codified
in his monograph, L ‘arte dell'antica eta della pietra, published in 1956 and translated
into several languages. By that time a few more sites were discovered in Italy, includ-
ing Gr. Polesini and Gr. dei Cervi at Levanzo. The main features of Graziosi’s “Medi-
terranean Province” included simplified animal representations, often quite rigid and
schematic, and the prevalence of geometric and abstract patterns.

In another monograph, L ‘arte preistorica in Italia, published in 1973, Graziosi in-
troduced some chronological refinements and restricted the “Mediterranean Prov-
ince” to the artistic production of the final Upper Paleolithic. He accepted that the
earlier art, that found in association with Gravettian industries, was a part of the broad
Franco- Cantabrian sphere. By this time his “Mediterranean Province” was becoming
increasingly restricted to Southern Italy.

The last 20 years have seen new discoveries and have resulted in a more articulate
knowledge of Paleolithic art, as well as a better understanding of the Italian Paleo-
lithic chronology. To date, we know the existence of engravings representing some
131 animals which have been found at 23 sites dating to the final Pleistocene through
the early Holocene (Final Epigravettian). The vast majority of these images depict
aurochs and equids (including the wild ass, Equus asinus hydruntinus, found at Sicil-
ian sites), followed by deer (Figure 11). Wolves, felids, ibexes, bisons, wild boars, la-
gomorphs, and birds, taken together account for a dozen of specimens. Some 24
depictions of humans are also found at six of these sites. Geometric art is even more
diffuse: In a preliminary inventory of decorated pebbles, for example, we counted
more than 160 specimens from 35 sites — more than one hundred of which had en-
graved or painted geometric patterns (Mussi & Zampetti 1992).

In the last version of his hypothesis about the “Mediterranean Province,” Graziosi
merged the Gravettian art of Italy with that from the rest of Europe. We now argue that
the same is true for later art as well, because Late Pleistocene art from Italy finds many

MUSSI AND ZAMPETTI

nN
Nn

‘papnyour ye jejaued pure ayqeuod yiog ‘sats ueniaavididgy yeuly wloy speumtue stewlayds pue onstpemnjeu pareisuq “|| FOI
%

uesyooiny
spinb3
S130q
spluodaq
spljed
suosigq
saxeq|
SJE0 PIIM
SOAIOM
/SPllq

PALEOLITHIC IMAGES FROM ITALY

parallels in the Franco-Cantabrian region. In the sphere of naturalistic art, both zoo-
morphic and anthropomorphic, we have noted several examples from Riparo Tagli-
ente, Vado all'Arancio, Gr. Polesini, Gr. Paglicci, and Gr. dei Cervi at Levanzo,
among others, which are similar to engravings discovered elsewhere in Western
Europe.

The same holds true for the schematic art as well. Recent discoveries and new
studies in Western Europe have yielded a better understanding of the Azilian and Epi-
paleolithic art. These works have added many animal engravings to the traditional
geometric repertoire (Aparicio Pérez 1990; Guy 1993; Lejeune 1986; Lorblanchet
1989; Martin 1989). The shapes of these images are rigid and poorly detailed, and fre-
quently filled, either with hatching or with dotting, as is the case with Italian sche-
matic art. The chronology, unfortunately, is not always firmly established on both
sides of the Alps. In both areas, however, the art is apparently more or less contempo-
rary and dates to the last two millennia of the Pleistocene as well as to the early Holo-
cene. The same is true for the classic Azilian pebbles discovered all over the peninsula
as well as in Sicily.

We suggest, therefore, that it is no longer possible to oppose the “Mediterranean
Province” to the Franco-Cantabrian one. In arguing this, we are not suggesting that
artistic production had the same meaning all over Europe during some 25,000 years or
more. There are clearly some major differences between the art found in Italy and in
Franco-Cantabria. First, the quantities are different, with only a few engravings found
at each Italian site. The animal depicted in Italy obviously do not include reindeer,
mammoths, saiga antelopes and other taxa which did not live in this southern region.
Furthermore, cave paintings are almost completely absent from Italy — a phenomenon
which still awaits explanation. Next, figurines are not found at sites dating to the final
Pleistocene. The lack of ivory and reindeer antler may partially explain their absence,
but only partially so.

The very end of the Upper Paleolithic and the early Holocene saw a unique devel-
opment in the art of southern Italy. The so-called “rectangular” animals are an exam-
ple of this. Some early Holocene sites in Southern Apulia have other unusual items
which we have not discussed, namely, a multitude of small irregularly shaped lime-
stone blocks, covered by tightly engraved linear patterns, mostly geometric in nature.
These have been found at Gr. delle Mura, Gr. delle Veneri, Gr. Marisa, Gr. di Porto
Badisco, and Gr. del Cavallo (Calattini 1992; Cremonesi 1992; Guerri 1992; Vi-
gliardi 1972).

We can understand the “art phenomenon” better if we link it to the local geogra-
phy and demography. First, Italy is effectively separated from the rest of Europe by
the Alps. It is further subdivided by high mountain ranges, such as the Apennines, as
well as by water, which produced islands. These natural barriers created problems for
the circulation of people, goods, and ideas. In spite of this, however, it is clear that ex-
changes over long distances did take place.

As we noted above, there is a marked increase in the number of sites in Italy
through time. Less than 30 sites date to the Gravettian or initial Early Epigravettian, a
period of 7,000 years. The 3,500 year time span following the glacial maximum, the
late Early Epigravettian and the Evolved Epigravettian, is represented at ca. 40 sites.
The subsequent Final Epigravettian, found over a timespan of some 4,500 years is
known from ca. 80 sites. In addition to the increase in site numbers, we also see a pro-
gressive expansion of occupation through time. Thus, for example, the Apennines be-
come occupied shortly after the glacial maximum while Sicily has sites dating after
14.500 BP. We assume that this increase in site density and areas occupied reflects not

234 MUSSI AND ZAMPETTI

only better preservation of younger deposits, but also mirrors a real increase in the
number of groups inhabiting Italy.

During the Gravettian and the initial Early Epigravettian, data on the compara-
tively sparse population extant in Italy, on the comparatively high number of works of
art, as Well as on ceremonial burials, suggest that the very sparse population engaged
in intense ritual activities. During this time it appears that symbolic activity, and the
ensuing “artifact mode of communication,” as defined by Wobst (1977), was used to
counterbalance the dangers of dispersion over a large territory. These dangers in-
cluded the threat of extinction of isolated populations — something which can occur
from sheer stochastic imbalance between sex and age classes (Mussi 1990; Mussi &
Zampetti 1988; Mussi et al. 1989).

The subsequent millennia saw a slow but steady rise in population density. This
more benign demographic situation apparently led to a decrease in ritual activities —
for this time period, works of art are rare and burials absent.

[he very end of the Pleistocene saw a further demographic expansion. This expan-
sion permitted, for the first time, the establishment of closed mating networks, sensu
Wobst (1976; Mussi er al. 1989). At this time there was also a marked regional diver-
sification in lithic industries and other technological and cultural activities. The in-
creased relevance of rituals, as expressed by art, indicates long-distance relationships
which were possibly geared towards counteracting exclusive access to restricted re-
gions. As has been pointed out by Wiessner (1983) and others, social storage of obli-
gation, done via exchange relationships and ensuing debts, assure access to distant
territories. We suggest that the intense and diversified artistic activities at the end of
the Pleistocene resulted from just such a pattern of exchange.

Literature Cited

Acanfora, O, 1967. Figurazioni inedite della grotta Romanelli. Presentazione preliminare.
Bull. di Paletnol. Ital. n.s. 18:7-67.

Aparicio Pérez, J. 1990. Chronologie de l'art mobilier paléolithique dans |'Espagne Mediter-
ranéenne. Pages 101-120 in Jean Clottes, ed., L'drt des Objets au Paléolithique,
vol. /. Picard, Paris, France.

Bagolini, B. & G. Dalmeri. 1983. Site Paléolithique tardif- Mésolithique du lac de Terlago
(Trento). Preistoria Alpina 19:189-196.

Bartolomei, G., A. Broglio, A. Guerreschi, P. Leonardi, C. Peretto & B. Sala. 1974. Una sepol-
tura epigravettiana nel Riparo Tagliente in Valpantena (Verona). Rivista di Sci. Preistor.
29: 101-152.

Bartolomei, G., A. Broglio, P. Cassoli, M. Cremaschi, G. Giacobini, G. Malerba, A. Maspero,
M. Peresani & A. Tagliacozzo. 199]- 1993, Risultati preliminari delle nuove ricerche al Ri-
paro di Fumane. L'dnnuario Storico della Valpolicella: 9-64

Bisson, M.S. & P. Bolduc. 1994. Previously undescribed figurines from the Grimaldi Caves.
Curr. Anthropol. 35:458-468

Boldue, P., J. Cing-Mars.& M. Mussi. 1996. Les figurines des Balzi Rossi (Italie): Une collec-
tion perdue et retrouvee. Prehist. Arieg. LI:15-53.

Borzatti von Lowenstern, E. & D’Eorzatti von Lowenstern, E. de 1965. Oggetti romanelliani
con testimonianze d'arte nella Grotta delle Prazziche (Novaglie, Lecce). Rivista di Sei. Pre-
istor, 20:303-306.

Broglio, A. 1992. Le pietre dipinte dell'Epigravettiano recente del Riparo Villabruna-A (Dolo-
miti Venete). Pages 223-237 in A. Revedin, ed. Atti della XXVIII Riunione Scientifica

PALEOLITHIC IMAGES FROM ITALY

tN
(os)
in

dell'Istituto Italiano di Preistoria e Protostoria. \stituto Italiano di Preistoria e Protostoria,
Firenze, Italy.

Broglio, A. & D.G. Lollini. 1982. | ritrovamenti marchigiani del Paleolitico superiore e del
Mesolitico. Pages 27-61 in / Convegno Beni Culturali e Ambientali delle Marche. Paleani,
Rome, Italy.

Buisson, D. & H. Delporte. 1988. Interét d'un raccord pour I'authentification d'une oeuvre d'art.
Bull. de la Soc. Préhist. Francaise 85:4-6.

Calattini, M. 1992. Oggetti di arte mobiliare dallo strato mesolitico di Grotta delle Mura BA.
Thesis. Pages 293-301 in A. Revedin, ed. Atti della XXVIII Riunione Scientifica dell'Isti-
tuto Italiano di Preistoria e Protostoria. \stituto Italiano di Preistoria e Protostoria, Firenze,
Italy.

Capitan, L. & D. Peyrony. 1928. La Madeleine, son Gisement, ses Industries, ses Eouvres
d'Art. Institut International d'Anthropologie, Paris, France.

Cardin, L. 1942, Nuovi documenti sull'antichita dell'uomo in Italia: Reperto umano del Paleo-
litico superiore nella Grotta delle Arene Candide. Razza e Civilita 3:5-25.

. 1971. Rinvenimenti paleolitici nella Grotta Giovanna (Siracusa). Pages 29-35 in Atti
della XIII Riunione Scientifica dell'Istituto Italiano di Preistoria e Protostoria, \stituto
Italiano di Preistoria e Protostoria, Firenze, Italy.

. 1972. Dipinti schematici antropomorfi della Grotta Romanelli e su ciottolo dei livelli
mesolitici della Caverna delle Arene Candide e della Grotta della Madonna a Praia a Mare.
Pages 225-235 in Atti della XIV Riunione Scientifica dell'Istituto Italiano di preistoria e
Protostoria, \stituto Italiano di Preistoria e Protostoria, Firenze, Italy.

Chollot, M. 1964. Collection Piette. Musées Nationaux, Paris, France.

Clottes, J., A. Beltran, J. Courtin & H. Cosquer. 1992a. La Grotte Cosquer (Cap Morgiou, Mar-
seille). Bull. Société Préhist. Frangaise 89:98-128.

Clottes, J., J. Courtin & H. Valladas. 1992b. A well-dated Paleolithic cave: The Cosquer Cave
at Marseille. Rock Art Res, 9:122-129.

Couraud, C. 1985. L'art azilien. Origine-survivance. XX Suppl. Gallia Préhist.C.N.R.S., Paris,
France.

Cremonesi, G. 1992. Manifestazioni d'arte mobiliare dai livelli epiromanelliani di Grotta delle
Veneri di Parabita e da Grotta Marisa presso Otranto (Lecce). Pages 303-315 in Atti della
XXVIII Riunione Scientifica dell'Istituto Italiano di Preistoria e Protostoria, Firenze, \taly

D'errico, F. 1989. L'Art Grave Azilien. Analyse Microscopique, Reconstitution Gestuelle, Sig-
nification. These a |'Institut de Paléontologie Humaine, Paris, France.

. 1994. Birds of Cosquer Cave. Rock Art Res. 11:45-57.

Fortes Perez, F.J. 1978. Arte paleolitico del Mediterraneo espanol. Trabajos de Prehistoria 35:
199-149.

Gioia, P. 1984-1987. L'industria litica di Fontana Nuova nel quadro dell'Aurignaziano Italiano.
Origini XU: 27-58.

. 1990. An aspect of the transition between Middle and Upper Paleolithic in Italy: The
Uluzzian Pages 241-250 in C. Farizy, ed., Paléolithique Moyen Recent et Paléolithique
Superieur Ancien en Europe, Mem. Musée de Préhistoire d'lle de France 3.

Graziosi, P. 1956. L'Arte dell'Antica eta della Pietra. Sansoni, Firenze, Italy.

. 1962. Levanzo. Pitture e Incisioni. Sansoni, Firenze, Italy.

. 1973. L'Arte Preistorica in Italia. Sanson, Firenze, Italy.

Grifoni, R. & A.M. Radmilli. 1964, La Grotta Maritza e il Fucino prima dell'eta romana. Rivista
di Sci. Preistor. \9:53- 127.

Guerreschi, A. & P. Leonardi. 1984. La fine del Paleolitico superiore (Epigravettiano finale).
Pages 243-281 in A. Aspes, ed., // Veneto nell'Antichita.Banca Popolare di Verona, Verona,
Italy.

236 MUSSI AND ZAMPETTI

Guerri, M. 1992. Scoperte di arte visiva paleolitica nella Grotta di Porto Badisco (Lecce).
(Scavie ricerche 1981-89). Pages 317-325 in A. Revedin, ed., Atti della XXVIII Riunione
Scientifica dell Istituto Italiano di Preistoria e Protostoria, \stituto Italiano di Preistoria e
Protostoria, Firenze, Italy.

Guy, E. 1993. Enquete stylistique sur l'expression figurative épipaléolithique en France: De la
forme au concept. Paleo 5:333- 373.

Lejune, M. 1986. Aspects peu connus de l'art paléolithique et mésolithique de Belgique. Syn-
these et problemes. L'Anthropologie 90:685-700.

Leonardi, P. 1972. Bisonte graffito e incision lineari e geometriche del deposito epigravettiano
del Riparo Tagliente net Lessini (Verona). Rivista di Sci. Preistor. 27:225-247

Leroi-Gourhan, A. 1972. Considerations sur l'organisation spatiale des figures animales dans
l'art parictal paleolithique Pages 281-300 in M, Almagro Basch & M. A. Garcia Guinea, Ac-
tas del Symposium Internacional de Arte Rupestre. Santander, Italy.

—__. 1986. Préhistoire de l'art occidental. 7th ed. Mazenod, Paris, France.

Leroi-Gourhan, Arl., & J. Allain. 1979. Lascaux Inconnu. XII Suppl. Gallia Préhistoire.
C.N.R.S., Paris, France.

Lollim, D. & M. Silvestrini. n.d. Museo Archeologico Nazionale delle Marche. Ministero per i
Ben Culturali e Ambientali, Soprintendenza archeologica delle Marche.

Lorblanchet, M. 1989, De l'art naturaliste des chasseurs de renne a l'art geéometrique du Meéso-
lithique dans le sud de la France. A/mansor 7:95-124.

Marshack, A. 1973. Analyse preliminaire d'une gravure a systeme de notation de la grotte du
Tai (Dréme). Etudes Préhistor. 4:13-16.

_. 1986. Une figurine de Grimaldi “redecouverte”: Analyse et discussion. L‘Anthro-
pologie 90:807-814.

Martin, Y. 1989. Nouvelles decouvertes de gravures a Gouy. L'Anthropologie 93:513-546.

Martini, F. 1992. | crottol dipinti di Grotta della Serratura (Salerno): Osservazioni sulla cro-
nologia e sul contest industrialt dell'arte “aziliana™ in Italia. Pages 261-275 in A. Revedin,
ed., Atti della NNVII/ Riunione Scientifica dell'Istituto Haliano di Preistoria e Protostorta.
Istituto Italiano di Preistoria e Protostoria, Firenze, Italy.

Mezzena, F. & A. Palma di Cesnola. 1972. Oggetti d'arte mobiliare di eta gravettiana ed epr-
gravettiana della Grotta Paglicc: (Foggia). Rivista di Sci. Preistor, 27:21 1-224.

—_. 1992. Nuove manifestazioni d'arte epigravettiana della Grotta Paglicci: nel Gargano.
Pages 277-292 in A. Revedin, ed., Atti della XXV///Riunione Scientifica dell'Istituto Itali-
ano di Preistoria e Protostorie. \stituto Italiano di Preistoria e Protostoria, Firenze, Italy

Minellono, F. 1985-1986. L'incisione di testa maschile dal Riparo di Vado all'Arancio (Gros-
seto). Rivista di Sci. Preistor. 40:1 15-135.

Movius, H. & S. Judson. 1956. The Rockshelter of La Colombiere. 4m. School Prehist. Re-
search Bull. 19, Peabody Museum, Cambridge.

Mussi, M. 1990. Continuity and change in Italy at the last glacial maximum. Pages 126-147 in
O. Soffer & C. Gamble, eds., The World at 18000 BP. Vol. 1; High Latitudes. Unwin Hy-
man, London, U.K.

. 1990-1991. Un ciottolo di Grotta Polesini con bovide inciso recentemente ricomposto.
Origini XV. 155-174.
_. 199]. L'utilisation de la stéatite dans les grottes des Balzi Rossi (ou grottes de Gri-
maldi). Gallia-Préhistoire 33:1- 16.
_. 1992. I] Paleolitico ei] Mesolitico in Italia. Popol: e Civilta dell'Ttalia Antica, vol. X.
Biblioteca di Storia Patria. Stilus BSP, Bologna, Italy
1995. Les statuettes itallennes de pierre tendre: Savignano et Grimaldi. Pages 169-185
in H. Delporte, ed., Actes du Colloque “La Dame de Brassempouy, ses Ancétres, ses Con-

PALEOLITHIC IMAGES FROM ITALY ye)

ies)
|

temporaines, ses Heritiéres’’. Etudes et Recherches Arcéhologique de l'Université de

Liege 74, Brassempoy, France.

. 1996. Problemes recents et decouvertes anciennes: La statuette de Savignano (Modene,
Italie). Prehist. Arieg. LI:55-79.

Mussi, M., D. W. Frayer & R. Macchiarelli. 1989. Les vivants et les morts. Les sépultures du
Paléolithique supérieur en Italie et leur interprétation. Pages 435-458 in I. Hershkovitz, ed.,
People and Culture in Change, B.A.R. Int. Ser. SO8(1).

Mussi, M. & D. Zampetti. 1988. Frontiera e confini nel Gravettiano e nell'Epigravettiano dell'I-
talia. Prime considerazioni. Scienze dell'Antichita 2:45-78.

. 1992. Ciottoli decorati, ciottoli utilizzati: Problemi di forma e di misura. 4 fi XAV/// Ri-
unione Scientifica dell'Istituto Italiano di Preistoria e Protostorie. 159-168. Pisa, Giardini

. 1993. Ciottoli decorati eciottoli utilizzati della fine del Paleolitico a Grotta Polesini (La-
z10). Bull. di Paletnol. Ital. 84:57-83.

Pales, L. 1976. Les Gravures de la Marche. II: Les Humains. Ophrys, Paris, France.

Palma di Cesnola, A. 1965. Il Paleolitico superiore arcaico (facies uluzziana) della Grotta del
Cavallo, Lecce. Rivista di Sci. Preistor. 20:33-62.

. 1966. I] Paleolitico superiore arcaico (facies uluzziana) della Grotta del Cavallo, Lecce.
Rivista di Sci. Preistor. 2\:3-55.

. 1972. La scoperta di arte mobiliare romanelliana nella Grotta del Cavallo (Uluzzo,
Lecce). Rivista di Sci. Preistor. 27:51-56.

1988. Paglicci. S.Marco in Lamis: Regione Puglia, Assessorato alla Pubblica Istruzione
e Cultura.

Radmilli, A.M. 1966. Le due “Veneri” di Parabita. Rivista di Sci. Preistor. 2\:123-133.

. 1974. Gli scavi nella Grotta Polesini a Ponte Lucano di Tivoli e la pitt antica arte nel
Lazio. Sanson, Firenze, Italy.

Rellini, U. 1912. Osservazioni e ricerche sull'etnografia preistorica delle marche. Ait della So-
cieta dei naturalisti e matematici di Modena, ser. \V, X1V:3-61.

Ricol, D. 1973. Les animaux en position “anormale” dans l'art parietal paléolithique. Cahiers
du Centre de Recherches Préhistoriques \:29-35.

Sacchi, D. 1972-1973. Une curieuse gravure d'oiseau dans le gisement magdalenien de Belvis
(Aude). Zephyrus XXUI-XXIV:189- 191.

Sieveking, A. 1987. 4 Catalog of Paleolithic Art in the British Museum. British Museum, Lon-
don, U.K.

Taschini, M. & A. Bietti. 1972. Quelques remarques typologiques sur les “pointes” du Paléo-
lithique supérieur de la Grotte Romanelli (Castro-Marina, Lecce, Italie). Quaternaria XVI.
271-287.

Thevenin, A. 1989. L'art azilien: Essai de synthése. L'Anthropologie 93:585-604.

Utrilla, P. 1990. Bases objectives de la chronologie de l'art mobilier paleolithique sur cote cant-
abrique Pages 87-99 in J. Clottes, ed., L’Art des Objets au Paléolithique, vol.1. Picard,
Paris, France.

Verbrugge, A.R. 1969. Le Symbole de la Main dans la Préhistoire. Verbrugge: Compiegne.

Vicino, G. 1984. Lo scavo paleolitico al Riparo Bombrini (Balzi Rossi di Grimaldi, Ventimi-
glia). Rivista Ingauna e Intemelia n.s. 39:1-10.

Vicino, G. & S. Simone. 1976. Gravures paléolithiques des Grottes de Grimaldi, Pages 148-
159 in H. de Lumley & L. Barra, eds., Sites Paléolithiques de la Région de Nice et Grottes
de Grimaldi, 1X Congrés U.I.S.P.P., Livret-guide de l'excursion BI, Nice, France.

Vighardi, A. 1972. Le incisioni su pietra romanelliane della Grotta del Cavallo (Uluzzo,
Lecce). Rivista di Scienze Preistoriche 27:57-115.

. 1982. Gli strati paleo-mesolitici della Grotta di Levanzo. Rivista di Sci. Preistor. 37:
79-134.

238 MUSSI AND ZAMPETTI

Weissner, P. 1983. Style and social information in Kalahari San projectile points. Amer. Antig-
uity 48:253-276.

Wobst, H.M. 1976. Locational Relationships in Paleolithic society. J. Hum. Evol. 5:49-58.

__. 1977. Stylistic behavior and information exchange. Pages 317-342 inC. E. Cleland, ed.,

For the director: Research essays in honor of James B. Griffin. Anthropological Papers

n.61. University of Michigan Museum of Anthropology.

Zampetti, D. 1987. L'arte zoomorfa del Paleolitico superiore in Italia. Scienze dell'Antichita |:

9.3

an

—__. 1990-199]. Post-scriptum: l'equide in miniatura inciso su un ciottolo di Grotta Polesini
(Lazio), Origini XV: 121-153.
.1993 La Venere del Trasimeno owvero la rappresentazione del corpo nel Paleolitico su-
periore. Origini 17: 89-106.

____. 1996. The body image in Paleolithic times. Preliminary remarks on the *Trasimeno Ve-
nus” (Central Italy). Pages 631-635 in M. Otte, ed., Actes du Colloque “Nature et Cul-
ture’, Etudes et Recherches Archeéologique de | 'Université de Liege 68/1 Liege, France.

Zorzi, F. 1962, Pitture panetali e oggetti d'arte mobiliare del Paleolitico scoperti nella Grotta
Paglicei presso Rignano Garganico, Rivista di Sci. Preistor, 17:123-137.

239

The Mutability of Upper
Paleolithic Art in Central and
Eastern Europe: Patterning and
Significance

Olga Soffer
Department of Anthropology
University of Illinois

109 Davenport Hall

Urbana, IL 61801

The study of Upper Paleolithic art, like the study of all Paleolithic things in gen-
eral, had its origins in western Europe — more specifically, in France. Even though
some 100 years have passed since the inception of this research, to this day most of the
treatises written about Upper Paleolithic art, at least those widely read by Anglo-
phone and west European scholars, emanate from western Europe or deal with mate-
rials from that part of Eurasia. This centering of research, in what I have argued was
the periphery of the Late Pleistocene oekumene (Soffer 1985, 1995), has led to some
serious distortions of the record and to some unwarranted conclusions about its evolu-
tionary implications. In this chapter I redress these errors by taking a broad look at art
in Upper Paleolithic Central and Eastern Europe, comprising roughly 50% of the
European landmass.

Bypassing epistemological and scholastic arguments about “what is art,” for the
purposes of this survey I consider as “art” — any decorated items, be they utilitarian or
nonutilitarian in function, as well as decorative items such as beads and pendants
which may have been used as items of personal adornment. Furthermore, I structure
the survey by region, time, and depositional context. Finally, because traditional pa-
rietal art is extremely sparse in Central and Eastern Europe, this chapter presents data
on the distribution of portable art. 1 do argue, however, that in the case of painted
mammoth bones found at some Upper Paleolithic sites on the Central European (or
Central Russian) Plain, the weight of these pieces is such as to more properly place
them in the “parietal” rather than the “portable” categories.

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

240 SOFFER

The Early Upper Paleolithic

I begin by examining art from sites occupied before the Last Glacial Maximum
(LGM), dating older than 20,000 BP.

As in western Europe, Central and Eastern data contain a small handful of objects
claimed as Middle Paleolithic “art” or at least “symbolic objects” (e.g., the famous
Tata plaque or the problematic “engravings” from Molodova I and Ternopol’)
(Chernysh 1978; Sytnik 1983). These, as elsewhere, are few in number and their
status as purposefully decorated objects most ambiguous.

Moravia

In this part of the world, the unambiguous appearance and proliferation of art is as-
sociated with the Gravettian technocomplex, appearing earliest in Moravia at sites as-
signed to the Pavlov Culture, dating some 28,000 to 23,000 BP (Figure 1). Sites
assigned to this culture, like the vast majority of those dating to all periods of the Up-
per Paleolithic in Central and Eastern Europe, are open air sites and art objects found
at them represent examples of portable art. Although caves exist in Moravia in close
proximity to the sites with portable art, no examples of parietal art have been found
here dating to this or any other Late Pleistocene period (Jelinek 1982; Svoboda 1986:
Svoboda ef al. 1994).

Moravian art comes from two contexts: a small percentage from burials and the
vast majority from living sites.

Funerary Art. Burials at Moravian sites are very diverse, found in habitation sites,
at their peripheries, as well as at some distances from them. They are quite impover-
ished in art or items of personal adornment, containing only a handful of perforated
arctic fox teeth and pierced fossil shells (Absolon & Klima 1977; Klima 1953, 1958,
1959, 1963, 1964, 1987, 1990, 1991: Svoboda 1991: Svoboda et a/. 1994, ). Those from
Dolni Vestonice, for example, both the one of the older woman at Dolni Vestonice I
and the triple burial at Vestonice I just contain a handful of perforated teeth of arctic
foxes (Klima 1987). The sole exception to this was the isolated Brno II burial, exca-
vated at the end of last century, where an adult male was buried with over 600 dentalia
beads, 14 perforated, often engraved, small disks made on a variety of media (stone,
ivory, bone, and mammoth molars), and a male ivory puppet that measured 20 cm in
height (Jelinek 1982; Svoboda et a/. 1994),

Art of the Living. The art of the living in Moravia, on the other hand, is extraordi-
narily rich and diverse (Figure 2). Like most of the funerary art, it is concentrated at
the large residential sites of Dolni Vestonice I, Pavlov I, and Predmosti. This inven-
tory includes:

1) nonfigurative beads and pendants made of perforated carnivore (mostly arctic
fox) canines, fossil shells, bone, ground stone and ivory, as well as ivory
rings, and diadems (Figure 2.4).

2) figurative ivory pendants, including those which are most likely synecdoches
for the female body (Figure 2.3) One scholar has recently argued that some of them
may be depicting male genitalia (Kehoe 1991) but, a hypothesis which, given the pro-
portions of the pieces (compared to those on the Bmo puppet, for example), this ap-
pears very ambiguous.

24]

CENTRAL AND EASTERN EUROPEAN IMAGES

uguns (¢] pue ‘says OAaYyIYysJOg-yuasoy ayy (Z{ ‘OAaapay ([ | ‘oaoutpn A “eyAououly (Q] ‘UIZaYy (6 ‘[] OAapAoyy
(8 -1YyotAsasi[y (7 ‘VIBYSAO]LUDY-OAary (9 “YOUIYyZayyy (¢ “NsoWpald (fp AAO] ‘AOTAv ‘adtuojsa A tUjO (¢-[ ‘Says o1ptpoayeg Jaddy passnostp ayy Jo uoRI07 *| AUNT

242 SOFFER

3) engravings on bone and especially on ivory, on both utilitarian pieces and non-
utilitarian ones, done in both geometric and curvilinear motifs. This inventory is also
rich in figurative pieces consisting ofanthropomorphic and animal figurines and figu-
rine fragments. It includes sparse examples of engraving of a highly schematized fe-
male form, found at Predmosti, as well as a flat sculpture of a lion from Pavlov I
(Jelinek 1982, 1990), and more numerous objects modeled in the round. Anthropo-
morphic forms are found sculpted out of mammoth metapodia, interpreted as seated,
perhaps female, figurines, as well as ubiquitous depictions of naked depersonalized
females wearing belts low on their hips, best exemplified by the clay figurine from
Dolni Vestonice I (Figure 2.2).

Moravian depictions of the female are extremely diverse, ranging from highly
variable realistic torsos, such as that from Petrkovice, to somewhat stylized depic-
tions accenting breasts and belies with belly buttons such as on the figurine from
Dolni Vestonice I (Figure 2.2), to extremely abstracted engraving such as on the en-
graved tusk from Predmosti (Figure 2.1). These faceless or headless figurines are
complimented by portraiture, both realistic and schematized, as on the carved and en-
graved ivory pieces from Dolni Vestonice | (Jelinek 1982, 1990; Klima 1955, 1986).

The largest part of the Moravian art inventory consists of animal figurines and
figurine fragments made in ivory, bone, and predominantly of fired clay. At Dolni
Vestonice, for example, whose identifiable fired clay inventory numbers 721 pieces,
98% depicts animals (Figure 2.5) (Soffer et a/. 1993).

While the rare depiction of human or female form accords well with the ratio of
humans to animals reported by Leroi-Gourhan (1965) for west European Upper Pa-
leolithic portable and parietal art, the species composition of Moravian animal figu-
rines differs dramatically from those found in southwestern France and Spain.
Moravian animal art is heavily dominated by depictions of carnivores, bears and lions
in particular. At Dolni Vestonice I, for example, where 77 figurines have been recov-
ered, Klima (1984) reports that 59% of those identifiable to a species, depict carni-
vores. This is in stark contrast to west European data where carnivores occur in well
under 10° of the depictions (Leroi-Gourhan 1965).

A comparison of proportions of herbivores to carnivores in the faunal inventories
to those found in art in southwestern France does mirror the infrequency of carnivores
among the skeletal remains recovered (Delporte 1990). While most of the organic re-
mains recovered at the Moravian sites await final quantification, those published for
Pavlov I (Musil 1955, 1958), where carnivores represent 50% of the individuals, come
close to the 45% of them depicted in the Pavlov I figurines. At present we are ina
midst of a study of these animal figurines; our preliminary observations suggest that
many depict young or immature carnivores (Seilt, personal communication, 1991).

The ivory, bone, and stone Moravian art objects were manufactured by a subtrac-
tive technique where their makers brought out the forms by carving, pressing, grind-
ing, whittling, and polishing. The ceramic figurines, on the other hand were made by
an additive process where small bits and pieces of clay were stuck or pushed together
with heads, ears, feet, and tails fashioned separately and added to the bodies (Soffer er
al. 1993). This represents a second example of the additive process in the production
of Upper Paleolithic art. The first — painting — is poorly represented in Moravia, and
reported only at Predmosti where Breuil (1924) noted a fragment of mammoth flat
bone decorated with ocher in a geometric and curvilinear design.

Our recent study of the technology used in making these ceramic objects led us to
conclude that the fired figurine fragments recovered from Moravia, where only one
complete piece is present in an inventory of ~10,000 pieces, were not fashioned as

CENTRAL AND EASTERN EUROPEAN IMAGES

sundry apetuiay (Z (YT amnsy “LL6] etulPy pue Uoposqy Jaye) nsoupaid Woy ysn}

(¢ aml y “EGGT 12 12 FAYJOS sayy) Aeyo
poy ‘nsowparg pur 7 aotuoysa A tu]od Wor] squawifeyy aurnay pue auumay jetty (¢ pue <(p amar ‘Z] d ‘og RUE Jaye) PootuosaA Tuyo Woy syuvpuad 10 speaq
AIoat aanvanary (p (1 nary ‘zd gg] BULIPY Jaye) [ atWosaA TPO Wor por AIOAL paaley
AIOATUO BULAR

(¢‘(, ainaiy gt d‘ogeT Buty Jaye) Appa party ‘fT aotuoysa A tUjod WoT
Bug (1 {says opooped Jaddy) ueravsoyy wody Aqjamafl pue wy °Z dU

244 SOFFER

permanent objects with long use value (Vandiver er al. 1990: Soffer et al. 1993).
Rather, what was important was the performance associated with their making and
firing. This performance often involved purposeful deformation of the animals as
well as piercing them with sharp objects. This puts the ceramic figurative inventory in
stark contrast to animal and human depictions in bone, ivory, and stone whose use-
life was infinitely greater, and suggests even greater complexity in the meaning that
Upper Paleolithic art had for its makers and users.

Aside from being found in greater numbers at the larger residential sites, the distri-
bution of the Moravian art, with the possible exception of the fired ceramics at Dolni
Vestonice I, is not patterned. The pieces came from the cultural layers both inside and
outside features such as dwellings. Data on the ceramics from Dolni Vestonice I indi-
cate that here these objects were found in greatest numbers near hearths and kilns used
to make them. The kilns were located near to, yet clearly sequestered from, the resi-
dential areas of the sites, suggesting that the performance associated with their pro-
duction was a special component of life at this sites. The fact that a small number of
figurative ceramic fragments has been recovered from Dolni Vestonice Il and Pavlov
II. sites with no other art objects, suggests that the production and use of ceramic figu-
rines Were far more ubiquitous than was the making or use of art made of ivory, stone,
and bone.

European Russia

The Russian Plain, some 2,000 km east of Moravia, is the second region with
abundant art predating the LGM. Archaeological inventories here, constituting a part
of the Gravettian technocomplex (Kozlowski 1986; Otte 1982), are assigned to a
myriad of archaeological cultures, with the Kostenki-Avdeevo cultural entity the
most widely represented and best known (Grigor’ev 1993: Soffer 1993). As in Mo-
ravia, art is found at just some of these sites, which cluster in the southeastern half of
the Plain with a few northern outliers.

Sungir’. Sungir’, one of the earliest of these outliers, dates to about 25,000 BP
(Figure |) (Bader 1965). It has been identified as a residential site, occupied during
warm Weather months. The site contained at least 6 burials, only three of which were
undisturbed by periglacial deformation that dislocated the cultural layer. These three,
one of a 60 year old male and a double one of a boy and a girl interred head to head ina
single grave, were found inside one of the dwellings, and contained the vast majority
of art objects. The grave goods consisted of 10,000 ivory beads sewn onto the de-
ceased's clothing, hundreds of perforated arctic fox teeth, stone pendants, ivory
bracelets, ivory animal figurines, and enigmatic latticed ivory disks.

Che art of the living at Sungir’ is more problematic. Objects found outside of the
burials were few and far between — a handful of ivory beads, belemnite pendants, a
latticed ivory disk and two animal figurine amulets in ivory, one complete and an-
other broken. The figurines, like those from the burials, show traces of having been
painted in red and black (White 1993). Since they came from a heavily disturbed cul-
tural layer, it remains unclear if their original use was associated with the living or
Whether they were parts of other disturbed burials.

The Kostenki - Avdeevo Sites. The predominantly funereal context of Sungir’ art
is counterbalanced by the abundance of art in the living areas recovered from sites as-
signed to the Kostenki-Avdeevo archaeological entity and affiliated sites, such as

CENTRAL AND EASTERN EUROPEAN IMAGES 245

Khotylevo II and Gagarino (Figure 1) (Boriskovskij 1984; Grigor’ev 1993: Gvozdo-
ver 1989a, b; Praslov & Rogachev 1982). Although problems remain with the dating
of these residential sites, they span a time period between some 24,000 and 20,000
BP, with most intensive occupation around 22,000 years ago (Svezhentsev 1993).
The typological similarity of the lithic inventories from these sites to those in Mo-
ravia, as well as similarities in the worked organic inventories, in art, and in site struc-
ture, has led numerous scholars to unite the two into the Willendorf-
Pavlov-Kostenki-Avdeevo cultural unity, the genesis of which is hotly debated.
Some see this similarity reflecting wide-spanning interaction networks (e.g., Gamble
1991), while others have argued that given the chronological disparity between the
two, with Moravian sites dating between 2,000 and 6,000 years earlier, demographic
shifts offer a more parsimonious explanation for the likenesses observed (Grigor’ev
1993: Soffer 1987, 1993).

Burials are infrequent here and their inventories disparate. The adult male buried
at Kostenki 18 (Khvoikovskaia) had no grave goods while the young boy at Kostenki
15 was interred with ~150 beads and perforated arctic fox teeth (Praslov & Rogache\
1982).

The art of the living at these sites, as in Moravia, is extremely rich and diverse,
executed in bone, ivory, and stone, produced by the reductive techniques of working
down these media into desired shapes, as well as by additive painting of many objects.
As at Moravia, this art is found in abundance at large residential sites where its distri-
bution is not associated with specific features or parts of the sites. The Avdeevo and
Kostenki inventories include (Abramova 1966; Grigor’ev 1993: Gvozdover 1989a,
b; Praslov & Rogachevy 1982):

1) beads and pendants made of ivory, perforated carnivore canines, stone, fossil

shell, and belemnites,

2) ivory pectorals and diadems incised in geometric, rarely curvilinear, designs,

3) a plethora of engraved utilitarian and nonutilitarian objects incised in flat, shal-

low geometric designs including short lines, crosses, checkworks, chevrons, Zig-

zags, and wedge-shaped notches. A few figurative engravings of schematized
animals have been found here as well.

The sites also contain anthropomorphic figurines, including a sitting figure made
of mammoth metapodia akin to the one recovered at Predmosti, and a few possibly
male ivory figurines.

Most of the human depictions from these sites, however, are of women — repre-
sented by 130 pieces in various stages of completeness. These are made of ivory as
well as of marl and, like their Moravian equivalents, are quite heterogeneous. All are
naked, some have elaborate hairdos but no faces, others both. Gvozdover (1989b) has
done an extensive study of them, identified the artistic cannon used to make them, as
well as delimited four different morphological types depicting different types of adult
women in the different stages of the reproductive cycle. These types are not specific
to particular sites, with more than one found at most of the sites. While G vozdover has
suggested that the four types had different semantic meanings, Delporte (1993) has
argued that they may also reflect different masters or different artistic schools, an in-
terpretation in need of firsthand future research.

Gvozdover (1989a,b), as well as Marshack (1991), have noted the presence of
belts on the chest or waists of many of the figurines as well as of bracelets engraved
with wedge-shaped notches or zigzags and slanted lines (Figure 3). The identification
of the same — in location and design employed — on shovel handles, awls with heads,
and decorated animal phalanges, permitted Gvozdover to identify these objects as

246 SOFFER

analogues of the figurines — namely, as highly stylized depictions of females on ob-
jects, many of which were, at the same time, both utilitarian tools and figurines.

Finally, perforated round marl and ivory disks or medallions recovered at the
Kostenki-Avdeevo sites have been classified as “vulvas” which, if such, represent
even more schematic female imagery (Marshack 1991 with references).

The Kostenki-Avdeevo art inventory also includes a great number of animal figu-
rines executed in bone, ivory, and marl. The larger, more complete, but less numer-
ous, ivory and bone pieces contrast sharply with the more numerous but much
smaller-sized and more fragmented marl inventory (Fradkin 1975). Although marl is
amore fragile material, a quality which may account for the more fragmented state of
these objects, given the Moravian data on intentional fracturing of ceramic figurines,
it is also possible that we are dealing with intentional fragmentation of images at the
Kostenki-Avdeevo sites as well.

The ratio of female to animal figurines at these sites, calculated from data in Abra-
mova (1966), differs dramatically from that observed not only in western but in cen-
tral Europe as well. The inventory at Kostenki I — layer 1, for example, has 57.6%

10 ih

Ficure 3. The location of decoration on the ivory figurines and other objects from the Kostenki and
Avdeevo sites (after Gvozdover 1989b, Figure 8)

CENTRAL AND EASTERN EUROPEAN IMAGES 247

depicting females — indicates that female figurines are more than 13 times as numer-
ous as in western Europe and some 28 times as frequent as at Dolni Vestonice I. The
species composition of the animal figurines also differs greatly from that encountered
at the Moravian sites. At Kostenki I-1, for example, herbivores account for 71.8% of
the depictions compared to the 40% at the Moravian sites. This proportion brings the
Russian sites closer to those observed in western Europe, and is in accord with the
herbivore to carnivore ratios observed among the number of identifiable specimens
noted among the osteological remains recovered at Kostenki I-] (Vereshchagin &
Kuz’ mina 1977).

As in Moravia, the depositional context of the Kostenki-A vdeevo art inventory is
indistinct. While Gvozdover (1989a, b; personal communication 1993) notes that all
the decorated implements with schematized female design (/.e., spatulas and awls) at
Avdeevo came from pits, female figurines were found both in the pits as well as in the
cultural layer. The estimation of how many of these are found inside or outside the
dwellings must await unambiguous reconstruction of the dwellings themselves,
something which is exceedingly hard to do at these extraordinarily complex sites.

The Late Upper Paleolithic

I next turn to the record dating at and after the LGM, between 20,000 and 10,000
BP.

Moravia

The Moravian record indicates depopulation of the area between some 21,000 and
15,000 BP, when the area was sporadically used for occasional hunting forays during
warm weather months by groups positioned elsewhere (Svoboda 1991; Svoboda er
al. 1994). The area became more permanently occupied after some 14,000 BP by mo-
bile Magdalenian groups who left behind a record of cave occupations. The more in-
tensely occupied caves, such as Pekarna, do contain some pieces of portable art
(Jelinek 1982; Klima 1974; Svoboda 1986; Svoboda e¢ al. 1994). This art is akin to
that known from Magdalenian groups elsewhere, consisting of engravings of herbi-
vores on utilitarian and nonutilitarian objects made of antler and bone. The sparseness
of these inventories, as well as the narrowing of the artistic context, when compared to
that preceding the LGM, may indicate a significant decline in artistic production anda
reorientation of the subjects represented from females and carnivores to herbivorous
prey species. Finally, human internments are not known from this time period (Svo-
boda et al. 1994).

