Vous êtes sur la page 1sur 444

origins of religion,

cognition and culture

Religion, Cognition and Culture

Series Editors: Jeppe Sinding Jensen and

Armin W. Geertz, Aarhus University

This series is based on a broadly conceived cognitive science of religion. It

explores the role of religion and culture in cognitive formation and brings
together methods, theories and approaches from the humanities, social sci-
ences, cognitive sciences, psychology and the neurosciences. The series is asso-
ciated with the Religion, Cognition and Culture (RCC) research unit at the
Department of Culture and Society, Aarhus University (www.rcc.au.dk).


The Burning Saints: Cognition and Culture

in the Fire-walking Rituals of the Anastenaria
Dimitris Xygalatas

Mental Culture: Towards a Cognitive Science of Religion

Edited by Dimitris Xygalatas and William W. McCorkle Jr

Past Minds: Studies in Cognitive Historiography

Edited by Luther H. Martin and Jesper Srensen

Origins of Religion, Cognition and Culture

Edited by Armin W. Geertz

Religion as Magical Ideology: How the Supernatural

Reflects Rationality
Konrad Talmont-Kaminski

Religious Narrative, Cognition and Culture:

Image and Word in the Mind of Narrative
Edited by Armin W. Geertz and Jeppe Sinding Jensen
Origins of Religion,
Cognition and Culture
Edited by
Armin W. Geertz
First published in 2013 by Acumen
Published 2014 by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
711 Third Avenue, New York, NY 10017, USA

Routledge is an imprint of the Taylor & Francis Group, an informa business

Editorial matter and selection Armin W. Geertz, 2013

Individual chapters individual contributors

This book is copyright under the Berne Convention.

No reproduction without permission.

All rights reserved. No part of this book may be reprinted or reproduced or utilised
in any form or by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying and recording, or in any information
storage or retrieval system, without permission in writing from the publishers.

Practitioners and researchers must always rely on their own experience and
knowledge in evaluating and using any information, methods, compounds, or
experiments described herein. In using such information or methods they should
be mindful of their own safety and the safety of others, including parties for whom
they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors,
or editors, assume any liability for any injury and/or damage to persons or property
as a matter of products liability, negligence or otherwise, or from any use or
operation of any methods, products, instructions, or ideas contained in the material

isbn: 978-1-84465-701-8 (hardcover)

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Typeset in Adobe Garamond Pro by JS Typesetting Ltd, Porthcawl, Mid Glamorgan.


Contributors ix

Introduction 1
Armin W. Geertz


1. Whence religion? How the brain constructs the world and
what this might tell us about the origins of religion,
cognition and culture 17
Armin W. Geertz
2. Why costly signalling models of religion require
cognitive psychology 71
Joseph Bulbulia
3. The prestige of the gods: evolutionary continuities in the
formation of sacred objects 82
William E. Paden
4. The evolutionary dynamics of religious systems: laying the
foundations of a network model 98
Istvn Czachesz
5. Art as a human universal: an adaptationist view 121
Ellen Dissanayake
6. The significance of the natural experience of a non-natural
world to the question of the origin of religion 140
Donald Wiebe


7. Religion and the emergence of human imagination 160

Andreas Lieberoth
8. The origins of religion, cognition and culture: the
bowerbird syndrome 178
Luther H. Martin
9. The will to sacrifice: sharing and sociality in humans, apes
and monkeys 203
Henrik Hgh-Olesen
10. Apetales: exploring the deep roots of religious cognition 219
Tom Sjblom


11. Cognition and meaning 241
Jeppe Sinding Jensen
12. Wittgenstein and the naturalness of religious belief 258
Mark Addis
13. Peekaboo! and object permanence: on the play of
concealment and appearance in cognition and religion 269
Thomas Hoffmann
14. Yogcra Buddhist views on the causal relation between
language, cognition and the evolution of worlds 285
William S. Waldron
15. A resource model of religious cognition: motivation as a
primary determinant for the complexity of supernatural
agency representations 301
Uffe Schjoedt
16. The recognition of religion: archaeological diagnosis and
implicit theorizing 310
Peter Jackson
17. Religion and the extra-somatics of conceptual thought 319
Mads D. Jessen
18. Tools for thought: the ritual use of ordinary tools 341
Pierre Linard and Jesper Srensen
19. Care of the soul: empathy in a dualistic worldview 365
Gretchen Koch


20. From corpse to concept: a cognitive theory on the

ritualized treatment of dead bodies 374
William W. McCorkle Jr
21. Anthropomorphism in god concepts: the role of narrative 396
Peter Westh

Index 415

This page intentionally left blank

Mark Addis is Professor of Philosophy at Birmingham City University,

Visiting Professor at the Department of Culture and Society at Aarhus Uni-
versity and a Research Associate at the Centre for Philosophy of Natural
and Social Science at the London School of Economics. His publications on
Wittgenstein include Wittgenstein: A Guide for the Perplexed (2006), Wittgen-
stein: Making Sense of Other Minds (1999) and he co-edited Wittgenstein and
Philosophy of Religion (2001).

Joseph Bulbulia is a Senior Lecturer at Victoria University of Wellington,

New Zealand and publishes widely in psychology, philosophy and evolu-
tionary religious studies. His work on the evolution of religion focuses on
costly signalling models and large-scale coordination problems. Among his
many publications are First Shots Fired for the Phylogenetic Revolution
in Religious Studies (co-authored, 2013) and Why Do Religious Cultures
Evolve Slowly? (2013), and he is co-editor of The Evolution of Religion:
Studies, Theories, & Critiques (2008).

Istvn Czachesz is Heisenberg Fellow and Privatdozent of New Testament

at the University of Heidelberg. He is author of Commission Narratives: A
Comparative Study of the Canonical and Apocryphal Acts of the Apostles (2007)
and The Grotesque Body in Early Christian Discourse: Hell, Scatology, and
Metamorphosis (2012), and the books he has co-edited include Mind, Morality
and Magic: Cognitive Science Approaches in Biblical Studies (2013).

Ellen Dissanayake is an Affiliate Professor in the School of Music at the

University of Washington and an independent scholar, author and lecturer
in many disciplines, including evolutionary biology, ethology, cognitive
and developmental psychology, cultural and physical anthropology, neuro-
science, and the history, theory and practice of the various arts. She is the


author of What Is Art For? (1988), Homo Aestheticus (1992) and Art and
Intimacy (2000), as well as over seventy scholarly and popular articles and
book chapters.

Armin W. Geertz is Professor in the History of Religions at the Department

of Culture and Society, Section for the Study of Religion, and Chair of the
Religion, Cognition and Culture Research Unit (RCC), Aarhus University,
Denmark. His publications in the cognitive science of religion range from
religious narrative and evolutionary theory to the neurobiology of religion.
His recent publications include Religious Narrative, Cognition and Culture (co-
edited, 2011). He is co-editor of Acumens Religion, Cognition and Culture
series, and senior co-editor of Journal for the Cognitive Science of Religion.

Thomas Hoffmann is Professor with special responsibilities at the Section

for Biblical Exegesis at the Faculty of Theology, Copenhagen University,
Denmark. He specializes in the study of the Quran and has published articles
on cognitive poetics-approaches to the Quran and is currently working on a
monograph on the cognitive Quran. His books include The Poetic Qurn:
Studies on Qurnic Poeticity (2008).

Henrik Hgh-Olesen is Professor in Social and Personality Psychology and

Head of the Department of Psychology and Behavioral Sciences, Aarhus Uni-
versity, Denmark. His research interests include evolutionary and comparative
perspectives on human mind and kind, human characteristics and the human
condition. His recent edited books are Human Characteristics: Evolutionary
Perspectives on Human Mind and Kind (co-edited, 2009) and Human Morality
and Sociality: Evolutionary and Comparative Perspectives (2010).

Peter Jackson is Professor of the History of Religions at Stockholm University,

Sweden. He specializes in the study of Indo-European religions, particu-
larly ancient Indian and Iranian religions, the religions of ancient Greece
and Rome, and Old Norse religion and works on more general theoretical
and conceptual concerns in the study of religion. Among his many publica-
tions are The Extended Voice: Instances of Myth in the Indo-European Corpus
(1999) and Verbis pingendis: Contributions to the Study of Ritual Speech and
Mythopoeia (2002).

Jeppe Sinding Jensen is Associate Professor at the Department of Culture and

Society, Section for the Study of Religion, and Coordinator of the Religion,
Cognition and Culture Research Unit (RCC), Aarhus University, Denmark.
His research interests include semantics and cognition in religious narrativity,
myth and cosmology, and method, theory and the philosophy of science in


the study of religion. He is the author of The Study of Religion in a New Key:
Theoretical and Philosophical Soundings in the Comparative and General Study
of Religion (2003), and he is editor-in-chief of Acumens Religion, Cognition
and Culture series.

Mads D. Jessen is a Project Researcher at the National Museum of Den-

mark. He is currently excavating the Late Viking Age royal monument at
Jelling, Jutland and studying pre- and proto-Christian rituals in south Scan-
dinavia. His publications include Material Culture, Embodiment and the
Construction of Religious Knowledge (2012). He is the co-editor of the new
journal Danish Journal of Archaeology.

Gretchen Koch received her PhD in the cognitive science of religion from
Aarhus University, Denmark, in 2009. She has published journal articles and
participates in interfaith dialogue on websites such as Religion Dispatches
and the Journal of Inter-Religious Dialogues State of Formation, and promotes
cognitive approaches to religion and morality to secular audiences on her blog
Cheap Signals.

Andreas Lieberoth is a PhD Fellow at the Department of Psychology and

Behavioral Science and the Centre for Functionally Integrative Neuroscience
(CFIN), Aarhus University. He studies phenomena connected with playful-
ness, imagination and the application of game design to serious contexts. His
publications include Similarity of Social Information Processes in Games
and Rituals: Magical Interface (co-authored, 2012).

Pierre Linard is an Assistant Professor at the University of Nevada, Las

Vegas where he teaches social and cultural anthropology as well as culture
and cognition. He has conducted ethnographic fieldwork in Eastern Africa.
His publications include Whence Collective Rituals? A Cultural Selection
Model of Ritualized Behavior (co-authored, 2006), Life Stages and Risk-
Avoidance: Status- and Context-Sensitivity in Precaution Systems (2011)
and Beyond Kin: Cooperation in a Tribal Society (2014).

Luther H. Martin is Professor Emeritus of Religion, University of Vermont.

He is the author of Hellenistic Religions (1987) and of numerous articles in
this field of his historical specialization and has co-edited several volumes
in cognitive theory and historiographical method, including Past Minds:
Studies in Cognitive Historiography (co-edited, 2011). He is a member of the
Honorary Board of Method & Theory in the Study of Religion, senior co-editor
of Journal of the Cognitive Science of Religion and a founding editor of the
Journal of Cognitive Historiography.


William W. McCorkle Jr served from 2011 to 2013 as Director of Experi-

mental Research at LEVYNA (Laboratory for the Experimental Research of
Religion) and Associate Professor and Research Specialist in the Department
for the Study of Religions at Masaryk University, Czech Republic. His publi-
cations include Ritualizing the Disposal of the Deceased: From Corpse to Concept
(2010), and he co-edited Mental Culture: Classical Social Theory and the
Cognitive Science of Religion (2013). He is the managing editor for the Journal
of Cognitive Historiography.

William E. Paden is Professor Emeritus of Religion at the University of Ver-

mont in Burlington, Vermont, where he taught for forty-five years and served
as department chair for two decades. He is the author of Religious Worlds:
The Comparative Study of Religion (1988, 1994) and Interpreting the Sacred:
Ways of Viewing Religion (1992, 2003), each appearing in numerous foreign
translations and editions.

Uffe Schjoedt is a Postdoc Researcher at the Department of Culture and

Society, in the Religion, Cognition and Culture Research Unit (RCC), Aarhus
University, Denmark. His research fields include cognitive science of religion,
psychology of religion and experimental cognitive neuroscience. His publica-
tions include Cognitive Resource Depletion in Religious Interactions (co-
authored, 2013).

Tom Sjblom is Docent in the History of Religions at the Department of the

Study of Religions, University of Helsinki, Finland. He is the author of the
book Early Irish Taboos: A Study in Cognitive History (2000). His interest areas
in cognitive science include narrative cognition, the deep roots of religious
cognition, emotional communication, and cognitive ethology.

Jesper Srensen is MINDLab Associate Professor at the Department of

Culture and Society in the Religion, Cognition and Culture Research Unit
(RCC), Aarhus University. He has published numerous articles on the cogni-
tive science of religion, in particular on magic, ritual and conceptual trans-
mission, as well as more general papers pertaining to theoretical issues in the
scientific study of religion. His publications include Magic Reconsidered:
Towards a Scientifically Valid Concept of Magic (2012), Cognitive Resource
Depletion in Religious Interactions (co-authored, 2013) and A Cognitive
Theory of Magic (2007).

William S. Waldron is Professor and Chair of the Department of Religion

at Middlebury College, where he teaches courses on South Asian Buddhism,
Hinduism, and the Study of Religion. He has written numerous publications


on the interface between Buddhist philosophy and modern theories of mind

from evolutionary biology, cognitive science and psychology, including
Buddhist Steps to an Ecology of Mind (2002) and The Buddhist Unconscious

Peter Westh is part-time Lecturer in the History of Religions at the Department

of Cross-Cultural and Regional Studies, University of Copenhagen, and
teaches religion, culture and social science at the Efterslgten upper-secondary
school in Copenhagen, Denmark. Recent publications include Illuminator
of the Wide Earth; Unbribable Judge; Strong Weapon of the Gods: Intuitive
Ontology and Divine Epithets in Assyro-Babylonian Religious Texts (2011)
and The Fire and the Sun: God Concepts in an Assyrian Incantation in a
Cognitive Light (2011).

Donald Wiebe is Professor of Philosophy at Trinity College, University of

Toronto. His primary areas of research interest are the history of the academic
study of religion, the philosophy of science, and method and theory in the
study of religion. He is the author of Religion and Truth (1981), The Irony
of Theology and the Nature of Religious Thought (1991), Beyond Legitimation
(1994), The Politics of Religious Studies (1999), and over two hundred aca-
demic articles, essays and reviews.

This page intentionally left blank
Armin W. Geertz

It has often been said that the origins of religion, cognition and culture are
beyond the ken of modern man. And yet, it is the most interesting, challeng-
ing and provocative of topics in the natural and human sciences. Perhaps
our origins are no longer as murky today as in previous centuries. The past
twenty-five years have startled the world with major advances in a number
of disciplines and sciences: evolutionary psychology, cognitive archaeology,
neuroscience, ethology, developmental psychology, social psychology, cogni-
tive linguistics, palaeoanthropology and genetics. These advances have signifi-
cantly impacted on comparative religion and have led to the establishment of
a burgeoning new field called the cognitive science of religion. This new field
has succeeded in casting new light on age-old problems.
Efforts to discover and explain the evolutionary origins of Homo sapiens
sapiens have led to a wide variety of hypotheses attempting to discover what is
particularly human about human beings. We have witnessed such memorable
ideas as the grandmother hypothesis, the aquatic ape, man the tool-maker,
man the hunter, woman the gatherer, crossing symbolic thresholds, the speak-
ing ape, social intelligence, the great hominid escape, mankinds epistemic
hunger, the hybrid mind, and so on. All of these attempts to understand the
origins of humanity have raised fundamental questions about the complex
relationships between cognition and culture. Are they two sides of the same
coin? Or is culture epiphenomenal to other more basic processes? And how
does religion fit into the picture?
Central to the debates on origins is the role of religion, religious ritual
and religious experience. What came first: individual religious (ecstatic) expe-
riences, collective observances of transition situations, fear of death, ritual
competence, magical coercion, mirror neurons or temporal lobe religiosity?
Together with the development of symbolic thinking, the roles of material
culture, written language and abstract thought in the development of reli-
gious systems are all central to the humanities and social sciences. Cognitive

armin w. geertz

scientists are now providing us with important insights on phylogenetic and

ontogenetic processes. Together with insights from the humanities and social
sciences on the origins, development and maintenance of complex semiotic,
social and cultural systems, a general picture of what is particularly human
about humans could emerge. Reflections on the preconditions for symbolic
and linguistic competence and practice are now within our grasp.
There are a growing number of anthologies and books on the origins of
human culture and religion. Most are written by scholars not familiar with
the comparative science of religion or its various disciplines (history of reli-
gions, sociology of religion, psychology of religion, etc.). The results are often
evidence of this. The two most talked about volumes are by two famous
atheists, Richard Dawkins (2006) and Daniel Dennett (2006). Some of their
insights are commendable. Their knowledge of religion and the academic
study of religion, however, is not (cf. Geertz 2009). Several anthologies are
more insightful (Bulbulia et al. 2008; Feierman 2009; Schloss & Murray
2009). This anthology is unique in that, with a few exceptions, the authors
are mainly professional scholars of religion.

Evolutionary theories in the cognitive science of religion

I have elsewhere addressed the major ideas in the cognitive science of religion
(Geertz 2004, 2008b, 2010b).1 As it turns out, virtually all of the theories in
the cognitive science of religion adhere to what Pascal Boyer has called the
naturalness of religion hypothesis (Boyer 1994). This hypothesis does not
mean that religion is inborn or hardwired, rather that religious behaviour
and ideas in all their startling cacophony thrive, survive and are passed down
through the millennia because they fit smoothly and naturally to intuitive,
hardwired psychological mechanisms and processes. They are not directly
selected for. They simply survive because they fit with evolutionarily selected
features. This is called the by-product hypothesis. Religion and culture are
sometimes described as cognitive spandrels: they fit nicely in-between arches,
but have no particular purpose. Others use the more onerous metaphor of
religion as a virus spreading through brain populations also called the epi-
demiology of representations by Dan Sperber (1996) or a parasite thriv-
ing on brain functions not originally designed for religious thoughts (which
Richard Dawkins called memes, i.e. information replicators that act like
genes; Dawkins 1976).
This hypothesis assumes certain theoretical commitments that have been
nicely systematized by Justin Barrett (2007: 59). First, all humans have the
same cognitive and psychological make-up no matter what culture they
come from. Second, the brain is a highly complex organ consisting of highly


specialized functional subsystems. Third, these subsystems shape and filter

in-coming perceptions of internal and external worlds. Fourth, they con-
strain and inform human thought and action, including religious thought
and action. And, fifth, recurrent features of religious thought and action can
be explained and predicted with reference to the basic dynamics of the mind.
Pascal Boyer was the first to identify a series of cognitive systems that are
essential to religious thought and behaviour. He has described them in crea-
tive and entertaining publications dating from the 1990s and most recently
in his book Religion Explained (Boyer 2001). What is it, he asks, that is so
remarkable about religious ideas that people remember them and are moti-
vated to pass them on to younger generations? This is where his minimally
counterintuitive ideas hypothesis comes in. Religious ideas fit well with our
cognitive machinery. They draw on ordinary, cognitive dispositions and add
one or two details that are counterintuitive to basic intuitive ideas about the
physical world, plants, animals, humans and natural and constructed objects.
Boyer adds, furthermore, the important point that not only do religious ideas
contain one or two details that are cognitively salient, they are also and more
fundamentally based on normal ideas and thus easy to remember and pass on.
One other set of ideas deserves mention, namely basic human agency
detection, which Stewart Guthrie drew attention to as early as 1980.2 Guthrie
claimed that this cognitive ability explains our tendency to see animate and
human-like qualities in everything around us. Thus animism and anthro-
pomorphism are the origins of religion. Justin Barrett has expanded on this
theme by exploring experimentally why we believe those counterintuitive
ideas described by Boyer. He claimed that the mental tool discovered by
Guthrie is characterized by being hyperactive or hypersensitive, which Barrett
called hypersensitive agency detection device (HADD). Together with the
characteristically human ability called theory of mind (i.e. the realization
that other people have minds and feelings just like us an ability that every
normal child above the age of four achieves), HADD naturally leads to god-
beliefs (Barrett 2004).
Evolutionary theories are not just concerned with ideas. The pioneering
team of E. Thomas Lawson and Robert N. McCauley introduced a highly
sophisticated account of how believers mentally represent ritual action
(Lawson & McCauley 1990). They are not, however, talking about believers
doing their rituals. Inspired by Noam Chomsky, Lawson and McCauley are
interested in identifying whether or not believers have a kind of intuitive rit-
ual grammar. How do they know that a ritual has been correctly carried out?
In other words, how do they judge the rightness of a ritual? Like everyone
else, believers have an inborn capacity to represent ordinary human action. By
adding a religious filter through years of socialization, this action representa-
tion system is transformed, and participants gain what Lawson and McCauley

armin w. geertz

call ritual competence. In an interesting debate with Harvey Whitehouse

(2004), Lawson and McCauley explored other dimensions of ritual behaviour
in which sensory pageantry plays an important role together with frequency
and ritual form (McCauley & Lawson 2002).
The other main ritual theory that has been circulating in the cognitive sci-
ence of religion is Harvey Whitehouses religious modes hypothesis (1995).
Whitehouse claimed that the frequency and types of ritual behaviour that
humans engage in stimulate evolutionarily selected memory systems, i.e. the
episodic and semantic memory systems. Rituals seldom performed but highly
arousing (violent initiation rituals for example) stimulate episodic memory
in much the same way that flashbulb memory is stimulated during traumatic
events. These rituals are called imagistic. Rituals that are performed often but
are much less arousing (like Christian Sunday services) stimulate semantic
memory. These rituals are called doctrinal. Both types of rituals and the way
they stimulate our memory systems have causal influence on the way that
people interpret their religious experiences and the kinds of social organiza-
tions that are built up around them.

The adaptationist approach

It should be obvious by now that the main assumption among theorists in

the standard cognitive science of religion is that causal explanations of human
behaviour and ideas are to be found in the mind. Thats why it can be charac-
terized as a mentalistic approach. Furthermore, the main emphasis is on indi-
vidual minds, and so it can also be characterized as an individualistic approach.
Other approaches, however, take foundational issue on two points. First,
anything that is as costly and dangerous as religious behaviour must have had
adaptive features that were selected for during our evolution. And, second,
the dynamics between culture and cognition are much more fundamental
and causally interrelated than the standard cognitive science of religion would
Jesse M. Bering is half-way between the by-product approach and the
adaptationist approach. He argued that religious belief is an exaptation, a
spandrel that turned out to be useful and so was subsequently selected for by
evolutionary pressures (Bering 2006a: 125). Bering, an experimental psy-
chologist, has conducted interesting experiments with children and adults
and has developed a novel theory of existential psychology (Bering 2006b).
Out of the conglomerate of human intuitions on causal attribution, moral
judgement, theory of mind, concept acquisition and teleological reason-
ing, Bering posited what he calls an organized cognitive system dedicated
to forming illusory representations of (1) psychological immortality, (2) the


intelligent design of the self and (3) the symbolic meaning of natural events
(ibid.: 453). This system evolved in response to the selective pressures of
human social environments.
As for the first issue, adaptationists relate their work to major advances
in evolutionary biology and evolutionary psychology. Borrowing from evo-
lutionary game theory and the biology of animal signalling, Joseph Bulbulia
and Richard Sosis have developed a costly signalling hypothesis of religion. In
order for human societies to work, you need to depend on reciprocal altruism.
But you also need to ensure that your partners are not cheating on you. One
of the solutions to the problem is to develop a public and costly institution,
like religion, where hard-to-fake god-fearing commitments ensure that peo-
ple can identify who is trustworthy and who is not. Joseph Bulbulia in fact
defines a religious deed as a costly signal capable of authenticating religious
commitment (Bulbulia 2004b: 669). Richard Sosis has tested this hypothesis
in a number of empirical studies (Sosis 2000; Sosis & Bressler 2003).
On the second issue, biologist and semiotician Terrence W. Deacon made
a strong case for epigenetics as a significant factor in evolution, even for non-
human animals (Deacon 1997). Animals carve out niches in the environment
which then put selective pressures on biological evolution. Deacon argued
that we are a self-domesticated species that through social evolution, the
transmission of symbolic communication and the elaboration of stone tools
created a radically different niche than that experienced by our non-symbolic
ancestors, the australopithecines and other apes (Deacon 2003: 93). We
are, Deacon argued, a symbolic species in a deeply physiological sense,
and until we understand the complex, emergent dynamics between genetic,
cultural and semiotic selection processes, we will remain in a prescientific
phase of cognitive and social science (ibid.: 95, 104).
Developmental psychologist Merlin Donald argued that humans have
evolved to live in cultural networks. Humans are incapable of using their
large brains without a supportive cultural trellis. The heuristic exercise of iso-
lating human cognition from culture, often practised by standard cognitive
scientists of religion, is meaningless from this point of view, since humans
have never been without culture, and their brains and cognition have been
radically redesigned by culture. Our evolutionary history, Donald claimed,
has released us from our animal solipsism, and we are not only able to hold
internal and external models of the self-in-the-world, but also to draw on
social connectivity and conventionality in order to do so (Donald 2001: 263).
Following this line of thought, Jeppe Sinding Jensen and I have argued
that religious models of the world are essential elements of cognitive mod-
els of the self. Much of the work being done by us and our colleagues is
concerned with demonstrating the top-down, bottom-up processes of cogni-
tion and the ways in which these processes interrelate with religious worlds.3

Table 1 Theories and approaches to the origins of religion, cognition and culture.

Name Discipline Topic Hypothesis (mechanism) Empirical evidence

Scott Atran (2002) anthropology evolution of religion cognitive and emotional psychology, ethnography,
systems archaeology
Justin Barrett (2004) psychology origins of religious ideas HADD and theory of mind psychology
Pascal Boyer (1994) anthropology transmission and survival of intuitive systems psychology, ethnography
religious representations
Boyer & Pierre Linard anthropology selection of ritual behaviour similar to obsessive psychiatry, ethnography
(2006) compulsive disorder and
hazard-precaution system
Joseph Bulbulia (2004a) study of religion evolution of religion costly signalling psychology, ethnography,
Richard Dawkins (1976, biology evolution of religion memes biology, evolutionary
2006) psychology
Terrence Deacon (1997) biosemiotics evolution of culture symbolic competence primatology, biology,
Merlin Donald (1991, 2001) psychology biocultural evolution hybrid consciousness psychology, neuroscience,
Armin W. Geertz (2008a) study of religion evolution of religion ape brains with symbolic archaeology, primatology,
competence neuroscience
Name Discipline Topic Hypothesis (mechanism) Empirical evidence
Stewart Guthrie (1993) anthropology evolution of religion intuitive animism and anthropology, psychology
E. T. Lawson & R. N. study of religion ritual representations ritual competence and ritual comparative religion,
McCauley (Lawson & form psychology
McCauley 1990, McCauley
& Lawson 2002)
David Lewis-Williams rock art research evolution of religion altered states of cave paintings
(2002) consciousness and
Steven Mithen (1996) archaeology material symbols modularity and fluidity archaeology, psychology
Jesper Srensen (2004) study of religion religious representations immunology and conceptual history of religion,
constraints evolutionary psychology
Richard Sosis (Sosis & anthropology evolution of religion costly signalling psychology, ethnography
Alcorta 2003)
Dan Sperber (1996) anthropology cultural representations epidemiology ethnography, evolutionary
Harvey Whitehouse (1995) anthropology transmission of religious ideas religious modes and psychology, ethnography
and rituals memory types
David Sloan Wilson (2002) biology evolution of religion group selection biology, ethnography,
comparative religion
armin w. geertz

Jensen has encapsulated this dynamic with the term normative cognition,
at which religions excel (Jensen 2010).
Table 1 gives an indication of the variety and theories, approaches and
assumptions of cognitive theorists concerning the origins of religion, cogni-
tion and culture. It is not my intention to describe them. I have mentioned
the few that have had an impact on current cognitive science of religion and
who will often be referred to in the chapters of this book. The rest of this
introduction will indicate how radically different contemporary approaches
are in comparison with the standard cognitive science of religion. And, in the
process, it is my wish and hope that the cognitive science of religion will be
transformed into a science of cognition and culture.

Issues and themes

Presented by an exciting group of internationally known scholars as well as

innovative younger scholars, these chapters explore the origins of religion,
cognition and culture and, in the process, put culture center-stage in the
cognitive science of religion. Each in its own way, every chapter presents
challenges to late-twentieth-century evolutionary psychology and cognitive
science of religion.
The chapters are organized in two parts. Part I consists of evolutionary
scenarios addressing the origins of religious behaviour and thought. Part
II consists of chapters that critically discuss some of the early theories and
hypotheses developed by the pioneers of the cognitive science of religion.

Evolutionary scenarios

Part I is a good indicator of the exciting creativity evident in new theories

on the origins of religion, cognition and culture. Chapters 14 explore pos-
sible evolutionary frameworks. My own Chapter 1 attempts to sketch out a
theory based on a biocultural, adaptationist approach. I argue that we need
to rethink religious thought and behaviour in terms of the kinds of brains
we have, based on current neurobiology. We are intelligent apes that are
highly emotional, easily spooked, very superstitious, extremely sensitive to
social norms and virtual realities, and, furthermore, we are equipped with
nervous systems that are vulnerable to influence from conspecifics and their
symbolic worlds. Our brains are constantly predicting and dwelling on the
future and our brains fill in quite a bit of what is missing, heavily supple-
mented (and even dictated) by the cultures we are socialized into. In looking
through the archaeological records, I argue that the kinds of minds that were


proto-religious most likely appeared long before Homo sapiens arose. The evi-
dence seems to indicate that proto-religious behaviour was evinced by Homo
In Chapter 2, Joseph Bulbulia argues that religiosity is an evolutionary
problem because such behaviour is often very costly in terms of reproduc-
tion and resources. In drawing on commitment-signalling theory with the
addition of cognitive psychology Bulbulia seeks to explain why religious
beliefs are emotionally expressed and why the very assumption of supernatu-
ral agents increases cooperation. He concludes that a cognitive science of
religious emotions can explain the cost problem that mind-blind behavioural
signalling theories require. Religious behaviour, he claims, enhanced group
survival, with subsequent reinforcement through selection because religious
people express their commitments in hard-to-fake ways. Based on a scope-
syntactic model, religious experience can be viewed as a kind of adaptive
William Paden (Chapter 3) approaches religious complexes as systemic
forms of enculturated prestige, and thus, in terms of human social compe-
tence, religious behaviour and the use of religious objects are integral to social
display. In this perspective, Paden urges us to change focus from religious
conceptions to religious behaviours (i.e. from conceptual cognition to social
cognition). His emphasis on ritual objects, furthermore, brings culture back
into centerfield.
In Chapter 4, Istvn Czachesz provides a framework for a network model
of the evolution of religious systems. Based on a formal understanding of
systems drawn from graph theory, Czachesz argues that a religious system
involves the bidirectional interaction between beliefs and artefacts, and
a dynamic network model would include the temporal dimension of this
interaction. Various systems, whether cultural, biological or ecological, have
evolved not by design but by tinkering (i.e. reusing existing bits and pieces
and adding new ones). Cross-cultural factors will drive the evolution of beliefs
and artefacts towards attractor positions that change with time in changing
environmental contexts.
Chapters 57 examine the roles of aesthetics and imagination in the evo-
lution of religion, cognition and culture. In Chapter 5, art historian Ellen
Dissanayake argues that art is an inherent psychobiological capacity of the
human species and, as such, had adaptive advantages in human evolution.
From an ethological perspective, art is a behavioural process that Dissanayake
calls artification, in other words, to make something into art or into some-
thing special. This process is dependent upon our aesthetic predisposition,
evinced in early mother-child engagement and assisted by emotional invest-
ment in events and states that we care about. Dissanayake argues that our
ancestors discovered the advantages of ritualized behaviours in uncertain and

armin w. geertz

stressful circumstances and in the resultant social reinforcement. This discov-

ery led to ceremonial rituals and to what we today understand as religion
and art.
Donald Wiebe (Chapter 6) is not convinced that current theories on our
cognitive capacities for religious conceptions have provided any convincing
explanation of the move from natural to supernatural thinking. Taking his
point of departure in Lewis-Williamss account of the origin of religion in the
Palaeolithic, Wiebe argues that the paintings are not just artistic depictions of
the natural world, rather they reflect the meanings, ideas and norms of their
human world, often based on visionary and trance experiences. The latter
must have been an important source of religious reasoning that was encour-
aged and developed in social contexts.
Andreas Lieberoth (Chapter 7) asserts that even though the cognitive
science of culture is based on the human brain, culture is fundamentally a
socially negotiated phenomenon. Religious ideas, however, must have sur-
faced at some point in our evolutionary history. How did such ideas arise,
assuming that they more or less arose in an environment without religious
ideas? Lieberoth claims that the answer may lie in relating religion to imagina-
tion, creativity, decoupling, pretend play and madness. A key process is imagi-
nation in that it allows us to conceptualize and analyse any acquired concept
whether supernatural or not. Off-line thinking, memory and external rep-
resentation, he argues, seem to be the cognitive basis for religious thought.
This occurred, Lieberoth believes, with the appearance of cave paintings some
30,000 to 40,000 years ago.
Chapters 810 compare human evolution with that of other vertebrates.
In Chapter 8, Luther H. Martin sidesteps the evolution of culture and cogni-
tion by investigating avian cognition and culture. Looking at domestic chicks,
passerines and bowerbirds, Martin argues that human culture is a social
exploitation of those phylogenetic dispositions, and, thus, that we would do
well to look much deeper than usual into our evolutionary vertebrate past to
probe the origins of cognition and culture. As for religion, Martin argues that
it might best be understood in terms of sexual selection. Psychologist Henrik
Hgh-Olesen (Chapter 9) also delves into characteristics that we share with
other animals, especially apes and monkeys. He asks whether our sociality
in terms of the pleasure of reciprocal cooperation and the anger at cheaters
and free-riders is similar. One difference, however, is that we seem to dislike
inequitable exchanges that benefit ourselves only. Furthermore, we evince the
willingness to sacrifice and actively share for the greater good. Thus, it is no
coincidence that sacrifice is universally central to religious practices.
Tom Sjblom (Chapter 10) argues that the origins of religion are insepa-
rable from the origins of the human capacity for culture. In focusing on sym-
bolic behaviour and the various theories and hypotheses about when it arose,


Sjblom claims that the origins of symbolic thought must be found long
before the so-called cultural explosion or revolution of the European cave art
period some 40,000 years ago. He argues that we must necessarily understand
human evolution in terms of a long, cumulative process. Following Jared
Diamond, Sjblom posits the third chimpanzee approach, and he uses it to
explicate the pervasiveness of snake beliefs. Indeed, he claims that the origins
of religion are to be found in the origins of emotional communication and
thus in our ape ancestry.

Cognitive theories

Part II takes a critical look at earlier theories of the origins of religion, cogni-
tion and culture. Chapters 1114 address philosophical issues. Jeppe Sinding
Jensen (Chapter 11) addresses an issue neglected in the cognitive science of
religion, namely, the question of meaning. This neglect, he argues, is based
on misunderstandings of profound importance to the future of the academic
study of religion. Jensen confronts the cultural reductionism and cultural
eliminativism that is widespread in standard cognitive science of religion
research and argues that if the methods they choose determine ontology, epis-
temology and the results of the research, then methodology has become
dogma. And so, the chosen discourse turns to immunization strategies which
conveniently avoid the possibility of falsification. As long as cognitivists deny
explanatory validity to culture, they will have little impact on the human
In Chapter 12, Mark Addis supports the naturalness hypothesis of the cog-
nitive science of religion as opposed to widespread assumptions in the study
of religion that the study of religion requires a special methodology because
religion is claimed to be special. Addis explores the similarities between
Wittgenstein and the cognitive science of religion and notes a clear resonance,
among other things, between Wittgensteins idea about certain things being
natural, given a particular biological form of life, and the widespread idea in
the cognitive science of religion that there are certain evolutionarily selected
cognitive capacities shared by all normal people that underpin religion.
Although Thomas Hoffmann (Chapter 13) believes that cognitive
approaches to the study of religion are the most promising in current research,
he argues that many cognitivists have forgotten their precursors. A number of
examples from early European and American philosophers and psychologists
show how fundamental their ideas have been for current cognitivist thinking.
By taking up Piagets concept of object permanence (together with Lakoff
& Johnsons containment schema), Hoffmann argues that the apparent
transition in the first year of life from being unable to being able to maintain

armin w. geertz

object permanence could support the hypothesis that our species underwent
this cognitive development, and that it might have had an important impact
on the origin of religious concepts and practices.
In Chapter 14, William S. Waldron analyses a model of the co-evolution of
culture and consciousness, centred on an explicit notion of unconscious mind
in Yogcra philosophy and shows how their proposal is relevant to discus-
sions of the origins of culture. Taking issue with ideas of causality in current
cognitive science of religion, Waldron argues that cognitive structures can
never cause cognition, culture or religion because they unavoidably depend
upon a variety of other, often implicit, causes in the first place. Causality in
living processes is multi-faceted and multi-dimensional. Thus, he argues, it
is neither genes nor beings that evolve in splendid isolation, but the pattern
of organisms-interacting-with-environment that together evolves.
But what do neurobiologists have to say about religion, cognition and
origins? Uffe Schjoedt (Chapter 15) introduces a theory quite unlike current
evolutionary psychological approaches to human cognition. His homeostatic
theory of religious behaviour posits human cognition in terms of embodied
motivation and intentionality. Drawing on current neurobiology, Schjoedt
argues that body and brain states modulate religious behaviour in an equally
dramatic way. The difference between this theory and current evolutionary
psychological theories is that it assumes only one coherent system (the organ-
ism) rather than a series of multiple, independent cognitive systems. Schjoedt
claims that such a simple and biologically rooted theory of homeostasis makes
hardcore cognitive typologies a less attractive alternative drawing on compara-
bly far more speculative and problematic hypotheses. He then illustrates how
this plays into the theories of E. Thomas Lawson and Robert N. McCauley.
Chapters 1618 address tools and material culture in the evolution of
human cognition and religion. In Chapter 16, with concern for general issues
in comparative religion, Peter Jackson discusses implicit notions about reli-
gion in prehistoric archaeology that may prove helpful to the study of religion.
He questions the assumptions that the Lwenmensch figurines found in caves
at two Swabian sites necessarily reflect the same traditions and, especially, are
evidence of shamanism. These figurines do, however, indicate capacities that
seem to be of a religious nature. Turning to religious action, Jackson argues
that the higher category behind such behaviour can be tentatively defined as
the mutual reiteration and counteraction of expediency.
Archaeologist Mads Jessen addresses the significance of material culture
on human cognition in Chapter 17. Not entirely satisfied with the somewhat
narrow cognitive science of religion field of enquiry, especially the neglect
of the role and importance of extra-somatic factors in human thought,
Jessen emphasizes the human technological entanglement with the world.
Jessen argues that the origins of religion must be found in its environmental


grounding and not simply in innate cognitive constraints. More specifically,

he argues that the transmission of certain (counterintuitive) concepts as
opposed to other concepts is generated in particular environments of natural
as well as cultural materialities.
Pierre Linard and Jesper Srensen (Chapter 18) ask why people resort to
ritual behaviour, more specifically, to the extraordinary uses that functional
tools are put to in ritual contexts. They posit that there may be universal
cognitive mechanisms responsible for the ritualistic manipulation of ordi-
nary tools and ritual tools. Linard and Srensen argue that the ritualistic
use of tools increases their cultural success, relies on the violation of basic
intuitions about tools and their functions, and the evocation of counter-
intuitive representations of tools is fast and easy, given the nature of human
tool knowledge.
Chapters 19 and 20 deal with moral psychology as they relate to religious
behaviour. In Chapter 19, Gretchen Koch claims that even if empathy is
adaptive, it may be that we are not designed to always empathize accurately
or automatically. It may be that we are designed to empathize (project) inac-
curately under certain circumstances, in order to pursue overriding interests.
What may be at play, Koch argues, is the cognitive capacity of self-awareness.
To empathize with another person requires a strong sense of self and of other.
Such representational abilities can break down in some diseases, but they can
also be deliberately inhibited. There are a number of top-down techniques
that can be used to avoid empathizing with others. She demonstrates how our
intuitive dualism leads to an understanding of humans as rational, responsible
agents, thus providing reasons for the conscious and even unconscious appli-
cation or inhibition of empathy.
William McCorkle (Chapter 20) takes up what is probably the oldest ritual
behaviour in our species, namely, actions and behaviours dealing with death
and the dead. McCorkle focuses on the cognitive processes that may be trig-
gered by dead bodies. Drawing on Pascal Boyers work, McCorkle posits that
five mental systems are at play: hyperactive agency detection device (HADD),
theory of mind (ToM), animacy, person-file and contagion/disgust systems.
Based upon experimental evidence, however, McCorkle concludes that indi-
viduals do not react the same way to dead bodies and this difference seems
to be a measurable index based upon certain personality traits that are innate,
developmental, and culturally constructed. Such individual cognitive proc-
esses might be involved in biological deviation in much the same way as
immune systems develop, thus evoking an evolutionary model that employs
difference in maintaining the survival of the species, or conspecific fitness.
The final chapter provides a detailed critique of experimental evidence
concerning HADD. Peter Westh (Chapter 21) investigates Justin Barretts
and Pascal Boyers critical reformulations of Stewart Guthries theories of

armin w. geertz

anthropomorphism and animism and how these reformulations are crucial

to their evolutionary assumptions. Westh demonstrates that they and oth-
ers have misunderstood Guthrie and that the cognitive mechanism, dubbed
HADD by Barrett, may make evolutionary common-sense, but remains to
be demonstrated in a controlled setting. Westh shows that the story compre-
hension studies by Barrett and Keil do not support widespread claims that
cognitive pressure explains the anthropomorphism evident in the compre-
hension task. On the contrary, he shows how the anthropomorphic bias may
well have been prompted by the task itself. He concludes that the processes
of narrative comprehension would be far more recent in evolutionary terms.
In fact, he argues that if the story comprehension experiments of Barrett
and Keil have any bearing on this issue, it is to suggest that the causal role of
culture in cognition is in fact a lot stronger than most cognitive theorists of
religion seem to think.


Earlier versions of these chapters were presented at a conference held in Aarhus, Denmark
in 2006. Ellen Dissanayakes chapter was invited as a valuable contribution to the collection.
Thanks to David Warburton for his careful linguistic improvements on the chapters, and
the former Faculty of Theology at Aarhus University for substantial funding of the Religion,
Cognition and Culture research unit (RCC) and the conference at which the papers were given
on which the chapters of this book are based. Thanks as well to the Aarhus University Research
Foundation for financial assistance and, especially, Carsten Riis for his continued support. My
gratitude is also extended to the Laboratory on Theories of Religion at the former Department
of the Study of Religion (now the Department of Culture and Society) which inaugurated the
RCC. Warm thanks are also extended to the students that took care of conference practicali-
ties Anne Pedersen, Jesper stergaard and Ulla April Petrea Winther and to our secretary,
Mikkel Pade. The final stages of copy-editing, proofreading and indexing were carried out by a
dedicated and highly professional team at Acumen. Under the benevolent guidance of Senior
Editor Tristan Palmer and Prepress Manager Kate Williams, the really hard work was accom-
plished by copy-editor Hamish Ironside, typesetter Jill Sweet, proofreader Coralie James and
indexer Angus Barclay. They have made my job immeasurably easier and have greatly improved
the manuscript. On behalf of the authors and myself, I extend our warmest gratitude.


1. This section on the main evolutionary theories is a slightly revised reprint of Geertz
(2008b: 923).
2. Guthrie (1980); see also Guthrie (1993).
3. See Jensen (2002) and Geertz (2008a, 2010a).



Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. Oxford: Oxford
University Press.
Barrett, J. L. 2004. Why Would Anyone Believe in God? Walnut Creek, CA: AltaMira Press.
Barrett, J. L. 2007. Is the Spell Really Broken? Bio-Psychological Explanations of Religion
and Theistic Belief . Theology and Science 5(1): 5772.
Bering, J. M. 2006a. The Cognitive Psychology of Belief in the Supernatural. In Where God
and Science Meet: How Brain and Evolutionary Studies Alter Our Understanding of Religion.
Volume 1: Evolution, Genes, and the Religious Brain, P. McNamara (ed.), 12333. Westport,
CT: Praeger Publishers.
Bering, J. M. 2006b. The Folk Psychology of Souls. Behavioral and Brain Sciences 29: 45398.
Boyer, P. 1994. The Naturalness of Religious Ideas: A Cognitive Theory of Religion. Berkeley, CA:
University of California Press.
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Boyer, P. & P. Linard 2006. Why Ritualized Behavior? Precaution Systems and Action
Parsing in Developmental, Pathological and Cultural Rituals. Behavioral and Brain Sciences
29: 595650.
Bulbulia, J. 2004a. Religious Costs as Adaptations that Signal Altruistic Intention. Evolution
and Cognition 10(1): 1942.
Bulbulia, J. 2004b. The Cognitive and Evolutionary Psychology of Religion. Biology and
Philosophy 19: 65586.
Bulbulia, J., R. Sosis, E. Harris, R. Genet, C. Genet & K. Wyman (eds) 2008. The Evolution
of Religion: Studies, Theories, and Critiques. Santa Margarita, CA: Collins Foundation Press.
Dawkins, R. 1976. The Selfish Gene. Oxford: Oxford University Press.
Dawkins, R. 2006. The God Delusion. London: Bantam Press.
Deacon, T. W. 1997. The Symbolic Species: The Co-Evolution of Language and the Human Brain.
London: Allen Lane.
Deacon, T. W. 2003. Multilevel Selection in a Complex Adaptive System: The Problem of
Language Origins. In Evolution and Learning: The Baldwin Effect Reconsidered, B. H.
Weber & D. J. Depew (eds), 81106. Cambridge, MA: MIT Press.
Dennett, D. C. 2006. Breaking the Spell: Religion as a Natural Phenomenon. New York: Viking.
Donald, M. 1991. Origins of the Modern Mind: Three States in the Evolution of Culture and
Cognition. Cambridge, MA: Harvard University Press.
Donald, M. 2001. A Mind So Rare: The Evolution of Human Consciousness. New York: W. W.
Feierman, J. R. (ed.) 2009. The Biology of Religious Behavior: The Evolutionary Origins of Faith
and Religion. Santa Barbara, CA: Praeger.
Geertz, A. W. 2004. Cognitive Approaches to the Study of Religion. In New Approaches to the
Study of Religion, Volume 2: Textual, Comparative, Sociological, and Cognitive Approaches, P.
Antes, A. W. Geertz & R. R. Warne (eds), 34799. Berlin: Walter de Gruyter.
Geertz, A. W. 2008a. From Apes to Devils and Angels: Comparing Scenarios on the Evolution
of Religion. In The Evolution of Religion: Studies, Theories, and Critiques, J. Bulbulia, R.
Sosis, E. Harris, R. Genet, C. Genet & K. Wyman (eds), 439. Santa Margarita, CA:
Collins Foundation Press.
Geertz, A. W. 2008b. Religion and Cognition: A Crisis in the Academic Study of Religion?.
Bulletin of the Council of Societies for the Study of Religion 37(4): 915.
Geertz, A. W. 2009. New Atheistic Approaches in the Cognitive Science of Religion: On

armin w. geertz

Daniel Dennett, Breaking the Spell (2006) and Richard Dawkins, The God Delusion
(2006). In Contemporary Theories of Religion: A Critical Companion, M. Stausberg (ed.),
24263. Abingdon: Routledge.
Geertz, A. W. 2010a. Brain, Body and Culture: A Biocultural Theory of Religion. Method
and Theory in the Study of Religion 22(4): 30421.
Geertz, A. W. 2010b. Too Much Mind and Not Enough Brain, Body and Culture: On What
Needs to Be Done in the Cognitive Science of Religion. Historia Religionum 2: 2137.
Guthrie, S. E. 1980. A Cognitive Theory of Religion. Current Anthropology 21(2): 181203.
Guthrie, S. E. 1993. Faces in the Clouds: A New Theory of Religion. Oxford: Oxford University
Jensen, J. S. 2002. The Complex Worlds of Religion: Connecting Cultural and Cognitive
Analysis. In Current Approaches in the Cognitive Science of Religion, I. Pyysiinen & V.
Anttonen (eds), 20328. London: Continuum.
Jensen, J. S. 2010. Doing It the Other Way Round: Religion as a Basic Case of Normative
Cognition. Method and Theory in the Study of Religion 22(4): 3229.
Lawson, E. T. & R. N. McCauley 1990. Rethinking Religion: Connecting Cognition and Culture.
Cambridge: Cambridge University Press.
Lewis-Williams, D. 2002. The Mind in the Cave: Consciousness and the Origins of Art. London:
Thames & Hudson.
McCauley, R. N. & E. T. Lawson 2002. Bringing Ritual to Mind: Psychological Foundations of
Cultural Forms. Cambridge: Cambridge University Press.
Mithen, S. 1996. The Prehistory of the Mind: A Search for the Origins of Art, Religion and Science.
London: Thames & Hudson.
Schloss, J. & M. J. Murray (eds) 2009. The Believing Primate: Scientific, Philosophical, and
Theological Reflections on the Origin of Religion. Oxford: Oxford University Press.
Srensen, J. 2004. Religion, Evolution, and an Immunology of Cultural Systems. Evolution
and Cognition 10(1): 6173.
Sosis, R. 2000. Religion and Intragroup Cooperation: Preliminary Results of a Comparative
Analysis of Utopian Communities. Cross-Cultural Research 34(1): 7788.
Sosis, R. & C. Alcorta 2003. Signaling, Solidarity, and the Sacred: The Evolution of Religious
Behavior. Evolutionary Anthropology 12: 26474.
Sosis, R. & E. Bressler 2003. Co-operation and Commune Longevity: A Test of the Costly
Signaling Theory of Religion. Cross-Cultural Research 37(2): 1139.
Sperber, D. 1996. Explaining Culture: A Naturalistic Approach. Oxford: Blackwell Publishers.
Whitehouse, H. 1995. Inside the Cult: Religious Innovation and Transmission in Papua New
Guinea. Oxford: Oxford University Press.
Whitehouse, H. 2004. Modes of Religiosity: A Cognitive Theory of Religious Transmission. Walnut
Creek, CA: AltaMira Press.
Wilson, D. S. 2002. Darwins Cathedral: Evolution, Religion, and the Nature of Society. Chicago,
IL: University of Chicago Press.

Whence religion? How the brain constructs the
world and what this might tell us about the
origins of religion, cognition and culture
Armin W. Geertz

Origins of religion

In cognitive theories of the origins of religion, it is held that religious thought

and behaviour are by-products of or even parasitic on more basic cognitive
processes.1 Religion is not an evolutionary adaptation, genetic or cultural,
even though it may occasionally help individuals and groups. In the follow-
ing, I take issue with this approach. It must be emphasized, however, that the
work of cognitive scientists of religion, such as Pascal Boyer and Scott Atran,
has raised issues and questions that have led to useful hypotheses, experiments
and insights on religion and human cognition. For instance, they agreed that
the age-old assumption that religion produces morals and values is neither
the only, nor the most parsimonious, hypothesis for religion. According to
Boyer and Atran, humans are more or less born with, or at least quite early
on have, default moral sensibilities (cf. Pyysiinen & Hauser 2010). The evi-
dence is strong that the origins of such sensibilities are to be found in basic
social cognition.
Boyer and Atran presented similar claims, which are useful to think with,
about what religion is not. Boyer summarized what religion is not in four

1. religion provides explanations;

2. religion provides comfort;
3. religion provides social order; and
4. religion is a cognitive illusion (Boyer 2001: 5).

Even though Boyer rejects these scenarios for the origin of religion, he claims
that they are not that bad (ibid.: 6), rather they point to phenomena that
need explaining. Atran argued that religion did not originate to:

armin w. geertz

1. cope with death;

2. keep social and moral order;
3. recover the lost childhood security of father, mother or family;
4. substitute for, or displace, sexual gratification;
5. provide causal explanations where none were readily apparent; or
6. provoke intellectual surprise and awe so as to retain incomplete, coun-
terfactual, or counterintuitive information (Atran 2002: 1213).