European Russia

The Russian Plain. A similar situation prevails in the eastern, Kostenki-A vdeevo,
part of the Russian Plain where there is also a decline in the intensity of occupation
and a significant decline in artistic production and use (Rogachev & Praslov 1982;
Soffer 1993). The artistic context is sparse, consisting of some beads and pendants,
but no female or animal figurines. These few remains are found in cultural layers of
habitation sites. The two burials dating from to this time period (Kostenki 2 and in
layer 3 of Kostenki 14 (Markina Gora) are devoid of grave goods (Praslov & Ro-
gachev 1982).

248 SOFFER

A dramatically different artistic record is in evidence, however, on the central part
of the Plain, where rich and diverse art inventories have been recovered from such
sites as Gontsy, Eliseevichi, Kievo-Kirillovskaia, Mezhirich, Mezin, Timonovka,
and Yudinovo (Figure 1) (Soffer 1985).

No human burials have been found here, with the art of the living found at just
some open air residential sites, those probably occupied during cold weather months
(Soffer 1985). Art and jewelry have been recovered from the cultural layers both in-
side and outside of the dwellings at the sites, and thus, as elsewhere, they are not ex-
clusively associated with specific features such as hearths, dwellings, or pits.

Although this inventory is smaller in size than found on the Plain before the LGM,
itis as diverse, and includes not only portable objects but parietal art as well.

Jewelry is represented by beads and pendants made of ivory, bone, fossil shell, and
amber as well as by engraved ivory bracelets (Figure 4).

Engraving is schematic, nonrepresentational, dominated by geometric “fish-
scales,” zigzags, chevrons, lines, ladders, rhomboids, and geometric meander or spi-
ral, motifs (Marshack 1979). Particular designs are not site-specific, and found on
pieces, most of which are nonutilitarian in nature. Curvilinear motifs are less frequent
and also not site specific. They are found on a few ivory plaques interpreted by some
scholars as schematic maps, as well as on an ivory tusk from Kievo-Kirillovskaia
(Soffer 1985).

Anthropomorphic figurines, on the other hand, are quite site-specific (Figure 5).
Ivory female figurines are far less frequent in number than at the Kostenki-A vdeevo
sites — with only 20 recovered altogether — and show increasing abstraction and sche-
matization through time (Soffer 1985). The oldest, from Eliseevichi (Figure 5.1), 1s
the most realistic and reminiscent of the classical *“Venuses” that predate the LGM.
The semantic focus here, however, has already shifted from emphasizing breasts and
bellies to emphasizing thighs and buttocks. The profile shows the triangular exag-
geration which, in time, comes to increasingly represent the female image across Up-
per Paleolithic Europe (Marshack 1991; Soffer 1985). This schematization and
triangulation is accompanied by the appearance of engraved pubic triangles or Vs on
the figurines. In time, this progressive schematization results in very abstracted fe-
males from Mezhirich (Figure 5.2), dating to about 15,000 BP — identifiable as such
because of the engraved pubic triangles (Softer 1985; Marshack 1991) .

Equally abstracted, but far more redundant and three-dimensional female forms
came from the coeval site of Mezin, dating also to ~15,000 BP, where 17 of these
ivory objects have been recovered (Figure 6) (Shovkoplyas 1965). They range in
length from 2.4 cm to 14.5 cm, in contour from tall and slender to short and squat,
have clearly delimited midsections or waists accented by horizontal lines, and feature
geometric engraving sometimes on their upper parts, sometimes on the lower parts,
but most often over both upper and lower and front and back surfaces. All are highly
polished from repeated use and have their bases broken. When found at the beginning
of this century, they were divided into two types, and on the basis of their general con-
tours identified as “phalluses” and “birds” (Abramova 1966; Volkov 1913). The pres-
ence of pubic triangles on the majority of these pieces led other scholars to consider
them as female figurines (Bibikov 1981; Shovkoplyas 1965; Marshack 1991). The
identification of these pieces as highly abstracted female figurines 1s strengthened
both by their shape, which accentuates the triangular buttocks region, as well as by the
nature of the engraved motifs decorating their surfaces. A number of scholars have
noted the presence of chevrons, rhomboids, and geometric meanders on both the figu-
rines and on the two bracelets found at Mezin (Marshack 1991; Stolyar 1985).

CENTRAL AND EASTERN EUROPEAN IMAGES 249

2 Ww KkllersS SSSHLLLLLIMLLLLE AL...
WiLddirssS Ss RS S Se 3)
( CK WH Zz SOOO LLLLLMMLLLLL.—_—
SSK OPIIIIII EERE
ENS AlLLLAS ALLA ESSSSSSSNU A MLL ALLL
LTT, D> TERROR pics

Pyyyyyyyya:

SSS Se YU SSSSWZ se4
PALME S\N POP ix SNe
WMO WALLLZAMUMU@UJ WSN LMM Ml?

m«x<_ oO yx<« z LEN

2
‘

Ficure 4. Ivory bracelets from Mezin (after Shovkoplyas 1965, Figure LI, LIL).

Using a semiological approach borrowed from Gvozdover (1989a) which traces
the duplication of the decorative motifs from the belts and bracelets found on the clas-
sical “Venus” figurines to highly schematized rendering of females on engraved ob-
jects, and recalling the presence of linearly engraved bracelets on the female figurines
from Willendorf to Kostenki, permits us to associate the wearing of engraved brace-
lets with females. The two complete bracelets recovered from Mezin (Figures 4.1,
4.2), feature slanted-line disjointed chevron design which is anchored by rhomboids
on each bangle, as well as a more complex version of the same motif, with some
rhomboids elaborated into geometric meanders found on the cuff bracelet (Figure
4.2). The same chevrons, rhomboids, and geometric meanders are found profusely
decorating the Mezin figurines, both those classified as “birds,” as well as the so-
called “phalluses” (Figure 7). This redundancy, together with the demarcation of the
belts on the waist as on all Eastern European classical “Venuses,” permits us to se-
curely identify the Mezin pieces as highly abstracted depictions of females.

Extending this decoding even further, into a more tenuous realm, I also suggest
that the slanted rectangles coming out of centered vertical lines found on the fronts of
some of them are highly schematic faces (Figure 8). The rationale for this is again
found in the classic Venuses predating the LGM. With the exception of the portrait
head from Dolni Vestonice I (which has detailing of forehead, brows, nose, and
mouth), females with faces (such as the head from Brassempoy), as well as the other
figurines from Dolni Vestonice, Kostenki, and Avdeevo, along with the head of the
male puppet from Brno II, all have detailed accentuation of the upper halves of the
face: forehead, brows, and noses, but no mouths. While the more realistic of these

250 SOFFER

L 1 1 1 1 J

PiGURE 5. Ivory female figurines from the Central Russian Plain: 1) Eliseevichi (after Soffer 1985, Figure
2.39) 2) Mezhirich (after Soffer 1985, Figure 2.72)

maintain the natural vertical and horizontal positions of the depicted facial features,
the most schematic of these faces, such as on the Dolni Vestonice “Venus,” slants the
merged brows and eyes and omits any features below them. This “face” does accent
the vertical planes of the face — by four very finely engraved straight lines which run
from the brow and eyes down the center of the face and terminate near the clavicles
(Figure 8.2). An analogous representation, as Marshack (1991) has pointed out, is
also found on the engraving from Predmosti (Figure 8.1). This combination of medial
vertical demarcation, originating at the top and branching into two laterally slanting
lines, leads me to suggest that at least parts of the slanting rectangles on the Mezin
figurines may depict faces (Figure 8.3).

Finally, going a step further into this realm, and generating rather than support-
ing a hypothesis, [ suggest that the number four itself may be associated with some
category of femaleness. Specifically, I note that figurines and figurine fragments
from DolIni Vestonice I have four holes punched on top of their heads, four “fat folds,”
two on each side of the back, and that the most complete one has four fine lines en-
graved down the center of its face (Figures 2.1, 8.2). The figurine from Eliseevicht

CENTRAL AND EASTERN EUROPEAN IMAGES 2

wn

Ep
Va

FIGURE 6. Ivory figurines from Mezin (after Soffer 1985, Figure 2.81)

(Figure 5.1) has the same four lines engraved on its chest and belly, with analogous
series of four engraved lines present on the Mezhirich piece as well (Figure 5.2).

While the female figurines from Mezhirich bear no traces of paint, the heavy
shellac coating on the Mezin pieces does not permit us to know if they were painted or
not. Painting, however, is well represented on decorated mammoth bones recovered
from Mezhirich, as well as from Mezin (Soffer 1985). Both inventories feature geo-
metric designs done in red ocher, and in both cases, because of the weights of the ob-
jects involved, we are dealing with nonportable or parietal art.

Finally, sites on the Central Russian Plain dating after the LGM contain a further
example of the parietal art form, which, fora lack ofa better word, I call architectural
art. Mammoth bone dwellings found here feature outside facades constructed of care-
fully selected skeletal elements. The most dramatic example comes from Mezhirich,
where the outside retaining wall of dwelling #1 was made of mammoth mandibles
stacked into a herringbone design. While, given the female-specific association of
this pattern at Mezin, it is tempting to interpret the semantic meaning of this dwelling
as signaling female also, the ubiquity of this geometric pattern makes me hesitant to
do so.

The placement of bones in other dwellings at Mezhirich and elsewhere indicates
also the use of the principles of symmetry and of mirror imagery not only in the con-
struction of individual dwellings but in their relationships to each other as well (Sof-
fer 1985). These observations have led me to argue that what we have here are cases of
architectural art — another example of Upper Paleolithic art which is nonportable but
parietal in nature.

The Urals. Finally, although traditionally defined parietal art such as cave paint-
ings and engravings is extraordinarily scarce outside of the Franco-Cantabrian re-
gion, even from areas with a record of cave occupations, it is not totally absent. The
engraving at Badanj in the former Yugoslavia, and paintings at Cucuilat in Rumania
are two examples (Bahn & Vertut 1988; Carciumaru 1987). The extreme eastern
fringe of Europe, the Urals, have two caves, Ignat'evskaia and Kapovaia, the later dat-
ing between 13,000 and 15,000 BP, which contain monochromatic paintings of mam-
moths and horses executed in red or black (Bader 1965; Petrin 1992). The study of this
art is still in progress, but the published preliminary reports indicate that the subject

SOFFER

(ZT RRp analy ‘SgG] Sedjdoyroys Jaye) (¢ pure “CU eRp ainaty “SoG SeATdoyroys Jaye) Cp (E09
ANAL] SOG] SeATdoyAoys Jaye) (E (Z'O9 UNBL S96 | svdtdoyaoys aye) (72°C 09 UNTLy “S96 | sepAdoyaoys Jaye) (| SSOULIN AL UIZOYY OY] UO FHOUL UISop aULL LOL

CENTRAL AND EASTERN EUROPEAN IMAGES

4

FIGURE 8. Schematic representation of “faces” on Upper Paleolithic engravings and figurines depicting females: 1) on the engraved Predmosti tusk (after Marshack 1992,
p pp g g g Pp g 2

Figure 9A);

) on the figurine from Dolni Vestonice 1; and 3) on the Mezin figurine

4

254 SOFFER

matter Is similar to that found in the Franco-Cantabrian region, with herbivorous prey
species being the dominant subject matter.

Implications and Conclusions

These data from Central and Eastern Europe carry the following implications:

|) First, the presence of both abstracted and realistic depictions of human forms
here, as well as the presence of three-dimensional “in the round” and flat sculptures,
of geometric and curvilinear design, of engravings, and of permanent and perform-
ance art. underscores that Gravettian art east of the Carpathians cannot, as some have
claimed, be equated with sculpture in the round depicting female forms (e.g, Del-
porte 1993).

2) Second, a comparison of the provenience of the Central to the East European art
show that Upper Paleolithic art here was clearly a matter of the living. Art of the dead

objects found in burials — is dominated by items of personal adornment, beads, pen-
dants, perforated carnivore teeth and such. Sparse as funerary data are, they also indi-
cate a decline over time in the wealth of the burial inventories — a phenomenon which,
I have argued elsewhere, reflects the entrenchment of hierarchical social relation-
ships here whose demarcation needed no longer to be hidden from daily view (for a
discussion see Soffer 1985).

3) The inventories of both regions also show the production of stylistically and se-
mantically diverse female figurines which, before the LGM, emphasized breasts and
bellies (Gvozdover 1989b). Although in Moravia recognizable depictions of women
are less numerous than those of animals, female figurines dominate the Russian port-
able art. They, like their Moravian counterparts, are very diverse. We can observe the
existence of both the geometric and the figurative systems of female representation.
While today we do not know if the two were used separately in Moravia, their use in
conjunction with each other has been clearly documented at the Kostenki-Avdeevo
sites. The heterogeneity in female depictions in form, medium, as well as in detailing
or abstraction, strongly argues against the validity of any single explanation — be it
“mother goddess,” “sex symbol,” or “interaction network” emblem for these objects.
Rather, as Munn (1973) has shown for the ethnographic present and Lewis-Williams
(1991) pointed out about Upper Paleolithic art in general, it speaks of semantic poly-
semy.

Itis also likely, as Conroy (1993) has recently noted, that the proliferation of stan-
dardized female depiction at this point of prehistoric time signals the gendering of so-
ciety. The absence of standardization in the depictions of males precludes us from
claiming that this gendering took the form familiar to us today.

4) A comparison of figurative art from the Willendorf-Kostenki-A vdeevo cultural
entity, which dates before the LGM, indicates the presence of many shared motifs.
First among these is the depiction of animals in art, with carnivores dominating Mo-
ravian inventories and herbivorous prey species those from the Russian Plain. Mo-
ravian ceramic data suggest that the production of this category of art objects was
more important for its performance value. The highly fragmented state of the Russian
marl figurines may reflect the same phenomenon as well. At the same time, both re-
gions have animals sculpted in ivory and bone. The high wear polish in evidence ona
number of these pieces suggest prolonged use. Thus, we see a diversity in the use of
animal motifs which likely reflects polysemy.

CENTRAL AND EASTERN EUROPEAN IMAGES 255

5) A diachronic look at female representation clearly shows progressive schemati-
zation and abstraction and the movement from the figurative to geometric system of
expression. While before the LGM we see both the figurative and geometric systems
used to depict females, the geometric one comes to dominate after some 18,000 BP.
As Morphy (1989) has pointed out, the two systems have dramatically different social
implications. The meaning of figurative depiction is more accessible to viewers: All
who saw the Dolni Vestonice or Kostenki *Venuses” knew them immediately to be
females. The meaning of geometric representation, on the other hand, requires decod-
ing. The key to this decoding — an explanation of what stands for what — represents
controlled knowledge and presupposes the existence of “decoders.” Here again, Up-
per Paleolithic art offers us yet another tenuous, but redundant and internally consis-
tent, line of evidence of increasing sociocultural complexity after 18,000 BP.

7) Elaborating on the issue of increasing social complexity, I also suggest that the
increasing complexity of the social world is reflected in the distribution of animal art
after the LGM. Naturalistic or recognizable depictions of animals either in the round
or as engravings all but disappear, continuing only at the sparsely populated peripher-
ies, in Moravia and the southern Urals. In the first case, this art is clearly the art of very
mobile recolonizers. I suspect an analogous colonizing scenario for the much more
poorly known regions of the Urals.

I read the Moravia to Mezin trajectory in animal vs. female depiction to mean the
decrease in the importance of influencing “nature” and the increase in the semantic
importance of manipulable “culture,” more specifically of the social, of people-to-
people relationships. I do not argue that this trajectory has a universal or evolutionary
significance, however. Coeval Urals and Moravian data contradict this. Rather, I see
it as a region specific phenomenon which reflects demographic and sociopolitical
factors, namely, increased density of occupation, increased group size, and increased
duration of occupation at the sites.

8) A diachronic look at the portability or permanence of artistic expression in Cen-
tral and Eastern Europe shows the late appearance of permanent or parietal art here. If,
following Root (1984) we see stylistic marking in portable media reflecting egalitar-
ian relationships, and the marking of place, such as is clearly the case with painted
mammoth bones and architectural art, as associated with hierarchical relationships,
this line of evidence shows that changes in sociopolitical relationships suggested by
other data are also reflected in the types of art produced. This relationship, however, is
clearly not a universal one either, because permanent locations, such as caves, can
serve as temporary aggregation sites and bear witness to corporate group demarca-
tion.

9) Art from Moravia and from the Russian Plain clearly indicates that the prolif-
eration of artistic expression in multiple media occurs from the Early Upper Paleo-
lithic onward. This negates the recent claims by some scholars that the “creative
explosion” — something interpreted as characterizing fully modern human behavior —
occurred only after the LGM, and dates younger than 20-18,000 BP (e.g., Lindly and
Clark 1991). While this pattern may be true for Western Europe and the Levant, as
demonstrated above, it is clearly not true for Central and Eastern Europe.

10) The intensity of this expression, as measured by diversity and volume, did not
increase in time. Central and East European regional records indicate a waxing and
waning in the production and use of art (Figure 9). This contrasts with the Franco-
Cantabrian record which does show an early appearance of beads and pendants, but
where portable and parietal art shows increasing intensity after the LGM.

256 SOFFER

Central and Eastern European data show peaks of artistic production occurring at
different times in different places. If we take art production to reflect cultural elabora-
tion. we see that such elaboration does not build upon itself. While the break in art
production in Moravia can be explained by invoking the depopulation of the region
around the LGM, the same does not hold true for the Russian Plain. This explanation
is also invalid for both regions if we compare across the Pleistocene-Holocene rubi-
con. Early Mesolithic hunter-gatherers were present in both regions: yet. by and
large. art was not (Bonsall 1990; Rybakov 1989). It is absent. I suggest. not because of
decreased capacity, or of prior cognition, or because the climate became benign and
conifers or oaks replaced grasslands, but because the social landscape had changed
once again to a more minimalist one.

11) Taking a global view at the distribution of Late Pleistocene art, we can enter
some corrections to the arguments that it is an example of cultural elaboration found
among climatically stressed groups of northern latitude hunter-gatherers (Gamble &
Soffer 1991). The proliferation of Moravian art and of art at the Kostenki-Avdeevo
sites occurred at interstadial times, before the LGM, and not at peak stadial condi-
tions. The same is true for art from the central Russian Plain which comes well after
the 20-18.000 BP cold vice. The climate has little to do with it directly. Moravian art
at 25,000 BP is not about climate but about dense human settlement.

The climate did affect demographic arrangements of people across the landscape.
Deteriorating conditions at the onset of the LGM did bring about demographic shifts
which probably ultimately led some groups from Central Europe onto the Russian
Plain (Grigor'ev 1993: Soffer 1993). In some, but not all cases, these demographic
shifts resulted in complex social landscapes negotiated through the production and
use of art. Yet population influxes, in and of themselves, did not guarantee the pro-
duction of art. Extant data from the Russian and Ukrainian southern steppe zone also
indicate possible population influx at as well as after the LGM (Leonova 1994). These
influxes, however, did not result in the proliferation of portable or parietal art.

Nor. clearly, are population influxes needed for the production of art. On the cen-
tral Russian Plain the linearization of the natural resources after the LGM led to the
linearization of hunter-gatherer settlements. This created more complex social land-
scapes which were mediated through, among other things, the manipulation of art
(Soffer 1985).

12) Furthermore, the artistic record of Central and Eastern Europe also under-
mines the recent universalizing claims about the isomorphy of artistic expression and
specific mechanisms of social interaction, namely, the existence of aggregation-
dispersal pulsations across the Late Pleistocene world (Barton er al. 1994). The
Gravettian data from Central and Eastern Europe indicate that while such pulsations
may have been part and parcel of settlement systems in Moravia ca. 26,000 BP and
the eastern part of the Russian Plain ca. 22,000 BP, they were absent on the central
part of the Russian Plain (Soffer 1985). Their absence, however, did not appreciably
affect the volume or diversity of the artistic output. The West European record clearly
can no longer be “writ large” to stand for the entire Late Paleolithic oekumene and we
would be well advised to stop viewing it as such.

13) Finally, comparing the west to east European Upper Paleolithic art records, we
can note after Hodder (1990), that Franco-Cantabrian art, dominated as it is by animal
motifs which decorate both cave walls and hunting implements is public art, art of the
hunter, art dealing with control of the landscape and of production. From this vantage
point, the dominance of female figurines, of decorated domestic objects such as
spoons, shovels, and needle cases, and finally painted mammoth bones found inside

CENTRAL AND EASTERN EUROPEAN IMAGES

SW Spain
& Spain

Italy

Germany
& Austria

Czech
Republic

Russian
Plain

a ee ee ey ee eee
30,000 20,000 10,000

Years B.P

FiGURE 9. The appearance and proliferation of art and jewelry across Upper Paleolithic Europe (after Softer
1995 Figure 1)

dwellings, which themselves are examples of architectural or parietal art, all can be
read as domestic art, reflecting the importance of controlling women and reproduc-
tion. The nature of Moravian Magdalenian art and of the art of the Urals negate sucha
facile west-east dichotomy in world-view and meaning, however. The twain do meet,
and this meeting has more to do with how regions are occupied and exploited, and for
how long, rather than with long-lasting ethnic or geographic semantic and cultural
traditions.

Acknowledgments

An earlier version of this chapter was presented at the First Oregon Archaeologi-
cal Retreat and at the symposium “What Means This Art? — Late Pleistocene *Crea-
tive Explosion’ in the Old World,” 58th Annual Meeting of the Society for American
Archaeology. The final version has greatly benefited from insightful comments by
Meg Conkey as well as by Marianna Davydovna Gvozdover, Natalia Borisovna Le-
onova, Alexander Marshack, Helaine Silverman, and all of the participants of the first

258 SOFFER

Oregon Archaeological Retreat: Ofer Bar-Yosef, Meg Conkey, Iain Davidson, Jean
Clottes, Dale Guthrie, David Lewis-Williams, John Parkington, Jean-Philippe Ri-
gaud, and Andrée Rosenfeld. Seminal as their comments may have been, the chapter's
shortcomings remain my own. Furthermore, I wish to thank Jean and Ray Auel who
generously hosted the participants of the first Oregon Archaeological Retreat and al-
lowed a bevy of archaeologists to invade their gracious home to discuss what Ayla.
Rynec, and Jondalar may or may not have made or seen.

My research in Central and Eastern Europe over the last 12 years, on which this
chapter is based, has been generously supported by the U.S. National Academy of
Sciences, the National Geographic Society, the National Endowment for the Hu-
manities, IREX, Fulbright Research Abroad and Faculty Lecturing and Research
Abroad Grants, and the University of Illinois through the good offices of the Research
Board, the Scholars’ Travel Fund, and the Russian and East European Center.

Literature Cited

Abramova, Z. A. 1966. /zobrazhenie Cheloveka v PaleoliticheskomIlskusstve. Nauka,
Moscow-Leningrad, Russia.

Absolon, K. & B. Klima. 1977. Predmosti, Ein Mammut-jagerplatz in Mahren. Academia,
Praha, Czech Republic.

Bader, N. O. 1965 Kapovaia Peshchera. Nauka, Moscow, Russia.

—__. 1978 Sungir’. Nauka, Moscow, Russia.

Bahr, P. G., & J. Vertut. 1988. /mages of the Ice Age. Facts on File, Inc., New York.

Barton, C. M.,G. A, Clark & A. E. Cohen. 1994, Art as information:explaining Upper Paleo-
lithic art in western Europe. World Archaeol. 26:185-207

Bibikov, S. N.. 1981. Drevneishii Muzikal'nyy Kompleks iz Kostei Mmamonta. Naukova
Dumka, Kiev, Russia.

Bonsall, C. (ed.) 1990. The Mesolithic in Europe. John Donald Publishers Ltd., Edinburgh,
U.K

Boriskivsky, P. 1. (ed.) 1984. Paleolit SSSR.. Nauka, Moscow, Russia

Breuil, H. 1924. Note de voyage Paleolithique en Europe centrale. L'Anthropologi 34: 515-

592.

Carciumaru, M. 1987. Marturii ale Artei Rupestre Preistorica in Romania. Editura Sport Tur-
ism, Bucuresti, Romania.

Chernysh, A. 1978. O vremeni vozniknoveniia paleoliticheskogoiskusstva v sviazi s issle-
dovaniiami 1976g. stoianki Molodova I. Pages 18-22 in R. S. Vasil'evskij, ed., Uistokov
tvorchestva. Nauka, Novosibirsk, Russia.

Conroy, L.P. 1993. Female Figurines of the Upper Paleolithic and the emergence of gender.
Pages 153-160 in H. duCross & L. Smith, eds., Women in Archaeology. Occasional Papers
in Prehistory No, 23, Department of Prehistory, Research School of Pacific Studies, The
Australian National University, Camberra, Australia.

Delporte, H. 1990. L'/mage des Animaux dans l'art Prehistoriques. Picard, Paris, France.

___. 1993. Gravettian female figurines: A regional survey. Pages 243-259 in H. Knecht, A.
Pike-Tay, & R. White, eds., Before Lascaux. CRC Press, Boca Raton, Florida.

Fradkin, E. E. 1975. Pamiatniki Paleoliticheskogo Iskusstva Kostenok 1. Avtoreterat Dissertat-
sii na Soiskanie Uchenor Stepeni Kandidataistorichekikh Nauk, Institut Etnografil, Lenin-
gradkaia chast', AN SSSR, Leningrad, Russia.

Gamble, C. 1991, The social context for European Paleolithic art. Proc. Prehist. Soc. 57:17-
32

CENTRAL AND EASTERN EUROPEAN IMAGES 259

Gamble, C. & O. Soffer. 1991. Introduction. Pleistocene polyphony: The diversity of human
adaptations at the Last Glacial Maximum. Pages 1-23 in O. Soffer & C. Gamble, eds., The
World at 18 000 BP, Vol. I. Unwin Hyman, London, U.K.

Grigor’ev, G. P. 1993. The Kostenki-Avdeevo Archaeological Culture and the Kostenki to
Clovis. Pages 31-50 in O. Soffer & N. D. Praslov, eds., From Kostenki to Clovis. Plenum
Publishing Corp, New York.

Gvozdover, M. D. 1989a. Ornamental decoration on artifacts of the Kostenki culture. Soviet
Anthrop. Arch. 27:8-31.

. 1989 b. The typology of female figurines of the Kostenki Paleolithic culture. Soviet An-
throp. Arch. 27: 32-94.

Hodder, I. 1990. The Domestication of Europe. Cambridge University Press, Cambridge, U.K.

Jelinek, J. 1982. Bol'shoi Iliustrirovanny Atlas Pervobytnogo Cheloveka. Artia, Prague,
Czech Republic.

. 1990. Artin the Mirror of Ages. Anthropos, Moravian Museum, Brno, Czech Republic.

Kehoe, A. 1991. No possible, probable shadow of doubt (fertility symbols of Palaeolithic
Europe). Antiquity 65:129-131.

Klima, B. 1953. Pavlov, nove paleoliticke sidliste na jizni Morave. Archeologicke Rozhledy. 6:
721-728, 753-758.

. 1955. Vysledky archeologickeho vyskumu na taboristi lovcumamutu v Petrkovicich,
okr. Ostrava v roce 1952 a 1953. Caspois Slezskeho Muzea 1V:1-35.

. 1958. Upper Paleolithic Art in Moravia. Antiquity XXXII:8-14.

. 1959. Objev paleolitickeho phrbu v Pavlove. Arkheologickerozhledy. X\:305-316.
. 1963. Dolni Vestonice. Nakladatelstvi Ceskoslovenske Akademie Ved, Praha, Czech
Republic.

. 1964. Paleoliticka reliefni plastika Ivici ze sidliste u Pavlov. Pamatky Arkheologicke
LV:82-90

. 1974. Archeologicky vyskum plosiny pred jeskyni Pekarnou. Studie Archeologickeho
Ustavu Ceskoslovenske Akademie Ved v Brne. Rocnik I: 1.

__. 1976. Die palaolitische Station Pavlov Il. Prirodovedene Praceustavu CSAV v Brne
(Acta Se. Nat. Brno). X,4. Academia, Brno, Czech Republic.

. 1984. “Les represantations animales du Paleolithique superieur de Dolni Vestonice
(Tchecoslovaquie). Pages 323-332 in H.-G. Bandi, ed., La Contribution de la Zoologie et
de l'Ethnologie al'Interpretion de l'Artdes Peuples Chasseur Prehistoriques. Ed. Universi-
taires, Fribourg, France.

. 1986. Najstarsi Osidleni Breclavska. Regionalni Muzeum v Mikulove, Mikulov,
Czech Republic.

. 1987. Mladepaleoliticky trojgrob v Dolnich Vestonicich. Archeologicke Rozhledy 39:
241-254, 353-357.

. 1990. Lovei Mamutu z Predmosti. Academia, Praha, Czech Republic.

. 1991. Die jungpalaelitischen Mammutjager-Siedlungen Dolni Vestonice und Pavlov in
Sudmahren-CSFR. Archaologie und Museum Heft 023:3-30.

Kozlowski, J. 1986. Gravettian in Central and Eastern Europe. Pages 201-251 in F. Wendorf &
A. Close, eds., Advances in World Archaeology. Vol. 5. Academic Press, Orlando, Florida.

Leonova, N. B. 1994. The Upper Paleolithic of the Russian Steppe Zone. J. World Prehist. 8:
169-210.

Leroi-Gourahan, A. 1965. Prehistoire de l'Art Occidental. Picard, Paris, France.

Lindly, J. M. & G. Clark. 1990. Symbolism and modern human origins. Curr. Anthropol. 31:
233-61.

Lewis-Williams, J. D. 1991. Upper Palaeolithic art in the 1990s: A southern A frican perspec-
tive. S. Afr. J. Sci. 87:422-429.

260 SOFFER

Marshack, A. 1979. Upper Paleolithic symbol systems of the Russian Plain: cognitive and
comparative analysis of complex ritual marking. Curr, Anthropol. 20:271-311.

__. 199]. The female image: a “time factored’ symbol. A study in style and aspects of image
use in the Upper Paleolithic. Proc. Prehist. Soc. 57:17-32.

Morphy, R. 1989. Introduction. Pages 1-20 i H. Morphy, ed., Animals into Art. Unwin Hy-
man, London, U.K.

Munn, N. 1973. Walbirt conography. Cornell University Press, Ithaca, New York.

Musil, R. 1955. Osteologicky material z Paleolitickeho sidliste v Pavlove. Prace Brnenske
Zakladny Ceskoslovenske Akademie Ved. Sesit 6, Spis 319. Rocnik XXVII, pp. 279-319.

____. 1958. Osteologicky Material z Paleolitickeho Sidliste v Pavlove. Anthropozoikum 8:
83-106, tab. |— IX.

Otte, M. 1982. Le Gravettien en Europe Central. De Tempel, Brugge, Belgium.

Petrin, V. T. 1992. Paleolitichekoe Svyatilishche v Ignatievskoj Peshcherena luzhnom Urale.
Nauka, Novosibirsk, Russia.

Praslov, N.D. & A. N. Rogachev eds. 1982. Paleolit Kostenkovsko — Borshchevskogo Raiona
na Donu 1879-1979, Nauka, Leningrad, Russia.

Root, D. 1984. Material Dimensions of Social Inequality in Non-stratified Societies — an Ar-
chacological Perspective. Ph.D. Dissertation. University of Massachusetts, Amherst.

Rybakov, B. A. ed. 1989. Mezolit SSSR. Nauka, Moscow, Russia.

Shovkoplyas, |. G. 1965. Mezinskaia Stoianka. Naukova Dumka, Kiev, Russia.

Soffer, O. 1985. The Upper Paleolithic of the Central Russian Plain, Academic Press, Orlando,
Florida.

—___. 1987. Upper Paleolithic connubia, refugia, and the archaeological record from eastern
Europe Pages 333-348 in O. Soffer, ed., The Pleistocene Old World. Plenum Publishing
Corp, New York.

_. 1993. Upper Paleolithic adaptations in central and eastern Europe and man-mammoth
interactions Pages 31-50 in O. Soffer & N. D. Praslov From Kostenki to Clovis. Plenum
Publishing Corp, New York.

—_.. 1995. Artistic apoggees and biological nadirs: Upper Paleolithic cultural complexity
reconsidered i M. Otte, ed., Nature et Culture. Etudes et Recherches Archéologique de
l'Université de LiPge. Liege, Belgium.

Soffer, O., P. Vandiver, B. Klima & J. Svoboda. 1993. The pyrotechnology of performance art:
Moravaian venuses and wolverines. Pages 259-276 in H. Knecht, A. Pike-Tay, & R. White,
eds., Before Lascaux. CRC Press, Boca Raton, Florida.

Stolyar, A.D. 1985. Proiskhozhdenie lxobrazitel'‘nogo Iskusstva. Iskusstvo, Moscow, Russia.

Svezhentsey, Y.S. 1993. Radiocarbon Chronology for the Upper Paleolithic Sites on the East
European Plain. Pages 23-30 in O. Soffer and N. D. Praslov, eds., From Kostenki to Clovis.
Plenum Publishing Corp, New York.

Svoboda, J. 1986. Mistri Kamenneho Dlata. Panorama, Praha, Czech Republic.

. 1991, Moravia during the Upper Pleniglacial. Pages 193-203 in O. Sotter & C. Gamble,
eds., The World at 18 000 BP. Vol 7. Unwin Hyman, London, U.K.

- ed. 1991. Dolni Vestonice I. Western Slope. E.R.A.U.L. no, 54.

Svoboda, J, 1., Czudek, T., P. Havlicek, V. Lozek, J. Macoun, A. Prychystal, H. Svobodova &
E. Vicek. 1994, Paleolit Moravy a Slezska. Dolnivestonicke studie, svazek 1, AUCR, Brno,
Czech Republic.

Sytmik, A. S. 1983. Gavirovanny) risunok na kosti s Must'erskoj stoiankipod Ternopol'em.
Pages 39-46 in R.S.Vasil'evski, ed..j Plastika i Risunki Drevnikh Kil'tur. Nauka, Novosi-
birsk, Russia.

CENTRAL AND EASTERN EUROPEAN IMAGES 261]

Vandiver, P., O. Soffer & B. Klima. 1990. Venuses and wolverines: The origins of ceramic
technology, ca. 26,000 BP. Pages 13-82 in W. D. Kingery, ed., Ceramics and Civilization,
Vol. V.

Vereshchagin, N. K., & I. 1. Kuz’mina. 1977. Ostatki mlekopitaiushchikhiz paleoliticeskikh
storanok na Donu i verkhnei Desne. Pages 77-110 in A. O. Skarlato, ed., Mamontovaiia
Fauna Russkoi Ravnini. Trudy Zoologicheskogo Insituta LXXII.

Volkov, T. 1913. Nouvelles decouvertes dans la station paleolithique de Mezin. Congres Inter-
national d'Anthropologie et d'Archeologie Prehistoriques, X1V session, Geneve, 1912.
White, R. 1993. Techological and Social Dimensions of “Aurignacian-A ge” Body Ornaments
Across Europ. Pages 277-300 in H. Knecht, A. Pike-Tay & R. White, eds., Before Lascaux.

CRC Press, Boca Raton, Florida.

Section IV

The Interpretive Process

Places of Art: Art and
Archaeology in Context

Silvia Tomaskova

Department of Anthropology
Peabody Museum

Harvard University
Cambridge, MA 02138

On my desk sits a row of volumes describing prehistoric representations, largely
drawn from Paleolithic Europe with supplements from Australia and Africa. To
frame the debates presented in these volumes, I have positioned two works at either
end: Sally Price's slender and eloquent Primitive Art in Civilized Places (1989), and
an issue of the art magazine Portfolio (1979/80), featuring Caillebotte's rainy-day
“Streets of Paris” on the cover and an article on the legacy of Altamira inside. At first
glance it may seem that these two publications would not be immediately germane to
a professional archaeologist. However, I would like to begin my discussion of “con-
text” with the local archaeology of my writing site, recalling the disciplinary location
of the topic at hand, and the theoretical frame through which we sift the evidence. In
the essay that follows I will argue for an understanding of archaeological context that
includes an awareness of the present as well as the past, and of historical geography as
well as the physical environment surrounding our finds. To begin with the assem-
blage of works before me then, is only fitting.

This essay discusses prehistoric representation from the perspective of a lithics
specialist working in East Central Europe, at two site complexes commonly evoked
in debates about “art”: Dolni Vestonice/Pavlov and Willendorf (a detailed evaluation
of the representational material from the sites is provided by Soffer, this volume). I
will suggest that in dealing with prehistoric materials, whether they are figurines or
stone tools, we need to pay close attention to the multiple contexts in which we posi-
tion them. These contexts, both the reconstructed frames of prehistory within ar-
chaeological practice, and the social frames of that research, have their own histories,
uses, and reasons for existence in the interpretive endeavor. The contrast of treatment
between art artifacts and lithic remains illustrates the manner in which archaeological
artifacts are subject to different interpretations depending on analytic context. The
case of Eastern Europe elucidates the importance of past and present geopolitics in ar-
chaeology, and is particularly relevant due to the large number of Paleolithic repre-
sentations found in this region. By focusing on the central importance of context for
archaeological practice, I not only wish to question the adequacy of certain features of

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

266 TOMASKOVA

an art historical approach to prehistoric materials, but also to propose that a greater
awareness of context in archaeological analysis will enhance our comparative meth-
odology, opening new avenues for approaching meaning of artifacts in past societies.

Contexts: Conceptual and Practical

The appearance of an article on Altamira in Poréfolio reflects a central tradition
within archaeological interpretations of the prehistoric record, a tradition that falls di-
rectly within the range of arguments presented in Primitive Art in Civilized Places.
The world of prehistoric art stretches between the murky depths of “primitive art,”
representation produced by traditions outside the scope of history, and the high skies
of “art history,” representation produced by traditions within the outlines of world
civilization (Davis 1993). Prehistoric art thus comprises a field of origin for both the
“primitive” and the “civilized,” providing each side of that lingering binary with a
frame in the past. The roots of Caillebotte cross with those of artists known by cultural
tradition rather than name, and terms drawn from the study of art as well as from the
study of material culture infuse archaeological explanation of prehistoric representa-
tion.

Geographic context plays a curious role in the ordering of both “high” and “low”
representation, at once central and empty. On the one hand we have 19th century
modernist rigor, the language of style (*Gothic,” “Baroque,” efc.) isolating objects
within rigid, and often transcendent categories of place, time and type. On the other
hand we have 20th century postmodern pastiche, consciously mixing materials and
contexts, challenging the notion of authenticity (Figure 1). Yetas Meyers (1994: 680)
and Price (1989) point out, both order and melange occur in distinctively Western lo-
cales. Bringing disparate items of art together into one place, be ita geographic center
(Paris or New York), or an intellectual frame (writing, presentations, or academic de-
bates), does not erase the historical trajectories of these places (Appiah 1991). It re-
mains pertinent to ask where rock paintings, figurines, or body ornaments allow
connection with present day artistic representations of ethnographically observed
people, and where they do not. In which regions of the world do we find the prehis-
toric and the historic united in one interpretive frame? And in which parts of the world
do we find modern terms of high art (such as “galleries,” “altars,” “‘artists”) used to
describe ancient relics? Decades of criticism notwithstanding, analyses of prehistoric
materials continue to use well-traveled routes of analysis, including the one running
between the language of the primitive and that of the civilized (Gamble 1993: 46-47).

Here I will address the issue of “context” on two levels, one conceptual and one
practical. On a conceptual level | question the appropriateness of the term “art” rela-
tive to prehistoric representations, suggesting that the category of art is not only inap-
propriate from an epistemological standpoint but also a hindrance to archaeological
research, due to the conceptual attachments that it has in fields such as art history or
aesthetics. Although some archaeologists who have dealt with prehistoric symbol
systems in the past have considered theories of art (e.g., Bahn & Vertut 1988; Lerot-
Gourhan 1965, 1982: Ucko & Rosenfeld 1967), I maintain that the integration of
theoretical frameworks current in art history at present does not appear particularly
fruitful for archaeological research without a significant modification of concepts
central in both domains. Art history, the disciplinary center for analysis of visual art
(as opposed to domains of creativity or performance), uses some conceptual tools and
theoretical approaches that are applicable to prehistoric materials. However, these

ART AND ARCHAEOLOGY IN CONTEXT 267

eee
{ é S
Noi ads _—

C73 yi

Figure |. Anexample of blurred temporal, stylistic and geographic traditions used in archaological illus -
trations of “art,” which collapse time, space, and ultimately identity under a uniform aesthetic category
Here artifacts from many sites and periods combine into one vast, transcendental ritual. (Illustration by |
K. Townsend of J. Pfeiffer, 1986.)

techniques play a different role in the analysis of representational systems, best exem-
plified by the divergent use of “context” in art aesthetics and in anthropology.

The second contextual theme involves a more literal ordering of the past in the
present: the greater maps of archaeological research. The geographical delineation of
research within contemporary boundaries of nation-states constructs a stage for our

268 TOMASKOVA

prehistoric narratives. A passive acceptance of present political borders frames mod-
els of the past, and limits the number of alternatives that archaeologists suggest in in-
terpreting the archaeological record. Using the case of Central Europe, I will illustrate
how present maps limit our vision of what comparisons are appropriate to make. In
concluding, I suggest that a renewed and reflective attention to analytic problems of
framing may open up contexts other than those to which we have become accus-
tomed.

The Meanings of Art

Although Conkey (1983, 1987, 1993, this volume) and White (1992, 1993) have
repeatedly and convincingly argued against the use of the term “art” in the analysis of
prehistoric representations, this practice persists, either with some qualifications
amid claims of convenience and parsimony (e.g., Talalay, 1993:7: “... art remains a
very convenient term to use...”’), or nonchalantly, as if some general, tacit agreement
about the concept has been accepted (most writings on the subject from the 1970's and
1980's, but also more recently, e.g., Barton et al. 1994). Most scholars who deal with
rock paintings or objects recovered from prehistoric contexts that cannot be explained
in utilitarian terms and are thus categorized as decorative, ritual or symbolic, are
aware of the trap posed by the term “art” (e.g., Bar- Yosef, Soffer this volume). Once
this regrettable circumstance has been noted, however, it is treated as a semantic
problem, something self-evident, marginal, and ultimately inconsequential for the in-
terpretation of prehistoric materials. Yet it is precisely such semantic nuances that re-
veal the outlines of our own cultural typology, and, as Price (1989) reminds us,
apportion value. The terms that we choose have linguistic contexts beyond our texts
that we cannot ignore, 1f we are to move beyond the descriptive stages of antiquarian-
ism and deepen our understanding of representation.

The historical shallowness of “art” as a uniquely aesthetic marker ts not difficult to
demonstrate. Following the English cognate, we find that in the 1 6th century, the term
art not only referred to the liberal arts (grammar, logic, rhetoric, arithmetic, geometry,
music, astronomy) but also was used to describe any skilled person. Only with the
17th century does the more specialized notion of fine art as composed of painting,
sculpture, engraving and drawing appear, but this use did not become dominant until
the 19th century. At this time an artist was contrasted to an artisan —a skilled manual
worker lacking intellectual, imaginative, or creative purpose (Williams 1983:41).
The creative and imaginative became a part of the classificatory vocabulary describ-
ing the separate space inhabited by those involved in art production, the unusual — and
superior — state of mind, body and/or spirit that was unlike that of ordinary people.
This 19th century understanding of creativity as a special separate practice that only
few endowed individuals engage in, separate from profane unimaginative skilled
work, remains with us to this day, and has deeply affected our investigation of prehis-
toric symbolic activities that we label art. This assumption allows the positioning of
the cave paintings as “masterpieces” that are treated separately from any other prehis-
toric production, as the “culmination” of a vast, pan-European artistic tradition last-
ing many millennia (e.g., Bahn & Vertut 1988, although a debate leading away from
this perception has occurred, e.g., Conkey 1984, 1993; Davis 1986 and comments:
the responses to the discovery of both caves Cosquer and Chauvet were a reminder of
the entrenched beliefs about the centrality of the French cave paintings). The histori-
cal track of the term “art” over the past three centuries suggests that it comes hedged
with preconceptions, leaving only a narrow corridor of possible interpretation. The

ART AND ARCHAEOLOGY IN CONTEXT 269

discussion of these issues by Lorblanchet (1992:115-6), nicely illustrates the concep-
tual conflation at stake. In an otherwise critical article, he defends the use of the term
* claiming that its contemporary meaning in French covers all artificial

“

art,”
(nonnatural) things, relying on a La Rousse definition of the term. However, only a
quick look in Le Grande Robert Dictionnaire de la Langue Frangaise shows the
lengthy etymological development from the 13th century meaning of art as “la tech-
nique manuelle” to the 1786 appearance of “arts et métiers,”’ that was only one step re-
moved from the 19th century common usage of “Beaux-arts,” influenced by
“Pesthetique allemande,” as the meaning of the term (Le Grande Robert 1985:568-
70). Yet the example Lorblanchet gives is a list of cultural and aesthetic styles
(“Greek art, Gothic art, Assyrian art, abstract art, primitive art, cave art’). In the
course of this list an older, general sense of craft slides easily into a modern categori-
zation of exceptional creative expression. The field of vision narrows, and the “use-
ful” products of prehistoric craft, such as stone tools, are left aside. Here we might
recall that such restricted senses of art take firmest hold at a point when the lingering
sense of the Greek root fechine that once encompassed seeds of both art and technol-
ogy fades, partly in association with industrialism and the rise of modern forms of sci-
ence (Williams 1983).

In addition to reminding ourselves of the linguistic novelty of our contemporary
definitions of “art,” itis even more important to consider how “context” has been con-
structed and understood in art history as well as archaeology. Although a multiplicity
of theoretical approaches inhabit contemporary art history, depending on the author's
conceptual orientation, subject matter, as well as the historical and geographical fo-
cus of the study, one of the themes frequently discussed is the role of context (Clark
1989; Eaton 1995; Freedberg 1994; Price 1989). Yet the use of the same term in an-
thropology and art history is misleading, since the investigation of context in art his-
tory involves two clear poles of production and consumption (the artist and the
audience) that remain uncertain in prehistoric settings. A brief discussion of the uses
of context in art and archaeology will further clarify the distinctions between them.

Context in Art and Archaeology

Art history addresses works of art through the lens of aesthetics, considering er
ther motive, message, a moment of the artist’s talent, or the (timeless) response that
any piece may evoke in its viewers (Eaton 1995; Freedberg 1994). The art object pro-
vides the means for an exploration of transcendental aesthetic quality, the attribute
that must infuse any such piece to provide it with meaning, and resonates with the sen-
sibilities of the audience. This connection is seen as an emotional response, outside
the rational reasoning of everyday activities. The focus is not on the contextual role of
the object, since the basic aesthetic function of the piece is not in doubt. As Price
(1989:83) reminds us: “For displays presenting objects as art, the implied definition
of what should ‘happen’ between object and viewer is relatively constant; the mu-
seum's visitor's task-pleasure, ... is conceptualized first and foremost as a perceptual-
emotional experience, not a cognitive educational one.” Objects that are considered
ethnographic, not art, /.e., mundane things, receive the explanatory treatment, and
cultural context is deemed relevant for their understanding. Nowhere is this ex-
pressed more strongly than in the distinction between ethnographic objects and art
when discussing “primitive art,” as shown in the 1995 debate over the standing of an
African art exhibit in the London Royal Academy (Riding 1995):

270 TOMASKOVA

Until now, African pottery, wooden carvings and textiles had been viewed essentially as
handicraft because, 1t was argued, the religious, military, sexual or decorative functions
of the works suggested that they had not been created as art, to be appreciated for their
own sake. Even after “primitive” African art inspired Picasso, Brancusi, Braque, Modi-
gliani and Henry Moore earlier this century, it was its magical and mystical quality that
counted most. But at the Royal Academy, objects made by African hands are separated
from their cultural context and can be judged simply as art.

The objects have been elevated to the status of art once they have been cleansed of
their social and cultural context, judged only by the response that they are capable of
evoking. This approach is rooted in a disregard of context, in the anthropological or
archaeological sense, focusing on a “deeper” connection between an art object and
the viewer: the pure context of beauty. Here lies one of the crucial divides between a
discussion of prehistoric art and pure art, as it is understood by traditional art history.
In archaeological research, context is essential for understanding the nature of the
particular prehistoric society in question. An archaeologist views context as the meth-
odological and interpretive basis for understanding the past: “The contextual ap-
proach is based upon the conviction that archaeologists need to examine all possible
aspects of an archaeological culture in order to understand the significance of each
part of it” (Trigger 1989:349).

The separation of an artist and an artisan in discussion of art is not only a semantic
distinction, but also provides entirely different theoretical and conceptual tools with
which products of each are approached. The belief persists that art is produced by
people with some special talent, outside common everyday experience. The emo-
tional response to a work of art is supposed to provide the connection between the
viewer and the artists across intervening cultures. Only on the basis of this assumption
can Nodelman (1979/80:55) write in his discussion of the Altamira cave in Portfolio
that “Interestingly enough, much in the art of the twentieth century — from the Cubist
denial that things can be in only one place at one time, to the later abandonment of the
frame and pedestal and the recent interest of artists in natural processes — presents
striking analogies to certain aspects of the [Paleolithic] system.” Such a statement is
possible only through the suspension of social context, and any role it may have
played in the creation of the Altamira images, as well as Cubist paintings, in favor ofa
belief in transcendental aesthetics. It is clear, if unstated, that a connection is being
made across time but not really across space, as the art of the twentieth century and the
prehistoric paintings are both located in Western Europe, thus assuming a traceable
link between the two, and making it possible to speak of “The Legacy of Altamira”
(Nodelman 1979/80). Archaeology, as it is practiced today, cannot quite comfortably
operationalize such a concept of art, and will be making much greater progress fol-
lowing suggestions of Conkey (1993) and White (1992; this volume, p. 97) in treating
prehistoric representations as artifacts to be investigated in an archaeological context,
rather than appealing to transcendental aesthetic sensibilities ofa Western tradition.

Art often is a category from our own histories and experiences that renders the past com-
fortingly familiar. Although it is certainly justifiable for us to call these images art, in
that they often strike a resonant chord with what we think art is all about, we have to in-
quire into these images in other terms and from other perspectives if we are not merely to
replicate the present (Conkey 1993:115).

ART AND ARCHAEOLOGY IN CONTEXT 271
Maps as Spatial Contexts of Art and Archaeology

The distinction of sacred and profane, art and artifact includes a spatial compo-
nent. The contexts upon which the study of symbolic representations draws, be they
local ethnographies as is the case of the South A frican or Australian rock paintings, or
the lifeways of distant circumpolar peoples sometimes invoked in connection with
Ice Age occupations, involve geographic analogy. In the case of “primitive art,”’ a his-
torical continuity is presumed between past and present on the basis of locale, while in
the case of European prehistory, historical continuity can be bypassed on the basis of
assumed similarity in environmental conditions. The present outside the West func-
tions as the past within it, understood through analogy elsewhere. Under the guise of
neutral geographic interpretation, we find persistent divisions of the world into re-
gions, “worlds,” continents, and cultures. These divisions allow an easier handling of
smaller spatial units, but at the same time carry qualitative markers. The context that
defines a particular region ts selectively created in the present, and weighted with the
historical conventions that surround it.

A specific example central to our discussion will help demonstrate the general
point. On most maps (archaeological as well as otherwise) Europe functions as a
clearly demarcated region, with boundaries separating it from other continents. Yet
the arbitrariness of the eastern boundary of the continent becomes obvious if one
looks at the map of the Eurasian landmass from its geometric center (Figure 2). The
center of modern history (Wolf 1982) is not an obvious center of the prehistoric
globe. Indeed, if one reads back through history one finds that different empires found
different boundaries. But archaeological research (and the institutional apparatus of
funding and regulation that underlies it) operates across a grid of contemporary na-
tional pasts, and their imagined origins. The continuing narrative of prehistoric art,
with its apex in the European (French) Paleolithic, affirms contemporary geography
beneath the cultural frames used to interpret it.

In an encouraging recent development, a number of studies have crossed the es-
tablished boundaries and looked across the border to ‘the other side.’ Soffer’s admira-

Figure 2. Europe or Eurasia: Dolni Vestonice/Pavlov and Willendorf II in Central Europe.

272 TOMASKOVA

ble coedited volumes The World at 18,000 BP (1990), and From Kostenki to Clovis
(1993) unite spaces from Eastern Europe across Asia to the Bering Strait and the New
World. The fact that these studies have appeared only after thaws in the Cold War is a
clear sign of the degree to which modern historical events provide a context in which
archaeology operates. As Gamble notes in his conclusion to Kostenki to Clovis (1993:
313): “We live in interesting political times. Because Paleolithic and Pleistocene ar-
chaeology have always reflected contemporary interests, changes in the political map
have usually met with a geographical shift in the cradle of human origins or the center
from which humanity sprang.” The fairly unhindered exchange of ideas, and the com-
parison of data on sucha scale between West and East would have been quite difficult
to even imagine just ten years ago.

Yet, only a quick look at the regions represented makes it very clear that the past
has been longer and more thoroughly under investigation in some areas than others.
Prehistory, and the human past in general, preoccupy the imagination in the West toa
much higher degree than in any other part of the globe. Among the twenty nine re-
gional articles in the two volumes of The World at 18,000 BP, Europe is represented
by eleven, Asia and Africa by six each, and Australia and the Middle East by three
each. The overrepresentation of Europe in most our discussions of world prehistory is
illustrated by these numbers, particularly if we consider ita western margin ofa much
larger Eurasian landmass, one that has been probably been occupied for quite some
time. The intensity of interest in the past takes a spatial form, represented by the
number of studies devoted to regions within a country, individual countries, larger re-
gions, and entire continents. A closer look at The World at 18,000 BP (1990) reveals a
familiar pattern within the emphasis on Europe: Prehistoric spaces are read through
modern places. France receives a thorough treatment in a discussion of its northern re-
gions (the Paris basin) (Schmider 1990), and the South (Rigaud & Simek 1990).
Spain (Strauss 1990) and Portugal (Zilhao 1990) are presented in separate chapters,
as are Germany (Weniger 1990), Italy (Mussi 1990), and the region of the former
Czechoslovakia known as Moravia (Svoboda 1990). The Balkans (Bailey & Gamble
1990), Central Europe (Kozlowski 1990), and the Russian plain (Soffer 1990) start
the trend of larger regions involving immense stretches of land in Asia (Northern
Eurasia, Central Asia), the entire New World (Frison & Walker 1990), as well as huge
segments of Africa (North Africa, tropical Africa, and South Africa), with the excep-
tion of Zimbabwe, which receives a chapter of its own (Walker 1990).

The “regional approach” that replicates a national treatment of prehistory is not
unique to these two volumes, and in choosing them I do not mean to specifically criti-
cize the authors or suggest some geographic conspiracy on their part. Rather, the reli-
ance on contemporary spatial divisions can be observed in most edited publications of
archaeological materials from Europe, regardless of whether they are data oriented or
focus on theoretical issues (e.g., Clark 1991, Knecht et a/. 1993). The context in
which prehistoric remains are presented 1s drawn from within modern geopolitical
boundaries and these borders are observed, and rarely crossed in explanation. The use
(or absence) of comparison, and the choice of supporting evidence for a claim color
any given scenario of the past, in turn presenting modern social and political relation-
ships as long-lasting and natural. Like different filters on a camera, national perspec-
tives emphasize some shades at the exclusion of others, offering vivid, yet highly
selective images.

Although most North American archaeologists view European traditions as heav-
ily influenced by national histories, and operating within the historical paradigm, they
themselves continue working in national contexts in their fieldwork, and present their

ART AND ARCHAEOLOGY IN CONTEXT 273

data within brackets of modern political borders (e.g., individual entries in Clark
1991: Dibble & Montet-White 1988; Knecht er a/. 1993; Soffer & Gamble 1990). Yet
what is the justification for outlining the Pleistocene record of Germany in a separate
context from that of France or Italy, besides the national tradition of archaeological
practice? Moreover, should we not consider what effect this division has on the im-
ages of the past that we present? It is plausible that in some places differences in the
archaeological record could fall conveniently along contemporary borders, but this
coincidence should be investigated and established rather than assumed.

Social Context and Prehistoric Research

The most explicit statement about the geographical circumscription of archaeo-
logical practice, and its effects on both fieldwork and interpretations of the prehistoric
record can be found in recent discussion focused on the Middle East. Rosen’s (1991)
remarkable demonstration of the confrontation of geopolitical reality and research
traditions highlights the subtlety of the influences, while at the same time reminding
us of the pervasiveness of such basic factors as geographic location and history:

Our biases are more in the realms of geographical interpretations, emphases on particu-
lar data sets, and in the semantic nuances of professional jargon. In some ways, these
“minor” tendencies are more problematic than the cruder tendencies ... since they are
not so obvious. The definition of discrete industries along a modern national border,
when it coincides with an apparently geographic one, may be legitimate. The characteri-
zation of a specific region as a “culture core” area, and an outlying regions as “periph-
eral”, might be justifiable. However, in both cases we must pause to consider the
authors’ assumption in their definitions and characterizations, and then consider our
own in evaluating theirs. The data themselves are only rarely unambiguous (Rosen
1991:308).

Following Rosen we begin to see how the shape of interpretive patterns in archae-
ology can rest on differences located beyond individual viewpoint: In intellectual tra-
ditions or schools of thought linked to training in a particular area, distinctions
between programs in North America or Britain, say, or between national approaches
in different European countries that can be recognized in patterns of writing, focus, or
research questions. Even more crudely, the simple communication of ideas, contact
between researchers, visits to field sites, and examination of archaeological materials
are bounded by contemporary considerations, some of them overtly political on a
macro scale. In the Levant this results in a complete lack of access to Syrian collec-
tions by Israeli scholars, and vice versa (Rosen 1990:315-16). Foreign relations of
countries in the region are projected onto the prehistoric materials, leaving us to won-
der how complete or incomplete our record of Pleistocene occupations, interactions,
and movements might be, and to what degree this skews the debates related to the nu-
merous “origins” questions that are posited within this particular area.

Our consciousness of the recent troubled history of the middle east should not mis-
lead us into considering this particular area to be an exception or as an aberration of
the normal functioning ofa society and science. Rather than an anomaly, I would sug-
gest that the Middle East is better understood as a window, a place where assumptions
lie exposed on the surface, and deeper patterns are revealed. French, Czech, or even
American archaeologists do not lead lives that would often get press coverage; never-
theless they too are subjected to more subtle but equally entrenched social and politi-
cal boundaries that they observe, even without the immediacy of a struggle for

274 TOMASKOVA

national borders. The weight of historical traditions shapes the frame of research, in-
fluencing both subject matter and context.

Maps and Contexts of Central/Eastern Europe

The history of archaeology in Central and Eastern Europe provides us with a pole
of comparison only one step removed from the Israeli situation. Itis worth describing
to illustrate that what lies at stake is not only a theoretical paradigm affected by the so-
ciopolitical context (a frequent assumption about bias in science in nonwestern socie-
ties), but also the very practice of archaeology: fieldwork, testing, sampling, and all of
the usual steps of “objective” science. Research inevitably has to operate within the
limits of modern concerns, filtering the past through the present.

Until the 1990's Eastern Europe in general presented a challenge for any interna-
tional collaboration in scientific projects, particularly in such a sensitive area as ar-
chaeology (Kohl 1992; Milisauskas 1990; Soffer 1983). The unavoidable
consequence was an alignment of the researchers in the region in the direction of the
East and the Soviet Union, and even that alignment was often strained by the fact of
political domination. The result of this political context was isolation from Western
theoretical developments, and a barrier against incorporating Western archaeological
data into comparisons beyond a local scale (Figure 3). A comparison of materials
across the border to the West was severely restricted, and consequently most models
focused on hypotheses that could be empirically tested within the accessible field.
Thus studies addressing exchange of raw materials examined local sources, or those
in the neighboring countries that could be visited. The inductive nature of the research
led to a design of hypotheses about prehistoric contacts on the basis of the empirical
evidence obtained through limited and local scientific interaction. The modern politi-
cal context thus translated into the prehistoric social context, explained through an
entirely legitimate and thorough scientific research. It would be a mistake to imagine
that research in Eastern Europe proceeded as a monolithic Stalinist version of science
subjected to crude political aims (Fortescue 1990). Rather, the unintended conse-
quences of the political system, and adjustments to the limited range of possibilities
became normalized, internalized, and used to the utmost. The resulting structures
were both solid and limited, and the prehistoric models developed within them for this
particular region share these characteristics. To better illustrate the ways that Central
European maps of past and present geographies have intertwined, I will present the
area with which Iam most familiar, that of Moravia, and the shifting northern borders
of the former Austro-Hungarian Empire.

Moravian “Art” Sites in Archaeological Context

The middle section of the former Czechoslovakia, the region known as Moravia
lies between the Czech region of Bohemia (to the West), Poland (to the North), Slova-
kia (to the East) and Austria (to the South). Distances to the German and Hungarian
borders (north/west and south/east) are also not great (Figure 3). Given this patch-
work of nearby political borders, it is noteworthy that discussions of Moravian sites
tend to operate in two disparate contexts, one local and one macroregional, where the
settlements are considered to have been the westernmost outposts of the eastern set-
tlement system (¢.g., Klima 1963; Soffer 1987, 1993; Svoboda 1990, 1991). The Up-
per Palaeolithic settkement is presented as local and independent, with a cultural

ART AND ARCHAEOLOGY IN CONTEXT

to
~
Nn

GERMANY ae POLAND

CZECH REPUBLIC

dv/p #3
mete SLOVAKIA

oO + - ms as

AUSTRIA
HUNGARY

Figure 3, Central Europe: Dolni Vestonice/Pavlov and Willendorf II in Europe 1992

tradition of its own: the “Pavlovian” (Klima 1963, 1983; Svoboda 1991, 1993; Va-
loch 1980). This “archaeological culture,” defined on the basis of stone tool typology,
is claimed to be an indigenous development, the only suggested connection being the
much later Kostenki sites in the Don basin, possibly derived from the Moravian sites
through eastward migration (Klima 1963, 1983; Kozlowski 1985; Soffer 1985).
These particular claims were first made in the Czech context of the early 1960's. They
were received at the time ina rather reserved fashion on the Russian side of things,
where local (Eastern) origins were then an equally favored explanation (Dolukhanov
1982: Boriskovskil 1980).

Yet if we consider other directions from Moravia besides the distant east, it be-
comes obvious that other choices for comparison are available, including ones much
closer at hand. The Paleolithic occupation of Western Slovakia along the Vah River,
for example, lies a minimal distance eastward from the neighboring Pavlov region.
The site at Moravany has shown a great promise with a wealth of materials, including
an ivory female figurine, drilled animal teeth and shells, fragments of clay figurines in
a settlement ofa substantial size (Barta 1965). Yet it 1s little known in the West partly
due to slow progress in excavation, analysis, and publication of even preliminary re-
sults, and partly through the lack of its incorporation into previous maps of Paleolithic
Europe. That archaeologists working in the neighboring Moravian region did not
consider interactions or connections with this site is worthy of greater note. The same
can be stated for Paleolithic settlements down river along the Danube in neighboring
Hungary, which are treated yet again as separate entities with an unspecified relation-
ship to other sites in the region: “The sites of the Danube Bend form a topographical,
chronological and ecological unit within the Upper Paleolithic settlements of Hun-
gary” (Dobosi 1991:197). In both cases, Slovakia and Hungary, the lack of compari-
son with the Moravian sites is not a result ofa lack of awareness of their existence or
unfamiliarity with the recovered materials, but rather an excessive stress on local de-
velopment, and, in the case of Dolni Vestonice or Predmosti, an implicit sense of su-
periority due to the wealth of the prehistoric artifacts recovered there since the 1920’s

276 TOMASKOVA

(see explicit and implicit statements to that effect ine.g., Absolon 1929, 1938: Klima
1963: Svoboda 1991, 1994). Recently Montet-White has remarked that **... as re-
search continues to expand in the neighboring areas, the Gravettian of the Pavlov hills
no longer appears as the only significant cultural manifestation in the region between
28,000 and 20,000 BP.” (Montet-White 1994:483). I would argue that the promi-
nence of the Moravian sites was not necessarily due to a /ack of other archaeological
data from the region, but rather due to a comparative frame of local development that
stressed indigenous cultures with minimal contacts in the wider region. As the Mo-
ravian sites produced a substantial amount of prehistoric materials, the implicit isola-
tion of the groups occupying them fit comfortably with the present. The habit of
staying at home appeared to have had a long tradition, and prehistory justified the lack
of contacts in the modern political context.

If close eastern and southeastern localities were not incorporated into the greater
Moravian context due to a lack of sufficient, or equally spectacular data as in the Pav-
lovian, we must still consider why comparisons were not made to the west and south-
west, and the influential sites of Lower Austria. A number of settlements in the same
ecological zone of the Central European Basin (which includes both modern Lower
Austria and the Moravian Plateau) lie geographically close to Pavlov and Dolni Ve-
stonice (less than 100 km away), including some that have been well known interna-
tionally since the turn of the century (e.g, Aggsbach, Krems, Willendorf) as well as
others actively worked in the 1980's and 1990's (e.g., Alberndorf, Grubgraben, Ro-
senburg, Stillfried, Stratzing) (Felgenhauer 1956; Otte 1981; Montet-White 1994;
Neugebauer-Maresch 1993). Here we find a more complex interplay of historical and
political contexts, involving archaeological traditions of typology, greater national
traditions of prehistoric research, and the tensions of the Cold War frontier severing
Austria from Czechoslovakia and Hungary (Tomaskova 1995). The Austrian sites
have been understood differently, and their materials treated differently. from the mo-
ment of their excavation until the very recent past.

In general accounts of the Paleolithic occupation of Europe, Willendorf occupies
a prominent place. though only a basic analysis and typological categorization of its
materials have been carried out (Broglio & Laplace 1966; Felgenhauer 1959; Otte
1981). An extensive stratigraphic sequence with dates for several layers, and a rela-
tively detailed description of the recovered materials did not lead to further contem-
plation of the place of Willendorf in relation to the Moravian sites. Rather, Willendorf
would seem to exemplify Soffer’s (1993:45) remark: “All of these scenarios, how-
ever, considered each site or even cultural layer as a total universe; one that can only
be related to others through chronological genetic ties” (Figure 4).

In the 1990’s the posited connection of Pavlov — Avdeevo — Kostenki extended to
include also Willendorf, and a new “cultural unity”, Willendorf— Pavlov—Avdeevo—
Kostenki, emerged in some writings (Soffer 1993:44: Grigorev 1993:5 1-52: Svoboda
1994). Yet none of these new studies specify why this “cultural” relationship has not
been considered before, or why it is being proposed at this point in time, 1f no addi-
tional analysis of the Willendorf material appeared, except for the reexamination of
the stratigraphic sequence (Haesaerts 1990). Previous regional focus preserved the
division by modern countries. and adhered to the division of Western and Eastern
Europe. with the uncertain position of Austria — a “Western” country located geo-
graphically more to the east than a large segment of the former Czechoslovakia — un-
questioned (Otte 1981: Soffer 1985). In the 1990's the borders in the East became
more porous, and Willendorf was “reunited” with its eastern neighbors, amid the dis-
covery of the recent Paleolithic sites in Austria (Montet-White 1994: Soffer 1993).

ART AND ARCHAEOLOGY IN CONTEXT 277

ALTSTEINZEIT

IM OSTEN OSTERREICHS

Ficure 4. Willendorf and Galgenberg statues in a world of mammoths and art (illustration by A. Schuma-
cher, cover page in C, Neugebauer-Maresch, 1993)

However, Willendorf was incorporated without being compared in any systematic
fashion, the national tradition lurking under the region.

I emphasize the modern, historical, geopolitical context of the research not as a ne-
gation of possible links between sites in the region, but as a cautionary note, a re-
minder that similarities and comparisons are made at present and are selective. It 1s
plausible that there are major discernible differences between sites that are separated
by a modern border, such as Willendorf and Pavlov, just as it is equally plausible that
they were occupied by groups that could have had some cultural ties, but either propo-
sition needs to be examined rather than accepted without any support by the archaeo-
logical evidence. As Rosen stresses (1991:316):

278 TOMASKOVA

A research proposal to conduct a regional analysis in effect identifies the region before-
hand on the basis of modern perceptions, before the credibility of the notion can be
tested against data. Once the assumptions are made, testing them is all the more difficult.
Obviously, amajor source of this sort of sampling bias is modern political boundaries.

The Pavlov/Willendorf Divide

As arecent study has shown, the Willendorf II female figurine has been he single
most commonly used Paleolithic image in textbooks and popular discussions of pre-
historic art, closely followed by the second most common representation, the female
figurine from Dolni Vestonice (Conkey & Tringham 1995). Considering the shared
context in which at least the “art” material from these sites has been presented, the
lack of scientific comparison between the sites is a genuine puzzle. This silence was
even more startling considering the fame and length of excavation at each site (e.g.,
Absolon 1925; Felgenhauer 1959; Klima 1963; Svoboda 1994). Only in general dis-
cussions of Upper Paleolithic sites in Europe did Willendorf Il and Dolni Vestonice
appear together, and such work placed them in the general European context without
necessarily pointing out their proximity (see general works on the European Paleo-
lithic, e.g., Gamble 1986; Otte 1981).

Between 1990-1994 I conducted research across the border between the former
Czechoslovakia (at Dolni Vestonice) and Austria (at Willendorf). The starting point
of my investigation was the puzzle presented by the striking absence in the archaeo-
logical literature of any detailed comparison between Willendorf II and Dolni Ve-
stonice/Pavlov, despite their geographical and temporal proximity. My initial
reaction was that this must be a simple case of a modern boundary projected into the
past, the border between Austria and Czechoslovakia (and simultaneously the
post-W.W. II border between the West and the East), overshadowing the study of pre-
history. It appeared that all that needed to be done was to examine the archaeological
materials from both sites, and look for similarities. The extended process of compari
son, however, revealed the complexities involved.

Selecting the Gravettian sequence from Willendorf IH (layers 5-9) that was tempo-
rally in the same range as the complex of sites in DoIni Vestonice and Pavlov (30, 000
- 25,000 BP), [compared the lithic artifacts (over 8000 pieces in all) from both locali-
ties in terms of typology and use-wear (Tomaskova 1995). Rather than accepting the
analytical units of the traditional typology, the study examined the consistency in
which these typologies were applied within and between the collections from both lo-
cations, so as to evaluate the classificatory process itself, before making any judg-
ment in terms of possible prehistoric differences or similarities. It was my intention to
show that the process of categorization itself, particularly when dealing with such a
cultural construct as the past, is complicated by a number of factors that do not sepa-
rate smoothly, and whose influence, even when identified, may not be so effortlessly
corrected. Upon close examination it became apparent that the use of the standard
Gravettian typology varied not only within the collections from Willendorf and Pav-
lov, but even more between them (Tomaskova 1995).

A subsequent use-wear study provided a different analytic field for the same mate-
rials, and shifted from the more common studies of production of stone tools to their
utility in consumption. The results of the research circumvented the simple distinc-
tion of sameness or difference, and provided information that, in combination with
other lines of evidence, opened new possible ways of looking at prehistory in the re-
gion. Short-and-long term occupations associated with various seasonally deter-
mined activities could be suggested at both locations, skirting the question whether

ART AND ARCHAEOLOGY IN CONTEXT 279

this would result in culturally perceived distinctions. However, at the same time the
examination of the historical context of research in the area, and its influence upon the
construction of prehistoric geographical regions, also supplied evidence of an unset-
tling range of layers of social contexts within which the prehistoric materials were
also lodged. In suggesting a new focus on consumption and actual use rather than
lithic form and perceived function, | would simultaneously call for increased scrutiny
of the social and political fields through which we operate, and out of which our disci-
plinary traditions flow.

Contexts and Comparisons

Although the starting point of my study has been questioning claims of difference
between Willendorf and Dolni Vestonice/Pavloy, at the end of it I did not find myself
in a position to advocate clear claims of similarity, difference or identity. Rather, one
of the issues raised in the examination of the artifacts and the histories of the two sites,
was precisely the problem of comparison as a case of polarization: the notion that phe-
nomena, in this case archaeological sites and by extension cultures, can only be either
of the same kind, or different. Once the question is posed in this fashion (“Are they
different or the same?) we are forced to eliminate all uncertainty, all ambiguous
cases, all gray areas in between, facing only options ofa striking similarity or a radical
difference. In this way they are like regions, bounded by fixed borders that do not al-
low crossing, or any state of in-betweenness and which are both posited and reaf-
firmed in analyses. The very categories that we use to contextualize individual
objects limit our vision.

The difficult role of comparison in anthropology is, of course, not a new Issue, as a
reading of such an ancestral figure as Franz Boas would attest (Boas 1948). His call
for the use of historical rather than linear comparative methodology made little head-
way beyond cultural anthropology, and direct comparison remained a basic tool of ar-
chaeological method (Eggan 1953; Hammel 1983). The sporadic archaeological
record for a particular temporal sequence seemed a sufficient justification for com-
parisons over vast geographical distances (e.g., Gamble 1982; Otte 1981; Soffer
1985). A shortage of sufficient data in one area becomes a basis for the incorporation
of material from any other area, related geographically, temporally, or ecologically
(Wylie 1982).

Yet a closer inspection of published analyses reveals that the comparison of the
prehistoric record is also its very construction. Choosing which areas to compare, and
which to ignore (e.g., in Germany, Hahn 1983; Weniger 1989), constructs a unique
cultural sequence on a territory located between neighboring areas with a generous
prehistoric record. Thus my initial question— why were these sites not compared —is
partly answered by showing that before any answer can be provided, we need to look
at the manner in which comparisons have been conducted, the frame of chosen re-
search questions and the methods applied. Before considering interpretations that are
offered for any archaeological pattern, we need to step back and examine what is be-
ing compared, and what were the options for the study. The comparative method is a
highly selective tool that involves choices on the part of the researcher, and can al-
ways serve more than one purpose, whether these are conscious or not. The presence
of data in any one area does not automatically lead to placing them into the same con-
text with that of an adjacent area, as it is viewed through a modern day lens, and de-
pends as much on the way of seeing, as on the existence of the evidence itself.

280 TOMASKOVA

If archaeological regions are circumscribed by modern political entities, and proj-
ect prehistoric versions of national pasts, it is essential to examine the relationships
that are suggested in such a Paleolithic European community. The presence of ar-
chaeological remains in most parts of Europe (beyond the extreme northern area)
opens a possibility for testing models of interaction at the local, midrange, as well as
long distance level.

This question has usually been investigated within the scope of research address-
ing diffusion, and as Clark notes: “Archaeological studies of diffusion usually as-
sume that interaction among groups in close proximity to one another is not only
likely but inevitable (Clark 1994:321).” In a detailed analysis he shows that an as-
sumption of diffusion and/or migration in European Paleolithic archaeology is fre-
quently taken as an explanation in the subsequent interpretation of archaeological
materials (Clark 1994). While it is true that in some trends of European Paleolithic ar-
chaeology migration and diffusion tend to be the favorite explanatory paradigm (e.g.,
Kozlowski 1992; Otte 1981; Otte & Keeley 1990), I would hesitate to polarize the
models of prehistoric occupation in Europe in terms of either migration or local inde-
pendent adaptation, as Clark seems to suggest (Clark 1994). Rather, | would argue
that both models are and have been used in European Palaeolithic research, but very
selectively, and that significant differences can be detected in national research tradi-
tions at different periods. Thus at any given point the role of the present context, that
in which the archaeological materials are uncovered or interpreted, cannot be ig-
nored.

Anexamination of current models of larger mobility trends in the European Paleo-
lithic brings us back to the interplay between this suggested mechanism of culture
contact, and the aforementioned sociopolitical context of geography (e.g., Gamble
1986, 1993; Montet-White & Holen 1991; Soffer 1987; Soffer & Gamble 1990).
While a substantial Paleolithic record is available in Italy (e.g., Palma di Cesnola
1993), or Greece (e.g., Jacobson 1990; Perles 1989), most migration models rely on
the East-West trajectory (e.g., Gamble 1983; Kozlowski 1992; Otte & Keeley 1990,
Softer 1987, 1990), leaving the southern regions to their own fate. Whether one ac-
cepts the model of the western refugium (Jochim 1987), or the eastern version of the
same (Soffer & Praslov 1993), it appears that the only route considered was between
these two points. While the presence of the ice sheets during the Pleistocene affected
the appearance of the map of Europe, it did not isolate any one part entirely, particu-
larly not in the South (for a detailed discussion see Velichko & Gabori 1990).

On one hand it may be fairly easy to dismiss migration and diffusion as outdated,
suggesting a certain degree of naiveté, as well as hints of expansive national chauvin-
ism along colonial lines (e.g., Renfrew 1989). Yet on the other hand, it is worthwhile
to consider models of culture contact and interaction, that cross the local borders and
suggest the possibility of social exchange on a larger than a local scale. The material
evidence of exotic items. whether they were shells, mollusks, or raw materials, as
early as the Aurignacian (e.g, Krems-Hundsteig, Tibava), not to mention the later
Gravettian and Magdalenian sites, does suggest either access to distant places, or an
exchange over a long distance (Hahn 1977). While social contacts may be chronically
difficult to “prove,” especially among Paleolithic populations, they should not for
that reason alone be utterly eliminated from models of prehistoric behavior (e.g.,
Rouse 1986). In comparison to some European propensities, North American insis-
tence on local adaptation, low population density, and consequent lack of interaction
between groups may appear to be a neutral model (Clark 1994:330), but its strong fo-
cus on rugged individualism and survival bears striking resemblance to contempo-

ART AND ARCHAEOLOGY IN CONTEXT 281

rary American ideology, in addition to resting on an illusion of certain boundaries
maintained over enormous stretches of time. We should never lose sight of how par-
tial and imperfect our evidence Is.

Conclusion: Recognizing Context

In approaching prehistoric representation, then, an increased awareness of
context brings conceptual anachronisms to light. It also opens the possibility of coun-
tering our assumptions by working around them. When viewed as “art,” the figurines
of Willendorf and Dolni Vestonice lift effortlessly out of their local settings into a
timeless pan-European tradition of symbolic expression. Associated “nonart” arti-
facts are left behind, and enter the archaeological record in sites separated by national
research traditions. In the first case transcendental aesthetic criteria obscure issues of
place, creating an analytic context inappropriate to archaeological analysis (Figure
5). In the second case the social realities of archaeological practice project contempo-
rary conventions onto the past. All too often, the practitioners of archaeology repro-
duce the maps through which they were taught, retaining political borders even when
transgressing them. Such biases of context cannot be simply eliminated by abstract
models or some massive program of random sampling. Beyond the naive impractical-
ity of sucha vision (perhaps only imaginable within the comfortable amnesia of post-
war North America), the shifting conditions of physical as well as political geography
and the continual necessity of framing information render it untenable. Contextuali-
zation is an inevitable element of any analytic act. Yet by increasing our awareness of
the contexts in which we operate, as well as considering paths not chosen, we can be-
gin to consider a wider range of comparison, including pieces of “art” with artifacts,

Figure 5. Willendorf Venus Statue by site: Art blown out of context (photograph by O. Soffer)

282 TOMASKOVA

and relating sites across national borders. The maps we live and think through cannot
be erased, but they can be rearranged.

Acknowledgments

| would like to thank the editors of this volume for the opportunity to present my
work and opinions in this volume. Iam especially grateful to Meg Conkey for her pa-
tience in editing my prose and to Olga Soffer for her generosity in providing the pho-
tograph of the Willendorf monument. The organizers and participants of the Wattis
symposium of 1995, especially Meg Conkey and Randy White, provided an encour-
aging environment in which a different version of this essay was tried out. Heather
Price, Peter Redfield and Barbara Voytek were of great assistance and support in the
preparation of this revised manuscript.

Literature Cited

Absolon, K. 1925. A discovery as wonderful as that of Tutenkhamens Tomb. Moravia over
20,000 years ago. in The Illustrated London News

—. 1927. Une Nouvelle et Importante Station Aurignacienne en Moravie. Pages 73-88 in
Revue Anthropologique v. 27.

1928. Uber die groke Aurignac-Station bei Unter-Wisternitz in Mahren. Pages 57-61 in
Tagungsherichte der deutschen Anthropologischen Gesellschaft, Bericht tiber die 49, Ver-
sammlung in Kéln, Leipzig, Germany.

—. 1935. Ein Anhangsel aus einem fossilen Menschenzahn. Zeitschrift fiir Rassenkunde
1:317

1938, Vezkwn diluvialni stanice loveu mamutu v Dolnich Vestonicich na Pavlovskych
kopcich na Morave. Pracovni zprava za prvni rok 1924, Palaeoethnologicka Serie c. 4.
Brno, Czech Republic.

Absolon, K. & B. Klima 1977. Predmosti, ein Mammutjdgerplatz in Madhren. Academia,
Praha, Czech Republic.

Albreeht, G. & H. Berke, 1988. The Bruder Valley near Engen/Hegau: Varying land use in the
Magdalenian. Pages 465-473 in M. Otte, ed., De la Loire a l'Oder. Actes du Colloque de
Li¢ge, décembre 1985, Volume ii. BAR International Series 444(11).

Appiah, K. A. 1991. Is the post- in postmodernism the post- in postcolonial? Critiqual Inquiry
17:336-357.

Arnold, B. 1990, The past as propaganda: totalitarian archaeology in Nazi Germany. Antiquity
64:464-478

Bahn, P. G. 1982. Intersite and interregional links during the Upper Palaeolithic: The Pyrenean
evidence. Oxford J. Arch. 1:247-268.

Bahn, P.G. & J. Vertut. 1988. /mages of Ice Age. Windward Press, London, U.K.

Bailey, G. ed. 1983. Hunter-gatherer Economy in Prehistory: A European Perspective. Cam-
bridge University Press, Cambridge, U.K.

Barta, J. 1965. Slovensko v Starsej a Strednej Dobe Kamennej. Vydavatelstvo Slovenske}
Akadémie Vied, Bratislava, Slovak Republic.

_ 1987. Problems of dwelling structures in the Palaeolithic of Western Slovakia. dnthro-

pologie 25:105-110.

1989. Trencianske Bohuslavice, a Gravettian settlement in western Slovakia. Anthro-

pologie 93:173-182.

ART AND ARCHAEOLOGY IN CONTEXT 283

Barton, M.C.,G. A. Clark & A. E. Cohen. 1994. Artas information: explaining Upper Palaeo-
lithic art in western Europe. World Arch. 26:185-206.

Boas, F. 1948. Race, Language and Culture. Macmillan, New Y ork.

Bosinski, G. 1982. The transition Lower/Middle Palaeolithic in north-west Germany. Pages
165-75 in A. Ronen, ed., The Transition from Lower to Middle Palaeolithic and the Origin
of Modern Man. BAR International Series S151, Oxford, U.K.

. 1988. Upper and Final Palaeolithic settlement patterns in the Rhineland. /n H. Dibble
and A. Montet-White, eds., Upper Pleistocene Prehistory of Western Eurasia. University
of Pennsylvania, Philadelphia.

Broglio, A. & G. Laplace. 1966. Etudes de typologie analytique des complexes leptolithiques
de I'Europe centrale. Les complexes aurignacoides de la Basse Autriche. Rivista di Scienze
Preistoriche. 2\:61-121.

Champion, T., C. Gamble, S. Shennan & A. Whittle. 1984. Prehistoric Europe. Academic
Press, London, U.K.

Chernysh, O. P. & V. P. Savich, 1987. Arkheologiia Prikarpatia, Volyni i Zakarpatia ; Kamen-
nyi Vek. Nauk. Dumka, Kiev, Ukraine.

Chmielewski, W. 1972. The continuity and discontinuity of the evolution of archaeological
cultures in Central and Eastern Europe between the 55th and 25th millenaries BC. Pages
173-9 in Bordes, F. ed., Origine del'Homme Moderne. Unesco, Paris, France.

Clark, G. A. ed. 1991. Perspectives on the Past. University of Pennsylvania Press, Philadel-
phia.

. 1994. Migration as an explanatory concept in Paleolithic archaeology. J. Archaeol
Method Theory. \/4: 305-343.

Conkey, M. W. 1984. To find ourselves: Art and social geography of prehistoric hunter gath er-
ers. Pages 253-272 in C. Schrire, ed., Past and Present in Hunter Gatherer Studies. Aca-
demic Press, New York.

. 1987. Interpretive problems in hunter-gatherer regional studies: The case of the Euro-
pean Upper Palaeolithic. Pages 63-77 in O. Soffer, ed., The Pleistocene Old World: Re-
gional Perspectives. Plenum Press, New York.

. 1989. The place of material culture studies in contemporary anthropology. Pages 13-31

in A.L. Hedlund, ed., Perspectives on Anthropological Collections from the American

Southwest. Arizona State University Anthropological Research papers No 40.

. 1993. Humans as materialists and symbolists: Image making in the Upper Paleolithic.
Pages 95-118 in D. T. Rasmussen, ed., The Origin and Evolution of Humans and Human-
ness. Jones and Bartlett Publishers, Boston.

Conkey, M. W. & R. E. Tringham. 1995. Archaeology and the Goddess: Exploring the con -
tours of feminist archaeology. Pages 199-247 in D. C. Stanton & A. J. Stewart, eds.,
Feminisms in Academy. University of Michigan Press, Ann Arbor.

Conkey, M. W. & S.H. Williams. 1991. Original narratives: The political economy of gender
in archaeology. Pages 102-139 in M. Leonardo di, ed., Gender at the Crossroads of Know!-
edge: Feminist Anthropology in the Postmodern Era. University of California Press, Ber-
keley.

Davis, W. 1986. The origin of image making. Curr. Anthropol. 27:193-215.

. 1987, Replication and depiction in paleolithic art. Representations 19:11-144.

. 1989. Finding symbols in history. Pages 179-189 in Morphy, H. ed., Animals into Art.
Unwin Hyman, London, U.K.

Davis, W. 1993. Beginning the history of art. J. Aesthet. Art Crit. 51/2:327-350.

Dolukhanov, P. M. 1982. Upper Pleistocene and Holocene cultures of the Russian Plain and
Caucasus: ecology, economy, and settlement patterns. Pages 323-358 in F. Wendorf &
A.E. Close, eds., Advances in World Archaeology, vol. |. Academic Press, New York.

284 TOMASKOVA

____. 1986. Natural environment and the Holocene settlement pattern in the north-western
part of the USSR. Fennoscandia Archaeologica 3:3-16.

Dobosi, V. T. 199]. Economy and raw material. Pages 197-203 in A. Montet-White & S,
Holen, eds., Raw Material Economies among Prehistoric Hunter-Gatherers. University of
Kansas, Lawrence, KS.

Feblot-Augustins, J. 1993. Mobility strategies in the Late Middle Palaeolithic of central
Europe and western Europe: elements of stability and vaniability. J. Anthrop. Arch
12:21 1-265.

Felgenhauer, F. 1956-59. Willendorf in der Wachau. Mitteilungen der Prahistorische Kom-
mission der Osterreichische Akademie der Wissenschaft 4.