According to Atran, It is not that these explanations of religion are all wrong.
On the contrary, they are often deeply informative and insightful. It is only
that, taken alone, each such account is not unique to, or necessary or suffi-
cient for, explaining religion (ibid.: 13).
I agree that the various elements Boyer and Atran reject are insufficient
as mono-causal explanations or scenarios for the origin of religion. I do not
think, however, that current science shows clear signs of being able to distin-
guish religious elements from the strands of the evolution of human cultural
traits. The position I argue for in this essay is that brain and cognition devel-
oped in a dialectical relationship with culture. If this holds, then it would
suggest that issues such as morality, identity, solidarity, meaning and death
must be considered from the beginning as religious issues and not secondary
to other matters. Boyer argued that having a normal human brain does not
imply that you have religion. All it implies is that you can acquire it, which is
very different (Boyer 2001: 4). But if you admit, as I do, that having a nor-
mal human brain implies that you have culture (otherwise you cant use your
brain), then how can religion be filtered out of the discussion? If religion is
something that was with us from the beginning, what would the advantages
be? Here is where I take issue with many cognitive scientists of religion and
evolutionary psychologists. All the elements in the catalogues of things that
were supposedly not the origins of religion should be found here: meaning,
dealing with uncertainty and death, developing and maintaining individual
and group identities, group mobilization and so on.
In the following, I will argue that we have brains that co-evolved with
culture with the result that we developed large brains with particular abilities
and peculiar capacities. Furthermore, every human is born with an unfinished
brain that takes decades to mature under the persistent and heavy influence
of social and cultural pressures right from day one. Similar to many other
species, however, our brains have evolved into predictive organs that help
our bodies adapt and survive in complex environments. In warm-blooded
animals, such brains are intricately driven by their affective systems. Humans,
having learned to more or less control these systems through the invention of
culture, were able to move out into hostile environments and ultimately gain
control over them.2 Human brains, however, have weaknesses and biases that

whence religion?

any talented magician, priest or politician can manipulate. These strengths

and weaknesses, I argue, are the origins of religion.
The cultural evolutionary model I favour hypothesizes that:

We are intelligent apes that are highly emotional, easily spooked, very
superstitious, extremely sensitive to social norms and virtual realities,
and equipped with nervous systems that are vulnerable to influence
from conspecifics and their symbolic worlds. These traits are prerequi-
sites for religious behaviour.

The default condition for humans in my cultural evolutionary model is non-

religious. Notably, other animals do not seem to evince anything approaching
human religious tendencies.3 But all warm-blooded animals can get spooked
and are highly tuned to social contexts and other animate creatures apes and
hominins have more of it. So the biological and psychological conditions are
present, but what more is needed? I suggest that with the expansion of the
prefrontal cortex, our added symbolic competence, self-reflection, a sense of
mortality and a sense of patterns and external forces began to dominate. This
may very well be the source of superstition and proto-magical behaviour
our supersense as Bruce Hood calls it (2009). In the following, I will present
the evidence for my hypothesis.

Culture and cognition

Gene-culture co-evolutionary approach

My approach to evolution shares the basic assumptions of recent insights

from biology.4 As evolutionary biologist Mary Jane West-Eberhard explained
it, The universal environmental responsiveness of organisms, alongside genes,
influences individual development and organic evolution (West-Eberhard
2003: vii). Thus she conceived of evolution in terms of nature and nurture,
genes and environment within the framework of a fundamentally genetic
theory of evolution. It is a theory that must include the ontogeny of all
aspects of the phenotype, at all levels of organization, and in all organisms.
What is new here is the addition of a developmental approach to evolutionary
biology. Evolutionary biology has generally restricted itself to adaption and
has traditionally not included proximate mechanisms (ibid.: viii). One of
the advantages of this approach is relevant for us:

A deep look at the evolutionary role of development reaches

beyond the issue of nature and nurture to illuminate such themes

armin w. geertz

as the patterns of adaptive radiation, the organization of societies,

and the origin of intelligence. For it is undoubtedly the assess-
ment and management of environmental and social contingencies
that has led to the evolution of situation-appropriate regulation,
with the eventual participation of the sophisticated device we call
mind. Indeed, seeing judgment and intelligence among other
mechanisms of adaptive flexibility helps explain why learned
aspects of human behaviour so closely mimic evolved traits.
(Ibid.: 20)

Learning, Eberhard argued, is just one among many environmentally

responsive regulatory mechanisms that coordinate trait expression and deter-
mine the circumstances in which they are exposed to selection (ibid.: 338).5
Learning is, of course, one of the primary social strategies used by humans
(Frith & Frith 2010: 16970), and it is not simply copying. It is, as biologists
Eva Jablonka and Marion J. Lamb have argued, a function- or meaning-sen-
sitive developmental process (Jablonka & Lamb 2005: 209); thus rejecting
the simplistic assumptions of meme theory.
Jablonka and Lamb have also argued for the new synthesis in evolutionary
biology, claiming that:

there is more to heredity than genes;

some hereditary variations are nonrandom in origin;
some acquired information is inherited; and
evolutionary change can result from instruction as well as selection
(ibid.: 1).

They claimed that assumptions about genetics in current neo-Darwinian the-

ory are incorrect. Jablonka and Lamb provided evidence showing that cells
can transmit information through epigenetic inheritance; that many animals
transmit information through behaviour; and that humans do so through
symbolic inheritance (ibid.). Thus, there are four types of hereditary systems.
As Jablonka and Lamb argued:

It is therefore quite wrong to think about heredity and evolution

solely in terms of the genetic system. Epigenetic, behavioral, and
symbolic inheritance also provide variation on which natural selec-
tion can act By adopting a four-dimensional perspective, it is
possible to construct a far richer and more sophisticated theory of
evolution, where the gene is not the sole focus of natural selection.
(Ibid.: 12)

whence religion?

In consequence, Jablonka and Lamb also rejected the widespread gene-

based evolved-module view that evolutionary psychologists have been pro-
moting until recently. The alternative view is to see human behaviour and
culture as consequences of hominids extraordinary behavioural plasticity
coupled with and enhanced by their powerful system of symbolic commu-
nication, thus increasing their adaptive abilities in the extremely variable
ecological and social environments that humans construct for themselves
(ibid.: 213).

Why did the brain evolve?

With these caveats in mind, let me proceed to the main question here: Why
did the brain evolve? One of the answers to this was the appearance of sys-
tematic tool use, in other words, the appearance of culture. This point was
already noted in the middle of the twentieth century, as indicated by the
conference anthology edited by J. N. Spuhler (1959). Scholars from a variety
of disciplines met at the Plenary Session of the Fifty-Sixth Annual Meeting
of the American Anthropological Association in 1957 to discuss the relation-
ship between tools and biological evolution. The conclusion, as expressed by
Spuhler, was unequivocal: there was no doubt that our heads, brains, and
faces reached their present shape following, rather than preceding, the making
of tools (ibid.: v). One of the participants, physical anthropologist Sherwood
L. Washburn at the University of Chicago, argued that bipedalism and tool
use were incremental to human evolution and that tools put new selection
pressures on biological evolution:

Tools changed the whole pattern of life bringing in hunting,

cooperation, and the necessity for communication and language.
Memory, foresight and originality were favored as never before,
and the complex social system made possible by tools could only
be realized by domesticated individuals. In a very real sense, tools
created Homo sapiens. (Washburn 1959: 31)

These insights were picked up shortly afterwards by anthropologist Clifford

Geertz in his seminal essays The Growth of Culture and the Evolution of
Mind (C. Geertz [1962] 1973) and The Impact of the Concept of Culture
on the Concept of Man (C. Geertz [1966a] 1973).6 As he noted in 1962:

Most crucially, it then becomes apparent that not only was cul-
tural accumulation under way well before organic development
ceased, but that such accumulation very likely played an active

armin w. geertz

role in shaping the final stages of that development. Though it

is apparently true enough that the invention of the airplane led
to no visible bodily changes, no alterations of (innate) mental
capacity, this was not necessarily the case for the pebble tool or
the crude chopper, in whose wake seems to have come not only
more erect stature, reduced dentition, and a more thumb-domi-
nated hand, but the expansion of the human brain to its present
size. Because tool manufacture puts a premium on manual skill
and foresight, its introduction must have acted to shift selection
pressures so as to favour the rapid growth of the forebrain as, in
all likelihood, did the advances in social organization, commu-
nication, and moral regulation which there is reason to believe
also occurred during this period of overlap between cultural and
biological change. Nor were such nervous system changes merely
quantitative; alterations in the interconnections among neurons
and their manner of functioning may have been of greater impor-
tance than the simple increase in their number. Details aside,
however and the bulk of them remain to be determined the
point is that the innate, generic constitution of modern man
(what used, in a simpler day, to be called human nature) now
appears to be both a cultural and a biological product in that it is
probably more correct to think of much of our structure as a result
of culture rather than to think of men anatomically like ourselves
slowly discovering culture. (C. Geertz [1962] 1973: 67).7

Thus, we can assume that two factors were incremental to the expansion of
the brain: the use of tools and the further development of social cognition.
Geertz wrote that human history has shown a fundamental dialectical rela-
tion between our evolutionary development (the expansion of the brain) and
the development of culture. The two go hand in hand. As Geertz argued,
in comparison with other animals, where genetic information plays a much
larger role in controlling behaviour patterns, humans are born with much
more general response capacities that allow far greater plasticity, but which
leave behaviour much less regulated.8 Culture is not an added ingredient to an
already completed animal, Geertz claimed, rather, it is centrally ingredient
in the production of that animal itself (C. Geertz [1966a] 1973: 47). There
is, in other words, no such thing as a human nature independent of culture
(ibid.: 49). These claims have greater impact today because of insights gained
through the development of advanced techniques during the past few dec-
ades in palaeoanthropology, archaeology, cognitive archaeology, evolution-
ary psychology and genetic analysis. My colleagues and I at the Religion,
Cognition and Culture research unit (RCC) in Aarhus emphasize the pivotal

whence religion?

claim that symbolic systems are not just important, they are, in fact, fun-
damentally formative to cognitive development. This claim is aligned with
recent advances in neuroscience.9
The engine of culture, Merlin Donald claims, is found in metacognitive
awareness (Frith 2012), but the patterns of culture, the mazes we must pen-
etrate, are generated by the cultural matrix itself The patterns that emerge
at the level of culture are not only real but dominate the cognitive universe
that defines what reality is (Donald 2001: 287). Also:

symbols of all kinds are the playthings of a fantastically clever,

irrational, manipulative, largely inarticulate beast that lives deep
inside each of us, far below the polished cultural surface we have
constructed. That passionate and devious intelligence is iso-
morphic with our conscious experience of the world This is
why the human brain cannot symbolize if it is isolated from a
culture The tension between cultural symbolic systems and
the underlying intelligences that use them determines the qual-
ity of our uniquely human modes of consciousness Culture
shapes the vast undifferentiated semantic spaces of the individual
brain. The brain takes on its self-identity in culture and is deeply
affected in its actions by culturally formulated notions of selfhood.
(Ibid.: 2856)

Human consciousness is endlessly self-curious and epistemically starved. It is

caught between the innate memory banks of the brain and the vastly com-
plex, externally available memory banks of culture. The external memory field
is, according to Donald, a mirror of consciousness, but it also changes the
ways that consciousness deals with its representations. The core of awareness,
then, is caught in a dynamic interrelation between two powerful cognitive
fields, the internal and external. The external field allows consciousness to
reflect on thought itself and to develop thinking into formal procedures and
greater abstractions, the goal of which is to improve and refine our ways of

The psychology of tool use

Recent studies in the psychology of tool use confirm these assumptions.

Cognitive archaeologist Lambros Malafouris has formulated what he called
the brainartefact interface (BAI) and demonstrated how artefacts are
constitutive in cognition (Malafouris 2010: 265; cf. Geertz 2010a). He
argued that:

armin w. geertz

the functional structure and anatomy of the human brain is a

dynamic construct remodeled in detail by behaviourally impor-
tant experiences which are mediated, and often constituted, by
the use of material objects and cultural artefacts which for that
reason should be seen as continuous integral parts of the human
cognitive architecture. (Malafouris 2010: 266)

In a special 2008 issue of Philosophical Transactions of the Royal Society B, reis-

sued in 2009, neurologists and archaeologists pooled their knowledge and
were in general agreement about:

the special roles that materiality, cultural practices and social inter-
action play in the shaping of the human mind throughout its long
evolutionary and developmental trajectories the hallmark of
human cognitive evolution may not be based on the ever-increas-
ing sophistication or specialization of a modular mind, but upon
an ever-increasing representational flexibility that allows for envi-
ronmentally and culturally derived plastic changes in the structure
and functional architecture of the human brain.
(Renfrew et al. 2009: x)10

Psychologist Margaret Wilson has argued that culture re-engineers cogni-

tion during ontogenesis, not only in terms of content but also in terms
of function through a process she called cognitive re-tooling which uses
both material and non-material objects (Wilson 2010: 18081). She argued
that these cognitive inventions become firmware, constituting a re-engi-
neering of the individuals cognitive architecture. She argued, furthermore,
that ontogenetic experience from ones cultural context serves to re-tool the
developing mind into a variety of disparate cognitive phenotypes (ibid.:
180). Following Tomasello (1999), Wilson claimed that what makes human
cognition smarter is to a large extent not a collection of evolved cognitive
modules for accomplishing all our unique cognitive tricks, but rather the
ability to re-engineer our existing cognitive resources in a flexible fashion
(Wilson 2010: 185; cf. Bulbulia 2008; Mesoudi 2011).
This new line of research confirms not only the intimate evolutionary rela-
tionship between cognition and culture, but also indicates the crucial role
that it continues to play in ontogenetic development.

whence religion?

The archaeological record

When did proto-religious beliefs and practices appear?11 Biologist Terrence

Deacon argued that some form of symbolic cultural communication must
have been in place right from the point that neuroanatomical changes (e.g.
increasing relative brain size) become evident in hominids (Deacon 2003:
94.) In other words, the creature we are searching for among our anteced-
ents is before Homo sapiens. It would be found in any of the other hominins:
Homo habilis, Homo ergaster, Homo erectus, Homo neanderthalensis and so on.
And their features would probably already have appeared in simpler form in
the australopithecines. Deacon argued that the transition would be between
Australopithecus and Homo habilis, which would place the first noticeable
changes some 2.2 million years ago. Neuropsychologist Merlin Donald, how-
ever, has argued that there must have been an intermediate stage before full
symbolic competence. That stage he termed mimetic, which he roughly
dates from the appearance of Homo erectus around 2 million to 400,000 years
ago. If we accept the appearance of stone tools as an indication of culture, the
oldest tools found were used by Australopithecus garhi 2.52.6 million years
ago (Asfaw et al. 1999; de Heinzelin et al. 1999). McPherron et al. (2010)
have even found evidence as old as 3.3 million years ago. This would make
sense in terms of evolution. The systematic use of tools evidenced by the first
Homo habilis didnt just happen without precedence.12 But when this species
did appear, things were changing and the brain was expanding (Tattersall
1998; Sarmiento et al. 2007; Lepre et al. 2011).

Fire control

Much more intriguing is the question of when fire was brought under con-
trol. There is evidence indicating that even the australopithecines and Homo
erectus had tamed fire by 1.4 million years ago in southern and eastern Africa.
But the evidence is shaky (Roebroeks & Villa 2011). There is evidence at
Qesem Cave in Israel of the habitual use of fire during the period of 400,000
to 200,000 years ago, maybe by Homo erectus inhabitants at first and ana-
tomically modern humans and/or Homo neanderthalensis later in that period.
There is evidence of hearths and a residential base scenario (Karkanas et al.
2007: 208).
Francesco Berna and colleagues, however, using new techniques coupling
micromorphology to FTIR spectroscopy have discovered the habitual use of
fire by Homo erectus in one-million-year-old sediments at Wonderwerk Cave,
Northern Cape province in South Africa (Berna et al. 2012), thus supporting
other evidence in Kenya (Bellomo 1994a, 1994b) and Israel (Goren-Inbar

armin w. geertz

et al. 2004). The authors argued that their evidence seems to support the
cooking hypothesis introduced by biologist Richard Wrangham (Berna et
al. 2012: E1220). That hypothesis concerned Homo ergaster (about 1.5 mil-
lion years ago) who became a savannah creature with the new Acheulian tool
style and the use of hand axes which evidently indicates mental progress. They
were the first hominins to live outside the tropical zone and therefore must
have made use of fire. Wrangham argues that the use of fire was a significant
factor in hominin evolution:

The extra energy gave the first cooks biological advantages. They
survived and reproduced better than before. Their genes spread.
Their bodies responded by biologically adapting to cooked food,
shaped by natural selection to take maximum advantage of the
new diet. There were changes in anatomy, physiology, ecology,
life history, psychology, and society. Fossil evidence indicates that
this dependence arose right back at the beginning of our time
on Earth, at the start of human evolution, by the habiline that
became Homo erectus We humans are the cooking apes, the
creatures of the flame. (Wrangham 2009: 14).

Anthropologist Frances Burton, however, argued that the australopithecines

were in contact with and used to fire. She argued that the control of fire is
not a by-product, but one of the most important factors that started the
hominin line in the first place (F. D. Burton 2009: 9). Her hypothesis, how-
ever, was not a cooking hypothesis. She argued that the use of firelight at
night introduced significant changes in brain structure and the hormone sys-
tem which in the long run significantly influenced social relations and, ulti-
mately, language (ibid.: 1011). Cooking was already used 1.7 million years
ago (Hoberg et al. 2001; F. D. Burton 2009: 17):

I suggest that the more that these hominins did something to

the environment by way of intervening with natural processes
in this case, artificially increasing the daily hours of exposure to
light the more they dampened the effect of the environment on
themselves. In consequence, they took over the direction of their
own evolution on many important fronts, the most significant of
which may be the development of mind. The acquisition of fire
is precisely the kind of intervention that could have initiated and
perpetuated this chain of events. (F. D. Burton 2009: 18)

These assumptions still need clear and systematic documentation, but the
evidence is promising despite the paucity of the ancient material remains.

whence religion?

Regardless of the date assigned to the origins of controlled fire and the pre-
cise scope accorded to its evolutionary effects, the key point is that fire is
undoubtedly much older than our lineage, and unquestionably affected the
course of human evolution in dramatic ways. Importantly, though fire is uni-
versal and ancient, fire technology is not encoded into our genes. It is rather
transmitted through cultural learning. Is symbolic thinking like that? We next
consider this question.

Symbolic objects

Until recently, most scholars agreed that the use of symbolic objects is a sin-
gular characteristic of Homo sapiens, and assume that full-blown symbolic
thought is first evidenced in the cave paintings in Spain and France dated
around 30,00035,000 years ago. The evidence, however, suggests that sym-
bolic thinking extends much further back in time in Africa.
Evidence from the Middle Awash valley in Ethiopia and the Central Rift
Valley of Kenya indicate major tool innovations in the transition from the
Acheulian to the Middle Stone Age somewhere around 200,000300,000
years ago. It is assumed that this transition coincided with the appearance of
Homo sapiens. Lawrence S. Barham suggested also that language played a cru-
cial role in these innovations (Barham 2001: 70). Sally S. McBrearty and col-
leagues, however, argued that there was no human revolution and that the new
tool styles were being used by several hominin groups (Johnson & McBrearty
2010; McBrearty & Brooks 2000; Tryon & McBrearty 2002; Weaver 2012).
The earliest evidence of Homo sapiens is a skull found in the Omo Valley of
southwestern Ethiopia in 1967. It lay together with the skull of a Homo erec-
tus. They have since been dated to 195,000 years ago (Garrigan & Hammer
2006: 669).
Evidence from Herto, Middle Awash, Ethiopia, dating radioisotopically
some 30,000 years later (160,000 and 154,000 years ago) is important in this
context because of the state of the three crania that were found. The crania
evidence post-mortem mortuary practices (White et al. 2003: 742). Clark
et al. discussed the pros and cons of this assumption. They found cut marks
on all three crania indicating defleshing and abundant superficial marks of
repetitive scraping that, in comparison, are not found in remains processed
for consumption (Clark et al. 2003: 751). Furthermore, the juvenile cra-
nium is smooth and polished indicating post-mortem manipulation. The
oldest indications of such behaviour have been documented for earlier hom-
inin species 600,000 years ago in both South Africa (Pickering et al. 2000)
and Ethiopia (White 1986). There is also evidence of such behaviour at the
Krapina Neanderthal site (Russell 1987).

armin w. geertz

The earliest evidence of burial behaviour among Homo sapiens is at

Qafzeh, dated 90,000120,000 years ago (Bar-Yosef 1998; Bar-Yosef et al.
1986; Vandermeersch 1981). One of the burials seems to be associated with
grave goods (Bar-Yosef & Vandermeersch 1993; McBrearty & Brooks 2000;
Vandermeersch 1981). There is evidence of other burials as well (McBrearty
& Brooks 2000: 51921). Many archaeologists assume that the shaft in Sima
de los Huesos, Atapuerca, Spain, with the skeletons of 32 hominids, dated
200,000320,000 years ago may have been a deliberate burial (Bischoff et al.
1997), although other interpretations have been put forward (Arsuaga et al.
1997; Andrews & Fernandez-Jalvo 1997; McBrearty & Brooks 2000: 519).
There is plenty of evidence from the Middle Stone Age (MSA) of the use
of ochre, shell beads and geometric engravings on red ochre (for summary see
Coulson et al. 2011: 18). An important example of religious symbolism and
ritual behaviour is a huge quartzite outcrop that can be seen as a zoomorphic
form, such as a snake or a tortoise head, carved with hundreds of cupules of
varying sizes and shapes from the site of Rhino Cave, Tsodilo Hills, Botswana.
Excavations at the base of this outcrop revealed an assemblage characterized
by an unexpectedly large number of MSA points, which are for the most part
produced in non-locally acquired raw materials. These points, which are col-
ourful, carefully and often elaborately made, were either deliberately burned
to the point where they could no longer be used, abandoned or intentionally
smashed (Coulsen et al. 2011). As these excavations are positioned adjacent
to the cave wall, they did not comply with the requirements for thermolumi-
nescent dating. Therefore, the MSA assemblage from this site had to revert
to a typological comparison to the nearby well-dated MSA sites of White
Paintings Shelter (66,400 6500bp and 94,300 9,400bp; Robbins et al.
2000: 1092) and the open-air pan site of Gi (77,000 11,000bp; Brooks
et al. 1990: 62), which is approximately 120 km southwest of the Tsodilo
What other evidence is there for symbolic behaviour and the use of sym-
bolic objects before Homo sapiens? Two stones with engraved lines have been
found in Kenya, dated 260,000420,000 years ago and the use of ochre has
also been established during this period. The proto-figure found in Tan-Tan,
Morocco, dated 300,000500,000 years ago, a natural stone object with
anthropomorphic grooves and traces of ochre (Bednarik 2003a: 96) is the
earliest example of symbolic art in Africa. Ian Watts has argued for the use of
ochre in connection with collective ritual from around 300,000 or more years
ago (Watts 1999, 2002, 2009).
Another figurine, found in the Golan Heights and dated between 233,000
470,000 years ago, is known as the Venus of Berekhat Ram. It is a naturally
shaped pebble with engraved lines (Bednarik 2003a: 93; 2003b). An inter-
esting stone with what seem to be three faces, found at Makapansgat, South

whence religion?

Africa, is 2.53 million years old (Bednarik 1998). It is a natural stone, but
it was transported by australopithecines from its place of origin, which might
indicate that the bearers appreciated and understood the likeness (Bednarik
2003a: 97; Dart 1974). The same can be said about the penis-shaped fossil
found in Erfoud, Morocco, dated to around 300,000 years ago (Bednarik
2002; Bednarik 2003a: 97).
All in all, this evidence indicates that Homo erectus and/or Homo ergaster
or Homo heidelbergensis (depending on the classification system used) showed
signs of symbolic and perhaps religious behaviour.

The kinds of brains it takes

In ways that are still unclear to us, humans developed consciousness and sym-
bolic competence through culture. One way of phrasing it, in tune with two
well-known books, is that brains made up their minds and started thinking
through culture (Freeman 1999; Shweder 1991). Taming fire for heat, light-
ing and cooking, developing tools and utensils for hunting, processing food,
producing clothes and making adornments, all happened during the phylo-
genetic process of what we have become, namely, the human clan.
In order to understand the kinds of brains it takes to become humans,
we need to push our evolutionary antecedents and possible default mecha-
nisms much further back than is usual. As Ursula Goodenough and Terrence
Deacon wrote:

To say that our brains have undergone critical reconfigurations as

they evolved their capabilities for symbolic (self )-representation
is not to say that our common-ancestor brains were left in the
dustbin we share strong cognitive and emotional homologies
with our primate cousins, and, to the extent that degradation or
reconfiguration went into generating our capacity for language,
it occurred in a primate brain that remains very much a primate
brain one of the things that we do with our symbolic minds is
experience our primate minds symbolically. Our primate minds
have not gone away (although some phylotypic instincts have
been lost and perhaps reconfigured), nor are they experienced as
apes would experience them. They are experienced as things are
experienced by human minds: symbolically.
(Goodenough & Deacon 2003: 81314)

Sociologist Jonathan H. Turner and anthropologist Alexandra Maryanski

have persuasively argued similarly that we are basically hang-loose apes with

armin w. geertz

a history of developing ever-more restrictive social systems. The function of

such systems was to improve group survival in dangerous and ever-more com-
plex environments:

the very first institutional systems of human societies kinship,

religion, and economy were not natural for an evolved ape
with a penchant for a hang-loose forest lifestyle. Indeed, proto-
human society represented from its very beginnings an uneasy
necessity that often stood in conflict with humans ape ancestry.
And, as societies developed from their simple hunting-and-gath-
ering formations into ever-more complex sociocultural systems,
this tension between our hominoid ancestry and the structures
in which humans were forced to exist remained and, indeed,
increased. Our view is that we cannot fully understand human
behaviour and the colossal sociocultural creations that human
hominoids have constructed without first appreciating that the
ape in us was not extinguished, nor obviated by culture; indeed,
our ancestry continues to place pressures on individuals and their
sociocultural creations. (Turner & Maryanski 2008: 34)

Despite the great advances made by pioneers in the cognitive science of reli-
gion, work still remains. We need to increase our time scales and pay close
attention to the growing field of social cognitive neuroscience. The latter is
necessary because neurologists and neuropsychologists are discovering a lot
about the brain that gives us a better foothold on the neurobiological and
cognitive correlates of religious ideas and behaviours.14 Thus, for a theory of
the origins of religion, cognition and culture, we not only need to pay close
attention to the interdependent causal links between culture and cognition,
and the role that religious symbolic systems play in it, we also need to pay
attention to the brain and the body. Cognition, I argue, is a combination of
embrainment, embodiment and enculturation (Geertz 2010b: 37).
So, what kinds of brains are we looking for in the archaeological record?
Boyer has summarized a handy list of human characteristics that are useful

1. a hypertrophy of social intelligence;

2. a capacity to evaluate potential cooperators and to detect potential
3. a repertoire of moral feelings;
4. the stability of some un-fakeable (or difficult to fake) signals;
5. easily produced self-deception; and
6. emotional rewards for gossip (Boyer 2000: 2034).

whence religion?

These are indeed basic characteristics of human beings, but I will argue that
we need to rearrange and in some cases reformulate these features. Thus, the
kinds of brains posited here consist of the following fundamental features:

A finely honed social cognition.

A drive to communicate and cooperate.
A self-deceptive brain.
A superstitious brain prone to unusual mental and/or emotional

I will discuss in detail each of these features in the following.

A finely honed social cognition

As noted by Clifford Geertz above, our social cognition and the invention
and use of tools must have had equal influence on the expansion of the brain.
This assumption is supported by the neurological evidence, as Chris and Uta
Frith have summarized so well in a number of papers, especially The Social
Brain (Frith & Frith 2010) and Social Cognition in Humans (Frith &
Frith 2007). For them, the singularly most important feature of human cog-
nition is our social brain.
Our social cognition is ready right from birth. We share our sociality with
other species, not least our ape cousins, but our sociality is more complex.
Frith and Frith summarized this sociality as evidenced in neurological experi-
ments in a more or less progressive series of features (Frith & Frith 2010).
First, whenever we move, we broadcast involuntary signals about ourselves.
These are cues indicating biological agency, gender, emotion and other infor-
mation critical to determining whether the agent is friend or foe. Second,
people perceiving a biological agent will automatically look for indications
of the agents intentions and goals. Studies have shown that people can and
will attribute intentions even to non-human or non-biological agents. Third,
we discern intentions by movement alone based on prediction. By predicting
what movement the agent will make, we can test our predictions. Prediction
error thus contributes to updating our information about that agent. Fourth,
we improve our predictions by imagining what the agent knows and does not
know from the agents point of view. This is known as mentalizing or having
a theory of mind and is present from an early age. The ability to mentalize
is crucial to communicating with other people. This ability also allows us to
deceive and manipulate other minds. Fifth, when two people communicate,
they tune in to each other by unconscious imitation of each others move-
ments. This is called the chameleon effect, and it increases empathy and

armin w. geertz

rapport. As Frith and Frith noted, this motor and emotional resonance is
rewarding and is modulated by knowledge, beliefs and strength of interaction
(Frith & Frith 2010: 168).
Sixth, social interactions depend on more than these features. Joint action
with a common goal increases the success of social interactions, and the abil-
ity to perform coordinated social interaction seems to rely on the much-
acclaimed mirror neurons that have been discovered in monkey brains. The
brains mirror system seems to be the mechanism behind empathy and emo-
tional contagion. Seventh, the main feature of human sociality is the deliberate
use of social signals (ostensive gestures) indicating the desire to communicate
something to someone else. These signals are mostly visual and there is a need
to close the loop as Frith and Frith call it, through reciprocal communica-
tion back to the sender. Thus, both parties can update their predictions about
each other (ibid.: 16970). Eighth, a significant feature is the human ability
to deliberately teach conspecifics. Not only is it essential for infants success-
ful entrance into the human world, infants are already geared to discern and
learn from ostensive gestures. Studies of adult/infant interactions show cross-
cultural, intuitive and stereotypical patterns of interaction that capture infant
attention. These patterns are crucial to the learning of language. As Frith
and Frith noted, pedagogy is a unique human ability that makes cultural
accomplishments possible in the first place (ibid.: 170). Ninth, economic
games have revealed the complexities of partner interaction. Key elements are
influence (tracking the influence of ones own actions on others), knowledge
of others past actions and knowledge of ones own past actions. Tenth, social
cooperation relies on communication and altruistic behaviour. One of the
forces behind the drive to cooperate, Frith and Frith argued, is the wish to
build a good reputation (ibid.: 171). This is the so-called audience effect,
i.e. we behave better if we are, or think we are, being observed by others.
Thus, ostensive signalling is the basis of reputation building. Eleventh, the
senders and receivers of signals close the loop and establish trust. If trust is
broken, ways must be found to repair it. Forgiving and cooperative behaviour,
despite defection, seem to work best. Prosocial and cooperative behaviour
seems to be our default mode of behavior when we are not thinking very
deeply about what we are doing (ibid.).
Mentalizing is a kind of mind reading process. It assumes that a person is
not only conscious, but is also self-conscious. It assumes that self-conscious-
ness also implies consciousness of others and that the primary use of thought
is to simulate activities, actions and consequences in relation to other people.
The amazing thing about our development as a species is that our conscious-
ness made it possible for us to be aware not only of ourselves, but also of oth-
ers. We think not only for ourselves, but also for others. We think not only
of ourselves, but also of others. We develop during the first two years of our

whence religion?

lives the art of looking in on our own mental states, or inner mindsight, as
evolutionary psychologist Nicholas Humphrey called it. Almost immediately,
Humphrey pointed out, we use this mindsight to make sense of other people
(Humphrey [1986] 2002: 94). This is the art of doing psychology, and it is
achieved, Humphrey argued, through empathy.
A final set of features is how the brain makes sense of the world. Chris
Frith explained in his book Making Up the Mind (Frith 2007) how the brain
constantly compares input from the environment in relation to its predic-
tions and mappings of the body and the world. The brain is not just a passive
computer waiting for data input so that it can produce behavioural output, as
many cognitivists claim. Numerous studies indicate that the brain not only is
way ahead of our conscious (and even unconscious) behaviour, it is way ahead
of the bodys senses. We do not see and feel the world out there. We see and
feel the world that our brain creates. As Frith argued:

Everything we know, whether it is about the physical or the men-

tal world, comes to us through our brain. But our brains connec-
tion with the physical world of objects is no more direct than our
brains connection with the mental world of ideas. By hiding from
us all the unconscious inferences that it makes, our brain creates
the illusion that we have direct contact with objects in the physi-
cal world. And at the same time our brain creates the illusion that
our own mental world is isolated and private. Through these two
illusions we experience ourselves as agents, acting independently
upon the world. But, at the same time, we can share our experi-
ences of the world. Over the millennia this ability to share experi-
ence has created human culture that has, in its turn, modified the
functioning of the human brain. (Frith 2007: 17)

Thus, when dealing with the world, we are dealing with the brain. When
dealing with the social world, we are in fact embedded in the mental world
of others just as we are embedded in the physical world (ibid.: 184). But we
keep working at trying to read other peoples minds.
Our brain is constantly predicting the future, not only in the physical
world, but also, and more importantly, in the social world. In order to do this,
it draws on maps that are imprinted in the brain (such as the sensory-motor
areas) or developed through associative learning or provided by our cultures.
In this way, perception is a prediction of what ought to be out there in the
world and this prediction is constantly tested by action (ibid.: 132).15 This
predictive activity frees us from the tyranny of our senses and environment
(ibid.: 136), but can, of course, also lead to really bad (and sometimes fatal)

armin w. geertz

A drive to communicate and cooperate

As mentioned, Nicholas Humphrey argued that mindsight, or the art of doing

psychology, is achieved through empathy. Empathy seems to be inborn, but it
is also nurtured through the complex interactions between the child and its
mother and closest family members. Through nursing, the child gains chemi-
cal and somatic access to its mothers inner world, according to neurologist
Daniel J. Siegel and others (Siegel 2001; Cozolino 2002). Already here, we
find the formative nature of religion and culture, since each culture has its
own nursing styles. There are interesting studies which show how nursing
styles reflect models for ideal individuals in different cultures (DeLoache &
Gottlieb 2000; Yovsi 2003; Yovsi & Keller 2003).
Other mechanisms which help develop the childs empathy are mirror neu-
rons in the brain which automatically stimulate areas that would normally be
stimulated if the child itself were performing the perceived action (Rizzolatti
et al. 2002). Children seem to be equipped with the ability to imitate their
mothers facial expressions even though they could not possibly have a con-
ception of their own facial expressions.16
Merlin Donald argued that the childs initial abilities are manipulated by
parents, siblings and friends to capture the infants attention and make it
aware of attentional flags and markers that will eventually lead it to the heart
of culture (Donald 2001: 255). Thus the reciprocal eye contact, the exagger-
ated vocalizations and gestures, the reduction of communicative speed, the
use of hugging and kissing, etc., are all attempts to regulate the childs imme-
diate sentience and teach it the cultural scaffolded system (ibid.: 2556).
Right from the beginning, the child is subjected to massive socialization
techniques which help internalize the objectivized assumptions and repre-
sentations of a particular culture. The kinds of ideal patterns of behaviour
are controlled innocently at first through parental approval and disapproval,
but later become more systematically routinized. An important device is the
use of narrative. Daniel Siegel has shown that parents ability to relate their
autobiographical narratives are crucial not only to the childs mental world,
but also to its physical brain. A process of neural and bicameral integration
occurs through the use of coherent narrative. Such narratives also allow the
child to begin formulating its own story. But for many years, the childs nar-
rative is controlled by its parents, its siblings and other authority figures. For
many years, it is a question of being in someone elses narrative. Such narra-
tives, seen in the canvass of human history, are mostly religious ones. They are
narratives about a childs struggle to become the ideal person (Geertz 2011a;
Ochs & Capps 1996; Ochs & Shohet 2006).
As the child grows older, stricter instruments replace imitation and rou-
tines such as persuasion, force, even punishment and violence in some cases.

whence religion?

Systematic education is introduced and when the individual becomes a legal

person, social instruments of power are employed: law enforcement, religious
and political rituals and public displays, communal activities, initiation ritu-
als, etc. The individual is not simply a private person. He or she takes on a
personality model and a social role. He or she becomes the values of a cul-
ture and a religion. Everything in an individuals life functions to that end
(Mandler 1988, 1992, 2005).17
As mentioned, we are born with certain capacities which allow the above-
mentioned instruments and techniques to function. We are born with social
intelligence and moral sensibility (Frith & Frith 2007, 2010). We have not
only the ability, but the very need, to take on worldviews (C. Geertz [1966b]
1973; Koltko-Rivera 2004). We have the capacities to deal with imaginary
worlds and make them real, with imaginary beings and make them personal,
with imaginary narratives and make them embodied. As I wrote elsewhere:

Religious narrative promulgates, extrapolates and investigates the

significance of virtual worlds in real time contexts while at the
same time refining them in all their majestic virtuality. Religious
narrative provides paradigms for human identity thereby provid-
ing narrative governance of human cognition and emotion.
(Geertz 2011a: 9)

It seems evident that language evolved to increase the potency of commu-

nication. Much of our communication concerns other people in the form
of gossip, small-talk and everyday conversation. We are concerned with
communicating states of mind and emotion as well as behaviour and the
consequences of behaviour (others as well as our own). This kind of commu-
nication is deeply satisfying to us in much the same way that grooming is for
apes and other animals. Psychologist Robin Dunbar has argued that gossip
and small-talk is a kind of grooming that releases opioids that enhance a sense
of well-being and belonging (Dunbar 1996). Gossip is a tool for monitoring
social relations and for reputation management (Cox 1970; Geertz 2011c:
37983; Goffman 1959).
One last aspect of the drive to communicate and cooperate nicely dovetails
with research on costly signals. As mentioned, costly signals have to do with
reputation management. Behavioural ecologist Richard Sosis and cognitive
scientist of religion Joseph Bulbulia pioneered the costly signalling hypothesis
in the study of religion. They conceived of costly signalling in its widest sense
as mechanisms that link information properties signals to mechanisms
that generate mutually benefiting social-interactive behaviours cooperation
(Bulbulia & Sosis 2011: 371). Bulbulia and Sosis argued that the use of costly
signals helps groups (and individuals) overcome the problems of establishing

armin w. geertz

ways to ensure trust among partners and to detect free-riders and defectors
(Bulbulia 2004a, 2004b). Based on the biology of indexical signals in various
species (Zahavi et al. 1997), Sosis and Bulbulia have shown in several studies
that costly signals do secure group survival benefits (Sosis & Bressler 2003;
Sosis et al. 2007; see also Bulbulia & Sosis 2011; Chen 2010; Sosis 2000,
2003; Sosis & Bulbulia 2011).
In agreement with the predictive error hypothesis promoted here, Bulbulia
and Sosis have argued persuasively that costly signals are integral to social pre-
dictions (Bulbulia & Sosis 2011: 372). In other words, groups, large or small,
must find ways of predicting individual and group social behaviour. Bulbulia
defined religious behaviour as a costly signal capable of authenticating reli-
gious commitment (Bulbulia 2004b: 669). As Bulbulia and Sosis argued:
We believe that cultural evolutionary scholars are correct to notice that
norms evolve to manage problems of cooperative prediction among strangers,
the solution of which is a condition for the possibility of cooperation in large
social worlds (Bulbulia & Sosis 2011: 3734). Bulbulia and Sosis call the
information properties of the systems involved charismatic ecologies that
evolve to compel relatively powerful and automatic cooperative responses
across large populations (ibid.: 376).
Psychologist Joseph Henrich has explored the biases that learners have in
identifying individuals in the social environment that can be depended upon.
Henrich and psychologist Francisco J. Gil-White posited a class of cognitive
mechanisms known as prestige-biased transmission (Henrich & Gil-White
2001), by which is meant model-based cues of prestige, success, skill, age,
ethnicity (marked by dialect, dress, etc.) and sex (Henrich 2009: 245). These
cues influence a wide range of domains even though they in principle have
nothing to do with each; for instance people might take on a well-liked actors
political views even though good actors are not necessarily astute political
observers. Henrich pointed out that a highly prestigious individual motivated
by self-interest could express a degree of commitment to a belief or opinion
different from her own, which once adopted by others could yield benefits
to her and costs to the learners (ibid.). Examples of this are legion.
Henrich proposed that learners have evolved to attend to credibility
enhancing displays (CREDs) that provide the learner with reliable measures
of the models actual degree of commitment to (or belief in) the representa-
tions that he has inexpensively expressed symbolically (e.g. verbally) (ibid.:
2445). CREDs are a kind of cultural immune system18 that in principle at
least can expose Machiavellian manipulators because of their great costs (ibid.:
247). CREDs are also essential in commitment to counterintuitive notions.
As Henrich pointed out, the counterintuitive representational content of reli-
gious conceptions are not enough in themselves to explain why people believe
in them. Their believability is enhanced by CREDs (ibid.: 246).

whence religion?

Henrich summarized the evidence indicating that:

1. beliefpractice (ritual) combinations are spread by cultural group selec-

tion (CGS);19
2. participation in costly rituals is associated with prosocial in-group
behaviour, because costly rituals transmit commitment to group-ben-
eficial beliefs/goals to participants; and
3. institutions requiring costly displays by members transmit higher levels
of belief commitment and thereby promote cooperation and success in
intergroup or interinstitution competition (ibid.: 245).

The emulation of prestigious models of social behaviour, it should be noted,

is freely conferred deference not dependent on force or the threat of force
(Henrich & Gil-White 2001). It is a much more powerful mechanism than
dominance and one that is selected for in our species.
Religion is a terrifyingly effective instrument for the mobilization of group
violence. Almost all wars and terrorism thrive on religiously motivated group
identity, no matter whether other factors lurk behind the scenes. Both Boyer
and Atran presented, I feel, solid arguments concerning fundamentalism
and violence (Atran 2010; Boyer 2001: 2926). Here the argument is that
religious behaviour must be demonstratively costly in order for individuals
to prove their commitment. In fact, Atran defined religion in such terms:
Roughly, religion is (1) a communitys costly and hard-to-fake commitment
(2) to a counterfactual and counterintuitive world of supernatural agents (3)
who master peoples existential anxieties, such as death and deception (Atran
2002: 4). Emotionally motivated self-sacrifice to the supernatural, Atran
argued, stabilizes in-group moral order, inspiring competition with out-
groups and so creating new religious forms (ibid.: 268).
In their essay Dying for an Idea, Charles A. Ziegler and Benson Saler
specifically address these issues in relation to suicide cults (Ziegler & Saler
2012: 6066). They argued that the ability to suppress species survival dispo-
sitions is inculcated in each generation from infanthood and on. They pro-
vide evidence of the possible neural mechanisms that augment inculcation,
namely, those involved in the production of trust (oxytocin), suggestibility
(dopamine) and conviction (polygenic inheritance).

A self-deceptive brain
We must seriously accept the bald fact of our credulity, i.e. proneness to
illusion, self-deception and confabulation (Hirstein 2005).20 We are, in fact,
incredibly credulous. We evince wonder and delight over apparent miracles
and are easy victims of sleights-of-hand. Furthermore, we see what we want

armin w. geertz

to see and miss what we ought to see. We produce illusions and are self-
deceptive. We are oppressively smug and certain at some times and filled
with techniques for warding off the unknown and dangerous at others (R. A.
Burton 2008). We are existential illusionists (Bering 2006, 2011), and some
experimental psychologists claim that we might have a tacit acceptance of
sympathetic magic (Hood et al. 2010).
Then there are certain features of our brains. Neuropsychologist Cordelia
Fine has summarized the quirky features of our brains in a wonderful little
book A Mind of Its Own: How Your Brain Distorts and Deceives (Fine 2006).
The chapter titles say it all: the vain brain, by which she means vain, self-serv-
ing and overconfident; the emotional brain, which is driven by our feelings;
the immoral brain, meaning here, post-hoc (mostly false) rationalizations;
the deluded brain, which produces illusions and delusions; the (stubborn)
pig-headed brain; the secretive brain, driven by unconscious motivations; the
weak-willed brain, as in the common expression, my mind has a mind of its
own: I cannot even control my own thoughts; the bigoted brain, which, as
is well-known, is filled with stereotypes and biases; and, finally, the vulnerable
brain which is an adroit manipulator of information.
More importantly, predictive minds are expectant minds. In other words,
we expect more than the sensory world gives us. Our expectations are con-
stantly being compared to the incoming sensory information. If we or others
use techniques to separate the connections between our expectations and the
senses, the brain can be manipulated to various ends. The new field of magi-
cology or neuromagic is in the process of uncovering the neural correlates
of illusions and illusion-making.
Neurologists Stephen L. Macknik and Susana Martinez-Conde together
with Sandra Blakeslee have analysed a large number of magicians tricks and
illusions in terms of how the brain processes input from the senses. One of
the major avenues into our sensual world is that of sight. What happens
when light strikes the eyes is that the system of photoreceptors convert light
into electrochemical signals that detect contrast. This ability, they noted, is
the basis of all cognition, including your capacity to see, hear, feel, think,
and pay attention. Without it, the world would have no boundaries and your
brain could make no sense of itself or anything outside itself (Macknik et
al. 2010: 11). The signals are sent through the optic nerve to the thalamus
after which the information is passed on to the primary visual cortex. At this
point, the brain begins to distinguish line orientations, edges, corners, etc. in
the scene being perceived. Then the signals move on to other neurons that
respond to contours, colours, motion and so on.
The remarkable thing is that from two-dimensional information input,
our brain creates a rich and nuanced three-dimensional world. And yet, each
eye is roughly equivalent to a one-megapixel camera (ibid.: 1213). How

whence religion?

does this happen? The answer is that what we see is not the world as it is, but
the world as simulated by the predictive brain filling in all the information
gaps. This is what magicians (and miracle-workers, prophets, faith-healers,
shamans, witches, etc.) exploit: the fact that your brain does a staggering
amount of outright confabulation in order to construct the mental simulation
of reality [is] known as consciousness (ibid.: 9).21
These features taken together within the framework of extreme sociality
produce a creature that can be desperately dependent yet incredibly independ-
ent, whimsical yet determined, self-serving yet startlingly altruistic, appall-
ingly credulous yet highly sceptical, debilitatingly insecure yet supremely
confident, devious yet true, devastatingly violent yet incredibly peaceful and
forgiving a curious and fearful creature inextricably woven into collective
webs of symbolic, virtual worlds (Bulbulia 2009).

A superstitious brain prone to unusual

mental and/or emotional experiences
The term superstition is problematical because it has served so long in a
pejorative sense. I understand the term, however, to mean not only the appar-
ently illogical behaviours associated with the performance of important acts,
such as hanging your shoes in the same way in the locker room before every
game or performing stereotypical movements before hitting the golf ball,
but also acts that are meant to prevent hazards, such as hanging diapers in a
certain way to prevent harm coming to the baby or lighting a candle before
your examination to ensure a good result (Boyer & Linard 2006; Linard
& Boyer 2006; Jahoda 1969; Vyse 1997).22 The more formal definitions pro-
duced by psychologists and others are difficult to use here, but psychologist
Gustav Jahoda proposed four categories of superstition: those forming part
of cosmologies or worldviews; other kinds of superstitions socially shared;
occult experiences; and personal superstitions (Jahoda 1969).23 The bounda-
ries between superstition, magic and religion are extremely difficult to demar-
cate, especially because the terms are used negatively and polemically by just
about everybody about everybody else (atheists about religious people, reli-
gious people about other religious people, colonialists about the colonized,
etc.). Scholars of religion are not very helpful either. For this essay, I will argue
that the three terms are aspects of one single phenomenon, namely, intuitive
assumptions about and/or beliefs in non-natural causality. By non-natural
I do not merely mean non-instrumental or opaque causality. I mean causality
that contravenes the laws of nature.24
In the following, I will discuss our proneness to unusual mental and/or
emotional experiences. Such experiences are essential to the self-confirming
mechanisms of religious belief and behaviour. World history is rife with

armin w. geertz

reports of unusual and highly emotional experiences. The evidence of spirit

possession, hallucinations, visions, witchcraft beliefs, mystical experiences,
mob hysterics, etc., attest to this tendency.25 One could see in such symptoms
the contours of certain types of personalities well known to the history of reli-
gions. Those who feel themselves possessed by demons or deities, might be
suffering from multiple personality disorder (MPD). Others who see visions
and ghosts or obey voices, could be suffering from schizophrenia. People
claiming that they remember being abducted by angels or aliens, or that
they have been subjected to satanic ritual abuse as children might be suffering
from false memory syndrome. Studies of these and other ailments are there-
fore highly relevant to the study of religion. I will briefly examine some of the
relevant evidence from current diagnostic nosology on dissociative disorders
involving alterations in consciousness that affect memory and identity.26
The problem with these phenomena is that social and cultural expecta-
tions and norms play a highly significant role in them. In her study of hallu-
cinations and sensory overrides, cognitive anthropologist Tanya Luhrmann
noted that within our bodily constraints, what she calls the cultural invita-
tion shapes extraordinary experiences a great deal. She argued that:

the local theory of mind the features of perception, intention,

and inference that the community treats as important and local
practices of mental cultivation will affect both the kinds of unu-
sual sensory experiences that individuals report and the frequency
of those experiences. Hallucinations feel unwilled. They are expe-
rienced as spontaneous and uncontrolled. But hallucinations are
not the meaningless biological phenomena they are understood
to be in much of the psychiatric literature. They are shaped by
explicit and implicit learning around the ways that people pay
attention with their senses. This is an important anthropologi-
cal finding because it demonstrates that cultural ideas and prac-
tices can affect mental experience so deeply that they lead to the
override of ordinary sense perception. That is a powerful impact.
(Luhrmann 2011: 72)27

There is a good deal of work on the socio-cultural contexts of MPD where

the conclusions tend to view such behaviour as a form of social strategy which
allows an individual to disavow actions attributed to other selves or entities
in the body. This approach harmonizes well with the idea that selves are co-
created by the brain and interpersonal relationships. A leading scholar of this
approach was psychologist Nicholas P. Spanos. His book, published posthu-
mously, on MPDs and false memories (Spanos 1996) provides insight into a
number of phenomena well known to scholars of religion.

whence religion?

In reviewing the literature, there is a relatively high frequency of multiple

self-enactments in the world, especially that of spirit possession. A leading
scholar on the subject, anthropologist I. M. Lewis, distinguished between
central and peripheral possession. The former involves possession by major
deities during public performances that confirm central religious tenets, and
the latter involves possession by capricious and amoral deities or spirits who
possess socially marginal and oppressed members of society (Lewis [1971]
1989, 1986).28 The politics of possession extends also to the fact that in
most traditional societies with possession beliefs, possession happens more
frequently to women than men, and, in some societies thus offers women
social opportunities that they otherwise would not have.
What is important here is to determine to what extent MPD can be used
to interpret spirit possession. Spanoss conclusion is that possession phe-
nomena in traditional societies are rule-governed and generally socially sanc-
tioned. The possessed usually have goals in mind, and these goals are pursued
in public or, if not, they at least serve other social functions. Most persons
who become possessed are normal, happy individuals. Only a few cases have
been reported that indicate possession as being symptomatic of severe stress
and accompanied by psychopathological symptoms (Spanos 1996: 155). On
the other hand, it is clear that historical reports of demonic possession in
Western, Christian contexts resemble the historical manifestations of MPD
(ibid.: 169). But, as Spanos noted, just as with Catholic diagnostic and treat-
ment procedures (exorcism), displays of multiplicity by MPD patients are
also often cued and reinforced by the procedures purportedly used to diag-
nose and treat that condition (ibid.) Also here we find the politics of pos-
session, not only in demonic possession, but in witch hunts as well (ibid.:
1812). This doesnt mean that people necessarily lie about their experiences,
nor that MPD doesnt exist, nor that diagnosing spirit or demon possession
as a form of MPD is irrelevant. The point, however, and once again, is that
cognition and psychology are not primarily matters of the brain.
Many of the humanistic and social science disciplines have long main-
tained a deep distrust of psychological and psychiatric explanations of ecstatic
religious behaviour. Such behaviour, if touched upon at all by psychologists
and psychiatrists, was quickly labelled neurotic or psychotic.29 Ethnographers
have, however, not escaped from similar assumptions. Much of the ethno-
graphic literature on Siberian shamanism is rife with diagnoses of arctic
hysteria, insanity and lunacy (Bogoras 1907: 415; Krader 1954; Devereux
1956; Ohlmarks 1939; Radin 1937).30 The same holds for other shamanistic
areas and possession cultures (Bateson & Mead 1942: xvi; Silverman 1967;
Langness 1965; Yap 1969; Linton 1956: 131ff.). But, as I. M. Lewis has
pointed out, there is an equal number of scholars who are better informed with
precisely the opposite view, namely, that most shamans are psychologically

armin w. geertz

healthy, highly socialized, conventionally accepted, and important members

of their communities (Lewis [1971] 1989: 165ff., citing Shirokogoroff 1935;
Anisimov 1963; Murphy 1964: 76; Nadel 1946; Wavell et al. 1966: 40).
In their book Trance and Possession in Bali, Luh Ketut Suryani, a Balinese
psychiatrist trained in the West, and Gordon D. Jensen, a Western psychia-
trist with wide experience of Balinese culture and mental health, are more
positive in their evaluation of the relationship between Western psychiatric
and brain disorders with trance and trance possession in Bali (Suryani &
Jensen 1995). They claim that trance and trance possession are forms of disso-
ciation. There are dissociative phenomena that are normal and those that are
abnormal. Dissociative disorders can be characterized by five core symptoms:

(1) amnesia, i.e. a specific and significant segment of time that

is unavailable to memory;31 (2) depersonalization, i.e. a sense of
detachment from self;32 (3) derealization, i.e. a sense that ones
surroundings are unreal;33 (4) identity confusion, i.e. a feeling of
confusion, uncertainty, or puzzlement regarding ones identity;34
(5) identity alteration, i.e. objective behaviour that indicates a
change in identity.35 All or some of these symptoms occur in the
various types of dissociative disorders (e.g. MPD and PTSD).36
(Ibid.: 278)

Dissociation is measured by using the Dissociative Experiences Scale (DES),

which lists 28 dissociative behaviours (Bernstein & Putnam 1986). Studies
indicate that over 25 per cent report a substantial number of dissociative
experiences (Ross et al. 1990).
But dissociation is also fairly widespread in normal behaviour. Examples
are driving a car and not remembering or being aware of details along the
route; reading a good book or watching an exciting movie completely oblivi-
ous to ones surroundings; doing absent-minded things like pouring coffee
in the soup bowl, leaving your keys in the refrigerator or forgetting a dentist
appointment; whiling the afternoon away daydreaming; and even dreaming
during sleep. Suryani and Jensen claim that such behaviour has great indi-
vidual and species survival value because they provide:

(1) escape from overwhelming reality, (2) a cathartic discharge of

feelings, (3) a resolution of irreconcilable conflicts, (4) an abil-
ity to perform some behaviours automatically thereby permitting
simultaneous conscious engagement in other behaviours, and (5)
a beneficial enhancement of the herd sense, i.e. the human incli-
nation to be affiliated with people engaged in similar activities.
(Suryani & Jensen 1995: 22)

whence religion?