Fortescue, S. 1987. The Communist Party and Soviet Science. John Hopkins University Press,
Baltimore.

Fowler, D. D. 1987. Uses of the past: Archaeology in the service of the state, Amer. Antiquity

Freedberg, D. 1994. Context, visuality and the objects of art history. Art Bulletin 76:394-397,

Gamble, C. S. 1979. Hunting strategies in the central European palaeolithic. Proc. Prehist. Soc.
45:35-52.

—__. 1982. Interaction and alliance in Paleolithic society. Man. 17:92-107.

1984. Regional variation in hunter-gatherer strategy in the Upper Pleistocene of
Europe. Pages 237-260 in R. Foley, ed., Hominid Evolution and Community Ecology. Aca-
demic Press, New York.

_ 1986. The Palaeolithic Settlement of Europe. Cambridge University Press, Cambridge,
Wi:

—__. 1987. Archaeology, geography and time. Progr. Hum. Geog. 11:227-246.

Grande, R. 1985. Dictionnaire de la Langue Frangaise. 2nd edition. Le Robert, Paris, France.

Haesaerts, P. 1990, Nouvelles Recherches au gisement de Willendorf (Basse Autriche). Bulle-
tin van het Koninklijk Belgisch Insitut voor Naturrwetenschappen, Aardwetenschappen,
60:203-218.

Hahn, J. 1972a. Das Aurignacien in Mittel- und Osteuropa., Acta Praehist. et Archaeol. 3:77-
107.

—. 1972b. Aurignacian signs, pendants and art objects in central and eastern Europe.
World Arch. 3:252-66.

—. 1977. Aurignacien: Sas dltere Jungpaldolithikum in Mittel- und Osteuropa. Funda-
menta Reihe A/9, KéIn, Germany.

—. 1986. Kraft und Aggression: Die Botschaft der Eiszeitkunst im Aurignacien Stid-
deutschlands? Verlag Archaeologica Venatoria, Tubingen, Germany.

—__. 1987. Aurignacian and Gravettian settlement patterns in Central Europe. Pages 251-
261 in O. Softer, ed., The Pleistocene Old World. Regional Perspectives. Plenum Press,
New York.

Jochim, M. A. 1976. Hunter-gatherer Subsistence and Settlement: A Predictive Model. Aca-
demic Press, New York.

—__. 1991. Archaeology as long-term ethnography. Amer. Anthropol. 93:308-321.

Klima, B. 1963. Dolni Vestonice. CSAV, Praha, Czech Republic

_ 1976. Die palaolithische station Pavlov II. /n Prirodovedné prace ustavu CSA} v Brne
X/4. Academia, Praha, Czech Republic.

Knecht, H., A. Pike-Tay, & R. White, eds. 1993, Before Lascaux: The Complex Record of the
Early Upper Palaeolithic. CRC Press, Boca Raton, Florida.

Kozlowski, J. K. 1973. The origin of lithic raw materials used in the Palaeolithic of the Carpa-
thian countries. Acta Archaeol. Carpathica 13:5-19.

ART AND ARCHAEOLOGY IN CONTEXT 285

. 1977. Prehistoric settlement in the northern part of central Europe in the light of palaeo-
geographic conditions prevailing during the earlier phase of the Wiirm Pleniglacial. Folia
Quater. 49:5-14.

. 1986. The Gravettian in Central and Eastern Europe. Adv. World Arch. 5:131-200

. 1990a. A multispectral approach to the origins of the Upper Palaeolithic in Europe.
Pages 419-437 in P. Mellars, ed., The Emergence of Modern Humans. Cornell University
Press, Ithaca.

. 1990b. Northern Central Europe c. 18,000 BP. Pages 204-227 in O. Soffer and C. Gam -
ble, eds., The World at 18 000 BP. Vol. I: High Altitudes. Unwin Hyman, London, U.K.

. 199 1a. Paleolithic quarry sites. Pages 1-6 in A. Montet-White & S. Holen, eds., Raw
Material Economies among Prehistoric Hunter-gatherers. University of Kansas, Law-
rence, Kansas.

. 1991b. Raw material procurement in the Upper Paleolithic of central Europe. Pages
187-196 in A. Montet-White & S. Holen, eds., Raw Material Economies among Prehis-
toric Hunter-gatherers. University of Kansas, Lawrence, Kansas.

_ ed., 1976. L'Aurignacien en Europe. U.1.S.P.P., IX Congrés, Colloque XVI, Nice,
France.

Leroi-Gourhan, A. 1965. Treasures of Prehistoric Art. Harry Abrams, New Y ork.

. 1982. The Dawn of European Art. Cambridge University Press, Cambridge, U.K.

Longino, H. 1990. Science as Social Knowledge: Values and Objectivity in Scientific Inquiry.
Princeton University Press, Princeton.

Lorblanchet, M. 1992. Le triomphe du natiralisme dans l'art paleolithique. Pages 115-135 in T.
Shay & J. Clottes, eds., The Limitations of Archaeological Knowledge. Etudes et Recher-
ches Archéologiques de l'Université de Liége, 49.

Lowenthal, D. 1985. The Pastis a Foreign Country. Cambridge University Press, Cambridge,
U.K.

Marshack, A. 1962. The Roots of Civilization: The Cognitive Beginnings of Man's First Art,

Symbol, and Notation. McGraw Hill, New York.

Mason, S. L. R., J. G. Hather & G. C. Hillman. 1994, Preliminary investigation of the plant

macro-remains from Dolni Vestonice II, and its implications for the role of plant foods in

Palaeolithic and Mesolithic Europe. Antiquity 68: 48-57.

Milisauskas, S. 1990. People's revolution of 1989 and archaeology in Eastern Europe. Antiq-

uity 64: 283-5.

Montet-White, A. 1991. Lithic acquisition, settlements, and territory in the Epigravettian of
Central Europe. Pages 205-220 in A. Montet-White & S. Holen, eds., Raw Material Econo-
mies among Prehistoric Hunter-gatherers. University of Kansas, Lawrence, KS.

. 1994. Alternative interpretations of the late Upper Paleolithic in Central Europe. Ann.
Rev. Anthropol. 23:483-508.

Montet-White, A. & S. Holen, eds. 1991. Raw Material Economies among Prehistoric
Hunter-gatherers. University of Kansas, Lawrence, KS.

Miiller-Beck, H.-J. 1991. An interpretation of the meaning of early Paleolithic sculptures. Sae-
culum 42:56-70.

ed. 1973. Neue paldolithische und mesolithische Ausgrabungen in der Bundesrepublik
Deutschland, zum 9. INQUA-Kongress, Tubingen, Germany.

Neugebauer-Maresch, C. ed. 1993. Niederdsterreich in Urgeschichte. Wien, Austria.

Nodelman, S. 1979-80. The legacy of the caves. Portfolio: The Magazine of the Visual Arts, |:
48-55.

Otte, M. 1981. Le Gravettien en Europe centrale, Dissertationes Archaeologicae Gandenses
20.

286 TOMASKOVA

—__. 1990, Revision de la sequence du Paléolithique Supérieur de Willendorf (Autriche).
Bulletin van het Koninklijk Belgisch Insitut voor Naturrwetenschappen, Aardwetenschap-
pen, 60:21 9-228.

. 199], Evolution in the relationship between raw materials and cultural tradition in the
European Paleolithic. #7 A. Montet-White & S. Holen, eds., Raw Material Economies
among Prehistoric Hunter-gatherers. University of Kansas, Lawrence, KS.

Otte, M. & L. H. Keeley, 1990. The impact of regionalism on Palaeolithic studies. Curr. An-
thropol. 31:577-582.

Perles, C. 1987. Les Industries lithiques taillées de Franchthi (Argolide, Grece). (7 T.W. Jacob-
sen, ed., Excavations at Franchthi Cave, Greece, Fascicle 3, 5. Indiana University Press,
Bloomington, IN.

Pfeiffer, J. E. 1986. Cro-Magnon hunters were really us, working out strategies for survival.
Smithsonian, pp. 75-85.

Praslov, N. D. 1977. Problemy Paleolita Vostochnot i Tsentralnoi Evropy. Nauka, Leningrad,
Russia.

Praslov, N. D. & A. N. Rogachey, 1982. Paleolit Kostenkovsko-Borschevskovo Raiona na
Donu 1879-1979. Nauka, Leningrad, Russia.

Price, S. 1989. Primitive Artin Civilized Places. Chicago University Press, Chicago.

Prichystal, A. 1991. Zdroje kamennych surovin. in J. Svoboda, ed., Paleolit Moravy a Slezska.
CSAV, Archeologicky Ustav Brno, Czech Republic.

Priifer, OH. 1958-59. The Upper Palaeolithic cultures of the lower Austrian loess region.
Quartar, 10:79-114.

Renfrew, C. 1989. Archaeology and Language. The Puzzle of Indo-European Origins. Pen-
guin, London, U.K.

Rosen S. A. 1991, Paradigms and politics in the terminal Pleistocene archaeology of the Le-
vant. Pages 307-321 in G. A. Clark, ed., Perspectives on the Past..University of Pennsylva-
nia Press, Philadelphia.

Schrire, C. 1984. Wild surmises on savage thoughts. Pages 1-25 in C. Schrire, ed., Past and
Present in Hunter Gatherer Studies. Academic Press, New York.

Schrire, C. ed. 1984. Past and Present in Hunter Gatherer Studies, Academic Press, New Y ork.

Soffer, O. 1985a. The Upper Palaeolithic of the Central Russian plain. Academic Press, New
York.

_ 1985b. Patterns of intensification as seen from the upper paleolithic of the Central Rus-
sian Plain. Pages 235-270 in D. Price & R. Brown, eds., Prehistoric Hunter-Gatherers.
Academic Press.

__. 1989. Storage, sedentism and the Eurasian Palaeolithic record. Antiquity 63:719-732.

_ ed. 1987. The Pleistocene Old World: Regional Perspectives. Plenum Press, New

York.

Soffer, O. & C. Gamble, eds. 1990. The World at 18 000 BP. Unwin Hyman, London, U.K.

Soffer, O. & N. D. Praslov, eds. 1993. From Kostenki to Clovis: Upper Paleolithic — Paleo-
Indian adaptations. Plenum Press, New York.

Soffer, O. & Vandiver, P. 1994. The ceramics. Pages 151-162 in Svoboda, J. ed., 1994. Pavlov
1952-53 Monograph Series in Archaeology, University of Liege, Belgium.

Svoboda, J. 1983. Raw material sources in Early Upper Palaeolithic Moravia. The concept of

lithic exploitation areas. Anthropologie 21:147-158.

_. 1989. Middle Pleistocene adaptations in Central Europe. J. World Prehist. 3:33-70.
__. 1990. Moravia during the Upper Pleniglacial. Pages 191-203 im O. Soffer & C. Gamble,
eds., The World at 18 000 BP. Vol. 1: High Altitudes, Unwin Hyman, London, U.K.

- ed. 1991a. Paleolit Moravy a Slezska. CSAV, Archeologicky Ustav Brno, Czech
Republic

ART AND ARCHAEOLOGY IN CONTEXT 287

.ed. 1991b. Dolni Vestonice Il. Monograph Series in Archaeology, University of Liege,
Belgium.

.ed. 1994. Paviov 1952-53 Monograph Series in Archaeology, University of Liege, Bel-
gium.

Svoboda, J. & K. Siman. 1989. Middle-Upper Palaeolithic transition in southeastern central
Europe. J. World Prehist. 3:283-322.

Tomaskova, S. 1995. Boundaries and differences: Central European Paleolithic under a mi-
croscope. PhD Dissertation. University of California, Berkeley.

Trigger, B. 1984. Alternative archaeologies. Nationalist, colonialist, imperialist. Man 19:355-
370.

. 1989. History of Archaeological Thought. Cambridge University Press, Cambridge,
U.K.

Ucko, P. & A. Rosenfeld. 1967. Paleolithic Cave Art. McGraw-Hill, New York.

Valoch, K. 1986a. Les questions du Pavlovien. Antiquites Nationales 18:55-62.

. 1986b. Stone industries of the Middle/Upper Palaeolithic transition. Pages 1-19 in M.
Day, R. Foley & Wu R.K., eds., The Pleistocene Perspective, vol. 1. World Archeological
Congress, Southampton, U.K.

. 1987. Raw materials used in the Moravian Middle and Upper Palaeolithic. Proceedings
of the International Conference on Prehistoric Flint Mining and Lithic Raw Material Iden-
tification in the Carpathian Basin. pp.263-268. Siimeg, Budapest, Hungary.

Vandiver, P., O. Soffer, B. Klima & J. Svoboda 1989. The origins of ceramic technology at
Dolni Vestonice, Czechoslovakia. Science 246:1002-1008.

Veit, U. 1989. Ethnic concepts in German prehistory: A case study on the relationship between
cultural identity and archaeological objectivity. Pages 33-56 in S. J. Shennan, ed., Ar-
chaeological Approaches to Cultural Identity. Unwin Hyman, London, U.K.

Weniger, G. C. 1989. The Magadalenian in western central Europe, settlement pattern and re-
gionality. J. World Prehist. 3/3:323-372.

. 1990. Germany at 18,000 BP. Pages 171-192 in O. Soffer and C. Gamble, eds., The
World at 18 000 BP. Vol. I: High Altitudes. Unwin Hyman, London, U.K.

. 1991. Notes on the mobility of groups of hunters during the Upper Paleolithic. Saecu-
lum, 42:82-103.

White, R. 1992. Beyond art: Toward an understanding of the origins of material representation
in Europe. Ann. Rev. Anthrop. 21:537-64.

Wobst, H. M. 1989. Commentary: a socio-politics of socio-politics in archaeology. Pages
136-140 in V. Pinsky & A. Wylie, eds., Critical Traditions in Contemporary Archaeology.
New Directions in Archaeology, Cambridge University Press, Cambridge, U.K.

. 1990. Minitime and megaspace in the Palaeolithic at 18K and otherwise. Pages 331-
343. in O. Soffer and C. Gamble, eds., The World at 18,000 BP, vol.1, High Latitudes .Un-
win Hyman, London, U.K.

Wobst, H. M. & A. S. Keene 1983. Archaeological explanation as political economy. in J.M.
Gero, D.M. Lacy & M. L. Blakey, eds., The Socio-politics of Archaeology. Dept. of An-
thropology, University of Massachusetts, Research report no. 23, Amherst.

Yoffee, N. & A. Sherratt , eds. 1993. Archaeological Theory. Who Sets the Agenda?

Cambridge University Press, Cambridge, U.K.

289

Archaeological Signatures of the
Social Context of Rock Art
Production

Andrée Rosenfeld

Department of Archaeology & Anthropology
Australian National University

P.O. Box 4 Canberra

ACT 2601, Australia

A significant trend in recent rock art studies in Australia has focused on the con-
cept of art as communication, and hence on identifying the extent of past ‘networks of
communication’ and their role in the adaptive strategies of hunter-gatherer societies
in varying environments (for a review of recent trends see Rosenfeld 1992a). Some
more recent studies have questioned the dominant role of environmental constraints
in the development of Australian prehistory (see, e.g., David and Cole 1990; David
1991; David and Chant 1995; McDonald 1994) but the concept of rock art as an ex-
pression of group identity and territoriality remains implicit, as does the model of ex-
tended networks in ‘high risk’ environments, and more bounded networks in fertile or
predictable environments.

Quite how rock art articulates territoriality or intergroup communication remains
poorly explored despite the continuing relevance of rock art (and other) sites in con-
temporary Aboriginal society. Anthropological studies of the symbolism and social
context of art production have concentrated on the diverse artistic modes in ephem-
eral or perishable material, that is, those that are not fixed in the landscape, while the
archaeological evidence is virtually restricted to rock art. As also emphasized in
Davidson’s paper, (this volume) different categories of art construct meanings in dif-
ferent spheres of sociality so that the insights gained from anthropological perspec-
tives need to be contextualized in their application to an understanding of rock art.
Other material culture from archaeological contexts consists almost exclusively of
stone artifacts that are by and large devoid of formal regularities other than those im-
posed by their technology, and while lithic raw material itself can have symbolic
value, this concept is difficult to use archaeologically. Archaeological considerations
of symbolic behavior have therefore focused almost exclusively on the development
of rock art. Important exceptions are McBryde’s (1986) analysis of the distribution of
greenstone axes in Victoria, Morwood’s (1987) integration of stone arrangements in

Beyond Art: Pleistocene Image and Symbol Memoirs of the Califomia Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

290, ROSENFELD

his analysis of southeast Queensland archaeology and Pardoe’s (1988) discussion of
the significance of cemeteries in maintaining closely bounded descent based social
entities in the Riverina region of western New South Wales and Victoria.

Although the chronology of Australian rock art is still based on very little secure
dating (Rosenfeld 1994), the evidence available indicates a chronological trend from
stylistic uniformity to increasing regionalism. During the later Pleistocene two major
stylistic zones can be recognized: Across the greater part of the Continent early rock
art appears to consist entirely of petroglyphs with a restricted range of formal motifs,
while in Arnhem Land (and probably extending westwards into the Kimberleys)
highly detailed figurative paintings are generally also considered to be of Late Pleis-
tocene age. By analogy with the low density and dispersed populations of the present
day arid zones, this uniformity of stylistic expression is thought to define broad re-
gions within which very extensive and open networks of relations operated among
small dispersed populations throughout the Pleistocene occupation of Australia — de-
spite the varying environmental conditions through time and space. Indeed the mod-
eling of much Pleistocene archaeology on present day Aboriginal adaptive strategies
has resulted in the formulation of an extraordinarily static social and ideational his-
tory for the 50,000 or so years of colonization of the Continent. An emphasis on the
adaptive significance of maintaining communication networks has led to the devel-
opment of a model which incorporates stylistic expression as an integral component
in the process of the initial colonization of the Continent (Smith 1992b), and this has
facilitated the uncritical acceptance of very early dates claimed for some Australian
rock art.

During the mid- to later Holocene these widespread early styles give way to a
more diverse body of rock art with spatially restricted areas of stylistic unity. This has
led some authors to suggest the development of more bounded networks as adaptive
strategies with changing environments (e.g., Lewis 1988 for West Arnhem Land) or
with the development of more specialized extractive technologies (e.g., Morwood
1991:4 for the Central Queensland Highlands) during the Holocene. Layton
(1992:235ff), on the other hand, while not working within the confines of a proces-
sual archaeology, also emphasizes the increased regionalism in rock art after the
early-mid Holocene climatic optimum. He examines these Holocene developments
in rock art as evidence for adaptive changes in territorial and social organization, pos-
sibly heralding the development of clan totemism.

While such functional approaches to the study of rock art have brought it firmly
into the arena of mainstream Australian archaeology (and hence of archaeological
funding), the analyses proposed, even on a regional scale, remain very coarse grained
chronologically and spatially. Equally problematic is their firm anchorage in a nar-
row and often ecologically focused processualism. This has required little considera-
tion of the particulars of stylistic expression or of choice of subject matter beyond
their use in the definition of stylistic unities (but see Lewis 1988), nor has the internal
patterning of art sites within a region received much attention (but see McDonald
1991). The full potential of Australian rock art as archaeological information remains
untapped, to a large extent because adequate methodologies for its exploitation are
only beginning to be explored (e.g., Layton 1992; McDonald 1991; Officer 1991; Ro-
senfeld 1992b). There has also been little examination of the putative role of rock art
in the operation of ‘networks ofcommunication’ nor on the very nature of rock artas a
behavioral manifestation.

In this paper I propose to examine more closely what rock art is, and in particular
the nexus between rock art and territoriality in Aboriginal Australia with a view to

SOCIAL CONTEXT OF ROCK ART 29]

finding archaeological correlates. As Forge (1991) has pointed out, the very category
rock art is a construct of western archaeology: Its cohesion or its unity as an indige-
nous cultural product is questionable, and the nature of rock art in relation to other
cultural practices therefore needs closer examination If rock art studies are to be inte-
grated meaningfully into archaeological analyses. In view of the emerging focus on
rock artas integral to current Australian archaeology, it is pertinent at this stage, to ex-
plore particularly what role(s) rock art plays in territorial and other social negotia-
tions, especially in relation to access to subsistence resources.

A first consideration is the proposition that not all features at rock art sites are the
same order of cultural manifestation. At least two very distinct categories of features
are comprised in what we normally think of as rock art. This issue is dealt with at
greater length in two other publications (Rosenfeld 1992b, 1993) and only a brief
summary is presented here.

On the one hand there are coherent sets of images the meanings of which are con-
veyed by virtue of their visual qualities as constrained by conventions of graphic con-
struction, Z.e., by style. These artistic systems, in referring to cosmological truths,
operate primarily to visually express and mediate supernaturally sanctioned power
relationships. Artist and viewer relate to them according to their place in the web of
social relations that operate the ideational system.

On the other hand there are also many visual marks at rock art sites, the form of
which is purely mechanically determined. While these can also be meaningful, the
meanings are not mediated by culturally construed conventions of form. These are the
hand stencils or prints, incised grooves, finger flutings, etc. As Forge (1991) has
pointed out the form of hand stencils is universal and referential meanings are medi-
ated through knowledge of the actor and act of their creation, even if the actors are
identified merely as ‘the old people’ (7.e., our ancestors). Their signification is to indi-
viduals acting within certain contexts, but without visual reference to the individual's
identity as underpinned by the ideational system. In other words, these rock markings
emphasize the individual identity of their maker in contrast to the supraindividual
identity that constrains the creation of rock art proper.

Current Aboriginal practice in relation to rock art is arguably our most relevant
source of information into its significance. Ethnographic observations, however, are
variable according to the context in which they were made, and in the depth of their
analyses, but significant conclusions can be reached (for an in depth discussion of the
relevance of current Aboriginal practice and perception see Layton 1992: Chapter 5;
Layton 1994). In the first place, it is clear that most, ifnot all Aboriginal rock art has in
some sense a meaningful relationship to the locale in which it is executed. There are
cultural constraints of varying severity on artists that influence who may paint and
what it 1s appropriate to paint at any locality. Furthermore, most rock art concerns
ideas of the dreaming either explicitly or evocatively, or by negation via the ‘trickster’
spirits that stand in opposition to the ordering principles of the dreaming (Layton
1992:82-86). Exceptions to these generalities are the hand stencils and incised
grooves (i.e., the rock markings as defined above) which are made as a personal com-
memoration, as well as some other personal commemorative images and some sor-
cery images. Beyond these generalizations, however, the nature of the relationships
between the artist, the locality and what is produced become diverse.

A brief overview of the present or recent context of rock art production in some of
the major rock art regions of the country explores how creating rock art may articulate
with aspects of identity and of territoriality in contemporary or recent Aboriginal so-
ciety. These regions include those that would fall clearly within the archaeological

292 ROSENFELD

notion of the stylistic homogeneity of a bounded network in a predictable environ-
ment, for instance the Kimberleys, and by contrast the situation in the Central Region
which is one of the inspirational models for the so called ‘open network’ of archaeo-
logical writing, where stylistic expression appears to be poorly differentiated over
very extensive tracts of country.

Areas of rock art that constitute geographically bounded regions with internally
coherent stylistic expression include for instance the Wandjina art of the west Kim-
berleys, the stencil art of the Central Queensland Highlands or the distinctive ‘x-ray’
art of west Arnhem Land. Although, as Officer (1992b) warns our definition of the ex-
tent of stylistic regions based on rock art are of necessity conditioned by the natural
geology and topography of the country, stylistic variation of sufficient degree enables
the definition of distinct foci of artistic expression for many parts of the country.
Closer focus on these regions, however, reveals that the degree of internal homogene-
ity varies, and that the rock art articulates with very different expressions of sociality
and territoriality. There are differences in terms of the significance of localities with
rock art and of the social contexts of rock art production. In other words there are dif-
ferences in the relationship between the nature of the rock art, the social significance
of the locality and the artist's affiliation to that locality.

The Central Queensland Highlands perhaps conform most closely to the model as
conceived archaeologically: A highly distinctive rock art emerges ca. 4000 years ago
at the same time as technological changes that permit a more efficient extraction of
the subsistence resources of the region. This is accompanied by increased site density
and a dispersed settlement pattern. By contrast the preceding period is characterized
by a low density site distribution anchored to reliable water sources and associated
with petroglyphs that delimit the eastern extent of the Pleistocene petroglyph style.
Morwood (1984. 1991) has interpreted this mid-Holocene development, and particu-
larly the emergence ofa regionally bounded and distinctive rock art, as an expression
of the closing up of networks of communication within a region that had formerly
been linked via open networks to other areas further afield.

Morwood (1984) found the ethnographic information for the region to be very in-
complete and to relate to a period of stress and conflict with western colonizers. It
does not inform directly on the practice or perceptions of rock art. In the recent past
the norms of sociality and territoriality were based on matrilineal descent, clan exog-
amy, and patrilocal residence for men. This insured a dispersed clan membership
among territorially focused foraging groups constituted according to male kin. Clan
based ceremonies that reaffirm socially constituted identities and spiritual responsi-
bility to territory normally required travel to totemically defined centers. This would
tend to require travel beyond a person's habitual foraging range and resulted in the
gathering of large numbers of people underwritten by the fruiting of a grove of
macrozamia nuts that provide a periodic glut of starch-rich food (Beaton 1982). Such
sporadic large gatherings functioned to articulate a multidirectional network of
interactions that bound the entire region. However, there is no evidence to suggest
that rock shelters or rock art played a role in these ceremonies. The rock art which
consists overwhelmingly of hand and other object stencils, occurs mainly in or near
camping sites where it marks the patrilineally determined territory, i.e., the habitual
foraging range and resources of closely related men and their families. Rock art is also
associated with burials placed in natural hollows in the walls of rock shelters, and
these paintings tend to comprise more formal motifs (thought perhaps to be clan
designs). The burials clearly mark an affiliation between the deceased individual and
the locality.

SOCIAL CONTEXT OF ROCK ART 293

In contrast to the Central Queensland situation the Wandjina paintings of the West
Kimberleys are focal localities in a patrilineal clan’s estate. These sites define a to-
temic locus for clan member’s ritual obligations and for the expression of their so-
cially constituted identity. Painting or renovation of the Wandjina figure and its
associated motifs is integral to these totemic rites and is restricted to appropriate indi-
viduals identified according to their assigned cultural identities (usually senior clan
men).

The network of communication between clans within each of the two patrimoie-
ties is a tightly structured, linear and directional system known as the Wunan. The
Wunan is also conceived of spatially as two interlocking arcs of territory that link the
contiguous estates of each moiety’s clans along a line of communication (Blundell &
Layton 1978). Ritual interactions, exchange, etc. operate up or down the line of the
Wunan, that is, between members whose clan territories, incorporating their
Wandjina art site, are contiguous geographically and symbolically. The system is
closed or bounded insofar as it encompasses a determined number of patriclans from
three distinct language groups.

A zigzag-like linkage between the two lines of the Wunan is insured by rules of
moiety exogamy. It is these affinal relationships that primarily determine a foraging
group's seasonal composition and its range. Foraging rights, therefore, extend well
beyond the range of a patriclan estate (Blundell & Layton, 1978; Layton 1992:33ff).
Thus here, territoriality as expressed in Wandyina rock art is a central organizing con-
cept of West Kimberley cognition (Blundell 1980) but it is not the primary determi-
nant of residence, mobility or foraging group composition. It determines a line of
communication according to men's structured identities. Wandjina rock art. however,
does not articulate the habitual interpersonal relations that are an important nexus of
the ‘network of communications’ that determine access to subsistence.

There are formal differences between Wandjina paintings but these are seen (by
archaeologists, at least) as minor variants upon a highly standardized motif, and this
conforms to the theoretical expectations for the symbolic expression of a closely
bounded network. Within the Central Queensland Highlands, the rock art is also re-
gionally distinctive, uniform in style and in its dominant motifs. The very different in-
ternal structures of sociality and land affiliation that operate in these two areas are not
evident from the stylistic analyses available. The stylistic unity of Wandjina rock art
articulates the interrelationship between people as social entities, but the directional
pattern of its integrative symbolism does not closely parallel the actualization of the
communication networks and residence patterns in relation to subsistence. However,
since each Wandjina is a totemic increase site the associated plant, animal and other
motifs depicted are highly site specific (Layton 1985:449). The clan based totemic fo-
cus should thus be recoverable from the geographical distribution of its less dominant
motifs. By contrast, the dominance of hand and object stencils in Central Queensland
rock art may be seen as an artifact of a largely individualized expression of habitual
residence, and not primarily concerned with the articulation of socially constituted
identity and ritual, clan based affiliation to territory.

In the rich monsoonal environment of west Arnhem Land a proliferation of
painted shelters determines a distinctive rock art province with complex internal sty-
listic structure. While the principal features of motif, style and color are widespread.
Tacon (1993) has identified some regional variation from north to south that corre-
sponds to three major linguistic zones, but the stylistic boundaries between the zones
appear to be clinal rather than exclusive. The essential features relevant to the issues
of the social context of rock art and territoriality lie in the distinction between the rit-

294 ROSENFELD

ual context of art ‘of the dreaming’ and the secular context of art ‘about the dream-
ing, and of anecdotal art (Tagon 1992:210-11).

The pattern of land use and custodianship has been well documented in the context
of claims under the Aboriginal Land Rights Act of 1976, particularly for the northern
groups, who live in the exceptionally rich environment of the wetlands and the river
gorges on the margins of the escarpment ( Keen 1980). Territorial responsibility 1s via
patrilineal affiliation which determines custodianship of a clan’s estate and its sacred
localities. Relationship between clans are expressed (in part) through the several lo-
calized activities of shared totemic ancestors and this can be manifest in composite
imagery: For instance the lines dividing the body parts that together constitute a
whole and that characterizes x-ray paintings (Taylor 1989). Sacred localities may
have paintings, but unlike the Wandjina art, the paintings do not seem to be a central
feature of the site’s significance. Rather they act as tangible evidence of ancestral
truths (Keen 1980). This art of the dreaming, however, marks patriclan owned locali-
ties and is subject to clan based controls of execution which may entail the participa-
tion of members from other clans according to totemically sanctioned relationships.
The execution or renovation of such sacred rock art articulated the cultural identities
and relationships of the participants.

Foraging rights are subject to very little control: Residence, mobility and foraging
group composition are loosely patterned, and individuals can operate an open net-
work of communication with the region, quite unlike the structured directionality of
the Wunan of the west Kimberleys. This fluidity of rules of access to territory and its
resources (Keen 1980) is paralleled ina fluidity of rights to paint nonsacred paintings
(Haskovec & Sullivan 1989; Layton 1992:235), These are executed mainly in wet
season camps when people forage widely away from the flooded wetlands. Certain
motifs, suchas the linear human figures of the Oenpelli area define such an essentially
secular style (Tagon 1992:205). However, there are other human or animal motifs that
can be either sacred or secular according to context and especially the form of their in-
fill design (Taylor 1989:384; Tagon 1989:238). Archaeologists have recognized the
different styles of infill, but have interpreted these as chronological variants, not as
distinct signifiers (Brandl 1973; Chaloupka 1984). The extent to which the rock art
marks territorial affiliation as constituted through social identity or through mere
rights of residence is, therefore, expressed in the visual qualities of the art but itis not
unambiguously evident from an archaeological perspective.

The large number of camp sites with paintings is one of the most striking charac-
teristics of this region. Many of these paintings commemorate the plentiful nature of
resources, either specifically in their subject matter, or more evocatively in the use of
the nonsacred x-ray design that connotes concepts of ‘rainbowness,” the essence of
life force (Tagon 1989). The personalized statement of territorial access here finds ex-
pression (in addition to hand stencils) ina visual style that denotes a generalized force
of the dreaming, not the specificity that articulates socially constituted identities. In
this case, the essentially secular use of formal imagery has blurred the archaeologist's
ability to distinguish between expressions of individual and supraindividual identi-
ties.

A similar problem arises in the Central Queensland rock art where formal motifs
are constructed from a pattern of many small stenciled units created with different ob-
jects, parts of objects or varying finger positions (Walsh 1983). This may also blur the
archaeologist's ability to distinguish between individualized expression and the for-
mal designs thought to relate to clan identities of the deceased.

SOCIAL CONTEXT OF ROCK ART 295

Preliminary study of the recent rock art in the Central Ranges suggests that, like
the Arnhem Land situation, there are at least two distinct classes of rock art sites. In
rock shelters used as camp sites rock art consists mainly of hand stencils and track
lines with only rare formal motifs such as circles, arcs, etc. The loosely structured
syntax of the track lines parallels the art of sand drawings that frequently illustrate or
punctuate verbal communication. It may refer to the dreaming but does so at its most
secular level (Munn 1973:61). A smaller number of sites that are not camping sites,
and some of which are known to be totemic increase sites, contain formally structured
designs often unique to the site (and see Layton 1992:236). These designs are created
from the same limited motif range of circles, arcs, and lines, usually with fewer tracks
(or at Ruguri consisting almost exclusively of emu tracks), and they seem to display
greater regularity in their spatial organization. The tighter patterning of these designs
suggests parallels with the syntax of the body and object art of ceremonial context
such as the range of guruwari designs of the Walpiri of the Tanami desert (Munn
1973: 119ff).

The use of this limited artistic vocabulary mainly of tracks, circles, arcs, efc., 1s
widely distributed through much of arid Central Australia, and also characterizes the
widely distributed petroglyphs that have their origin in the Pleistocene. Smith
(1992a) has related the stylistic uniformity of present day Central Australian art mo-
tifs to the extensive networks of totemic links that operate here, and she contrasts this
to the heterogeneity of the artistic vocabulary that permits the identification of dis-
tinct art styles among the more closed social networks of Arnhem Land. However, a
recent study by Dickins (1992) demonstrates how superficial the appearance of artis-
tic homogeneity can be. Dickins has been able to distinguish geographical and con-
textual styles in the shield designs of Central Australia according to their syntax.

Among the Aranda speakers of the Central Ranges patrilineal totemic affiliation is
normally dominant, but individuals may choose to emphasize their totem of place of
conception or of place of birth in negotiating social relations and access to territory or
resources. Furthermore the extensive travels of many totemic ancestors may be called
upon to create additional links between persons whose dreamings are at different lo-
calities on the same ancestor’s travels. Ultimately these can create indirect links that
cross the greater part of the Continent from north to south and east to west, and inci-
dentally, also extend into regions with different art styles. In practice it seems those
very distant links operated only rarely, but they were central in articulating some of
the long distance trading expeditions across the Continent.

This system of different modes and degrees of affiliation opens a multidirectional
network of relations for individuals as members of several differently constituted
groups (Strehlow 1947). In addition, an emphasis on the system of dual ritual reci-
procity insures that individuals have both ‘ownership’ responsibilities to their own
patrilineal totemic sites, and ‘management’ responsibilities via the matriline.

Thus an individual has potential interests of varying degree to a range of locales or
tracts of country, and he/she shares these interests with different sets of other indi-
viduals. Group affiliation and territoriality are negotiable according to an individual’s
life situation and political status. (Habitual residence is often a primary factor in de-
ciding the principal affiliation that is exercised [Peterson 1983]). In situations such as
these, knowledge ts a particularly valuable commodity, and there is a very high de-
gree of ritual secrecy throughout the Center. Indeed the most effective way of settling
territorial disputes is a demonstration of ritual knowledge usually by recounting rele-
vant myths and songs.

296 ROSENFELD

Relatively little published material is, as yet, available on the rich and varied rock
art of this region, and my discussion is based on an ongoing research project and refer-
ence to unpublished reports made in the context of site management and site protec-
tion legislation. The archaeological analysis of the more recent rock art in this area is
beginning to reveal two contrastive patterns within the same region. One with mean-
dering track lines and hand stencils as its dominant motifs is widely distributed. This
art is associated with archaeological evidence of camping sites and thus relates to
people's habitual foraging ranges. Another of localized, perhaps unique expressions
and with more elaborated motif structure, generally not associated with camping sites
appears in its nonsecular context to be expressive of formal relationships. These two
categories of rock art sites seem to reflect very different aspects of territorial affilia-
tion.

This parallelism for the dual contextual traits of rock art in the two contrastive en-
vironments of the Central Ranges and west Arnhem Land suggests that our models of
style in relation to environmental adaptation are inadequate to deal with the full com-
plexity of the realities of rock art production.

So far the discussion has been restricted to a static view of rock art in relation to
varying facets of territoriality and the identity of the artist. A further quality of rock art
is its relative durability. Once executed it transforms a locality ina manner that is con-
tinually socially reconstituted in the process of its observation and, possibly, its re-
painting or augmentation. The dynamics of the meaning of rock art have been
demonstrated for instance in Wardaman country by Merlan (1989). She shows how in
the process of rock art perception and understanding features are isolated for exegesis
according to the observers’ knowledge of the significance of the locality, and this is
expressed in terms of social relations to territory as defined through the dreaming.
Similar observations are reported by Lewis & Rose (1988) in the Victoria River Dis-
trict and are discussed more extensively by Layton (1992, Chapter 5). Clearly the
meanings of rock art, while seen as a legitimization of social territorial affiliations are
open to constant renegotiation according to the extent of an individual's cultural
knowledge and interests.

With the passage of time, marks that were understood in their individualizing or
secular context become assigned to the creation. They acquire the potency of the
dreaming in the manner of (what we call) natural features. In this way, both technical
marks and images of foreign style can be incorporated into the system of meanings
that is contemporary and coherent.

To summarize, rock art can be variously an ancestral creation, a human creation
about the dreaming ora human creation about human concerns (e.g., love magic, efc.,
or commemorative).

The creation or maintenance of ancestral images is restricted to individuals ac-
cording to their structurally determined identities. The art created in such contexts is
narrowly constrained in terms of style and motif. Subject matter and style emphasize
the nature of the artist's contextualized relationship to a locale. In its reference to ap-
propriately expressed concepts of the dreaming it expresses his/her structurally deter-
mined interests in the locale as defined via the legitimacy of the ideational system of
social relations. The creation or renovation of such paintings is an act of expressing
this classificatory and contextual identity, and of relationship to place via the spiritual
power of the ‘law.’ This rock art therefore expresses and mediates social relations
rather than explicitly territorial affiliation. It is an affirmation of territorial affiliation
via social relations as constructed through cosmological principles.

SOCIAL CONTEXT OF ROCK ART 297

The execution of rock markings and of nonsacred designs is generally concurrent
with rights of residence that are much more inclusive. In these contexts individual
presence or concern may be celebrated in art, but this art does not evoke rights of con-
trol over place nor of ritual affiliation to it. It makes reference to the artist’s individual
identity but it does not visually define his/her socially constructed identity.

To that extent rock art in its formal variation may be thought to express and deter-
mine a social landscape. But it is one of far greater complexity than that subsumed un-
der the simple model of reciprocity of access to resources between members of
territorial groups.

Conclusion

Among the members of a residential group, i.e., one that habitually exploits a par-
ticular foraging range, individuals are likely to have differing ritual affiliation to the
localities within that range — differing at least between marriage partners, but in some
situations also between related same sex members of the group. Furthermore, indi-
viduals may belong to several differently constituted networks of communication and
have differing territorial interests. Stylistic expression at any rock art site, as well as
the artistic mode chosen serve to express the artist’s particular level of affiliation ac-
cording to the social context in which the rock art is produced. Different rock art sites
within a region may fulfill different roles, and be subject to varying constraints on ar-
tistic expression. Furthermore, as Smith anticipated (1992:40), it is in the art of secu-
lar contexts that the uniformity of stylistic expression is likely to be found most
widely distributed, while unique or regionally localized motifs may serve to identify
internal social/territorial differentiation within a stylistic region. This conclusion sug-
gests there are serious limitations to the analytical potential of the normative descrip-
tions, often validated by statistics, that are widely used in the identification of the
characteristics of regional rock art assemblages. The potential analytical significance
of rare or highly localized traits is lost in such studies, as is any analysis of internal di-
versity and of organisational principles.

As far back as 1962, Hiatt wrote about Aboriginal territoriality, * it is important to
distinguish two kinds of relationships between people and land, ritual relationships
and economic relatiohships’ (Hiatt 1962:284). From the survey presented here, it is
clear that some aspects of these separate types of relationships are evidenced in rock
art: in the motifs used and in the stylistic conventions used, as illustrated notably in
Tacon’s study of Arnhem Land rock art. Clearly in an archaeological context it will be
difficult to distinguish the internal stylistic boundaries that define contextual mean-
ings. However, with the identification of hand stencils and other mechanical marks as
visual statements of the artist’s individual rather than social identity, it becomes pos-
sible to explore some aspects of past territoriality and sociality as mediated via rock
art. Furthermore a preliminary assessment of the rock art in the Central Ranges sug-
gests that the internal patterning of motifs within a site, such as the distinction be-
tween structured designs and meandering track lines may also be pertinent to issues of
social context.

I would conclude, therefore, that the identity and/or territorial relationships ex-
pressed via the several modes of rock art are diverse. In particular that formalized art
systems, i.e., those that conform to style conventions operate primarily, though not
exclusively, to relate socially constructed identity-place relationships. In their refer-
ence to the dreaming, they operate to negotiate the varying degrees of territorial ac-

298 ROSENFELD

cess as determined via the constructs of social identities. Mechanically made marks
(and also the personal commemorative images) refer to a more individualized rela-
tionship, one that finds expression in the interpersonal relations of habitual foraging
group compositions and ranges.

In emphasizing the distinction between images that validate the artist’s social
identity and the individualizing marks, it becomes possible to explore some aspects of
past territoriality and sociality as mediated via rock art. While a few years ago Layton
(1992:237) concluded from his analysis of the diversity of contextualized rock art ex-
pressions that “All these considerations tend to reduce the probability of correctly re-
constructing the cultural context of past rock art traditions,” IT would now maintain
that focused regional archaeologies of rock art which combine a consideration of the
diversity of form and of syntax, of ‘art’ versus ‘mark,’ and of archaeologically de-
fined contexts lead to a more optimistic view of the interpretative potential of rock art
in archaeology.

Acknowledgments

1am indebted to Meg Conkey and Olga Soffer who organized the Oregon sympo-
sium at which this paper was first presented, and to all participants whose contribu-
tions and discussions created such a stimulating (and enjoyable) environment that
necessarily influenced the final formulation of this paper. also want to acknowledge
valuable discussions with colleagues, and particularly with lan Keen and with Nic Pe-
terson in relation to current views on patterns of Aboriginal sociality and territorial-
ity.

I also particularly want to thank Jean Auel for her superb hospitality at the First
Oregon Archeological Symposium and for travel assistance.

Literature Cited

Beaton, J. 1982. Fire and water: Aboriginal management of cycads. Arch. Oceania. 17:51-58.

Blundell, V. 1980. Hunter-gatherer territoriality: Ideology and behavior in northwest Austra -
lia. Ethnohistory 27:103-117.

Blundell, V. & R. Layton. 1978. Marriage, myth and models of exchange in the West Kimber-
leys. Mankind 11:231-45.

Brandl, E. 1973. Australian Aboriginal Paintings in Western and Central Arnhem Land: Tem -
poral Sequences and Elements of Style in Cadell River and Deaf Adder Creek Art AIAS,
Canberra, Australia.

Chaloupka, G. 1984. From Palaeoart to Casual Paintings. Northern Territory Museum of Arts
and Sciences, Darwin, Australia,

David, B. 1991. Fern Cave, rock art and social formations: Rock art regionalization and demo-
graphic models in southeastern Cape York Peninsula. Arch. Oceania 26:41-57.

David, B.& N. Cole, 1990. Rock art and inter-regional interaction in northeastern Australian
prehistory. Antiquity 64:788-806.

David, B. & D. Chant. 1995. Rock art and regionalization in North Queensland prehistory
Memoirs of the Queensland Museum, 37:360-528.

Dickins, J. 1992. Shielding the Dreaming: A Formal Analysis of Central Australian Shield De-
signs. B.A. Thesis. Australian National University, Canberra, Australia.

SOCIAL CONTEXT OF ROCK ART 299

Forge, J.A. 1991. Handstencils: Rock art or not art. Pages 39-44 in P. Bahn & A. Rosenfeld,
eds., Rock Art and Prehistory: Papers Presented to Symposium G of the AURA Congress,
Darwin 1988. Oxbow Press, Oxford, U.K.

Godwin, L. 1991. Little big men: Alliance and schism in northeastern New South Wales during
the late Holocene. Archaeology and Linguistics: Understanding Ancient Australia (unpub-
lished conference papers). 2:200-215.

Keen, I. 1980. A/ligator Rivers Stage Two Land Claim: Claim book. Northern Land Council
Darwin, Australia.

Hascovek, I. & H. Sullivan 1989. Reflections and rejections of an Aboriginal artist. Pages 57-
74 in H. Morphy, ed., Animals into Art. Unwin Hyman, London, U.K.

Hiatt, L. 1962. Local organisation among Australian Aborigines. Oceania, 32:267-286.

Layton, R. 1985. The cultural context of hunter-gatherer rock art. Man(n.s.) 20:434-453.

. 1992. Australian Rock Art: A New Synthesis. Cambridge University Press, Cambridge,
U.K.

Lewis, D. 1988. The Rock Paintings of Arnhem Land, Australia. BAR International Series 415.
Oxford, U.K.

Lewis, D. & D. Rose, 1988. The Shape of the Dreaming: The Cultural Significance of Victoria
River Rock Art. ATAS Canberra, Australia.

Mcbryde, I. 1986. Artefacts, language and social interaction: A case study from southeastern
Australia. Pages 77-93 in G. Bailey & P. Callow, eds., Stone Age Prehistory. Studies in
Memory of Charles McBurney. Cambridge University Press, Cambridge, U.K..

Mcdonald, J. 1991. Archaeology and art in the Sydney region: Context and theory in the analy -
sis of a dual medium art style. Pages 78-85 in P. Bahn & A. Rosenfeld, eds., Rock Art and
Prehistory: Papers Presented to Symposium G of the AURA Congress, Darwin 1988, Ox-
bow Monograph 10, Oxford, U.K.

. 1994. Dreamtime Superhighway: An Analysis of Sydney Basin Rock Art and Prehis-
toric Information Exchange. PhD dissertation. Australian National University.

Merlan, F. 1989. The interpretive framework of Wardaman rock art: a preliminary report. dus-
tralian Aboriginal Studies \989/2:14-24.

Morwood, M.J. 1984. The prehistory of the Central Queensland Highlands. Adv. World Arch
3:325-80.

. 1987. The archaeology of social complexity in southeast Queensland. Proc. Prehist.
Soc. 53:337-349.

Morwood, M. 1991. Introductory essay on ethnography and the archaeological study of art.
Pages |-6 in M. Morwood & D. Hobbs, eds., Rock Art and Ethnography: Proceedings of the
Ethnography Synmposium (H) AURA Congress, Darwin 1988. Occasional AURA Pubhi-
cation No. 5, Melbourne, Australia.

Munn, N. D. 1973. Walbiri Iconography. Graphic Representation and Cultural Symbolism in
a Central Australian Society. Cornell University Press, Ithaca, NY.

Officer, K. 1991. What's in an anthropomorph question? Pages 112-119 im P. Bahn & A. Ro-
senfeld, eds., Rock Art and Prehistory: Papers Presented to Symposium G of the AURA
Congress, Darwin 1988. Oxbow Monograph 10, Oxford, U.K.

. The edge of the sandstone: Style boundaries and islands in south-eastern New South
Wales. Pages 6-14 in J. McDonald & I. Haskovec, eds., State of the Art: Regional Rock Art
Studies in Australia and Melanesia. Occasional AURA Publication No.6, Melbourne,
Australia.

Pardoe, C. 1988. The cemetery as symbol. The distribution of prehistoric Aboriginal burial
grounds in southeastern Australia. Arch. Oceania 23:1-16.

300 ROSENFELD

Peterson, N. 1983. Rights, residence and process in Australian territorial organisation. Pages
134-145 in N. Peterson & M. Langton, eds., Aborigines, Land and Land Rights. A\AS,
Canberra, Australia.

Rosenfeld, A. 1992a, Recent developments in Australian rock art studies. Pages 231-238 in R.

Laffineur & J. Crowley, eds., E/KON: Aegean Bronze Age iconography: Shaping a meth-
odology, Proceedings of the 4th International Aegean Conference, Hobart 1992. Universite
de Liege, Liege, France.

—_. 1992b. Rock art and rock markings, Paper presented at the 1992 AURA Conference,

Cairns, Queensland, Australia.

. 1993. A review of the evidence for the emergence of rock art in Australia. Pages 71-80
inM. Smith, M. Spriggs & B. Frankhauser, eds., Sahu/in Review. Late Pleistocene Archae-
ology in Australia, New Guinea and Island Melanesia. Prehistory RSPacS, Australian Na-
tional University, Canberra, Australia.

Smith, C. 1992a. The use of ethnography in interpreting rock art: A comparative study of Arn-
hem Land and the Western Desert of Australia, Pages 39-45 in M.J. Morwood & D.R.
Hobbs, eds., Rock Art and Ethnography, Proceedings of the Ethnography Symposium, Aus -
tralian Rock Art Research Association Congress, Darwin 1988. Occasional AURA Publi-
cation No. 5, Melbourne, Australia.

—_. 1992b. Colonising with style: reviewing the nexus between rock art, territoriality and

the colonisation and occupation of Sahul. Austr Arch. 34:34-42.

Strehlow,T.G.H. 1947. Aranda Traditions. Melbourne University Press, Melbourne, Austra -
lia.

Tagon, P. 1989. Artand the essence of being: Symbolic and economic aspects of fish among the
peoples of western Arnhemland. Pages 236-252 in H. Morphy, ed., Animals into Art. Un-
win Hyman, London, U.K.

_ 1992. Somewhere over the rainbow: An ethnographic and archaeological analysis of re-
cent rock paintings of western Arnhem Land. Pages 202-215 in J. McDonald & I. Hask-
ovec, eds., State of the Art: Regional Rock Art Studies in Australia and Melanesia.
Occasional AURA Publication no. 6, Melbourne, Australia.

Tagon, P. 1993. Regionalism in the recent rock art of western Arnhem Land, Northern Terri-
tory. Arch. Oceania, 28:1 12-120.

Taylor, L. 1989. Seeing the inside: Kunwinjku paintings and the symbol of the divided body.
Pages 371-389 in H. Morphy, ed., Animals into Art. Unwin Hyman, London, U.K.

Walsh, G. 1983. Composite stencil art: Elemental or specialised? Austral. Aborig. Stud
1983/2:34-44.

301

Processions and Groups: Human
Figures, Ritual Occasions and
Social Categories in the Rock
Paintings of the Western Cape,
South Africa

John Parkington
Department of Archaeology
University of Cape Town
Private Bag. Rondebosch
Cape 7700, South Africa

Anthony Manhire

Department of Archaeology
University of Cape Town
Private Bag. Rondebosch
Cape 7700, South Africa

‘Egyptian artists, as well as many others, construct their representations according to
mental images which, in their view, summarize the essential physical character of the
objects depicted as opposed to their appearance” (Schafer, Principles of Egyptian Art
1974:x1).

“Mediation between the opposite sexual spheres centers around a comprehensive meta-
phor linking eating and intercourse. Men hunt and eat women as carnivores prey on her-
bivores” (Biesele, Women Like Meat 1993:196).

Our archaeological work in the western Cape is aimed at the writing of a regional
history, an account of the lives of people who were eventually substantially impacted
and then decimated by colonial events. Painted rock shelters are far more common in
the area than other archaeological sources such as cave deposits, shell middens or
stone tool scatters, but they are not easy to read. Although paintings have the advan-

Beyond Art: Pleistocene Image and Syinbol Memoirs of the Califomia Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

302 PARKINGTON AND MANHIRE

tage of not having been moved from their position of original manufacture, most are
undated palimpsests or very residual versions of formerly more detailed pictures.

Despite poor preservation and complex metaphor, we argue here that particular
social occasions can be recognized in the groupings of human figures and that par-
ticular social categories of person can be read from the conventional depictions. The
paintings are not about hunting and gathering but about being a hunter or a gatherer
and the assumption of adult roles at the transition from adolescence to adulthood. The
suggestion that initiation events and painting conventions referring to initiation are
prominent in western Cape paintings prompts us to consider the authorship of paint-
ings and the location of appropriate rock surfaces. The frequent use among ethno-
graphic groups (Biesele 1993; McCall 1970) of verbal metaphors drawn from hunting
behavior to refer to sexual and social relations points to a likely source of meaning for
the paintings.

The search for meaning in rock art studies in southern Africa has been led by
Lewis-Williams. In his early work (198 La) he stressed the role of the eland as “animal
de passage” in ritual and social events, and was prepared to identify such events and
argue their significance in the interpretation of paintings. Later (Lewis-Williams &
Dowson 1989), he has been inclined to give central significance to the actions and ex-
periences of shamans, arguing that even isolated and residual paintings probably re-
flect the visions of trance, downplaying the recognition of actual occasions. We
prefer the earlier formulation because it captures more successfully the significance
of the eland in surviving San commentaries (Parkington er a/. in press). The emphasis
on social roles and ritual occasions is consistent with the ethnography of Biesele
(1993) and the stories recorded by Bleek and Lloyd (Bleek 1931, 1932, 1933, 1935,
1936: Lloyd 1911), both of which underline the centrality of the tensions of social re-
lationships in San or Bushman expressive culture.

Recently, Solomon has pointed to the significant role of gender in both the stories
and the paintings of southern African San (1989, 1991, 1992). We here support her
position and agree that much of the painted imagery refers to an expressive vocabu-
lary that is far wider than trance. Our position is that many paintings in the western
Cape, but certainly not all, make reference to the social roles allocated to men and
women, young and old, and, presumably, though at present less persuasively recog-
nized, to further social distinctions. The paintings, like the stories, become vehicles
for expressing appropriate behavior but also opportunities for challenging the estab-
lished codes.

The Western Cape
Paintings

In recent years we have built a number of bridges between the study of rock paint-
ings and that of excavated materials. The discovery, for example, of paintings show-
ing people hunting with nets and armed with fully recurved (triple-curved) bows
(Manhire et a/. 1985) — neither of which have been documented in excavated collec-
tions or early traveler reports — feeds into the broader inquiry into hunting methods
and diet, along with analysis of faunal remains. The spatial and temporal patterns of
images, which we record by plotting locations, registering superpositioning and iden-
tifying chronologically sensitive imagery (such as domestic animals and colonial pe-
riod artifacts) can be linked to the spatial and temporal patterning in depositional

ROCK PAINTING OF THE WESTERN CAPE 303

settlement histories (Yates et a/. 1994). Rubbing, smearing and repainting of images
identifies them securely as artifacts with enduring significance for people who lived,
literally, within arm's length of many of the paintings (Yates & Manhire 1991).

In an earlier paper (Manhire et al. 1983) we have argued that the distribution of
compositions involving large numbers of human figures reflects an uneven spread of
the social events depicted and symbolized. Thus compositions that can be seen as
dances or conflict scenes are very much restricted to the mountain folds and are rare
on the rock surfaces of the sandy coastal plains. This distinction between mountain
folds and coastal plains, which we now recognize to involve both time and place, has
also emerged in our analyses of ostrich eggshell beads, stone artifact assemblages and
faunal remains. These latter, however, have the advantage of being better dated than
the paintings (Yates et a/. 1994). We can show that, after the appearance of pastoral-
ism in the region between about 1800 and 1500 years ago, traditional patterns of stone
tool and bead manufacture were retained longer and changed more slowly and less
dramatically in the mountains. Along with the higher concentrations of domestic
stock and potsherds in coastal plain sites, we have interpreted this as the retention of
important elements of prepastoralist settlement and behavior patterns in the more iso-
lated parts of the landscape, where the impact was less severe (Parkington et a/. 1986;
Yates et al. 1993: Yates & Smith 1993). It is possible that the depictions of groups of
humans in recognizable social contexts also date to this period of residual but threat-
ened survival.

Rock paintings of the western Cape differ substantially from European Paleolithic
paintings as well as from many American and Australian rock paintings in being
dominated by human figures. This is partly the result of the fact that most human fig-
ures are painted in groups (Figures 1-3). Such compositions are, arguably but cer-
tainly not simply, reflections of social events or episodes. Following the pioneer work
of Maggs (1967, 1971a, 1971b), we try to recognize the character of the social events
and derive social information from the paintings in the area. This means exploring the
conventional ways in which people of various social categories were depicted and in
which groups of human figures were composed. We assume that the painter and con-
temporary viewer shared the vocabulary of painted conventions.

Motifs are usually recognizable where preservation is good, but, because there are
no direct commentaries on the making and meaning of the paintings, we have to de-
pend extensively on archaeological, ethnographic and historical records for the deri-
vation of themes (Panofsky 1962:14). We have recently argued (Parkington ef a/. in
press), for example, that the conventionally ‘tidy’ depiction of a cloak or kaross on
male figures is a device to reveal to the viewer the status of the individuals depicted
“as opposed to their appearance” (Schafer 1974:xi). Reference to ethnographic pho-
tographs and drawings shows that the painted cloaks are not literal images of gar-
ments, but resemble the equally conventionalized depiction of eland torsos, albeit
turned through 90°. A reading of the texts closest in space and time to the precolonial
hunter gatherers of the western Cape (Bleek 1923) supports the notion that the eland
was the animal through which young men became hunters (Lewis-Williams 198 La).
Stories involving the mantis and the eland are related to the creation of the social cate-
gory of hunter rather than of people or animals in general. Our suggestion 1s, then, that
cloaked males should be read as initiated men and that the eland should be read as ini-
tiator of men, creator of hunters and, because only initiated men may marry, of hus-
bands (Parkington ef a/. in press). Figure | is an explicit reference to a male initiation
school, and the best evidence we have that such schools were held in the western
Cape.

PARKINGTON AND MANHIRE

04

ry
3

MO]JAA Svare aTYM ay} uasaidal sear papeys ayy ‘aul] IaMOy ayy FO Yay ay}

‘ Tike

aR

ROCK PAINTING OF THE WESTERN CAPE

305

\ line of female human figures with a smaller male figure, probably a bowman, at right. Color is red except for the lower diamond shaped forms which are yellow

cation: Be

yontjieskloot

IGURE 2

Li

306 PARKINGTON AND MANHIRE

i,

Figure 3. A line of female human figures with a smaller male bowman at right. Color red. Location

ZA,
Treg

Ute

id
RG

ES

Boontyieskloot

Perhaps one of the most enigmatic characteristics of western Cape rock art 1s the
almost total absence of hunting scenes (Maggs 1967:103). Despite the existence of
thousands of paintings of men, bows and animals, there are only a handful of paint-
ings of men actively hunting animals with bows. This is surprising given the fre-
quency with which hunts are retold in the camp by men, but perhaps predictable given
the social taboo on boasting, and consistent with the modesty demanded of a success-
ful hunter. It may also point to an important distinction between painter and hunter.
Apart from the net hunts, in fact, there are few narratively connected associations be-
tween human figures and animals. More common are juxtapositions which imply a
symbolic rather than a narrative connection. It seems inescapable that the bows, ar-
rows and quivers are shown for some purpose other than that of illustrating their func-
tional context. Along with other conventions, hunting equipment serves to identify
people in particular social categories (Parkington ef a/. in press).

Processions
[here are many kinds of compositions involving numbers of humans to the extent

that a simple typology breaks down. Here we concentrate first on processions, de-
fined loosely as those lines of human figures that share a single directionality, often

ROCK PAINTING OF THE WESTERN CAPE 307

use acommon but undepicted baseline and show a coherence of form, colorand scale.
Such processions grade off into more informally grouped humans, but when strongly
linear seem often to imply a dance context. Remarkably, processions usually face
right and, as noted by Maggs (1967:101) some time ago, are almost always made up
of a single sex, although the exceptions are particularly interesting. He pointed out
that this corresponds well with the clearly differentiated gender roles ascribed to
males and females in San ethnography. We go further and suggest that the proces-
sions refer to the most strongly marked gender separations in San ritual life, the initia-
tion or puberty ceremonies held for young men and women.

As with other components of the archaeological record, the most effective ap-
proach is to begin with resolved pictures rather than complex palimpsests. Whereas
Figure | depicts an event from which women are excluded, Figures 2 and 3 are pre-
dominantly processions of female figures, each, interestingly, with one or two male
figures painted, at least in one case, with opposed directionality. The women are iden-
tified by breasts but also have a characteristic posture, not shared with males, which
includes an arm outstretched and holding a short stick, spherical and exaggerated but-
tocks and some details of the delineation of back and thighs. The men have visible pe-
nises, carry drawn bows but are rather stick-like in form, unlike the shapely contours
of the women. Ethnographically recorded occasions which conform to this dramatis
personae are not hard to find. The flirtatious behavior of younger women at a girl's
first menstruation dance includes an interaction with an older male or males (Bleek
1928:23). On this occasion, the Eland Bull dance, the women play the roles of female
eland, discard their normal dress code, exposing their buttocks, and present to the
male ina blatantly sexual manner. The performance seems designed to welcome the
new maiden into the company of adult women. In the paintings the males with drawn
bows recall the metaphor between hunting and sexual activity (McCall 1970). Solo-
mon has drawn attention to exaggerated buttocks (1989, 1992) and Maggs has long
noted the local convention of the spherical buttock on women, which we can suggest
denoted the category of sexually mature women and thus ‘prey.’ This theme is re-
peated with similar conventions in other compositions in the western Cape, most of
which are processional (Figure 4).

In Figure 5 we illustrate a line of male figures processional in the sense of sharing
scale, orientation and base line and rendered even more coherent by the repetition
along the line of two very specific postures. Although the men are equipped with
hunting gear, and are cloaked, the effect of these postures is to suggest a dance rather
than a hunt. Barnard (1992:72, referring to Heinz [1966]) reports that among the !Xo
“the initiation dance has unique music and dance steps,” a comment that recalls this
particular procession. At the rear of the line are two female figures that are arguably
part of the composition, as might be the elephants below. In this case the women have
none of the features ascribed to prey and could be taken as older, perhaps post-
menopausal. Whatever their status, they do not confront the men, and are clearly allo-
cated a segregated position in the rear of the line. This picture most likely makes refer-
ence to the ritual life of men and perhaps the dances that accompany male initiation
(Schapera 1930).

Groups
If we turn our attention away from processions to so-called group scenes, we see

some other conventions. Maggs (197 1a) first described both the limited distribution
and typological coherence of these paintings, which have subsequently been referred

308 PARKINGTON AND MANHIRE

(_BHEHEEEEEE ES&S

Ficure 4. A female human figure with exaggerated buttocks. Color red. Location: Boontjieskloot

to elsewhere (Manhire et a/.1983; Parkington 1989; Parkington ef a/. in press; Yates
etal. 1985). At the core of this coherence is typically a picture of several human fig-
ures, males and females, tightly grouped together, most of them seated but often with
a single standing figure, shown under a set of hanging bags and other equipment,
sometimes with a clear overarching boundary line. There is some variety of expres-
sion, which Maggs (197 1a) took as “typological sequence,” but which may rather re-
flect the flexible use ofa series of conventions, some of which are used in other kinds
of compositions. The key elements seem to be the hanging equipment and the close
juxtaposition, even merging, of human figures. Figures 6 and 7 illustrate the set well.
Spacing is significantly tighter than in processions, there is no attempt to impart a di-
rectionality to the composition and some pictorial closure is often attempted. The par-
ticipation of several social categories is made clear by the recognizable sex of figures,

309

ROCK PAINTING OF THE WESTERN CAPE

peasy ay} Jo YoOoy sq} UlYWM a ) S9dBI} AALY SajeuUlaf OM} 9] PUB YeO]I MO]JOA B JO SadeI} JULBJ SEY }fd] dUIdI}X9 ay} UO AINSI IY, “Pal Jojo) aay UMOYS JOU ale

2uUINY JO duty Buoy ve jo wed “¢ AUNDIA

gory syueydaya ait jsea] ye pue sana UeLUNY aTOW sapnjoul pur pa arasaid Ajiood st aued ay) Jo 1sad ay], ayseeEys ing WO samMsi}

CU

310 PARKINGTON AND MANHIRE

by associated postures and actions, as well as by the variety of equipment painted.
Here again, as in processions, equipment is painted to make a social rather than func-
tional point, in this case that a complete group is intended, located at home in a rock
shelter. We should note that the depiction of children is not well substantiated, partly
because the use of scale has not yet been adequately researched.

Although Maggs (1971a:219) was reticent about identifying the activity implied
in these group scenes, Yates and others (1985:74) have convincingly argued that they
are scenes of trance or healing occasions. Women clap and a single male figure usu-
ally exhibits features long shown by Lewis-Williams (198 1a, 1981b, 1983a, 1983b)
to be conventions related to trance experiences. Reference to San ethnography re-
veals that fundamental features of trance dances are that all members of the group at-
tend, that roles are specified by gender and that the event is held in camp. The
composition of group scenes is organized to emphasize precisely these points, with
the hanging bags and surrounding painted line adding a local detail in the use of acave
as campsite.

We may ask, as did Maggs (197 1a:219), whether the actual numbers of people de-
picted carry meaning? Maggs’ histogram of group sizes (1971a:219, his Figure 1)
shows that most groups are smaller than Kalahari ethnography might have predicted,
but the Karoo material collected by Bleek and Lloyd (Bleek 1931, 1932, 1933, 1935,
1936; Lloyd 1911) refers more frequently to trancing in domestic contexts than in
large gatherings. These details will obviously in part depend on local demographic
patterns. What may be worthy of comment is the complete absence of any animal as-
sociation at these group scenes. No attempt is made to link domestic healing occa-
sions with a particular animal. In fact it is our impression that animals hover near
linear dance processions but rarely next to group healing occasions.

Figures 8, 9 and 10, as well as Figure 1, show how elements of the group scene
paintings are reused and elaborated to construct other pictures. The first of these is
particularly interesting in that the hanging bags and sitting people are placed as a
background to a large group of trancing men. These latter are arranged in registers
rather like an Egyptian tomb painting, presumably an example of Schafer’s aspective
as distinct from perspective viewpoint. Surely the men are not meant to be read as
‘moving to the right in three lines,’ but as ‘dancing in large number,” perhaps in a cir-
cular formation. In Figure 9, by contrast, the composition is simplified to include two
standing, perhaps trancing, males and three seated, clapping females. Note the simi-
larity of the sitting position of these women and at least one of those in Figure 6.

In Figure 10 the painted line that signifies cave or rock shelter is placed around a
natural indentation in the rock. Here is also a reminder that, although bows are rarely
raised in anger against animals, they are more often shown being used against other
people. The suggestion (Yates er al. 1985:78) that this may not be a literal fight may
be true, but the fact remains that conflict scenes outnumber hunting scenes in the
western Cape.

Exclusion and Exposure

Depictions of subsistence scenes, such as the hunting of animals, the gathering of
shellfish or the collection of plant foods are almost completely absent from the west-
ern Cape, despite paintings of the necessary equipment. Instead some details of the
anatomy of people and equipment worn are used to denote different categories of per-
son, particularly the distinctions between men and women and between adolescents

ROCK PAINTING OF THE WESTERN CAPE

3]

itioned above male and female figures. Color dark red

roup scene from Sevilla with bags pos

Ag

iWWRE ©

Fic

ws)
i)

PARKINGTON AND MANHIRE

Figure. 7. A group scene from Keurbos with human figures partially enclosed by a line and with equipment

positioned above. Color red

and adults. The recognition that some of these occasions are initiations, from which
some social categories of people were excluded, raises the questions who painted
these events, where did they put the paintings and why commit them to material, and
thus exposed, form?

[he isolation of ritual occasions is often achieved socially rather than spatially,
with little distance required between domestic and ritual sites. More important, per-
haps, the secrecy of events is often more notional than real, with the emphasis on play-
ing out the role of ignorance rather than ‘really’ being ignorant. Men may know what
women do ata girl’s first menstruation dance, but should not flaunt such knowledge
or even admit to it. Painting such an occasion would seem a thoroughly material
flaunting of knowledge and thus tantamount to attendance. We doubt that those not
permitted to attend an event were permitted to paint it. Because both male and female
events are depicted, does this mean that both men and women painted? It may do, al-
though older men are acommon element in the list of participants. Older men, who no
longer expect to kill large animals, who may hunt outside the metaphor with snares

ROCK PAINTING OF THE WESTERN CAPE

Ww

_ = = = =

FIGURE 8. The upper part of the panel shows elements of atypical group scene with the lower part filled with
trancing men most of whom are shown in a hand-to-mouth position. Color red. Location: Sevilla

and nets rather than bows, and, thus, for whom young women pose no threat, play key
roles in both male and female initiations. They are the shamans, the scarifiers, the
teachers and the story tellers. Their contribution to the ritual rather than economic
well-being of society is a good parallel for the content of rock art, which seems to em-
phasize ritual rather than economic behavior. Postmenopausal women, again no
threat to the successful hunt, may also be shamans and are often key story tellers, but
play no role in male initiations (but see Figure 5). Older people seem likely to have
been the painters, a parallel to the reference at the beginning of stories to phrases such
as “Our mothers told us...” (Bleek 1931, 1932, 1933, 1935, 1936).

Ww

PARKINGTON AND MANHIRE

314

‘JooTysaruoog uo

OT par Jojo ‘sayeut Fursuey Ajqeqoad ‘samay yurey AIDA OM) SplVMO}] Fuldey sopeulay surddeyo ‘payeos aaIyf °6 IUNYIA

ROCK PAINTING OF THE WESTERN CAPE

d {

group of figures partially enclosed within a natural indenta-

1ows a line of human figures on the left facing towards a tightly packed §

ipparent conflict scene sl

ck. Color

re 10. This

red and white

316 PARKINGTON AND MANHIRE

Painting an initiation scene effectively allows the viewer to attend, which raises
the question of location. We have previously described paintings as ‘domestic arti-
facts’ (Parkington ef a/. in press), noting that they are often found within meters of
sleeping and cooking areas. We may be right at one spatial scale but wrong at another.
Painted rock shelters are extremely common in the rocky ridges where people prefer-
entially sought living space, so that at the scale of hundreds of meters or a kilometer
they are intermingled. But most painted sites have little or no deposit and few stone
artifacts or potsherds. We cannot hope to resolve the overprinting of several thou-
sands of years of changing decisions and schedules, but clearly some separation, at
the scale of tens or hundreds of meters, is possible between domestic and painted lo-
cations, if that were necessary. If a painting reflects an occasion which all are ex-
pected to attend, no separation may be required. We might predict such a situation for
the group scenes. But, if reference is made to an occasion restricted to people of spe-
cific social categories, then a site a few tens of meters away achieves a notional isola-
tion.

There 1s, of course, another means to achieve isolation, which is by developing
esoteric painted conventions that are so transformed, so opaque, as to require verbally
transmitted keys. This may be the mechanism adopted in Australia and California, but
it does not seem to have been the one used in the western Cape, at least not exclu-
sively. The same conventions are used in both group and procession scenes, although
it may be that some of the trance-related imagery, crenellated motifs, for example,
had a more restricted distribution in the community.

Here we encounter the major obstacle in using paintings for the writing of histo-
ries: Their poor chronological context. With what other changing components of the
archaeological record are we to link them? Without an understanding of the changing
subject matter and context of manufacture and reuse we can only give them an over-
arching significance, and one that runs the danger of underestimating change. The
evidence we have described here shows that some, at least, of the painting was di-
rected at the underlining of the relationships that were appropriate between male and
female, young and old, initiated and uninitiated, animal and human. As well as con-
firming existing values, the paintings would have been useful vehicles for challenge
and change. Whether this was the case for particular time periods more than others, or
for particular regions more than others remains to be researched.

The Extended Eland Metaphor

What seems to underpin the ritual roles of men and women in San ethnography, as
Lewis-Williams (198 1a) and Solomon (1989, 1992) among others have shown, is the
concept of threat, or danger (Douglas 1966). Despite abundant evidence to the con-
trary in reality, men in San thought are perceived as the critical providers and women,
after their first menstruation, are identified as a threat to male success. The generation
of threat and its containment are played out through the metaphor of the eland, with
men and women playing both the roles of male and female eland and those of hunter
and prey. The eland is thus the pivotal symbol by which people understood their so-
cial and sexual relations (Lewis-Williams 1981a). They were People of the Eland
(Vinnicombe 1976). Critical events selected as the signal that individuals have
moved into their adult roles are, thus, the first menstruation and the first (eland) kill. A
variety of sexual metaphors then develop in relation to hunting and hunting equip-
ment, which prescribe the roles men and women, older men and older women play in

ROCK PAINTING OF THE WESTERN CAPE Sy

ritual occasions. Many of these metaphors are discussed in Solomon’s work (1989,
1992).

There are differences in the treatment as well as the frequencies of animals in the
rock paintings of the western Cape. Eland is the most commonly identified species,
followed by elephant, although there are large numbers of paintings of small bovids
that are difficult to identify to species or genus. Eland depictions are often in proces-
sions, almost never figure in compositions involving people and show almost no vari-
ety of posture. Elephants are often painted in family groups that include young, are
conflated with humans to form elephant headed men and are often associated with
surrounding lines and crenellations (Maggs & Sealy 1983; Manhire et al. 1983).
Small bovids are sometimes associated with nets in hunting scenes (Manhire er al.
1985). Because the paintings are a visual component of the expressive life of the pain-
ters, we attempt here to account for these differences with reference to Biesele’s
(1993) work on verbal expressive life among related hunter gatherers in the Kalahari.

The title of her book Women Like Meat “implies a world of integrated collabora-
tions between the sexes involving work and foodstuffs, marriage and procreation —
negotiations whose continuing success assures the ongoing viability of Ju/hoan [San
or Bushmen] society itself’ (Biesele 1993:1). These negotiations are conducted in-
cessantly and in metaphorical terms with “men’s hunting ... often symbolically op-
posed not to the complementary female activity of gathering but rather to women’s
reproductive capacity” (1993:41). The rites of menarche for girls and the first kill
ceremony for boys mark the readiness of young people to adopt these roles.

“Animals are used as metaphorical operators” (Biesele 1993:88), particularly the
larger animals which have n!ao, a concept “linking mens’ and womens’ great pro-
creative powers — childbirth and hunting — to the vitally important polarities of the
weather” (1993:87). Smaller bovids such as “(steenbok and duiker) generally do not
possess n!ao” (1993:108). The eland, the largest and fattest of the bovids, often fills
the metaphorical role of mediator and 1s specifically implicated in the initiation rites
of both sexes. Girls become women (mothers and sexual partners) by becoming eland
and boys become men (hunters and sexual partners) by becoming eland. “Ideas re-
garding hunting involve sympathetic identification of hunters and prey and/or hunt-
ers’ wives and prey. A sex/food equation is pervasive, and men/animal/deity
relationships are ritually important and elaborate” (1993:41).

“The peculiarly intimate identification between hunter and prey in Bushman be-
lief is traceable in part to the period of hunting during which a man can do no more,
but must wait for the arrow poison to do its work” (Biesele 1993:90). The metaphori-
cal potential here for allusion to the sexual relations between men and women, where
once again, men may introduce the substance (semen/poison) and await the outcome
(childbirth/death of prey), is obvious, but points to the urgent need to understand the
history of poison use and changes in weaponry (Noli 1993). We suggest that the asso-
ciation of menstruation with hunting disaster is not that “she is obviously at the height
of her female procreative power’ (Biesele 1993:93) at that time, but that, by the meta-
phor used, menstruation is a ‘failed kill’ and will sympathetically bring hunting fail-
ure. Adult women bleed and recover, whereas men hope to despatch their prey
through a process which involves fatal blood loss.

The interpretations of western Cape paintings we have built up (Manhire e¢ al.
1985; Parkington 1989; Parkington ef a/. in press) are consistent with this extended
eland metaphor. Figure 1, for example, includes an eland oriented with the initiates
and also underlines the identification of the cloaks of initiated men with the bodies of
eland. Cloaked men are thus a kind of therianthrope, conflating human and eland. The

318 PARKINGTON AND MANHIRE

use of the spherical buttock to depict sexually active women Is a painted parallel to the
exposure of buttocks at the eland bull dance performance. Such women are thus also
therianthropes ofa kind. Paintings of overt sexual behavior are as rare as the depiction
of the hunting of choice food animals, with which they could easily be identified. “It is
hard to tell, even in... everyday discourse, which meat-animal or woman- ts being dis-
cussed. The metaphors tying women to the enchanted hunted prey are so intricate as
to defy untangling” (Biesele 1993:197). Paintings such as Figure 11 show that there
are visual parallels to the verbal conflation of women and animals that are equally
hard to untangle!

Vinnicombe suggested (1976) that the Bushmen of the Drakensberg were People
of the Eland, in the sense that they identified closely and spiritually with eland.
Lewis-Williams later (1988) felt that the connection was through the shamans who
took on the power of eland in trance. We feel closer to the former position, arguing
that the eland, as noted by Lewis-Williams earlier (198 la), provides the metaphor for
relating social identities to the two preoccupations reported by Biesele (1993) — sex
and food. From this we suggest that paintings of eland are more likely metaphors for
social categories of person than for shaman. We also suggest that, as shown by Bie-
sele (1993), the metaphors in both verbal and painted imagery refer to conflations that
include, but are not restricted to, trance. Trance, we believe, is the medium, the nego-
tiation of social identities the message, of western Cape rock paintings.

Figure 11. Two therianthropic figures, the one on the left is clearly female. Color red. Location: Horlosie-

kop

ROCK PAINTING OF THE WESTERN CAPE 319
Acknowledgments

We acknowledge with thanks the many contributions of Royden Yates to our
thinking on the meaning of western Cape rock paintings. Candice Goucher made
some very insightful comments on an earlier draft. Neither, of course, can be held ac-
countable for what we have written here.

Literature Cited

Barnard, A. 1992. Hunters and Herders of Southern Africa. A Comparative Ethnography of the
Khoisan Peoples. Cambridge University Press, Cambridge, U.K.

Biesele, M. 1993. Women Like Meat. Witwatersrand University Press, Johannesburg, South
Africa.

Bleek, D. 1923. The Mantis and his Friends: Bushman Folklore. Maskew Miller, Cape Town,
South Africa.

. 1928. The Naron: A Bushman Tribe of the Central Kalahari. Cambridge University
Press, Cambridge, U.K.

. 1931. Customs and beliefs of the /Xam Bushmen. Part I: Baboons. Bantu Stud.5:167-
79.

. 1932. Customs and beliefs of the /Xam Bushmen. Part II: The hon. Part Il: Game ani-
mals. Part 1V: Omens, wind-making, clouds. Bantu Stud. 6:47-63, 233-49, 321-42.

. 1933. Customs and beliefs of the /Xam Bushmen. Part V: The rain. Part VI: Rain-
making. Bantu Stud. 7:297-312, 375-92.

. 1935. Customs and beliefs of the /Xam Bushmen. Part VII: Sorcerors. Bantu Stud
9:1-47.

. 1936. Customs and beliefs of the /Xam Bushmen. Part VIII: More about Sorcerors and
charms. Bantu Stud. 10:131-62.

Douglas, M. 1966. Purity and Danger. Penguin Books, Harmondsworth, U.K.

Heinz, H.J. 1966. The social organization of the !Ko Bushmen. M.A. thesis. University of
South Africa.

Lewis-Williams, J.D. 1981a. Believing and Seeing: Symbolic Meanings in Southern San Rock
Paintings. Academic Press, London, U.K.

. 1981b. The thin red line: Southern San notions and rock paintings of supernatural po -
tency. S. Afr. Arch. Bull. 36:5-13.

. 1983a. The Rock Art of Southern Africa. Cambridge University Press, Cambridge, U.K.

. 1983b. Introductory essay: science and rock art. S. Afr. Arch. Soc. Goodwin Series
4:3-13.

. 1988. ‘People of the eland’: An archaeo-linguistic crux. Pages 203-11 in T. Ingold, D.
Riches & J. Woodburn, eds., Hunters and Gatherers 2: Property, Power and Ideology.
Berg Publishers, Oxford, U.K.

Lewis-Williams, J.D., & T.A. Dowson. 1989. /mages of Power: Understanding Bushman
Rock Art. Southern Book Publishers, Johannesburg, South Africa.

Lloyd, L. 1911. Specimens of Bushman Folklore. Allen, London, U.K.

McCall, D.F. 1970. Wolf Courts Girl: The Equivalence of Hunting and Mating in Bushman
Thought. Ohio University Occasional Papers in International Studies, Africa Series No.7.

Maggs, T.M.O'C. 1967. A quantitative analysis of the rock art from a sample area in the west-
ern Cape. S. Afr. J. Sci. 63:1 00-4.

320 PARKINGTON AND MANHIRE

—__. 197 1a. Microdistribution of some typologically linked rock paintings from the weste m
Cape. Pages 218-20 in H.J. Hugot, ed., Proceedings Sixieme Congres Panafricain de Pre-
histoire, Dakar, Senegal.

____. 1971b. Some observations on the size of human groups during the Late Stone Age. S
Afr. J. Sci. Spec. Publ. 2:49-53.

Maggs, T.M.O'C. & Sealy, J. 1983. Elephants in boxes. S. Afr, Arch. Soc. Goodwin Series
4:44-8.

Manhire, A., J. Parkington & W. Van Rujssen. 1983. A distributional approach to the interpreta -
tion of rock art in the south-western Cape. S. Afr. Arch. Soc. Goodwin Series 4:29-33.
Manhire, A., J. Parkington & R. Yates. 1985. Nets and fully recurved bows: rock paintings and

hunting methods in the western Cape, South Africa. World Arch. 17:161-74.

Noh, H.D. 1993. 4 Technical Investigation into the Material Evidence for Archery in the Ar-
chaeological and Ethnographical Record of Southern Africa. Ph.D. thesis. University of
Cape Town.

Panofsky, E. 1962. Studies in Iconology. Humanistic themes in the art of the Renaissance.
Harper and Row, New York.

Parkington, J. 1989. Interpreting paintings without a commentary: meaning and motive, con-
tent and composition in the rock art of the western Cape, South Africa. Antiquity 63: 13-26.

Parkington, J., A. Manhire & R. Yates. (In press) Reading San images. /n J. Deacon & T. Dow-
son, eds., Bleek and Lloyd 1870-1991. Volume 2. Witwatersrand University Press, Johan -
nesburg, South Africa.

Parkington, J., R. Yates, A. Manhire & D. Halkett. 1986. The social impact of pastoralism in
the south-western Cape. J. Anthrop. Arch.. 5:313-29.

Schater, H. 1974. Principles of Egyptian art. Clarendon Press, Oxford, U.K.

Schapera, I. 1930. The Khoisan Peoples of South Africa: Bushmen and Hottentots. George
Routledge and Sons, London, U.K.

Solomon, A. 1989. Division of the Earth: Gender, Symbolism and the Archaeology of the
Southern San. MA thesis. University of Cape Town.

. 199]. Gender, metaphor and meaning tn San art. Pages 138-43 in Current Perspectives
in South African Art and Architecture. Proceedings of the sixth annual conference of the
South African Association of Art Historians.

___. 1992. Gender, representation and power in San art and ethnography. J. Anthropol.
Arch. 11:291-329,

Vinnicombe, P. 1976. People of the Eland. Natal University Press, Pietermaritzburg, South
Africa.

Yates, R., J. Golson & M. Hall. 1985. Trance performance: The rock art of Boontjieskloof and
Sevilla. S. Afr. Arch. Bull. 40:70-80.

Yates, R. & A. Manhire. 1991. Shamanism and rock paintings: Aspects of the use of rock art in
the south-western Cape, South. Africa. S. Afr.Arch. Bull. 46:3-11.

Yates, R., A. Manhire & J. Parkington. 1993. Colonial era paintings in the rock art of the
south-western Cape: Some preliminary observations. S| Afr. Arch. Soc. Goodwin Series
7:59-70

_. 1994. Rock painting and history in the south-western Cape. Pages 29-60 in T.A. Dow-
son & J.D. Lewis-Williams, eds., Contested Images: Diversity in Southern African Rock
Art Research, Witwatersrand University Press, Johannesburg, South A frica.

Yates, R. & A. Smith. 1993. Ideology and hunter/herder archaeology in the south-western
Cape. S. Afr. Field Arch. 2:96-104.

Harnessing the Brain: Vision and
Shamanism in Upper Paleolithic
Western Europe

J.D. Lewis-Williams

Department of Archaeology
University of the Witwatersrand
Johannesburg 2050

South A frica

A glance down the list of chapters that constitute this volume reveals something of
the character of present-day Upper Paleolithic art research — or, as some writers prefer
to say, Upper Paleolithic “image-making” or, still more circumspectly, “mark-
making.” Two to three decades ago more researchers would have been interested in
the grand sweep of Upper Paleolithic art — despite some of its inappropriate connota-
tions, I retain the handy monosyllable. Art for art's sake, totemism, sympathetic
magic and binary oppositions, all explanations for the art as a whole, would have been
debated. Then, following the death in 1986 of the eminent French prehistorian Andre
Leroi-Gourhan, there was a feeling that research of this kind had attempted to fly too
high. Tacitly, a moratorium on explanation was declared. Certainly no one was pre-
pared to attempt another grand scheme to replace Leroi-Gourhan's and Annette
Laming-Emperaire's binary structuralist hypothesis (Leroi-Gourhan 1968; Laming-
Emperaire 1962); indeed, the very notion of an all-encompassing explanation was
discredited. Attention therefore began to focus on empirical issues.

Data and Explanation

West European Upper Paleolithic research has produced and still is producing
data that are indispensable to new hypothesis construction, and nothing I say here
should be seen as critical of good empirical work. Research requires both theory and
data. But, in this post-Leroi-Gourhan age, we have to guard against what Richard
Bradley (1985:86) has called Mr. Micawber archaeology: Keep at it long enough and
something is bound to turn up. When we have “enough” facts their meaning will be-
come evident; in other words, data precede and lead to theory. Empiricism, the philo-
sophical name for Micawberism is, however, not an infallible program for the

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copyright ©1997

322 LEWIS-WILLIAMS

production of explanations (for discussions of empiricism in rock art research see
Lewis-Williams 1984, 1989, 1990; Lewis-Williams & Loubser 1986; Tilley 1991).
We need some other methodological routes to the uncovering of ‘meaning.’

Insouthern Africa, a historical trajectory markedly different from the European
one led not to. a moratorium but to an intensification of work aimed at explaining rock
art. This difference between European and southern African research Is partially ex-
plained by the existence of directly relevant southern African ethnography. It was a
close reading of this ethnography that led to an understanding of the art that has, in the
first place, specific and detailed explanatory power and, secondly, broad methodo-
logical implications that extend beyond southern A frica.

In short, the demonstrably multiple fit between aspects of nineteenth- and
twentieth-century San ethnography and the highly detailed rock art images of the sub-
continent led to a concomitantly detailed explanation: San rock art was, at any rate in
large measure, associated with the beliefs, cosmology, experiences, diverse rituals,
notions of supernatural power, changing social relations, metaphors, and symbols of
San shamanism (e.g., Lewis-Williams 1980, 1981; Lewis-Williams & Dowson 1989;
Dowson 1994). At first there was some resistance to this view because two classic in
terpretations, art for art's sake and sympathetic magic (both inherited from European
Upper Paleolithic art research), were deeply entrenched in popular thought and politi-
cal strategies (Lewis-Williams 1995b). Today, however, the shamanic interpretation
is widely accepted, though healthy debate continues on just how much of the art is
shamanic and in what sense it is shamanic, and, further, on the nature of other mean-
ings that may be encoded in the art (see papers in Lewis-Williams & Dowson 1994).