Thus dissociation is a basic psychological process not only in normal behav-

iour but also in altered states of consciousness (ASCs), such as hypnosis,
trance and possession,37 as well as in mental disorders such as MPD. Suryani
and Jensen conclude that the psychobiological mechanism and phenomenol-
ogy of MPD is similar if not the same as possession both in the West and in
Bali (ibid.: 219). Both physiologically and psychologically, their experiences
are similar: some other entity or part of the self that to them has a real exist-
ence takes over (ibid.: 21920), and the goals are the same, i.e. to the expres-
sion of feelings and behaviours not otherwise permitted by the individuals
ego or conscious state (ibid.: 220). The difference, however, is in the thought
content of these two conditions. MPDs are highly negative and result from
traumatic experiences of early childhood, whereas trance possession in Bali is
highly valued culturally and socially.38
ASCs play an important role in a wide variety of religious behaviours.
Whether induced by hypnosis, hallucinogenic plants and drugs, meditation,
rituals, solitude, fasting, music, chanting, singing or dancing, the symptoms
are fairly uniform in psychophysiological terms. The differences one finds are
due to circumstances, beliefs, role models, concepts of person, and so on.
Altered states of consciousness involve modifying or distorting the monitor-
ing and controlling functions of mental alertness and awareness. ASC can be
defined by four features:

(1) operationally, as the product of a particular induction tech-

nique; (2) phenomenologically, as an individuals subjective report
of altered awareness or voluntary control; (3) observationally,
as changes in overt behaviour corresponding to a persons self-
report; and (4) physiologically, as a particular pattern of changes
in somatic functioning. (Kihlstrom 1994: 207)

Unfortunately, although the literature on ASC is voluminous, the experimen-

tal evidence that supports claims made for performance-enhancing qualities,
such as with meditation, are inconclusive either because the results have been
negative or because positive results have come out of experiments lacking
critical controls (Cahn & Polich 2006; Kihlstrom 1994: 247).39
Drawing on psychological and neuroscientific studies of dreaming, jog-
ging, meditation, daydreaming, hypnosis, and drug-induced states, neurosci-
entist Arne Dietrich concluded that the unifying neuroanatomical feature of
these various ASCs is that they are due to transient prefrontal cortex deregu-
lation. The phenomenological uniqueness of each state, he argues, is the
result of the differential viability of various frontal circuits (Dietrich 2003:
231). The prefontal cortex constitutes half of the frontal lobe and is primarily
involved in executive functions, not that it is the seat of consciousness, but

armin w. geertz

that it enables the top layers of consciousness by contributing the highest-

order cognitive functions to the conscious experience (ibid.: 232). The fron-
tal cortex uses highly processed information from the other brain areas to
integrate and enable a self-construct,40 self-reflective consciousness,41 com-
plex social function,42 abstract thinking,43 cognitive flexibility,44 planning,45
willed action,46 and theory of mind47 (ibid.: 2323). Furthermore, three
other cognitive functions (working memory,48 temporal integration,49 and
sustained attention50):

provide the infrastructure to compute these complex cognitive

functions by providing a buffer to hold information in mind and
order it in space-time.51 It is this superimposing of already highly
complex mental constructs that dramatically increases cognitive
flexibility and permits a unified phenomenological experience.
(Dietrich 2003: 233)

The theory promoted by Dietrich is a hierarchy of consciousness, whereby

full-fledged consciousness is a global function, but not all areas of the brain
contribute equally to it. ASCs are induced by subtle modifications of func-
tions in the top layers. Damage to lower levels are much more serious and
can destroy consciousness altogether (ibid.: 235). ASCs operate on the basis
of transient prefrontal deregulation, which by using phenomenological
subtraction (Hobson 2001), reduce certain prefrontal circuits thus contrib-
uting to the uniqueness of each altered state. For meditation, hypnosis and
daydreaming, attentional resources are controlled:

in order to eliminate extraneous information from being proc-

essed consciously. This intentional blockage permits specific pre-
frontal circuits to be run in safe mode. While in meditation and
hypnosis attention is redirected, daydreaming accomplishes this
feat by reducing attentional ability. (Dietrich 2003: 249)

Spanos has convincingly demonstrated how manipulating attentional control

and sociocognitive expectations can result in university psychology students
experiencing a variety of dissociative states as well as past life regression, mul-
tiple identities, false memories, amnesia and so on (Spanos 1996). Spanos
showed just how easy it is to manipulate emotional and cognitive states
through suggestion and other means. He also showed, however, that people
being manipulated play an active role in their manipulation through as-if,
goal-directed enactment (ibid.: 389).
With this in mind, our teams have been designing experiments to test ways
by which expectation and predictive error monitoring can be manipulated in

whence religion?

religious situations. Uffe Schjoedts fMRI experiments on prayer have dem-

onstrated how expectations concerning charismatic authorities influence sub-
jects receptivity to prayer (Schjoedt et al. 2008, 2009, 2011). Else-Marie
Jeginds clinical and field experiments have demonstrated how religious
expectations influence the experience of pain intensity and unpleasantness
(Jegind et al. 2013b). These expectations, furthermore, superseded interfer-
ence caused by injections of the drug naloxone which effectively blocks cen-
tral opioid receptors (Jegind 2012: 44).52
Following up on Michael Persingers temporal lobe experiments (the so-
called God-helmet experiments), which could not be replicated by a Swedish
team,53 Marc Andersen, Uffe Schjoedt, Kristoffer Nielbo and Jesper Srensen
initiated experiments using a non-active helmet. The preliminary results indi-
cate that the RCC team could reproduce the same success rate claimed by
Persinger simply by suggestion and deprivation (Andersen 2012; Andersen et
al. in preparation).54
From a different perspective, Kristoffer Nielbo and Jesper Srensen in a
series of experiments on ritualized behaviour, have demonstrated that ritual-
ized actions systematically affect how humans process actions by hindering
the smooth integration of individual action gestalts (e.g. gripping, lifting)
into causally structured events schemas (e.g. drinking). They hypothesize
that this lack of predictability affects a redirection of attention towards low
features of the actions thereby depleting cognitive resources from other tasks
(Nielbo 2012; Nielbo & Srensen 2011). This, in turn, facilitates not only
representations of rituals as having direct or magical efficacy (Srensen 2006,
2007; Srensen et al. 2006), it also makes them prime candidates for subse-
quent reinterpretation, realigning individual experience of the ritual to cultur-
ally sanctioned religious narratives (Schjoedt et al. 2013).
Religious rituals and social behaviour modulate the neural regulatory
systems. The ritualized use of the body stimulates particular mental states.
Rituals can serve to produce attunement in a group (Konvalinka et al. 2011).
Public rituals can help groups of individuals attune themselves to a common
mental state, whether it is terror, ecstasy or violence (Deeley 2004; Geertz
2010b). Rituals and the modulation of regulatory systems play a significant
role in therapeutic systems around the world (Krippner & Achterberg 2000).
Clinical psychiatrist Wolfgang G. Jilek investigated shamanism and the
Spirit Dance among the Salish on Vancouver Island, and concluded that the
mechanisms of sensory stimulation (i.e. pain stimulation, hypoglycemia and
dehydration, forced hypermotility, temperature stimulation, acoustic stimula-
tion, seclusion and restricted mobility, visual-sensory deprivation, sleep dep-
rivation, kinetic stimulation, and hyperventilation) that produce ASCs have
their underlying metabolic mechanism in endogenous opiates (Jilek 1982a:
33641; cf. Jilek 1982b).

armin w. geertz

There are strong indications of a connection between auto-suggestion and

the endorphins in fire-handler cults,55 trance, and spirit possession;56 however,
hypnotic trance and drum-and-dance trance may depend on two different
physiologies. The former produces what is called faith analgesia and the lat-
ter endorphin-mediated analgesia (R. H. Prince 1982: 41112).
This brings us to the placebo effect. Although this effect tells us a lot about
the mind/brain/body interfaces, we hardly know as yet what processes are
involved. Neurologist Fabrizio Benedetti and his team have argued in their
review of the placebo literature that:

Placebos are not inert substances, as thus far believed. They are
made of words and rituals, symbols, and meanings, and all these
elements are active in shaping the patients brain A real placebo
effect is a psychobiological phenomenon occurring in the patients
brain after the administration of an inert substance, or of a sham
physical treatment such as sham surgery, along with verbal sugges-
tions (or any other cue) of clinical benefit.
(Benedetti et al. 2011: 339)

This effect has been well known throughout human history, and yet, until
recently it has been treated as an irritation factor in clinical tests of pharma-
cological substances. New research has revealed many interesting aspects of
placebo. Benedetti and his team argue that the neural networks responsible
for this effect in pain and Parkinsons disease are (1) the opiodidergic-chole-
cystokinergic-dopaminergic modulatory network in pain and (2) the basal
ganglia circuitry in Parkinsons disease (ibid.), but they emphasize the impor-
tant fact that there are many placebo effects based on differing mechanisms
and therapies for different diseases:

Sometimes it is anxiety that is modulated, at some other times

reward mechanisms are involved, and in some other circumstances
different types of learning, or even genetic variants, may take place
in placebo responsiveness. In this sense, the placebo effect is a
melting pot of neuroscientific concepts and ideas, ranging from
anxiety and reward mechanisms to Pavlovian conditioning and
social learning, and from neurogenetics and neurophysiology to
clinical practice and neuroethics. (Ibid.: 340)

Expectation and learning play a central role in most types of placebo, thus
once again supporting the hypothesis promoted in this essay (cf. Benedetti
2009). Because therapy has played and still plays an important role in reli-
gion, placebo research will be of singular importance to biocultural analyses.

whence religion?

Ted J. Kaptchuk at Harvard Medical School views alternative therapies as

placebo dramas designed to maximize patient expectation modulation. The
efficacy of these methods he called performative efficacy, which is founded
on the power of belief, imagination, symbols, meaning, expectation, per-
suasion, and self-relationship (Kaptchuk 2002: 818). The placebo drama,
Kaptchuk claimed, consists of five elements, namely patient, practitioner,
patient-practitioner interaction, the nature of the illness and the treatment
and setting (ibid.: 818). Alternative medicine, he argued, offers a charged
constellation of expectations that elicits patients magical anticipation
(ibid.: 818; cf. Mattingly 1998).
The brain is embodied in a nervous system that seamlessly connects the
body, brain and mind in one comprehensive system. We can therefore begin
to understand how environmental contexts through the continuous avenues
of primes gain access to our innermost selves in socialization contexts, polit-
ical contexts, religious contexts, etc. We are especially vulnerable to the influ-
ences of peers and authorities. Neuropsychologists Ramsey M. Raafat, Nick
Chater and Chris Frith call this feature herding:

Herding can be broadly defined as the alignment of the thoughts

or behaviours of individuals in a group (herd) through local inter-
actions rather than centralized coordination. In other words, the
apparent central coordination of the herd is an emergent property
of local interactions. (Raafat et al. 2009: 420)

There are many, incredibly complex and subtle transmission mechanisms of

this phenomenon. They range from information cascades, quorum sensing,
chemosensory signalling, priming, imitation, mimicry, contagion, empathy,
vocalizations, postures, movements, emotional manipulations and so on.
Raafat et al. noted that humans herd like other animals in a ripple effect
that influences group dynamics and individual emotions and cognitions. This
effect, they argued:

provides an evolutionary perspective into these behaviours and

suggests that they are crucial for the maintenance of societal norms
Thus, the notion of emotional contagion can be extended to
the broader concept of social contagion: the tendency to automat-
ically mimic and synchronize expressions, vocalizations, postures
and movements with those of another person leading to behav-
ioural convergence. (Ibid.: 425)

This reads like a quote from mile Durkheim, and well it should. Many
of Durkheims central insights, intuitive as they may have been, are being

armin w. geertz

vindicated empirically in contemporary psychological and neurological

research. From the above-mentioned understanding of the brain, religious
ritual can be seen as a set of herding tools that merge individual and collective
minds within the framework of the greater realities of their virtual worlds.
For Durkheim, this process of convergence occurs during the effervescence
produced by hyper-exciting religious rituals.

It is not difficult to imagine that a man in such a state of exalta-

tion should no longer know himself. Feeling possessed and led on
by some sort of external power that makes him think and act dif-
ferently than he normally does, he naturally feels he is no longer
himself. It seems to him that he has become a new being And
because his companions feel transformed in the same way at the
same moment, and express this feeling by their shouts, move-
ments, and bearing, it is as if he was in reality transported into a
special world entirely different from the one in which he ordinar-
ily lives, a special world inhabited by exceptionally intense forces
that invade and transform him It is in these effervescent social
milieux, and indeed from that very effervescence, that the religious
idea seems to have been born. (Durkheim [1912] 1995: 220)57

As I have argued elsewhere, rituals, whether collective or individual, manipu-

late somatic and mental states in order to influence our minds in particular
ways. Especially ritual pageantry draws on a wide variety of driving tech-
niques, which have physical influence on our bodies and brains, and thus
on our minds (Geertz 2010a: 307ff.). Besides dance, song, clapping, postur-
ing, etc., techniques such as mortification, fasting, over-exertion and drugs,
are also used to manipulate the brains expectations and decrease or increase
sensory input. Thus ceremonies involving such techniques tug deeply at the
psychological and somatic foundations of each and every individual and have
the ability to arouse, shape and form emotions and mental states, thus allow-
ing the transfer and sharing of norms and ideals (ibid.: 307). The empirical
evidence for this is growing in several areas of research, such as neurological
and psychological studies of music,58 synchrony59 and dancing.60 The studies
conducted by RCC research teams on firewalking in Spain and on Mauritius
have also been testing the prosocial and effervescent aspects of collective,
religious rituals.61
These ritual techniques also play a fundamental role in healing rituals,
shamanistic sances, magical rituals, etc. with the express purpose of manipu-
lating the bodies and minds of participants, patients and patrons.62 On the
dark side of human behaviour, these techniques used together with careful
manipulation of the brain, as described above, are essential ingredients in

whence religion?

thought control. Kathleen Taylor called it the traitor in your skull (Taylor
2004: x).
After this discussion of the kinds of minds we have, the question now is,
when did religion appear?

When did proto-religion appear?

So, when did religion appear? A crucial problem of course is what we mean
by religion. I have argued elsewhere that religion is social and cultural. It is a
socially imposed hermeneutical device that draws on the cultural competence
of individuals with reference to conceived transempirical powers or beings
(Geertz 1999: 460). How this phenomenon arose can, of course, only be
speculated upon. But whatever the most reasonable scenario might be, reli-
gion cannot have played a merely subsidiary role. The powerful social and
cultural processes of which we have knowledge today could not have been
developed or conceived outside of a religious framework. And the reason for
this, again, recalling Goodenough and Deacon (2003), is that what sets our
ape brain off from the brains of other apes is symbolic competence.63
Because we cannot assume that earlier species, or even the earliest Homo
sapiens, had minds completely like ours, we will have to concede that the first
activities and ideas that we might conceivably recognize as being religious or
spiritual and here I am not talking about theologically correct or coherent
ideas or highly developed priesthoods, shamans or other religious specialists
were what might be called partially formulated superstitions or proverbial
responses, feelings and intuitions. The question is: when did these ideas and
behaviours first appear?64 I will argue here that there are good reasons to
assume that they arose long before Homo sapiens appeared on the scene.
What antecedents would be necessary for a proto-religion to develop? My
bet would be on the following items:

1. Quite a lot of the cognitive and social functions must be present in

order to share normative ideas and behaviours.
2. Certainly self-consciousness (and not just consciousness or intentional-
ity, which all animals have).
3. Self-reflection and identity.
4. Empathy and sympathy.
5. The ability to communicate symbolically.
6. There must be a social group that shares values and norms.
7. There must be social structures for solving common problems (altruism,
alliances, cooperation, etc.).
8. Transgenerational passing on of knowledge (tradition) is essential.

armin w. geertz

9. Authority, control and morality are ever-present.

10. Rituals would be used (like other animals) producing entrainment,
emotional contagion, a sense of identity, and so on.
11. Superstition, supersense and a sense of forces in nature would most
likely be at play.

The evidence presented in this essay indicates that the species evincing these
qualities was Homo erectus, perhaps more specifically Homo heidelbergensis
between 600,000 and 400,000 years ago (Tattersall 2012: 13543). This spe-
cies with its big, strikingly modern brains (with, however, a slightly different
positioning of the prefrontal cortex) with expanded Brocas areas, sophisti-
cated wooden spears, symbolic behaviour and objects, control of fire, primi-
tive huts and superior stone tools, and that left Africa for temperate climates,
must have had proto-religious traits. They were, at any rate, smarter than their
predecessors, according to Ian Tattersall (ibid.: 137). Tattersall did not think
that they evinced symbolic thought processes, claiming that the evidence of
the use of symbolic objects is too sketchy (ibid.: 142). Nevertheless, he was
very positive about their abilities and their amazing technological and cul-
tural ingenuity (ibid.). But if he concedes the possibility of spoken language,
then he must necessarily concede the presence of symbolic thought. I stand
clearly with Terrence Deacon and Merlin Donald on this issue: symbolic
thought led to language and not the other way around. But even without
language, it is difficult to see how their other achievements could happen
without symbolic competence.
It could be argued that Homo ergaster, the inventor of Acheulian lithics,
the cooking species and perhaps speaking a proto-language, evinced qualities
that would allow proto-religious behaviour and thought. That would push the
time-frame back to over 1.5 million years ago. The apparent burial in Sima
de los Huesos at Atapeurco, however, might indicate religious behaviour by
a species after Homo ergaster and before Homo heidelbergensis, although there
is some discussion about what species was involved. They were either Homo
heidelbergensis or most likely the ancestors of Homo heidelbergensis and Homo
neanderthalensis. I think, however, that we are on firmer ground with the big-
brained Homo heidelbergensis. If at some point more is learned thus pushing
the time-frame further back to earlier species, it would not be deleterious to
my hypothesis quite the contrary!

Of what would this proto-religion therefore consist?

It seems clear from the archaeological evidence that death and the dead were
primary objects of symbolic concern. Furthermore, attention to human

whence religion?

fertility may have been important. One could assume that important cul-
tural items and their production were of symbolic value. These and other
creature-related concerns would have been shared by a social unit engaged
in networking with other units exchanging items, goods and females. These
groups may have engaged in ritual dance. My guess, then, would be the fol-
lowing proto-religious behaviour:

1. It would reflect the social group, centred near the fire, with implied
gender-related domains, social status and control.
2. It would also be concerned with social events such as life transition
events, sexuality and female biological cycles and fertility.
3. It would be concerned with death and the dead as well as misfortune
and evil.
4. It would reflect symbolic thought by signs and omens and a sense of
forces behind plants (food, medicinal), animals (game, fish) and natural
phenomena (storms, drought, floods, eruptions, earthquakes).
5. It would involve magical thinking (sympathetic and contagious).
6. It would involve rituals, dance, music (and/or song) and various tech-
niques to manipulate both body and mind.
7. It would involve experiences such as altered states of consciousness
(dreams, fasting and hunger, psychopharmica, self-inflicted pain, etc.).
8. It would be concerned with expressing group feelings, values, traditions
(i.e. tool-making) and narratives (pre-linguistic, mimetic).


In this chapter, I have taken issue with a number of assumptions promoted

by cognitive scientists of religion concerning the origins of religion, cogni-
tion and culture. First, I took issue with the assumption that culture is irrel-
evant and religion even more so in human evolution. I have argued instead
that recent insights from neurobiology indicate that we need to work with
biocultural evolutionary scenarios. Furthermore, the claim that the brain
evolved like a computer into systems of dedicated modules ignores insights
from neurology. In fact, what is special about our brain is its predictive func-
tions and flexibility. Another special feature of the brain is that it is depend-
ent on culture in order to function properly because it has evolved not only
together with culture but because of it. The growth of the brain was appar-
ently also due to the growing complexity of human social structures. A final
key feature of the brain is that it is incomplete in human infants and is first
fully matured after two decades of social and cultural life: the unfinished
animal that becomes completed through a particular culture. Taking these

armin w. geertz

basic features as our starting point, it becomes obvious that mentalistic, indi-
vidualistic and culture-blind theories of human cognition are curious at best.
Looking more closely at the neuropsychological evidence of our brains and
behaviour, I have identified four key human features:

a finely honed social cognition;

a drive to communicate and cooperate;
a self-deceptive brain; and
a superstitious brain prone to unusual mental and/or emotional

In unpacking these features, it became evident not only that we need to

rethink how to understand religious thought and behaviour in terms of the
kinds of brains we have, but we also need to incorporate these insights into
our evolutionary scenarios. With these four features, it is possible to account
for the hows and whys of religion and to relate them to some of the current
evolutionary theories. Thus, I claim, we are intelligent apes that are highly
emotional, easily spooked, very superstitious, extremely sensitive to social
norms and virtual realities and equipped with nervous systems that are vul-
nerable to influence from conspecifics and their symbolic worlds. We and
our brains are constantly predicting and constantly dwelling on the future in
our attempts to navigate social and natural environments. Our brains fill in
quite a bit, but our cultures are also in our brains and around them, filling
in many more things.
These traits are prerequisites for religious behaviour. In looking through
the archaeological records, with all the caveats and disagreements in mind,
and yet, in accordance with an evolutionary theory of slowly emerging steps, I
argue that the grand narrative of the cultural explosion of our own subspecies
has blinded us to the fact that nothing emerges out of nothing. Or, to put it
in more positive terms, our traits and functions and abilities and worlds are
accumulations of prior ones. There can be no doubt, although there are many
who have it, that symbolic behaviour is evident in the archaeological record
before the appearance of Homo sapiens. Therefore, in our search for the origins
of religion, cognition and culture, we will need to look beyond ourselves into
the deep past, even before the rise of the hominins, to understand who we are
and where we came from.
As mentioned, the most likely candidate that evinced proto-religious
behaviour is Homo heidelbergensis, although it would not surprise me if the
evidence will persuade us to look further back in time. I find this thought
comforting. I also find the idea that proto-religious behaviour emerged gradu-
ally, long before the appearance of Homo sapiens, to fit much better with the
Darwinian framework of this essay.

whence religion?


This contribution was originally written for this volume. A version of it was, however,
reprinted in Geertz (2010c) (there were some good jokes in it, which have been dutifully
removed here), way before this volume ever saw the light of day. In preparation for this pub-
lication, it has been substantially revised. The ideas presented here were made possible by the
unique interdisciplinary environment at Aarhus University, especially the MINDLab UNIK
initiative funded by the Danish Ministry of Science, Technology and Innovation, and the
Religion, Cognition and Culture Research Unit (RCC) at the Department of Culture and
Society, Section for the Study of Religion, Aarhus University. My special thanks are extended to
neurologists Merlin Donald, Chris Frith and Uta Frith, my close MINDLab colleagues, neu-
rologist Leif stergaard and anthropologist-biologist Andreas Roepstorff, as well as the young
team of predictive-coding enthusiasts at the RCC (Jesper, Uffe, Ella, Kristoffer and Marc).
Their ideas have had an indelible impact on my thinking. My sincere thanks are extended to
friends and cohorts Joseph Bulbulia and Jesper Srensen whose careful reading of this chapter
has improved it greatly, to Sheila Coulson for her corrections in the section concerning Rhino
Cave and to Chris Frith for his comments on crucial neurological issues. None of the above-
mentioned is to be blamed for my idiosyncrasies.


1. See my reviews in Geertz (2004, 2008b). See also Bulbulia (2004b).

2. This is the key argument put forward by Jonathan Turner and Alexandra Maryanski in
their important book On the Origin of Societies by Natural Selection (2008).
3. Which is why I have persistently argued that it is theism that needs explanation and
not atheism (Geertz & Marksson 2010; Geertz 2012). See also Gervais et al. (2011:
4. See Geertz (2008a, 2010a, 2012).
5. West-Eberhard is clearly aligned with the growing literature on the importance and
functions of epigenesis (cf. Durham 1982; Gottlieb 2007; Jablonka & Lamb 2005;
Oyama 2000; Sol & Goodwin 2000).
6. See my coming essay on Clifford Geertz and the cognitive science of culture in Geertz
(in press).
7. The quoted passage at the end is to Washburn (1959).
8. This is of course opposite to Tooby and Cosmidess claims (Tooby & Cosmides 1992),
which have been characterized by one critic as modularity gone mad (Fodor 1987).
Philosopher of biology and cognitive science Robert C. Richardson argued that evo-
lutionary psychology in this guise is maladapted psychology that is more speculative
than scientific (Richardson 2007: 38). Tooby and Cosmides, by the way, were also very
wrong about Clifford Geertz in their diatribes against what they labelled the standard
social science model. See Geertz (in press) for my essay on this subject.
9. See for instance the growing literature on social and affective neuroscience (Adolphs
2003; Barsalou et al. 2003; Cacioppo et al. 2002; Cacioppo & Berntson 2005; Donald
2001; Kitayama & Park 2010; Losin et al. 2010; Renfrew et al. 2009; Varela et al. 1991;
Vogeley & Roepstorff 2009).
10. Several special issues on cognition and materiality have appeared recently (e.g. Chiao
2010; Knappett & Malafouris 2008; Renfrew & Scarre 1998).

armin w. geertz

11. The sections Fire control and Symbolic objects are loosely based on Geertz (2011b:
415), with substantial additions.
12. Evidence indicates that monkeys, as well as apes, use stones as tools.
13. I am indebted to Sheila Coulson for this information and for her critical reading of this
14. See Geertz (2010a, 2010b).
15. For the neurological details of prediction in mental processes, see the special issue of
Philosophical Transactions of the Royal Society B, volume 364, especially Bar (2009),
Gilbert and Wilson (2009), Mitchell (2009), Moulton and Kosslyn (2009), and
Schacter and Addis (2009).
16. Besides Meltzoff and Moores pioneering work in 1977, just to name a few: Beebe
and Lachmann (1988), Ekman et al. (1983), Ekman et al. (1990), Dissanayake (2000:
19ff.), Donald (2001: 255ff.), Murray and Andrews (2000), Murray and Trevarthen
(1985), Trevarthen (1993) and Trevarthen et al. (1999). Recent research indicates, how-
ever, that more work needs to be done on this issue (see Jones 2009).
17. See Charles A. Zieglers and Benson Salers excellent discussion of the social, neurologi-
cal and genetic mechanisms behind this process (Ziegler & Saler 2012).
18. Jesper Srensen presented an immunology hypothesis to counter-balance the epidemiol-
ogy models by, among others, Boyer and Sperber (Srensen 2004). Henrich seems not
to have been aware of that article.
19. The issue of group selection has been eloquently addressed in several publications by
David Sloan Wilson (2002, 2007, 2008).
20. See Benson Salers excellent chapter on the evolutionary advantages of credulity (Saler
2009: 13346).
21. See Koch (2004) for the technical details of how this happens. Because a good deal
of the literature on consciousness is based on the visual system, Chris Frith has
argued that the experimental literature on the motor system may reveal more. This
literature, he claimed, shows what consciousness is for rather than what it is (Frith
2010: 498).
22. Damisch et al. provide experimental evidence showing that superstition contributes to
performance enhancement (Damisch et al. 2010).
23. Jahoda (1969) is discussed by Vyse (1997: 20ff.). Jahodas own definition, after a number
of pages criticizing other definitions, is not very helpful: the kind of belief and action a
reasonable man in the present-day Western society would regard as being superstitious
(Jahoda 1969: 10). What he actually mentions are behaviours ranging from lucky or
unlucky numbers, days or colours via astrology and other occult systems to witches,
ghosts and sorcerers (ibid.). The consequences of including the last three items in this
catalogue are that a good Christian who really believes that Satan works in the world is
superstitious, just like a Siberian indigene who believes that his shamans tutelary spirits
can heal illnesses or an American Indian who is convinced that he has been bewitched
by an evil sorcerer. These examples I would categorize squarely as religious.
24. In this, I concur with biologists Jan Becks and Wolfgang Forstmeiers Bayesian approach
to superstition, which they argued supports the observation of coincidence and search
for causal patterns, thus improving survival rates (Beck & Forstmeier 2007).
25. Cf. Cohen (2007), Targ et al. (2000), Taves (2009) and Wulff (2000).
26. A brief overview can be found in Kihlstrom et al. (1990).
27. On the role of cultural structures and genetic variation, see Luhrmann (2012).
28. Cf. Bourguignon (1976), Oesterreich (1974) and Smith (2001).
29. See, for instance, texts that would have us believe that the great religious figures in world

whence religion?

history were either epileptics (Devinsky & Lai 2008), schizophrenics (Polimeni & Reiss
2002) or in need of oxygen (Arzy et al. 2005).
30. See also Emma Cohens (2007: 7997) excellent review of the literature on ASC in rela-
tion to possession.
31. Steinberg et al. (1986), Steinberg et al. (1990).
32. Mayer-Gross (1924), Steinberg (1991).
33. Steinberg (1991).
34. Ibid.
35. Steinberg et al. (1986), Steinberg et al. (1990).
36. PTSD is post-traumatic stress disorder.
37. This fact was recently demonstrated in Else-Marie E. Jeginds study of the Thaipasam
ceremony on Mauritius. Her analyses show that reduction in the subjective inten-
sity and unpleasantness of the painful rituals was more evident in participants that
reported dissociative states than those that did not (Jegind et al. 2013b; cf. Seligman
& Kirmayer 2008).
38. See anthropologist Bambi Chapins enlightening study of a Sri Lankan possession priest-
ess (Chapin 2008) placed in the context of the debate between Melford Spiro (1997)
and Gananath Obeyesekere (1990).
39. Uffe Schjoedt and I have argued that much of the neurological research on religious
experiences and on religion and health is religiously motivated and addresses religious
audiences, and is thus of minimal relevance to the scientific study of religion (Geertz
2009; Schjoedt 2009).
40. Keenan et al. (2000), Vogely et al. (1999).
41. Courtney et al. (1998), Vogeley et al. (2001).
42. Damasio (1994).
43. Rylander (1948).
44. Lhermitte (1983), Lhermitte et al. (1986).
45. Norman and Shallice (1986), Shallice and Burgess (1991).
46. Frith and Dolan (1996).
47. Frith and Frith (2001), Povinelli and Preuss (1995) and Stone et al. (1998).
48. Fuster (2000) and Goldman-Rakic (1992).
49. Fuster (1995), Knight and Grabowecky (1999) and Kolb (1984).
50. Posner (1994) and Sarter et al. (2001).
51. Dehaene and Naccache (2001) and Duncan and Owen (2000).
52. Neurologist Fred H. Previc, however, has argued that the ventromedial dopaminergic
systems are active during hyperreligiosity and may even be linked to the evolution of
religion (Previc 2006).
53. See Cook and Persinger (1997) and Persinger (1999); critique by Granqvist et al.
(2005); Persinger and Korens reply (2005); and counter-reply by Larsson et al. (2005).
Granqvist argued that the Persingers results were most likely due to suggestive influ-
ences on the subjects.
54. Colleagues in London and Belfast are working on similar projects. Neuropsychiatrists
Quinton Deeley and Eamonn Walsh have successfully produced automatic writing
in normal subjects. In Belfast, Lauren Swiney is experimenting with suggestion and
thought insertion (Swiney in preparation).
55. Cf. Kane (1982), Rawcliffe (1959: 2916), Holm (1982: 23), Tinterow (1970), Barber
(1958) and R. J. Prince (1968).
56. Cf. Henry (1982) and Walker (1972).
57. See the excellent summary of this process by Chris Shilling and Philip A. Mellor

armin w. geertz

(1998). See also Douglas A. Marshalls interesting sociological model of ritual based on
Durkheim (Marshall 2002).
58. Avanzini et al. (2003), Brown and Volgsten (2006), Garza Villarreal (2012), Green et
al. (2012), Hauser and McDermott (2003), Janata and Grafton (2003), Levitin (2007),
Vuust et al. (2011) and Wallin et al. (2000).
59. Cohen et al. (2010), Konvalinka et al. (2010) and Wiltermuth and Heath (2009).
60. Berrol (2006), Brown and Parsons (2008), Jola et al. (2011), Winters (2008) and
Worthen-Chaudhari (2011).
61. Konvalinka et al. (2011) and Xygalatas et al. (2011, in 2013a, in 2013b).
62. Barsalou et al. (2003), Barsalou et al. (2005), Frank (1977), Jegind (2012), Jegind et
al. (2013a), Kaptchuk (2002), Niedenthal et al. (2005) and Petrovic et al. (2002, 2005).
63. See Deacon (1997) as well.
64. I have introduced my views in several papers, such as Geertz (2008a, 2010a, 2010b),
but more recently in systematic fashion in Danish: Geertz (2011b).

Adolphs, R. 2003. Cognitive Neuroscience of Human Social Behaviour. Nature Reviews
Neuroscience 4: 16578.
Andersen, M. N. 2012. The God Helmet Reloaded: En ny eksperimentel tilgang til religions-
psykologiske studie af mystiske, religise og spirituelle oplevelser [A New Experimental
Approach to the Study of Mystical, Religious and Spiritual Experiences]. Masters thesis,
Aarhus University.
Andersen, M. N., U. Schjoedt, K. L. Nielbo & J. Srensen in preparation. Predicting God:
A New Theoretical and Experimental Framework for the Study of Mystical, Religious and
Spiritual Experiences.
Andrews, P. & Y. Fernandez-Jalvo 1997. Surface Modifications of the Sima de los Huesos
Fossil Humans. Journal of Human Evolution 33: 191218.
Anisimov, A. F. 1963. The Shamans Tent of the Evenks and the Origin of the Shamanistic
Rite. In Studies in Siberian Shamanism, H. N. Michael (ed.), 84107. Toronto: University
of Toronto Press.
Arsuaga, J.-L., I. Martnez, A. Gracia, J. M. Carretero, C. Lorenzo, N. Garca & A. I. Ortega
1997. Sima de los Huesos (Sierra de Atapuerca, Spain); The Site. Journal of Human Evolu-
tion 33: 10928.
Arzy, S., M. Idel, T. Landis & O. Blanke 2005. Why Revelations Have Occurred on Moun-
tains? Linking Mystical Experiences and Cognitive Neuroscience. Medical Hypotheses 65:
Asfaw, B., T. White, O. Lovejoy, B. Latimer, S. Simpson & G. Suwa. 1999. Astralopithecus
garhi: A New Species of Early Hominid from Ethiopia. Science 284: 62935.
Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. Oxford: Oxford
University Press.
Atran, S. 2010. Talking to the Enemy: Faith, Brotherhood, and the (Un)making of Terrorists. New
York: HarperCollins.
Avanzini, G., C. Faienza, D. Minciacchi, L. Lopez & M. Majno 2003. The Neurosciences and
Music. Annals of the New York Academy of Sciences 999. New York: New York Academy
of Sciences.
Bar, M. 2009. The Proactive Brain: Memory for Predictions. Philosophical Transactions of the
Royal Society B 364: 123543.

whence religion?

Barber, T. X. 1958. Toward a Theory of Pain: Relief in Chronic Pain by Prefrontal Leucotomy,
Opiates, Placebos, and Hypnosis. Psychological Bulletin 56: 43060.
Barham, L. S. 2001. Central Africa and the Emergence of Regional Identity in the Middle
Pleistocene. In Human Roots: Africa and Asia in the Middle Pleistocene, L. S. Barham &
K.Robson-Brown (eds), 6580. Bristol: Western Academic & Specialist Press.
Barsalou, L. W., P. M. Niedenthal, A. K. Barbey & J. A. Ruppert 2003. Social Embodiment.
The Psychology of Learning and Motivation: Advances in Research and Theory 43: 4392.
Barsalou, L. W., A. K. Barbey, W. K. Simmons & A. Santos 2005. Embodiment in Religious
Knowledge. Journal of Cognition and Culture 5(12): 1457.
Bar-Yosef, O. 1998. On the Nature of Transitions: The Middle to Upper Palaeolithic and the
Neolithic Revolution. Cambridge Archaeological Journal 8: 14163.
Bar-Yosef, O. & B. Vandermeersch 1993. Modern Humans in the Levant. Scientific American
4: 94100.
Bar-Yosef, O., B. Vandermeersch, B. Arensburg, P. Goldberg, H. Laville, L. Meignen, Y. Rak,
E. Tchernov & A.-M. Tillier 1986. New Data on the Origin of Modern Man in the
Levant. Current Anthropology 27: 634.
Bateson, G. & M. Mead 1942. Balinese Character: A Photographic Analysis. New York: New
York Academy of Sciences.
Beck, J. & W. Forstmeier 2007. Superstition and Belief as Inevitable By-products of an
Adaptive Learning Strategy. Human Nature 18(1): 3546.
Bednarik, R. G. 1998. The Australopithecine Cobble from Makapansgat, South Africa.
South African Archaeological Bulletin 53: 48.
Bednarik, R. G. 2002. An Outline of Middle Pleistocene Palaeoart. Purakala 13: 3944.
Bednarik, R. G. 2003a. The Earliest Evidence of Palaeoart. Rock Art Research 20(2): 89135.
Bednarik, R. G. 2003b. A Figurine from the African Acheulian. Current Anthropology 44(3):
Beebe, B. & F. M. Lachmann 1988. The Contribution of Mother-Infant Mutual Influence to
the Origins of Self- and Object-Representations. Psychoanalytic Psychology 5(4): 30537.
Bellomo, R. V. 1994a. Methods of Determining Early Hominid Behavioural Activities
Associated with the Controlled Use of Fire at FxJj20 Main, Koobi-Fora, Kenya. Journal of
Human Evolution 27(13): 17395.
Bellomo, R. V. 1994b. Early Pleistocene Fire Technology in Northern Kenya. In Society,
Culture, and Technology in Africa, S. T. Childs (ed.), 1628. Philadelphia, PA: University
of Pennsylvania.
Benedetti, F. 2009. Placebo Effects: Understanding the Mechanisms in Health and Disease,
Oxford: Oxford University Press.
Benedetti, F., E. Carlino & A. Pollo 2011. How Placebos Change the Patients Brain. Neuro-
psychopharmacology 36: 33954.
Bering, J. M. 2006. The Folk Psychology of Souls. Behavioral and Brain Sciences 29: 45398.
Bering, J. M. 2011. The Belief Instinct: The Psychology of Souls, Destiny, and the Meaning of Life.
New York: W. W. Norton.
Berna, F., P. Goldberga, L. K. Horwitzc, J. Brink, S. Holt, M. Bamford & M. Chazang 2012.
Microstratigraphic Evidence of in situ Fire in the Acheulean Strata of Wonderwerk Cave,
Northern Cape Province, South Africa. Proceedings of the National Academy of Sciences
Plus E121520.
Bernstein, E. M. & F. W. Putnam 1986. Development, Reliability, and Validity of a Dis-
sociation Scale. Journal of Nervous and Mental Disease 174: 72735.
Berrol, C. F. 2006. Neuroscience Meets Dance/Movement Therapy: Mirror Neurons, the
Therapeutic Process and Empathy. The Arts in Psychotherapy 33(4): 30215.

armin w. geertz

Bischoff, J. L., J. A. Fitzpatrick, L. Lon, J. L. Arsuaga, C. Falgueres, J. J. Bahain & T. Bullen

1997. Geology and Preliminary Dating of the Hominid-Bearing Sedimentary Fill of the
Sima de los Huesos Chamber, Cueva Mayor of the Sierra de Atapuerca, Burgos, Spain.
Journal of Human Evolution 33: 12954.
Bogoras, W. 1907. The Jesup North Pacific Expedition, vol. 7, part 2: The Chukchee Religion,
F.Boas (ed.). Leiden: Brill.
Bourguignon, E. 1976. Possession. San Francisco, CA: Chandler & Sharp.
Boyer, P. 2000. Functional Origins of Religious Concepts: Ontological and Strategic Selection
in Evolved Minds. Journal of the Royal Anthropological Institute 6(2): 195214.
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Boyer, P. & P. Linard 2006. Why Ritualized Behavior? Precaution Systems and Action
Parsing in Developmental, Pathological and Cultural Rituals. Behavioral and Brain Sciences
29: 595650.
Brooks, A. S., P. E. Hare, J. E. Kokis, G. H. Miller, R. D. Ernst & F. Wendorf 1990. Dating
Pleistocene Archeological Sites by Protein Diagenesis in Ostrich Eggshell. Science, New
Series, 248(4951): 6064.
Brown, S. & L. M. Parsons 2008. The Neuroscience of Dance. Scientific American 299(1):
Brown, S. & U. Volgsten (eds) 2006. Music and Manipulation: On the Social Uses and Social
Control of Music. New York: Berghahn Books.
Bulbulia, J. 2004a. Religious Costs as Adaptations that Signal Altruistic Intention. Evolution
and Cognition 10(1): 1942.
Bulbulia, J. 2004b. The Cognitive and Evolutionary Psychology of Religion. Biology and
Philosophy 19: 65586.
Bulbulia, J. 2008. Meme Infection or Religious Niche Construction? An Adaptationist
Alternative to the Cultural Maladaptationist Hypothesis. Method and Theory in the Study
of Religion 20: 67107.
Bulbulia, J. 2009. Religiosity as Mental Time Travel: Cognitive Adaptations for Religious
Behavior. In The Believing Primate: Scientific, Philosophical and Theological Perspectives on
the Evolution of Religion, J. Schloss & M. Murray (eds), 4475. Oxford: Oxford University
Bulbulia, J & R. Sosis 2011. Signalling Theory and the Evolution of Religious Cooperation.
Religion 41(3): 36388.
Burton, F. D. 2009. Fire: The Spark That Ignited Human Evolution. Albuquerque, NM:
University of New Mexico Press.
Burton, R. A. 2008. On Being Certain: Believing You Are Right Even When Youre Not. New
York: St Martins Press.
Cacioppo, J. T. & G. G. Berntson (eds) 2005. Social Neuroscience: Key Readings. Hove:
Psychology Press.
Cacioppo, J. T., G. G. Berntson, R. Adolphs, C. S. Carter, R. J. Davidson, M. K. McClintock,
B. S. McEwen, M. J. Meaney, D. L. Schacter, E. M. Sternberg, S. S. Suomi & S. E. Taylor
(eds) 2002. Foundations in Social Neuroscience. Cambridge, MA: MIT Press.
Cahn, B. R. & J. Polich 2006. Meditation States and Traits: EEG, ERP, and Neuroimaging
Studies. Psychological Bulletin 132(2): 180211.
Chapin, B. L. 2008. Transforming Possession: Josephine and the Work of Culture. Ethos
36(2): 22045.
Chen, D. 2010. Club Goods and Group Identity: Evidence from Islamic Resurgence during
the Indonesian Financial Crisis. Journal of Political Economy 118: 300354.

whence religion?

Chiao, J. Y. 2010. At the Frontier of Cultural Neuroscience: Introduction to the Special

Issue. Social Cognitive and Affective Neuroscience 5(23): 10910.
Clark, J. D., Y. Beyene, G. WoldeGabriel, W. K. Kart, P. R. Renne, H. Gilbert, A. Defleur,
G. Suwa, S. Katoh, K. R. Ludwig, J.-R. Boisserie, B. Asfaw & T. D. White 2003.
Stratigraphic, Chronological and Behavioural Contexts of Pleistocene Homo sapiens from
Middle Awash, Ethiopia. Nature 423: 74752.
Cohen, E. 2007. The Mind Possessed: The Cognition of Spirit Possession in an Afro-Brazilian
Religious Tradition. Oxford: Oxford University Press.
Cohen, E. E. A., R. Ejsmond-Frey, N. Knight & R. I. M. Dunbar 2010. Rowers High:
Behavioural Synchrony Is Correlated with Elevated Pain Thresholds. Biology Letters 6(1):
Cook, C. M. & M. A. Persinger 1997. Experimental Induction of the Sensed Presence in
Normal Subjects and an Exceptional Subject. Perceptual Motor Skills 85(2): 68393.
Coulson, S., S. Staurset & N. Walker 2011. Ritualized Behavior in the Middle Stone Age:
Evidence From Rhino Cave, Tsodilo Hills, Botswana. PaleoAnthropology 2011: 1861.
Courtney, S. M., L. Petit, J. V. Haxby & L. G. Ungerleider 1998. The Role of Prefrontal
Cortex in Working Memory: Examining on the Contents of Consciousness. Philosophical
Transactions of the Royal Society of London Series B 353: 181928.
Cox, B. A. 1970. What Is Hopi Gossip About? Information Management and Hopi Factions.
Man 5(1): 6898.
Cozolino, L. J. 2002. The Neuroscience of Psychotherapy: Building and Rebuilding the Human
Brain, New York: W. W. Norton.
Damasio, A. R. 1994. Descartes Error: Emotion, Reason, and the Human Brain. New York:
Avon Books.
Damisch, L., B. Stoberock & T. Mussweiler 2010. Keep Your Fingers Crossed! How
Superstition Improves Performance. Psychological Science 2010: 17.
Dart, R. A. 1974. The Waterworn Australopithecine Pebble of Many Faces from Makapansgat.
South African Journal of Science 70: 1679.
Deacon, T. W. 1997. The Symbolic Species: The Co-evolution of Language and the Human Brain.
London: Allen Lane.
Deacon, T. W. 2003. Multilevel Selection in a Complex Adaptive System: The Problem of
Language Origins. In Evolution and Learning: The Baldwin Effect Reconsidered, B. H.
Weber & D. J. Depew (eds), 81106. Cambridge, MA: MIT Press.
Deeley, P. Q. 2004. The Religious Brain: Turning Ideas Into Convictions. Anthropology and
Medicine 11(3): 24567.
Dehaene, S. & L. Naccache 2001. Towards a Cognitive Science of Consciousness: Basic
Evidence and a Workspace Framework. Cognition 79: 137.
de Heinzelin, J., J. D. Clark, T. White, W. Hart, P. Renne, G. WoldeGabriel, Y. Beyene &
E. Vrba 1999. Environment and Behavior of 2.5-Million-Year-Old Bouri Hominids.
Science 284(5414): 6259.
DeLoache, J. & A. Gottlieb (eds) 2000. A World of Babies: Imagined Childcare Guides for Seven
Societies. Cambridge: Cambridge University Press.
Devereux, G. 1956. Normal and Abnormal: the Key Problem of Psychiatric Anthropology.
In Some Uses of Anthropology: Theoretical and Applied, J. B. Casagrande & T. Gladwin (eds),
2348. Washington, DC: Anthropological Society of Washington.
Devinsky, O. & G. Lai 2008. Spirituality and Religion in Epilepsy. Epilepsy and Behavior
12: 63643.
Dietrich, A. 2003. Functional Neuroanatomy of Altered States of Consciousness: The
Transient Hypofrontality Hypothesis. Consciousness and Cognition 12: 23156.

armin w. geertz

Dissanayake, E. 2000. Art and Intimacy: How the Arts Began. Seattle, WA: University of
Washington Press.
Donald, M. 2001. A Mind So Rare: The Evolution of Human Consciousness. New York: W. W.
Dunbar, R. 1996. Grooming, Gossip and the Evolution of Language. London: Faber & Faber.
Duncan, J. & A. M. Owen 2000. Common Regions of the Human Frontal Lobe Recruited
by Diverse Cognitive Demands. Trends in Neuroscience 23: 47583.
Durham, W. H. 1982. Interactions of Genetic and Cultural Evolution: Models and Examples.
Human Ecology 10(3): 289323.
Durkheim, . [1912] 1995. The Elementary Forms of Religious Life, K. E. Fields (trans.). New
York: Free Press.
Ekman, P., R. W. Levenson & W. V. Friesen 1983. Autonomic Nervous System Activity
Distinguishes among Emotions. Science 218: 120810.
Ekman, P., R. J. Davidson & W. V. Friesen 1990. Emotional Expression and Brain Physiology:
II. The Duchenne Smile. Journal of Personality and Social Psychology 58: 34253.
Fine, C. 2006. A Mind of Its Own: How Your Brain Distorts and Deceives. New York: W. W.
Fodor, J. 1987. Modules, Frames, Fridgeons, Sleeping Dogs and the Music of the Spheres.
In Modularity in Knowledge Representation and Natural-Language Understanding, J. Garfield
(ed.), 2636. Cambridge, MA: MIT Press.
Frank, J. D. 1977. Persuasion and Healing: A Comparative Study of Psychotherapy, Revised
Edition, 2nd edn, Schocken Paperbacks on Psychology. New York: Schocken Books.
Freeman, W. J. 1999. How Brains Make Up Their Minds. London: Weidenfeld & Nicolson.
Frith, C. D. 2007. Making Up the Mind: How the Brain Creates Our Mental World. Oxford:
Blackwell Publishing.
Frith, C. D. 2010. What Is Consciousness for? Pragmatics and Cognition 18(3): 497551.
Frith, C. D. 2012. The Role of Metacognition in Human Social Interactions. Philosophical
Transactions of the Royal Society B 367: 221323.
Frith, C. D. & R. Dolan 1996. The Role of the Prefrontal Cortex in Higher Cognitive Func-
tions. Cognitive Brain Research 5: 17581.
Frith, C. D. & U. Frith 2001. Cognitive Psychology: A Biological Basis. Science 286: 16925.
Frith, C. D. & U. Frith 2007. Social Cognition in Humans. Current Biology 17: R72432.
Frith, U. & C. D. Frith 2010. The Social Brain: Allowing Humans to Boldly Go Where No
Other Species Has Been. Philosophical Transactions of the Royal Society B 365: 16576.
Fuster, J. M. 1995. Temporal ProcessingStructure and Function of the Human Prefrontal
Cortex. Annals of the New York Academy of Sciences 769: 17381.
Fuster, J. M. 2000. Executive Frontal Functions. Experimental Brain Research 133: 6670.
Garrigan, D. & M. F. Hammer 2006. Reconstructing Human Origins in the Genomic Era.
Nature Reviews Genetics 7: 66980.
Garza Villarreal, E. A., E. Brattico, L. Vase, L. stergaard & P. Vuust 2012. Superior Analgesic
Effects of Mental Arithmetic versus Unfamiliar Music and Sounds: The Role of Emotional
Impact and Cognitive Styles. PLoS One 7(1): 18.
Geertz, A. W. 1999. Definition as Analytical Strategy in the Study of Religion. Historical
Reflections/Rflexions Historiques 25(3): 44575.
Geertz, A. W. 2004. Cognitive Approaches to the Study of Religion. In New Approaches to
the Study of Religion, Volume 2: Textual, Comparative, Sociological, and Cognitive Approaches,
P.Antes, A. W. Geertz & R. R. Warne (eds), 34799. Berlin: Walter de Gruyter.
Geertz, A. W. 2008a. From Apes to Devils and Angels: Comparing Scenarios on the Evolution
of Religion. In The Evolution of Religion: Studies, Theories, & Critiques, J. Bulbulia, R. Sosis,

whence religion?