But the southern A frican research trajectory did not end with that specific, ethno-
graphic explanation. Two simple yet crucial points with wider methodological imphi-
cations became apparent. First, all shamanism is posited on certain kinds of institu-
tionalized altered states of consciousness. Secondly, the nervous system that
produces those states is a human universal. Because all human beings have the same
nervous system, they have the potential to experience similarly structured visual,
aural, somatic, olfactory and gustatory hallucinations, though the meanings ascribed
to hallucinations and much of their content are, of course, culturally and historically
contingent (Knoll ef al. 1963; Klaver 1966; Bourguinon 1974; Eichmeier & Hifer
1974; La Barre 1975; Siegel & West 1975; Siegel 1977; Asaad 1980; Asaad &
Shapiro 1986). I argue that the southern African work that led to these two observa-
tions established a methodology that, together with the empirical work that is now be-
ing done in western Europe, will allow us to escape from a purely empiricist approach
to Upper Paleolithic art and enable us to essay a bold conjecture, or hypothesis, that is
empirically based and that can be evaluated (Lewis-Williams 1991).

The methodology to which I refer eschews the empiricist notion that research
moves linearly from data to explanation; instead, this methodology consists in the in-
tertwining of independent, mutually constraining and reinforcing strands of evidence
(Wylie 1989). In western European research, these strands include Upper Paleolithic
empirical data of many different kinds, neuropsychological research on altered states
of consciousness, and multiple analogical arguments that draw on a range of hunter-
gatherer ethnography and are founded on strong relations of relevance (Wylie 1982,
1985, 1988). Because this methodology has been published elsewhere (Lewis-
Williams 1991, 1995a), 1 do not rehearse its details here. In essence, it has informed
the argument that much Upper Paleolithic parietal imagery was, in some measure, in
some ways, shamanic.

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE

Ww
to
ee)

This explanation is, of course, not new; it has long been in the air (e.g., Reinach
1903; Levy 1963; Lommel 1967; La Barre 1970; Eliade 1972; Marshack 1972:280,
1976:278-9; Eichmeier & Hifer 1974; Furst 1976; Halifax 1980:3, 17; Pfeiffer 1982;
Hedges 1983; Bednarik 1984, 1986; Bahn & Vertut 1988:157-8; Goodman 1988;
Smith 1992). Yet it has never been generally accepted, as were art-for-art's-sake,
sympathetic magic or, more recently, Leroi-Gourhan's and Laming-Emperaire's
structuralist interpretations. The reasons for the relegation of the shamanic interpreta-
tion to the periphery of academic interest are no doubt numerous and complex. I sim-
ply mention four. First, it has lacked a powerful, articulate and charismatic proponent,
like the Abbe Henri Breuil, who remained for many decades committed to the sympa-
thetic magic explanation. Secondly, in more recent times the position has run counter
to an important trend in Upper Paleolithic research — adaptation. Rather than granting
social, individual and, much worse, psychological factors a prominent formative po-
sition, adaptationist writers have preferred to see the making of art as a response to
Upper Paleolithic climatic conditions and resultant demography. Thirdly, the sha-
manic view has never meshed as neatly with contemporary Western philosophical
thought as did (at least for a while) Leroi-Gourhan's and Laming-Emperaire's struc-
turalism. Finally and most importantly, the shamanic view has never been fully devel-
oped, its exact implications made clear and the arguments in favor of it — especially
those recently derived from neuropsychological research — fully set out.

I begin this broad overview by listing the principal features of shamanism; this
provides some overall orientation for the sections that follow. Then, having given rea-
sons for suspecting some sort of shamanism in the Upper Paleolithic, I assess the po-
tential of the explanation by considering four specific features of the underground
parietal imagery of southwestern France. Next, I comment briefly on some of the
ways in which Upper Paleolithic shamans exploited the widely varying topographies
of the caves. Finally, I consider, in a preliminary way, the role of shamanism in the
changes that took place during the Upper Paleolithic.

Shamanism

Although societies practicing other modes of production sometimes evince ele-
ments of shamanism (e.g., the Maya; Freidel et a/. 1993), I narrow the field to hunter-
gatherers and propose ten central characteristics of shamanism as it is practiced in
such societies.

In the first place, hunter-gatherer shamanism is fundamentally posited on a range
of institutionalized altered states of consciousness. Secondly, the visual, aural and so-
matic experiences of altered states of consciousness give rise to conceptions of an al-
ternative reality that is frequently tiered. Thirdly, people with special powers and
skills, the shamans, are believed to have access to this alternative reality. Fourthly, the
behavior of the human nervous system in certain altered states creates the illusion of
dissociation from one's body (less commonly understood in hunting and gathering
shamanic societies as possession). Shamans use dissociation and other experiences of
altered states of consciousness to achieve at least four ends; these ends constitute the
next four features of hunter-gatherer shamanism. Shamans are believed to contact
spirits and supernatural entities; they heal the sick; they attempt to control the move-
ments and lives of animals; they are believed to have the ability to change the weather.
Ninthly, these four functions of shamans, as well as their entrance into an altered state
of consciousness, are believed to be facilitated by a variously conceived supernatural

324 LEWIS-WILLIAMS

potency, or power. Lastly, this potency is commonly associated with animal-helpers
that assist shamans in the performance of their tasks.

In compiling these ten characteristics of hunter-gatherer shamanism I have
ex-cluded features that some writers consider important, 1f not essential, for the clas-
sification of a religion as shamanic. I do not, for instance, link shamanism to mental
illness of any sort, though some shamans may well suffer from epilepsy, schizophre-
nia, migraine and a range of other pathological conditions. Nor do I stipulate the
number of religious practitioners that a shamanic society may have; some societies
have many, others only a few, often politically powerful, shamans. Nor do I stipulate
any particular method or methods for the induction of altered states of consciousness.
Altered states may be induced by ingestion of psychotropic drugs, rhythmic and
audio driving, meditation, sensory deprivation, pain, hyperventilation, and so forth.

If we allow only the ten distinguishing features I have given, the word ‘shaman’
can be freed from its central Asian Tungus origin and be more widely applied (c/. Eli-
ade 1972: Lewis-Williams 1992), even, I argue, to the Upper Paleolithic of western
Europe.

Upper Paleolithic Shamanism

There are two major reasons for suspecting some form of shamanism in the Upper
Paleolithic.

First, there is evidence to suggest that the ability of the human nervous system to
enter altered states and to generate hallucinations is of great antiquity. We do not, of
course, know exactly what chimpanzees, baboons, monkeys, cats and dogs experi-
ence, but they and many other creatures apparently do hallucinate, not only when psy-
chotropic drugs are administered to them but also sometimes under natural
circumstances (Siegel & Jarvik 1975:81-104). The ability to hallucinate is therefore
probably a feature not just of the human but of the mammalian nervous system. Con-
sequently, it seems likely that australopithecines hallucinated, highly probable that
Neanderthals hallucinated, and certain that at least some of the anatomically modern
human beings of the Upper Paleolithic also hallucinated. What australopithecines,
Neanderthals and Upper Paleolithic people made of their hallucinations is another
question altogether.

Secondly, I point to the ubiquity of shamanism among hunter-gatherer communi-
ties (see, among many others, Eliade 1972; Winkelman 1986; Harner 1973; Vitebsky
1995). Whatever differences there may be, hunter-gatherer communities throughout
the world and on all continents have religious practitioners who enter altered states of
consciousness to perform the tasks I have listed. The widespread occurrence of sha-
manism results not from diffusion but (in part) from universal neurological inheri-
tance that includes the capacity of the nervous system to enter altered states and the
need to make sense of the resultant hallucinations within a foraging community.
There seems to be no other explanation for the remarkable similarities between sha-
manic traditions worldwide. It is therefore probable that some form of shamanism —
not necessarily identical to any ethnographically or historically recorded type of sha-
manism — was practiced by the hunter-gatherers of Upper Paleolithic Europe.

That probability is increased by a more specific consideration of some of the fea-
tures of Upper Paleolithic imagery. We need to know if components of the imagery of
Upper Paleolithic art have anything in common with the mental imagery of altered

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 325

states of consciousness. I leave aside mobile and open air art and concentrate on parie-
tal cave art.

Entirely independently of work on Upper Paleolithic art, neuropsychological
re-search on altered states of consciousness has identified types of mental imagery
and the sequence in which the types are often, but not ineluctably, experienced (see
Lewis-Williams & Dowson 1988 for an overview). In the first and ‘lightest’ stage of
altered consciousness people may experience geometric visual percepts that include

dots, grids, zigzags, nested catenary curves and meandering lines (e.g., Klaver 1926;
Eichmeier & Hifer 1974; Siegel & Jarvik 1975; Asaad & Shapiro 1986). These per-
cepts flicker, scintillate, expand, contract and combine with one another; they are in-
dependent of light from an exterior source. They can be experienced with the eyes
closed; with open eyes, they are projected onto and partly obliterate veridical visual
perception. Because they are wired into the human nervous system, all people, no
matter what their cultural background, have the potential to experience the same
forms. They are known variously as phosphenes, form constants and entoptic phe-
nomena (on nomenclature, see Tyler 1978). As subjects go further, or deeper, into al-
tered consciousness they enter a second stage in which they sometimes try to make
sense of these geometric forms by seeing them as objects or experiences familiar or
important to them (Horowitz 1964:514; 1975:177, 178, 181). This process of con-
strual is, unlike the forms themselves, culturally situated. The third and deepest stage
is frequently entered via a vortex or tunnel (Siegel 1977; Horowitz 1975:178). In this
stage, the geometric forms are peripheral to but sometimes combined with iconic im-
ages of animals, people, monsters and highly charged emotional situations (Siegel &
Jarvik 1975:127, 143; Siegel 1977:136). The animals, people and objects seen in this
stage are often distorted in various ways, but they can also have a startling, life-like ‘r-
eality.’ Eventually, subjects become part of their own imagery, and they feel them-
selves to be blending with both geometric and iconic imagery (Klaver 1942:181,
182). It is in this final stage that people sometimes feel themselves to be turning into
animals (e.g., Siegel & Jarvik 1975:105) and undergoing other frightening or exalting
transformations.

The utility of this neuropsychological model to identify graphic imagery derived
from altered states of consciousness may be assessed by slotting it against arts known
ethnographically, and therefore independently, to be associated with shamanic al-
tered states of consciousness, such as southern A frican San rock art (Lewis-Williams
1980, 1981; Lewis-Williams & Dowson 1989), the rock art of the Cosos (Great Basin,
North America; Hedges 1982; Whitley 1987, 1988, 1992, 1994a, 1994b), the art of
the Tukano (South America; Reichel-Dolmatoff 1972, 1978) and Huichol art (Cen-
tral America; Berrin 1978). In the iconography of each of these four arts there are geo-
metric motifs isomorphic with the geometric forms of stage one, construed
geometrics referable to stage-two, and stage-three iconic images that are sometimes
therianthropic and sometimes combined with geometric forms (Lewis-Williams &
Dowson 1988). Arts known independently to be associated with shamanic altered
states of consciousness thus display a complex set of features — not merely geometric
motifs — that fits the set of features established by laboratory research for the mental
imagery of altered states of consciousness. If the model is tested against an art not as-
sociated with altered states of consciousness, say Rembrandt’s work, no such fit will
be found (cf Dronfield 1994, 1995).

Having supported the utility of the neuropsychological model, we can apply it to
Upper Paleolithic parietal art, an art that is, of course, not known a priori to be sha-
manic. Because this has been done in detail elsewhere (Lewis-Williams & Dowson

326 LEWIS-WILLIAMS

1988, 1992; Lewis-Williams 1991), 1 give only a briefsummary of the main points. In
the first place, many of the so-called ‘signs’ of Upper Paleolithic art are referable to
one or other category of stage-one geometric mental percepts: dots, zigzags, grids,
meandering lines and, less frequently, nested catenary curves (so-called festoons) oc-
cur in Upper Paleolithic parietal art. On the other hand, some Upper Paleolithic motifs
that are at present usually classified as ‘signs’ do not appear to be derived from geo-
metric mental percepts. These include spearlike forms and the so-called claviforms
and tectiforms (Lewis-Williams & Dowson 1988:Figure 3). What these excluded
motifs may represent does not concern us at the moment, but it needs to be said that
their exclusion points to a strength rather than a weakness in my argument because It
renders invalid the potential criticism that virtually any mark can be interpreted as
having been derived from a geometric mental percept.

In any event, the fit between neuropsychologically identified elements of mental
imagery and the motifs of Upper Paleolithic parietal art does not end with geometric
forms. Upper Paleolithic images referable to stage two and, especially, stage three are
also present. By their very nature, stage-two construals are difficult to identify be-
cause, if the construal is far advanced, the original geometric form is masked by the
representational image into which it is transformed. Possible Upper Paleolithic exam-
ples of construal include depictions of ibexes with greatly exaggerated curved horns
that recall the nested catenary curves of stage one. Again, it must be emphasized that
the meaning or meanings of such depictions is another issue altogether. Stage-three
Upper Paleolithic images are clearer. They include therianthropes, anthropomorphs
and, of course, animals, many of which are superimposed on or by and juxtaposed
with geometric forms — a characteristic of stage-three mental imagery. Sometimes
depictions show distorted animals and ‘monsters,’ but many images are, on the other
hand, markedly ‘realistic’ (Clottes e¢ al. 1994). Comparable realism also character-
izes much southern A frican San rock art (Vinnicombe 1976; Lewis-Williams 1981):
yet a small, easily overlooked feature sometimes suggests that a San image does not
depict a “real” animal, as a superficial glance may suggest (Lewis-Williams & Dow-
son 1989). It is important to note that the argument I am outlining does not concern
geometric motifs alone. It is the presence in Upper Paleolithic parietal art of imagery
referable to all three stages that makes the argument compelling (Lewis-Williams
1991).

In sum, the antiquity and ubiquity of altered states of consciousness, the
widespread occurrence of shamanism among hunter-gatherers, and formal parallels
between elements of the mental imagery of altered states and Upper Paleolithic parie-
tal imagery are three points that suggest that at least some — not necessarily all — parie-
tal art was probably associated with institutionalized hallucinations. In other words, it
seems highly probable that some yet to be precisely defined forms of shamanism were
present at, probably, all periods of the Upper Paleolithic of western Europe. This very
general conclusion contributes significantly and economically (in the sense of eco-
nomical explanations) to an understanding of numerous otherwise puzzling features
of Upper Paleolithic art. I refer to four such features.

Explanatory Power

The first has exercised researchers for a long time. It is indeed one of the key fea-
tures of Upper Paleolithic parietal art. Frequently, depictions of animals are adjacent
to, superimposed by or superimposed on geometric ‘signs.’ Panels, such as the Apse

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 327

at Lascaux (Leroi-Gourhan & Allain 1979) or parts of the Sanctuary at Les Trois
Frares (Begouan & Breuil 1958), are so densely covered with both representational
and geometric images that some sort of relationship between the two types of image
seems inescapable. At one time it was thought that representational imagery evolved
out of geometric imagery, but it is now accepted that both types continued to be made
throughout the Upper Paleolithic. The contemporaneity of the two kinds of image has
led to the suggestion that they derived from two distinct, parallel, perhaps comple-
mentary, graphic systems (e.g., Leroi-Gourhan 1968; Marshack 1972). In terms of
this view, geometric and representational images are comparable to text and graphs in
a book: both may be saying the same thing, though in different ways. Although there
may be an element of truth in this two-systems explanation, it does not explain why
people maintained the two systems throughout the Upper Paleolithic, why the two
types of image crowd one on top of another in a way that baffles and confuses modern
Western viewers, nor, significantly, why both geometric and representational images
are found in shamanic arts throughout the world.

The answer to the problem is, I argue, straightforward. As we have seen, in certain
altered states, the human nervous system produces these two kinds of images — geo-
metric and representational —and they are sometimes superimposed and blended with
one another (Siegel 1977:134). The two types are therefore not as different as they ap-
pear to modern Westemers: both derive from the same source — the nervous system in
certain altered states. Their intimate association in a shamanic art 1s therefore not a
surprise; on the contrary, as the neuropsychological model I have outlined suggests, it
is to be expected.

This observation resolves the postulated dichotomy between geometric and repre-
sentational Upper Paleolithic imagery that has formed the basis and starting point for
all classifications of Upper Paleolithic art hitherto devised. Both types of image are in
fact representational, for they both represent (by way of complex processes) mental
imagery. Ifa taxonomic distinction is to be drawn between them, it should be posited
on the different stages of altered consciousness in which the nervous system gener-
ates them and the iconographic contexts which those stages create, not on a supposed
distinction between their “representational” and “non-representational” functions.

The second feature of Upper Paleolithic art that the shamanic hypothesis explains
is the placing of some parietal images deep underground. Some images are in or near
entrances to caves, in shallow rock shelters or in the open air (Bahn 1995) and must
therefore have been well or partially lit by natural light; during the Upper Paleolithic,
before weathering processes took their toll, there were doubtless many more than
there are today. Other images, especially (but not exclusively) those of the Mag-
dalenian, are found a kilometer and more underground in totally dark and silent cham-
bers and passages. Sometimes these images are in small diverticules and chimneys
that are hard to find and to which access is sometimes extremely difficult. Moreover,
the danger involved in getting to some remote locations should not be underesti-
mated: Deep chasms have to be avoided, and labyrinthine passages have to be negoti-
ated. Upper Paleolithic people had, of course, to traverse these routes assisted by only
flickering torches or tallow lamps. These remote, subterranean images have to be
seen in the context of shamanic cosmology.

To speak of shamanism in the Upper Paleolithic is to make a statement about
cosmology, not just religious belief. All life, economic, social and religious, takes
place within and interacts reciprocally with a cosmology. As I have mentioned, the
shamanic cosmos is conceived, in the first instance, as comprising two realms, this
world and a spirit world. Often, the spirit world is immanent, interdigitating with the

328 LEWIS-WILLIAMS

real world. At the same time, these two realms are frequently conceived of as subdi-
vided and layered, the more complex the shamanic society, the more tiered subdivi-
sions of the cosmos (Eliade 1972). I argue that the widespread shamanic notions of a
spirit world and a layered cosmos derive, in the first instance, from experiences of al-
tered states of consciousness that are universal: They include a whole range of hallu-
cinations and dreams as well as sensations of flying, rising up, entering a vortex, and
passing underground and through water. Altered states of consciousness thus not only
create notions of a tiered cosmos; they also afford access to and thereby validate the
various divisions of that cosmos. The shamanic cosmos and “proof” of that cosmos
constitute a closed system of experiential creation and verification. Universal neuro-
logically generated and verified experiences account, at least in part, for the ubiquity
of the shamanic cosmos in one manifestation or another.

The layers of the shamanic cosmos are accessible to shamans as they explore their
altered states of consciousness in pursuance of their various tasks. The route of their
explorations, what anthropologists sometimes call the axis mundi (e.g., Eliade 1972),
is conceived of differently in different shamanic societies: it may, for instance, be
thought ofas a hole in the ground at the back ofa shaman's dwelling, the roots ofa tree
or, significantly, acave. Always it is some sort of tunnel, a notion that derives, I argue,
from the neurologically produced vortex that seems to draw subjects into the third and
deepest stage of trance, the stage in which they inhabit a vivid, hallucinatory, but for
them intensely real, world. I argue that the neurologically generated concepts of a
tiered cosmos and entry into a vortex or tunnel make it highly probable that entry into
a cave was, for Upper Paleolithic people, entry into part of the spirit world.

Further, the constriction and sensory deprivation of narrow subterranean
passages may not only have replicated the vortex; they may also have contributed to
the induction of altered states. Under such circumstances, the experience of the pas-
sage and the experience of the vortex may have become inextricably interwoven as
‘spiritual’ experiences were given topographical materiality. Here is one of the rea-
sons why Western binary oppositions of spirituality and reality, sacred and profane, a
material realm and a nonmaterial realm, together with all their various permutations,
probably did not obtain in the Upper Paleolithic.

The subterranean passages and chambers were therefore places that afforded
close contact with, even penetration of, a spiritual, nether tier of the cosmos. The im-
ages that people made there related to the chthonic world. Images were not so much
taken underground and ‘placed’ there as obtained there and fixed there. The halluci-
natory, or spirit, world, together with its painted and engraved imagery, was thus in-
vested with materiality and precisely situated cosmologically. Moreover, acts of
image making did not merely take place in the spirit world: They also informed and
incrementally created that world. There was thus a complex interaction between the
topography of the caves, mental imagery and historically situated image making by
individuals and groups that, through time, built up and changed the spirit world both
conceptually and materially.

This understanding brings me to a third and related feature of Upper Paleolithic art
that is difficult to explain outside of the shamanic hypothesis. One of the best known
and most consistent features of Upper Paleolithic art is the use that artists made of fea-
tures of the rock surfaces on which they placed their images (e.g., Graziosi 1960;
Ucko & Rosenfeld 1967; Bahn & Vertut 1988). Almost every cave contains exam-
ples; I cite but a few.

At Labastide, for instance, the natural contours of the rock provide the dorsal line
of a bison; the rest of the animal has been suggested by the addition of a few features

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 329

(Figure 1; Omnas 1982: Figure 154; see also Figure 147, PI.XIX, no.1). Upper Paleo
lithic depictions are also often placed so that a small, seemingly insignificant nodule
or protuberance forms the eye ofan animal. Some of these nodules are so insignificant
that one supsects that they were identified and selected by touch rather than by sight.
Fingers lightly exploring the walls may have discovered a nodule, and the mind, pre-
pared for the discovery of animals, took it to be an eye.

On a larger scale, a natural rock shape at Comarque seems to have suggested a
remarkably realistic horse’s head, complete with nostrils and mouth (Leroi-Gourhan
1968: Figure 13); an engraver completed and added details to the form. Human fig-
ures occasionally also make use of natural features of the rock. At Le Portel, for exam-
ple, a red outline human figure is painted so that a protuberance becomes its penis
(Bahn & Vertut 1988: Figure 52).

The importance of natural features of the rock is particularly clear at Castillo
where a depiction of a bison has been painted to fit undulations in the surface of a sta-
lagmite: the back, tail and hindleg of the depiction fit the shape of the rock. But in or-
der to use the rock in this way the artist had to position the bison vertically (Figure 2;
Bahn & Vertut 1988: Figure 53). In doing so, he or she expressed a highly significant
difference between the animals of mental imagery and the animals of the real world
that, in the nature of things, normally assume horizontality. What was important to the
Castillo artist was fitting the bison into the natural features of the rock, not orienting
the image so that it would call to mind a real, standing bison.

So far, I have described depictions that present, principally, lateral views of
animals. By contrast, in the Salon Noir at Niaux an artist added antlers to a hole in the

Ficure |. Labastide (Hautes-Pyrenees, France). Natural rock shape used as dorsal line ofa bison. Redrawn
from Omnis (1982: Figure 154)

330 LEWIS-WILLIAMS

rock that looks something like the head of a deer as seen face on (Figure 3; Clottes
1995: Figures 142 and 164). At Altamira, in one of the deepest sections of the cave,
natural shapes in the rock have been transformed by the addition of painted eyes and,
in one case, a black patch that may represent a beard (Leroi-Gourhan 1968: Figures
402-404; Freeman ef al. 1987:224-233). The same technique has been employed at
Gargas (Breuil 1952:Figure 271). At Montespan, a natural rock formation has been
similarly transformed into what appears to be an animal head (Leroi-Gourhan 1968:
Plate 36). The same effect has been achieved at Rouffignac where a remarkable
horse's head has been painted on a flint nodule that juts out from the wall of the cave;
the body of the horse seems to be behind the wall (Figure 4). Fora final example I cite
a depiction in the Chimney at Bernifal: Eyes, nostril and mouth have been added to a
natural edge to produce a human face. The effect created by all these images is of hu-
man and animal faces looking out of the rock wall, the rest of the bodies being con-
cealed behind the surface of the rock. The figures are not merely painted onto the
surface: they become part of and, at the same time, construct the walls of the caves in
specific ways.

All the examples I have so far described point to an interaction between the maker
of an image and natural features of the rock face. My last examples are especially im-
portant because they imply interaction between, not just makers of images and their

FiGuRE 2. Castillo (Santander, Spain). Natural shape of stalagmite used to give form toa vertical bison
Redrawn from Ripoll, in Bahn and Vertut (1988: Figure 53)

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 331

seer

ed
~

t

wa

F;
Z
.

fd

hae

US

"ld

Ficure 3. Niaux (Midi-Pyrénées, France). Antlers added to natural cavity to give effect of a deer seen from
the front. Redrawn from Clottes (1995: figs 120, 142, 164)

handiwork, but also between viewers and images. Sometimes an undulation in the
rock surface becomes the dorsal line ofan animal if one’s light is held ina specific po-
sition; an artist has added legs and some other features to the shadow. By moving
one’s lamp the image can be made to disappear and reappear. At Niaux, for example,
an undulation in the rock has been used as a bison’s back; this is especially clear when
the light source is held to the left of and somewhat below the image. An artist has
added a head, legs, a belly line anda tail. But, like the bison at Castillo, this Niaux ani-
mal is positioned vertically in order to exploit natural features of the rock (Clottes
1995: Figures 177 and 180). On a larger scale, the head of one of the well known
“spotted horses” at Pech Merle is suggested by a natural feature of the rock, especially
when the source of light is in a certain position (Figure 5). But, in this case, the artist
distorted the painted horse's head, making it grotesquely small; the rock shape 1s in
fact more realistically proportioned than the painted head. It is as though the rock sug-
gested “horse,” yet the artist painted not a “real” horse but a distorted horse, perhaps a
“spirit-horse.”” Many further examples could be given. Indeed, Freeman (Freeman ef
al. 1987:105) points out that the technique of using shadows to complete a depiction
“is more common than is usually supposed.”

An important reciprocality is implied by these images born of shifting chiaro-
scuro. On the one hand, the creator of the image holds it in his or her power: A move-
ment of the light source can cause the image to appear out of the murk; another
movement causes it to disappear. The creator is master of the image. On the other
hand, the image holds its creator in its thrall: if the creator (or subsequent viewer)

vs)
[vs)
i)

LEWIS-WILLIAMS

xs

Ficure 4. Rouffignac (Dordogne, France). Horse’s head painted on a flint nodule that juts out of the cave
wall. Redrawn from Graziosi (1960: Plate 202d)

wishes the image to remain visible, he or she is obliged to maintain a posture that
keeps the light source ina certain position. Relax, and the image retreats into the Sty-
gian realm from which it was coaxed. Perhaps more than any other Upper Paleolithic
images, these “creatures” (creations) of light and darkness point to a complex interac-
tion between person and spirit, artist and image.

All the intimate and complex relationships between images and rock surfaces that
I have described, together with the placing of these images in subterranean locations,
are understandable in the light of ethnographically known practices. In some sha-
manic societies, such as those of North America, but not in others, such as that of the
San, the notion ofa vision quest is crucial to becoming a shaman and, sometimes, re-
peated vision quests are required to sustain a shaman’s power. A North American
quester usually repairs to a remote, isolated place, sometimes a high cliff-top, some-
times a cave (Eliade 1972:50-51, 110-14; Halifax 1980:6), to fast, meditate and in-

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 333

on Bs \
ey 9°
a SE:
2 F) ‘
axe a %, ae eg ;

© aide ars wk

%2g\9

tH
oP oy

Ficure 5. Pech Merle (Lot, France). Natural shape to right (broken line) suggests a horse’s head. Redrawn
from Leroi-Gourhan (1968: Figure 64), Graziosi (1960: Plate 196), and other photographs

duce the altered state of consciousness in which he or she will “see” the animal helper
that will impart the power necessary for shamanic practice. When the vision of the
sought-after animal comes, it bestows shamanic power. It is a vision, it should be
noted, that bestows the power, not a real animal.

Upper Paleolithic evidence suggests that parts of the caves, especially the deep
passages and small, hidden diverticules, were places where vision quests took place
(cf. Hayden 1993; Pfeiffer 1982). Certainly, the sensory deprivation afforded by the
remote, silent and totally dark chambers, such as the Chamber of Felines in Lascaux,
induces altered states of consciousness (La Barre 1975:14: Walker 1981:146; Pfeiffer
1982:211; Siegel & Jarvik 1975). In their various stages of altered states, questers
sought, by sight and touch, in the folds and cracks of the rock face visions of power-
full animals. It is as if the rock was a membrane between them and one of the lowest
levels of the tiered cosmos; behind the rock lay a realm inhabited by spirit-animals,
and the passages and chambers of the cave penetrated deep into that realm.

Such beliefs and rituals also account for the fourth feature of Upper Paleolithic art
to which I refer: the various ways in which the walls of numerous Upper Paleolithic
caverns were touched and otherwise treated. In some sites, such as Grotte Cosquer,
finger-flutings cover most of the walls and parts of the ceilings to a considerable
height (Clottes et a/. 1992:586; Clottes & Courtin 1994). At Hornos de Ja Pena the
so-called finger-flutings are more restricted in distribution: A finger-traced ‘grid’
surrounds and appears to issue from a natural cavity (Ucko 1992: Plate 9). In another
remarkable treatment of cave walls, cavities at Hornos de la Pena were filled with
mud and then punctured, apparently with fingers or sticks (Ucko 1992:158, Plates 10
and 11). Ucko remarks, “It is ... inconceivable to us today to understand the nature of
such action” (ibid.). But, if we allow that Upper Paleolithic people believed that the
spirit world lay behind the thin, membranous walls of the underground chambers and
passages, the evidence for this and much otherwise incomprehensible behavior can
be understood in rational, if not absolutely precise, terms. Ina variety of ways, people
touched, respected, painted and otherwise ritually treated the cave walls because of

334 LEWIS-WILLIAMS

what existed behind their surfaces. The walls were not a meaningless support. They
were part of the images, a highly charged context.

The four features of Upper Paleolithic art that I have discussed are thus
economically explained by a hypothesis that invokes altered states of consciousness
and a shamanic cosmos. Moreover, features that may otherwise be considered unre-
lated are shown, economically, to be manifestations of a single religious activity —
some form of shamanism.

Space and Time

As I pointed out at the beginning of this chapter, there is today an understandable
reluctance to accept explanations that purport to cover the entire span of Upper Paleo-
lithic art in all its diversity. Such comprehensive explanations are rightly dismissed as
“monolithic.” Conkey (1987:414-415) remarks that “there can no longer be a single
“meaning” to account for the thousands of images, media, contexts, and uses of what
we lump under the term ‘Paleolithic art’.” I have therefore confined my discussion to
west European cave art and have not considered portable art or the arts of other re-
gions. But it also needs to be shown that the shamanic explanation, even within west-
ern Europe, is sensitive to temporal, geographical, social and iconographic diversity.
Indeed, the shamanic explanation is an effective tool for uncovering and exploring
such diversity, not for concealing it. To demonstrate the heuristic potential of this ex-
planation and its sensitivity to diversity, and so refute a potential charge of mono-
lithism, I briefly consider, first, geographic diversity and then temporal diversity.

Individual Caves

Upper Paleolithic people did not “process” caves ina rigid, formularistic way, as
Leroi-Gourhan (1968) suggested. Rather they explored and adapted each cave in ac-
cordance with its peculiar topography (cf. Vialou 1982, 1986) and, most importantly,
the particular expression of shamanic cosmology and social relations as it existed at a
given time and ina given region. At the same time, we must recognize that the prehis-
toric entrances to caves are not always known. Moreover, in some caves, the distribu-
tion of imagery as it is found today is a palimpsest resulting from changing
exploitation of the topography of the cave and changing shamanic beliefs and social
relations, as, for example, the dating of the recent discoveries at Grotte Cosquer
shows (Clottes er a/. 1992). Italso seems that, even at a given time, Upper Paleolithic
people did not adhere rigidly to a single set formula; as in many shamanic societies,
there were always idiosyncratic interventions (cf: Dowson 1988). In some instances
new dating techniques (see Watchman and Clottes, this volume) will sort out what is a
function of the passage of time and what is contemporary idiosyncrasy, but the
chronological resolution in many instances will not be fine enough to make the neces-
sary discriminations.

Le Gabillou (Dordogne) serves as a fairly simple example of how the distribution
of imagery within caves may be understood in terms of the shamanic explanation.
This cave is said to date to the early Magdalenian and so to be contemporary with Las-
caux (Breuil 1952:310-311; Gaussen 1964; Leroi-Gourhan 1968:317). Unlike many
caves that are divided into complex interlinking passages, chambers and diverticules,
Le Gabillou comprises only an entrance chamber, that was probably at least partially
open to natural light in prehistoric times, and a single straight tunnel that extends from

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 335

the entrance chamber for approximately 33 yards (Leroi-Gourhan 1968:317;
Gaussen et ail. in L’ Art des Cavernes 1984:225-231). The two parts of the cave need to
be considered separately.

It seems that the entrance chamber (now partly destroyed by the construction of a
cellar) was decorated with carefully painted images as well as some “simpler” en-
gravings. Some of the painted images may have been communally produced, various
classes of people participating ritually and differentially, perhaps hierarchically, in
the procurement of pigment and binders, mixing the paint and applying it to the rock
(Lewis-Williams, 1995c; see also Vandiver 1983). This seems almost certainly to
have been the case in Lascaux, where the remains of scaffolding have been found near
very large paintings (Leroi-Gourhan & Allain 1979). The entrance chamber at Le Ga-
billou may therefore have been a kind of vestibule in which group ceremonies were
performed. These communal ceremonies and the paintings themselves may, on some
occasions, have prepared the minds of vision questers for what they were to see at the
climax of their initiation. The culturally informed component of the deepest stage of
trance derives from memory, and the novices were being shown not just pictures of
animals but recreations of spirit-animals of the kind that they themselves hoped to
see. Their memories were being stocked with desirable images. The entrance cham-
ber was therefore a staging post on the route from this world to the lower tier or tiers of
the cosmos.

The tunnel that leads off the entrance chamber at Le Gabillou was excavated in
recent times; during the early Magdalenian visitors to the passage were obliged to
crawl, and movement was very restricted. The passage images are engraved, often
with only a few, sure strokes. There are no elaborately painted images in the passage,
although there do appear to be some patches of ocher that may be the remains of paint.
The comparatively “simple” execution of the passage images implies that less time
was expended on them than on the imagery in the entrance chamber. If, as | suggest,
communal, preparatory rituals took place in the entrance chamber, it seems that soli-
tary vision quests were undertaken in the narrow tunnel, where questers, isolated
from the community, made more swiftly executed engravings.

The engravings in the tunnel are, moreover, strung out along its entire length.
Some sections of the tunnel are slightly more densely engraved than others, but, al-
though some images are overlain by other marks (said by some to be “magical
strokes”; Breuil 1952:311), no section is nearly as densely engraved as, say, the Apse
at Lascaux.

During the Magdalenian, the Apse was reached by means of a painted and
engraved 15 m passage in which it was necessary to crawl or crouch very low (Leroi-
Gourhan & Allain 1979: Figure 42). This passage led from the large, richly decorated
Hall of the Bulls, into which the prehistoric entrance almost certainly led and in which
communal rituals probably took place, past the Apse and on to the Nave and, finally,
the very narrow Chamber of Felines, the most remote part of the cave (Leroi-Gourhan
& Allain: Figure 313). The Apse was thus an alcove, or side-chamber, on the way
from the Hall of the Bulls to the Chamber of Felines, a route that may have been fully
traversed only rarely. The Apse is moreover, situated above the Shaft (Leroi-Gourhan
& Allain: Figure 53), at the bottom of which is the celebrated group comprising an ap-
parently bird-headed anthropomorph, a “wounded” bison, what may be a bird-headed
staffand other images (for a shamanic interpretation of this “scene” see Davenport &
Jochim 1988; some researchers now question whether the Apse was connected to the
Shaft during the Magdalenian).

336 LEWIS-WILLIAMS

In the Apse there seems to have been a determined and sustained effort to place
principally engraved but also some painted images ina single area. The complexity of
the superimpositions in the comparatively small Apse suggests that the people who
made the images were sharing, or desired to share, the acquisition of a special, topo-
graphically situated power (Leroi-Gourhan & Allain 1979: 220-88). This power was,
I argue, related in complex ways to the different kinds of images in the other, topo-
graphically and iconographically distinct, areas of Lascaux.

By contrast, the separated images in the tunnel at Le Gabillou suggest that the
acquisition of visions there was of a more individual nature: The people at Le Gabil-
lou did not seek to associate their visions intimately with the visions of others. The
different ways of placing imagery in the Apse and the Le Gabillou tunnel thus point to
different kinds of social relations between vision questers themselves and, by exten-
sion, between shamans and the wider community.

The approach that I advocate is thus able to suggest the different uses to which
parts of Le Gabillou (and other caves) may have been put and how various parts of the
cave related to the shamanic cosmos.

Image and Change

Social contexts are, of course, not immutable. As the Upper Paleolithic pro-
gressed, the historically and geographically situated forms of shamanism changed,
and the associated practice of making imagery came to be linked to growing social
complexity. Evidence for such change may be found in Upper Paleolithic art and its
placement in the caves.

At the beginning of the Upper Paleolithic, Aurignacian “artists” engraved and
painted their images in the entrances to caves or, at any rate, as far in as light pene-
trated, though the deep art at La Grotte Chauvet shows that much variation existed
even at that time (Chauvet eral. 1995). At this time, access to religious imagery (if not
the actual experience) may have been open to all, and the shamanic cosmos may not
have been conceived of as tiered. Later, they ventured into totally dark depths, and,
certainly by the Magdalenian a few people were traveling considerable distances un-
derground in order to make images, sometimes squeezing through narrow openings
or scaling slippery chimneys to reach hidden niches (e.g., Bahn & Vertut 1988; Ucko
& Rosenfeld 1967). The leaving of daylight areas to go deeper into the dark caves
suggests growing interest in and insistence on spatially distinct ritual areas. The
source of increasingly esoteric religious knowledge was being placed farther and far-
ther away from daily life: The deeper into the caverns, the more restricted the access
to altered states of consciousness or, more precisely, to those altered states that were
defined as the ones that really mattered, a tendency that probably paralleled growing
social differentiation (cf. Bender 1989; Thomas 1991) and increasing cosmological
complexity. The spectrum of altered states of consciousness, ranging from ‘light’
euphoria to deeply hallucinatory conditions, was being divided up and socially allo-
cated. Upper Paleolithic shamanism was developing through time.

In many instances, it seems that the more structured the social dimension, the
more likely it is that the deeper altered states of consciousness will be regarded as dan-
gerous (Douglas 1973:104). Certainly, it is hard to imagine that penetration of the
deep Upper Paleolithic caves could have been regarded as anything but perilous. In-
creasing the physical danger and discomfort further restricted access to the ultimate
religious experience and the power that this experience bestowed. At the same time,
the markedly restricted movement that some of the smaller decorated areas (e.g., the

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 337

tunnel at Le Gabillou, the Chamber of Felines in Lascaux and the Chimney in Berni-
fal) permit suggests increasing restrictions on the ways in which ideologically de-
fined altered states of consciousness might be experienced and a concomitant
codification of beliefs concerning such states. Access to certain altered states, and
hence to certain kinds of knowledge, was being restricted. Altered states were being
appropriated, defined, and differentially allocated in attempts to reproduce and to
transform relations of power (Lewis-Williams 1995c).

The essential historicity of this process explains why the making of shamanic
rock art was not a universal practice, despite the universality of shamanic mental im-
agery. In southern Africa, for instance, some shamanic San communities made rock
paintings and engravings, while others, such as those living in the Kalahari Desert,
did not. There are, of course, few rock surfaces in the flat, sandy Kalahari on which
rock art could be made, but the shamanic functioning of the San communities that live
there does not, even today, seem to be impaired by an absence of image-making. The
experiences of altered states of consciousness are, in fact, a necessary but not a suffi-
cient condition for the making of shamanic art. Those experiences and their associ-
ated mental imagery constitute a resource that, under certain historically specific
social circumstances can be exploited. As Bourguignon (1974:234) pointed out, vi-
sions acquired in altered states of consciousness are “raw materials for potential cul-
tural utilization.” The need to reify on rock these experiences and visions is not
ineluctable: there is no neuropsychological imperative. In some communities, sha-
mans weave their experiences into tales, songs and mythology; sometimes they also
depict them on perishable materials, such as skin drum coverings and clothing (e.¢.,
Brodzky et al. 1977). Each instance of the making of rock art is therefore a specific
historical question. Neuropsychological research into altered states of consciousness
clarifies the substance and nature of a resource, not the social conditions under which
that resource is accessed. Why people living at the beginning of the west European
Upper Paleolithic began to make rock art is a question that still needs to be addressed.
The reasons why they continued to make rock art throughout the Upper Paleolithic
may well have changed as historical circumstances changed.

The historical specificity of each Upper Paleolithic image and the kind of
shamanic context in which it was made point to an important mechanism for change
in the cosmological, social, religious and iconographic frameworks of the period.
Each painted or engraved image should be seen as an individual or group interven-
tion. Human agency (see, for example, Bourdieu 1977; Giddens 1984; Johnson
1989), not impersonal, hardwired cognitive structures, accounts for change. As I have
pointed out, altered states of consciousness are a resource on which actors can draw as
they negotiate their social and political statuses. Unfortunately, it is a resource that is
usually omitted from archaeological accounts of change, perhaps because of current
values in Western academia and a (one hopes implacable) resolve to exclude ‘New
Age science’ with which “mystical” things like altered states are associated. The seri-
ousness of this omission is highlighted by those studies that do consider altered states.
For instance, in an account of change in the iconography of southern A frican rock art,
Dowson (1994) argues that different types of images were manipulated by politically
emergent shamans to negotiate new statuses. Citing Barth (1987), Riches (1994)
makes a similar point. Religious specialists among the Mountain Ok of New Guinea
effect incremental shifts in the connotations of particular symbols. In Riches’ apt
phrase, the shaman is a “cosmology maker.” I argue that, as individual Upper Paleo-
lithic shamans pursued their personal and group interests, the cosmology that they
created and modified (partly by their art) both constrained and enabled social change.

338 LEWIS-WILLIAMS

A Way Ahead

As I pointed out at the beginning of this chapter, there is today a deep and justifi-
able suspicion of all-encompassing explanations for Upper Paleolithic art — those that
cover all the different forms of art that people made during the entire 20, 000 year span
of the period. In demonstrating the explanatory power of the shamanic hypothesis, I
have therefore considered only the parietal cave art of western Europe. I have, moreo-
ver, emphasized spatial and temporal diversity.

There are, of course, unresolved problems, and the explanation and its implica-
tions need to be fully worked out. But that is true of the most abundantly confirmed
and widely accepted hypotheses; evolutionary theory 1s a case in point. One does not
have to explain everything in order to explain something.

I simply argue that no other explanation currently before researchers has equal
evidential support. Nor does any other explanation have comparable explanatory
power and potential; the shamanic explanation resolves some of the most intractable
problems of Upper Paleolithic research in a parsimonious way. In doing so, it creates
an entirely new way of categorizing Upper Paleolithic images and addressing their
changing deployment within the caves. Moreover, it places all Upper Paleolithic ac-
tivity, whether economic, social or ritual, within an evolving cosmology.

By intertwining diverse empirical data, multiple ethnographic analogies with
strong relations of relevance, and the results of neuropsychological research on al-
tered states of consciousness, the approach that I advocate breaks out of the current
impasse created by the collapse of earlier explanations and by a tendency to think that
new data, by themselves, will ultimately provide explanations. The despair of ever
understanding at least something about Upper Paleolithic art that characterizes much
writing on the period is unjustified.

Acknowledgments

Iam grateful to colleagues who offered useful comments on drafts of this paper:
Geoff Blundell, Meg Conkey, Anne Holliday and Tom Huffman. Denise Voorvelt
and Anne Holliday kindly typed successive drafts. The illustrations were prepared by
Anne Holliday who also wrestled with references. Iam especially grateful to numer-
ous people who made it possible for me to visit Upper Paleolithic caves in France in
1972, 1989, 1990 and 1995 and with whom I had useful discussions: They include
Norbert Aujoulat, Paul Bahn, Robert and Eric Bégouan, Brigette Delluc, Jean
Gaussen, Yanik Le Guillou, André Leroi-Gourhan, Michel Lorblanchet, Jaques
Omnis, Aleth Plenier, Jean-Philippe Rigaud, Yoan Runeau, Dominique Sacchi,
Georges Simonnet, Denis Vialou, Luc Wahl and, especially Jean Clottes. I am in-
debted to Jean Auel who generously facilitated a highly productive colloquium in her
home, and to the Wattis Foundation for sponsoring the California Academy of Sci-
ences Symposium at which this paper was presented. Both Mrs. Auel and Mrs. Wattis
have shown a personal interest in the work. The Rock Art Research Unit is funded by
the Centre for Science Development (Human Sciences Research Council) and the
University of the Witwatersrand; neither institution is responsible for the views ex-
pressed here.

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE

ies)
ie)
No}

Literature Cited

L'Art des cavernes (Atlas des Grottes Ornces Paléolithiques francaises). 1984. Ministire de la
Culture, Paris, France.

Asaad, G. 1980. Hallucinations in Clinical Psychiatry: A Guide for Mental Health Profession-
als. Brunner/Mazel, New York.

Asaad, G., & B. Shapiro. 1986. Hallucinations: Theoretical and clinical overview. Amer. J.
Psych. 143:1088-97.

Bahn, P.G. 1995. Cave art without caves. Antiquity 69:23 1-37.

Bahn, P.G. & J. Vertut. 1988. /mages of the Ice Age. Windward, London, U.K.

Barth, F. 1987. Cosmologies in the Making. Cambridge University Press, Cambridge, U.K.

Bednarik, R.G. 1984. On the nature of psychograms. The Artefact 8:27-32.

. 1986. Parietal finger markings in Europe and Australia. Rock Art Res, 3:30-61.

Bégouan, H. & H. Breuil. 1958. Les Cavernes du Volp, Trois-Fréres, Tuc d'Audoubert, O
Montesquieu-Avantds (Aridge). Arts et Métiers Graphiques, Paris, France.

Bender, B. 1989. The roots of inequality. Pages 83-93 in D. Miller, M. Rowland & C. Tilley,
eds., Domination and resistance. Unwin and Hyman, London, U.K.

Berrin, K. 1978. Art of the Huichol Indians. Abrams, New Y ork.

Bourdieu, P. 1977. Outline of'a Theory of Practice. Cambridge University Press, Cambridge,
U.K.

Bourguignon, E. 1974. Cross-cultural perspectives on the religious uses of altered states of
consciousness. in I. Zaretsky & M. Leone, eds., Religious Movements in Contemporary
America. Princeton University Press, Princeton.

Bradley, R. 1985. Consumption, Change and the Archaeological Record. Department of Ar-
chaeology, University of Edinburgh, Edinburgh, U.K.

Breuil, A. 1952. Four Hundred Centuries of Cave Art. Centres d'étude et de Documentation
Préhistoriques, Montignac, France.

Brodzky, A.T., R. Danesewich & N. Johnson, eds. 1977. Stones, Bones and Skin: Ritual and
Shamanic Art. The Society for Art Publications, Toronto, Canada.

Chauvet, J-M., E.B. Deschamps, & C. Hillaire. 1995. La Grotte Chauvet, O Vallon Pont-
D'Arc, éditions du Seuil, Paris, France.

Clottes, J. 1995. Les Cavernes de Niaux: Art Préhistorique en Aridge. Editions du Seuil, Paris,
France.

Clottes J., A. Beltran, J-Courtin & H. Cosquer. 1992. The Cosquer Cave on Cape Morgiou,
Marseilles. Antiquity 66:583-98.

Clottes, J. & J. Courtin. 1994. La Crotte Cosquer: Peintures et Gravures de la Caverne
Engloutie, éditions de Seuil, Paris, France.

Clottes, J., M. Garner & G. Meury. 1994. Magdalenian bison in the caves of the Ariage. Rock
Art Res. 11:58-70.

Conkey, M.W. 1987. New approaches in the search for meaning? A review of research in *Pa-
laeolithic art.” J. Field Arch. 14:413-30.

Davenport, D. & M.A. Jochim. 1988. The scene in the Shaft at Lascaux. Antiquity 62:558-62.

Douglas, M. 1973. Natural Symbols: Explorations in Cosmology. Penguin, Harmondsworth,
U.K.

Dowson, T.A. 1988. Revelations of religious reality: The individual in San rock art. World
Arch, 20:1 16-28.

Dowson, T.A. 1994. Reading art, writing history: rock art and social change in southern Africa.
World Arch. 25:332-45.

340 LEWIS-WILLIAMS

Dronfield, J. 1994. Subjective Visual Phenomena in Irish Passage-tomb Art: Vision, Cosmol-
ogy and Shamanism. Ph.D. Dissertation. University of Cambridge.

Dronfield, J. 1995. Subjective vision and the source of Irish megalithic art. Antiquity 69:539-
49.

Eichmeier, J. & O. Hifer. 1974. Endogene Bildmuster. Urban and Schwarzenberg, Munich,
Germany.

Eliade, M. 1972. Shamanism: Archaic Techniques of Ecstasy. Routledge and Kegan Paul, New
York.

Freeman, D., J.G. Echegerey, F. Bernaldo de Quiros & J. Ogden. 1987. Altimira Revisited and
other Essays on Early Art. Institute for Prehistoric Investigations, Chicago.

Freidel, D., L. Schel & J. Parker. 1993. Maya Cosmos. Three Thousand Years on the Shaman's
Path. William Morrow, New York.

Furst, P.T. 1976. Hallucinogens and Culture. Chandler and Sharp, Novato, CA.

Gaussen, J. 1964. La Grotte Ornée de Gabillou. Memoir 4, Inst. de Préhistoire de l'Universite
de Bordeaux, Bordeaux, France.

Giddens, A. 1984. The Constitution of Society: Outline of the Theory of Structuration. Univer-
sity of California Press, Berkeley.

Goodman, F.D. 1988. Ecstasy, Ritual, and Alternate Reality. Indiana University Press, Bloom-
ington.

Graziosi, P. 1960. Palaeolithic Art. Faber and Faber, London, U.K.

Halifax, J. 1980. Shamanic Voices. Penguin Books, Harmondsworth, U.K.

Harner, M.J., ed. 1973. Hallucinogens and Shamanism. Oxford University Press, New York .

Hayden, B. 1993. The cultural capacities of Neandertals: A review and re-evaluation. J. Hum
Evol. 24:113-46.

Hedges, K. 1982. Phosphenes in the context of Native American rock art. Pages 7-8:1-10 in
F.G. Bock, ed., American Indian Rock Art. American Rock Art Research Association, El
Toro,

—_. 1983. The shamanice origins of rock art. Pages 46-59 in J.A. Van Tilburg, ed., Ancient
Images on Stone: Rock Art of the Californias. Rock Art Archive, UCLA, Los Angeles.

Horowitz, M.J. 1964. The imagery of visual hallucinations. J. Nerv. Ment. Dis. 138:513-23.

1975. Hallucinations: an information-processing approach. Pages 163-95 in R.K.
Siegel & L.J. West, eds., Hallucinations, Behaviour, Experience, and Theory. John Wiley
and Sons, New York.

Johnson, M.H. 1989. Conceptions of agency in archaeological interpretation. J. Anthropol
Arch. 8:189-211.

Knoll, M., J. Kugler, O. Hifer, & S.D. Lawder. 1963. Effects of chemical stimulation of
electrically-induced phosphenes on their bandwidth, shape, number and intensity. Confinia
Neurologica 23:201-26.

Klaver, H. 1926. Mescal visions and eidetic vision. Amer. J. Psych. 37:502-15.

—__. 1942. Mechanisms of hallucinations. Pages 175-207 in Q. McNemar, & M.A. Mermill,
eds., Studies in Personality. McGraw Hill, New York.
1966. Mescal and Mechanisms of Hallucinations. Chicago University Press, Chicago.

La Barre, W. 1970. The Ghost Dance: Origins of Religion. Doubleday, Garden City.

—_. W. 1975. Anthropological perspectives on hallucination and hallucinogens. Pages 9-52
in RK. Siegel & L.J. West, eds., Hallucinations: Behaviour, Experience, and Theory. John
Wiley and Sons, New York,

Laming-Emperaire, A. 1962. La Signification de l'Art Rupestre Paleolithique. Picard, Paris,
France.

Leroi-Gourhan, A. 1968, The Artof Prehistoric Man in Western Europe. Thames and Hudson,
London, U.K.

SHAMANISM IN UPPER PALEOLITHIC WESTERN EUROPE 341

Leroi-Gourhan, A. and J. Allain. 1979. Lascaux /nconnu. Centre National de la Recherche Sci-
entifique, Paris, France.

Levy, G.R. 1963. Religious Conceptions of the Stone Age. Harper and Row, New York.

Lewis-Williams, J.D. 1980. Ethnography and iconography: Aspects of southern San thought
and art. Man 15:467-82.

. 1981. Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings. Aca-
demic Press, London, U.K.

. 1984. The empiricist impasse in southern African rock art studies. S. Afr, Arch. Bull
39:58-66.

. 1989. Theory and data: A brief critique of A. Marshack's research methods and position
on Upper Palaeolithic shamanism. Rock Art Res. 6:38-53.

_. 1990. Documentation, analysis and interpretation: Dilemmas in rock art research. Re-
view of “The rock paintings of the Upper Brandberg. Part I: Amis Gorge”. S. Afr. Arch
Bull. 45:126-136.

. 1991, Wrestling with analogy: A problem in Upper Palaeolithic art research. Proc. Pre-
hist. Soc. 57:149-62.

. 1992. Ethnographic evidence relating to ‘trancing’ and ‘shamans’ among northern and
southern Bushmen. S. 4fr. Arch. Bull. 47:56-60.

. 1995a. J.D. Seeing and construing: The making and ‘meaning’ of a southern African
rock art motif. Cambridge Arch. J. 5:3-23.

. 1995b. Some aspects of rock art research in the politics of present-day South Africa. im
K. Helskog & B. Olsen, eds., Perceiving Rock Art: Social and Political Perspectives. No-
vus Forlag, Oslo, Norway.

. 1995¢. Modelling the production and consumption of rock art. S| Afr Arch. Bull.
50:143-54.

Lewis-Williams, J.D. & T.A. Dowson. 1988. Signs of all times: Entoptic phenomena in Upper
Palaeolithic art. Curr. Anthropol. 29:201-45.

. 1989. Images of Power: Understanding Bushman Rock Art. Southern Book Publishers,
Johannesburg, South A frica.

. 1992. Art rupestre San et Paleolithique superieur: Le lien analogique. L ‘Anthropologie

96:769-90.

. 1994. eds. Contested Images: Diversity in Southern African Rock Art Research. Univer-
sity of the Witwatersrand, Johannesburg, South Africa.

Lewis-Williams, J.D. & J.H.N. Loubser. 1986. Deceptive appearances: A critique of southern
African rock art studies. Pages 253-89 in F. Wendor & A.E. Close, eds., Advances in World
Archaeology. Academic Press, New York .

Lommel, A. 1967. Shamanism: The Beginnings of Art. McGraw-Hill, New York.

Marshack, A. 1972. The Roots of Civilization. Weidenteld and Nicolson, London, U.K.

. 1976. Some implications of the Palaeolithic symbolic evidence for the origin of lan-
guage. Curr. Anthropol. \7:274-82

Omnis, J. 1982. La Grotte ornée de Labastide (Hautes-Pyrenees). Lourdes: J. Omnes, France.

Pfeiffer, J.E. 1982. The Creative Explosion: An Enquiry into the Origins of Art and Religion.
Harper and Row, New York.

Reichel-Dolmatoff, G. 1972. The cultural context of an aboriginal hallucinogen: Banisteriop-
sis Caapi. Pages 84-113 in P.T. Furst, ed., Flesh of the Gods: The Ritual Use of Hallucino-
gens. Allen and Unwin, London, U.K.

. 1978. Beyond the Milky Way: Hallucinatory imagery of the Tukano Indians. UCLA
Latin America Centre, Los Angeles.

Reinach, S. 1903. L'art et la magie: O propos des peintures et des gravures de l'Age du Renne.

L'Anthropologie 14:257-266.

342 LEWIS-WILLIAMS

Riches, D. 1994. Shamanism: The key to religion. Man (N.S.) 29:381-405.

Siegel, R.K. 1977. Hallucinations. Sei. Amer. 237:132-40,

Siegel, R.K. & M.E, Jarvik, 1975. Drug-induced hallucinations in animals and man. Pages 8 |-
lol in R.K. Siegel & L.J. West, eds., Hallucinations: Behaviour, Experience, and Theory.
Wiley, New York.

Siegel, R.K. & LJ. West eds. 1975. Hallucinations: Behaviour, Experience, and Theory. John
Wiley and Sons, New York.

Smith, N.W. 1992. 4n Analysis of Ice Age Art: Its Psychology and Belief System. Peter Lang,
New York.

Thomas, J. 1991. Rethinking the Neolithic. Cambridge University Press, Cambridge, U.K.

Tilley, C. 1991. Material Culture and Text: The Art of Ambiguity. Routledge, London, U.K.

Tyler, C.W. 1978. Some new entoptic phenomena. Vision Res. 18:1633-39.

Ucko, P.J. 1992. Subjectivity and recording of Palaeolithic cave art. Pages 141-180 in T. Shay,
& J. Clottes, eds., The Limitations of Archaeological Knowledge. Etudes et Recherches Ar-
cheologiques de l'Universite de Lidge, Liige, Belgium.

Ucko, P.J., & A. Rosenfeld. 1967. Palaeolithic Cave Art. Weidenteld and Nicholson, London,
U.K.

Vandiver, P. 1983. Paleolithic Pigments and Processing. MSc Thesis. M.L.T.

Vialou, D. 1982. Niaux, une construction symbolique magdalémienne exemplaire. Ars Prae-
historica 1:19-45.

_ 1986. L'Art des Grottes en Aridge Magdalenienne. editions du CNRS, Paris, France.

Vinnicombe, P. 1976. People of the Eland: Rock Paintings of the Drakensberg Bushmen as a
Reflection of their Life and Thought. Natal University Press, Pietermaritzburg, South
Africa.

Vitebsky, P. 1995. The Shaman. Macmillan, London, U.K.

Walker, J. 1981. The amateur scientist: About phosphenes. Sci. Amer. 244:142-52.

Whitley, D-S. 1987. Socio-religious context and rock art in east-central California, J. An-
thropol. Arch. 6:159-88.

. 1988. Comment on J.D. Lewis-Williams and T.A, Dowson “The signs of all times: en-
toptic phenomena in Upper Palaeolithic art”. Curr. Anthropol, 29:238.
—_. 1992. Shamanism and rock art in far western North America. Cambridge Arch. J. 2:89-
113.

—__. 1994a. By the hunter, for the gatherer: Art, subsistence and social relations in the pre-
historic Great Basin. World Arch. 25:356-373.

_. 1994b. Shamanism, natural modeling and the rock art of far western North American
hunter-gatherers. Pages 1-43 in S.A. Turpin, ed., Shamanism and Rock Art in North Amer-
ica. Rock Art Foundation, Inc, San Antonio, TX.

Winkelman, M. 1986. Trance states: A theoretical model and cross-cultural analysis. Ethos
14:174-203.

Wylie, A. 1982. An analogy by any other name is just as analogical: A commentary on the
Gould-Watson dialogue. J. Anthropol. Arch. 1:382-401

Wylie, A. 1985. The reaction against analogy. Adv. Arch. Method Theory 8:63-111.

—_. 1988. ‘Simple analogy’ and the role of relevance assumptions: Implications of archaeo-
logical practice. /ntl. Stud. Phil. Sci. 2:134-150.

—__. 1989, Archaeological cables and tacking: The implications of practice for Bernstein's
‘Options beyond objectivism and relativism.” Phil, Soc. Sci. 19:1-18.

Beyond Art and Between the
Caves: Thinking About Context
in the Interpretive Process

Margaret W. Conkey

Department of Anthropology
University of California

232 Kroeber Hall

Berkeley, CA 94720

It is obvious that the study of early art has been changing. I am acutely aware of
this primarily because I have been trying to finish a book on the topic of European Pa-
leolithic art for the past ten years. To take this long to complete the book has been both
a good and problematic time lag. While the project began as a summary of the extant
interpretations for the imagery (c/. Conkey 1987), it has shifted into a study of the in-
terpretive histories and processes. The shifts in our own intellectual, social, and po-
litical contexts since the early 1980's — such as changes in art history, in the
anthropology of art, the challenges to empiricism and the “interpretive turn” in an-
thropology and archaeology, even the mobilization of the field of “rock art research,”
to say nothing of our own more self-reflexive research contexts — have necessarily
impacted upon my project. In this final chapter to this volume of Beyond Art, 1 will
consider some of the issues that have emerged for me in completing this book primar-
ily because these are some of the very same issues that are implicated by many of the
papers in this volume. In particular, I want to take up the issue of “context” because of
its role in interpretation.

I know that I have been seduced by some of the ongoing intellectual history of this
field; [have become as interested in the constituting contexts (after Gero 1996) for the
research as in the contexts of the prehistoric imagery. It has become very clear to me
that what an analyst thinks about this term — “‘context” — before or as they work makes
a distinct and specific impact on their interpretations; that the so-called “constituting
contexts” of how one does research are crucial; and that certain accounts for the im-
agery readily follow from the conceptual referents of “context.” “Context” and “in-
terpretation” then are inextricably entangled. To understand and to assess the
interpretations is directly related to the conceptual referents of context — that is, what
do people mean by “context” ?

Beyond Art: Pleistocene Image and Symbol Memoirs of the California Academy of Sciences
Editors, M. Conkey, O. Soffer, D. Stratmann, & N.G. Jablonski Number 23, Copynght ©1997

344 CONKEY

There have been three proximate motivations for me to take up a consideration of
the concept of “context”:

1) because I have had some of my own notions of this taken-for-granted concept
jolted by some perceptive comments by David Lewis-Williams . To summarize only
briefly at this point, he reminds us that contexts are not inherent but constructed;
“contexts are made out of contexts” (Lewis-Williams 1990:133).

2) because Homo sapiens sapiens (anatomically-modern humans) now appear to
have established themselves outside of Europe a considerable time before the appear-
ance there of preserved Upper Paleolithic imagery at some 32,000 years ago, and be-
cause — as many papers in this volume discuss — there is increasing support for
image-making or at least ‘marking’ at comparably early dates elsewhere (in southern
Africa and Australia, at least). Thus, the Upper Paleolithic imagery of Eurasia 1s liber-
ated from what have been persistent, almost “vitalistic” concepts of “context” in
which the art “happens” with modern humans, and the European materials are the or7-
gins of art. We are now at the point in the history of research where we decouple the
imagery from grand vitalistic and evolutionary schemes for the imagery, which can
now be fruitfully approached as local, historical phenomena. While these phenomena
may be linked to more macroprocesses (for one account, see Davidson, this volume),
nevertheless, the specific forms, shapes, raw materials, and transformations through
time and space are not likely to be explained by such processes. This has been a cru-
cial liberation in that it has undermined the prevalent tendency to spatialize time (cf.
Fabian 1983), which has presented the Upper Paleolithic as synonymous with the
Eurasian archaeological record; and

3) because I find myself pursuing a research project that is intended to grapple
with what I have constructed to be one “context” for the making and use of materials
and images, specifically during the Magdalenian period of the French Midi-Pyrénées.
Thus, I must, if | can, turn my critical eye on my own research concepts as well as on
those of others.

This chapter then, will first consider some connotations of “context” in the study
of Upper Paleolithic imagery, then consider this idea of “constituting contexts,” and
lastly turn to the particular research project as an example of one kind of contextual
analysis of/for the imagery as social practice.

Connotations of Context

As Lewis-Williams suggests in his chapter, we are in the so-called post-Leroi-
Gourhan era of interpretation, at least in regard to interpreting the imagery of the late
Pleistocene of Eurasia. To Lewis-Williams, this involves the idea that contemporary
researchers, especially those in the European intellectual tradition, have retreated
from the somewhat grand and inclusive interpretations, such as Leroi-Gourhan’s
structuralist/mythogram account (e.g., Leroi-Gourhan 1965, 1982; Conkey 1989) for
the making and placing of specific images in the caves. There are many important
legacies both of the detailed empirical work that Leroi-Gourhan did, and of the theo-
retical implications embedded in his publications, which drew as much from the
thinking of Laming-Emperaire and Raphael as from Levi-Straussian structuralism
(Chesney 1993; Conkey 1992a, in press; Delporte 1990). While Leroi-Gourhan’s im-
portant work undermined the idea of painting/image making as just a “mode of cogni-
tion,” it had the problem of then defining the meaning of the sign (taken in the
semiotic sense, whatever image or image-sets) “entirely by formal means ... within an

CONTEXT IN THE INTERPRETIVE PROCESS 345

enclosed system” (Bryson 1983: xi1). This can be seen most baldly in the formulae he
proposes for the various panels or figure groups in Lascaux (in Arl. Leroi-Gourhan &
Allain 1979:348). What was lacking, of course, is any grappling with “how signs
{again , in the semiotic sense of signs] interact with the world outside of their internal
system” (Bryson 1983:x11). Image making 1s “in constant touch with signifying forces
outside of (it), forces that cannot be accounted for by ‘structuralist’ explanations” (af-
ter Bryson 1983: xii-xiti, who 1s critiquing not Leroi-Gourhan, but certain approaches
and assumptions in the field of art history).

While “contextualist” is a term that would not be used to describe a structuralist
account such as that of Leroi-Gourhan, which held sway for many decades in the
study of the European cave art, the structuralist account is, however, at one scale,
“contextual” in that it depends upon inferring the relationships among component
parts of the system. In the case of Leroi-Gourhan's study of the topographic place-
ments of different images (e.g., bison, horse, hand imprint), it did matter that a bison
was ina specific locale and/or in some sort of “association” with another particular
image. In contrast to this kind of context, many anthropologically trained archaeolo-
gists of early art are likely to think first of “context” as the social group(s) within
which the imagery was made. What this illustrates is that the consideration of context
can be at many levels, or multiscalar. While each researcher may have their preferred
scale or analytical level, it will no doubt be from the dialectics — from playing off one
level against another — that we may find our most fertile interpretive terrain. Of
course, we must be explicit about the level of context at which we are working, and
this should structure the level of interpretive reference: an analysis of the associations
of geometric designs on specific decorated objects tells us about one of the most lim-
ited levels of context, and — without the appropriate linking arguments — cannot be
used to imply higher or more inclusive orders of context, such as social groups or the
work of an individual maker.

Perhaps the most obvious body of work that has gone on in the realm of one kind of
context for the study of ancient “art” and image making has been under the rubric of
“the archaeology of decorated caves”; in fact, since so much of this has been spear-
headed by French researchers, it could be most appropriately called, L ‘Archéologie
des Grottes Ornées. Within this emergent and energetic field there have been basi-
cally two kinds of studies: 1) the archaeology of the images, usually within a given
cave: footprints, hearths, traces of the most minute prehistoric activities: and 2) the
studies of the production and reproduction of the images themselves — the pigments
used, the engraving tools, efc. — that imply or evoke aspects of productive contexts,
such as grinding pigments (see Clottes 1993; this volume; and Clottes et a/. 1990).
Just within the French Pyrenees, the work of Jean Clottes and colleagues has brought
to light fascinating aspects of this kind of site-level analysis. At Fontanet, there are the
hand prints, the “drawing” on the cave floor next to the footprint of a covered foot, and
there are reindeer antler “markings”; in the Reseau Clastres, there are the stalactites
broken off, presumably by ancient visitors to the cave, as well as some footprints. In
some caves there are pigments on the floor, or implements, such as the bone “spatula”
at Montespan used to sculpt the bear of clay.

In other caves, a combination of archaeological evidence and experimental
work is suggestive about different kinds of “visits,” such as at Pech-Merle, where
Lorblanchet (1980) has replicated the so-called Black Frieze to suggest that it could
have been done quickly (in less than an hour) and by one individual, which he uses in
combination with fragmentary pollen evidence to suggest that this cave area was only
visited infrequently and was not the “sanctuary of the masses” that Breuil once sug-

346 CONKEY

gested. What is compelling about this kind of work is that the researchers are often us-
ing several lines of converging evidence when any one line of evidence alone is not
strong enough to support an interpretation.

However, as Clottes (1993, personal communication) has pointed out, among
most European (and especially French, see also GRAPP 1993) researchers working
with “Paleolithic art,” the notion of “context” is associated primarily with studies like
these: of and within specific caves, more so than than in terms of sociopolitical forma-
tions, or the social relations of production (for some European exceptions, see
Gonzalez-Morales 1991, this volume; Criado & Penedo 1989; for a consistent at-
tempt to frame Paleolithic imagery in quite wide and inclusive social contexts, see
Gamble 1982, 1991). Many of these researchers who favor the delimited, within-cave
approaches tend to hold that the image making is somehow (but unspecifiably ) “relig-
ious” (Delporte, personal communication), and the kind of more explicitly theorized
“context” that is so seductive for Anglo-American anthropologists and/or Marxist ar-
chaeologists is not central to (nor even very knowable, epistemologically, in ) their
research agendas. These latter do include continued — and quite valiant and sophisti-
cated — work to render the image making comprehensible in structuralist and semiotic
terms (Sauvet 1988: Sauvet & Wlodarezyk 1992); and the increasingly sophisticated
analyses of the forensic sort, such as inferring pigment mixes and techniques of
“painting”; considerations of the chaine opératoire in the manufacture of forms,
shapes, and decorated objects; and consistently inspired work directed at chronologi-
cal refinements and spatial relationships.

If one were to begin to compile a kind of “typology” of the different kinds of
“context” that different researchers (using examples here from the European Paleo-
lithic) have pursued, it might start out like this:

1. Immediate and proximal contexts of imagery

la. internal analysis: the close inspection of images on the walls or bone/antler in
association with each other or with attributes of the medium and/or the very gestures
that brought a ‘mark’ into existence in the first place (e.g., d'Errico 1989, 1992; see
Marshack, this volume), or close inspection of figures associated within a cave
(Leroi-Gourhan 1958, 1966)

1b. site-level: the use(s) ofa cave or shelter site, which includes studies of the na-
ture of the “visits” related to imagery pigments, scaffolding (Lascaux), footprints,
“snacks,” a bone stuck here or there into the cave floor, or actual occupation (e.g.,
Enléne, Gazel, La Vache, Fontanet); also, tracking these “uses” through time (see
Clottes, this volume).

2. The occurrence of imagery as spatial ‘marks’; both decorated caves and depos-
its of “portable art” (engraved/carved bone/antler) in certain sites (and not others)
may be considered as ‘marked’ and ‘marking’ features of the cultural landscape, right
into the present (e.g., Bahn 1978; Conkey 1980, 1992b; Jochim 1983)

3. The relationships among and between image-sets: such as the stylistic contexts
(when images can be ‘related’ stylistically) or the contexts of form and materials (see
White, this volume, p. 97). In certain “sets” of imagery — such as the contours de-
coupées (Bellier 1984), the female figurines (Gamble 1982), or the animal-ended
spear-throwers (Cattelain 1980) — that appear distributed differentially across the
landscape but suggest “connections” and formal similarities, these could be ap-
proached as “stylistic contexts” (see Davis 1990).

To advocate that we vigorously pursue the study of the context(s) of Paleolithic
(or any other prehistoric/ethnographic) imagery is simultaneously banal — of course
we are interested in context! — and deeply philosophical and complicated. “Context”

CONTEXT IN THE INTERPRETIVE PROCESS 347

is one of those pervasive taken-for-granted notions, but it is only rarely agreed upon,
especially when it comes to what the appropriate or relevant contexts for interpreta-
tion are. In his recent review of Pager's monumental study of the Brandberg rock art,
Lewis-Williams (1990) includes a most succinct and crucial insistence on the fact that
contexts are not inherent, but constructed; a “context” is a theoretical assumption, and
“contexts are made out of contexts”:

All contexts are assemblages of selected features; they are value-laden constructs made
by archaeologists acting in specific social circumstances. It is (of course) the archaeolo-
gists who decide which features of the past shall be deemed “the context of art” (Lewis-
Williams 1990:133),

As noted above, there are many different kinds of context that are often at differ-
ent scales of analysis and interpretation: In looking at the relations between and
among incisions on an engraved bone or plaquette, Marshack’s “internal context” is
different from the contextual understandings of bone pieces stuck in the floors and
crevices of some decorated caves — a kind of “proximate context” — in the French
Pyrénées. And it is yet another kind of context to infer a group of at least four indi-
viduals to have constructed the purported scaffolding for painting at Lascaux, and an
even other kind of context to infer that by late Upper Paleolithic times, there were se-
lect individuals who were somehow “authorized” — due to any number of social, cul-
tural and/or psychological factors — by a wider social group to not only engage in
behaviors involving altered states of consciousness but also to engage in the transfer-
rence of experiences into material or visual forms (an argument that Lewis-Williams
is, ineffect, making; this volume ), or —as Soffer suggests (this volume) — that the so-
cial situation in later Upper Paleolithic central European sites suggests there were cer-
tain specialists who could decode an increasingly schematized visual repertoire.
What one then takes to be the relevant evidence for each of these (and many other
kinds of ) “contexts” will vary, and it is not unusual for each analyst to be a strong ad-
vocate for “their” kind of context.

Of course, what we need are many levels and many kinds of contexts. The more
formal and proximate contexts listed above about marks and individual caves are one
part — but a crucial part — of the engagement with the wider social formations within
which imagery was made, caves were used, and material meanings were generated
and transformed. What we need as much is the critical assessment of what we assume
about context, and precisely how we construct the contexts we prefer, the contexts we
pursue, and the contexts that we endorse or dismiss.

Constituting Contexts

Practices of Research

One often overlooked “context” is the notion of “constituting contexts,” which is
often discussed in the literature of the sociology of science and has been recently
elaborated by Joan Gero (1996) in her analysis of archaeological excavation prac-
tices. Two aspects to this notion of “constituting contexts” are relevant here: 1) prac-
tices of research, and 2) the entanglements of our own modes of thought with those
(modes of thought) we seek to interpret or represent. | want to draw heavily from
Gero’s account here for I think there are important implications for the entanglement
of “context” and “interpretation.”

348 CONKEY

Our knowledge of the past is entrenched in the everyday unquestioned practices, the ac-
cepted use of (certain) specific instruments, and the carefully inculcated technical abili-
ties [such as tracing rock art imagery] that we teach and expect of practitioners, and that
we build into, and count on, as the unquestioned features and context of doing archae-
ology (Gero 1996).

These, she reminds us, have become the “right” thing(s) to do; they proceed
without question and thus * they constitute their own kind of rationality.” Gero ar-
gues that in asking — as we have in recent years (see Tomaskova, this volume) —“how
is archaeological knowledge related to the cultural politics in which such knowledge
is accumulated?” that we have “taken a false turn.” “We have looked for a context
apart from and external to the knowledge itselfand have tried to see how that context”
~ such as the interest in cybernetics and systems theory approaches in the 1960's, the
1970's focus on population issues and problems, and the rise of business metaphors —
“acted upon practitioners of (archaeological) science and upon data” (Gero 1996).
What she suggests is that it is not so much the seemingly external cultural politics or
social movements that are primarily structuring our production of knowledge, but it is
archaeological practices themselves (that) “constitute a context of rational account-
ability”: They create, reinforce, and replicate our own kind of rationality. That 1s, ar-
chaeological practices — how it is we “do” our archaeology — are themselves a
constituting context. We are, she points out — in our interactive ways — “each others’
context.” In our symposia, for example, we are anticipating, speaking to, reacting to,
and diverting or reinforcing the views, aims, and opinions of each other. The volume
of papers here is as mucha product of this interactive context as it is some “objective”
presentation of the state-of-the-art. It is not, then, just ethnographers who must con-
front and deal with the intersubjective nature of their enterprise. There are two (at
least) directions to take this observation.

First, I find it somewhat suggestive that theroutines of discovery and exploration —
especially the “original” and pre-1950’s ones — of much European Paleolithic “cave
art” are reproduced — often quite directly, in the interpretive scenarios for the “art.”
These are most obvious in artists' reconstructions that are usually somehow (even if
through passive acceptance) sanctioned by archaeologists, 1f not usually drawn by
them. For example, many of the early caves were discovered (or immediately
authenticated”) by a discovery/exploration party of a small select group of men often
led by a religious specialist (e.g., Breuil). It is not surprising to see many of these
means of encounter, exploration and discovery replicated in the visual interpretations
for the imagery (Figure 1, compare the 1940 photograph of M. Ravidat with his torch
exploring the cave of Lascaux with the 1986 drawing of the “shaman’/religious
leader in Lascaux). In Figure 2, the expressions on the faces of those (men) who are
encountering the clay bison at Le Tuc d’ Audoubert are unambiguously expressions
of awe, which was itself, understandably in 1912, one of the characteristic expres-
sions of the Begouen brothers, their father and the Abbe Breuil upon their discovery
(Begouen & Breuil 1958). Ofcourse, Breuil and other priests were supported by soci-
ety; they had both the time and funds where few others did. Yes, as Lewis-Williams
has suggested (personal communication), priests were motivated to be among the
first discoverers, to perhaps “control” the discoveries, where were, after al to them
about the “origins of religion.”

While there is indeed some tantalizing evidence for a kind of scaffolding at Las-
caux, the routing of the Abbé Glory in tracing the (relatively unknown and unappreci-
ated but abundant) engravings in Lascaux — on an elaborate scaffolding (Figure 3a) —

CONTEXT IN THE INTERPRETIVE PROCESS 349

Figure la. M. Ravidat shows the paintings at Lascaux with the aid of his carbon lamp in 1940 (Leroi-
Gourhan and Allain 1979:54, Figure 35)

is replicated in the artists’ reconstructions for “comment on fait Lascaux” (Figure
3b): how they made the Lascaux imagery.

The absence of women as image makers in both artists’ reconstructions and in the
discussions about the making and using of imagery has recently been noted (Mack
1992; Russell 1987, 1992; Gifford-Gonzalez 1993; Conkey 1992c) (but see Figure 4
for two of the rare exceptions). And despite the fact that, for example, Annette Lam-
ing did join the cave exploring and documenting group of Leroi-Gourhan, Vertut and
Brezillon (all men) in the 1950’s (Laming-Emperaire 1962), there are almost never
women in the photos of the archaeological discoverers or students of imagery. This
absence of a female presence in the archaeological practices and in the production of
the imagery is an uncanny parallelism that cannot merely be referable to the prevail-
ing generalized androcentrisms of the last century.

As well, the constituting contexts of how (Paleolithic) imagery is made are
patently right out of 19th century easel painting, which was then, after all, the para-
mount definition of “art.” Compare the woodcut published in 1870 (Figure 5a) — bef-
ore the 1879 discovery of the first parietal art at Altamira — that elevates the decorated
stone plaque as if it were itself on an easel, with any number of artists’ reconstructions
for the cave wall image-making (Figure 5b). And it is a much longer discussion and

350 CONKEY

Ficurelb. “Against the backdrop of the painted cave [wall at Lascaux]a clan leader imparts the oral tradi-
tion of his people .... ‘illustrations without books’” (Pfeiffer 1 986:84-85, drawing by Llo yd Townsend; re-
printed with permission). Note that this “clan leader” is standing more or less in the same location in

Lascaux as is Ravidat (Figure la) with his torch. Itis also of interest but problematic that around the edge of
this drawing is a selection of portable art items, which derive from archaeological sites from central Europe
to France, and date between |2-20,000 years ago, thus collapsing regional and temporal variations into a
singe “Paleolithic art” (see Tomaskova, this volume)

analysis to show how the constituting contexts for the initial (and continuing) inter-
pretation of female imagery are embedded in such issues as the 19th and early 20th
century notion of the artist as male with female as the subject (Garb 1993) and the
19th century revaluing of the “erotic ideal” to be female (replacing the male) (Conkey
1992c, 1997; Mack 1992; Solomon-Godeau 1997).

CONTEXT IN THE INTERPRETIVE PROCESS

ys)
wn

FiGure 2. Three or four men “in awe” of the clay bison at the end of Le Tuc d'Audoubert (Ariege, France)
Drawing from Lascaux et son Temps by Véronique Ageorges and Henri de Saint-Blanquat (1989)

We must seek out the ways in which we reproduce the routines of exploration and
research into the interpretive frames, but this is not to expose ourselves as somehow
fraudulent or merely pawns within cultural contexts. It is much more complicated
than that. First, as Gero argues, our rationalizations for our interpretations are as much

Bs)
wn
i)

CONKEY

FiGuRE 3a. The use of scaffolding by the Abbé Glory to reach the figures on the ceiling in order to trace
them, an archaeological “routine” of rock art research (Leroi-Gourhan & Allain 1979:219)

product of our own immediate archaeological routines and practices as of some wider
generalized cultural “forces” or preferences. Second, we ourselves are most familiar
with these disciplinary practices and should thus be most informed on how they work
and why. Third, while most of the audiences for the study of “early art” invite (or de-
mand) attention to the interpretations of and for the imagery — “but what does it
mean?” — because these do not exist independently from the practitioners and the
structuring contexts within which we/they work, any consideration of interpretations
must take such practices into account in direct relation to the interpretive “results.”

The Entanglement of Modes of Thought

Furthermore, just as our practices, our routines of exploration, discovery and re-
porting, are never separate from interpretive assumptions and from our interpretive
projections, we cannot claim that we can infer, discuss, or describe the modes of

CONTEXT IN THE INTERPRETIVE PROCESS

Ww
n
Ww

FiGuRE 3b. An artist's reconstruction of the scaffolding inferred (Arl. Leroi-Gourhan 1982) for painting in
Lascaux (from Ageorges & Saint-Blanquat 1989). The label for this illustration is “How they made Las -
caux.”

thought of others as if they were somehow “out there,” separate from modes of
thought in which we participate, to which we subscribe, and through which we in-
scribe “their” modes of thought . We all readily succumb to this, to inscribing “their”
modes of thought through those of our own. As a brief illustration of this point, we can
look at the case of André Leroi-Gourhan and the legacy of “anthropological human-
ism.

354 CONKEY

Figure 4a. One of the very rare pictorial artist's reconstructions for Paleolithic cave art that features a
woman as an active image-maker (as commissioned by M. Conkey for Science Year 1987) (Conkey 1988)

No one can challenge the fact that the publications of Leroi-Gourhan, beginning
in the late 1950's (e.g., 1958) forever changed and affected the way we think about
Paleolithic art, especially in Europe (e.g., Audouze & Schnapp 1992; BSPF 1987). As
one tries to trace out the conceptual crucible within which Leroi-Gourhan's ideas and
interpretive preferences took shape, some interesting influences and intellectual rela-
tionships can be noted. For example, it is the case that Leroi-Gourhan and Claude
Levi-Strauss were the two assistant directors to Rivet, the director of the newly-
established Musée de |’ Homme. In the mid-1930’s, coincident with this new mu-
seum, had emerged a cluster of philosophical ideas that — along with the museum it-
self— was to celebrate ‘man’: The museum itself was to be a monumental unification
of ‘humanity’ under one roof (Clifford 1988:138). In describing this intellectual and
cultural movement, Clifford notes that:

“anthropological humanism begins with the different and renders it — through naming,
classifying, describing, interpreting — comprehensible. It familiarizes” (1988:145).

For most of the scholars that Rivet assembled at the new Musée —e.g., L. Dumont,
M. Griaule, M. Leenhardt, Leroi-Gourhan — the “connection,” Clifford argues, “be-
tween art and ethnography was crucial” (1988:138). Their vision of humanism was an
expanded view, forcing an opening out of what had been very local conceptions of hu-
man nature. Art was a “universal essence” (Clifford 1988:144). From this, it is not
surprising that Leroi-Gourhan took up the idea of European Paleolithic art as one of
(and equal to) the great artistic movements of the historical world, inseparable from a
coherent religious system. His visions of and for the Paleolithic fit well within the lib-

CONTEXT IN THE INTERPRETIVE PROCESS

ies)
nN
n

Ficure 4b. The other rare reconstruction of an active female in the process of “decorating caves.” Note that
here, with a barefoot, almost-hopping or perhaps dancing youth/child in the background, two assumed
stereotypes are reinforced: a) that women are consistently associated with children; and b) that footprints in
the caves are linked to youth dancing and initiation ceremonies (illustration by V. Ageorges in Ageorges &
Saint-Blanquet 1989).

356 CONKEY

eral synthetic image of ‘man’ that (as Clifford has suggested) inspired and guided the
creation of the Musée de Homme.

In describing her collaborative work with Leroi-Gourhan as she completed
her thesis (in 1957) and drew Leroi-Gourhan into the analysis of the content of Paleo-
lithic art, Annette Laming wrote of it as a “constructive period where we had the feel-
ing of seeing reborn before us a new and more coherent Paleolithic world ”
(Laming-Emperaire 1962: 2, my translation and emphasis added). This view of Pa-
leolithic peoples — as seen through their religion (Leroi-Gourhan 1964: Les religions
de la prehistoire) and their art (Leroi-Gourhan 1965: Treasures of Prehistoric Art),
can be seen to have derived from and constituted a rubric of “complete humanity”:

“All people create, love, work, worship” (Clifford 1988:144).

The point of this brief foray into intellectual history is to remind us how entangled
are our own modes of thought with those that we claim to be discovering. The legacy
of this “universal humanism,” especially in reference to the European Paleolithic art,
is still with us. A poster for an exhibition in France on human evolution proclaims
“Nous Tous Avons 400,000 Ans” (We are all 400,000 years old): the postage meter in
the famous “capital of prehistory” in France, Les Eyzies, cancels stamps with the say-
ing, *200,000 Ans de la Qualité de la Vie” (200,000 years of the high quality of life); a
children's book on prehistory presents Paleolithic life and art as “Les grottes sont les
eglises” (the caves are the churches); the Hewlett-Packard company advertises its

FiGuRE Sa. Lithograph published by Figuer (1870) before the ‘discovery’ of Altamira (1879) and Pleisto-
cene cave wall art in Europe. Note that the then-known portable art was imagined, in part, as “easel” image

making

CONTEXT IN THE INTERPRETIVE PROCESS 357

Figure 5b. One of Z. Burian's (Augusta & Burian 1960) well-known reconstructions of Paleolithic life
Note that the (male) artist (with male helpers) approaches his work in an easel-posture

358 CONKEY

own expertise with a drawing of presumed Paleolithic men, grouped together in a
cave, making typical hand prints on the cave wall, and the ad caption reads: “Leaders
have always known good advice.”