E. Harris, R. Genet, C. Genet & K. Wyman (eds), 4349. Santa Margarita, CA: Collins
Foundation Press.
Geertz, A. W. 2008b. Religion and Cognition: A Crisis in the Academic Study of Religion?
Bulletin of the Council of Societies for the Study of Religion 37(4): 915.
Geertz, A. W. 2009. When Cognitive Scientists Become Religious, Science is in Trouble: On
Neurotheology from a Philosophy of Science Perspective. Religion 39: 31924.
Geertz, A. W. 2010a. Brain, Body and Culture: A Biocultural Theory of Religion. Method
and Theory in the Study of Religion 22(4): 30421.
Geertz, A. W. 2010b. Too Much Mind and Not Enough Brain, Body and Culture: On What
Needs to Be Done in the Cognitive Science of Religion. Historia Religionum 2: 2137.
Geertz, A. W. 2010c. Reflections on the Origins of Religious Thought and Behavior. In
Chasing Down Religion: In the Sights of History and the Cognitive Sciences Essays in Honor
of Luther H. Martin, P. Pachis & D. Wiebe (eds), 97114. Thessaloniki: Barbounakis.
Geertz, A. W. 2011a. Religious Narrative, Cognition and Culture: Approaches and Defini-
tions. In Religious Narrative, Cognition and Culture: Image and Word in the Mind of Narra-
tive, A. W. Geertz & J. S. Jensen (eds), 929. Sheffield: Equinox.
Geertz, A. W. 2011b. Om religion og evolution: En introduktion [On Religion and
Evolution: An Introduction]. Religionsvidenskabeligt Tidsskrift 56: 2360.
Geertz, A. W. 2011c. Hopi Indian Witchcraft and Healing: On Good, Evil, and Gossip.
American Indian Quarterly 35(3): 37293.
Geertz, A. W. 2012. Why Should Atheists Be for Anything? On the Collective Idiosyncracies
and Illusions of Cognitive Scientists of Religion. Religion, Brain and Behavior 2(1): 735.
Geertz, A. W. in press. The Meaningful Brain: Clifford Geertz and the Cognitive Science of
Culture. In Mental Culture: Towards a Cognitive Science of Religion, D. Xygalatas & W. W.
McCorkle Jr (eds). Sheffield: Equinox.
Geertz, A. W. & G. I. Marksson 2010. Religion Is Natural, Atheism Is Not: On Why
Everybody Is Both Right and Wrong. Religion 40: 15265.
Geertz, C. [1962] 1973. The Growth of Culture and the Evolution of Mind. In his The
Interpretation of Cultures: Selected Essays, 5583. New York: Basic Books.
Geertz, C. [1966a] 1973. The Impact of the Concept of Culture on the Concept of Man. In
his The Interpretation of Cultures: Selected Essays, 3354. New York: Basic Books.
Geertz, C. [1966b] 1973. Religion as a Cultural System. In his The Interpretation of Cultures:
Selected Essays, 87125. New York: Basic Books.
Gervais, W. M., A. K. Willard, A. Norenzayan & J. Henrich 2011. The Cultural Transmission
of Faith: Why Innate Intuitions Are Necessary, But Insufficient, to Explain Religious
Belief . Religion 41(3): 389410.
Gilbert, D. T. & T. D. Wilson 2009. Why the Brain Talks to Itself: Sources of Error in
Emotional Prediction. Philosophical Transactions of the Royal Society B 364: 133541.
Goffman, E. 1959. Presentation of Self in Everyday Life. Garden City, NY: Doubleday.
Goldman-Rakic, P. S. 1992. Working Memory and the Mind. Scientific American 261:
Goodenough, U. & T. W. Deacon 2003. From Biology to Consciousness to Morality. Zygon:
Journal of Religion and Science 38(4): 81314.
Goren-Inbar, N., N. Alperson, M. E. Kisleve, O. Simchoni, Y. Melamed, A. Ben-Nun &
E.Werker 2004. Evidence of Hominin Control of Fire at Gesher Benot Yaaqov, Israel.
Science 304: 7257.
Gottlieb, G. 2007. Probabilistic Epigenesis. Developmental Science 10(1): 111.
Granqvist, P., M. Fredrikson, P. Unge, A. Hagenfeldt, S. Valind, D. Larhammar & M. Larsson
2005. Sensed Presence and Mystical Experiences Are Predicted by Suggestibility, Not by

armin w. geertz

the Application of Transcranial Weak Complex Magnetic Fields. Neuroscience Letters 379:
Green, A. C., K. B. Brentsen, H. Stdkilde-Jrgensen, A. Roepstorff & P. Vuust 2012.
Listen, Learn, Like! Dorsolateral Prefrontal Cortex Involved in the Mere Exposure Effect
in Music. Neurology Research International 2012: 111.
Hauser, M. D. & J. McDermott 2003. The Evolution of the Music Faculty: A Comparative
Perspective. Nature Neuroscience 6(7): 6638.
Henrich, J. 2009. The Evolution of Costly Displays, Cooperation and Religion: Credibility
Enhancing Displays and Their Implications for Cultural Evolution. Evolution and Human
Behavior 30: 24460.
Henrich, J. & F. J. Gil-White 2001. The Evolution of Prestige: Freely Conferred Deference as
a Mechanism for Enhancing the Benefits of Cultural Transmission. Evolution and Human
Behavior 22: 16596.
Henry, J. L. 1982. Possible Involvement of Endorphins in Altered States of Consciousness.
Ethos 10(4): 394408.
Hirstein, W. 2005. Brain Fiction: Self-Deception and the Riddle of Confabulation. Cambridge,
MA: MIT Press.
Hoberg, E. P., N. L. Alkire, A. de Queiroz & A. Jones 2001. Out of Africa: Origins of the
Taenia Tapeworms in Humans. Proceedings of the Royal Society B 268: 7817.
Hobson, J. A. 2001. The Dream Drugstore. Cambridge, MA: MIT Press.
Holm, N. G. 1982. Ecstasy Research in the 20th Century: An Introduction. In Religious
Ecstasy: Based on Papers Read at the Symposium on Religious Ecstasy Held at bo, Finland, on
the 26th28th of August 1981, N. G. Holm (ed.), 726. Stockholm: Almqvist & Wiksell.
Hood, B. M. 2009. SuperSense: Why We Believe in the Unbelievable. New York: HarperOne.
Hood, B. M., K. Donnelly, U. Leonards & P. Bloom 2010. Implicit Voodoo: Electrodermal
Activity Reveals a Susceptibility to Sympathetic Magic. Journal of Cognition and Culture
10: 3919.
Humphrey, N. [1986] 2002. The Inner Eye. Oxford: Oxford University Press.
Jablonka, E. & M. J. Lamb 2005. Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral,
and Symbolic Variation in the History of Life. Cambridge, MA: MIT Press.
Jahoda, G. 1969. The Psychology of Superstition. London: Allen Lane.
Janata, P. & S. T. Grafton 2003. Swinging in the Brain: Shared Neural Substrates for Behaviors
Related to Sequencing and Music. Nature Neuroscience 6(7): 6827.
Jegind, E.-M. Elmholdt 2012. Pain and Coping in the Religious Mind. PhD dissertation,
Aarhus University.
Jegind, E.-M. Elmholdt, L. Vase, J. Charles Skewes, A. Juhl Terkelsen, J. Hansen, A. W. Geertz,
A. Roepstorff & T. S. Jensen 2013a. Expectations Contribute to Reduced Pain Levels
During Prayer in Highly Religious Participants. Journal of Behavioral Medicine 36: 41326.
Jegind, E.-M. Elmholdt, L. Vase, J. Jegind, & A. W. Geertz 2013b. Pain and Sacrifice:
Experience and Modulation of Pain in a Religious Piercing Ritual. International Journal
for the Psychology of Religion 23(3): 17187.
Jilek, W. G. 1982a. Altered States of Consciousness in North American Indian Ceremonials.
Ethos 10(4): 32643.
Jilek, W. G. 1982b. Indian Healing: Shamanic Ceremonialism in the Pacific Northwest Today.
Surrey, BC: Hancock House.
Johnson, C. R. & S. McBrearty 2010. 500,000 Year Old Blades from the Kapthurin
Formation, Kenya. Journal of Human Evolution 58: 193200.
Jola, C., S. Ehrenberg & D. Reynolds 2011. The Experience of Watching Dance: Phenomen-
ological-Neuroscience Duets. Phenomenology and the Cognitive Sciences 11(1): 1737.

whence religion?

Jones, S. S. 2009. The Development of Imitation in Infancy. Philosophical Transactions of the

Royal Society B 364(1528): 232535.
Kane, S. M. 1982. Holiness Ritual Fire Handling: Ethnographic and Psychophysiological
Considerations. Ethos 10(4): 36984.
Kaptchuk, T. J. 2002. The Placebo Effect in Alternative Medicine: Can the Performance of a
Healing Ritual Have Clinical Significance?. Annals of Internal Medicine 136(11): 81725.
Karkanas, P., R. Shahack-Gross, A. Ayalon, M. Bar-Matthews, R. Barkai, A. Frumkin, A.
Gopher, M. C. Stiner 2007. Evidence for Habitual Use of Fire at the End of the Lower
Paleolithic: Site-Formation Processes at Qesem Cave, Israel. Journal of Human Evolution
53: 197212.
Keenan, J. P., M. A. Wheeler, G. G. Gallup & A. Pascual-Leone 2000. Self-Recognition and
the Right Prefrontal Cortex. Trends in Cognitive Sciences 4: 33844.
Kihlstrom, J. F. 1994. Altering States of Consciousness. In Learning, Remembering, Believing:
Enhancing Human Performance, D. Druckman & R. A. Bjork (eds), 20748, 35164.
Washington, DC: National Academies Press.
Kihlstrom, J. F., D. J. Tataryn & I. P. Hoyt 1990. Dissociative Disorders. In Comprehensive
Handbook of Psychopathology, 2nd edn, P. J. Sutker & H. E. Adams (eds), 20334. New
York: Plenum (unedited manuscript available at http://socrates.berkeley.edu/~kihlstrm/
Kitayama, S. & J. Park 2010. Cultural Neuroscience of the Self: Understanding the Social
Grounding of the Brain. Social Cognitive and Affective Neuroscience 5(23): 11129.
Knappett, C. & L. Malafouris (eds) 2008. Material Agency: Towards a Non-Anthropocentric
Approach. New York: Springer.
Knight, R. T. & M. Grabowecky 1999. Prefrontal Cortex, Time, and Consciousness. In
The Cognitive Neurosciences, 2nd edn, M. S. Gazzaniga (ed.), 131937. Cambridge, MA:
MIT Press.
Koch, C. 2004. The Quest for Consciousness: A Neurobiological Approach. Englewood, NJ:
Kolb, B. 1984. Functions of the Prefrontal Cortex in the Rat: A Comparative View. Brain
Research Review 8: 6598.
Koltko-Rivera, M. E. 2004. The Psychology of Worldviews. Review of General Psychology
8(1): 358.
Konvalinka, I., P. Vuust, A. Roepstorff & C. D. Frith 2010. Follow You, Follow Me: Cont-
inuous Mutual Prediction and Adaptation in Joint Tapping. Quarterly Journal of Experi-
mental Psychology 63(11): 222030.
Konvalinka, I., D. Xygalatas, J. Bulbulia, U. Schjdt, E.-M. Jegind, S. Wallot, G. Van Orden
& A. Roepstorff 2011. Synchronized Arousal Between Performers and Related Spectators
in a Fire-Walking Ritual. Proceedings of the National Academy of Sciences 108(20): 851419.
Krader, L. 1954. Buryat Religion and Society. Southwestern Journal of Anthropology 10:
Krippner, S. & J. Achterberg 2000. Anomalous Healing Experiences. In Varieties of Anomalous
Experience: Examining the Scientific Evidence, E. Cardea, S. J. Lynn & S. Krippner (eds),
35395. Washington, DC: American Psychological Association.
Langness, L. 1965. Hysterical Psychosis in the New Guinea Highlands: a Bena-Bena example.
Psychiatry 28: 25877.
Larsson, M., D. Larhammar, M. Fredrikson & P. Granqvist 2005. Reply to M. A. Persingers
and S. A. Korens Response to Granqvist et al. Sensed Presence and Mystical Experiences
Are Predicted by Suggestibility, Not by the Application of Transcranial Weak Complex
Magnetic Fields. Neuroscience Letters 380: 34850.

armin w. geertz

Lepre, C. J., H. Roche, D. V. Kent, S. Harmand, R. L. Quinn, J.-P. Brugal, P.-J. Texier,
A.Lenoble & C. S. Feibel 2011. An Earlier Origin for the Acheulian. Nature 477: 825.
Levitin, D. J. 2007. This Is Your Brain on Music: The Science of a Human Obsession. London: Plume.
Lewis, I. M. [1971] 1989. Ecstatic Religion: A Study of Shamanism and Spirit Possession, 2nd
edn. New York: Routledge.
Lewis, I. M. 1986. Religion in Context: Cults and Charisma. Cambridge: Cambridge University
Lhermitte, F. 1983. Utilization Behaviour and Its Relation to Lesions of the Frontal Lobes.
Brain 106: 23755.
Lhermitte, F., B. Pillon & M. Serdaru 1986. Human Autonomy and the Frontal Lobes, PartI:
Imitation and Utilization Behavior; A Neuropsychological Study of 75 Patients. Annals of
Neurology 19: 32634.
Linard, P. & P. Boyer 2006 Whence Collective Rituals? A Cultural Selection Model of
Ritualized Behavior. American Anthropologist 108(4): 81427.
Linton, R. 1956. Culture and Mental Disorders. Springfield, IL: Charles C. Thomas.
Losin, E. A., M. Dapretto & M. Iacoboni. 2010. Culture and Neuroscience: Additive or
Synergistic?. Social Cognitive and Affective Neuroscience 5(23): 14858.
Luhrmann, T. M. 2011. Hallucinations and Sensory Overrides. Annual Review of Anthro-
pology 40: 7185.
Luhrmann, T. M. 2012. When God Talks Back: Understanding the American Evangelical
Relationship with God. New York: Knopf.
Macknik, S. L., S. Martinez-Conde & S. Blakeslee 2010. Sleights of Mind: What the Neuroscience
of Magic Reveals about Our Everyday Deceptions. New York: Henry Holt.
Malafouris, L. 2010. The Brain-Artefact Interface (BAI): A Challenge for Archaeology and
Cultural Neuroscience. Social Cognitive and Affective Neuroscience 5(23): 26473.
Mandler, J. M. 1988. How to Build a Baby: On the Development of an Accessible Repres-
entational System. Cognitive Development 3: 11336.
Mandler, J. M. 1992. How to Build a Baby: II. Conceptual Primitives. Psychological Review
99(4): 587604.
Mandler, J. M. 2005. How to Build a Baby: III. Image Schemas and the Transition to Verbal
Thought. In From Perception to Meaning: Image Schemas in Cognitive Linguistics, B. Hampe
& J. E. Grady (eds), 13763. Berlin: Mouton de Gruyter.
Marshall, D. A. 2002. Behavior, Belonging, and Belief: A Theory of Ritual Practice. Socio-
logical Theory 20(3): 36080.
Mattingly, C. 1998. Healing Dramas and Clinical Plots: The Narrative Structure of Experience,
Cambridge: Cambridge University Press.
Mayer-Gross, W. 1924. Selbstschilderungen der Verwirrtheit: die oneirode Erlebnisform, Berlin:
Springer Verlag.
McBrearty, S. & A. S. Brooks 2000. The Revolution That Wasnt: A New Interpretation of the
Origin of Modern Human Behavior. Journal of Human Evolution 39: 453563.
McPherron, S. P., Z. Alemseged, C. W. Marean, J. G. Wynn, D. Reed, D. Geraads, R. Bobe
& H. A. Barat 2010. Evidence for Stone-Tool-Assisted Consumption of Animal Tissues
before 3.39 Million Years Ago at Dikika, Ethiopia. Nature 466: 85760.
Meltzoff, A. N. & M. K. Moore 1977. Explaining Facial Imitation: A Theoretical Model.
Early Development and Parenting 6: 17992.
Mesoudi, A. 2011. Cultural Evolution: How Darwinian Theory Can Explain Human Culture
and Synthesize the Social Sciences. Chicago, IL: University of Chicago Press.
Mitchell, J. P. 2009. Inferences about Mental States. Philosophical Transactions of the Royal
Society B 364: 130916.

whence religion?

Moulton, S. T. & S. M. Kosslyn 2009. Imagining Predictions: Mental Imagery as Mental

Emulation. Philosophical Transactions of the Royal Society B 364: 127380.
Murphy, J. M. 1964. Psychotherapeutic Aspects of Shamanism on St. Lawrence Island,
Alaska. In Magic, Faith, and Healing: Studies in Primitive Psychiatry Today, A. Kiev (eds),
5383. New York: Free Press.
Murray, L. & L. Andrews 2000. The Social Baby. Richmond: C. P. Publishing.
Murray, L. & C. Trevarthen 1985. Emotional Regulation of Interactions between Two-
Month-Olds and Their Mothers. In Social Perception in Infants, T. Field & N. Fox (eds),
17798. Norwood, NJ: Ablex.
Nadel, S. F. 1946. A Study of Shamanism in the Nuba Hills. Journal of the Royal Anthro-
pological Institute 76: 2537.
Niedenthal, P. M., L. W. Barsalou, P. Winkielman, S. Krauth-Gruber & F. Ric 2005. Embodi-
ment in Attitudes, Social Perception, and Emotion. Personality and Social Psychology
Review 9(3): 184211.
Nielbo, K. L. 2012. Spontaneous and Hierarchical Segmentation of Non-functional Events.
PhD dissertation, Aarhus.
Nielbo, K. L. & J. Srensen 2011. Spontaneous Processing of Functional and Non-Functional
Action Sequences. Religion, Brain, and Behavior 1(1): 1830.
Norman, D. A. & T. Shallice 1986. Attention to Action: Willed and Automatic Controls
of Behavior. In Consciousness and Self-Regulation, R. S. Davidson, G. E. Schwartz &
D.Shapiro (eds), 118. New York: Plenum.
Obeyesekere, G. 1990. The Work of Culture: Symbolic Transformation in Psychoanalysis and
Anthropology. Chicago, IL: University of Chicago Press.
Ochs, E. & L. Capps 1996. Narrating the Self . Annual Review of Anthropology 25: 1943.
Ochs, E. & M. Shohet 2006. The Cultural Structuring of Mealtime Socialization. New
Directions for Child and Adolescent Development 111: 3549.
Oesterreich, T. K. 1974. Possession and Exorcism: Among Primitive Races, in Antiquity, the
Middle Ages, and Modern Times, D. Ibberson (trans.). New York: Causeway Books.
Ohlmarks, . 1939. Studien zum Problem des Schamanismus. Lund: Lund University.
Oyama, S. 2000. Evolutions Eye: A Systems View of the BiologyCulture Divide. Durham, NC:
Duke University Press.
Persinger, M. A. 1999. Increased Emergence of Alpha Activity over the Left but Not the
Right Temporal Lobe within a Dark Acoustic Chamber: Differential Response of the Left
but Not the Right Hemisphere to Transcerebral Magnetic Fields. International Journal of
Psychophysiology 34: 1639.
Persinger, M. A. & S. A. Koren 2005. A Response to Granqvist et al. Sensed Presence and
Mystical Experiences Are Predicted by Suggestibility, Not by the Application of Transcranial
Weak Complex Magnetic Fields. Neuroscience Letters 380: 3467.
Petrovic, P., E. Kalso, K. M. Petersson & M. Ingvar 2002. Placebo and Opioid Analgesia:
Imaging a Shared Neuronal Network. Science 295: 173740.
Petrovic, P., T. Dietrich, P. Fransson, J. Andersson, K. Carlsson & M. Ingvar 2005. Placebo
in Emotional Processing: Induced Expectations of Anxiety Relief Activate a Generalized
Modulatory Network. Neuron 46: 95769.
Pickering, T. R., T. D. White & N. Toth 2000. Cutmarks on the Plio-Pleistocene Hominid
from Sterkfontein, South Africa. American Journal of Physical Anthropology 111: 57984.
Polimeni, J. & J. P. Reiss 2002. How Shamanism and Group Selection May Reveal the
Origins of Schizophrenia. Medical Hypotheses 58(3): 2448.
Posner, M. 1994. Attention: The Mechanism of Consciousness. Proceedings of the National
Academy of Sciences 91: 7398403.

armin w. geertz

Povinelli, D. J. & T. M. Preuss 1995. Theory of Mind: Evolutionary History of a Cognitive

Specialization. Trends in Neuroscience 18: 41824.
Previc, F. H. 2006. The Role of the Extrapersonal Brain Systems in Religious Activity.
Consciousness and Cognition 15: 500539.
Prince, R. H. 1982. The Endorphins: A Review for Psychological Anthropologists. Ethos
10(4): 30316.
Prince, R. J. (ed.) 1968. Trance and Possession States. Montreal: R. M. Bucke Memorial Society.
Pyysiinen, I. & M. Hauser 2010. The Origins of Religion: Evolved Adaptation or By-
Product?. Trends in Cognitive Sciences 14(3): 1049.
Raafat, R. M., N. Chater & C. Frith 2009. Herding in Humans. Trends in Cognitive Sciences
13(10): 42028.
Radin, P. 1937. Primitive Religion: Its Nature and Origin. New York: Viking.
Rawcliffe, D. H. 1959. Illusions and Delusions of the Supernatural and the Occult. New York:
Renfrew, C. & C. Scarre (eds) 1998. Cognition and Material Culture: The Archaeology of
Symbolic Storage. Cambridge: McDonald Institute for Archaeological Research.
Renfrew, C., C. Frith & L. Malafouris (eds) 2009. The Sapient Mind: Archaeology Meets
Neuroscience. Oxford: Oxford University Press (originally published as a special issue of
Philosophical Transactions of the Royal Society B, 2008).
Richardson, R. C. 2007. Evolutionary Psychology as Maladapted Psychology. Cambridge, MA:
MIT Press.
Rizzolatti, G., L. Fadiga, L. Fogassi & V. Gallese 2002. From Mirror Neurons to Imitation:
Facts and Speculations. In The Imitative Mind: Development, Evolution, and Brain Bases, A.
N. Meltzoff & W. Prinz (eds), 24766. Cambridge: Cambridge University Press.
Robbins, L. H., M. L. Murphy, G. A. Brook, A. H. Ivester, A. C. Campbell, R. G. Klein,
R. G. Milo, K. M. Stewart, W. S. Downey & N. J. Stevens 2000. Archaeology, Palaeo-
environment, and Chronology of the Tsodilo Hills, White Paintings Rock Shelter, North-
west Kalahari Desert, Botswana. Journal of Archaeological Science 27(11): 1085113.
Roebroeks, W. & P. Villa 2011. On the Earliest Evidence for Habitual Use of Fire in Europe.
Proceedings of the National Academy of Sciences 108(13): 52095214.
Ross, C. A., S. Joshi & R. Currie 1990. Dissociative Experiences in the General Population.
American Journal of Psychiatry 147: 154756.
Russell, M. D. 1987. Mortuary Practices at the Krapina Neandertal Site. American Journal
of Physical Anthropology 72: 38197.
Rylander, G. 1948. Personality Analysis before and after Frontal Lobotomy. Research Public-
ations Association for Research in Nervous and Mental Disease 27: 691705.
Saler, B. 2009. Understanding Religion: Selected Essays. Berlin: Walter de Gruyter.
Sarmiento, E., G. J. Sawyer, R. Milner, D. C. Johanson, M. Leakey & I. Tattersall 2007. The
Last Human: A Guide to Twenty-Two Species of Extinct Humans. New Haven, CT: Yale
University Press.
Sarter, M., B. Givens & J. P. Bruno 2001. The Cognitive Neuroscience of Sustained Attention:
Where Top-Down Meets Bottom-Up. Brain Research Reviews 35: 14660.
Schacter, D. L. & D. R. Addis 2009. On the Nature of Medial Temporal Lobe Contributions
to the Constructive Simulation of Future Events. Philosophical Transactions of the Royal
Society B 364: 124553.
Schjoedt, U. 2009. The Religious Brain: A General Introduction to the Experimental
Neuroscience of Religion. Method and Theory in the Study of Religion 21: 31039.
Schjoedt, U., H. Stdkilde-Jrgensen, A. W. Geertz & A. Roepstorff 2008. Rewarding
Prayers. Neuroscience Letters 443: 1658.

whence religion?

Schjoedt, U., H. Stdkilde-Jrgensen, A. W. Geertz & A. Roepstorff 2009. Highly Religious

Participants Recruit Areas of Social Cognition in Personal Prayer. Social Cognitive and
Affective Neuroscience 4: 199207.
Schjoedt, U., H. Stdkilde-Jrgensen, A. W. Geertz, T. E. Lund & A. Roepstorff 2011. The
Power of Charisma: Perceived Charisma Inhibits the Frontal Executive Network of Believers
in Intercessory Prayer. Social Cognitive and Affective Neuroscience 6: 11927.
Schjoedt, U., J. Srensen, K. L. Nielbo, D. Xygalatas, P. Mitkidis & J. Bulbulia 2013. The
Resource Model and the Principle of Predictive Coding: A Framework for Analysing
Proximate Effects of Ritual. Religion, Brain and Behavior 3(1): 7986.
Seligman, R. & L. J. Kirmayer 2008. Dissociative Experience and Cultural Neuroscience:
Narrative, Metaphor and Mechanism. Culture, Medicine and Psychiatry 32: 3164.
Shallice, T. & W. Burgess 1991. Deficits in Strategy Application following Frontal Lobe
Damage in Man. Brain 114: 72741.
Shilling, C. & P. A. Mellor 1998. Durkheim, Morality and Modernity: Collective Effer-
vescence, Homo Duplex and the Sources of Moral Action. British Journal of Sociology
49(2): 193209.
Shirokogoroff, S. M. 1935. Psychomental Complex of the Tungus. London: Kegan Paul.
Shweder, R. A. 1991. Thinking Through Cultures: Expeditions in Cultural Psychology. Cambridge,
MA: Harvard University Press.
Siegel, D. J. 2001. Toward an Interpersonal Neurobiology of the Developing Mind: Attach-
ment Relationships, Mindsight, and Neural Integration. Infant Mental Health Journal
22(12): 6794.
Silverman, J. 1967. Shamans and Acute Schizophrenia. American Anthropologist 69: 2131.
Smith, F. M. 2001. The Current State of Possession Studies as a Cross-Disciplinary Project.
Religious Studies Review 27(3): 20322.
Sol, R. & B. Goodwin 2000. Signs of Life: How Complexity Pervades Biology. New York: Basic
Srensen, J. 2004. Religion, Evolution, and an Immunology of Cultural Systems. Evolution
and Cognition 10(1): 6173.
Srensen, J. 2006. A Cognitive Theory of Magic. Lanham, MD: AltaMira Press.
Srensen, J. 2007. Acts That Work: A Cognitive Approach to Ritual Agency. Method and
Theory in the Study of Religion 19: 281300.
Srensen, J., P. Linard & C. Feeny 2006. Agent and Instrument in Judgements of Ritual
Efficacy. Journal of Cognition and Culture 6(34): 46382.
Sosis, R. 2000. Religion and Intragroup Cooperation: Preliminary Results of a Comparative
Analysis of Utopian Communities. Cross-Cultural Research 34(1): 7788.
Sosis, R. 2003. Why Arent We All Hutterites? Human Nature 14(2): 91127.
Sosis, R. & E. Bressler 2003. Cooperation and Commune Longevity: A Test of the Costly
Signaling Theory of Religion. Cross-Cultural Research 37(2): 1139.
Sosis, R. & J. Bulbulia 2011. The Behavioral Ecology of Religion: The Benefits and Costs of
One Evolutionary Approach. Religion 41(3): 34162.
Sosis, R., H. C. Kress & J. S. Boster 2007. Scars for War: Evaluating Signaling Explanations
for Cross-Cultural Variance in Ritual Costs. Evolution and Human Behavior 28: 23447.
Spanos, N. P. 1996. Multiple Identities and False Memories: A Sociocognitive Perspective, Wash-
ington, DC: American Psychological Association.
Spiro, M. E. 1997. Gender Ideology and Psychological Reality: An Essay on Cultural Reproduction.
New Haven, CT: Yale University Press.
Spuhler, J. N. (ed.) 1959. The Evolution of Mans Capacity for Culture. Detroit, MI: Duke
University Press.

armin w. geertz

Steinberg, M. 1991. Systematizing Dissociation: Symptomatology and Diagnostic Assess-

ment. Paper presented at the Dissociation Workshop for the Research Network in Mind
Body Interactions, Stanford, California.
Steinberg, M., F. Howland & Dante Cicchetti 1986. The Structured Clinical Interview for
DSM-III-R Dissociative Disorders: A Preliminary Report. In Proceedings of the Third Inter-
national Conference on Multiple Personality and Dissociative States, B. Braun (ed.). Chicago,
IL: Rush Presbyterian Hospital.
Steinberg, M., B. J. Rounsaville & D. Cicchetti 1990. The Structured Clinical Interview for
DSM-III-R Dissociative Disorders: Preliminary Report on a New Diagnostic Instrument.
American Journal of Psychiatry 147: 7682.
Stone, V. E., S. Baron-Cohen & R. T. Knight 1998. Frontal Lobe Contributions to the
Theory of Mind. Journal of Cognitive Neuroscience 10: 64056.
Suryani, L. K. & G. D. Jensen 1995. Trance and Possession in Bali: A Window on Western
Multiple Personality, Possession Disorder, and Suicide. Oxford: Oxford University Press.
Swiney, L. in preparation. Mind Control and Telepathy: What Can Reports of Unusual
Cognitive Abilities Tell Us about the Folk Psychology of the Agentive Mind.
Targ, E., M. Schlitz & H. J. Irwin 2000. Psi-Related Experiences. In Varieties of Anomalous
Experience: Examining the Scientific Evidence, E. Cardea, S. J. Lynn & S. Krippner (eds),
21952. Washington, DC: American Psychological Association.
Tattersall, I. 1998. Becoming Human: Evolution and Human Uniqueness. New York: Harcourt
Tattersall, I. 2012. Masters of the Planet: The Search for Our Human Origins. Basingstoke:
Palgrave Macmillan.
Taves, A. 2009. Religious Experience Reconsidered: A Building-Block Approach to the Study of
Religion and Other Special Things. Princeton, NJ: Princeton University Press.
Taylor, K. 2004. Brainwashing: The Science of Thought Control. Oxford: Oxford University Press.
Tinterow, M. M. 1970. Foundations of Hypnosis: from Mesmer to Freud. Springfield, IL: Charles
C. Thomas.
Tomasello, M. 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard
University Press.
Tooby, J. & L. Cosmides 1992. The Psychological Foundations of Culture. In The Adapted
Mind: Evolutionary Psychology and the Generation of Culture, J. H. Barkow, L. Cosmides &
J. Tooby (eds), 19136. Oxford: Oxford University Press.
Trevarthen, C. 1993. The Function of Emotions in Early Infant Communication and
Development. In New Perspectives in Early Communicative Development, J. Nadel &
L.Camaioni (eds), 4881. London: Routledge.
Trevarthen, C., T. Kokkinaki & G. A. Fiamenghi Jr 1999. What Infants Imitations Com-
municate: With Mothers, with Fathers, and with Peers. In Imitation in Infancy, J.Nadel
& G. Butterworth (eds), 12785. Cambridge: Cambridge University Press.
Tryon, C. A. & S. McBrearty 2002. Tephrostratigraphy and the Acheulian to Middle Stone
Age Transition in the Kapthurin Formation, Kenya. Journal of Human Evolution 42:
Turner, J. H. & A. Maryanski 2008. On the Origin of Societies by Natural Selection. Boulder,
CO: Paradigm Publishers.
Vandermeersch, B. 1981. Les Hommes Fossiles de Qafzeh (Israel). Paris: CNRS.
Varela, F. J., E. Thompson & E. Rosch 1991. The Embodied Mind: Cognitive Science and
Human Experience. Cambridge, MA: MIT Press.
Vogeley, K. & A. Roepstorff 2009. Contextualising Culture and Social Cognition. Trends in
Cognitive Sciences 13(12): 51116.

whence religion?

Vogeley, K., M. Kurthen, P. Falkai & W. Maier 1999. Essential Functions of the Human Self
Model Are Implemented in the Prefrontal Cortex. Consciousness and Cognition 8: 34363.
Vogeley, K., P. Bussfeld, A. Newen, S. Herrman, F. Happe, P. Falkai, W. Maier, N. J. Shah,
G.R. Fink & K. Zilles 2001. Mind Reading: Neural Mechanisms of Theory of Mind and
Self-Perspective. Neuroimage 14: 17081.
Vuust, P., M. Wallentin, K. Mouridsen, L. stergaard & A. Roepstorff 2011. Tapping Poly-
rhythms in Music Activates Language Areas. Neuroscience Letters 494: 21116.
Vyse, S. A. 1997. Believing in Magic: The Psychology of Superstition. Oxford: Oxford University
Walker, S. S. 1972. Ceremonial Spirit Possession in Africa and Afro-America: Forms, Meanings,
and Functional Significance for Individuals and Social Groups. Leiden: Brill.
Wallin, N. L., B. Merker & S. Brown 2000. The Origins of Music. Cambridge, MA: MIT Press.
Washburn, S. L. 1959. Speculations on the Interrelations of Tools and Biological Evolution.
In The Evolution of Mans Capacity for Culture, J. M. Spuhler (eds), 2131. Detroit, MI:
Duke University Press.
Watts, I. 1999. The Origin of Symbolic Culture. In The Evolution of Culture: An Inter-
disciplinary View, R. Dunbar, C. Knight & C. Power (eds), 11346. New Brunswick, NJ:
Rugters University Press.
Watts, I. 2002. Ochre in the Middle Stone Age of Southern Africa: Ritualised Display or Hide
Preservatives? The South African Archaeological Bulletin 57(175): 114.
Watts, I. 2009. Red Ochre, Body Painting, and Language: Interpreting the Blombos Ochre.
In The Cradle of Language, R. Botha & C. Knight (eds), 6292. Oxford: Oxford University
Wavell, S., A. Butt & N. Epton 1966. Trances. London: Allen & Unwin.
Weaver, T. D. 2012. Did a Discrete Event 200,000100,000 Years Ago Produce Modern
Humans?. Journal of Human Evolution 63: 1216.
West-Eberhard, M. J. 2003. Developmental Plasticity and Evolution. Oxford: Oxford University
White, T. D. 1986. Cut Marks on the Bodo Cranium: A Case of Prehistoric Defleshing.
American Journal of Physical Anthropology 69(4): 5039.
White, T. D., B. Asfaw, D. DeGusta, H. Gilbert, G. D. Richards, G. Suwa & F. Clark Howell
2003. Pleistocene Homo sapiens from Middle Awash, Ethiopia. Nature 423: 7427.
Wilson, D. S. 2002. Darwins Cathedral: Evolution, Religion, and the Nature of Society. Chicago,
IL: University of Chicago Press.
Wilson, D. S. 2007. Evolution for Everyone: How Darwins Theory Can Change the Way We Think
about Our Lives. New York: Delacorte Press.
Wilson, D. S. 2008. Evolution and Religion: The Transformation of the Obvious. In The
Evolution of Religion: Studies, Theories, and Critiques, J. Bulbulia, R. Sosis, E. Harris,
R.Genet, C. Genet & K. Wyman (eds), 239. Santa Margarita, CA: Collins Foundation
Wilson, M. 2010. The Re-Tooled Mind: How Culture Re-Engineers Cognition. Social
Cognitive and Affective Neuroscience 5(23): 18087.
Wiltermuth, S. S. & C. Heath 2009. Synchrony and Cooperation. Psychological Science
20(1): 15.
Winters, A. F. 2008. Emotion, Embodiment, and Mirror Neurons in Dance/Movement
Therapy: A Connection across Disciplines. American Journal of Dance Therapy 30(2):
Worthen-Chaudhari, L. C. 2011. New Partnerships Between Dance and Neuroscience:
Embedding the Arts for Neurorecovery. Dance Research 29(2): 46996.

armin w. geertz

Wrangham, R. W. 2009. Catching Fire: How Cooking Made Us Human. New York: Basic
Wulff, D. M. 2000. Mystical Experience. In Varieties of Anomalous Experience: Examining
the Scientific Evidence, E. Cardea, S. J. Lynn & S. Krippner (eds), 397440. Washington,
DC: American Psychological Association.
Xygalatas, D., I. Konvalinka, A. Roepstorff & J. Bulbulia 2011. Quantifying Collective Effer-
vescence: Heart-Rate Dynamics At a Fire-Walking Ritual. Communicative and Integrative
Biology 4(6): 7358.
Xygalatas, D., U. Schjdt, J. Bulbulia, I. Konvalinka, E.-M. Jegind, P. Reddish, A. W. Geertz
& A. Roepstorff 2013a. Autobiographical Memory in a Fire-Walking Ritual. Journal of
Cognition and Culture 13: 116.
Xygalatas, D., P. Mitkidis, R. Fischer, P. Reddish, J. Skewes, A. W. Geertz, A. Roepstorff
& J.Bulbulia 2013b Extreme Rituals Promote Prosociality Psychological Science online,
2013: 14.
Yap, P. M. 1969. The Culture-Bound Reactive Syndromes. In Mental Health Research in Asia
and Pacific, W. Caudell & T. Lin (eds), 3353. Honolulu, HI: East-West Center Press.
Yovsi, R. D. 2003. Ethnotheories About Breastfeeding and Mother-Infant Interaction: The Case of
Sedentary Nso Farmers and Nomadic Fulani Pastorals with Their Infants 36 Months of Age in
Mbven Sub Division of the Northwest Province of Cameroon. Mnster: LIT Verlag.
Yovsi, R. D. & H. Keller 2003. Breastfeeding: An Adaptive Process. Ethos 31(2): 14771.
Zahavi, A., A. Zahavi, N. Zahavi-Ely & M. P. Ely 1997. The Handicap Principle: A Missing
Piece of Darwins Puzzle. Oxford: Oxford University Press.
Ziegler, C. A. & B. Saler 2012. Dying for an Idea: Voluntary Death, Afterlife Beliefs, and
Their Role in the Quest for Immortality. In Science and the Worlds Religions, Volume 3:
Religions and Controversies, P. McNamara & W. J. Wildman (eds), 4572. Santa Barbara,

Why costly signalling models of
religion require cognitive psychology
Joseph Bulbulia

Who benefits from religion?

Call religiosity beliefs and practices respecting gods, along with the cultural
and psychological architecture that supports religion. Religiosity presents a
striking evolutionary problem. From a biological perspective, it would appear
that humans should have developed allergies to religion. The burdens that
religion imposes its opportunity costs, resource outlays, and risks appear
in vivid contrast to natures thrift. We should have evolved to recoil from
religion, as Richard Dawkins recoils from religion, yet religiosity is common-
place. Naturalists interested in the evolution of humans are faced with a cost
problem (Atran 2002; Bulbulia 2004b; Dennett 2006).
Surprisingly, given the ubiquity of religiosity, little is known about the
evolutionary dynamics that sustain its propagation (Dennett 2006; Wilson
2002). Increasingly, naturalists are approaching religiosity as an evolved com-
mitment-signalling device (see Alcorta & Sosis 2005; Bulbulia 2004a.; Cronk
1994; Iannaccone 1992; Irons 1996a, 1996b, 1996c, 2001; Schelling 1960,
2001; Sosis 2003). On this view, religion is the effect of systems elaborated
by genetic and culture selection and conserved from the benefits these sys-
tems bring to those who are religious. Signalling theory holds that religion
enables motivated cooperators to discover each other amid opportunists who
would exploit them. Religion evolved to support stable, mutually reinforcing
cooperative exchange.
This new position echoes past observations. Theorists of religion have long
noticed that the pomp and ritual expressions of religion, and the affective
delirium that accompanies it, enhance coordination and solidarity. Durkheim

The general conclusion of the book is that religion is some-

thing eminently social. Religious representations are collective

joseph bulbulia

representations which express collective realities; the rites are a

manner of acting which take rise in the midst of the assembled
groups and which are destined to excite, maintain or recreate cer-
tain mental states in these groups. (Durkheim [1915] 1964: 22)

It is easy to see how stable cooperation may enhance reproduction, by gen-

erating goods unavailable to singletons. The lone honeybee weaving its way
through a forest is an evolutionary dead end. It is widely understood that,
on the whole, humans are massively cooperative animals (Fehr & Fishbacher
2005; Henrich et al. 2004). Our primate lineage has been cooperative since
before we were human (Boehm 1999), a core feature of our success (Boyd &
Richerson 2005). That religion evolved to bolster cooperation should come
as no surprise. We are adapted to getting along with others, the most biologi-
cally relevant features of our worlds. We should expect a multi-faceted coop-
erative toolkit. Yet the religion-as-cooperation explanation still leaves us with
an apparent cost problem. It remains puzzling that selection did not strike
upon a cheaper means for enabling cooperative exchange without the epis-
temic, opportunity, and material costs of religion. As Daniel Dennett (2006)
puts it, Cui bono? (Who benefits?). Signalling theory views the costs of
religion as themselves adaptations.

Simple games and commitment devices

To understand religion as a signalling system, we need to first understand

ways in which cooperation may be a problem. Many interactions that gen-
erate collective benefits through cooperation also invite defection. The pris-
oners dilemma (PD) models core features of common cooperation problems.
Imagine an interaction between two unrelated people. Each person benefits
from mutual cooperating over going it alone. However withholding aid to
a cooperator (defection) benefits any focal agent more than mutual coop-
eration. In a non-repeated PD, this incentive structure will lead to mutual
defection. There is an incentive to cooperate but an even greater incentive
to cheat cooperation: no matter what the other person does, defection pays
best. Where rewards correlate to fitness, over the course of evolutionary time,
PDs will lead to defection. The moral of this story is that cooperation cannot
be a PD.
Biologists have discovered mechanisms that convert the payoff structures
of PDs to enable the evolution of stable cooperation.1 Signalling theorists
consider religion to be one such mechanism. Signalling theorists urge that
religiosity is evolutionarily ratifiable because religious costs allow for the
verification of cooperative commitment among unrelated partners (Irons

why costly signalling models require cognitive psychology

2001). Religious costs are configured to make it hard for free riders to draw
collective benefits without contribution. For example, entry costs will pre-
clude joining more than one religion; after all, one can only go to so many
churches. Entry costs enable ritual goers to form reliable exchange groups
(Iannaccone 1992). The costs of going to church do not exceed the benefits
of belonging to one, but they do prohibit members from multiple affiliations.
This makes it harder to belong to more than one group at time. Consider
theology. Mastery of esoteric knowledge in a theological domain not only
gives evidence of commitment: it also precommits agents to a theological
group. Learning non-utilitarian knowledge brings no practical benefit, burns
time and clogs memory. Investment in such learning leaves one committed
to one theological group, because one cannot master all theologies (easily):
Janes expertise in the Upanishads will get her nowhere in an Islamic the-
ocracy, so Janes interests converge to those of her Hindu group (Mahoney
2008). Moreover, permanent marking strategies (tattooing, scarification, cir-
cumcision, etc.) tie the fate of the individual to that of the group for there
it is inconvenient to live among strangers when perceptibly marked. Indeed,
in their survey of the ethnographic record, Sosis et al. (2007) have shown
that permanent marking correlates positively with frequency of war. This
fact is well explained on the signalling model; groups that pray together
stay together, but only if the prayerful cannot easily join, defect, and leave.
There has been growing experimental evidence supporting signalling theory;
it appears that religious communities are especially cooperative, and that
costs correlate with commitment (see Sosis 2000; Sosis & Bressler 2003;
Bulbulia & Mahoney 2008; Soler 2008). Yet signalling theory still faces
many challenges.
Scott Atran observes that signalling approaches lack a plausible psychol-
ogy: they are mind blind (Atran 2002). Signalling theory grew out of
human behavioural ecology, a discipline that abstracts away from proximate
mechanisms. It uses optimality modelling to study behaviour under specified,
quantifiable constraints. Human behavioural ecologists do not need to study
minds because this is not their subject. However, it is interesting, from the
vantage point of cognitive psychology, to consider why religious costs as
opposed to other forms of display would motivate cooperation. Whether
Peter cuts his penis for the god Zugroo or for his friend directly appears irrel-
evant. Why does Peter not dispense with the astral middleman? Irons (2008)
does not see any special property of religion to distinguish it from other forms
of group-binding solidarity (nationalism, ethno-centrism, sporting affiliates,
etc.) a view similarly expressed by Atran (2002: ch. 6). These authors argue
that the class of ritual-bound cooperative communities is larger than the class
of religious communities, yet a cost problem remains for religion. Why reli-
gious costs? Certainly, if we lavish resources on the gods, we may be signalling

joseph bulbulia

our commitment to the moral doctrines that the gods are thought to support.
An alternative explanation might be non-utilitarian entertainment: people
waste money at the movies, on vacations and drugs why not religion? In
these examples, costs are indicators of a preference. Alternatively, expenditures
might present a hard-to-fake signal of prosperity, status, or skill, rather than
of cooperative commitment. The Ferrari parked in our garage signals wealth,
not virtue. Religious excess may be configured to get the girl, or to impress
our peers rather than to prove our moral mettle (Pyysiinen 2008; Slone
2008). Why should religious costs signal cooperative commitments (Bulbulia
2008)? The ability to communicate through resource expenditure varies with
wealth, not moral commitment. Moreover, where costs are wasted, the mean
fitness of groups declines. In conditions where solidarity is needed the most
for groups at war costly signals of commitment will be most damaging.
If commitment-signalling theory is to cast light on the cost problem, we first
need to understand how beliefs and practices respecting gods motivate coop-
erative behaviour. We need a signalling theory of religion grounded in the
cognitive science of religion.

The importance of psychological frames

Social decision making is affected by contextual cues (Nisbett & Wilson

1977). Moreover people are conditional cooperators; if we suspect others
will cooperate then so will we. If not, we defect (Fehr & Fishbacher 2005).
A heuristic approach, cooperation focuses on the relationship between con-
textual cues and social behaviours (Bicchieri 2006). Caildini et al. (1990)
observed responses to littering in environments subject to specific informa-
tional parameters. In nine separate studies, the authors measured littering in
response to a task (disposal of rubbish) under varying conditions. In the first
studies, they manipulated environments (clean/dirty) to establish a base rate
for littering. Then they modified informational and normative contexts by
rendering either the environment more salient (clean/dirty) or by allowing
participants to first observe confederates act to either litter or clean. To avoid
confounding results, the confederates exited immediately after performing
their scripted behaviours. The findings were intriguing, showing that con-
textual cues impact significantly on decision-making processes. Subjects
littered more in dirty as compared with clean environments (3240% and
1118%, respectively). Moreover, littering escalated in already dirty environ-
ments where participants observed confederate dumping (54%). Yet only 6
per cent persisted in littering when observing an agent picking up garbage
from an otherwise pristine environment. The experiment suggests that there
is much variation among agents with respect to following civic norms, with

why costly signalling models require cognitive psychology

only a small minority of subjects choosing to violate norms across conditions.

Agent behaviour reliably covaried with salient features of natural and social
environments. Thus, norms appear to rely on informational triggers for their
activation (see Bicchieri 2006: 647 for discussion). In this case, agents act in
accordance with the norms and expectations we predict for our neighbours.
We do both as the Romans do and as the Romans expect.
In thinking about beliefs and practices respecting gods, we also need to
consider how immersions in contexts where such beliefs and practices are
common activate cooperative norms. Consider Boyers observation that gods
are full access strategic agents, or FASAs (Boyer 2001). On Boyers view, we
posit gods as ad hoc explanations for moral feelings and intuitions. We dont
know why we think incest is wrong and honesty is right, so we invent gods as
reasons (ibid.: ch. 5). Yet there is evidence that we respond differently when
we think others are present or watching. In his discussion of religion in the
Descent of Man, Darwin writes:

The tendency in savages to imagine that natural objects and agen-

cies are animated by spiritual or living essences, is perhaps illus-
trated by a little fact which I once noticed: my dog, a full-grown
and very sensible animal, was lying on the lawn during a hot and
still day; but at a little distance a slight breeze occasionally moved
an open parasol, which would have been wholly disregarded by
the dog, had any one stood near it. As it was, every time that the
parasol slightly moved, the dog growled fiercely and barked. He
must, I think, have reasoned to himself in a rapid and uncon-
scious manner that movement without any apparent cause indi-
cated the presence of some strange living agent, and no stranger
had a right to be on his territory. (Darwin [1871] 1981: 67)

For Darwins dog, the route from stimulus to response is not mediated by
deliberation. The dog does not philosophize before acting out (for compari-
son see Guthrie 1993). More important, it acts out in a way it would not
have acted had it perceived itself to be alone. Darwin himself does not draw
the connection to morality, but recent experimental evidence suggests that we
act less selfishly when we perceive our actions to be observed. For example,
controlled experimental games show that participants who are knowingly
observed behave more cooperatively than those who perceive their actions
to be unmonitored (Haley & Fessler 2005). Bering has recently shown that
priming participants to reflect on the ghost of a dead undergraduate reduced
cheating in otherwise anonymous games with monetary payoffs (Bering et
al. 2005). That is, merely thinking of a ghost tends to support cooperation
in unmonitored contexts. Moreover, Norenzayan and Shariff (2007) have

joseph bulbulia

recently demonstrated substantial positive correlations between implicit

supernatural primes (words that, when scrambled, suggest religious themes)
and prosocial behaviour in dictator games. Surprisingly, the effect holds for
theists and atheists alike. Consider further that the cues need not be of fully
anthropomorphized agents. Bateson and her colleagues have demonstrated
significant increases in anonymous gifting to common goods resources pools
where participants are presented with stimuli that only vaguely resemble
actual persons (Bateson et al.).1 Extra giving was found in an office environ-
ment where an image of a pair of eyes (no head, no face, no body) was placed
above a collection box for milk and coffee. This effect is striking because the
images were clearly of unreal observers mere A5 photographs of faceless
human eyes, varied and alternating each week with images of flowers for the
control condition. Average gifting increased in the eyes condition by 274
per cent.
The psychological relevance of context to decision making has an imme-
diate bearing on the analysis of religion as a solidarity device and noticing
the cooperative effects of anthropomorphism we are now in a better position
to understand why religion motivates cooperation. Merely contemplating
anthropomorphically has scope for enhancing norm-abiding behaviour. If
such agents are imagined to be (potentially) everywhere if they are FASAs
then effective policing may be secured relatively cheaply. The actual payoffs
of cooperation and defection in any given instance are not the relevant predic-
tors of cooperation. What matters are the perceptions and conditional pref-
erences of agents who interact (Bulbulia 2004a; Harsanyi & Selton 1988).
Religious persons do not face cooperation dilemmas because they do not
perceive defection as paying better than cooperation. The presence of a super-
natural other when primed will tend to trigger norm obedience.
Yet and this is the important point supernatural framing alone cannot
conserve the policing exchange over time. Assuming PDs abound, tendencies
to supernatural policing cannot become common when the anthropomor-
phic trait is rare. Against defectors, the religiously committed would only get
exploited (in PDs). Where religion is common, religious persons would also
remain exploitable by defectors. Put another way, religious cooperators face
a recognition constraint (Bulbulia 2005, 2006). Religiously motivated coop-
eration is evolvable only if co-religionists can find each other, while at the
same time spotting religious impostors. Signalling theory remains relevant.
And we have elucidated the crucial property that religious signals identify.
That property is a psychological property the perception of FASAs, and the
disposition to act on an awareness of them. Practices respecting gods certify
beliefs respecting gods.

why costly signalling models require cognitive psychology

Religious emotions as signals for cooperative futures

Costly signals identify wealth and reduce mean fitness. Religious signals are
likely to be hard-to-fake, not economically costly. Elsewhere I have urged that
emotions which indicate the perception of FASAs facilitate both cheap and
effective commitment signalling (Bulbulia 2004a). There are five properties
of emotions that jointly enable emotions to signal cooperative intentions.