But to admit the entanglement of ow modes of thought within and for “their”
modes of thought is not saying that we should give up and its not to say that it is all
constructed. Yes, the viewer is an interpreter, and yes, interpretation changes as the
work changes and thus we cannot claim final or absolute knowledge. Many of the
core concepts that we invoke and depend upon (in our interpretive work with “early
art”) are not just cultural constructions, but are cultural concepts that have changed
over time: “art,” “symbolism,” “meaning,” “consciousness,” to name but a few. Not
only, for example, has the notion of “art” (and its handmaiden, “aesthetics”) changed
since the 18th century, but the concepts ofand for it have been (and still are) contested
(Murray 1991; see also, Tomaskova, this volume).

Catherine Lutz (1992) has brilliantly elucidated how the particular genre of
Euro-American “consciousness” and what constitutes mental life and thought entail a
whole sweep of linked features that are regularly assumed and implicated in our many
inquiries into the mental worlds of others, including those in prehistory. The linked
features — which are not inherent in “consciousness” — include the notions of objec-
tivity, control of attentional processes in the interest of solving technical problems,
non-emotionality, linear thought, and rationality, among others (Lutz 1992:73). Few
question, for example, that stone tool technologies and the linear production se-
quences that we infer for them should constitute relevant evidence for the evolution of
human symbolic behavior or of consciousness (e.g., Chase & Dibble 1987).

We thus need to “excavate” our own core concepts and the constitution of unques-
tioned background assumptions that set up evidential relations that, in turn, structure
the selection of not just relevant data but what then constitutes relevant “context.” But
this is only to force a more informed and self-reflexive engagement with the interpre-
tive process. By revealing ambiguities and uncertainties, or by revealing the entangle-
ments of our modes of thought with those we “infer” for the past, we are not
necessarily accusing views — such as those of Leroi-Gourhan or others of crucial im-
portance to the field — as being fraudulent; we are only challenging the authority with
which such views often are or have been held (Daston 1991).

Between the Caves: The Social Geography of
Paleolithic Visual Culture

In this section, I want to suggest one line of contextual analysis for under-standing
arelatively delimited set of image-making — what I am calling a “visual culture”- that
characterizes the archaeological record of the late Pleistocene of what is today the
Midi-Pyrénées region of southern France.

My own push into this arena of considering “context,” has come in part from an at-
tempt to put into practice some theoretical notions about the “social geography”
(Conkey 1984, 1990) within which mobile populations of Late Pleistocene Europe
were embedded. I believe it is plausible to think that the peoples who left what we call
Magdalenian period sites in this region were peoples who experienced a mosaic of so-
cial formations and a more-or-less mobile life throughout this region (and beyond).
Given that there are some dozens of cave sites with paintings and engravings (e.g.,
Bedeilhac, Niaux, Les Trois Fréres) and some hundreds of engraved bone, antler,

CONTEXT IN THE INTERPRETIVE PROCESS 359

ivory and plaquettes, it is clear that at least some of these people were engaged with
the production and reproduction of what appears to be a rich visual culture.

I have been inspired to probe into the “social geography” within which this took
place by at least two separate influences. On the theoretical side, I take seriously that
there is much to be gained from exploring “the referential contexts of social action”
(Hodder 1982:16): why would the making of imagery have been meaningful and to
whom, in what contexts? These are questions that are well considered in the chapters
in this volume by Rosenfeld, Parkington and Manhire, and Lewis-Williams. The in-
terpretive insights each offers derive from their being able to approximate some of the
specific referential contexts of social action — territoriality (Rosenfeld), social and
sexual relationships and statuses (Parkington and Manhire), and shamanism (Lewis-
Williams).

On the side of research practice, have been inspired by the widespread
recognition within the field that, in this so-called post-Leroi-Gourhan era with more
global and differential manifestations of early art, other emphases and approaches are
called for. “There are no masters”; we “have moved away from inclusive accounts”
(cf. GRAPP 1993; Clottes 1995).

Most simplistically, I am interested in what we can infer about the social
relationships among the people who lived in what we identify as a certain region — the
Midi-Pyrénées of France — who produced an admittedly distinctive Paleolithic “art” —
on cave walls and on bone and antler. There is, taking these images and materials to-
gether, a justification for a regional approach: “L'art contribue puissamment a indi-
vidualiser le Magdalénien des Pyrénées” (Clottes 1989:325). That is, the corpus of
imagery from the Magdalenian period - some 14-1 1,000 years ago — can be character-
ized as somewhat distinctive, and it is the art forms that invite us to conceptualize the
Magdalenian period of the Pyrénées as a culturally meaningful entity of some sort.
For example, there are certain relatively abundant and widespread characteristic ob-
jects of portable art, such as the decoupées (or animal-head bone cutouts), although
these are not only found in this region. Because there is a certain coherence to the im-
ages made, even within the wider recognizable manifestation that we call “Paleolithic
art” in Eurasia, it is likely that those who made, used, and left behind these images
were somehow participants in an “informing context” (Lewis-Williams 199 1a), that
is, an interactive, communicative, productive, and even cosmological context within
which the practices and products of imagery were meaningful, which informed the
art. How do we ever begin to access such contexts?

The research that is motivated by such questions is only beginning, but I present it
here as an example of the kind of work that lies ahead of us in the task of interpreting
the imagery that has taken hold of our attention. The intent of mentioning here this
particular — and very much “in progress” (Conkey 1994, Conkey ef al. 1995) — re-
search project is to bring into this discussion of “context” what some of my own theo-
retical assumptions are about the possible informing contexts for the production of
this visual culture we call “Paleolithic art.” And it is crucial to consider assumptions
in all approaches, no matter what the conceptual referent of “context” may be. This
work, with its own informing background and constituting assumptions will lead to-
wards certain kinds of “results”; different from work informed otherwise.

And thus, we have embarked on a two-part study, each of which is quite
traditional archaeology. However, I would like to think that it is traditional archae-
ology in the service of redefined interpretive goals. On the one hand, [have begun an
ongoing stylistic type of analysis of the bone and antler imagery among and between
sites, which I hope can lead to inferring a “stylistic geography”. That is, I hope to be

360 CONKEY

able to use very precise notions of artifact production, stylistic treatments, and pat-
terns of use of the engraved bone and antler that are often abundant in the cave sites,
where such materials are often well preserved. Dobres (1995) has already shown that,
for the unengraved bone/antler artifacts there are some distinctive patterns in the
making, repair, or abandonment of artifacts, yet that, overall, there is a notable fluid-
ity and flexibility in how such artifacts are made, repaired, and/or reused. She argues
convincingly that this flexibility was engendered by relatively fluid social forma-
tions; that the social relations of production were more open than closed, more oppor-
tunistic than rigidly structured. It will be interesting, for example, to see if the imaged
ones — with geometric and/or animal figures — are more rule-bound, made according
to more productive constraints. Will there be notable stylistic “linkages” between cer-
tain locales/sites, such as one already assumes for the sites of Mas d’ Azil and Enlene,
and what can be inferred from this about the social formations within which image-
making and using took place?

On the other hand, we have initiated a regional survey that 1s especially focused
on understanding the distributions of late Paleolithic (especially, if possible, of Mag-
dalenian, ca. 11,000-14,000 years ago) peoples within and across the landscape — es-
pecially focused on the Ariege region within which are located many important
Pyrenean cave sites. And it is the cave sites that have yielded almost all of the pre-
served archaeological evidence to date. Before this survey began, no systematic sur-
vey for evidence of Paleolithic peoples in the region — that is, in open air locations —
had been carried out, other than Simmonet’s (1981) work on locating potential
sources of flint raw materials for stone tools. What we have found to date 1s that,
across the region, there are differential distributions of distinctly Upper Paleolithic
stone tools: some locations have yielded a recurrent density that will lead to probable
coring and testing for subsurface evidence. These “lithic landscapes” (after Prine
1996) are already suggestive of particular presences — on the plateau top above the
decorated cave of Marsoulas; on the rise where one can look west towards the Mar-
soulas plateau and north towards the ridge of the Pre-Pyreéneées rising above Ausseing

and allow us to begin to “fill in” the heretofore vacant spaces “between the caves”
and to provide a sense of the pathways and lines of sight that may have linked caves,
materials, and people.

Given that these were mobile peoples, given that all our archaeological evidence
derives from the cave sites in the region, given that there is no evidence for the year-
round occupation of any cave sites, we have all at least conceptualized that people
moved in and through the region. But our archaeological reconstructions and our in-
terpretive imaginations have not included what is “between the caves.” How was the
landscape experienced, and in what ways were the landscapes imbued with all sorts of
biographical and cultural significance? How might they have been a source of/for
memory or sentiment (e.g., Myers 1986 )? This kind of a “*geography of social action”
looks at how people “used” the region, at the landscapes of “being in” and experienc-
ing the late Paleolithic world at this regional scale. How might these movements and
memories been part of the “informing contexts” within which a rich and abundant
material and visual culture was produced? The two “geographies” that I envision we
might infer are intended, in analysis, to be played off, dialectically, against each
other: How do the patterns of each inflect upon the other? Where are the resonances,
the dissonances?

The “social geography” approach is based ona theoretical idea about what human
symbolic behavior — such as making images, pictures, signs or whatever — is about.
Rather than taking human consciousness (or, even altered states of consciousness),

CONTEXT IN THE INTERPRETIVE PROCESS 361

symbolism, and meaning-making to be reified properties of the mind, I take symbolic
behavior and meaning-making to be an emergent property in and of social action.
Thus, I have deliberately chosen to problematize the nature and dynamics of social
formations and social life during the Magdalenian in the Midi-Pyrénées, as expressed
through, experienced, and constituted by the liaisons and movements through and
within a regional landscape. Here indeed, there are differential patternings in the en-
gagement with material culture (e.g., Clottes 1989; Dobres 1995); there is little re-
dundancy from site to site (Conkey 1993). Each cave site has yielded its own
relatively unique combination of imagery and media, yet there is something charac-
teristically “Pyrenean” about the Magdalenian materials.

Although outside the immediate scope of my research, this is a region for which
diverse archaeological evidence might well be marshaled to begin to make the requi-
site case for the kinds of sociopolitical contexts that Lewis-Williams (this volume and
elsewhere, e.g., 1982, 1991b) has invoked as those within which shamans might have
taken over the control and production of arcane knowledge, which included certain
image-making in, or left in, certain caves. In this volume, Lewis-Williams does note
that despite the rather inclusive appeal of the art-as-shamanistic interpretation that he
has been developing, nonetheless, this must be demonstrated within specific and his-
toricized instances, as he suggests for the site of Gabillou (Dordogne, France).

The interpretation of a shamanistic art for the European Upper Paleolithic cannot
be sustained without wedding the analysis of the images per se and their immediate
contexts of cave wall shapes and galleries (through which one wriggles) to the socio-
political contexts and social formations. These can only be inferred archaeologically:
that is, from information about sites, landscapes, movements, activities, technolo-
gies, group dynamics, and other patternings. This is why what the inquiry is today is
one that is “beyond art.” As Lewis-Williams himself has argued, one must work to-
wards specifying the “informing context” of the imagery, which “specifies the social
relations and the cognitive processes that informed the art and linked it to broader
[San] social formations” (Lewis-Williams 1990:133; 199 1a). But even where there 1s
more information on the “informing contexts,” such as in the pioneering work on the
shamanism of southern African rock art, the interpretation is still being debated and
refined (Parkington & Manhire, this volume; Manhire et a/. 1985; Solomon 1992).

Assessing Context

While proclaiming we must then be more “contextual” and pay more attention to
“it” (Le., context), there is possible delusion here (see Davis 1989). While it is easy to
end our evaluations and critiques of past research attempts and shortcomings with a
coda of “contextualize, contextualize, contextualize,” this could be somewhat of a
false promise. Is this a false, thinly-veiled positivist kind of promise that if only we
could “find” context (s), we'd “have” meanings; we would “find” meanings that con-
text would (somehow) reveal to us?

Of course, we — as in art history— can justify the call for (self-critical) contexts
onthe grounds that it is a suitable response to the ahistorical or noncontextualizing ap-
proaches of structuralism or other homogenizing or utopian accounts that have pre-
vailed in the interpretation of Paleolithic art. Note Rosenfeld’s critique (this volume)
of how normative descriptions of motifs and images have perhaps precluded the inter-
pretive potential of rock art. In art history, the call for contexts is justified as a valid
response to the formalistic connoisseurship (and its ahistorical procedures) that has

362 CONKEY

dominated that field (Davis 1989:25). To a great extent, the breathless celebration
that has long characterized many approaches to Paleolithic cave art, for example, is
but another variant of this connoisseurship (Conkey 1997).

While I think it is an extremely valid, if not crucial, question to ask for the Euro-
pean Paleolithic imagery, “What were people doing in the caves?”, one must ask is the
cave. as site, fhe unquestioned, bounded entity for analysis and interpretive frame?
Further, as Ucko (1992) has argued, and as a few really detailed studies of individual
caves (e.g., Le Magdeleine [Rouzaud et a/. 1989]; Villars [Delluc & Delluc 1974];
Aldene [Vialou 1979)]) have shown, just interpreting the “use” and “use-life” of a
cave 1s more complex than ever imagined. While the cave-as-context, for example,
should provide additional lines of evidence in our quest for understanding, what we
must avoid is the false hope that once we define context, then meaning will follow.

Thus, the focus on context and thus, a view of “art” as a social activity and
practice, doesn't necessarily move away from the view that these are objects or im-
ages with certain (e.g., aesthetic or technological) qualities or a certain content. But it
is always easy to lean too far, especially for me, as an anthropologist who is trained to
and almost instinctively prefers to lean towards the social and cultural. There is the
potential problem of considering “art” as primarily or on/v as the trace or symptom of
those (social) facts” (Davis 1990), neglecting serious consideration of such things as
the aesthetic or spiritual experiences. Perhaps this is what Alpers (in Schoch 1988)
meant in differentiating between the making of pictures and the circumstances of pic-
tures, or between those images that are more or less profitably studied in terms of the
culture around them.

This differentiation has been, of course, a long-standing tension in the study of im-
ages and “art,” with two ‘camp’: those who study “art” as object, as aesthetic produc-
tion and those who study it as social practice, as symbolic production. This is why,
when visiting painted caves in France, a contemporary painter once demanded of me,
and with exasperation, “Was the imagery really only about the differing strategies?”
that I was presenting to him as the various explanations for Upper Paleolithic art: in-
formation exchange, population refugia, adaptive signalling of the social order. Of
course, what is needed is a dialogue, not a choosing of sides, an either/or. The making
of pictures is not outside the realm of the circumstances of picture; we may feel an
analytical need for separation but one is not possible without the other.

If there is any one (albeit disappointing) thing to be said about the study of
“meaning,” especially in regard to Paleolithic imagery, it is nor that 1f we can only
circumscribe context, the meaning(s) will emerge or be ‘revealed.’ As Sperber wrote,
meaning is not a useful “descriptive category,” buta ‘misleading metaphor”: “Itis not
possible to circumscribe the notion of meaning in sucha way that it could still apply to
the relationship between symbols and their interpretation” (Sperber 1975: 11,33). Al-
though the pursuit of contexts will not “reveal” meaning, which has no stable identity
anyway, we have little to lose from not probing into varying contextual circumstances
but also expanding our ideas as to what we mean by the concept of “context(s).”

Closing Thoughts

The idea of recognizing “context-as-cultural-construction” is not to abandon any
hopes of understanding about the contexts and cultural frames within which meaning,
consciousness, symbolism and so forth came into play in the lives of past peoples. By
recognizing the role of culture im the construction of the experience of self, other, and

CONTEXT IN THE INTERPRETIVE PROCESS 363

the world (after Lutz 1992), I think we will actually find that that which we can be
talking about is much wider and deeper than what we have done so far.

By taking the “interpretive turn” in the human sciences — an explicit retreat from
positivist science and from the false hope that the human sciences will someday reach
“an age of paradigm’ as in the natural sciences — our attention, to paraphrase Rabinow
and Sullivan (1987: 5-7). is refocused on “the concrete varieties of cultural meaning,”
where we recognize that our baseline practices in the human sciences are socially
constituted actions that are intersubjective, in which one “denies and overcomes the
almost de rigeur opposition of subjectivity and objectivity.” What the “interpretive
turn” means for an archaeology of imagery is not that it is just another “methodologi-
cal option among a growing number of investigative tools” (Rabinow & Sullivan
1987: 2). Rather, the interpretive turn is a process that has “the capacity to challenge
the practices of knowing in our culture. Knowledge is not — despite the enthusiasm
and persistence with which it has been taken to be — a ‘technical project’.” Rather is
“inescapably practical and historically situated” (Rabinow & Sullivan 1987:2).

While we are all easily seduced into studying these ancient images and their mak-
ers in order to extract or “find” some mythical ideals, what we must do instead is take
up ancient visual culture as a subject of engaged study in which we ourselves are re-
minded of our own historical specificity rather than loosing ourselves in mythical ide-
als (after Kampen 1993). If we can do this — with both empirical depth and
imaginative power — 1am convinced that we will learn something about the past and
the present, and about the relation between the two.

Acknowledgments

This paper incorporates ideas that have been presented ina variety of conferences
and symposia, and I want to thank the organizers of these for having invited and sup-
ported my participation: especially, the Interpretive Archaeologies Conference
(Cambridge, England, 1990) with thanks to lan Hodder, Michael Shanks and Alexan-
drea Alexandrei; and the Alta Conference on Rock Art (Alta, Norway, 1992), with
thanks to Knut Helskog and Bjonar Olsen. Various commentaries on my presenta-
tions have been incorporated into this paper, and I am especially grateful to Jean
Clottes, Betty Meehan, and David Lewis-Williams. The paper also incorporates ideas
from the presentations and commentaries at the Society for American Archaeology
session (1993) and from the outstanding pre-SAA meeting, the Oregon Archaeologt-
cal Retreat, so graciously and enthusiastically hosted by Jean and Ray Auel: and from
the closing commentary at the Wattis Symposium, California Academy of Sciences
(March 1995). 1am most grateful to those at the Academy forall they did to make that
symposium possible and enjoyable.

I have also profited directly from discussions in South Africa during a visit there
arranged primarily by John Parkington and the University of Cape Town, with assis-
tance from the University of the Witwatersrand (David Lewis-Williams, Lyn Wad-
ley. and Tom Huffman), from Aron Mazel, and especially supported by the Human
Sciences Research Council. Although the particular configuration of ideas presented
here is my own. there would be little to say without the thoughts and writings of a mul-
titude of scholars — and especially recent Berkeley students Marcia-Anne Dobres,
Justin Hyland, Heather Price and Silvia Tomaskova— to whom Lam deeply indebted.

364 CONKEY
Literature Cited

Ageorges, V. & H. de Saint-Blanquat. 1989. Lascaux etson Temps Casterman, Paris, France.

Augusta, J. & Z. Burian. 1960. Prehistoric Man. P. Hamlyn, London, U.K.

Audouze, F., & A. Schnapp. 1992. Dossier 1: Un homme, une oeuvre: Andre Leroi-Gourhan.
Les Nouvelles de l'Archéologie 48-49:5-54.

Bahn, P. 1978. Water Mythology and the Distribution of Palaeolithic Parietal Art. Proc. Pre-
hist. Soc. 44:125-134.

Bégouén, H. & H. Breuil. 1958. Les Cavernes du Volp: Trois Freres, Tuc d'Audoubert, a
Montesquieu-Avantes (Ariege). Arts et Meétiers Graphiques, Paris, France.

Bellier, C. 1984. Contribution a l'étude de l'industrie osseuse préhistorique: les contours de -
coupées du type ‘tétes d'herbivores’. Bull. Soc. Roy. Belge d'Anthrop. Préhist. 95:2\-34,

Bryson, N. 1983. Vision and Painting: The Logic of the Gaze. Yale University Press, New Ha-
ven.

BSPF: Bulletin de la Sociéte Préhistorique Frangaise. 1987. Hommage de la SPF a Andre
Leroi-Gourhan. Bull. Soc. Préhist. Frangaise 84:291-425.

Cattelain, P. 1980, Quelques aspects des propulseurs paléolithiques. Bull. de la Soc. Préhist
Francaise 77:131-1306.

Chase, P. & H. Dibble. 1987. Middle Palaeolithic symbolism: A review of current evidence and
interpretations. J. Anthropol. Arch. 6:263-296.

Chesney, S. 1991. Max Raphael's contribution to the study of prehistoric symbol systems.
Pages 14-22 in P. Bahn & A. Rosenfeld, eds., Rock Art and Prehistory. Oxbow Mono-
graphs # 10. Oxbow Press, Oxford, U.K.

Clifford, J. 1988. On ethnographic surrealism. Pages 117-151 in J. Clifford, ed., The Predica-
ment of Culture. Harvard University Press, Cambridge

Clottes, J. 1989. Le Magdalénien des Pyrénées, Actes du Colloque de Mayence (1987), Etudes
et Recherches Archéologique de I'Université de Liege, 38. Universite de Liege, Liege, Bel-
gium.

1993. Paint analyses from several Magdalenian caves in the Ariége region of France. J.
Arch. Sci. 20:223-235.

___, ]995. Perspectives and traditions in Paleolithic rock art research. /n K. H. A. B, Olsen,
ed., Social and Political Perspectives on Rock Art, The Alta Conference. Instituttet fur
Sammenlignende Kulturforskning, Tromso, Norway.

Clottes, J., M. Menu & P. Walter. 1990. New light on the Niaux paintings. Rock Art Res. 7:21-
26.

Conkey, M. 1980. The identification of prehistoric hunter-gatherer aggregation sites: The case
of Altamira. Curr, Anthropol. 21:609-630.

___. 1984. To find ourselves: Art and social geography of prehistoric hunter gatherers.
Pages 253-276 in C. Schrire, ed., Past and Present in Hunter Gatherer Studies. Academic
Press, Orlando.

_ 1987. New approaches in the search for meaning? A review of research in “Palaeolithic
Art”. J. Field Arch. 14:413-430.

_ 1988. Images from the Ice Age. /988, Science Year: The World Book Annual Supple-
ment. pp. 40-55. World Book, Inc, Chicago.

__. 1989. The structural analysis of Paleolithic art. Pages 135-154 in C. C. Lamberg-
Karlovsky , ed., Archaeological Thought in America. Cambridge: Univerity of Cambridge
Press, U.K.

CONTEXT IN THE INTERPRETIVE PROCESS 365

. 1990. L'Art mobilier et l'établissement de geographies sociales. Pages 163-172 in L'Art
des Objets au Paléolithique. Tome 2: Les voies de la recherche. (Actes des Colloques de la
Direction du Patrimoine, Foix-le Mas d'Azil, 1987). Ministre de la Culture, Paris, France.

. 1992a. L'approche structurelle de l'art paléolithique et I'heritage d'André Leroi-
Gourhan. Les Nouvelles d'Archéologie 48-49:4 1-45.

. 1992b. Les sites d'agregation et la réparation de l'art mobilier, ou: Y-a-t-il des sites d'a-
gregation magdalénien? Pages 19-25 in J.-P. Rigaud, H. Laville and B. Vandermeersch,
eds., Le Peuplement Magdalénien: Paléogeographie Physique et Humaine. Editions du
Comité des Travaux Historiques et Scientifiques. Documents Préhistoriques, 2: Actes du
Colloque de Chancelade (October 1988), Paris, France.

.1992c¢. Mobilizing ideologies: The archaeo-logics of Paleolithic “art”. Paper presented
at the Annual Meetings of the American Anthropological Association, San Francisco.

. 1993. Defining “contexts”: Interpreting the “art” of the Magdalenian. Paper presented
at the Annual Meetings of the Society for American Archaeology, St. Louis.

. 1994. Magdalénien de Pyrénées: Ariége, Haute-Garonne. Operations Interdépartemen-
tales, Midi-Pyrénées. Bilan Scientifiques 1993:204-205. Direction Regionale des Affaires
Culturelles Midi-Pyrénées, Service Regional de l'Archéologie, Toulouse, France.

. 1995. Making things meaningful: Approaches to the interpretation of the Ice Age 1m -
agery of Europe. Pages 62-77 in I. Lavin, ed., Meaning in the Visual Arts: Essays in Honor
of E. Panofsky. Princeton University Press, Princeton.

. 1997. Mobilizing ideologies: Paleolithic ‘art’ gender trouble, and thinking about alter-
natives.Pages 172-207 in L. Hager, ed., Women in Human Evolution. Routledge, London,
UK.

_in press. Paleovisions: Interpreting the Imagery of Ice Age Europe. W.H.Freeman and
Co, New York.

Conkey, M., L. Rowntree & V. Andrieu. 1995. Magdalénien des Pyrenees: Ariége, Haute-
Garonne. Opérations Interdépartementales, Midi-Pyrénées. Bilan Scientifiques 1994,
223-224. Direction Regionale des A ffaires Culturelles Midi-Pyrénees, Service Regional de
l'Archéologie, Toulouse, France.

Criado-Boado, F. & R. Penedo Romero. 1989. Cazadores y salvajes: una contraposicion entre
el arte paleolitico y el arte postglaciar Levantino. Munibe 41:3-22.

d'Errico, F, 1989. L'Art Grave Azilien. Analyse Microscopique, Reconstitution Gestuelle, Sig-
nification. Ph. D. Dissertation, Museum National d'Histoire Naturelle, Paris, France.

. 1992. Technology, Motion, and the Meaning of Epipalaeolithic Art. Curr. Anthropol
33:94-109,

Daston, L. 1991. Marvelous facts and miraculous evidence in early modern Europe. Criticial
Inquiry 18:93-124.

Davis, W. 1989. Review of Object and Intellect: Interpretations of Meaning in African Art, ed-
ited by H. J. Drewal. Afr. Arts XX11:24-32.

. 1990. Style and history in art history. Pages 18-31 in M. Conkey & C. Hastorf, eds., The
Uses of Style in Archaeology. Cambridge University Press, Cambridge, U.K.

Delluc, B. & G. Delluc. 1974. La grotte ornée de Villars (Dordogne). Gallia Preéhistoire
17:1-67.

Delporte, H. 1990 L'Image des Animaux dans L'Art Préhistorique. Picard, Paris, France.

Dobres, M.-A. 1995. Gender in the Making: Late Magdalenian Social Relations of Production
in the French Midi-Pyrénées. Ph.D. Dissertation, Department of Anthropology, University
of California at Berkeley.

Fabian, J. 1983. Time and the Other: How Anthropology Makes its Object. Columbia Univer-
sity Press, New York.

Figuer, L. 1870. L'Homme Primitif. Hachette, Paris, France.

366 CONKEY

Gamble, C. 1982. Interaction and Alliance in Palaeolithic Society. Man (NS) 17:92-107.

. 1991. “The Social Context for European Paleolithic Art.” Proc. Preshist. Soc, 57:3-15.

Garb, T. 1993. The forbidden gaze: Women artists and the male nude in late nineteenth-century
France. Pages 33-42 in K.Adler & M. Pointon, eds., The Body Imaged: The Human Form
and Visual Culture since the Renaissance. Cambridge University Press, Cambridge, U.K.

Gero, J. 1996. Archaeological practice and gendered encounters with field data. #7 R. Wright,
ed., Gender in Archaeology: Essays in Research and Practice. University of Pennsylvania
Press, Philadelphia.

Giftord-Gonzalez, D. 1993. You can hide, but you can't run: Representations of women's work
in illustrations of Paleolithic life. Vis. Anthrop. Rev. 9:3-21.

Gonzalez-Morales, M. 1991. From hunter-gatherers to food producers in northern Spain:
Smooth adaptive shifts or revolutionary change? Pages 204-216 in G. A. Clark, ed., Per-
spectives on the Past: Theoretical Biases in Mediterranean Hunter-gatherer Research,
University of Pennsylvania Press, Philadelphia.

GRAPP (Groupe de Réflexions sur |'Art Paliétal Paléolithique), ed. 1993. L'Art Parietal Paléo-
lithique: Techniques et Méthodes d'Etude. Comité des Travaux Historiques et Scienti-
fiques, Paris, France.

Hodder, I. 1982. Theoretical Archaeology: A reactionary view. Pages I-16 i 1. Hodder, ed.,
Symbolic and Structural Archaeology. Cambridge University Press, Cambridge, U.K.
Jochim, M. 1983. Palaeolithic cave art in ecological perspective. Pages 212-219 in G. Bailey,
ed., Hunter-gatherer Economy in Prehistory: A European Perspective. Cambridge Univer-

sity Press, Cambridge, U.K.

Kampen, N. 1993. Ancients and moderns. Review of Pornography and Representation in
Greece and Rome, edited by Amy Richlin (New York: Oxford University Press, 1992).
Women's Review of Books X:\ 1-12.

Laming-Emperaire, A. 1962. La Signification de l'Art Rupestre Paléolithique. Picard, Paris,
France.

Leroi-Gourhan, A. 1958. Repartition et groupement des animaux dans l'art parietal paleo-
lithique. Bull. de la Soc. Préhist. Francaise 55:515-527.

1965. Treasures of Prehistoric Art. Harry Abrams, New York,

1966. La religion des cavernes: Magie ou métaphysique. Sciences et Avenir 228

1982. The Dawn of European Art: An Introduction to Paleolithic Cave Painting . Cam-
bridge University Press, Cambridge, U.K.

Leroi-Gourhan, Arl. 1982. The archaeology of Lascaux Cave. Sci. Amer. 246:104-113.

Leroi-Gourhan, Art. & J. Allain. eds. 1979. Lascaux Inconnu. X11 Supplement a Gallia Préehis-
toire. CNRS, Paris, France.

Lewis-Williams, J. D. 1982. The Economic and Social Context of Southern San Rock Art.

Curr. Anthropol. 23:429-449.

_. 1990. Documentation, analysis, and interpretation: Dilemmas in rock art research. A re-
view of The Rock Paintings of the Upper Brandberg. Part I: Amis Gorge, by H. Pager
(1989). S Afr Arch, Bull. 45:126-136.

—. 199 1a. Upper Palaeolithic art in the 1990s: A Southern African perspective. S Afr. J.
Sci. 87:422-429.

—. 1991b. Wrestling With analogy: A methodological dilemma in Upper Palaeolithic art
Research. Proc. Prehist. Soc. 57:149-162.

Lorblanchet, M. 1980. Peindre sur les parois de grottes. Dossiers de I'Archéologie 46:33-39.

Lutz, C. 1992. Culture and Consciousness: A problem in the anthropology of knowledge Pages
64-87 inF. Kessel, P. Cole & D. Johnson, eds., Se/fand Consciousness: Multiple Perspec-
tives. Lawrence Erlman Associates, Hillsdale, NJ.

CONTEXT IN THE INTERPRETIVE PROCESS 367

Mack, R. T. 1992. Gendered site: Archaeology, representation, and the female body. Pages
235-244 in A. S. Goldsmith, S. Garvie, D. Selin & J. Smith, eds., Ancient Images, Ancient
Thought: The Archaeology of Ideology. Archaeological Association, University of Cal-
gary, Calgary, .

Manhire ,T., J. Parkington & R. Yates. 1985. Nets and fully-recovered bows. World Arch. 17:
161-174.

Murray, S. 1991. The Anthropology of Art. ms. on file: Department of Anthropology, Univer-
sity of California at Berkeley.

Myers, F. 1986. Pintupi Country, Pintupi Self Sentiment, Place, and Politics anong Western
Desert Aborigines. Smithsonian Institution Press, Washington, DC.

Pfeiffer, J. 1986. Cro-Magnon hunters were really us, working out strategies for survival.
Smithsonian 17:74-85.

Prine, E. 1996 The Ethnography of Landscape. Ph.D. Dissertation. University of California,
Berkeley.

Rabinow, P. & W. Sullivan. 1987. The interpretive turn: A second look. Pages 1-30 in P. Ra bi-
now & W. Sullivan, eds., /nterpretive Social Science: A Second Look. University of Cali-
fornia Press, Berkeley.

Rouzaud, F., A. Bisio, J. Lautier & L. Soulier. 1989. Grotte de la Magdeleine-des-Albis, a
Penne (Tarn). Bull. Soc. Préhist. Ariége-Pyrénées 44:21-69.

Russell, P. 1987. Women in Upper Paleolithic Europe. M. A. Thesis. University of Auckland,
Auckland, New Zealand.

. 1992. Men only? The myths about European Paleolithic artists. Pages 346-351 in D. W.
Willows, ed., The Archaeology of Gender. Proceedings of the 22nd annual Chacmool con -
ference. Archaeological Association of the University of Calgary, Calgary, Canada.

Sauvet, G. 1988. La communication graphique paléolithique: De l'analy se quantitative d'un
corpus de données a son interpretation sémiologique. L ‘Anthropologie 92:3-16.

Sauvet, G. & A. Wlordarezyk. 1992. Structural interpretation of statistical data from European
Palaeolithic Cave Art. Pages 223-234 in A. S. Goldsmith, S. Garvie, D. Selin & J. Smith,
eds., Ancient Images, Ancient Thought: The Archaeology of Ideology. Archaeological As-
sociation, University of Calgary, Calgary, Canada.

Schoch, R. 1988. California Q and A: A Conversation with Svetlana Alpers. Calif, Monthly
September: 12, 13, 16.

Simmonet, R. 1981. Cartes des gites 4 silex des Pré-Pyrénées. Congres Préhstorique XX!
(1979). Quercy, France. 1:308-323.

Solomon, A. 1992. Gender, representation and power in San art and ethnography. J. of An-
throp. Arch. 11:291-329.

Solomon-Godeau, A. 1997. Male-trouble: A Crisis in Representation. Thames and Hudson,
London, U.K.

Sperber, D. 1975. Rethinking Symbolism. Cambridge University Press, Cambridge, U.K.

Ucko, P. 1992. Subjectivity and the recording of Palaeolithic cave art. Pages 141-180 i J.
Clottes & T. Shay, eds., The Limitations of Archaeological Knowledge. Etudes et Recher-
ches Archéologique de |'Universiétde LiPge, 49. Université de Liege, Service de Préhis-
toire, Liege, Belgium.

Vialou, D. 1979. Grotte de l'Aldéne a Cesseras (Herault). Gallia Préhist. 22:1-85.

Index

37]

Index

A

Abri Blanchard 98
Abri Castanet 98
Abri de la Souquette 98
Acheulian Industry 54
Ain Mallahal 65
Altamira 30, 37, 46, 149, 193, 203, 270
AMS (accelerator mass spectrometry) dating, see archaeometry, methods 24
Amud 57
Anatolia, sites in 172
anatomically modern humans, see Homo sapiens sapiens 53
animals, depictions of 62, 65, 114, 129, 180, 190, 220, 226, 329
carnivores 242
felines 337
herbivores 242
See Also individual animals (aurochs, bison, etc.)
related to subsistence patterns 191 - 192
anthropomorphic figurines, see also female images 107, 196
Gravettian 107
Grimaldi 114
Levant 167
raw materials 109
relationship to object form 114
archaeometry, methods 19
cation ratio method 21, 28, 39
radiocarbon 19, 207
14C AMS dating 24, 26, 28, 39, 45 - 46, 49, 138, 149
architectural art 251
Arene Candide (delle) 218
Arnhem Land 293, 295
art, concept of 125, 269 - 270, 281, 362
Aude-Heérault 206
Aurignacian Period 10, 53
European 53, 95, 217
Levantine 60
personal ornaments 95, 167
aurochs, depictions of 68, 168, 222
Australia, art of 39, 136, 290
Azilian tradition 228

B

Balzi Rossi 218
bas-reliefs 207
beads 95, 254

372 INDEX

basket-shaped 98, 107
ivory 96
stone 98
use in the Levant 166
Becov 54
Bedeilhac 75, 359
Berekhat Ram 57
Bernifal 203
bison, depictions of 65
bracelets 248
Breuil, Henri 37, 203, 323, 34
Brno 249

n

&

Cafer Hoyuk 172
Cangyuan 39
Castillo 30, 329
Catal Huyuk 161, 172
cation ratio method, see archaeometry, methods 28
caves, light levels in 205, 336
caves, topography of 203, 328, 333
Cayonti 172
ceramic objects 242
chaine operatoire 10,94, 346
in fabrication of female scultpures 108
in fabrication of personal ornaments 95
Chalcolithic 161
Levant 165
Chatelperronian Period 53
Chauvet 20, 30, 149, 176, 336
age of images 149, 207
climatic stress 256
Coa Valley (Foz Coa) 21
Coa Valley, see petroglyphs 27
Comarque 329
Combarelles 203
context, concept of 128, 267, 279, 281, 343, 346, 359
Cosquer 20, 30, 39, 46, 150, 176, 206, 333
age of images 207
Cougnac 30, 44, 46

D

dei Cervi 224

Dell’ Addaura 226

delle Veneri 218

demography of Late Pleistocene populations 256

INDEX

designs used in cave paintings 42, 195, 208
See Also animals, depictions of
See Also female images
geometric signs 42, 93, 226, 326
claviforms 42, 326
geometric 327
Devil’s Lair 138
Dolni Vestonice 240
Dolni Vestonice 265, 278

jot

El Castillo 46

El Wad 75

eland, depictions of 302, 316

elephants, depictions of 317

En Gedi 179

Enlene 38, 210

engraving 31,57, 142, 203
Australian 31
Franco-Cantabrian 208
Hayonim 77
Quneitra 57

entoptic phenomenal 76

Epigravettian Period 220

Epipaleolithic Period 70 - 71
Levant 161

F

female images 75, 108, 114, 144, 172, 179, 218, 254, 307
"Venus figurines" 108, 144, 196, 250, 254 - 255, 278
concept of femaleness 250
fertility symbols 93
pregnancy, depictions of 114, 117
Fontanet 38, 41, 206, 345
Font-de-Gaume 37, 203
Foz Céa, see petroglyphs 27
Franco-Cantabrian region, cave paintings of 5, 7, 12, 68, 176, 189, 220
Funerary art 240

G

Gabillou (Le) 334
Gagarino 245
Gargas 46, 203
GeiBenklésterle 98

374 INDEX

gender, role of 302

geopolitical considerations 272, 277

Giant Horse site 30

Gilgal 171

Giovanna 224

Golan Heights 54

Gravettian Period 10, 54, 204, 218
Central and Eastern Europe 240

Graziosi, Paolo 231

Grimaldi 109, 218

H

hallucination(s) 324
hand stencils 150, 207, 210, 212, 291, 296
Hayonim 60, 71, 166 - 167

Hayonim block 75, 84, 170, 178

notation 86

healing, depictions of 310
Homo sapiens sapiens 53,175, 344
horse, depictions of the 62, 178, 220
humans, images of 194, 213, 222, 226, 231, 242, 303, 306, 308, 326
hunter-gatherer societies 289, 303, 322 - 323
hunting 195, 302, 316

hunting magic 223

implements 302, 307

scenes 306, 310

I

Ignat'evskaia 25]
initiation scenes, depictions of 316 - 317
Isturitz 98, 208
ivory 100, 102, 105
See Also mammoth tusk
physical properties of 102

J

Jericho 171
Judd’s Cavern 136

K

Kapovaia 251
Karain 67

INDEX

be
~—
n

Kebara 57, 178

Kharaneh 164

Khotylevo 245

Koonalda cave 136, 211
Kostenki-Avdeevo Culture 244,275
Ksar 'Akil 178

L

Laming-Emperaire, Annette 204, 321
language 126, 151

Lascaux 2, 37, 337, 348
Laugerie-Haute 37

Laurie Creek 136

Le Reseau Clastres 38, 41
Leroi-Gourhan, Andre 6, 20, 45, 62, 94, 127, 196, 204, 321, 344, 354
Les Mallaetes 130

Les Trois-Freres 38, 65, 210
lighting, effects of cave 205, 211
lion, depictions of 222

M

Madeleine (La) 64
Magdalenian Period 12, 204, 210
Cantabrian 190
Malakunanja 138
male images 307
mammoth tusks 100, 102, 105
mammoth, depictions of 64 - 65
Mandu Mandu 138
Marsoulas 203
Mas d'Avil 75
Mauss, Marcel 94
Mesolithic 190
Middle Paleolithic 57
minimum number of individuals (MNI) 192
Moravia 240, 274

Pavlov Culture 240
mortuary practices 164

Levant 164
Mousterian 95
Mouthe (La) 37, 203
Mungo, Lake 138
Mureybet 171, 173

376 INDEX
N

Nahal Ein Gev 164
Nahal Hemar 165, 172, 179
Natufian tradition 71, 75, 79, 164, 168, 179
Nauwalabila 138
Neandertals 53
Levantine 54
Neolithic 161
Early 171
iconography 171
Levant 164
Netiv Hagdud 171
neuropsychology 325, 337
Neve David 164
Niaux 20, 30, 37, 42, 149, 203, 330, 359
Salon Noir 38, 211
Nisula 27

O

Ohallo 164

Okiizini 67, 73

oxalate salts, use in dating 26, 29, 31
oxygen plasma oxidation 29

P

Paglicci 220

paints used in cave paintings 23, 37
charcoal particles 23, 26, 47
oil paint 41
organic binders 23, 31, 37

DNA content of 39

paint analyses 39

paints used in cave paintngs
organic binders 31

Parpallo 129, 133

Pavlov 240, 265

Pavlov Culture 275, 278

pebbles, painted 228, 230

personal ornamentation, see also beads 95, 140, 166
methods of manufacture 102
raw materials used in 98

mammoth ivory 100

petroglyphs 21, 290
Coa Valley 21, 23, 39

phallic images 75, 80-81, 171

INDEX

plaquettes 129, 147, 359
Polesini 223

Predmosti 240

Prince (Grotte du) 109, 218

Q

Qafzeh 166
Quercy caves 38, 214

R

radiocarbon dating, see archaeometry, methods 19
red ocher 105
use in the Levant 166
Réseau Clastres 206
rock art 127
Australia 127, 136, 289, 295
Cantabrian 190
South Africa 301, 322
rock shelters 301
Russia (European) 244

S

Saint Jean de Verges 98
Salibiya 171
San ethnography 316, 322
San 337
San(t) Gregori 130
seasonality, adaptations to 81
Natufian 82
shamanism 321 - 323, 327, 334, 361
shamans 361
depictions of 313
silica coverings of rock paintings 26
Skhul 166
subsistence, depictions of 310
Sungir 244
Swabia 140
symbolism and symbolic behavior 161, 189, 194
symbols 127, 140, 142, 153

378 INDEX
T

Taphonomy 142, 203

Tata 54

Tel Ramad 172

territoriality 289, 294, 297
trance(s), depictions of 310, 325
Treasure Cave 179

Trois Freres (Les) 359

U

Uluzzian 217

Upper Paleolithic 9, 14, 133, 136
Australian 140
Cantabrian 189
iconography 75,93
Levant 161
transition 57, 86
western European 64, 125

Urkan er-Rubb 168

Vache (La) 38, 42
Vogelherd 53, 140

W

Wandjina art 292 - 293
Willendorf 265, 276, 278
Wunan 293

Xx
x-ray art 294
r

Yarmukian culture 173

vA

Zagros 178
Zubialde 40

a

IAN INSTITUTION LIBRARIES

I

wilt

3 9088

CALIFORNIA
ACADEMY
SCIENCES

GOLDEN GATE PARK

ARCHAEOLOGY / ANTHROPOLOGY / ART HISTORY

"This is the book which puts the interpretation of
Paleolithic art where it sensibly belongs—in the hands of archaeologists.”
Clive Gamble, author of The Paleolithic Settlement of Europe

“A benchmark publication that sets a new standard for studies
of our evolving cognitive capacities... Must reading for all those
interested in the emergence of human symbolic behavior.”
Geoffrey A. Clark, author of Perspectives on the Past

Margaret W. Conkey is Professor of Anthropology at the University of California at Berkeley
g y I
Olga Soffer is Professor of Anthropology at the University of Illinois

Nina G. Jablonski is Irvine Chair of Anthropology at the California Academy of Sciences

ISBN 0-940e28-37-8

9°780 HM 228

Distributed by the University of California Press ;
Printed in the United States of America