1. Emotions simultaneously index how a person perceives the world and

how a person is motivated, in light of that perception (Schachter &
Singer 1962). We feel a certain way because we perceive ourselves in our
world to be a certain way, and these feelings are linked to motivations.
2. Emotional signals are public. We wear many emotions on our face:
fear, joy, anger, sadness, shock, disgust, shame, pride, and others, mark
us quite vividly, in characteristic ways (Darwin [1872] 1965; Ekman
1994). And no one teaches us how to read these basic emotional sig-
natures; and though they may be suppressed (i.e. though they are not
infallible indicators) strong emotion is often suppressed imperfectly
(Ekman 1971).2
3. Emotional signals are honest. The language of the heart does not (easily)
lie. Suppression requires effort. Some emotions are extremely difficult
to suppress. The absence of fear as the lions charge the Coliseum floor
is possible only if one believes in another world that will not end in a
lions jaw. To generalize, many emotions cannot be easily faked. This
is evident in the Duchenne smile.3 The muscular orchestration of a
smile involves the intricate synchronization of dozens of facial mus-
cles. The movements are produced outside of the neocortex, and so are
not subject to conscious manipulation, at least not directly. As a result,
nearly everyone finds it difficult to produce genuine smiles on demand
(Levenson et al. 1990). Emotions are not merely cues about how an
emoting agent thinks and feels, they offer credible cues.
4. Emotional signals predict future behaviours. This follows from the link
between emotions and motivational states. Motivations drive actions,
hence reliable information about a persons motivations improves
behavioural forecasting (Schelling 1960).
5. Emotional signals can solve recognition problems. This follows because
emotions speak prophetically and accurately, because emotions are
expressed publically, and because emotions identify likely behav-
iours. Where emotions identify cooperative commitments, emotional
signalling can benefit partners who jointly display cooperative emo-
tions. Said differently, cooperative emotions need not be construed as
mere responses to environmental conditions; emotional signalling may

joseph bulbulia

actively alter the conditions in which partners transact, converting PDs

into coordination problems (Hinde 1985). Notably, emotional signal-
ling does not carry an intrinsic price tag.

A cognitive science of religious emotions can explain the cost problem that
mind-blind behavioural signalling theories require. The genetic and cultural
systems that enable religious cognition co-evolved to enable the signalling
of cooperative intentions. Beliefs in moralizing supernaturals, if shared, will
motivate reciprocal exchange if such beliefs can be authenticated. Emotional
signals establish a pathway to authentication. Given the cooperation-
enhancing prospects of believing in a moral universe, emotional responses to
anthropomorphic agents (FASAs) will offer hard-to-fake evidence of coopera-
tive intentions. In turn, the benefits of cooperation will ratify systems that
perpetuate religious beliefs and that facilitate the projection and amplification
of religious emotions. Put simply, emotional signalling offers an inference to
the best explanation for religions ubiquity and conservation. Religious people
are able to cooperate relatively cheaply because they (i) believe in moral reality
and (ii) express these commitments in manners that are hard to fake. Through
their cooperative efforts, religions endure. To better understand the systems
that support religion requires a cognitive science of religion grounded in the
evolutionary biology of animal communication.

1. Where agents are closely related, cooperation can evolve where relatedness (r) exceeds
the costbenefit ratio: r > c/b. Where agents frequently interact, cooperation is favoured
where the likelihood of future interaction (w) exceeds the costbenefit ratio: w > c/b
(Hamilton 1964; Trivers 1971). Indirect reciprocity will favour cooperation where the
likelihood of knowing an agents image score (q) exceeds the cost to benefit ratio q >
c/b (Alexander 1987; Nowak & Sigmund 2005). Population structure also will favour
reciprocity where benefit-to-cost ratio exceeds the average number of neighbours (k): b/c
> k (Lieberman et al. 2005). Finally, group selection will favour cooperation, assuming
weak selection and rare group division, where b/c > 1+(n/m), with n = maximum group
size and m = number of groups. For a discussion of this and the other cases see Nowak
(2006). Furthermore, many psychological dispositions (for example memory, indigna-
tion, language and gossip, duty, conscience) and social institutions (for example, the
state, its police) secure and maintain these conditions.
2. Complex emotional expressions arise through culturally specific display rules (Ekman
1972; Griffiths 1997). For example, Ekman and colleagues found that Japanese and
American students expressed similar facial responses to stressful films when unob-
served, but only the Japanese students suppressed these expressions in the presence of
experimenters. Ekman attributes this effect to cultural differences respecting authority
(Ekman 1971). So development matters to the emergence of complex emotional pheno-
types. Nevertheless, the semantic conventions governing the meaning of basic emotions

why costly signalling models require cognitive psychology

are not culturally specific. Such expressions of glee at entertainment, and harsh moral
approbation, remain universal.
3. The natural smile is named after the nineteenth-century French neurologist Guillaume
Duchenne, who discovered, by electrocuting subjects faces, that natural smiles require
the fine muscular coordination of many muscles, including the orbicularis oculi muscle.
Electrical stimulation of the zygomatic major muscle alone the major muscle used
to hoist the ends of a smile looks rather like well, pain. Duchenne was one of the
first neurologists to use photography in his work, and is also credited with discovering
muscular dystrophy.

Alcorta, C. & R. Sosis 2005. Ritual, Emotion, and Sacred Symbols: The Evolution of Religion
as an Adaptive Complex. Human Nature 16(4), 32359.
Alexander, R. 1987. The Biology of Moral Systems. New York: Aldine de Gruyter.
Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. New York: Oxford
University Press.
Bateson, M., D. Nettle & G. Roberts 2006. Cues of being watched enhance cooperation in
a real-world setting. Biology Letters 2(3): 41214.
Bering, J. M., K. McLeod & T. K. Shackelforth 2005. Reasoning about Dead Agents Reveals
Possible Adaptive Trends. Human Nature 16: 36081.
Bicchieri, C. 2006. The Grammar of Society: The Nature and Dynamics of Social Norms. Cam-
bridge: Cambridge University Press.
Boehm, C. 1999. Hierarchy in the Forest. Cambridge MA: Harvard University Press.
Boyd, R. & P. Richerson 2005. The Origin and Evolution of Cultures. New York: Oxford
University Press.
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Bulbulia, J. 2004a. Religious Costs as Adaptations that Signal Altruistic Intention. Evolution
and Cognition 10(1): 1938.
Bulbulia, J. 2004b. The Cognitive and Evolutionary Psychology of Religion. Biology and
Philosophy 18(5): 65586.
Bulbulia, J. 2005. Are There Any Religions? Method and Theory in the Study of Religion 17(2):
Bulbulia, J. 2006. Natures Medicine: Empirical Constraint and the Evolution of Religi-
ous Healing. In Where Man and God Meet: The New Sciences of Religion and Brain,
P.MacNamara (ed.). Westwood, CT: Greenwood Publishers.
Bulbulia, J. 2008. Free Love: Religious Solidarity on the Cheap. See Bulbulia et al. (2008),
Bulbulia, J. & A. Mahoney 2008. Religious Solidarity: The Hand Grenade Experiment.
Journal of Cognition and Culture 8(34): 295320.
Bulbulia, J., R. Sosis, R. Genet, E. Harris, K. Wyman & C. Genet (eds) 2008. The Evolution
of Religion: Studies, Theories, and Critiques. Santa Margarita, CA: Collins Foundation Press.
Caildini, R., C. Kallgren & R. Reno 1990. A Focus Theory of Normative Conduct: A
Theoretical Refinement and Reevaluation of the Role of Norms in Human Behaviour.
Advances in Experimental Social Psychology 24: 20134.
Cronk, L. 1994. Evolutionary Theories of Morality and the Manipulative Use of Signals.
Zygon 29(1): 81101.

joseph bulbulia

Darwin, C. [1871] 1981. The Descent of Man and Selection in Relation to Sex. Princeton, NJ:
Princeton University Press.
Darwin, C. [1872] 1965. The Expression of The Emotions in Man and Animals. Chicago, IL:
University of Chicago Press.
Dennett, D. C. 2006. Breaking the Spell: Religion as a Natural Phenomenon. New York: Viking.
Durkheim, E. [1915] 1964. The Elementary Forms of the Religious Life. London: George Allen
& Unwin.
Ekman, P. 1971. Universal and Cultural Differences in Facial Expressions of Emotion. In
Nebraska Symposium on Motivation 4, J. K. Cole (ed.), 20783. Lincoln, NE: University
of Nebraska Press.
Ekman, P. 1972. Emotions in the Human Face. New York: Pergamon Press.
Ekman, P. 1994. Strong Evidence for Universals in Facial Expression: A Reply to Russells
Mistaken Critique. Psychological Bulletin 115: 26887.
Fehr, E. & U. Fishbacher 2005. The Economics of Strong Reciprocity. In Moral Sentiments
and Material Interests, H. Gintis, S. Bowles, R. Boyd & E. Fehr (eds), 15192. Cambridge,
MA: MIT Press.
Griffiths, P. 1997. What Emotions Really Are: The Problem of Psychological Categories. Chicago,
IL: University of Chicago Press.
Guthrie, S. 1993. Faces in the Clouds: A New Theory of Religion. New York: Oxford University
Haley, K. J. & D. M. T. Fessler 2005. Nobodys Watching? Subtle Cues Affect Generosity in
an Anonymous Economic Game. Evolution and Human Behaviour 26(3): 24556.
Hamilton, W. 1964. The Evolution of Altruistic Behaviour. American Naturalist 97:
Harsanyi, J. & R. Selton 1988. A General Theory of Equilibrium Selection in Games. Cambridge,
MA: MIT Press.
Henrich, J., R. Boyd, S. Bowles, C. Camerer, E. Fehr & H. Gintis (eds) 2004. Foundations of
Human Sociality: Economic Experiments and Ethnographic Evidence from Fifteen Small-Scale
Societies. New York: Oxford University Press.
Hinde, R. A. 1985. Expression and Negotiation. In The Development of Expressive Behaviour,
G. Zivin (ed.), 10316. New York: Academic Press.
Iannaccone, L. R. 1992. Sacrifice and Stigma: Reducing Free-Riding in Cults, Communes,
and Other Collectives. Journal of Political Economy 100: 27191.
Irons, W. 1996a. In Our Own Self-Image: The Evolution of Morality, Deception, and Rel-
igion. Skeptic 4: 5061.
Irons, W. 1996b. Morality as an Evolved Adaptation. In Investigating the Biological Found-
ations of Morality, J. P. Hurd (ed.), 134. Lewiston, NY: Edwin Mellen Press.
Irons, W. 1996c. Morality, Religion, and Evolution. In Religion and Science: History, Method,
and Dialogue, W. Wildman & W. M. Richardson (eds), 37599. New York: Routledge.
Irons, W. 2001. Religion as Hard-to-Fake Sign of Commitment. In Evolution and the Cap-
acity for Commitment, R. Nesse (ed.), 292309. New York: Russell Sage Foundation.
Irons, W. 2008. Why People Believe (What Other People See As) Crazy Ideas. See Bulbulia
et al. (2008), 518.
Levenson, R. W., P. Ekman & W. V. Friesen 1990. Voluntary Facial Action Generates
Emotion-Specific Autonomic Nervous System Activity. Psychophysiology 27: 36384.
Lieberman, E., C. Hauert & M. Nowak 2005. Evolutionary Dynamics on Graphs. Nature
433: 31216.
Mahoney, A. 2008. Theological Expressions as Costly Signals of Religious Commitment. See
Bulbulia et al. (2008), 1616.

why costly signalling models require cognitive psychology

Nisbett, R. E. & T. D. Wilson 1977. Telling More than We Can Know: Verbal Reports on
Mental Processes. Psychological Review 84: 23159.
Norenzayan, A. & A. Shariff 2007. God is Watching You: Supernatural Agent Concepts
Increase Prosocial Behaviour in an Anonymous Economic Game. Psychological Science
18: 8039.
Nowak, M. 2006. Five Rules for the Evolution of Cooperation. Science 314: 156063.
Nowak, M. & K. Sigmund 2005. Evolution of Indirect Reciprocity. Nature 437: 12918.
Pyysiinen, I. 2008. Ritual, Agency, and Sexual Selection. See Bulbulia et al. (2008), 17580.
Schachter, S. & J. Singer 1962. Cognitive, Social, and Physiological Determinants of Emo-
tional State. Psychological Review 69: 37999.
Schelling, T. 1960. The Strategy of Conflict. New York: Oxford University Press.
Schelling, T. 2001. Commitment: Deliberate Verses Involuntary. In Evolution and the Cap-
acity for Commitment, R. Nesse (ed.), 4856. New York: Russell Sage Foundation.
Slone, D. J. 2008. The Attraction of Religion: A Sexual Selectionist Account. See Bulbulia
et al. (2008), 1818.
Soler, M. 2008. Commitment Costs and Cooperation: Evidence from Candomble, and Afro-
Brazilian Religion. See Bulbulia et al. (2008), 16774.
Sosis, R. 2000. Religion and Intragroup Cooperation: Preliminary Results of a Comparative
Analysis of Utopian Communities. Cross-Cultural Research 34(1): 7788.
Sosis, R. 2003. Why Arent We All Hutterites?. Human Nature 14(2): 91127.
Sosis, R. & E. Bressler 2003. Co-operation and Commune Longevity: A Test of the Costly
Signaling Theory of Religion. Cross-Cultural Research 37(2): 1139.
Sosis, R., H. Kress & J. Boster 2007. Scars for War. Evolution and Human Behaviour 28:
Trivers, R. 1971. The Evolution of Reciprocal Altruism. Quarterly Review of Biology 46:
Wilson, D. S. 2002. Darwins Cathedral: Evolution, Religion, and the Nature of Society. Chicago,
IL: University of Chicago Press.

The prestige of the gods: evolutionary
continuities in the formation of sacred objects
William E. Paden

This is an attempt to reflect on some evolutionary connections between the

formation of religious objects and what can be called prestige dispositions.
I approach the topic as a historian of religion concerned with recurrences in
pan-human behaviour, and also with integrative ways of explaining those
recurrences such that compatibilities between biological and cultural frames
of analysis can be exposed.
The study of religion shows patterned behaviours affected by the pres-
ence of stereotypical social representations. In terms of evolutionary theory,
these are habitation behaviours that could be considered human versions of
environment construction (Odling-Smee et al. 2003) and emergent symbolic
cultures (Chase 1999), as well as essentialized cues that amount to dense
forms of social eco-capital. Historians of religion and Durkheimian sociolo-
gists call them sacred objects and institutions, and forms of worldmaking.
These objects have been given analytical value in terms of agency inference
and relevance (Boyer 2001; McCauley & Lawson 2002), ritual invariance
(Rappaport 1999), commitment devices (Sosis & Alcorta 2004), category
boundaries as information-processing cues (Anttonen 2004), pollution avoid-
ance (Boyer 2001: 21215, 23740), status (Milner 1994) and earlier, in the
work of social anthropologists, kin affiliation and collective order (e.g. Mary
Douglas). In broad terms, and in a way that might complement the above, I
consider here the evolution of religious complexes as systemic forms of encul-
turated prestige. Perhaps it adds one more piece to the puzzle.
Prestige attribution and status negotiation are part of the architecture and
functionality of the human mind, a mind that is thoroughly social, built for
communicative display and status behaviours, and thus intuitively geared to
respond to religious objects. My aim here is to show that religious sacredness
can be modelled among other ways in terms of permutations of social
status display. As such, I shift the spotlight from conceptual cognition to
social cognition, from thought, knowledge, beliefs and representation

the prestige of the gods

to social communication, interaction, and behaviour patterns. Gods, in this

sense, are strategic relationships.
In the evolutionary psychology of religion, much advance has been made
studying the mechanisms of intuitional ontologies, agency detection, forms
of memory, ritual competencies and conceptual inference generally. A second
phase of the movement is exploring the factor of culture and its objects, the
microprocessing mind being also an embedded social being, subject to role-
playing cues in fields of social value fields and inputs that do not simply
download into blank minds but that both play upon dispositions for social
relationships and recreate them.

Attributing prestige

Prestige is a status attribution applied to entities that gives them strategically

high standing in relation to certain areas of performance. Its dynamics and
adaptivity are relative to different social domains, circumstances and ontogenic
programs. In our own culture charismatic figures and prestige institutions are
found in sports, politics, technology, business and so forth. Religion, for its
part, seems to be a highly specialized, systemic form of prestige attribution
and manipulation, where status and sacredness are intertwined. In it, prestige
is connected with gods and their representations.
The evolutionary study of status behaviour has tended to associate it with
dominance, mate attraction, competition, or information transmission.
Barkow, for example, emphasizes how selection transformed agonistic pri-
mate dominance into human symbolic prestige and that sexual selection
was the key process in that move (Barkow 1989: 6, 183). When dominance
prestige transitions to symbolic prestige, it accordingly takes on the function
of a resource investment for social security and productivity. Geoffrey Miller
(2000) has laid out the continuity between mate attraction display and forms
of human status display, though he did not include religious behaviour in
his otherwise extensive treatment of the subject. Human ethology scholars
generally have focused on prestige economies evidenced in forms of sta-
tus and rank competition, such as the potlatch or other displays of standing
(Eibl-Eibesfeldt 1989: 297314). Still others (Henrich & Gil-White 2001;
Richerson & Boyd 2005: 1246), have concentrated on the prestige mecha-
nism as an information transmission enhancer insofar as it favored social
learners who could evaluate potential models and copy the most successful
among them (Henrich & Gil-White 2001: 165). For example:

In order to improve the fidelity and comprehensiveness of such

ranked-biased copying, social learners further evolved dispositions

william e. paden

to sycophantically ingratiate themselves with their chosen mod-

els, so as to gain close proximity to, and prolonged interaction
with, these models. Once common, these dispositions created, at
the group level, distributions of deference that new entrants may
adaptively exploit to decide who to begin copying. This generated
a preference for models who seem generally popular. Building
on social exchange theories, we argue that a wider range of phe-
nomena associated with prestige processes can more plausibly be
explained by this simple theory than by others In addition,
we distinguish carefully between dominance (force or force threat)
and prestige (freely conferred deference). (Ibid.)

These and other authors stress that among humans prestige ethologies are
less a matter of force than of excellence in valued domains of activity.
It is interesting that many of those who write about status behaviours,
along the lines of Barkow, Miller or even Eibl-Eisenfeldt and the ethologists,
stop short of applying this line of analysis to religion. Hence my concern to
examine some linkages. In most cultures, after all, religious prestige appears to
be the highest form of status, whether manifest in the gods and ancestors or
in their representative objects. I suggest here some ways of explaining this and
then describe how religious systems might look from this point of view. How
is it that religious objects, as it were, become new versions of alpha objects?

Sources of religious prestige

We may consider several factors as sources of religious prestige, all variants on

the disposition for structured attentiveness to socially ranked objects and roles.
Religious systems build on that attentiveness in environmentally shaped ways.

Deference and paying attention to social dominance

It is a phylogenetic legacy and cognitive channel to pay attention to those

around us who can effectively determine our success, fate and any vital inter-
ests. Social animals there are thousands of species with social alphas
pay special focus on those above them in rank and status within the group.
Experiments show that male monkeys will give up food for the chance just to
look at a picture of a higher-ranking individual (Adler 2004: 45), as people
will pay thousands of dollars to attend a reception for a presidential candidate.
Currently, a teenage boy meditating in a Nepalese jungle and believed to be the
reincarnation of the Buddha draws more than ten thousand observers a day.

the prestige of the gods

That some religious behaviours may be understood as legacies of the sub-

mission strategies surrounding social alphas is a judgement made by many,
not only popularizing ethologists (Morris 1984: 1467), but also recently
by E. O. Wilson (1999: 2834), Burkert (1996: 80101) and Atran (2002:
127). The behaviours referred to here include displays of various kinds of
appeasement and deference in the face of the dominant individual. Here is
E. O. Wilson:

Behavioral scientists from another planet would notice immedi-

ately the semiotic resemblance between animal submissive behav-
ior on the one hand and human obeisance to religious and civil
authority on the other. They would point out that the most elabo-
rate rites of obeisance are directed to the gods, the hyperdomi-
nant if invisible members of the human group. And they would
conclude, correctly, that in baseline social behavior, not just in
anatomy, Homo sapiens has only recently diverged in evolution
from a nonhuman primate stock.
It would be surprising to find that modern humans had man-
aged to erase the old mammalian genetic programs and devise
other means of distributing power. All the evidence suggests that
they have not. True to their primate heritage, people are easily
seduced by confident, charismatic leaders, especially males. That
predisposition is strongest in religious organizations. Cults form
around such leaders. Their power grows if they can persuasively
claim special access to the supremely dominant, typically male
figure of God. (Wilson 1999: 2834)

Wilsons cultural and theological allusions aside, it would have been an adap-
tive strategy to generate behavioural techniques for avoiding, at low cost,
being the object of aggression or shame in the face of social superiors (Krebs
& Janicki 2004: 134), including superiors capable of rending considerable
harm if not properly submitted to. Submission and deference thus include
displays of loyalty, gratitude, gift-giving and sacrifice, propitiation, and atone-
ment for offenses constant exchanges of social capital and constant ways of
gaining approval and avoiding trouble.

The prestige of collective role function

Rank status is not just a matter of one-on-one dominance and submission

behaviours, but can also evolve as a function of social specialization, castes
and guilds, and their role authority in short, as a function of social structure.

william e. paden

In human cultures, leadership patronage functions may be hereditary and

involve specialty lineages, developing essentialized categories such as shaman,
chieftain, priest, or patriarch. Here, as the object is institutionalized, deference
becomes more routinized. We could hypothesize then that gods, spirits and
ancestors, and their lineages, become castes of human cultures, precipitates
of the process of role differentiation and function. They become the brah-
mans, patrons and kings, requiring tribute displays that keep the hierarchic
world in place. Objectivized and sedimented in the world of language, given
social reality by deference behaviours, and anchored in an artefactual world of
space and object, they function as live agencies. As the patrons and defenders
of the group, arbiters of justice, producers of fate and success, the gods inhabit
virtual, psychic space.
The prestige of the social insect queens is a case in point about collective
status roles and their strategic importance in the behavioural choices of the
workers. Queens, as the egg-laying entities, draw the servicing and defensive
attentions of the workers. For the most part they are a socially constructed
function and not just an individual who happens to have dominance char-
acteristics. Indeed, queens, in most cases, do not have any different genetic
constitution than the workers: they are formed into their role (and size) by
being served special foods, much the way human groups make certain of
their individuals, through ritual behaviours toward them, into chiefs, phar-
aohs, popes, presidents and Dalai Lamas. If one of these dies, another is
made. As with the insect queens, their status indicators pheromones in the
case of queens activate dispositions in individuals to serve and protect them.
Religions systematize status indicators and are structurally full of examples
of such reciprocal practices as the Theravadin laity donating food daily to
the monks vehicles of the prestige of the Buddha and receiving merit in
To serve the vehicles of high role prestige the church, the gods, the rites
is to cultivate and defend investments in group-specific symbolic capital
and its reproduction. In the face of competition or threats to group honour
there would be a natural tendency to sacralize, perhaps to the point of a
bloating effect,1 the prestige of the patron object or its symbol. That prestige
of the flagship god or symbolic capital becomes the loyalty-inducing prestige
of the in-group. Kin selection perspective, for its part, would indicate that
readiness to sacrifice for an imagined that is, socially constructed kin
group is an evolved program that may be activated circumstantially. As well,
the notion of the inviolate transmission of certain social institutions as a kind
of cultural DNA points to the dynamics of survival strategies (Rappaport
1999: 418).

the prestige of the gods

Artefact prestige

The extension of prestige to objects is natural with the coming of human arte-
fact cultures (Dissanayake 1992). Human groups attribute values to objects
in egregious excess of the objects material worth. This is what gives special
status to an ancient piece of furniture, a winning football, a certain cancelled
postage stamp, the bones or ashes of the deceased, or even a piece of monetary
currency itself. Weapons, among the ancients, often had magical force and
status. Archaeologists have argued that:

the emergence of an economy of prestige goods provided the

means for leaders in chiefdom-level societies to attract followers
and establish hierarchical relations with elites in neighboring poli-
ties, leading to the formation of permanent social ranking catego-
ries and hierarchical political structure prestige goods originally
appear as a response to increased competition for prestige, and
operate as costly signals of high levels of skill and knowledge.
(Plourde 2006)

Religious objects are variants on this process. It is an intelligible series of steps

to the institutionalized holiness of totemic churingas, the Kaabas and shrines,
the holy books of the religions. In relation to the complexity and size of their
populations, and hence the emergence of social ranking in group behaviours,
human cultures built systematic one might say hypertrophic forms and
contexts of social attentiveness to these scaffolds, surrogates and props for the
gods.2 To that extent religious systems would represent the epitome of sym-
bolic culture understood as an emergent evolutionary environment (Chase
1999: 42). Thus, thousands of holy objects have sat side by side on the planet,
each a priceless currency for its people, each empowering, and each typically
irrelevant or nonexistent in other social landscapes.

Trait display and enhancement

Prestige traits can also be understood as part of the process of communica-

tive display enhancement. Many experiments show that animals are more
interested in a pronounced or exaggerated copy of some biological signal or
key stimulus than in the real thing. The cue could be visual, behavioural
or pheromonal.

Males of the silver-washed fritillary [a kind of butterfly] appear to

have evolved to prefer the strongest expression of certain stimuli

william e. paden

they encounter, with no upper limit. The phenomenon is wide-

spread in the animal kingdom. While experimenting with anole
lizards of the West Indies a few years ago, I found that males dis-
play enthusiastically to photographs of other members of the same
species, even though the images are the size of a small automobile.
Other researchers have learned that herring gulls ignore their own
eggs when presented with appropriately painted wooden models
so large they cannot even climb on top of them.
(Wilson 1999: 252)

The adult female gull has an orange spot on her beak, at which her
chicks instinctually peck, to stimulate the female to regurgitate
and feed them. Tinbergen showed that chicks would peck even
more readily at exaggerated cardboard models of the orange spot,
so-called supernormal stimuli. (Dennett 2006: 122)

Religious objects, including gods, may be such supernormal stimuli:3

enhanced, exaggerated, or otherwise strong expressions of social status in the
ways they are represented and regarded. They become the stereotypic orange
spots or painted wooden models built for maximum effectiveness in activat-
ing dispositions of social respect. Religious spots and models, however, are
not necessarily exaggerated in physical size, though they can be grand, but are
characteristically enhancements of signs of authority status and ritual precau-
tions. Humble physical objects that nevertheless have high symbolic associa-
tions, such as relics or certain representative emblems of gods and ancestors,
can therefore command the greatest prestige. Myth, icon and ritual, for their
part, become the ultimate display-language enhancers for religious cultures.
The content of the stimulus signs will be mostly culture specific.
Note that the receiver of the signal is not always other humans. Humans
also display to gods. Is that not the religious syndrome itself? Religious behav-
iour is tuned to impress its object, as in the communicative lineage of the
bowerbird. We build and perform to attract and impress. Where the behav-
iours are costly statements to impress the god with signals of loyalty, praise,
or sincerity, adherents would expect the benefits to be greater. The statements
are an investment in religious capital (Stark & Finke 2000: 281). St Simeon
Stylites, living his life atop a pillar, clearly signalled his fitness for the mythic
kingdom of God world renunciation being high currency in that fifth-
century environment.
The handicap principle (Zahavi & Zahavi 1997) applies here. The
Zahavis have tried to show that some extreme behaviours among animals are
deliberately risky and costly in order to communicate status and excellence
of qualities. Those qualities for example, the ability to defend or provide

the prestige of the gods

for others in turn redound to reproductive chances and social rank. Their
point is that in order to be effective signals have to be reliable; in order to
be reliable, signals have to be costly(ibid.: xiv). For example, by manag-
ing to find food and avoid predators despite its enormous tail, a peacock
proves that he is the high-quality mate that the peahen is seeking to father
her future chicks (ibid.: xiv). Displays of high quality religious behaviour
and faith, for their part, would seem to redound, in the mental world of the
adherents, to chances for eternal life and prestige among the gods. Notably,
and the subject of much current research, they may also function as adaptive
commitment devices regulating group cooperation (Sosis & Alcorta 2004;
Bulbulia 2004).

Prestige and other social dispositions

Prestige is buoyed by a number of other evolutionary dispositions. A con-

formity bias seems to be at work here an adaptive mechanism that tradi-
tionally functions to make information acquisition more efficient (Boyd &
Richerson 2005: 8397). Michael Tomasellos work on the sociogenesis of
attention sharing would apply (Tomasello 1999), as would costbenefit or
rational choice analysis, and even mirror-neuron theory. The cognitive opti-
mum draw of counter-intuitive objects, so well described by Pascal Boyer,
might have applications to the formation of charisma attribution. And in
Sperbers epidemiological terms cultural representations replicate by causing
those who hold them to produce public behaviours that cause others to hold
them too (Sperber 1996: 100). But memetic theory would also be relevant.
Prestige is catchy. It is its own cachet, its own meme. If we are fascinated by
people, it is normally because they are already deemed fascinating by oth-
ers; we are intrigued by famous people not only because we are attracted to
their special qualities, but also by the very fact that they are famous. Indeed,
that may be the special quality itself. In the colloquial, nothing succeeds
like success. So successful was Simeons prestige that 35,000 pilgrims would
arrive daily to witness it and one of the largest basilicas in the world was
built over his display site. Finally, Plotkin and others review a number of
other psychological mechanisms that select for or induce group representa-
tions, such as described in social agreement theory, habitualization theory,
docility theory, social force theory, automaticity of everyday life theory, and
theories of various kinds of morality modules activated by moral norms
and social situations (Plotkin 2003: 24890; Krebs & Janicki 2004). While
individuals may ultimately be the only agents, the world of prestige objects
is given social existence by individuals agreeing on what is there and what
its status is.4

william e. paden

Religions as prestige systems

Among other things, religions can be seen as forms of culture that enhance
objects by endowing them with superhuman attributes and that engender
interactive relationships with those objects. To that extent they are contextual-
ized by the processes described above. Here, in summary form, are some more
illustrative ways that the prestige model can apply to religious behaviours.

Myth and ritual as prestige enhancers

Mythicization and ritualization, in this framework, can be understood as ways

that religions signal-enhance the social status and honour of their objects.
This includes, most fundamentally, language that connects or endows the
objects with supernatural attribution or miracle.
Myth eternalizes its objects, a fundamental strategy for depicting value.
The object is pictured as always existing, perhaps from the beginning of time
or even before the beginning of time. It becomes an archetype not eroded by
temporality. When a guru is described as an avatar of Krishna this is a high
mythic attribution and social enhancement. The most sacred cemetery in the
Shiite world, the Valley of Peace in Najaf, is represented in legend as dedi-
cated from the creation, and the emperor Justinian had believed the same of
his Hagia Sophia. Many scriptures are regarded as eternal blueprints rather
than as historically produced. Where founders and saints have historicity they
are nevertheless linked with genealogies or lineages that go back to original,
founding times. We see this in royal lines of kingship, too. Countries have
their eternal flames honouring the sacrifices of their ancestors, and even
folk heroes like Elvis Presley become ensconced with legends of immortal-
ity. Myth, then, preserves and insulates the status of its objects in a virtual,
archetypal, eternal world.
In addition, mythic prestige employs the language of honorific and super-
lative idealization. Protocols for addressing gods follow exceptional linguistic
etiquette appropriate to their special social rank often it is terminology that
is only applied to the god, or it is language specially transmuted into forms
of chant and song. In short, mythicization which could be called one of our
grand phenotypic traits uses the resources of language in every way to bol-
ster the prestige of its material, and thus bestows a kind of alphahood on it.
Much ritualization is also an obvious prestige-booster, with its physical
and emotional enhancement scenes and employment of group attention tech-
niques. It differentiates states of purity and impurity at every level, and regu-
lates status through offerings and gift-giving. Worship itself has been studied
as a status process (Milner 1994: 17288). Ritual can intensify the focus on

the prestige of the gods

an object by creating special configurations of space altars, for example, or

the holy of holies and requiring degrees of behavioural preparedness to
approach them. The Grand Shrine of Ise in Japan houses its sacred symbols
at the centre of seven concentric circles of access; every synagogue houses the
Torah in a marked place of honour; cathedrals or pyramidal temples emphasize
asymmetric, vertical relationships between the below and the above. For their
part, festivals and other marked times give temporal punctuation to the pres-
tige of the object by making it choreographed as a time like no other time.
Special display behaviours also enhance the prestige of the object: kneel-
ing, prostration, bowing, decorating, undergoing pain, processing, making
pilgrimage, making special offerings of goods, dressing up in distinctive ways.
The actions, the physical scenes, the ideal controlled environments, the fre-
quency or infrequency of the observances all shape the relative honour of
the object just as they correspond to it.

Religious prestige is essentialized

Religious objects can take on the nature of a transmittable substance. The

power of the prestige, the substantiveness of the honour, make the object an
entity, a thing, an ontological kind, an object. So understood, the creation
of the gods is not a mystery: with a nod to St Anselm, if the object has the
greatest prestige, it must exist. The social chemistry of honour becomes trans-
formed into a world of representation, prototype and concept, and essen-
tialization is enhanced when the object is given an artefactual representation
along with a linguistic one.

Prestige as distributable and shareable

The prestige of objects can be spread by contact and association in a network

of relics, rites, saints, places, authorities, and ultimately individuals. We speak
of someone giving their prestige to such and such an occasion, cause, or
group. In a million daily Masses around the globe the prestige of the god is
substantively passed on in the consecrated bread and wine, and in millions
of recitations of the Qurn each day, the holy words of the god are dissemi-
nated. When the Ise Shrine is periodically disassembled its parts are distrib-
uted to other Shinto shrines through the land, just as pieces of the Kiswah,
the black cloth that covers the Holy Kaaba in Mecca and is replaced each year,
are annually carried to Muslim groups around the world.
This is to say that prestige is not only something we attribute to objects,
but also something we receive from them, something we benefit from, in a

william e. paden

circle of mutual gain and contact. To some extent this acquisition of mental
blessings, of grace, or darshan, is at the heart of religious behaviour, ritual
or mystical. It is shown directly in the phenomenon of faith healing, where
actual contact with the prestigious object becomes the highpoint of religious
reciprocity. Ritual structures thus set up ways of contacting pieces of the
object and its charisma, where the part nevertheless contains the power of the
whole. A holy woman believed to be the incarnation of the Mother Goddess
and who has gained an international following by her habit of hugging
devotees sometimes thousands at a single event thus gives the followers
dramatically tangible contact with their spiritual object and its phenotypic
The founders of comparative religion called this the contagion of the
sacred, noting its analogues with electricity. One might also think back to
the analogies with pheromones secreted by social insect queens, where worker
bees, for example, will lick it off the queens body or get it indirectly from
other workers who have had contact with her (Zahavi & Zahavi 1997: 158).5
The pheromone for humans, here, is social prestige.

Prestige status is not just a function of projection onto objects,

or contact with them, but also an acquisition of individuals

We attribute prestige, we also want it in the form of status, recognition and

approval, and we need to avoid its opposite, shame. Individuals cultivate
status by conducting themselves with group-defined genres of reciprocity,
holiness, purity or honour. True piety is then manifest in the social matrix
of conspecifics and the watching gods. It was William James, in the Varieties
of Religious Experience (James [1901] 1960), who emphasized the point that
saints, so-called, are such by virtue of the different kinds of social environ-
ments they inhabit. At the same time, note that the gods, spirits and ancestors
themselves constitute a virtual, if physically absent social reference group
for individual behaviour (Barkow 1989: 192). Much of the history of reli-
gious values and behaviour can be construed in terms of differential qualities
signalled to the gods relative to social contexts, whether those social contexts
are on earth or in the theosphere. It is as though there were a kind of mate-
choice effect going on here and in some cases, a runaway effect between
gods and people.
In religious contexts, display signals indicate any kind of phenotypic qual-
ity, such as patience, generosity, or altruism, but reliable signalling the cur-
rency of value and prestige is often in the form of self-sacrifice, martyrdom,
suffering or other extreme forms of devotion and self-denial. The initiatory
ordeals of shamans and the perceived difficulty of their trance journeys are

the prestige of the gods

prestige producers; and the primary symbol of the Christian tradition is a

demonstration of the handicap principle in the form of a man-god who will-
ingly undertakes crucifixion and death to show his celestial power, quality and
reliability. The desert fathers practised egregious humility behaviours, which
displayed the sacrifice of pride in a kind of inverse potlatch. The need for
religious status and approval has its hypertrophic peaks in various forms of
ascetic altruism (Lopreato 1984: 188).
The social opposites of prestige are shame and its variants such as guilt,
impurity, sin, bad karma. Insofar as religious programmes are largely designed
to counteract these, most noticeably in the more overt theologies of redemp-
tion, it is often the sharing of the prestige of the alpha as with accepting the
power of the Vow of Amida Buddha that brings this about.

The history of religions can be understood in

terms of the pivoting of prestige attribution

Religious prestige can attach, alternatingly, to any number of domains within

a culture ancestors, current leaders, domestic sacra, public rites as indi-
viduals find themselves behaving in shifting social environments and circum-
stances. It is also expressed in varying kinds of phenotypic traits. Dramatic
prestige transmutations occur over historical time, witnessed in reformations
and new religious movements where what is sacred one day is profane the
next. Anti-hierarchical dispositions, which go back to the primate herit-
age and huntinggathering days (Boehm 1999: 1012), take on a phylo-
genetic resonance. And the attribution of highest value to concepts such as
human rights, individual conscience and political liberty still continues to be
anchored in notions that these are to be honoured as divine, inalienable fea-
tures of human life. Status ascriptions can alter or challenge any previously
existing dichotomies about what constitutes a prestigious institution and its

Concluding points

Broadly sketched, I have suggested here some ways that prestige sociality
might be modelled as a link between religious behaviour and evolving social
strategies, and how it might be productive to examine that continuity more
fully. Among other things, the status behaviour model has the function of
recontextualizing in biocognitive terms a concept that has been at the heart
of the history of religions field, namely, sacredness. Much religion is about the
honour of its objects and negotiating the status of that honour, and much of

william e. paden

myth and ritual is about prestige enhancement. At the same time, I am not
presenting this as a theory of religion in any generic sense. It would be short-
sighted to miss the point that religion, which is only a conceptual umbrella
term, encompasses an enormous set of phenotypic traits, any one of which
may have a different evolutionary pedigree. I have focused on just one trait,
which could itself be parsed further.
Social objects are full of information, and their prestige is a piece of it. Yet
I have primarily drawn my analysis in terms of behavioural strategies rather
than in terms of information processing mechanisms per se. At the same
time, depicting a theatre of social displays, with its imagined environments
of spiritual status objects, is not simply a phenotypic gambit (Smith 2000:
30) that thereby avoids adaptive or dispositional factors. It seems to me that
the study of pan-human social behaviour patterns, as instanced above, is able
to capture some meaningful species-typical uniformities.6
Responding to religious prestige which is variably relevant in shifting
environmental circumstances is not a function of a single, task-specific
program or decision-making formula. Inferential programs overlap and are
subject to socio-ecological variables; adaptations build upon adaptations;
environmental cues occur within other environmental cues; social frames and
responses vary throughout the day; and any information-processing deci-
sion is apt to become a cascading chain of many kinds of decisions. Any
given unit of behaviour, as Whitehouse summarizes, involves the activation
of numerous nuclear and global systems (Whitehouse 2006: 22).
The notion of prestige bias invites integration of theoretic resources and
levels of analysis.7 Notions of charisma, status and sacrality as analysed in
Weberian and Durkheimian traditions could find more direct evolutionary,
ethological grounding here.8 All in all, this model can point up some interac-
tions of cultural, social and psychological levels of the evolutionary process
while relating them directly to behavioural strategies in the construction of
religious worlds.


1. One thinks here of the work of the social theorist Ernest Becker, who gave classic for-
mulation to the notion that culture, for which religion is the prototype, functions as a
bloated veneer of self-enhancement including illusions of sacrality, power and immor-
tality built upon the fear of death or extinction. The negative side of this status seek-
ing is the logic of killing others deemed to be threats to it. See his The Denial of Death
(Becker 1973) and Escape from Evil (Becker 1975).
2. Matthew Day assesses the theory held by Steven Mithen that such artefacts were nec-
essary cognitive compensations for the computational challenges that are introduced
when the gods appear on the scene (Day 2004: 250). Day inclines to think that the

the prestige of the gods

artefacts are not cognitively essential for dealing with gods, but rather likely to be
content-fixing elements in religious cognition, an additional input class that exists
alongside the traditional vehicles (narrative and ritual) for generating and transmitting
religious knowledge (ibid.: 253).
3. The expression supernormal sign stimuli is found in N. Tinbergens The Study of
Instinct (Tinbergen 1951: 446).
4. Hence, this is not just a simple matter of downloading social norms into otherwise
blank minds. Groups may be made up of small self-interested components, as evolu-
tionary psychologists point out, but within those units are mechanisms for represent-
ing and respecting agreed upon rules and prestige institutions. Groups may thus be in
some senses perpetually reconstructed output fictions of individual minds but among
those fictions are powerful ideologies and authority attributions that powerfully influ-
ence behaviour and thus can even have deadly causal force. Sacred norms may just be
intermittent signposts of behavioral tendencies rather than shared norms that auto-
matically replicate in individuals, as put in Atrans terms (Atran 2002: 199), but the
message on the sign can incite holocausts.
5. The Zahavis even suggest that having queen pheromone is a vehicle of relative prestige
among workers, somewhat analogous to money in human societies (Zahavi & Zahavi
1997: 159), though there is apparently no experimental basis for this yet. Along these
lines, I recall that certain followers of the Aum Shinrikyo sect were known to don elec-
tric headsets supposedly synchronized with the brain of their founder.
6. Tooby and Cosmides hold that behavioural levels of analysis ordinarily deal with too
kaleidoscopic a range of phenomena to identify universal, functional mental architec-
ture (Tooby & Cosmides 1992: 64). Yet communicative status behaviours are them-
selves grounded in adaptive mechanisms of a general kind.
7. Laland and Brown (2002) impressively interconnect the five major strands of evolu-
tion theory (human sociobiology, evolutionary psychology, human behavioural ecology,
memetics, and gene-culture co-evolution). For a helpful sorting out of the explanatory
complementarity between the approaches of evolution psychology and behavioural
ecology see Smith (2000). Extending the evolutionary psychology model to include
more attention to externalized, cultural forms of cognition and their constraints is
clearly underway among religion scholars (for example, Pyysiinen 2004; Srensen
2004; Day 2004).
8. Prestige, as social respect, as honour attributed to objects, is a concept related to
Durkheims notions of social force, the totemic principle and mana. In the Elementary
Forms these are products of the representations of many individuals (Durkheim 1995:
21011), thus giving them imagined autonomy, and it is the psychic properties of social
respect that give it power (ibid.: 209). There is therefore significant conceptual linkage
between the model described in this paper and Durkheims notion of the energy that
social respect attributes to sacred objects. Connections could also be made with Webers
three types of authority legitimation (tradition, charisma, legality). Milner points out
that no sustained work has developed the connection of status and sacrality (Milner
1994: 291), though his own book, albeit without evolutionary reference, opens up the
issue nicely. Lopreato (1984) and Sanderson (2001) also begin to close the gap between
social and evolutionary theory.

william e. paden


Adler, J. 2004. Mind Reading. Newsweek (5 July): 446.

Anttonen, V. 2004. Pathways to Knowledge in Comparative Religion: Clearing Ground for
New Conceptual Resources. In Religion as a Human Capacity, T. Light & B. C. Wilson
(eds), 10519. Leiden: Brill.
Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. New York: Oxford
University Press.
Barkow, J. H. 1989. Darwin, Sex, and Status: Biological Approaches to Mind and Culture.
Toronto: University of Toronto Press.
Becker, E. 1973. The Denial of Death. New York: Free Press.
Becker, E. 1975. Escape from Evil. New York: Free Press.
Boehm, C. 1999. Hierarchy in the Forest: The Evolution of Egalitarian Behavior. Cambridge,
MA: Harvard University Press.
Boyd, R. & P. J. Richerson 2005. The Origin and Evolution of Cultures. New York: Oxford
University Press.
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Bulbulia, J. 2004. The Cognitive and Evolutionary Psychology of Religion. Biology and
Philosophy 19: 65586.
Burkert, W. 1996. Creation of the Sacred: Tracks of Biology in Early Religions. Cambridge, MA:
Harvard University Press.
Chase, P. G. 1999. Symbolism as Reference and Symbolism as Culture. In The Evolution of
Culture: An Interdisciplinary View, R. Dunbar, C. Knight & C. Power (eds), 3449. New
Brunswick, NJ: Rutgers University Press.
Day, M. 2004. The Ins and Outs of Religious Cognition. Method and Theory in the Study of
Religion 16(3): 24155.
Dennett, D. C. 2006. Breaking the Spell: Religion as a Natural Phenomenon. New York: Viking.
Dissanayake, E. 1992. Homo Aestheticus: Where Art Comes From and Why. New York: Free
Durkheim, E. 1995. The Elementary Forms of Religious Life, K. E. Fields (trans.). New York:
Free Press.
Eibl-Eibesfeldt, I. 1989. Human Ethology. New York: Aldine de Gruyter.
Henrich, J. & F. J. Gil-White 2001. The Evolution of Prestige: Freely Conferred Deference as
a Mechanism for Enhancing the Benefits of Cultural Transmission. Evolution and Human
Behavior 22: 16596.
James, W. [1901] 1960. The Varieties of Religious Experience. London: Collins.
Krebs, D. & Maria Janicki 2004. Biological Foundations of Moral Norms. In The Psycho-
logical Foundations of Culture, M. Schaller & C. S. Crandall (eds), 12548. Mahwah, NJ:
Lawrence Erlbaum Associates.
Laland, K. N. & G. R. Brown 2002. Sense and Nonsense: Evolutionary Perspectives on Human
Behaviour. Oxford: Oxford University Press.
Lopreato, J. 1984. Human Nature and Biocultural Evolution. Boston, MA: Allen & Unwin.
McCauley, R. N. & E. T. Lawson 2002. Bringing Ritual to Mind: Psychological Foundations of
Cultural Forms. Cambridge: Cambridge University Press.
Miller, G. 2000. The Mating Mind: How Sexual Choice Shaped the Evolution of Human Nature.
New York: Doubleday.
Milner, M. Jr, 1994. Status and Sacredness: A General Theory of Status Relations and an Analysis
of Indian Culture. New York: Oxford University Press.

the prestige of the gods

Morris, D. 1984. The Naked Ape. NewYork: McGraw-Hill.

Odling-Smee, F. J., K. N. Laland & M. W. Feldman 2003. Niche Construction. The Neglected
Process in Evolution. Princeton, NJ: Princeton University Press.
Plotkin, H. 2003. The Imagined World Made Real: Towards a Natural Science of Culture. New
Brunswick, NJ: Rutgers University Press.
Plourde, A. 2006. Prestige and the Origins of Social Inequality, www.ucl.ac.uk/ceacb/
Pyysiinen, I. 2004. Magic, Miracles, and Religion. Walnut Creek, CA: AltaMira Press.
Rappaport, R. A. 1999. Ritual and Religion in the Making of Humanity. Cambridge: Cambridge
University Press.
Richerson, P. & R. Boyd 2005. Not by Genes Alone: How Culture Transformed Human Evolution.
Chicago, IL: University of Chicago Press.
Sanderson, S. K. 2001. The Evolution of Human Sociality: A Darwinian Conflict Perspective.
Lanham, MD: Rowman & Littlefield.
Smith, E. A. 2000. Three Styles in the Evolutionary Analysis of Human Behavior. In Adapt-
ation and Human Behavior: An Anthropological Perspective, L. Cronk, N. Chagnon &
W.Irons (eds), 2748. New York: Aldine de Gruyter.
Srensen, J. 2004. Religion, Evolution, and an Immunology of Cultural Systems. Evolution
and Cognition 10: 6173.
Sosis, R. & C. Alcorta 2004. Signaling, Solidarity, and the Sacred: The Evolution of Religious
Behavior. Evolutionary Anthropology 12: 26474.
Sperber, D. 1996. Explaining Culture: A Naturalistic Approach. Oxford: Blackwell.
Stark, R. & Roger Finke 2000. Acts of Faith: Explaining the Human Side of Religion. Berkeley,
CA: University of California.
Tinbergen, N. 1951. The Study of Instinct. Oxford: Clarendon Press.
Tomasello, M. 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard
University Press.
Tooby, J. & L. Cosmides 1992. The Psychological Foundations of Culture. In The Adapted
Mind: Evolutionary Psychology and the Generation of Culture, J. H. Barkow, L. Cosmides &
J. Tooby (eds), 19136. New York: Oxford University Press.
Whitehouse, H. 2006. The Cognitive Parsing Model: Nuclear and Global Psychological
Systems in the Transmission of Culture, www.artsci.wustl.edu/~pboyer/RelCognWebSite/
Wilson, E. O. 1999. Consilience: The Unity of Knowledge. New York: Random House.
Zahavi, A. & A. Zahavi 1997. The Handicap Principle: A Missing Piece of Darwins Puzzle. New
York: Oxford University Press.

The evolutionary dynamics of religious systems:
laying the foundations of a network model
Istvn Czachesz

This chapter aims at laying the foundations for the study of religions as sys-
tems, which would enable scholars to produce formalized and quantitative
explanations and predictions about the inner causal structure and possible
developmental tracks of religions. Whereas the notion of a system has been
formerly used in connection with culture and its various aspects (cultural
systems, symbol systems, thought systems, belief systems and even ritual and
religious systems), these accounts have not been based on a shared, formal,
let alone mathematical, definition of systems and did not therefore provide
scholars with appropriate tools to develop quantitative explanations and pre-
dictions about culture or religion.
Frustrated by the loose, metaphorical and ultimately not very produc-
tive talk about systems in cultural studies, sceptics have recently raised their
voices against too easily presuming the existence of systems where there might
be none. For example, Pascal Boyer (1994: 229) has written about the false
theologism that takes the existence of connections among religious assump-
tions for granted. Benson Saler (2001, 2005; and personal communication)
has argued that beliefs do not constitute a system, because there are appar-
ently numerous beliefs that we can remove from the set of an individuals or
cultures beliefs without affecting any other belief. Cognitive anthropologist
Roy DAndrade suggested that culture is a collection of units, rather than an
entity. Cultural items in the minds of people do not constitute a thing
because they are lacking entitativity. DAndrade compares this with the col-
lection of items on his desk:

the collection of things on my desk doesnt really make much of

a thing because the items on my desk arent in immediate con-
tact with each other, arent made of the similar stuff, dont have
much of a common fate, dont strongly resist dispersion, and dont
interact strongly. Basically, the collection as a whole has no causal

the evolutionary dynamics of religious systems

properties. In my opinion, the situation with respect to the

entitativity of the collection of cultural items found in the minds
of people living on Bali is not much better than that for the things
on my desk [Clifford] Geertzs opinion of the matter notwith-
standing. (DAndrade 2001: 252)

DAndrades arguments raise interesting questions as to the extent to which

one can compare a collection of items on ones desk with beliefs in an indi-
viduals mind, as well as about the systemic properties of both examples.
But even scholars who did think about culture as a system, such as Clifford
Geertz, whose ideas DAndrade is criticizing in the passage quoted above,
did not necessarily want to make quantitative, scientific predictions about
culture or religion. Geertz, in spite of his own claims to the contrary, has
been accused of rejecting anthropology as a scientific endeavour (Pals 2006:
2857), and he famously claimed (C. Geertz 1973: 5) that cultural analysis
is not an experimental science in search of a law, but an interpretive one in
search of meaning.
It has to be noticed, however, that not only scholars of culture and religion
have been lacking a shared definition of systems. The problem is not that
there have been no definitions around, but rather that there have been too
many, and they were too diverse and too unconstrained (Backlund 2000).
Among the many possible ways to discuss and analyse systems, I have chosen
network theory, which has developed into an independent research field in
recent years from mathematical graph theory, incorporating insights from the
study of networks in a number of disciplines, such as chemistry, biology, ecol-
ogy, sociology, economics and computer science. This article will therefore
introduce concepts that have not been used in cultural studies or religious
studies traditionally, but we will keep the use of technical terms and math-
ematical notation to a minimum.
In the first part of the article, I will introduce the concept of systems and
its exact mathematical definition in terms of graph theory. In the second part,
I will consider the possible levels and angles of analysis that would allow us
to create system theoretical models of religion. Finally, I will give examples
of how the dynamics of a religious system can be studied in the context of
system theory.

Defining a system

Ludwig von Bertalanffy (1950: 143), the father of general system theory,
has given the following definition of a system: A system can be defined as
a complex of interacting elements Interaction means that the elements

istvn czachesz

stand in a certain relation, R, so that their behaviour in R is different from

their behaviour in another relation, R. Similar definitions have been pro-
posed by other scholars. For example, according to J. G. Miller (quoted by
Backlund 2000: 444), a system is a set of interacting units with relationships
among them. As Alexander Backlund (ibid.: 445) has recently pointed out,
the problem with most definitions is that whereas they include some kind
of systems, they exclude others, and do not exclude everything that is not a
system. Consider a set of elements (a, b, c, d), where a interacts with b and
c interacts with d (in the sense of Bertalanffys definition). Do the four ele-
ments and the relations among them constitute a system? Since a and b are
not affected by any of the other two elements (and vice versa), it might be
perhaps better to speak of two systems instead of one in this case. And what
shall we say about a and b if the behaviour of a influences the behaviour of
b but not the other way around? We could answer the latter question by the
clarification that interaction means a bidirectional relation, but this would
not yet fix the previous problem. Shall we also require that every element of
a system interacts with every other element? That would seem redundant: if
a interacts with b and b with c, then an interaction between a and c is not
needed any more (but rather emerges spontaneously). If we do require, how-
ever, that a interacts with c, the interactions do not have to be bidirectional
any more: the three elements can influence each others behaviour in a chain-
like manner (abca) and still constitute a system. The bottom line
seems to be that relations within the system should allow for every element
(potentially) affecting the behaviour of every other element but this again is
a very imprecise definition that does not allow for any conclusions about how
systems look like and how their elements are actually connected.
Instead of multiplying such clarifications, in this article we will proceed
from a more compact and unambiguous, mathematical definition of systems
that relies on the set-theoretical definition of Alexander Backlund (2000) and
its graph-theoretical extension by Igor Gazdk (2006). Let us start by defin-
ing a couple of concepts that we will need for our definition of systems: set,
relation, graph and path.

1. A set is a collection of objects. For example, the collection of objects on

a desk mentioned in the previous section can be understood as a set.
2. As a next step, we can define relations among the elements of a set. For
example, on the non-empty set M = {1, 2, 3} we can define the relation
is greater than, which is a set containing the following ordered pairs:
R = {(2, 1), (3, 1), (3, 2)}. In other words, an ordered pair (a, b) is an
element of the relation R defined on the set M, if both a and b are ele-
ments of M and a is in relation with b (relation being is greater than
in our example).

the evolutionary dynamics of religious systems


Contract Contract
Figure 4.1 Graph representing the relation R = {(2, 1), (3, 1), (3, 2)} on set M = {1, 2, 3}.

3. A relation defined on a set can be represented as a graph, provided that

we are speaking of finite set. Graphs are mathematical objects consist-
ing of vertices connected by arcs (or edges). The vertices of the graph in
Figure 4.1 represent the elements of the set M = {1, 2, 3} in our exam-
ple, and the edges connect the ordered pairs that appear in the relation
R = {(2, 1), (3, 1), (3, 2)}, pointing from the first element toward the
second element of each ordered pair. In graph theory, edges that have a
direction are called directed edges.
4. Finally, we can define the concept of path using another example.
Imagine that the vertices in the graph are shops and the directed edges
are one-way streets connecting them. If we can drive from shop a to
shop b (driving on any street only in the permitted direction) we can
say that a path exists from a to b. In our example, there is a path from
vertex 3 to vertices 1 and 2 as well as from vertex 2 to vertex 1.

For the purposes of the present article we adopt Backlunds (2000: 448) defi-
nition with some simplification and define a system as follows. A system con-
sists of a set M, and relations on M, R. Two conditions have to be satisfied
by M and R: (1) M contains at least 2 elements and (2) from every member
of M there is a path to every other member of M. For example, if a influ-
ences the behaviour of b, b influences the behaviour of c, and c influences the
behaviour of a, then we can say that there is a path from a to c through b, a
path from b to a through c, and a path from c to b through a (for a graphical
representation see Figure 4.2). The definition has the important implications
that [1] if an element affects parts of the system but is not affected by it,
then it is outside the system, and [2] if an element is affected by parts of the
system but does not affect any part of the system, then it is outside the sys-
tem, too (Backlund 2000: 448). For example, if elements d and e are added
to the previous example (Figure 4.3) so that the behaviour of d influences the
behaviour of a (but d is not connected to any other element of the system)

istvn czachesz


Contract Contract
Figure 4.2 Graph representing a system.

and the behaviour of b influences the behaviour of e (but e is not connected to

any other element of the system), then neither d nor e belongs to the system.
Applying the analogy that we have used above, if the vertices of the graph in
Figure 4.2 were shops and the directed edges were one-way streets, it would
be possible to reach the shops a, b and c from every other shop, but shop d
could not be reached from any other shop and we would be stuck once we
have reached shop e. The graph represents a system only if every shop can be
reached from every other shop.
Using graph theory, Igor Gazdk (2006) has extended Backlunds defini-
tion of systems. Graph theory can be used to analyse systems in many differ-
ent ways, such as studying subsystems or identifying elements that are crucial
to the functioning of the system. Vertices of a system can be connected by
more than one edge, or edges can connect more than two elements (hyper-
graph), which helps to capture the complexity of interactions within a system.
The concept of a graph is analogous, for our purposes, to the concept of a
network (in that case we might speak of connected nodes rather than vertices).
Modelling religious systems as graphs (or networks) will enable us to apply
insights from graph theory (and network theory) to the study of religion: for
example, we will be able to answer questions about the systemic nature of
beliefs by employing such a method.


Contract Contract

Figure 4.3 Vertices d and e are not parts of the system.

the evolutionary dynamics of religious systems

Religious systems

Now we can give a definition of a religious system a simple one to start with.
Let the system consist of two vertices, V1 = beliefs and V2 = artefacts, con-
nected by two edges, E1 directed from V1 to V2 and E2 directed from V2 to
V1, as shown on Figure 4.4.

Contract Contract

Figure 4.4 The simplest (religious) system.

This system is the simplest one possible: it contains the minimally required
two vertices, and there is a path from any vertex to any other vertex. It
involves bidirectional interaction between religious beliefs and artefacts: reli-
gious beliefs, such as beliefs about gods, spirits, objects and places, facilitate
the creation of artefacts, such as texts, objects, architecture, instruments and
performances. Artefacts, in turn, generate beliefs in the minds of people who
use them. The interaction between beliefs and artefacts is somewhat similar
to the interplay of internal and external (or private and public) representa-
tions described by Dan Sperber (1996: 7797), but artefacts in this system do
not need to represent (or stand for) beliefs in any direct way (and vice versa).
Further, the concept of a religious system is often used in the sense of a sys-
tem of religious beliefs (properly a set of religious beliefs) about gods, spirits
and the like. The religious system under discussion is not to be interpreted as
a system of beliefs that is secondarily manifested in pieces of religious art. All
four components of the system, that is, artefacts, beliefs and the two edges
connecting them, are equally important, and after removing any of them the
system would cease to exist. Numerous examples of how this system functions
in practice can be mentioned from various religions. Consider the European
Reformation, where changes in beliefs resulted in spectacular changes in the
production and use of religious artefacts, including Church interiors, music,
Bible translations, and literature. The new set of artefacts had a further effect
on religious beliefs. Another example can be taken from Harvey Whitehouses
(1995) ethnography of the Pomio Kivung movement in Papua New Guinea,
where the new ideas of the splinter group resulted in the construction of a
new community building (the round-house) in ancestral style, which then
served as the spot of various dramatic events and the (trans-)formation of
beliefs within the group.
There are three questions in particular that we will ask about our simple
religious system in the rest of this section:

istvn czachesz

1. Is it meaningful to break down the system to further components?

2. Are there components that have to be added to the system?
3. Are there factors outside of the system that are relevant for its functioning?

Let us start with the first question.

Is it meaningful to break down the system to further components?

The plurals artefacts and beliefs already suggest that both of these com-
ponents can be broken down into further components. Although our system
remains functional if it contains only a single religious belief, practically it is
hard to imagine any religious belief that would not be connected to several
other religious beliefs. For example, believing that a person or object has
supernatural qualities or powers implies many other beliefs about the ori-
gins, history and possible future consequences of that quality. Such a belief
is likely to be shared with other people, resulting in various artefacts, such
as texts and pieces of art, which in turn generate similar religious beliefs in
other people.
We can reasonably argue that at least some beliefs are connected to each
other. In such a case, beliefs and connections between them can be repre-
sented as networks. A simple way to represent networks of beliefs is a word
association network, where two words are connected if people associate them
with each other. For example, if you enter umbrella into the online query
form of the Edinburgh Word Association Thesaurus (Science and Technology
Facilities Council undated) as a stimulus, the database will tell you that 60
per cent of people associated it with rain, 8 per cent with stand, 4 per cent
with black, and so on, the total number of words associated with umbrella
being 22. In terms of graph theory, umbrella is a node of the association
network that is connected to 22 other nodes (in other words, it has a degree
of 22). If you now enter rain, you get 41 associated words, starting with
snow. You can again pick any of the 41 words and continue to explore the
network. It is also possible to experiment with words connected to religion.
The stimulus religion, for example, will be associated with God, Church,
Christian, faith, Catholic, sex, belief, Bible, book, Church of England,
Christianity, cross, Jesus, Jew, just to mention the top of a list of 43 items.
More sophisticated networks that connect concepts include association net-
works that distinguish various types of associations (such as synonyms and
antonyms) and semantic networks, where concepts are connected by different
kinds of relations such as has or is a; for example parrot in a semantic
network is connected to bird by the relation is a and to feather by the
relation has: a parrot is a bird and has feathers.

the evolutionary dynamics of religious systems

Does the fact that beliefs form networks mean that they also form systems
or a system? Not necessarily. All systems can be represented as graphs (or
hypergraphs), but not all graphs represent systems: for example, if we take
away the edge between vertices c and a from the graph shown in Figure 4.2
(or turn around its direction), the resulting graph will not contain a system
any more. It is easy to see that at least some beliefs affect the behaviour of
other beliefs. For example, priming the brain with one belief can facilitate
the recall of another belief. Association networks can be mapped out by a
similar procedure: after hearing or reading a word, associated words come
to mind more easily than other (non-associated) words. A little more experi-
mentation with the Edinburgh Word Association Thesaurus shows that the
stimulus God yields the response Christian and vice versa. Each of these
two vertices thus influence the behaviour of the other vertex, and therefore
they constitute a system. Is there a way to know whether the association net-
work contains also larger subsystems, or whether the whole network forms a
system? Given a set of data, this is an empirical question. There are different
methods to analyse networks and gain insights about their components, of
which I will give some examples in the final part of the article. Probably some
of these methods could be adjusted to find systems that fulfill the definition
given in the previous section but unfortunately we have no ready-made
tools that would give an immediate answer to the question. Recently Gergely
Palla et al. (2007) have analysed communities in directed association net-
works (Nelson et al. 1998), but they have excluded from their search cycles
like {a, b, c} in Figure 4.1. Yeon-Mu Choi and Hyun-Joo Kim (2007) have
published an interesting study of directed connections among mythological
figures in Greco-Roman mythology. Whereas this study reveals interesting
facts about the network of mythological figures, the directions given to the
edges is not very helpful in deciding on the systemic nature of beliefs (edges
in Choi & Kims study reflect cross-references between lexicon entries). It is
likely that all beliefs belong to some larger, interconnected network (that is,
they are connected at least to some other beliefs), and based on the samples
that we have examined we can also hypothesize that they also form systems,
many of which are probably connected into larger systems (suprasystems). Yet
we cannot say with absolute certainty at this point that all religious beliefs in
a religious system form a belief system or are parts of a belief system.
Also artefacts are more than just collections of items on a desk (referring
again to DAndrades analogy). Some artefacts evidently form a system. The
elements of any architectural construct are good examples: removing pillars
from a cathedral will change the distribution of forces in the structure, even-
tually resulting in the collapse of the building. Even the much more modest
round-house in Whitehouses (1995: 13) ethnography is a system rather than
a collection of components. We can, however, remove altars, organs, statues,

istvn czachesz

and many other components from a cathedral, as indeed the Reformation did,
without affecting the building from the static point of view. One might object
that the artefacts that we have removed change the function of the building:
in order that it can be used for a Catholic mass, for example, it needs to have
an altar. Yet in this case it is the religious system that changes, rather than the
system of artefacts. Religious systems thus seem to contain subsystems of arte-
facts, but this includes artefacts that do not form a system among themselves.
Some artefacts that are not parts of a system of artefacts still can be thought
about as parts of a network of artefacts. Many of them are parts of particular
configurations and cannot be found apart from that configuration normally.
It can be said that networks of artefacts are often themselves artificial, in the
sense that their arrangement is not due to natural laws but cultural conven-
tions. Thus networks of artefacts are themselves artefacts. Whereas in the brain
beliefs are represented on a neural network that is determined by anatomic
structures as well as by electric and chemical systems of communication,
artefacts are not normally interconnected by such constraints and systems of
communication. However, there are at least two important arguments for the
view that the principles that underlie the organization of artefacts and beliefs,
respectively, are not all too different from each other. First, the way networks
of artefacts are formed and maintained in the framework of cultural systems
is comparable to the creation and maintenance of belief systems. One may
object that most artefacts belong to networks and systems only because beliefs
in our minds establish such relations among them. It can be argued, however,
that many beliefs as well as connections among them that are represented in
our nervous system can only be maintained because a certain configuration
of artefacts exists in our environment. We live in a cultural environment that
provides our mental representations with indispensable scaffolding, without
which many of the beliefs represented in our brains would diminish. Second,
in (post-)industrial societies more and more artefacts also form physical net-
works and interact by using their own language of communication. The arte-
facts that surround us are becoming increasingly systemic. In sum, as was the
case with beliefs, one may hypothesize that many artefacts belong to intercon-
nected networks, in a way that is quite comparable to the networks of beliefs,
though certainly not all such networks are systems.
How can we include networks and systems of beliefs and artefacts in our
initial model? Subsystems have to be connected to suprasystems by at least
two edges (one directed from the subsystem toward the suprasystem and
another one directed from the suprasystem toward the subsystem) in order to
ensure that the behaviour of the subsystem both influences the behaviour of
the suprasystem and is influenced by its behaviour. For two networks to be
connected, a single edge between them is sufficient. Networks and systems of
beliefs and artefacts are obviously connected in complex ways. Various beliefs

the evolutionary dynamics of religious systems

Contract Contract

Contract Contract

Contract Contract
Figure 4.5 Random example of beliefs and artefacts.

can be attached to various artefacts, irrespective of how those beliefs and arte-
facts are connected among themselves. A random example is shown in Figure
4.5 (the direction of edges among artefacts and the direction of edges among
beliefs is ignored). Actual networks can be extremely complex, and cannot be
analysed without the assistance of specialized software. However, we might
be able to make some general suggestions about networks based on empirical
research that has been previously done in various fields.

Are there components that have to be added to the system?

Our second question concerned the addition of further components to the

system. Let us start with emotions, which have played an important role in
recent cognitive theorizing about religion (e.g. Pyysiinen 2003: 13042).
One possibility would be to add emotions as a further subsystem, which is
connected both with artefacts and beliefs. It seems questionable, however,
that emotions form a system or even a network in the technical sense, as the
concept is used in this article. Another possibility is to add emotions to the
system as a set of beliefs. This approach is supported by the fact that both
beliefs and emotions are implemented in the brain even though emotions
are closely connected to the archaic brain and somatic processes (Ward 2006:
30935). A third possibility is to think about emotions as processes that act
on the components of the system. Emotions can result in the creation of
beliefs as well as they strengthen or weaken, establish or delete links among
beliefs and between beliefs and artefacts. To use a somewhat banal example,
love creates links between the mental representation of the beloved one and
mental representations of a number of pleasant things in the world. A fourth
possibility is to include emotions both as processes and beliefs. The con-
cept of mental representations of pleasant things in the previous suggestion
requires that emotions are understood not only as processes but also as per-
manent (components of ) mental representations. Without further discussing

istvn czachesz

the problem at this point, I suggest that it is helpful to make a difference

between emotions as processes (playing an important role in rituals, which
we will briefly discuss in the final part of the article) and lasting mental repre-
sentations of emotions that can be regarded as parts of the network of beliefs.
So far we have discussed religious systems as networks of beliefs and arte-
facts. But one can object that religious systems in reality do not look like that:
they rather consist of people acting and interacting in particular ways, per-
forming rituals and participating in different kinds of joint action. Evidently
people hold beliefs but whose beliefs are we interested in? Is it the beliefs of
the priest or of the congregant that are included in the network? The religious
system under discussion is a model that enables a particular representation
of data in order to make it possible to gain insights about religion. Models
are always limited in their scope, and so is our religious system. A major
simplification involves the representation of types rather than of instances
of beliefs. Network models used in scientific research typically have to make
a choice between representing either types or instances (Santos et al. 2007).
For example, mapping out the interactions of all animals living in a habitat is
practically impossible: you might be able to follow a family of lions day and
night, but hardly every single animal with which they interact. Ecological net-
works therefore include species (types) rather than single animals and plants
(instances). Now consider the task of mapping out a terrorist network (e.g.
Moon & Carley 2007). In that case, we want to know a social network of
individual terrorists, that is, of instances. Association networks and semantic
networks are dealing with types, mapping out an average version of beliefs
that several individuals hold. Theoretically, we could also include different
types of religious beliefs in a single system: for example, an expert type
and a lay type could be defined where experts try to change the belief sys-
tems of lay people by interacting with them and a range of artefacts (such
as religious texts). Such complex interactions could be studied with the help
of computer modelling1. For the time being, however, we do not pursue the
possibility of such models, and regard social networks as another dimension
of the religious system.
In spite of the fact that social networks have been studied for decades, what
they are and how we can measure them is far from being self-evident. In social
networks, a link between two people is thought to exist if they interact with
each other (with some regularity or in particular ways). Sometimes, however,
we rather want to know what kind of beliefs people maintain about each
other (e.g. when we draw a sociogram based on a questionnaire). In the study
of social networks, it is usual (Krackhardt 1987) to make sharp distinction
between cognitive (the latter type) and behavioural networks (the former
type). Whereas it is useful to make such a distinction in empirical research
(e.g. Pittinsky & Carolan 2008), we have to recognize that behavioural

the evolutionary dynamics of religious systems

networks are based on and closely related to cognitive networks. In fact,

all social networks that are manifested in interactions among people materi-
ally exist as beliefs and artefacts: people hold particular beliefs about each
other and typically they interact by means of artefacts. Whether two people
interact with each other is largely determined by the beliefs they hold and/
or the artefacts that make interaction possible, such as being friends or allies,
living in the same street or village, etc. Social interactions, in turn, can result
in the modification of beliefs and artefacts, such as making new friends or
sending postcards. Most importantly for our purposes, social networks influ-
ence how beliefs are created and changed. Social networks have a number of
features that allow us to make predictions about how beliefs will spread in a
population, and whether they will disappear or survive in the long term. Also
this feature of social networks is reciprocal: there are beliefs (also religious
beliefs), that can directly influence the formation of edges in social networks.
For example, if one follows Christianity and obeys Jesus command to love
ones neighbour as oneself (Matthew 21:39, based on Leviticus 19:18), this
evidently leads to the creation of social ties that shape social networks in ones
Finally, we can ask whether social networks related to religious systems
form subgraphs (or subsystems) of social networks in general in a similar
way as religious artefacts and beliefs form subgraphs and subsystems of a cul-
tural system (Figure 4.4). It seems difficult to give a general answer to that
question. In contemporary Western societies, religious communities can be
relatively easily distinguished from other communities and the rest of soci-
ety but this is hardly the case in various traditional societies, where religion
permeates all aspects of life. An ancient Greek city, for example, was simulta-
neously a political and a cultic community, and this also seems to be true for
todays pre-industrial societies. A closer look at a Christian or Jewish congre-
gation in a modern, secular state will also reveal that social relations outside of
the domain of religion still very much influence the formation of a religious
congregation and vice versa. It might be the case that social connections and
networks related to religious systems are so tightly integrated into larger social
networks that we cannot map out and separately study a subgraph or subsys-
tem of social networks specific to religion.

Are there factors outside of the system that

are relevant for its functioning?

The third question that we have raised in this section concerns the role of
external factors: are there factors outside of the system that are relevant for
its functioning?

istvn czachesz

First, religious beliefs are only a subset of all beliefs that we maintain and
religious artefacts are only a subset of all artefacts people produce. We can
hypothesize that a relation similar to the one found between religious beliefs
and artefacts exists between beliefs and artefacts in general, and therefore
a cultural system can be defined in a similar way as we define a religious
system. We can further accept (without further analysis at this time) that
religious systems influence the behaviour of cultural systems and vice versa.
In that case, we can say that a religious system is a subsystem of a cultural
system. A system that interacts with another system that is not its subsystem
is called an open system (Backlund 2000: 450). Religious systems are there-
fore open systems: they interact with cultural systems, which are not their
Second, it is obvious that producing any artefact, even a humble arrow or
stone tool, costs energy. Producing and maintaining networks and systems
of artefacts (such as building and decorating a sanctuary) can be extremely
costly. But also acquiring and maintaining beliefs is not without costs: the
weight of the human brain makes up only two per cent of the total body-
weight, but the brain consumes twenty per cent of the bodys energy. Since a
religious system can hardly generate the resources it needs (religious artefacts
cannot produce energy, they cannot be eaten), resources must come from an
external system, particularly from food production. Food production can be
regarded as another subsystem of cultural systems. It is also probable that
religious systems influence beliefs that also play a role in food production and
vice versa (food taboos, gods and spirits connected with vegetation, influence
on productivity etc.). Religious systems and food production are therefore
two interconnected subsystems of cultural systems (Figure 4.6).


Let us now add a temporal dimension to our study of religious systems. At

any given point of time, the system can be described by describing its vertices
and edges. We can examine, for example, how many vertices the system has,
what kind of vertices they are (beliefs, artefacts or their possible subcatego-
ries), how many edges each vertex has on average, how edges are distributed
among vertices, and so on. Then the dynamics of the system can be defined
as the change of such variables as a function of time. If we repeatedly exam-
ine a religious system as time goes on, we can expect to find that some new
beliefs and artefacts are added to or removed from the system, and new rela-
tions among them appear or existing relations disappear. The dynamics of a
religious system relies on the interactions that have been outlined in the pre-
vious section. At the level of instances (from the viewpoint of the individual

the evolutionary dynamics of religious systems

believer) the system is in perpetual motion: beliefs (mental representations)

facilitate the production of artefacts (external representations) and artefacts
generate beliefs. People are born and die, learn and forget, and make new
artefacts, which can also perish with time. If we take a look at the dimension
of social networks, we can see that connections may change all the time and
nodes (individuals forming social networks) can be replaced. At the level of
types, however, this might remain unnoticed, similarly as the fate of indi-
vidual animals or plants does not normally change the overall picture of an
ecological system: lions and antelopes die or migrate to new territories, yet
the ecological system can remain unchanged. Such a view of the religious
system corresponds to the observation that most real-life religions around us
seem to operate in a fairly constant and continuous way, even on a historical
How can we explain the seemingly unchanging character of religious
systems around us? It has been suggested that religious beliefs occupy ideal
or close to ideal positions in the space of possible beliefs, particularly due
to their minimally counterintuitive structure (Boyer 1994: 48, 121; Barrett
2008) that makes them memorable as well as it allows the mind to make
rich inferences from them. Following Dan Sperbers (1996) use of the system
theoretical notion of attraction, it has been suggested that religious beliefs
occupy attractor positions. One might suggest that if religious beliefs are
(nearly) optimal beliefs, this will guarantee the stability of religious systems
(cf. Sperber & Hirschfeld 2004). For the purposes of this article, we can
ignore the debate that has developed around experimental findings about
counterintuitive ideas (Boyer & Ramble 2001; Barrett & Nyhof 2001; Atran
2002: 100107; Norenzayan & Atran 2004; Gonce et al. 2006; Upal et al.
2007), and accept that some form of counterintuitiveness does contribute
to the memorability and stabilization of concepts. Would this guarantee the
stability of religious systems? A comparison with ecological systems might
be helpful at this point. All species that live on earth today evidently evolved
to occupy some kind of optimal position in their environment: this follows,
after all, from the mechanism of natural selection. This optimum, how-
ever, is relative to (a) the natural conditions of its habitat (such as climate
and landscape; cf. Mayr 2001: 1523), (b) the ecological system to which
it belongs (e.g. it can optimally predate on some species and defend itself
against other species) and (c) the options that were left open by its previous
developmental history (e.g. being a fish, bird or mammal; cf. ibid.: 14043).
Speaking of beliefs, their (assumed) optimality has been achieved relative to
a similar set of constraints: (a) religious beliefs have developed in a natu-
ral habitat, which includes the natural environment and human anatomy;
(b) they are part of an ecological system, consisting of other elements
of culture; and (c) finally, their development is determined by previous

istvn czachesz

developmental stages (although cultural bits might change more freely and
rapidly than do biological species; cf. Sperber 1996). Especially important
for our purposes are the first and second criteria: if we assume that there are
attractor positions toward which beliefs tend to develop, such positions are
relative to cultural niches (consisting of natural environments and cultural
systems). But in that case cultural niches cannot remain unchanged, since
that would preclude the change of the species that constitute them. The
behaviour of each part in the system influences (to different degrees) the
behaviour of the whole system and therefore the behaviour of every other
part of the system. In other words, both ecological and cultural systems
display complex behaviours, which cannot be obviously understood from
the properties of their parts (Nicolis & Prigogine 1989; Chu et al. 2003;
Mainzer 2004).2 We cannot go into more details about cultural complexity
at this place (Denton 2004; Czachesz 2007), but have to notice that most
real-life systems are complex systems.
We will therefore consider the problem of stability and change from the
perspective of the study of complex systems, which will also provide us with
clues about the expected changes and possible evolutionary trajectories of
religious systems. There are three different kinds of states in which systems
can remain for a longer period of time (Nicolis & Prigogine 1989: 66). The
first one, mechanical equilibrium, is only possible in so-called conservative sys-
tems, which preserve their total energy, translational momentum and angular
momentum (ibid.: 4650). An example of a conservative system is a pendu-
lum operating under the idealized conditions of classical mechanics, which
ignore friction. A pendulum has two equilibrium states: an unstable equilib-
rium at the highest point of its path and a stable equilibrium at the lowest
point. The pendulum can rest at its stable equilibrium or move periodically
without any change as long as it is left alone. If its motion is disturbed, it
will stop or switch to a different amplitude (ignoring the case of an over-
damped system). If a religious system were conservative, we would be able
to observe an endless generation of beliefs from artefacts and vice versa. This
would, however, require that it remains uninfluenced by its (cultural and nat-
ural) environment. As soon as the slightest influence from the environment
occurred, however, the system would not be able to return to its previous
state. Such a religious system would be completely at the mercy of its envi-
ronment. Since the environment inevitably impacts religious systems all the
time (as Figure 4.6 shows), if their dynamics were analogous to the dynam-
ics of conservative systems, their behaviour would be constantly and directly
reflecting changes in their environment.
Most real-life systems are dissipative (Nicolis & Prigogine 1989: 5054):
for example, friction is part of real mechanical systems. A dissipative system
can be an isolated system, one that does not exchange matter or energy with

the evolutionary dynamics of religious systems


Contract Contract

Figure 4.6 Religious and cultural systems.

its environment. Such a system will irreversibly evolve toward a second type
of stability, the final state of thermodynamic equilibrium, in which parts of a
system have achieved their maximum level of disorder. Religions do not seem
to be in such a state. Again, we can refer to Figure 4.6, which shows that our
religious system is interacting with a complex environment. Religious sys-
tems are therefore best studied as dissipative, non-isolated (or open) systems.
Such systems can exchange matter and energy with their environment, and
have two kinds of stable states. The first option is to reach a thermodynamic
equilibrium (our second type of continuous state; see above), in which the
system stops exchanging energy or matter with its environment, similarly
as an object that has reached the temperature of its environment and stops
exchanging heat with it. Using the notion of thermodynamic equilibrium
analogically, rather than in a technical sense, we can imagine that a religious
system is in equilibrium with its cultural environment in some ways. A reli-
gious system might, for example, share the belief system of its cultural envi-
ronment and vice versa. Differences in beliefs between religion and the rest
of the cultural system have not been much of a concern, by and large, until
the arrival of modernity. In contemporary Western theology, in contrast, an
exchange of ideas with science, different worldviews, and cultural theories
is a major source of the dynamics of the belief system of religion. To take
another example, a medieval monastery might have been independent to a
large degree from its cultural (but not its natural) environment in terms of
food and material needs. Yet such examples are very limited both in extent
(e.g. monks did depend on artisans, building materials, supply of novices)
and scope (e.g. monasteries interacted with culture and the rest of the reli-
gious system in many ways). The notion of thermodynamic equilibrium in
open systems is thus also of limited use to understand stability in religious
Another option for dissipative, non-isolated systems to acquire stability
for some time is to be in a stationary non-equilibrium state (the third type of
continuous state in our discussion). In such a state, the system does exchange

istvn czachesz

matter or energy with its environment, but its internal and external param-
eters are related in such a way that the system enjoys some level of stability.
If matter on Earth were in a chemical equilibrium (which is a component
of thermodynamic equilibrium), for example, 99 per cent of its atmosphere
would consist of carbon dioxide and the salt content of its oceans would
be 13 per cent the current figures being 0.03 per cent and 3.5 per cent,
respectively (Lovelock 1979). The biosphere of Earth, like many real-world
systems, operates far from thermodynamic equilibrium: the influx of energy
from without the system (from Sun) makes the persistence of such a non-
equilibrium state possible. The biosphere has reached its current state through
a long process of complex self-organization (balancing myriads of interde-
pendent effects). Now we can apply this model to human culture, which has
gone through another amazing process of self-organizing, in which mental
processes and artefact production can be seen as two major factors. Religion
occupies an interesting position within culture, because of its continuous
existence in the midst of cultural change. How can we explain this continuity
and how can we account for change in religious systems?
Religious systems are complex, self-organized systems, rather than engi-
neered systems with a straightforward structure and easily detectable subsys-
tems (modules) that have clearly distinguished functions. It seems reasonable
to think that culture has developed similarly as did biological and ecological
systems, that is, through a long history of evolutionary tinkering, reusing
existing bits and pieces and adding new ones, often resulting in redundancy
or seemingly dysfunctional units. It is questionable if parts of culture such
as religion can be rightly called spandrels (Gould & Lewontin 1979;
Atran 2002: 435). Spandrels (ornaments filling up space in gothic architec-
ture) are clearly not part of the system of gothic architecture from the static
point of view: they are connected to the system by outgoing edges (they exert
gravitational force on some elements around them) but do not have incom-
ing links (gravitational forces are not distributed to them). Religious systems,
in contrast, are parts of cultural systems, connected to them by both incom-
ing and outgoing edges. (It has to be noticed that although spandrels are not
parts of the static system of buildings, they are certainly parts of the cultural
system of medieval culture.) What we think about the adaptiveness or evolu-
tionary path of religion is a completely different matter: naturally developed
subsystems may contain some level of redundancy (possibly serving as backup
systems, as it were) or their function might remain largely hidden, due to the
complexity of the system, including unclear boundaries between modules and
circular interactions among individual parts and the whole system.
In sum, religious systems can be thought about as subsystems of culture
that inhabit particular niches of cultural landscapes. Its remarkable continu-
ity through history in the midst of cultural change does not mean it is in

the evolutionary dynamics of religious systems

an equilibrium state, but can be rather characterized as a stationary non-

equilibrium state (that is, a state not showing radical changes on a given
timescale). Yet such states are never actually steady. In real-life contexts, the
environment communicates matter, momentum or energy to a system all
the time in cultural systems we might consider such influences at a more
abstract level and it is practically impossible to control for every state vari-
able of the system with unlimited precision. (State variables such as tempera-
ture or pressure can be used to characterize states of thermodynamic systems;
for cultural and religious systems other variables have to be identified.) Also,
quite independently from the environment, complex systems all the time
exhibit small deviations from the reference state (called fluctuations or back-
ground noise). The continuous state (or reference state) of a system can be
therefore interpreted as the most probable state within a range of continuous
perturbations (that is, local, small-scale changes). There are various possible
reactions of a system in response to perturbations (Nicolis & Prigogine 1989:
6671), of which we only mention a few important options at this place.
Sometimes the system remains all the time in a given vicinity of a continuous
state. In other cases, perturbations initiate a sequence of states, through which
the system returns to the reference state (orbital stability). It is also possible
that in reaction to perturbations the system steadily approaches a reference
state (asymptotic stability), which in such a case can be called an attractor.
Attractors can be local, when the system returns to the reference state only
if the perturbation is below a certain limit, or they can be global, when the
system will return to the attractor after any perturbation, regardless of its size.
Finally, perturbations might grow rapidly and drive that system away from
the reference state (which is then called and unstable reference state).
If we apply these different classes of system behaviours to religious sys-
tems, we gain a new perspective on continuity and innovation. Whereas per-
turbations in engineered systems are normally negative, undesirable effects,
in self-organized systems they are essential for the development of the sys-
tem. Cultural stability is desirable for a certain amount of time, but being
bound by an attractor forever would be detrimental for human culture, since
it would prevent us from experimenting. In such a case the only way to facili-
tate cultural change would be radical impulses from the environment such
as a major climate change for which we would be then completely unpre-
pared. It seems that human culture is staying near to reference states with
considerable perturbations, which occasionally result in abandoning the old
reference state and the emergence of a new reference state (such as occurs in
major historical transitions).
Within this general framework, religion seems to occupy a special posi-
tion. In spite of the major cultural transitions of European culture in the
last two millennia, the two major religions within this culture, Judaism and

istvn czachesz

Christianity, have remained almost unchanged since antiquity. We have

already rejected the idea that the assumed stability of religious beliefs could be
responsible for the stability of religions. Can we perhaps adopt a milder form
of this hypothesis? In spite of the great variety of cultures, there is a good deal
of cross-cultural agreement among religious beliefs. It seems indeed reason-
able to think that cross-cultural factors such as anatomy or the natural envi-
ronment will drive the evolution of beliefs or components of beliefs toward
some attractor positions. Artefacts are no exceptions: anatomy, purpose and
environmental constraints define attractor positions for their development.
This does not mean, however, that all beliefs and tools globally end up in the
same attractor positions. The evolution of beliefs and artefacts, as well as of
cultural or religious systems in general, is constrained by their history (simi-
larly as we have noticed about biological evolution). Once you have a tool,
options for further improvement are restricted by its current shape. Another,
somewhat related, issue is the effect of local attractors. Local attractors might
trap a system if perturbations practically never achieve a level at which the
system would not return to its reference state anymore but would start to
migrate toward another attractor, preventing the system from arriving at a
new reference state, which might be more advantageous from some perspec-
tive. Of course such a concept already presupposes that we have criteria for
prioritizing various states of a system. In the case of cultural systems one can
think of some ideologically set preferences, such as cultures favouring social
justice or individual freedom, but there are also more general criteria, such as
stability or evolvability (the capacity of finding better attractors).
Modelling religious systems as networks allows us to study stability and
change in them with the help of parameters that describe networks. At this
point we can only give some tentative examples, but mapping out the net-
works with the help of computerized tools will enable us to make quantita-
tive measurements and predictions. In our religious system, beliefs constantly
generate artefacts and artefacts generate and modify beliefs. Rituals, to which
we have paid little attention in this article, obviously play an important role
in this process. They can be introduced to the network model as hyperedges,
that is, edges that connect more than two vertices (beliefs or artefacts). Rituals
may rearrange edges in the religious system, influencing beliefs, artefacts or
social networks. If we assign weights to the edges (in our imaginary net-
work connecting shops this would be comparable to streets having different
numbers of lanes), we can think of changes that involve the weakening or
strengthening of connections. Since rituals are themselves represented in the
religious system as beliefs and artefacts, rituals can also modify rituals. It is to
be expected that such interactions occur within the system in ways that main-
tain the stability of the system in the midst of environmental influences. For
example, aging is compensated by transmitting beliefs to a new generation

the evolutionary dynamics of religious systems

of believers; events that negatively or positively affect the community are

interpreted in ways that help to maintain the belief system (e.g. attributing
changes in the environment to benign or hostile gods and spirits). It is also
obvious that the religious system will show perturbations due to internal and
external factors. For example, environmental effects will be changing and
beliefs will have a natural tendency to fluctuate. At this point, the system will
follow one of the developmental trajectories that we have described in the
case of thermodynamic systems, ultimately either remaining in the neigh-
bourhood of its previous, continuous state, or migrating toward a new one.
Modelling religious systems as graphs provides us with possibilities to make
observations and predictions about their evolution. Graph theory and the
study of real-life networks in various disciplines (as mentioned in the intro-
duction of this article) have established various regularities with respect to
how networks behave across different domains. Quite interestingly, observa-
tions about networks in one domain often prove themselves to be relevant for
another domain: graphs are capable of modelling the organization of things
in a fairly universal way. Let us take the concept of the diameter of a graph,
which is defined as the maximum of the shortest paths between any two
vertices in the graph (Berge 1973: 66): for any two vertices in the graph, we
calculate the length of the shortest path between them (that is, their distance);
then, the diameter is the length of the longest of these shortest paths, which
gives the distance of the farthest two points in the graph. A famous example
of the practical application of this concept is the hypothesis of six degrees
of separation, which is a more precise formulation of the commonplace that
we are living in a small world. It was not a sociologist, but Hungarian
writer Frigyes Karinthy (1929) who first suggested that any two humans on
Earth are connected by not more than five acquaintances. Three decades later,
Stanley Milgram (1967) formulated the idea of six degrees of separation as
a scientific hypothesis and successfully demonstrated its truth in an experi-
ment conducted in the United States. Networks that have a small diameter
are called small world networks, and one of their important characteristics
is that information spreads in them rapidly (Watts & Strogatz 1998). If reli-
gious systems are small worlds, fluctuations can spread quickly in them,
and fluctuations initiated at various points of the system can combine in
unpredictable ways. A similar prediction can be made about social networks
that are connected to religious systems: if they are small worlds, new beliefs
can spread on them rapidly. At some stages of development, religious systems
indeed show rapid change, whereas most of the time they do not: above we
have made the observation that religious systems show relative stability in
the midst of cultural and environmental change. This suggests that the net-
work structure of religious systems is changing from time to time, regulating
the systems capacity of change (or evolvability). Of course diameter (or the

istvn czachesz

empirically more accessible average shortest path) is not the only variable that
is responsible for the spread of information in a network but the investiga-
tion of those factors has to be postponed to future contributions.

I would like to thank the Helsinki Collegium for Advanced Studies for providing me with
a wonderful research environment to work on the project Religion in Dynamic Systems.
I am thankful to Robert N. McCauley and Benson Saler for their feedback on an earlier
version of this article, presented during the Origins of Religion, Cognition, and Culture
conference. Tams Birs critical reading of the manuscript and his comments on my use
of system theory and mathematical formalism helped me tremendously in formulating the
final version of the text.


1. Agent-based modelling is a computerized method to study such interactions. Whereas

previously most agent-based modelling has focused on the interaction of agents having
a simple cognitive architecture (capable of representing a small set of simple beliefs),
recently more sophisticated models have been designed, for example by Ron Sun (2006,
2. Using a more precise mathematical language, these complex systems are analogous to
non-linear systems, the behaviour of which cannot be understood as a linear sum of the
behaviour of their parts.


Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. Oxford: Oxford Uni-
versity Press.
Backlund, A. 2000. The Definition of System. Kybernetes, 29(4): 44451.
Barrett, J. L. 2008. Coding and Quantifying Counterintuitiveness in Religious Concepts:
Theoretical and Methodological Reflections. Method and Theory in the Study of Religion
20 (4): 30838.
Barrett J. L. & M. A. Nyhof 2001. Spreading Non-natural Concepts: The Role of Intuitive
Conceptual Structures in Memory and Transmission of Cultural Materials. Journal of
Cognition and Culture 1: 69100.
Berge, C. 1973. Graphs and Hypergraphs. Amsterdam: Elsevier.
Bertalanffy, L. von 1950. An Outline of General System Theory. The British Journal for the
Philosophy of Science 1(2): 13465.
Boyer, P. 1994. The Naturalness of Religious Ideas: A Cognitive Theory of Religion. Berkeley, CA:
University of California Press.
Boyer, P. 2001. Religion Explained: The Human Instincts that Fashion Gods, Spirits and Ancestors.
New York: Basic Books.

the evolutionary dynamics of religious systems

Boyer P. & C. Ramble 2001. Cognitive Templates for Religious Concepts: Cross-cultural
Evidence for Recall of Counter-intuitive Representations. Cognitive Science 25: 53564.
Choi, Y.-M. & H.-J. Kim 2007. A Directed Network of Greek and Roman Mythology.
Physica A 382: 66571.
Chu, D., R. Strand & R. Fjelland 2003. Theories of Complexity: Common Denominators
of Complex Systems. Complexity 8(3): 1930.
Czachesz, I. 2007. Theory Forming in Biblical Studies: Contributions to an Interdisciplinary
Dialogue. In Complexity, Interdisciplinary Communications 2006/2007, W. streng (ed.),
558. Oslo: Center for Advanced Study.
DAndrade, R. 2001. A Cognitivists View of the Units Debate in Cultural Anthropology.
Cross-Cultural Research 35(2): 24257.
Denton, T. 2004. Cultural Complexity Revisited. Cross-Cultural Research 38: 326.
Gazdk, I. 2006. Modelling Systems by Hypergraphs. Kybernetes 35(9): 136981.
Geertz, C. 1973. Thick Description: Toward an Interpretive Theory of Culture. In his The
Interpretation of Cultures: Selected Essays, 231 31. New York: Basic Books.
Gonce, L., M. A. Upal, D. J. Slone & R. D. Tweney 2006. Role of Context in the Recall of
Counterintuitive Concepts. Journal of Cognition and Culture 6: 52147.
Gould, S. J. & R. C. Lewontin 1979. The Spandrels of San Marco and the Panglossian
Paradigm: A Critique of the Adaptationist Programme. Proceedings of the Royal Society of
London B 205(1161): 58198.
Jablonka E. & M. J. Lamb 2005. Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral,
and Symbolic Variation in the History of Life. Cambridge, MA: MIT Press.
Karinthy, F. 1929. Lncszemek. In his Minden mskppen van, 8590. Budapest: Atheneum.
Krackhardt, D. 1987. Cognitive Social Structures. Social Networks 9: 10934.
Lovelock, J. 1979. Gaia: A New Look at Life on Earth. Oxford: Oxford University Press.
Mainzer, K. 2004. Thinking in Complexity: The Computational Dynamics of Matter, Mind, and
Mankind, 4th edn. Berlin: Springer.
Mayr, E. 2001. What Evolution Is. New York: Basic Books.
Milgram, S. 1967. The Small-World Problem. Psychology Today 1: 6067.
Moon, I.-C. & K. M. Carley 2007. Modeling and Simulating Terrorist Networks in Social
and Geospatial Dimensions. IEEE Intelligent Systems 22(5): 4049.
Nelson, D. L., C. L. McEvoy & T. A. Schreiber 1998. The University of South Florida Word
Association, Rhyme, and Word Fragment Norms, http://w3.usf.edu/FreeAssociation.
Nicolis, G. & I. Prigogine 1989. Exploring complexity: An Introduction. New York: W. H.
Norenzayan, A. & S. Atran 2004. Cognitive and Emotional Processes in the Cultural
Transmission of Natural and Nonnatural Beliefs. In The Psychological Foundations of
Culture, M. Schaller & C. S. Crandall (eds), 14969. Mahwah, NJ: Lawrence Erlbaum
Palla, G., I. J. Farkas, P. Pollner, I. Derenyi & T. Vicsek 2007. Directed Network Modules.
New Journal of Physics 9(6): 186.
Pals, D. 2006. Eight Theories of Religion. New York: Oxford University Press,
Pittinsky, M. & B. V. Carolan 2008. Behavioral versus Cognitive Classroom Friendship
Networks: Do Teacher Perceptions Agree with Student Reports?. Social Psychology of Edu-
cation 11: 13347.
Pyysiinen, I. 2003. How Religion Works: Towards a New Cognitive Science of Religion. Leiden:
Saler, B. 2001. On what we may Believe about Beliefs. In Religion in Mind, J. Andresen (ed.),
4769. Ithaca, NY: Snow Lion Publications.

istvn czachesz

Saler, B. 2005. Finding Way Religion. Historical Reflections/Reflexions Historiques 31(2):

Santos, F. C., H. Bersini & T. Lenaerts 2007. Networks Regulating Networks: The Effects
of Constraints on Topological Evolution. In Advances in Artificial Life: Proceedings of the
Ninth European Conference on Artificial Life (ECAL 2007), F. Almeida e Costa (ed.), 956
65. Berlin: Springer.
Science and Technology Facilities Council undated. Edinburgh Word Association Thesaurus,
Sperber, D. 1996. Explaining Culture: A Naturalistic Approach. Oxford: Blackwell.
Sperber, D. & L. A. Hirschfeld 2004. The Cognitive Foundations of Cultural Stability and
Diversity. Trends in Cognitive Sciences 8(1): 4046.
Sun, R. (ed.) 2006. Cognition and Multi-Agent Interaction: From Cognitive Modeling to Social
Simulation. Cambridge: Cambridge University Press.
Sun, R. 2007. Cognitive Social Simulation Incorporating Cognitive Architectures. IEEE
Intelligent Systems 22(5): 339.
Upal, M. A., L. O. Gonce, R. D. Tweney & D. J. Slone 2007. Contextualizing Counter-
intuitiveness: How Context Affects Comprehension and Memorability of Counterintuitive
Concepts. Cognitive Science 31: 41539.
Ward, J. 2006. The Students Guide to Cognitive Neuroscience. Hove: Psychology Press.
Watts, D. J. & S. H. Strogatz 1998. Collective Dynamics of Small-World Networks. Nature
393: 44042.
Whitehouse, H. 1995. Inside the Cult: Religious Innovation and Transmission in Papua New
Guinea. Oxford: Clarendon Press.

Art as a human universal: an adaptationist view
Ellen Dissanayake

In order to consider art as a human universal, it is of course necessary to

decide what is meant both by the term universal and by the word art.
Universal may imply that a feature (e.g. art) is untaught and appears spon-
taneously, is latent in all normal individuals, has been invented by all cultures,
or is a product of some people (e.g. artists) that has been important in all
societies. These meanings arise from different assumptions and carry incom-
patible implications.1
Similarly, the familiar one-syllable word art drags behind it a long, shad-
owy train or tail of theory, definition, qualification and contention an
appendage that has become only more elaborated and unmanageable over the
past century. Many unexamined assumptions are tucked into its folds, and
one who looks for universals must begin by carefully sorting through these
beguiling, yet confusing, embellishments.
For example, the word art is often tacitly restricted to the visual arts (e.g.
paintings, sculptures, drawings), especially to fine art and thereby denied
to craft, to decoration and to the artistic efforts of untrained or untalented
persons. A notion of fine art implies that there is a qualitative distinction to
be made between art and non-art, or between good and bad art in other
words, that art is a kind of essence that inheres in some works and is lacking
in others. What comprises that essence? Can it be defined so that one knows
art when one encounters it? Does that essence inhere in arts form or content?
In its function (or non-functionality)?
Frequently the category art is extended to include other arts music,
dance, poetry, literature, drama and their subdivisions which may, like the
visual arts, lend themselves to distinctions of quality (or essence) that make
some music or literature art and other examples of music or literature not
art. To consider art as a superordinate category subsuming several arts requires
that one be prepared to say what characteristics these arts have in common.
What does a symphony have in common with a sonnet, or a folk dance with

ellen dissanayake

a novel or Ming vase, that justifies placing them in one conceptual category?
Is the common denominator to be found in formal attributes, in their func-
tion, or in some other feature? Beauty has been considered by many as a
necessary feature of art, or good art. What about examples of the arts that
are not beautiful?
Such questions and distinctions (about art both as visual art and as a gen-
eral category) have been the subject matter of philosophical aesthetics in the
West for more than two centuries. Although Western aesthetics has been
typically concerned with arts of the western European tradition, a universal-
ist position must include the arts (however defined) of people everywhere. A
worthwhile effort in this vein is that of Dutton (2002), who in the spirit of
Weitz (1956) and Munro (1963), used a family resemblance notion of art,
and made a provisional list of seven characteristics which, in whole or large
part, will apply to the practice of art across cultures and throughout historical
time: expertise or virtuosity, non-utilitarian pleasure, style, criticism, imita-
tion, special focus, and serving as an imaginative experience for both pro-
ducers and audiences.
Such a list is a valiant and useful attempt to delineate universal character-
istics of the arts across cultures, but five of the features (i.e. specialized skill,
styles and rules, critical evaluative language, representation and imaginative
embodiment) characterize examples of nonart as well. Only intrinsic pleasure
(self-reward) and bracketing (special focus) seem more or less restricted to art
or art-like activities (such as play and make-believe, or ritual behaviour see
Dissanayake 1988, 1992).
More commonly in recent decades, many philosophers of art have given
their attention to non-essentialist (and nonuniversalist) matters as if alto-
gether abandoning the possibility of sorting out the confusions inherent in
the subject of the nature and purpose of art. In general, the climate in aesthet-
ics and the arts at the end of the late twentieth century is a vaporous one of
cultural constructivism or cultural relativism (see below) that claims that
anything can be art if one (or an artworld) chooses to see it as such. If this
is the true state of affairs, then looking for universals in art is doomed, for
there is no reason to look across cultures or in the past for something that
can be anything.
My own view of art, which will be described more fully below, emerges
from a naturalistic and specifically Darwinian or adaptationist approach.
As such, it considers art like language or toolmaking to be an inherent
psychobiological capacity of the human species, an evolved component of
human nature that is in some respects untaught and spontaneous and in oth-
ers latent in every individual. To approach art (or any human attribute) evo-
lutionarily requires that one specify not only what the capacity consists of, but
why it should exist at all: what it is for. What did art provide for our ancestors

art as a human universal

for whom it was adaptive? I will claim that only the evolutionary perspective
can satisfactorily establish art as a human universal that is, can suggest why
people universally have, engage in, make or experience art.

What art does: seven views

Before describing my evolutionary or adaptationist view of art, let us first

examine some more familiar notions about what art is for that is, answers
to such questions as what its function is, why people do it, and what art
accomplishes for artists or experiencers of art. Here again there are a number
of complementary, overlapping or incompatible views, each generally stem-
ming from a larger theoretical position.
For example, there is what might be called the theological view, held implic-
itly or overtly in theocratic societies as in the European Middle Ages. Art, in
such societies, reflects the power or beauty and goodness of God or the gods:
it manifests or reveals or gives human access to the divine. Not many scholars
hold a theological position today, but it has characterized perhaps the major-
ity of human societies and is implicit in small-scale traditional societies where
the arts were made and used primarily in religious ceremony or observance.
A sociological or socio-cultural perspective typically considers art to be an
instrument of power whether this power be economic, or of class, gender
or race. Art then reflects, asserts and consolidates privilege or vested interests
of the powerful, whether rulers and the nobility, the church, the state, the
upper classes, the bourgeoisie or white European males. In the historical
past, few questioned what art was it was what the powerful said it was.
Over the last century, and especially in recent decades, a variant of this view,
called cultural constructivism or cultural relativism, has become predomi-
nant in the academy. It considers art to have no objective essence of its own,
but to be simply a cultural construct or label given to objects or practices by
interested parties. High and low art, heathen art, tribal art, good
and bad art all are (or have been) labels assigned by one group to its
own or anothers art. An extreme position of this sort is the institutional
theory of art, which claims that art is what the artworld of dealers, critics
and curators says it is (Danto 1964; Dickie 1974). Another is that art can
be anything and anything can be art (e.g. Andy Warhol and Joseph Beuys,
cited in Danto 1996: 110).
To a Freudian psychoanalytical perspective, art is a product of delusion or
lack a symptom of neurosis. It arises from psychological defence measures
such as sublimation or projection, and serves both in the artmaker and
in the respondent to art as the disguised fulfillment of a forbidden and
repressed (unconscious) wish, a substitute for something else.

ellen dissanayake

Other psychological views consider art more positively as self-expressive or

therapeutic rather than palliative or neurotic. Art is a means to personal indi-
viduation, to creativity and fulfillment. It expresses and communicates mood
and personality, and may aid self-knowledge and self-acceptance, which are
themselves considered to be good things.
Experimental psychological studies approach the arts empirically as stimuli
that humans consider pleasing or beautiful. Such studies investigate aes-
thetic perceptual and cognitive preferences that reveal something about
what people universally like and dislike (e.g. Seashore 1938; Valentine 1962;
Berlyne 1971; Gestalt psychologists). Such studies may have practical conse-
quences for advertisers, interior decorators and fashion designers. They also
contribute to understanding human minds and how they work.2
In anthropological views, art reflects a cultural system. It is an instance
of and repository for symbolic meaning, and embodies and conveys impor-
tant cultural truths to people of that culture. Palaeoarchaeologists who study
human prehistory have also typically claimed, or presumed, that art is an
instance of human symbol-making ability (e.g. Mithen 1996).
As described earlier, philosophical views of art examine longstanding ques-
tions about beauty, quality, taste and judgement the subject area that is
traditionally called aesthetics. The nature and purpose or function of art
are among those questions.
Each of these perspectives on arts nature and purpose is or has been use-
ful at one time or another for addressing various specific problems clinical,
commercial, ethnographic, even interpretive. Many serious and gifted think-
ers have devoted years of their lives to investigating them, and have left valu-
able insights. Nevertheless, none of the approaches can satisfactorily address,
much less answer, the question of whether or not art is a universal.
The cultural relativist view of course eschews the very possibility of uni-
versals, so that we need not mention it again. Yet even those views which
one might assume would pretend to universality fail because they cannot be
applied to the arts of all individuals or all societies. Theological views rest on
their adherents belief in a particular divinity, not in a universal deity that
every human society would accept. The psychoanalytical and psychological
views are inapplicable to cultures or individuals where sublimation and wish-
fulfillment are absent in the arts or where self-expression and creativity are not
fostered or valued. Traditional philosophical aesthetics is based on Western
Enlightenment and post-Enlightenment ideas (e.g. disinterested apprecia-
tion) and concepts (e.g. beauty, taste, a defined entity or essence of art) that
one can show do not apply across cultures. Although anthropological views
are concerned with the arts in a variety of cultures, they generally uphold a
cultural relativist position that emphasizes individuality and uniqueness, and
denies universality in any cultural product.3 By assuming that art is necessarily

art as a human universal

symbolic, they beg (or ignore) the question of what makes an artistic symbol
different from a nonartistic symbol; additionally, they thereby disregard or
deny the possibility that there are presymbolic origins or instances of the arts.
Experimental psychological and neurological approaches do address uni-
versals, but their formulations omit important considerations. Although they
are concerned with universal aesthetic preferences or cognitive capacities
that contribute to art, they are generally silent regarding how these isolated
elements (shapes, colours, musical intervals, motifs, or particular regions of
the brain) are used in actual instances of artmaking and art experience. Actual
instances of art arguably involve something more than a collection of prefer-
ences or capacities (Dissanayake 1998). They have characteristics and effects
that are different from those of any individual component.
In addition to insufficiently sorting through the manifold assumptions
inherent in the term art, or failing to appreciate the complexity of an
individual instance of art or the variety in art practices cross-culturally, the
approaches just described typically ignore the question that is fundamental to
understanding art as a universal: why does art exist at all? Where did it come
from and why? They may tacitly hold assumptions about arts origin and rea-
son for existence that inhere in their particular view, e.g. that art was created
by (their own) God, or that individuals made it up, or that this is the way
brains and minds and societies just are. An adaptationist view is not satisfied
with this complacence.

Art as a universal behaviour: the ethological approach

Because life in modernized societies is so recent in the enormous time span

of hominid or even human evolution, it is misleading to generalize about
human nature or human universals by looking only at the way contemporary
people lead their lives. Working from an adaptationist view, we should keep
in mind that art is likely to be broader than, or different from, common-
sense ideas that emerge from the cultural biases of modern and postmodern
Westerners. That is, art will not necessarily or automatically be such things as
works in museums, products of creativity and self-expression, embodiments
of beauty or anything at all that one chooses to call art. An adaptationist view
will always ask of any offered statement or conclusion about a human activity
(such as art): is this likely to have been the case in ancestral societies? Because
we have no ancestral societies to observe, we will have to ask instead whether
a statement or conclusion is likely to be applicable to small-scale, traditional,
pre-industrial, primarily foraging (huntergatherer) societies whose way of
life is closer than ours to that in which human nature evolved and to which
it was adapted.

ellen dissanayake

As an alternative to the traditional, yet ultimately restricted, perspectives

described in the previous section, I will here use an ethological perspective
to suggest both what art is and what it does its origin, nature, and reason
for existing.4 Unlike the other views, ethology allows one to approach art as
a human universal. What is more, it subsumes rather than nullifies the
other seven approaches.
Developed as a branch of biology over the past half-century, the science
of ethology adopts a Darwinian adaptationist (or evolutionary) view that
humans, like other animal species, have acquired over time, through natu-
ral selection, a congeries of adaptive traits which helped individuals who
possessed these traits to survive and reproduce more successfully than indi-
viduals who lacked them or displayed them to a lesser extent. Ethologists
specifically consider an animal speciess characteristic psychology and behav-
iour like its anatomy and physiology as having evolved to fit, or
adapt, them to a particular way of life. For example, within a particular
family (say, Felidae), some species (lions) have a way of life (on the open
savannah) that promotes sociality, while other species way of life (tigers
in dense jungles) fosters asociality. Not only behavioural systems of social
interaction but of mating, parenting, acquiring food, and defence indi-
vidually called behaviours or behavioural mechanisms evolved to suit
the members of a species to their particular environmental niche and its
required way of life. Behaviours generally require a facilitating environment
in order to develop smoothly but they are inherited, with greater and lesser
degrees of lability in expression, as predispositions. They are not (or are
rarely) mechanically determined.
With regard to art, the ethological approach will ask whether it is justi-
fiable to consider it like language, toolmaking, forming social bonds or
parenting as a particular kind of behaviour that universally characterizes the
human species. By viewing art as a behaviour, the ethological approach
unlike the previously described approaches considers art as a process rather
than as an outcome or product of the process, or a feature (such as beauty)
of that product.
Just as languages, tools, social practices and parenting styles vary from cul-
ture to culture yet all humans are born with a predisposition to speak, make
and use tools, form bonds with others, and care for their young expressions
of the arts may also vary, yet still rest upon universal predilections. Like these
other biological predispositions, art requires cultural facilitation.
Thinking of art as an evolved that is, adaptive characteristic of human
nature provides a new set of criteria to apply when considering claims by
others about its nature or universality. If art is adaptive, it is, by definition,
universal, and (a) there will be evidence of this behaviour at some point in our
ancestral hominid past; (b) it will be observable cross-culturally in members

art as a human universal

of all known societies regardless of their degree of economic or technological

development; and (c) its rudiments will be detectable or easily fostered in the
behaviour of young children. Like other adaptations, (d) art will appear under
appropriate conditions or circumstances.
Also like other adaptations, (e) art will be generally a source of pleasure.
Most people will willingly devote time, effort, thought and other resources to
it, as they do with other adaptive behaviours such as mating, parenting, find-
ing and preparing and eating food, socializing and gaining social acceptance,
talking, seeking out and staying in safe (familiar) surroundings, and learn-
ing information that is useful for their way of life. Additionally, an account
of art as a universal behaviour (f ) will distinguish between its motivation
and immediate effect (the proximate reasons for the behaviour) and its
ultimate or adaptive value, although the motivations will be of emotional
importance to those who engage in it.5
Clearly the ethological suggestion that art is an adaptive behaviour opens a
new pathway for understanding art as a human universal. Let us now examine
what a behaviour of art might be.

Artification: making special

I have shown that earlier views have regarded art as objects, entities, an
essence or a label. None of these approaches is translatable to an ethological
perspective, which will necessarily conceptualize art as a behaviour what one
might call artification. Although European languages do not have a verb to
art, it should not be difficult to understand what this word might mean: to
make something art. But what could this be?
It is easier to conceptualize art as a behaviour if we think of art as music
(chanting, singing, playing an instrument) or performing (dancing, reciting,
miming, acting, telling stories), since these arts take place, like behaviour,
in time. In a similar way, one can also think of the plastic or visual arts as
making, marking, image-making, decorating, adorning (in any medium)
that is, as the process or activity rather than the product or outcome of the
artifying. But it is not immediately evident what if anything all these
activities have in common.
In earlier publications (Dissanayake 1988, 1992, 1995), I offered a com-
mon denominator for a behaviour of art that I called making special. That
is, I claimed that in all art (here artifying) in all times and places, ordi-
nary experience (i.e. ordinary objects,6 materials, movements, sounds, words,
utterances, the surrounding environment, even ideas) is transformed, is made
extraordinary. The notion is congruent with similar formulations by others
e.g. the notion of bracketing mentioned above, or defamiliarization

ellen dissanayake

(making strange) and foregrounding in literary studies (e.g. Shklovsky

[1917] 1965; Mukarovsky [1932] 1964; Miall & Kuiken 1994a, 1994b).7
Based on this characterization of art, I advanced a theoretical position that
suggested how making special would have been adaptive, and could thus be
considered a universal feature of human species nature. In my most recent
work and in the present essay, I refine and extend the earlier position. My
argument has three strands: aesthetic predisposition; emotional investment
(or care); and the invention of ceremonial ritual.

Aesthetic predisposition

In recent work (e.g. Dissanayake 1999, 2000, 2001), I describe universal

features that can be observed in early interactions between human mothers
and their infants. Despite cultural variations, mothers all over the world talk
to their small infants in a characteristically soft, high-pitched, undulant voice
which babies have been shown to prefer to typical adult conversational
speech. Along with special vocal behaviour to infants, mothers (and other
adults) engage infants attention by the use of rhythmic body movements
(touching, patting, stroking, hugging and kissing the baby), exaggerated facial
expressions (long looks, sustained smiles, widened eyes, raised eyebrows), and
characteristic head movements (bobs, nods and wags) in an almost ritual-
ized way. These vocalizations, expressions and movements are often repeated,
sometimes with dynamic variations (louder and softer, faster and slower) in
what can be called a multimedia performance.
Yet it is more than an individual performance. Painstaking analysis of vide-
otaped engagements of mothers and babies show that the pair are interacting
in remarkably close temporal unity responding to each other in subtle yet
precise ways (see, for example, Stern 1971; Beebe et al.1977; Papousek &
Papousek 1981; Beebe et al. 1988; Nadel 1996). The mother varies her pace
and rhythm in order maximally to fit in with the babys emotional state and
to help it achieve equilibrium. The baby in turn responds to the mothers
signals with kicks, hand and arm movements, facial expressions, head move-
ments, and vocalizations of its own often as if participating in a mutually-
negotiated rhythmic beat with complementary dynamics. The pair engage
and disengage, synchronize and alternate, practising their attunement over
the first five or six months of the infants life.
Such behaviour has been shown to have many practical effects for the
babys development of emotional homeostasis (Hofer 1990) and later sociali-
zation (Papousek & Papousek 1979; Schore 1994; Aitken & Trevarthen
1997), language learning (Fernald 1992), cognitive development (Papousek
& Papousek 1981; Trevarthen 1997), and acquisition of parental culture. Yet

art as a human universal

it is rarely pointed out that the very components of the interaction are fun-
damentally aesthetic.
Repetition, patterning, exaggeration, dynamic variation, elaboration and
surprise in visual, vocal and kinesic modalities are used by the arts in
order to gain attention and create expectancy. The performative, temporal,
and dynamic features of mother-infant engagement can be viewed as aesthetic
(or protoaesthetic) elements to which, for adaptive reasons (infant survival and
maternal reproductive success), humans are innately sensitive.8
There are other innately-appealing (or aesthetic) sensory features that
originally appeared in nonaesthetic contexts e.g. bright, true, clear colours,
vigorous or graceful movements, significant motifs (e.g. eyespots and zigzags;
see Aiken 1998; her 1991), or cognitively interesting and satisfying musical
intervals or visual shapes and patterns (as described by Gestalt psychologists
and cognitive neuroscientists). These are immediately attractive or salient to
humans, insofar as they signal beneficial or possibly harmful states or events
(e.g. ripeness, youth, health, strength, danger, interest and cognitive mastery).
Individuals who attended to and valued such signals would have enjoyed
greater survival and reproductive success than individuals who did not. Such
signals would become inherent perceptual and cognitive preferences, as
described by experimental psychologists and cognitive neuroscientists.
I suggest that the innate predisposition to take note of or positively like
such protoaesthetic visual and aural signals, as well as the inborn capacities
and sensitivities that predispose adults to make and babies to respond to the
protoaesthetic temporal and dynamic manipulations that were described above,
existed as a sort of reservoir from which early humans could draw when at
a later point in evolution they began deliberately to artify (to make special).

Emotional investment (care)

Humans, more than any other animals, use their wits rather than their
instincts to address the problems of their lives. For our species, what to do
and how to live are not instinctive, but must be learned. Over the millen-
nia of hominid evolution, the mind increasingly became a making sense
organ: interrelated powers of memory, foresight and imagination gradually
developed and allowed humans to stabilize and confine the stream of life by
making mental connections between past, present and future, or among
different experiences or observations.
Humans could remember good and bad things, and imagine them hap-
pening again. One cost of this awareness of the desired possibilities and inevi-
table unpredictability of life, greater in humans than in other animals, was
uncertainty, even anxiety. I suggest that uncertainty leading to emotional

ellen dissanayake

investment or care was the original motivating impetus for the human
invention of religion and its behavioural expression, art (or artification).
Usually religion and art are treated as aspects of culture, which according
to conventional anthropological theory is opposed to biology. An adapta-
tionist view, however, considers the various components that are called cul-
ture for example (as described earlier), language and toolmaking to be
outgrowths of evolved psychobiological predispositions. Here I view religion
and art similarly as cultural behaviours that originally were based on the wish
to influence the outcomes of circumstances that were especially important,
but uncertain.
In general, cultural knowledge and practices direct our attention to par-
ticular biologically significant things ways to become a competent adult, to
make a living, to rear children and to maintain social relationships. Language
and traditions of toolmaking and subsistence practice are among these
ways. Additionally, our ancestors had to care about the outcome of biologi-
cally significant and valuable events and states that were not always certain
of attainment e.g. assuring or restoring safety, prosperity, fecundity, health
and victory, or successfully dealing with the bodily changes and emotional
concomitants of sexual maturity, pregnancy, birth and death.
Other animals in uncertain circumstances frequently engage in displace-
ment activities or ritualized behaviours whose components are drawn from
ordinary bodily movements used in everyday contexts such as grooming,
locomotion, or nestbuilding (e.g. preening, preparing for flight, or plucking
grass). In the new uncertain context, these ordinary movements become more
stereotyped that is, exaggerated, patterned and repeated. Such ritualized
movements signal to conspecifics that the sender of the signal is agitated
or anxious. They also serve to reduce the tension of the displaying animal
(Tinbergen 1952; Eibl-Eibesfeldt 1988).
I suggest that in uncertain circumstances that did not call for immediate
pragmatic action (that is, were not matters of immediate fight, flee or freeze
responses), our early human ancestors at some point found that performing
repetitious, stereotyped, exaggerated sounds and movements provided some-
thing to do, felt comforting, and ultimately eased tension particularly
when performed jointly among members of a group. I further suggest that
individuals in groups that responded to uncertainty in stressful circumstances
with such practices would gradually have gained survival advantage over those
in groups where each person behaved individually or randomly.
It is important to note that such stereotyped, ritualized behaviours
were culturally invented (not, as in other animals, biologically programmed)
and would therefore vary among groups, even though like toolmaking, lan-
guage, and other cultural activities they were based on inherent predisposi-
tions (for example, body rocking and repetitious rhythmic vocalizations are

art as a human universal

spontaneous self-comforting behaviours of emotionally distressed individuals

and even of captive animals; Charmove & Anderson 1989). Unified group
behaviour, even more than individual activity, would create the illusion that
the disturbing situation was being coped with e.g. Mead ([1930] 1976)
described how the people of Manus huddled together during a frightening
storm and chanted charms to abate the wind.
The mammalian response to physiological and psychological stress is an
adaptive answer to potential or actual physical danger: glucocorticoids and
adrenaline are secreted and help the body to react immediately (Sapolsky
1992; Flinn et al. 1996). Worry and unfocussed anxiety, as well as direct
psychological and social stress, may also provoke the response, which if long-
standing or excessive negatively affects immunity, growth, reproduction, mus-
cle action and cognition.
One of the psychological variables that modulates the stress response is
to have a sense of control or predictability. It is healthier more adaptive
to feel that one knows how to deal with uncertain events. Behaviour that is
controlled that is, patterned, repeated, exaggerated, and performed with
deliberation and care is a physical expression that mimics, and feels like,
psychological control.
Although we cannot observe ancestral humans responding to uncertainty
with stereotyped visual, vocal and kinetic patterns, repetitions and exag-
gerations, some palaeoarchaeologists (Brody 1977; Taon 1983; Taon et
al. 1994; Taon & Brockwell 1995) have found evidence of a conspicuous
increase in artistic activity during periods of environmental stress; McNeill
(1995: 89) notes that preaching and song combined with rhythmic muscu-
lar movement are conspicuous in times of trouble and among distressed

The invention of ceremonial ritual

I have just suggested that the earliest forms of what we today call religion
and art arose together during human evolution as ways to address the inevi-
table uncertainties of life that became increasingly evident to intelligent, fore-
sightful people. In my ethologically-plausible reconstruction, a behaviour of
art may have originated in the psychobiological tendency, in circumstances
of perceived uncertainty and its concomitant psychological stress or anxiety,
to attempt to gain control of desired outcomes by means of controlled pat-
terned, repetitive, exaggerated, elaborated, dynamically varied vocal and
kinetic actions that were a behavioural analogue of psychological coping. As
in motherinfant interactions, such activities may well have been presymbolic
and preverbal to begin with, eventually acquiring symbolic significance.

ellen dissanayake

Today we call such behaviour ceremonial ritual, but we could just as

well call it artification. That is, in ceremonies individuals use protoaesthetic
sensory and cognitive features (which, as described in the previous section,
are innately noteworthy because they were already adaptive in other nonaes-
thetic contexts) in temporally and spatially controlled patterned, repeated,
rhythmic, exaggerated, elaborated, dynamically varied ways. Because ritual-
ized motherinfant interaction had already prepared humans to be especially
sensitive to dynamic temporal and spatial manipulations as a way of creating,
expressing and sustaining emotional accord, further shaping and elaborating
(artifying) of the components of the interaction in visual, vocal and kinesic
modalities would be additionally affecting and effective.
It seems likely that in their origins, the arts of music, dance and mime
would have been performed together as one multimedia activity, as they occur
in motherinfant interactions. Perhaps protoaesthetic visual elements (say, in
body adornment) were added to make the performance even more striking.
Over time, individual arts could be additionally developed and even emanci-
pated from ceremonies, from religion and from uncertainty. Once artifying
in ceremonial ritual became part of an individuals or cultures repertoire, its
various features could be further artified or manipulated and used in a variety
of other, even secular and celebratory, contexts.
Inherent in an ethological view is the premise that culturally created cer-
emonial rituals were biologically adaptive. The fact that we are emotionally and
behaviourally susceptible to elaborated movements in time, visual compelling-
ness, skillful execution and the structuring and manipulation of our sensory
experiences made it more likely that we would engage in the socially reinforcing
ceremonial behaviours, remember the messages that these practices transmitted
and become emotionally convinced of their truth and effectiveness. Without
such biologically adaptive reactions, artification of existing protoaesthetic sig-
nals would not have become an important universal human behaviour.

Artification as a human universal

In conclusion, I will consider my ethological or adaptationist view of art as

a human universal with respect to conceptual and other issues mentioned in
earlier sections of this essay. I will also refer, when appropriate, to the other
seven approaches that were described. Unlike some of these, an ethological
approach includes the arts of people in all societies and all times. That is, it is
not restricted to fine art, and accepts decoration, much craft, the perform-
ing arts and even unskilled or careless examples of all these. Determinations
of what is good and bad are not relevant considerations, and are left to
art critics.9

art as a human universal

To summarize, art, in its origins, is regarded as an inherent psychobiologi-

cal capacity to artify: that is, to use (and respond to) protoaesthetic visual,
vocal, and/or kinesic behaviours and features which occurred originally in
other adaptive contexts in a considered (i.e. made special exaggerated or
formalized or elaborated) way, thereby demonstrating serious regard (care)
for biologically important life concerns. In circumstances that provoke con-
cern or care, it seems particularly human to enlist exceptional, attention-
getting, emotion-affecting, memorable elements and activities as a sort of
demonstration of serious regard correlative to the biological significance
and value of the things cared about.
As described here, the original psychobiological motivation for artification
was the desire to affect or control through extraordinary effort and execu-
tion the outcomes of uncertain (hence anxiety-provoking) biologically-
important occasions about which people rightly cared. Whether or not an
individual ceremony achieved its particular or proximate purpose (say, secur-
ing game, placating a powerful spirit, or expressing ones resolve), its ultimate
effect was to relieve individual anxiety by providing an illusion of coping,
thereby contributing to survival and reproductive success.
Of equal or even greater importance, however, was an associated benefit.
Ceremonial participation instilled general coordination, cooperation and
feelings of affiliation among members of the group, additionally enhancing
the fitness of individuals. Through cultural ritualization and elaboration, the
behavioural mechanisms that were first evolved in motherinfant mutuality
the repetitions, patterning, dynamic variation, visual, vocal and kinesic dis-
play became adaptive means for arousing interest, riveting joint attention,
synchronizing bodily rhythms and activities, conveying messages with convic-
tion and memorability, and ultimately indoctrinating and reinforcing right
attitudes and behaviour in members of a group. By being especially compel-
ling, beautiful, rare, painstaking and astonishing, a peoples arts are emblems
of how much they care about the sacred beliefs that bind and preserve them.
As such, the arts in traditional societies reinforce a groups communality
and solidarity: they manifest and celebrate who they are. By carefully making
and then responding to these constructions, group members transmit and
reinforce the values the emotional dispositions on which their cohesive-
ness depends. Their belief furthers commitment to long-term interests that
arouse and satisfy needs for shared emotional meaning, to be distinguished
from equally important short-term interests that serve immediate physical
subsistence and preservation. Humans evolved to require satisfaction of both.
The adaptationist account presented here shows that predispositions to
artify are untaught and spontaneous. That is to say, all humans are innately
receptive to protoaesthetic sensory and cognitive features in the environment
(identified, as described, by experimental psychologists and neuroscientists

ellen dissanayake

as particularly pleasurable, satisfying and compelling) and to the temporal

manipulations inherent in motherinfant interaction. Similarly, all chil-
dren are predisposed to play and make-believe that is, to participate in
and acknowledge a bracketed or special, extraordinary dimension to expe-
rience.10 Young children also readily show and observably enjoy the rudi-
ments of art practice, usually without being taught. That is, they will make
marks and images, dance, sing, play with words and language, dress up,
make-believe, and be receptive to specially-crafted stories. Without example
and encouragement from other people, specific art interests and abilities may
remain latent and undeveloped. In premodern societies, however, the arts are
valued and performed by most or all adults, and children grow up experienc-
ing, valuing and performing them also. Although historically, as sociologists
point out, artification has served political and personal power and specialists
(artists) often make the arts that reflect or consolidate that power, in small-
scale, traditional societies, art is rarely confined only to specialists. Artification
is practised by all.
The work of cultural anthropologists not only makes clear the wide diver-
sity of arts in all societies, but implicitly supports an adaptationist view that
in all societies people artify when they care about important things.11 That is,
anthropologists report how ceremonies, with their constituent arts, embody
and give potency to the cultural meanings of societies the meaningful sys-
tems and stories by which religions explain the world and join their adherents
in common cause. At the same time, insofar as it regards art as a ritualized
behavioural counterpart of religion, my adaptationist view incorporates the
theological view of art, since devotees of religious practices everywhere view
their aesthetic actions and artefacts as inevitably associated with their deity
or deities.
In some contemporary socio-cultural or cultural constructivist views, art
can be about anything and anything can be art. Nevertheless, even in mod-
ern environments that are very different from the subsistence societies in
which human nature evolved, people still tend to artify in circumstances
about which they care (e.g. when they wish to impress someone else, mark
an important event, or show love and regard). That is, although art occupies
a variety of new and different roles, it also continues to appear under appro-
priate conditions. The fact that we today artify by purchasing (rather than
ourselves making) self-adornment, holiday decorations and gift presentations
does not negate the wish to effect important outcomes.
Artification still provides pleasure, and people willingly devote large
amounts of time, effort and thought to it, notwithstanding the fact that con-
temporary arts are primarily devoted to popular entertainment, distraction,
and through advertisements the promotion of consumerism.12 Art as
practised by individuals still relieves anxiety, and in alienated modernized

art as a human universal

societies where individualism is valued and fostered it can be a means of self-

expression and self-validation. Various forms of arts therapy whether visual
art, music, dance or drama are acknowledged as ways to deal not only with
sublimation of forbidden wishes but of giving form to, thereby articulating
and resolving important individual problems.
By establishing that art is a human universal, the adaptationist view pre-
sented in this essay implies that art (as making special or artification) has
been and continues to be integral to our lives. More than a revelation of
the divine, a manifestation of political power, the satisfaction of unfulfilled
desires, an expression of the creative self or an agglomeration of perceptual
and cognitive preferences, art as described here emerges from our funda-
mental nature as humans and for untold millennia has been essential to our
life in the world.

This essay is adapted from its original appearance as Kunst als menschliche Universalie: Eine
adaptionistische Betrachtung in Universalien und Konstruktivismus, edited by Peter M. Hejl
(2001, Frankfurt am Main: Suhrkamp Verlag). I am grateful to Dr Hejl for granting permis-
sion to publish this English version. Although I did not always follow their suggestions to the
letter, I have gained much from discussions about the ideas in this essay with Joseph Carroll,
Denis Dutton, Brian Hansen, Joel Schiff and Robert Storey. I gratefully acknowledge their
insights and criticisms.

1. See Brown (1991, 1996) for broad and illuminating discussions of human universals
and their implications.
2. Related to experimental psychology is the neurological view of art as a brain/mind or
cognitive phenomenon. It identifies and describes areas of the brain that are involved
in art-like capacities such as pattern perception, visual thinking, spatial abilities, manual
or kinesthetic abilities, musical abilities, and metaphoric and imagistic abilities. The
recent fields of neuroaesthetics and evolutionary aesthetics (e.g. Voland & Grammar
2003), imply that art can be understood as a collection of cognitive capacities or percep-
tual preferences, and they generally ignore its motivational, emotional and functional
aspects (see Brown & Dissanayake 2009).
3. A notable exception is Anderson (1990).
4. The well-known Austrian ethologist Eibl-Eibesfeldt (1988; 1989: 665702) also
addresses human art, arriving at some of the same ideas and suggestions as those which
follow, although in a less systematic manner. I acknowledge the inspiration of his pio-
neering work. See also Eibl-Eibesfeldt and Stterlin (2007).
5. Non-ethological accounts of the motivation for or function of artistic behaviour (i.e.
that it is self-expression or wish fulfillment or projection or individuation) have
not demonstrated how these proximate functions are ultimately related to ultimate sur-
vival or reproductive success.

ellen dissanayake

6. By object, I include such things as written works (e.g. a novel or musical score), or a
reading or performance of such a work an entity.
7. Dictionaries give subtly different meanings of the word special, not all of which apply
to my use of the adjective. In the Houghton-Mifflin Canadian Dictionary of the English
Language, the first meaning of special is surpassing what is common or usual; excep-
tional. It is this sense in which I have adopted the term. Criticisms that consider
special to be imprecise because it can refer to non-artistic things are employing other
dictionary meanings distinct among others of a kind (singular), primary, peculiar to
a specific person or thing (particular), having a limited or specific function or scope,
arranged for a particular occasion or purpose, esteemed or close. Each of the arts makes
special (surpasses what is usual or ordinary). In dance, for example, ordinary bodily
movements of everyday life are exaggerated, sustained, repeated, patterned; in poetry,
ordinary speech is formalized, rhymed, made striking through alliteration, assonance,
unusual vocabulary and word order; in song, the prosodic features of human vocal
utterance are formalized into fixed intervallic patterns and more regular metre, exag-
gerated with sustained vowels and given notable dynamic emphasis; in the visual arts,
ordinary materials are treated with colour and pattern, or transformed by formalizing
and elaborating; stories are given shape and emphases that surpass the bare facts of
their plot. One might call the actions of sustaining, repeating, exaggerating, patterning,
formalizing, or adding vividness through colour and dynamic variation ways of artifying
aesthetic actions. There may be other aesthetic actions that I have not named. Rather
than list aesthetic actions anew each time, I prefer to unite them in one overarching
concept: making special (as surpassing what is common or usual) or artifying. This
should distinguish artistic making special from specialness for purposes of identification
or esteem or denoting a specific function, scope or application. The notion of making
special is not meant to account for everything about art: it is offered as the ancestral
activity that gave rise to the arts, an activity that continues to imbue all instances of
8. One might ask why human mothers and infants developed such an elaborate and
complex interactive behaviour. We know that walking on two rather than four legs
demanded a number of anatomical changes, including a narrower pelvis. At birth homi-
nid infants (whose head size was also gradually becoming larger than any other primate)
would have to be smaller (more immature) than their ape cousins simply to pass success-
fully through the birth canal without endangering themselves or their mothers. Indeed,
it has been estimated that to be of a comparable maturity at birth as an ape baby, todays
human infants would have to be in the womb a full year longer than nine months, and
would weigh twenty-five pounds (Leakey 1994: 44). Such an immature and helpless
infant would be well advised to be perceived as being as lovable as possible so that its
mother would be motivated to care for it for the requisite longer period of dependence.
I suggest that the rhythmic, patterned mother-infant interaction that we observe today
is based on a ritualized behaviour that co-evolved between early hominid mothers and
their infants to foster emotional attunement and interdependence, thereby enhancing
the babys survival and the mothers reproductive success. The facial expressions, move-
ments and sounds that mothers use with their infants are exaggerations of expressions,
movements and sounds used by other primates, as well as human adults, in contexts of
friendliness and affiliation. By using these, the mother not only communicates affec-
tion to her infant, but also reinforces positive emotion in her own neural circuits. When
structured in a patterned, rhythmic way, so the two can respond and counter-respond,
the interaction becomes a ritualized expression and sharing of a positive emotional state
(see Dissanayake 1999).

art as a human universal

9. Although my adaptationist notion of art as a human universal is not concerned with

evaluations of good and bad art, it nevertheless is concerned with value in art in
terms of motivation and activity rather than product or result. In ancestral times it
would not have been adaptive to make any old thing special. The occasions and arte-
facts for which artifications were considered necessary were of importance to biological
(including psychobiological) survival. Such value is not, and need not be, a considera-
tion in art practice today where subsistence is not an issue.
10. It seems certain that early hominids, like other primates and many higher animals,
would play, and thereby acknowledge a dimension or sphere of activity that is not for
real in that it does not directly affect the subsistence activities of obtaining real food,
evading real predators, fighting real competitors, finding real mates and so forth.
11. It has been said that we assume that all cultures have art, but no one has counted
(e.g. Anderson 1993). My view is that the burden of disproof is on the doubters. I invite
information about human cultures that do not practice some form of artification as it
has been described in this essay.
12. As societies modernize, they emphasize short-term (or material subsistence) values.
Buying and selling, getting and spending, and the quick gratification of immediate
needs or desires become scaffolded upon the fundamental human requirements for
physical survival, but far surpass these. In the ensuing assessment of cost to benefit, the
long-term (emotionally meaningful or spiritual) values so necessary to the cohesion
and perpetuation of truly subsistence societies are forsworn. Thus the arts can become
separate from life, made for their own sake or for reasons that are nonadaptive.


Aiken, N. 1998. The Biological Origins of Art. Westport, CT: Praeger.

Aitken, K. J. & C.Trevarthen 1997. Self/Other Organization in Human Psychological
Development. Development and Psychopathology 9: 65377.
Anderson, R. L. 1990. Calliopes Sisters: A Comparative Study of Philosophies of Art. Englewood
Cliffs, NJ: Prentice-Hall.
Anderson, R. L. 1993. Review of Homo Aestheticus. American Anthropologist 95(1): 183.
Beebe, B., D. Stern & J. Jaffe 1977. The Kinesic Rhythm of MotherInfant Interactions. In
Of Speech and Time: Temporal Speech Patterns in Interpersonal Contexts, A. W. Siegman and
S. Feldstein (eds), 2334. Hillsdale, NJ: Erlbaum.
Beebe, B., D. Alson, J. Jaffe, S. Feldstein & C. Crown 1988. Vocal Congruence in Mother
Infant Play. Journal of Psycholinguistic Research 17(3): 24559.
Berlyne, D. E. 1971. Aesthetics and Psychobiology. New York: Appleton-Century Crofts.
Brody, J. J. 1977. Mimbres Painted Pottery. Albuquerque, NM: University of New Mexico
Brown, D. E. 1991. Human Universals. New York: McGraw-Hill.
Brown, D. E. 1996. Human Universals and Their Implications. In Encyclopedia of Cultural
Anthropology, vol. II, D. Levinson & M. Ember (eds.), 60713. New York: Henry Holt.
Brown, S. & E. Dissanayake 2009. The Arts are More than Aesthetics: Neuroaesthetics as
Narrow Aesthetics. In Neuroaesthetics, M. Skov & O. Vartanian (eds.), 4357. Amityville,
NY: Baywood.
Charmove, A. S. & J. R. Anderson 1989. Examining Environmental Enrichment. In
Housing, Care and Psychological Well Being of Captive and Laboratory Primates, E. F. Segal
(ed.), 183202. Park Ridge, NJ: Noyes Publications.

ellen dissanayake

Danto, A. C. 1964. The Artworld. Journal of Philosophy 61(15 October): 57184.

Danto, A. C. 1996. From aesthetics to art criticism and back. Journal of Aesthetics and Art
Criticism 54(2): 10515.
Dickie, G. 1974. Art and the Aesthetic: An Institutional Analysis. Ithaca, NY: Cornell University
Dissanayake, E. 1988. What Is Art For? Seattle, WA: University of Washington Press.
Dissanayake, E. 1992. Homo Aestheticus: Where Art Comes From and Why. New York: Free
Dissanayake, E. 1995. Chimera, Spandrel, or Adaptation: Conceptualizing Art in Human
Evolution. Human Nature 6(2): 99117.
Dissanayake, E. 1998. Komar and Melamid Discover Pleistocene Taste. Philosophy and
Literature 22(2): 48696.
Dissanayake, E. 1999. Antecedents of Musical Meaning in the MotherInfant Dyad. In
Biopoetics: Evolutionary Explanations in the Arts, B. Cooke & F. Turner (eds), 36798. New
York: Paragon.
Dissanayake, E. 2000. Art and Intimacy: How the Arts Began. Seattle, WA: University of
Washington Press.
Dissanayake, E. 2001. Aesthetic Incunabula. Philosophy and Literature 25(2): 33546.
Dutton, D. 2002. Aesthetic Universals. In The Routledge Companion to Aesthetics, B. Gaut &
D. M. Lopes (eds), 26777. London: Routledge.
Eibl-Eibesfeldt, I. 1988. The Biological Foundations of Aesthetics. In Beauty and the Brain:
Biological Aspects of Aesthetics, I. Rentschler, B. Herzberger & D. Epstein (eds), 2969.
Basel: Birkhuser.
Eibl-Eibesfeldt, I. 1989. Human Ethology. Hawthorne, NY: Aldine de Gruyter.
Eibl-Eibesfeldt, I., & C. Stterlin 2007. Weltsprache Kunst: Zur Natur- und Kunsthistoriche
bildlicher Kommunikation. Vienna: Verlag Christian Brandsttter.
Fernald, A. 1992. Human Maternal Vocalizations to Infants as Biologically Relevant Signals:
An Evolutionary Perspective. In The Adapted Mind: Evolutionary Psychology and the Gen-
eration of Culture, J. H. Barkow, L. Cosmides & J. Tooby (eds), 391428. New York:
Oxford University Press.
Flinn, M. V., R. Quinlan, M. Turner, S. A. Decker & B. G. England 1996. MaleFemale
Differences in Effects of Parental Absence on Glucocorticoid Stress Response. Human
Nature 7(2): 12562.
Hofer, M. A. 1990. Early Symbolic Processes: Hard Evidence from a Soft Place. In Pleasure
Beyond the Pleasure Principle: The Role of Affect in Motivation, Development, and Adaptation,
vol. I, R. A. Glick & S. Bone (eds), 5578. New Haven, CT: Yale University Press.
Leakey, R. 1994. The Origin of Humankind. New York: Basic Books.
McNeill, W. H. 1995. Keeping Together in Time: Dance and Drill in Human History. Cambridge,
MA: Harvard University Press.
Mead, M. [1930] 1976. Growing Up in New Guinea. New York: Morrow.
Miall, D. S. & D. Kuiken 1994a. Beyond Text Theory: Understanding Literary Response.
Discourse Processes 17: 33752.
Miall, D. S. & D. Kuiken 1994b. Foregrounding, Defamiliarization, and Affect: Response to
Literary Stories. Poetics 22: 389407.
Mithen, S. 1996. The Prehistory of the Mind: The Cognitive Origins of Art, Religion, and Science.
London: Thames & Hudson.
Mukarovsky, J. [1932] 1964. Standard Language and Poetic Language. In A Prague School
Reader on Esthetics, Literary Structure and Style, P. L. Garvin (ed.), 1730. Washington, DC:
Georgetown University Press.

art as a human universal

Munro, T. 1963. Evolution in the Arts and Other Theories of Culture History. New York: Abrams.
Nadel, J. 1996. Early Interpersonal Timing and the Perception of Social Contingencies.
Infant Behavior and Development 19 (Special ICIS issue): 202.
Papousek, H. & M. Papousek 1979. Early Ontogeny of Human Social Interaction: Its Bio-
logical Roots and Social Dimensions. In Human Ethology, L. von Cranach, K. Foppa, W.
Lepenies & D. Ploog (eds), 45678. Cambridge: Cambridge University Press.
Papousek, H. & M. Papousek 1981. Musical Elements in the Infants Vocalization: Their
Significance for Communication, Cognition, and Creativity. In Advances in Infancy
Research, vol. I, L. P. Lipsitt & C. K. Rovee-Collier (eds), 163224. Norwood, NJ: Ablex.
Sapolsky, R. M. 1992. Neuroendocrinology of the Stress Response. In Behavioral Endo-
crinology, J. B. Becker, S. M. Breedlove & D. Crews (eds), 287324. Cambridge, MA:
MIT Press.
Schore, A. N. 1994. Affect Regulation and the Origin of the Self: The Neurobiology of Emotional
Development. Hillsdale, NJ: Erlbaum.
Seashore, C. 1938. Psychology of Music, New York: McGraw-Hill.
Shklovsky, V. [1917] 1965. Art as Technique. In Russian Formalist Criticism: Four Essays,
L.T. Lemon & M. J. Reis (eds., trans.), 324. Lincoln, NE: University of Nebraska Press.
Stern, D. 1971. A Microanalysis of MotherInfant Interaction. Journal of the American
Academy of Child Psychiatry 10: 50117.
Taon, P. S. C. 1983. Dorset Art in Relation to Prehistoric Culture Stress. tudes Inuit/Inuit
Studies 7(1): 4165.
Taon, P. S. C. & S. Brockwell 1995. Arnhem Land Prehistory in Landscape, Stone, and
Paint. Antiquity 69: 67695.
Taon, P. S. C., M. Wilson & C. Chippindale 1994. Australias Ancient Warriors: Changing
Depictions of Fighting in the Rock Art of Arnhem Land. Cambridge Archaeological Journal
4: 21148.
Tinbergen, N. 1952. Derived Activities: Their Causation, Biological Significance, Origin, and
Emancipation During Evolution. Quarterly Review of Biology 27(1): 132.
Trevarthen, C. 1997. Fetal and Neonatal Psychology: Intrinsic Motives and Learning
Behavior. In Advances in Perinatal Medicine: Proceedings of the XV European Congress of
Perinatal Medicine, F. Cockburn (ed.), 28291. New York: Parthenon.
her, J. 1991. On Zig-zag Designs: Three Levels of Meaning. Current Anthropology 32:
Valentine, C. W. 1962. The Experimental Psychology of Beauty. London: Methuen.
Voland, E. & K. Grammer (eds) 2003. Evolutionary Aesthetics. Berlin: Springer.
Weitz, M. 1956. The Role of Theory in Aesthetics. Journal of Aesthetics and Art Criticism
15(1): 2735.

The significance of the natural experience
of a non-natural world to the question
of the origin of religion
Donald Wiebe

Human cognitive capacities and the problem of the supernatural

Many students in the field of religious studies who have adopted a cognitive
science of religion approach to understanding religious phenomena seem
to think that once we have come to understand the brain as a collection
of cognitive capacities, formed in our evolutionary development for deal-
ing with the data-processing needs in relationship both to our physical and
social environments, that we can then easily provide a naturalistic, empiri-
cal account of the emergence and transmission of religious ideas and beliefs.
Although I have no doubt that the cognitive sciences are essential to achiev-
ing a naturalistic explanation of religion, I have not found such accounts
for the minds move from the natural to the supernatural realm simply in
terms of such cognitive capacities as a theory of mind (ToM), a hyperac-
tive agency detection device (HADD), an innate dualism, and the innate-
ness of teleological thinking and the like, persuasive. Such human cognitive
capacities mechanisms by which the mind obtains knowledge of the world
make it possible to conceive of supernatural powers or beings, and may
even predispose us to becoming religious, but I do not see how that sheer
possibility actually generates a mental move from the natural to the super-
natural. It seems to me, that is, that something more than the ordinary natu-
ral world is a necessary condition for that predisposition to religion to be
effected; a set of conditions that necessitates a tweaking of the normal human
cognitive capacity humans have for detecting agency in the environment
that ultimately amounts to a radical transformation of the ordinary everyday
notion of agency into supernatural agency. In my judgement, one can find
a persuasive argument to that effect in David Lewis-Williamss attempt to
provide an account of the meaning of the cave paintings of our Palaeolithic
forebears which he believes are essentially religious and the earliest available
expressions of religion. Before setting out Lewis-Williamss account of the

the natural experience of a non-natural world

origin of religion in the Palaeolithic, however, I think it might be helpful

to point out what I find to be the weakness in the current cognitivist views
on the origin or emergence of religion in human thought and practice, and
for this I shall focus on two of the more influential explanatory cognitive
accounts of religion in our discipline: Pascal Boyers Religion Explained: The
Evolutionary Origins of Religious Thought (2001) and Justin Barretts Why
Would Anyone Believe in God? (2004). Neither, in my judgement, however,
provides a coherent and persuasive account of the relationship between our
cognitive capacities and belief in the gods.
In Religion Explained Boyer writes that religion is now just another set
of difficult but manageable [intellectual] problems [and t]he explanation
of religious beliefs and behaviours is to be found in the [natural] way all
human minds work (Boyer 2001: 2). But he does not mean by this to say
that religious ideas are innate that is, for Boyer, people are not born with
the notions of powerful gods and spirits (ibid.: 237). He maintains, moreo-
ver, that our mental predispositions permit the acquisition of such religious
notions but do not determine them (ibid.: 34). There is, he insists, neither
a metaphysical (ibid.: 298) nor a religious (ibid.: 309) centre in the brain by
means of which such ideas are generated. Neither are such supernatural con-
cepts created by religious virtuosos (ibid.: 210).
At one stage in his argument he claims that people get those ideas from
other people, from hearing what they say and observing how they behave
(ibid.: 237), but this clearly does not provide an answer to the question of the
origin of religious belief for, if it is the case that such ideas are gained from
other people, we are still left with the question of how the other people
who are the source of these ideas came to have them. But Boyer, I think, fails
to provide a persuasive response to that question. In Religion Explained he
suggests rather vaguely that the origin of the notion of the supernatural
is the result of a successful activation of a whole variety of mental systems
(ibid.: 298).
Trying to overcome the acknowledged vagueness of his account, he then
maintains that religious cues trigger activation of these cognitive systems
that produce the religious ideas. The problem with this claim, however, is that
he seems to beg the question by referring to religious cues in his explana-
tion of the emergence of religious phenomena. And his caution to the reader
that all this is not so much caused as made more likely by the cognitive proc-
esses [he] described (ibid.: 298; emphasis original), moves him no closer to
having an explanation.
Boyer tries to set out a clearer account of all this in a later essay entitled
Religious Thought and Behavior as By-Product of Brain Function (Boyer
2003), but the argument here also appears to involve begging the question
in that, as he puts it, [r]eligious notions are products of the supernatural

donald wiebe

imagination (ibid.: 119). However, no account is given of the origin of the

so-called supernatural imagination. Moreover, to tell us that supernatural
concepts are informed by very general assumptions from domain concepts
(ibid.) such as artefact, living-thing and person concepts, that make those
concepts inferentially productive, does not really inform us as to how the
notion of the supernatural is derived solely from natural concepts. He admits
that some see such ideas deriving from personal experience (ibid.: 120) but,
unfortunately, he does not follow up on that idea. He rather simply claims
that others find the notion of supernatural agents plausible without such
experience, which still leaves us without an explanation for the emergence
of the concept of god(s), as does the fact, as he puts it, that the notion of
supernatural agents involves a logic of social exchange that is active in non-
religious contexts (ibid.: 122), or the fact that people find culturally acquired
descriptions of such agents intuitively plausible (ibid.: 123). Such comments
leave us with questions that need answering; for example: What precisely in
the non-religious contexts triggers the transition in concept formation from
the natural to the supernatural? Even though our cognitive equipment, with
a bit of tweaking that is, with some violation of the implicit assumptions
that govern the representation of non-religious domains makes this pos-
sible, that simple possibility does not in itself account for the fact of the
emergence of such supernatural notions. None of the mental systems are
about the supernatural and Boyer admits that they therefore require some
kind of external activation, as he puts it in his essay Gods and the Mental
Instincts That Create Them (Boyer 2005).
These cognitive systems will support such religious belief and behaviour, as
Boyer maintains, but they can not, it seems to me, automatically or sponta-
neously create them as do such cognitive capacities that produce an intui-
tive knowledge of the physical, biological and psychological realms. That is,
for Boyer, and most other cognitive scientists, there is no folk religion in
the sense that there is a folk physics, folk biology and folk psychology.
Indeed, it is difficult to see how random deviations or misfirings of these
mental capacities in the non-religious or natural domain can add up to a
socially viable concept of the supernatural. The notions of automaticity and
spontaneity, therefore, are inappropriate in seeking the origin of religious or
supernatural concepts given the generally accepted conclusion that religion
is not an innate capacity in human beings. In my judgement, then, there
must be some positive external force (i.e. external to the cognitive capaci-
ties themselves, although not necessarily external to the brain) that motivates
the tweaking of the mental capacities a something beyond the natu-
ral everyday world that motivates the violations of the implicit assumptions
that govern the representation of non-religious domains of reality to which
Boyer refers.

the natural experience of a non-natural world

If this brief analysis of Boyers argument is on the mark, then it seems

that the gods and religion are by no means an inevitable by-product of our
everyday ordinary cognitive capacities. And yet Boyer, paradoxically in my
judgement, seems also to suggest that supernatural notions and ideas of gods
are almost automatically or spontaneously generated by these cognitive tools.
This is suggested, for example, in his Gods and the Mental Instincts that
Create Them, although we get no account from him of the origin of the
supernatural imagination that apparently creates them (Boyer 2005: 243).
Finally, returning to comments in Religion Explained, he writes that we
have many ideas that require no reasonable reason, as it were, except that
they are the inevitable results of the way the machine works, (Boyer 2001:
95; emphasis added). And he also claims that we do not have the cultural
concepts we have because they make sense or are useful but because the way
our brains are put together makes it difficult not to build them (ibid.: 164).
Both claims, however, seem to suggest that we may indeed be in possession
of a folk religion which is not the understanding of religion he ultimately
To sum up then: according to Boyer, the way our brains are put together
including mental/cognitive capacities such as HADD and ToM (ibid.: 164),
along with the brains native dualism (Bloom 2004, 2005) and native teleology
(Kelemen 1999, 2004; Kelemen & Rosset 2009) may predispose the mind
to belief in the gods and the construction of religion but it is also clear that
that predisposition to holding such beliefs cannot by itself account for push-
ing the mind beyond natural to some kind of supernatural conceptualization.
Justin Barrett, in similarly paradoxical fashion, argues in Why Would Anyone
Believe in God? that most beliefs people hold arise from a collection of non-
conscious mental tools automatically generating assumptions about the way
things are in the world (Barrett 2004: 30; emphasis added). Nonreflective
beliefs, he writes, seem to spontaneously generate [assumptions] in each and
every mundane moment (ibid.: 78; emphasis added), and to form beliefs
simply by looking at the world around us (ibid.: 8). It is important to
notice, however, that the beliefs spoken of here are about the natural world,
an umwelt relevant to the evolutionary wiring of the brain. Barrett also claims
to have shown that the same holds for nonreflective beliefs about God; that
is, he claims that he has shown how belief in gods comes naturally from the
way our minds function in the ordinary world (ibid.: 61; emphasis added).
Mental tools like the HADD and a ToM device, together with a natural bent
toward teleological thinking, he argues, disposes us to find agents around us,
including supernatural ones, given fairly modest evidence of their presence
(ibid.: 31; emphasis added). As he puts it: Sometimes HADDs tendency to
attach agency to objects contributes to the formation of religious concepts
(ibid.: 33). Unfortunately, these claims are made without actually accounting

donald wiebe

for the shift from the natural to the supernatural except by way of arguments
in a conscious, reflective effort to explain some aspect or other of the world of
human experience in or of the world. And the reference to modest evidence
of the existence of the presence of supernatural agents seems to suggest as
much. The fact that the HADD and ToM may make us open to ideas of
god(s) does not provide a motive force that can account for the emergence
of those ideas, so, in effect, there really is no accounting for how one arrives
at having supernatural concepts. Nor does the idea that the existence of the
gods may be parasitic upon a universal system of intuitive expectations gen-
erated by our cognitive capacities constitute evidence for a violation of those
expectations that gives rise to concepts of the supernatural. Whereas for Boyer
the character of our mental capacities predisposes the mind toward the super-
natural, for Barrett it positively encourages belief in gods generally and God
in particular (ibid.: 90); indeed, for Barrett, such concepts are automatically
and spontaneously generated by the brain and therefore are available for use
in explanations of states of affairs in the natural world. This comes through
more clearly in a later essay by Barrett on Cognitive Science, Religion, and
Theology, where he maintains that even though HADD, ToM and other
cognitive capacities are not themselves the origin of the concept of the super-
natural, if they are married with otherwise inexplicable dramatic events
they will prompt belief in gods (Barrett 2009). He does not specify whether
these otherwise inexplicable events are natural events or not and herein lies
the problem. If they are natural events he has not provided an account for
the jump to a supernatural level, and if they are non-natural in the sense
of supernatural, he must explain how they came to be judged as such. He is
clearly aware of the problem that faces him in his search for the origin of the
concept of the supernatural, but he opts for a peculiarly religious/Christian
way to resolve it: he simply asserts that there is a real supernatural reality that
accounts for the existence of the concept in the human mental repertoire.
As he puts it: God has revealed Himself in a variety of ways that trigger
these cognitive capacities to suspect superhuman agency (ibid.: 98), and
he accounts for the imperfect conception of the formation of that concept
in terms of the Christian faith as due to the effects of sin and corruption on
the human character (ibid.: 97). As problematic as his solution is in a search
for an adequate explanation for the origin of religion, it indicates a clear rec-
ognition that one cannot simply move on from a world of natural cognitive
capacities that provide us with an understanding of natural agency in the
world to a conception of supernatural agency without some adequate, but
scientifically believable, motivating force.
If my critical assessments of these arguments is on the mark, then nei-
ther Boyer nor Barrett provides a clear, unambiguous and persuasive scientific
account of how our forebears moved, in thought, from the natural to the

the natural experience of a non-natural world

supernatural realm. I have not, moreover, found a persuasive account of this

movement from the natural to the supernatural in human thought in other
accounts inspired by Boyer and/or Barrett.

Palaeolithic cave art and the problem of religion

In two recent papers I have argued that David Lewis-Williams (and colleagues;
Lewis-Williams 1981, 2002, 2010; Lewis-Williams & Pearce 2005) makes an
important contribution to the explanation of religion (Wiebe, 2009, 2010);
an account that complements and completes the arguments found in Boyer
and Barrett by pointing to the existence of an appropriate umwelt that
makes possible a natural/reasonable account of the conceptual move from the
natural to the supernatural wholly in terms of the ordinary cognitive capaci-
ties of the human mind, which capacities can therefore be seen as playing an
active role in the creation of religion. I will briefly recapitulate that argument
below because I think both that in its essentials it is correct and because
it can very nicely supplement the explanation of religion from within the
framework of the cognitive sciences. However, Lewis-Williamss contribution
was made by way of an attempt to understand the significance of the ancient
rock art of the San of South Africa, and, more particularly, of the Palaeolithic
cave paintings, and this has come under severe criticism that could well deter
scholars of religion from taking his argument seriously.
Although I think I provided an adequate response to some of the criti-
cism of Lewis-Williamss thesis, at the time I wrote those essays, I was not
then aware of R. Dale Guthries massive study of the Palaeolithic material in
his book The Nature of Paleolithic Art (2005) in which he rejects the wide-
spread acceptance of an interpretation of Palaeolithic cave art as essentially the
expression of religious thought and practice and, in particular, he rejects the
interpretation of this material offered by Lewis-Williams. I find it important
to give serious consideration to Guthries argument, moreover, because of
Pascal Boyers positive review of Guthries work and the negative implications
that holds for both my criticisms of Boyers views on the origin of religion and
my espousal of Lewis-Williamss position on that topic. According to Boyer,
for example:

[Guthrie] shatters our most cherished and deeply entrenched

beliefs about rock art, demonstrating for instance that most of it
was not terribly good, that it was probably not very important to
Palaeolithic people and to top it off, that these awesome paintings
had less to do with metaphysics than with testosterone-fuelled young
mens feverish imaginations. (Boyer 2009; emphasis added)

donald wiebe

Although Guthries book is a marvellous work of devoted scholarship and a

delight to read, I nevertheless do not find his claims about the lack of a sig-
nificant relationship between Palaeolithic art and religion persuasive; indeed,
I find a number of his claims problematic or simply unsound. In his fail-
ure to see the weaknesses in Guthries views about Palaeolithic religion
or perhaps, better, the strengths of Lewis-Williamss theory I think Boyer
missed an opportunity to strengthen his own account of religion. Following
my response to Guthrie, I will provide a brief account of the role experience
plays in Lewis-Williamss account of the emergence of religion and show how
it supplements the account of religion given by Boyer, Barrett and most other
cognitive scientists of religion who follow their lead.
I begin this section with a few preliminary comments. Although Guthrie
claims that he does not mean to deny the existence of supernatural
themes [in Palaeolithic art], but only to lift their tyranny (Guthrie 2005:
10), it appears that for him, the presence of such themes in some Palaeolithic
art did not really amount to the emergence of religion. Indeed, throughout
the early sections of the book Guthrie clearly suggests full-blown religion only
emerged in the period of the Holocene. The magico-religious paradigm in
Palaeolithic art, he insists, has derailed rock art research, and he claims that
Palaeolithic art has been accounted for in an obtusely symbolic language
(ibid.: ix) that makes any possible connection between Palaeolithic art and
the supernatural appear exaggerated and contrived, and that at its worst it
has presented early peoples in a distorted light as superstitious dolts totally
preoccupied with mystical concerns (ibid.: 10), and as savages (ibid.: 419).
Although Guthrie, late in the book, admits that Lewis-Williams, along with
others, has made great progress in explaining [this material] with the help
of ethnography of existing Bushman and aboriginals (ibid.: 427), it appears
nevertheless that he has included Lewis-Williams among the nut-case inter-
preters of Palaeolithic art (ibid.: 811, 460). But he has done so without
spending any time assessing the character of the arguments in favour of the
magico-religious interpretations or, indeed, without discussing any of the
Palaeolithic masterpieces themselves that, as William H. McNeill has pointed
out in his review of Guthries book, have been central to virtually all the other
interpreters of Palaeolithic cave art (McNeill 2006: 223). As McNeill com-
ments, this makes Guthries repudiation of magical and religious motivation
in the making of the masterworks of cave art implausible.
In the early chapters of the book Guthrie also maintains that it is only in
the artwork of the Holocene period that we have any real indication of the
emergence and establishment of religion which he sees as a mode of thought
characterized by supernaturalness and irrationality neither of which, in his
early judgement, is characteristic of Palaeolithic life and thought. Indeed, I
shall suggest below that Guthrie holds a kind of fall from grace-view of

the natural experience of a non-natural world

our Cro-Magnon forebears in the radical shift in thought and practice that
took place in the transition from Palaeolithic to Holocene culture. However,
I will argue below that his thought on these matters in the later sections of
the book contradict his early remarks and suggestions about both Palaeolithic
and Holocene art and religion. I will show, that is, that Guthries account
of Palaeolithic art when properly assessed in light of his comments about
Holocene art, will have no negative import for Lewis-Williamss answer to
the question as to how the idea of deity (the gods/god) entered the human
mind. According to Guthrie, not only is the art of the Palaeolithic bands
not focused on an obtusely symbolic language about another world (Guthrie
2005: ix), their art seems more focused on complicated earth-bound sub-
jects, diverse everyday interests and wonders (ibid.: 5), and portray[s] a peo-
ple in close touch with the details of a complex earth (ibid.: 10). Moreover,
claims Guthrie, [h]unting behavior is a central unifying feature of Paleolithic
art (ibid.: 49) because, as he puts it, there is something about the connec-
tion between large-mammal hunting and the human psyche that is unique,
because this behavior involves interacting with beings similar to [them-
selves] (ibid.: 49). Thus he writes: Animals were the Pleistocene libraries,
newspapers, comics and videos, classrooms, shop floors, soccer matches and
churches (ibid.: 91). And making these images was not limited to master
artists but was engaged in by adolescents and even by children. Indeed, he
maintains that works by young people constitute both a disproportionate
and largely unrecognized fraction of preserved Paleolithic art (ibid.: 115) and
that much of it is of a sexual and erotic nature the claim that has earned his
work the title of the testosterone theory of Palaeolithic art. He also main-
tains in criticism of Lewis Williamss claim that caves containing paintings
by the Palaeolithic huntergatherers may have been used as places of vision
quests or of worship of some kind or other that it is unreasonable to argue
that it was mostly Paleolithic adults who went back into the caves [for]
most Paleolithic visitors seem to have been young people (ibid.: 124). This
may not, however, be as big a problem as Guthrie thinks it is if, as McNeill
points out, the paintings are not connected to a kind of public worship but
should rather be seen as a form of silent insurance against danger to humans
from the angry displaced spirits of the animals they had killed (McNeill
2006: 23). A similar problem exists with Guthries view of the meaning of
handprints on cave walls. For him the handprints are simple graffiti, like
leaving ones name to indicate I was here. Brian Fagan suggests rather that
by way of the handprints people acquired some kind of power from contact
with dark rock faces beneath the earth (Fagan 2010: 211). As he puts it:
The participants used red iron oxide or black manganese oxide to outline
their hands, which created the impression that their hands had melted into
the rock. Once the hands were withdrawn, the impression would ensure vivid

donald wiebe

proof of contact with the supernatural world (ibid.: 212). Nevertheless, the
conclusions Guthrie draws from his various observations are that [w]e must
recognize that the Paleolithic art we have is not uniformly great art (Guthrie
2005: 144); and that most of the cave (and mobile) art is adolescent graffiti
that can be compared to modern adolescent graffiti (ibid.: 12930), generated
in large part because of a surfeit of testosterone.
It seems to me obvious, then, that Guthries suggestion that the cave
paintings could have been made by adolescents and children (ibid.: 126)
is far-fetched, as several reviewers of the volume have pointed out. Randall
White (2006), for example, claims that the comparison of what Guthrie calls
Palaeolithic graffiti with modern graffiti underestimates the cultural context
of both the [Palaeolithic] figurines and the Playboy photographs, and Paul G.
Bahn (2006) says that Guthries claims about the ubiquity of vulvas and men
with erections are exaggerations, and that the evidence that these works were
created mostly by children and adolescents is flimsy. Nevertheless, Guthries
claim that Paleolithic art is not simply a collection of masterworks (Guthrie
2005: 91) is no doubt right when he is talking about the number of pieces
of Palaeolithic art to which we have access, and if we take into consideration
the taphonomic principle that insists that we recognize the many biases in the
preservation of their art that can have a considerable distorting influence on
our interpretations of it (ibid.: 13).
Given the attention that Guthries reference to testosterone-driven graffiti
found in Palaeolithic cultural productions has received one might get a dis-
torted impression of Guthries views about the nature of all Palaeolithic art.
I think it important, therefore, to bring to attention the fact that Guthrie
sees Palaeolithic art as fitting into several different categories that is, art
produced by children, by testosterone-impelled adolescents, by developing
artists, and by seasoned adult artists. Guthrie makes a good case for seeing
much of the artwork found in Palaeolithic remains as childrens art made
by children themselves either as toys or as playing at making art (ibid.: 142).
By graffiti Guthrie means quick images and thoughtless scrawls made
by adolescents motivated by an inclination to make some visual mark to
leave a personal scratched or painted trace [of their existence] behind in pri-
vate spots (ibid.: 140), or to express erotic thoughts and desires (ibid.: 141).
Although he views this as part of a larger temporal phenomenon, he does
not see these Palaeolithic productions to be exactly like modern graffiti
because they do not, for example, contain the sense of anger, pique, and
ugliness so often found in modern graffiti. He also insists that modern graf-
fiti is not the same as Palaeolithic art (ibid.: 198), and, more importantly,
he warns that Palaeolithic graffiti must not be confused with fine art. As he
puts the latter point: They drew in ways that were often free, careless, cas-
ual, alive, gritty, and occasionally erotic (ibid.: 141) indeed, splattered

the natural experience of a non-natural world

with youthful testosterone (ibid.: 198) and much of what they drew could
be considered vandalism in that it involved modifying, marking over, or
scratching out previous art (ibid.: 141, 198). And Guthrie distinguishes this
irregular scatter[ing] of graffitiesque images from the few fine works of art
such as those to be found in Lascaux and other Palaeolithic cave sites (ibid.:
459). These masterworks of art would have been the work of proficient adult
artists experts, one presumes, at painting large mammals and carving intri-
cate figures. As developing artists they would have produced thousands of
artefacts on their way to acquiring proficiency (ibid.: 144) that would be
found among the many works of children and adolescents but those pieces
would not, it seems obvious to me, hold the significance of the masterful cave
art ultimately produced by this adult class of artists.
For Guthrie, then, Palaeolithic artwork is essentially a rational kind of
exercise, both in the sense that it is rationally explicable and in the sense
of the artists being naturalists rather than supernaturalists (religionists) and
therefore more concerned with animals than spiritual realities. According to
Guthrie, that is, it was important for them to understand animal behav-
iour, and that came by way of close empirical observation which was then
expressed in their art (ibid.: 52). Images of large animals in a certain style,
he writes, constitute the chief unity of Paleolithic art since they played an
essential role in our evolution, either because they were sources of food and
general welfare or objects of hunting done for the love of it (ibid.: 209, 226;
see also ibid.: 247). In either case, he maintains, the cave art is meant as a
celebration of the splendour of their prey (ibid.: 236), not as an expression of
an encounter with spiritual realities of some sort or other. And that motiva-
tion for Guthrie is the guy rope that will help ground Paleolithic art (ibid.:
250), not the childrens art or the Palaeolithic graffiti. This, he claims, is the
bigger story of Palaeolithic life, not that of dogmatic religions and fright-
ening shamanistic practices (ibid.: 460). But the legacy of the sum total of
Palaeolithic art he claims really tells a story of past halcyon days around the
campfires with skewered roasts and extravagant sunsets that tell us to get on
with life as did our Palaeolithic forebears; the message of their art being that
one would be wise to play: play physically, play mentally, and, above all, play
artfully (ibid.: 460). (This statement, as McNeill has appropriately pointed
out in his review of the book, is clearly a secular confession of faith; McNeill
2006: 20).
It is clear, then, that early in the volume Guthrie sees the Palaeolithic
huntergatherers essentially as modern rational people free from magic and
religion, and therefore very much like us. And for him, Palaeolithic art is:

a testimony to the fact that our Pleistocene ancestors were neither

cruel brutes nor noble savages. People of their times, individuals

donald wiebe

like ourselves, they did mainly what evolution had them do best.
They must have found goodness, beauty, and happiness in that
life-way, along with excitement, difficulty, and danger differing
only in degree to what we feel today, for we are essentially the same
creature, however different the context of our lives.
(Guthrie 2005: 224; emphasis added; see also ibid.: 408)

They had to be rational, he insists, because in tracking, neither magic nor

appeals to omniscient authority are possible (ibid.: 423, 224). Hunting
made [them] the cooperative ape (ibid.: 237) which required a sophis-
ticated intelligence (ibid.: 249). They nurtured their children in the same
fashion we do ours, trying to guarantee them a sound education (ibid.: 250),
and in their play there is a deep evolutionary agenda (ibid.: 375) that was
important in the making of an animal that is intelligently flexible and oppor-
tunistically able (ibid.: 385). Again, as he puts it:

To survive and prosper, Paleolithic peoples needed to construct a

lifeway that allowed them to be deeply empirical, rational, flexible,
and imaginative. [And] Paleolithic works of art, [he continues],
reveal the artists absorption in observed details about their living
subjects, quite different from the works arising from traditions of
visually codified and highly symbolic images. (Ibid.: 433)

This is a quite marvellous and attractive view of our Palaeolithic ancestors,

but I think it is also quite unbelievable with respect to the psychological and
emotional proximity in which it places the Palaeolithic to the mind-set and
life-style of historically modern human beings compared to the psychologi-
cal and emotional distance it places between the Palaeolithic huntergather-
ers and their Holocene descendants. As I noted above, it almost seems as if
Guthrie espouses a kind of theory of a fall from grace to account for the
difference between the Palaeolithic and the Holocene. He talks, for example,
of a major qualitative shift in life-way between the two groups of people given
the jump in size of human communities from bands to tribes, chiefdoms and
nation-states (ibid.: 412). He believes that some psychological threshold was
crossed with that increase in size but does not elaborate how this would have
effected a transition from a rational way of life to one steeped in magic and
religious supernaturalism (ibid.: 414). He provides criteria that can be used to
show that [c]onnections between art and the supernatural seem to have been
exaggerated and more common at higher Holocene densities (ibid.: 426)
but in claiming this he provides no real explanation for that development
but merely points to the correlation between increased population densities
and increased interest in the supernatural. The main difference seems to be a

the natural experience of a non-natural world

change in religion from an unorganized communal level of activity compared

to the structured activities of religion at an organized institutional level. As he
puts it: At the end of the Pleistocene when tribal organizations emerged, they
often took the supernatural and elaborated it into something else, becoming
organized religions, institutions that dominated everyday life (ibid.: 427).
And this, it seems, suggests to Guthrie that visions of Palaeolithic cave art as
fundamentally religious held by people like David Lewis-Williams and Jean
Clotte (ibid.: 10) inappropriately makes the Palaeolithic huntergatherers
appear to be superstitious dolts (ibid.: 10), although for Guthrie this conclu-
sion actually does apply to the Holocene. He objects to what he claims are
their negative views of the Palaeolithic as dolts by claiming that [d]espite the
counter-tugs of hormones, dreams, and personal and cultural histories, we
each have the ability to reason objectively (ibid.: 12) and that the Palaeolithic
responded in that rational fashion to the tasks required to accomplish staying
alive but he does not extend that charitable interpretation to the people of
the Holocene. In tracking, neither magic nor appeals to omniscient author-
ity are possible (ibid.: 423) he writes of the Palaeolithic. But the fact that
empirically rational observation is essential to tracking and hunting is not
sufficient grounds on which to justify the claim that our Palaeolithic forebears
were just like us. If that were the case Guthrie would also have to count our
Holocene forebears as rational moderns in the same sense since the Holocene
agriculturalists too required a great deal of empirically sound knowledge to
guarantee success in the agricultural venture on which they embarked. But he
draws no such conclusion about our Holocene forebears. Indeed, he presents
an altogether different picture of them, claiming that, with them, religion
with all its ugly irrational supernaturalness emerges. Yet at the same time that
he castigates the irrationality of the Holocene interest in the supernatural,
he acknowledges the existence of interest in the supernatural in at least some
Palaeolithic art.
Given the tenor of Guthries treatment of Palaeolithic art in the early chap-
ters of the book, it appears that he sees in the transition from Palaeolithic
to Holocene lifeways a fall from using his terminology rationality and
reasonableness into magic, superstition and ugly supernaturalism that is,
from rationality into doltishness. However, in trying to account for the evo-
lution of the supernatural in his final chapter, he acknowledges that the
supernatural does indeed, must characterize the Palaeolithic lifeway even
though he insists that supernaturalism did not permeate the Palaeolithic art
to the same extent and degree that it did the art of the Holocene (ibid.: 428);
art, that is, was less meaningful, less belief-based, and more a matter of
individual perception and experiment (ibid.: 433) in the Palaeolithic than
in the Holocene. It must characterize Palaeolithic art because, as he puts it,
[b]ehavior centering on the supernatural has life consequences that are too

donald wiebe

important for natural selection to have ignored (ibid.: 438). That is, access
to the supernatural is a powerful psychic force in maintaining emotional
health and well-being which is utterly essential in providing an inner core
of confidence in the face of the frailties of human existence (ibid.: 438).
And he sees this as a universal tendency, a proclivity that is a part of our
makeup, a genetic program within us (ibid.: 439). But it is also more than
merely a proclivity for he also calls it an adaptive fix to the perils of our
rational system that short-circuited our attempts to deal with such issues as
meaning, loss, illness, isolation and death (ibid.: 444). As he puts it: So while
it is our specialty, humans are not totally creatures of rationality and reason
(ibid.: 446). And the fix was a form of spirituality that he calls the emollient
of the supernatural. He notes:

With that, the engine of the rational brain not only works but
runs better most of the time. The fix was to isolate a few of the
processes of reason which dealt with meaning and purpose and to
make those sectors accept meaning by grace, authority by revela-
tion, and an alternative cast to reality outside the natural world.
(Ibid.: 444)

The problem with his fall from grace theory, however, is that given his
views of the character of the human brain it cannot be restricted in applica-
tion only to the people of the Holocene since all our forebears, including the
Palaeolithic, operate with the same intellectual strengths and weaknesses. And
if this is the case then, surely, religion, with its ugly and irrational supernatu-
ralness would also have characterized Palaeolithic society even if it had not
been a dominating organized religion. And, as a matter of record, Guthrie
recognizes that this has to be the case since, as he put it, we all use [the fix]
in different ways and degrees, nobody is exempt, it is part of being human
(ibid.: 444). What is even more problematic with Guthries view of the evolu-
tionary development of religion, then, is the implication that a form of auton-
omous, reflective empirical and rational thought about the natural world (not
simply a common-sense engagement with the natural world which charac-
terizes all animals engagements with the natural world) preceded belief in and
reliance upon the emollient of the supernatural. Emergence from the grip
of a powerful psychic force in maintaining emotional health and physical
well-being (ibid.: 438) makes far more evolutionary sense than does a devo-
lutionary idea of the emergence of a psychic fix to originally wholly rational
reflective capacities. This is especially so in light of the fact, as Brian Fagan
points out, that there has never been a huntergatherer society on earth cre-
ated by Homo sapiens that did not possess a complex set of supernatural beliefs
or consider itself as living in an intensely symbolic realm (Fagan 2010: 151).

the natural experience of a non-natural world

There are other, more specific issues that I think make Guthries criticisms
of Lewis-Williamss and Jean Clottes views on Palaeolithic art problematic.
There is, for example, the problem of a lack of consistency in his claims.
For example, he admits early on in his argument that [r]eligious images
probably do occur in Palaeolithic art (Guthrie 2005: 10). Given this turn
in Guthries views of Palaeolithic art it is surprising that, as McNeill puts it,
Guthrie never explains why the hunters supernatural beliefs he never spe-
cifically describes do not show up in the graffiti and paintings left behind
(McNeill 2006: 223). Guthrie further admits that it is not the case that the
many rock art researchers who see [Palaeolithic] art as magic and mystical are
totally wrong (Guthrie 2005: 10). He argues that they have extended those
interpretations universally (ibid.: 10), but this is a judgement that he does
not establish here and one which Lewis-Williams in particular denies. And
Guthrie seems to acknowledge this in his recognition that Lewis-Williams
made good progress in explaining San rock art as religious expression (ibid.:
427) even though San religion is not an organized religion. And even if
Guthrie insists that strange anthropomorphs make up only a small frac-
tion of Palaeolithic art, he nevertheless admits that this still indicates that a
supernatural fix was needed for his otherwise (that is, originally) wholly
rational Palaeolithic hunters.
Some might wish to argue here that Guthries admission about the need
for a psychic fix does not undermine his earlier arguments about the
nature of Palaeolithic art given his claim that even if Palaeolithic hunter
gatherers were religious, they were not religious in the ugly supernatural-
ist way of the people of the Holocene (ibid.: 4278). However, Guthries
complaint about the ugliness of religion is not that it is organized and insti-
tutionalized, but that it accepts and submits to the supernatural. And it
is this ugliness of the supernatural that he rejects as being characteristic of
the rational Paleolithic. But that claim would only be justified if it were
the case that the ugliness of the supernatural emerges only with the insti-
tutional organization of religion and no argument to that effect is presented
by Guthrie. Indeed, quite to the contrary, as Brian Fagan points out, the
supernatural is every bit as alive in the Palaeolithic as it is in the Holocene
as can be seen in the sculpture of the Lion Man figurine from Hohlenstein-
Stadel, Baden-Wrttemberg, Germany which, he maintains, is indicative of
a partnership between humans and the supernatural realm of animals. Fagan,
who is familiar with the views of Lewis-Williams and Guthrie, as well as
fully aware of the continuing controversy over the meaning of Cro-Magnon
art ever since its discovery (Fagan 2010: 149), insists that the existence of
the figurine is far more than merely an art-object (ibid.: 137, 149). [W] ith
his leonine head and human limbs, he writes, the figurine bridged the
chasm between living and supernatural realms (ibid.: 1) and indicated a

donald wiebe

burgeoning and obviously complex range of spiritual beliefs (ibid.: 137)

and a strong hint that at least some transformation rituals unfolded among
the Aurignacians of the Upper Danube (ibid.: 142). In the later stages of
his argument, moreover, Guthrie admits that the Holocene functioned with
the same mental and cognitive equipment possessed by their Palaeolithic
ancestors which can only mean that the Palaeolithic were also given over to
the experience of the ugliness of the supernatural as described by Fagan.
Consequently, Guthries comments about the transition from Palaeolithic
religion to the organized and institutionalized religion of the Holocene is
clearly irrelevant to his argument about the character of Palaeolithic thought
and art.
There are two important conclusions that one can warrantably draw from
this account of Guthries views on Palaeolithic art and his reflections on reli-
gion. The first is that his views on the relationship of religion to Palaeolithic
cave art is muddled at best, and simply incoherent at worst. McNeill has a
much more balanced view when, at the end of his review of Guthries book
he writes:

cave art derives both from the natural world of flesh, blood, and
brain that once existed on the Mammoth Steppe, and from an
imaginary world of invisible spirits, embodied and disembodied
who, the artists believed, controlled, directed, and inspired ani-
mal and human behaviour both above and below ground. Only
by positing such an imaginary world can we begin to understand
the paradoxical mix of serene and accurate masterworks with the
multitude of free and spontaneous scribbles that together com-
prise the art of the caves. (McNeill 2006: 23)

The second is that even though he espouses a wholly rationalist account of

Palaeolithic art, wholly free from the taint of irrational religion, in the early
portions of his book, he ultimately recognizes that religion in some form or
other characterized human life in the Palaeolithic and expressed itself to
some degree in their art even as it did in the life of Homo sapiens in the
Holocene and beyond.

Natural experience of a non-natural

(but not supernatural) world

If we accept the conclusions I have reached about Guthries views on

Palaeolithic art, then his criticisms of David Lewis-Williamss interpretation
of the masterpieces of their cave art as representations of religious beliefs and

the natural experience of a non-natural world

practices appears unfounded and misdirected. Guthrie is highly critical of

those scholars who seem to think that all Palaeolithic artwork possesses a reli-
gious character, and justifiably so, but Lewis-Williams makes no such claim
about all Palaeolithic art; indeed, for the most part, he restricts his analyses
to the masterpieces of Palaeolithic art found on cave ceilings and walls,
and in no way can his work be seen as presenting the Palaeolithic artists as
being superstitious dolts. Furthermore, even though Guthrie believes organ-
ized religion did not emerge until the Holocene, he does not as I have shown
deny that Palaeolithic huntergatherers were also religious and therefore
entertained notions of the supernatural, but only that they did not exhibit the
ugly irrational supernaturalness he sees in organized religion. Guthrie, that
is, clearly acknowledges that humans are not totally creatures of rationality
and reason (Guthrie 2005: 446), and that they, like their Holocene succes-
sors, required a fix to the perils of our/their rational system by what he
calls the emollient of the supernatural (ibid.: 446). But if the lifeway of the
Palaeolithic huntergatherers had been as rational in its everyday activities
as Guthrie suggests, what could the source of the supernatural have been?
Guthrie does not provide a clear and persuasive answer to that question, but
it is obvious to him that the Palaeolithic like their Holocene successors and
historically modern human beings would have experienced some kind of
supernatural reality (ibid.: 438).
That even the Palaeolithic huntergatherer would have created a religion,
moreover, should not come as a surprise to any cognitive scientist since these
people were equipped with all of the cognitive capacities of historically mod-
ern Homo sapiens and that would have predisposed them to doing so. Hence
my surprise at Boyers firm endorsement of what has been called Guthries
testosterone theory of Palaeolithic art as a convincing account of the awe-
some paintings, that we find in the caves. With McNeill and others, I find
such an account of the cave masterpieces implausible; this is not the kind of
work one can sensibly attribute to children and adolescents. It seems to me,
rather, that accounting for it as an expression of an extraordinary experience
of something beyond the natural realm that is, of a supernatural realm
is not bizarre, nor is it to perceive our Palaeolithic forebears as doltish. And
David Lewis-Williams, I suggest, provides a reasonable alternative account of
that cave art as expressions of religion of a supernatural outlook or world-
view that is wholly consistent with our understanding of how it is possible for
our cognitive capacities to conceive of supernatural powers and beings. Unlike
Guthrie, and most cognitive scientists of religion, Lewis-Williams points to
a source for the experience of the supernatural (where the supernatural
does not exist, as it does for Barrett, but is nevertheless psychologically and
socially real) that ultimately gets referred to or expressed in Palaeolithic art
and (in all likelihood) narrative.

donald wiebe

Lewis-Williamss theory of religion stems ultimately from his attempt to

understand the nature of the rock art of the San people of South Africa.
By combining a comparative analysis of artistic expressions that had only
received isolated attention, and by comparing those expressions with nine-
teenth century ethnographic data as well as contemporary ethnographic data
on the !Kung in Namibia and Botswana, Lewis-Williams was able to show
that the paintings and engravings did not primarily signify natural objects in
the world but were essentially expressions of the meanings, values and spir-
itual beliefs of the San; that they depicted realities with supernatural poten-
cies that were experienced in altered states of consciousness in which the San
were in contact with a non-natural (i.e. a pseudo-supernatural) realm of
existence. The artwork therefore was a way for them to see what it is they,
through their trance experiences, came to believe they could see. As Lewis-
Williams puts the point in his published PhD dissertation Believing and
Seeing: Symbolic Meanings in Southern San Rock Paintings the paintings ren-
dered visible a complex set of beliefs and made it possible for the San to see
what they believe (Lewis-Williams 1981: 83, 131). The paintings, therefore,
were reified visions of a non-natural world of a supernatural spirit world.
Given the many similarities between Southern African San rock paint-
ings and those of the Palaeolithic artists, Lewis-Williams thought that even
though they had no corresponding ethnographic data on which interpreta-
tions of their art could be based, their art could nevertheless be accounted
for in terms of similar experiences that the Palaeolithic people must have had
given that their brains were wired in the same way as the brains of the San
experiences that they would have had to account for both individually and
socially. Without the same neuronal potential in Palaeolithic people that is
to be found in all historically modern human beings, that is, there would
have been no revolution in experience and belief that he thinks is expressed
in their artwork.
So for Lewis-Williams (and colleagues) religion and Palaeolithic master-
pieces of art which in this case is religion expressed are rooted in the
electro-chemical functioning of their brains. The spectrum of human con-
sciousness, he points out, ranges along a continuum from the consciousness
we experience in our wakened state, in which we are alert and outward-
focused, to deep trance experiences which involve inward-directed states of
mind that go way beyond the waking problem-oriented thought of ordi-
nary everyday living. These altered states of consciousness produce sensations
(that is, hallucinations) of extra-corporeal travel in alternate worlds and a
sense of encounter with supernatural powers, which Lewis-Williams refers to
as spiritual experience. He does not conclude that these experiences alone
amount to religion, but rather argues that such experiences would have called
for an explanation and response in a way that was different from practical

the natural experience of a non-natural world

thinking in the material world (Lewis-Williams 2010: 157), and may well
have been viewed as non-natural resources for dealing with the exigencies
of everyday life. These experiences, that is, seemed to make people aware of
two realms of existence: the everyday ordinary natural world and a paral-
lel non-natural/supernatural (i.e. non-everyday) world suggested by the
weird, non-real experiences that their brains sometimes generated (ibid.).
Lewis-Williams refers to these experiences as constituting a consciousness-
contract, but points out that making sense of these experiences would not
have been a wholly individual or purely psychological affair since all people
have to live with and accommodate the products of their brains in a society
of other brains and bodies (Lewis-Williams & Pearce 2005: 36). According
to Lewis-Williams, that is: people do not confront questions about the sig-
nificance of the inner world as individuals [but i]nevitably discuss their
mental experiences with other people (Lewis-Williams 2010: 158). And in
doing so, he maintains that they were engaged in a process properly referred
to as a social contract ( la Rousseau). And that social contract, he main-
tains, was essential to the formation of religion which depends upon human
agreement on what that other realm is like (beliefs) and on the demands it
makes on them in the world of ordinary, everyday experience (religious prac-
tice). Achieving this means, of course, that some of these weird experiences
will be ignored while others will be cultivated. And in the process there can
be no doubt that some people laid claim to special knowledge or insight that
set them apart from others and allowed them to rise to positions of influence
and power, ultimately leading to organized religion in larger tribal societies
and beyond. Thus he writes: Once the existence of a supernatural realm is
accepted, religious devotees, whether they themselves have had religious expe-
riences or not, begin to create and establish religion and to build theological
superstructures (ibid.: 165).
In this respect it is helpful to recognize, as Merlin Donald has shown, that
the Upper Palaeolithic was the era of mythic invention in which our Homo
sapiens forebears moved beyond the episodic perception of events, beyond
the mimetic reconstruction of episodes, to a comprehensive modelling of the
entire universe (Donald 1991: 214). They leaped into a narrative mind-
set, as he puts it in a later work (Donald 2001: 295). And even though
symbolic invention may have been the work of single minds, Donald never-
theless argues persuasively that its full exploitation was a collective enterprise
that made possible a consensual definition of a shared reality that is the core
of oral culture (ibid.: 296). In the creation of virtual worlds, then, our fore-
bears, Donald argues, escaped the constraints of the nervous system making it
impossible to find all our explanations of and about our species simply within
the confines of the brain-mind (ibid.: 151). The full implications of this argu-
ment, however, will not be followed up on here.

donald wiebe


In conclusion I offer the following brief comments. First: central to a fully

scientific explanation of religion for Lewis-Williams is the illusory spiritual
experiences generated in the electro-chemical firings of the human brain and
the social negotiation of those experiences as well as their expression in myth,
ritual, music and other works of art. Second: as I have attempted to show
here, his work positively complements the search for an explanation of reli-
gion sought by cognitive scientists and those students of religion who find
the cognitive sciences helpful in understanding various religious phenomena.
Indeed, without this addition to their explanatory arsenal, it seems to me
that their explanations merely suggest the possibility of a wholly naturalis-
tic explanation of the origin of religion without actually realizing it. What
the electro-chemical firings of the human brain provide to the explanatory
project is a genuine basis upon which such human cognitive capacities as the
hyperactive agency detection device and theory of mind are, so to speak, able
to produce (mythically and/or doctrinally) supernatural beings (gods, God,
etc.) by deviating from the implicit assumptions that govern the representa-
tion of ordinary everyday agents, as well as a genuine motivating force for
the creation of new worlds by extension of the native dualism characteristic
of human minds. Such experiences are also hospitable to the native teleology
of the human mind that can account for belief in the meaningfulness of life
and the cosmos, and, finally, given the literary and narrative character of the
human mind referred to above, embody it all in myth, belief and, ultimately,


This paper is a revised version of a public lecture delivered at Masaryk University in Brno
(Czech Republic) on 26 October 2010. I wish to thank the Department for the Study of
Religion for the invitation to deliver the lecture, and for their wonderful hospitality. I also wish
to thank the auditors for their appreciative reception of the paper, as well as for their critical
comments and questions that have helped me in clarifying various issues raised in the paper.


Bahn, P. G. 2006. Sex and Violence in Rock Art. Nature 441: 5756.
Barrett, J. 2004. Why Would Anyone Believe in God? New York: AltaMira Press.
Barrett, J. 2009. Cognitive Science, Religion, and Theology. See Schloss & Murray (2009),
Bloom, P. 2004. Descartes Baby: How the Science of Child Development Explains What Makes Us
Human. New York: Basic Books.

the natural experience of a non-natural world

Bloom, P. 2005. Is God an Accident?. Atlantic Monthly 296(5): 10512.

Bloom P. 2009. Religious Belief as an Evolutionary Accident. See Schloss & Murray (2009),
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Boyer, P. 2003. Religious Thought and Behavior as By-products of Brain Function. Trends
in Cognitive Science 7(3): 11924.
Boyer, P. 2005. Gods and the Mental Instincts that Create Them. In Science and the Human
Experience, J. D. Proctor (ed.), 23759. New York: Oxford University Press.
Boyer, P. 2009. Paleolithic Art: Awesome But Not Religious. Pascals Blog (24 February):
html (accessed 18 October 2010).
Donald, M. 1991. Origins of the Modern Mind: Three Stages in the Evolution of Culture and
Cognition. Cambridge, MA: Harvard University Press.
Donald, M. 2001. A Mind So Rare: The Evolution of Human Consciousness. New York: W. W.
Fagan, B. 2010. Cro-Magnon: How the Ice Age Gave Birth to the First Modern Humans. New
York: Bloomsbury Press.
Guthrie, R. D. 2005. The Nature of Paleolithic Art. Chicago, IL: University of Chicago Press.
Kelemen, D. 1999. Beliefs about Purpose: On the Origins of Teleological Thought. In The
Descent of Mind: Psychological Perspectives on Hominid Evolution, M. C. Corballis & S.E.G.
Lea (eds), 27894. Oxford: Oxford University Press.
Kelemen, D. 2004. Are Children Intuitive Theists?: Reasoning About Purpose and Design
in Nature. Psychological Science 15: 295301.
Kelemen, D. & E. Rosset. 2009. The Human Function Compunction: Teleological Explana-
tion in Adults. Cognition 111: 13843.
Lewis-Williams, D. 1981. Believing and Seeing: Symbolic Meaning in Southern San Rock Art.
London: Academic Press.
Lewis-Williams, D. 2002. The Mind in the Cave: Consciousness and the Origins of Art. London:
Thames & Hudson.
Lewis-Williams, D. 2010. Conceiving God: The Cognitive Origin and Evolution of Religion.
London: Thames & Hudson.
Lewis-Williams, D. & D. Pearce. 2005. Inside the Neolithic Mind: Consciousness, Cosmos, and
the Realm of the Gods. London: Thames & Hudson.
McNeill, W. H. 2006. Secrets of the Cave Paintings. New York Review of Books (19 Oct-
ober): 213.
Schloss, J. & M. Murray. 2009. The Believing Primate: Scientific, Philosophical, and Theological
Reflections on the Origin of Religion. Oxford: Oxford University Press.
Wiebe, D. 2009. Roots in the Brain: On David Lewis-Williams and David Pearce, Inside the
Neolithic Mind. In Contemporary Theories of Religion: A Critical Companion, M. Stausberg
(ed.), 195208. London: Routledge.
Wiebe, D. 2010. Recovering Religious Experience in the Explanation of Religion. In
Chasing Down Religion: In the Sights of History and the Cognitive Sciences, Essays in Honor
of Luther H. Martin, P. Pachis & D.Wiebe (eds), 51130. Thessaloniki: Barbounakis
White, R. 2006. Looking for Biological Meaning in Cave Art. American Scientist (JulyAugust):

Religion and the emergence
of human imagination
Andreas Lieberoth


Relating the emergence of religion directly to single minds has grown hope-
lessly out of fashion, and even though the cognitive science of culture delves
deep into the human brain, we recognize that religions are first and foremost
socially negotiated phenomena. All religious ideas, however, must first surface
somehow. At some point in evolutionary history, an ancestor or close rela-
tive to the human race had the first proto-religious idea. We have no way of
knowing what the content of this thought might have been, what part of life
in the Pleistocene inspired it, or if this first believer ever shared it with oth-
ers of his kind. What we do have, however, are advances in cognitive science
and evolutionary psychology that allow us to venture certain guesses about
what kind of mind may have afforded this first religious thought, a mind so
powerful and playful that it grabbed bits and pieces out of thin air, to create a
marvel of its own devising. Imagination had entered the scene, and the world
would never be the same again
Imagery and make-believe are words commonly used by critics of reli-
gion. Many aspects of ritual involve object substitutions and juggling of facts,
which we know well from fiction and childrens play. The abilities to embrace
stories or be carried away by mystical experiences are hallmarks of spiritual-
ity all over the world. The ability to imagine things that one has never seen,
explore cognitive space or effortlessly reverse roles and identities, is the very
same that allows us to make new inventions by exploring and indeed breach-
ing conceptual domains, infer details about thoughts, feelings and motiva-
tions of our fellows, and discuss matters of cognitive science, for instance, in
the abstract.
In the following, I attempt to gather some of the many strings relating
religion to imagination, creativity, decoupling, pretend play and even mad-
ness. I believe that the study of human imagination in its many forms will

religion and the emergence of human imagination

allow us a glimpse of some of the important processes for the emergence of

religion in the past and today.
Imagination in informal use carries many connotations, including crea-
tivity, imagery, play, propositional reasoning, dream and delusion. In some
respects all of these are the subject of this study, because they may well
prove to be a posy of very different functions, emerging as the phenom-
enological and linguistic cluster, that we commonly name imagination. In
cognitive science, imagination is often investigated in terms of mental rep-
resentation, cognitive fluidity, decoupling and mental imagery, but more
often than not, a study will focus on a single narrow aspect of the phe-
nomenon. This is a good scientific habit, but doesnt contribute much to
the overall understanding of a phenomenon such as imagination, which is
far more than its part in everyday experience. Here I will therefore afford
the luxury of a sightseeing tour through varying aspects of the subject, to
divulge how the study of imagination might best benefit the field of reli-
gion, cognition and culture.
In the following, I present some of the central defining attributes of imag-
ination, and try to explain these and some of the related theories, within the
framework of cognitive science. Notably, I focus on experience, representa-
tion and decoupling within the frameworks put forward by Gerald Edelman
and Giulio Tononi (Edelman & Tononi 2001; Edelman 2004) and Alan
Leslie (1994). Further, I devote some space to the sharing and controlling
of imaginative content, and the roles of belief, context and intentionality. I
think that imagination can be viewed as a mental faculty, since it is well-nigh
impossible to shake off the bonds of linguistic definitions and everyday expe-
rience, but in that case it must be viewed as an epiphenomenon of several
co-developed and co-dependent processes. This chapter is an exploration of
some of these interlinked components, as well as their impact on religious

An initial definition of imagination

I will start this exploration by offering my own working definition of imagi-

nation. The word shows up many times over in the MIT Encyclopaedia of
Cognitive Sciences, but remarkably it does not have a formal entry. In the
Oxford Advanced Learners Dictionary, on the other hand, it is defined as:

1. the ability to create pictures in your mind; the part of your

mind that does this 2. something that you have imagined
rather than something that exists 3. the ability to have new
and exciting ideas. (Oxford University Press 2000)

andreas lieberoth

Further, the verbal form, imagining, is stated to mean both believing and not
believing. Despite its shortcomings, this definition covers the informal con-
notations of the word quite well, and it even captures the notion of imagi-
nation as at once a cognitive space, a process and a function. We probably
see these many connotations exist because imagination covers a wide array
of functions that are not necessarily the work of a single faculty, even if so
implied by language. For the sake of consistency, Id like to suggest some basic
features of the phenomenon. This is not meant to be an exhaustive defini-
tion, but simply to keep us all on the same page. My initial characterization
consists of three defining points:

1. decoupling from present reality;

2. conscious (but not necessarily intentional);1 and
3. potentially creative/generative.

Even some children with autism can relate to the concept of pictures in the
head (Harris & Leevers 2000). Imagination or imagery at least, is a fun-
damentally personal phenomenon, and therefore I will accept introspective
experience as a central point. It is separate from external stimuli, but still sen-
sitive to them, and can generate new meaningful content. It allows the crea-
tion of new inventions, and notably, thinking with-and-about existing ones.
Seen from the angle of cognitive (neuro)science, imagination as a whole
can, I think, be understood in terms of activation in the neural structures
used for whatever is imagined such as perceptual systems for visual imagery,
language-areas for semantics, motor-strip for movement, etc. In other words,
imagination is generated by novel combinations of existing neural mappings.
It is highly context-dependent, and changes unnoticed within the blink of an
eye. To reflect this, the cognitive basis for imagination may be:

4. activation of applicable neural systems/structures;

5. source inputs from the senses and their systems; and
6. ties to memory systems/processes.

To offer a less formalized definition, I think that imagination must be viewed

as several things: the process of recalling and consolidating items and qualia
into integrated wholes, the conscious experience when this is achieved, and
the manipulation of such mental content over shorter or longer periods of
time. Thus imagination would allow me to scrutinize any acquired super-
natural concept in my mind's eye, invent fantasy-creatures of my own from
the building blocks at my disposal, and/or employ such concepts in thinking.

religion and the emergence of human imagination

Cognitive explanations of imagination

A cognitive review of imagination is no easy task. The term appears often

enough, but the lack of definitions or formal investigation is striking at first
glance. Pretence, decoupling, imagery and creativity have been studied and
discussed in detail, and imagination is even described as one of the primary
sources of mental models, but most theories seem to focus on limited ele-
ments of the phenomenon. For this reason, this study started out by gathering
and skimming through the literature on pretend play, make-believe, theory
of mind, creativity, decoupling and imagination. These gave rise to the points
summarized above, and insight into many cultural and historical conceptions.
Before moving on to religion, I will try to explain the theoretical background
for choosing my six defining points.


Decoupling is probably the most salient descriptor of imagination. It can be

viewed as a level of thought completely separate from primary conscious-
ness, but reliant on the same cognitive mechanisms. A nave realist mind, or
primary conscious, is by default dependent on distally proximate stimuli
to interact with the world and act on information (Leslie 1994; Cosmides
& Tooby 2000; Edelman & Tononi 2001). Such primary representations
are experiences bound to the here-and-now. Decoupling is the freeing of the
mind from immediate perceptual inputs, and the ability to entertain meta-
representations at various levels, propositionals, memories, pretence and any
other form of contemplation unrelated to direct inputs from the periphery
(Leslie 1994; Boyer 2001). Studies in pretend play by Alan Leslie show that
children are very good at distinguishing fact from fancy at a very early age
even on a theoretical level (Leslie 1994; Friedman & Leslie 2007). Cosmides
and Tooby (2000) suggest that source- and attitude-tags attached to all
entries in the mental database help keep track of information, and create a
barrier between our knowledge stores, and doubtful, suppositional or imagi-
nary ideas. Something to this effect does seem to exist in the human mind,
but the theory of scope syntactical tags may be too much of a filing-system
metaphor to really fit the way the brain works. I imagine that it could be akin
to somatic markers, and/or co-activation of separate value systems for, for
instance, autonoetic memory.

andreas lieberoth

Memory processes

When looking for neural or cognitive correlates of imagination, both the

process and its components must be taken into consideration. The work-
ing memory (WM) theory (Baars & Franklin 2003; Baddeley 1986; Miller
et al. 1960) offers a neat, if perhaps overly schematic model of retrieval and
processing, which in itself explains many features of imagination. Directing
the mind towards decoupled content may, simply put, involve retrieving stuff
from long-term memory stores and semantics, and then manipulating it.
The theory has the virtue of explicitly linking what we might call decoupled
thought-sequence to several levels of memory processes, based on solid experi-
mental evidence of short-term memory operations and capacity. Schematic or
computational models like this one may, however, be problematic, as meta-
phors of connected spaces and pipelines are often taken too literally. Thoughts
do not travel along neurons; activation of each neuron carries its own sig-
nificance, and gives rise to activation of others, creating a very complex and
open relationship between various parts of the brain. Thus, the spaces for
memory manipulation are distributed all over the cortex and only peripher-
ally managed by subcortical structures.


It is hard to say whether all imagination is conscious, since claims in this

direction are often based on introspection in a circular fashion (for instance,
McGinn 2004). Notions of cognitive incubation, the resting state network,
the Zeigarnik effect,2 the dynamic core hypothesis (Edelman & Tononi 2001)
and Freudian theories of the Unconscious certainly seem to suggest otherwise,
and I will not stress the point. There do seem to be, however, clear connections
between mental imagery and perceptual systems; imaginative scenarios can
even leave episodic memory imprints. Thus, at least some aspects of imagina-
tion, namely mental imagery and processes of a sufficiently high order, seem
to be unquestionably linked to conscious experience in various sense-modali-
ties. Even the swirls and geometrical patterns sometimes seen in altered states
of consciousness, called entoptic phenomena, seem to beckon the attention of
the minds eye even supposing they are just images of the neural structures
in area V1 of the visual cortex, as suggested by Lewis-Williams (2002). I will
not discuss, and definitely not try to explain, consciousness. This grand sub-
ject I leave to specialists such as Gerald Edelman and Giulio Tononi (Edelman
2004; Edelman & Tononi 2001) for reference. In very simplified terms, their
theory of neural group selection (TNGS) and the dynamic core states that
conscious experience is generated when various families of neurons in the

religion and the emergence of human imagination

brain fire in sufficiently intense patterns, which must also include certain tha-
lamocortical systems. When such neural mappings are activated, they can be
related to semantics and previous experience. Lots of processes are going on at
the same time, competing for prevalence in an evolution-like race to become
conscious and/or strengthen their connections for future activation. In much
the same way that socks in a tumble-drier might form coloured circles at high
speeds, loops of re-entrant activation distributed over various parts of the
brain compete for prevalence, and if sufficiently intense, generate discernable
experience. In Edelmans framework, higher-order consciousness liberates the
mind from the here-and-now by relating experience to memory, language and
an advanced sense of self. The theory thus includes elements akin to meta-
representation and decoupling, and presents a clear-cut framework for the
conscious experience of mental activity.


Over fifteen years ago, Gregory Currie (1995) suggested that to become
immersed in fictions such as pretence or reading a book, you must neurocog-
nitively emulate the mindset of the protagonist, or replicate how you yourself
would feel in the imaginary situation. This idea has since been supported by a
multitude of studies in simulation,3 showing how facial muscles are activated
in sympathy with even slight facial changes of others, and how recorded activ-
ity in motor cortices during finger tapping sequences are re-activated when
subjects remember or just imagine performing them. Understanding thought
and imagination as sequential and re-entrant activation of neural firing-pat-
terns, as suggested by Edelman, supports implications of theories on neural
simulation. Since any neuronal pathways may be activated along with basic
thalamocortical structures, the cortices used for sensing or feeling primary
inputs may also be assimilated into higher-order conscious experience. In
short, I suggest that imagination is made possible by using and manipulating
circuits already in place in the brain be they recently coded memories or
inherent motor functions on a decoupled level.

Inputs and building blocks

Generative thought is based on activation and recombination of neural map-

pings already in place, and it is quite possible that all meaningful conscious
experiences, including memory recall, are pieced together as the mind goes
along (Boden 1998; Mithen 1998; Modell 2003; Edelman & Tononi 2001).
Thus, the content of private representations or unusual ideas is not conjured

andreas lieberoth

out of thin air. Thoughts (with certain exceptions4) are directed at objects
(Sartre 1966), which contain imagistic, conceptual or culturally control-
led meaning, of some sort. Mental scenes, for instance, are composed from
many parts, which must be loaded separately into meaningful experience,
or remain implicit until activated by intentional scrutiny (Lieberoth 2006).
They are not recalled in their entirety from long-term memory, but created for
the occasion from bits and pieces fitting the overall sense of meaning and con-
sistency. The process of generating new meaning from things and constraints
already in place is so fundamental to the way we experience thought that only
notably innovative or weird ideas from children, artists and madmen really
strike us as imaginative.


Radically new combinations only rarely surface in our conscious experience,

because the mind is fundamentally wired to handle inferences about the dull
everyday world (Byrne 2005). Cognitive constraints, such as heuristics and
cognitive domains, let us draw quick inferences about the world (Boden
1998; Mithen 1998; Boyer 2001). This saves time and resources in everyday
life, but also forms the proverbial boxes that one must think out of in
order to be innovative. All minds exercise creative potential, but since cogni-
tion keeps to certain general parameters, only rarely with never-before-seen
results. Some innovations are new to the entire world, while most are just
new to a single mind; a toddler is unlikely to invent anything that hasnt been
seen many times over in human history, but an idea might be genuinely new
to himself, none the less. Steven Mithen (1998) suggests that cognitive fluid-
ity, which basically means breaching and exploring the bounds of rationality,
is the defining point of our great evolutionary leap forward (Mithen 1998;
Boden 1998). When breaches happen, new and strange combinations appear,
and if they can be transferred into outer representation, ideas can even be
shared. Imagination thus affords personal and cultural innovation by creat-
ing a free space for thinking with-and-about cognitive content, but as space
which normally respects the day-to-day rules of reason. This might be why
madness can be a great catalyst for creative discovery, as explored by Daniel
Nettle (2001).
It is hard to say what imagination really is, apart from something in our
linguistic and introspective experience. Maybe an all-new faculty for imagi-
nation emerged from our forefathers existing mental faculties interwoven
into more complex functions; and maybe there is no imagination, except
a palette of more basic systems. I think we can assume a cognitive faculty
for imagination, but we must always consider the individual attributes as

religion and the emergence of human imagination

unrelated cognitive functions of memory, meaning and experience among

others. More experimental work clearly needs to be done at this point, but
some theories like that of Edelman seem to be on the right track.

Sharing and controlling imagination

Like religion, imagination is sometimes degraded as childish folly, separate

from rational reflection. We can normally tell mental imagery from reality
quite naturally, but this is a crude distinction which cannot easily be made
on the level of cognitive systems. As shown by Ruth M. Byrne (2005), hypo-
thetical thought and imaginative fancy are better viewed as one and the same
process, with different objects. Byrne explored in detail how inferences on
possible worlds differed according to human dispositions towards counter-
factual reasoning about what might be, or had been, moral judgement, etc.
Banishing imagination to the realms of childhood, poetry and madness, thus
seems to be a rather unfortunate by-product of Western enlightenment.
Many things may be explored by looking at childrens play. Childrens
structures of plausibility are less cemented than those of adults a fact that
has led to some mistaken notions that adult brains lose their general capacity
to learn but at 1824 months, they become able to successfully distinguish
pretend and simulacra from reality, and readily perform mental transforma-
tions and imagined scenarios with high levels of complexity. This emergence
of decoupling can be understood as an extra level of representation or con-
ceptual space, which inferences respect, and within which they work (Leslie
1994; Friedman & Leslie 2007). Danish preschoolers joint games are com-
monly bracketed by the verbal tag then we just said that to introduce
the premises of play, or open up for a representational negotiation, almost like
preliminary acts in religious ritual. In normal children, pretend play is gradu-
ally developed to a higher degree of independence from external cues, and
towards prolonged sustainability and tracking of pretend scenarios (Schwebel
et al. 1999). Also shared fantasy becomes more sophisticated and common
with age, including numerous real and pretend actors.
I have investigated elsewhere how verbal inputs influenced the flow of pri-
vate representation in fiction-immersion and role-playing (Lieberoth 2006,
2007). In fantasy role-playing, all players contribute bits and pieces to the
same story. My study of this process confirmed how perceived and imagined
scenes are very malleable, and gradually constructed by independent build-
ing blocks, as suggested by Edelman & Tononi (2001). I call this process
representational negotiation because active imagery is constantly subjected
to contradictions between existing repertoires of generalities and various inner
and outer contributions. This sort of deep fiction-immersion can probably be

andreas lieberoth

compared to a novice shamans guided spirit journeys (Lewis-Williams 2002),

although psychotropic substances play a part in these as well, which (to my
knowledge, at least) wasnt the case in my study. The process of role-playing
shows how easily visual mental scenes may be modified, by verbally calling
attention to certain diegetic5 features and qualia (in my study disguised as
questions like what colour is the roof of the house in your minds eye?).
The imaginer often fails to realize that his mental construction is being modi-
fied by the outside input, and instead simply feels that attention is shifted to
something already there. Moreover, studies in modularity (Modell 2003)
suggest that visual imagination might be more readily stimulated by specific
content such as colour, motion and contrasting shapes, even if in the form of
input from different modalities, such as spoken words in my example. The dif-
ference is a matter of milliseconds, but such experiments suggest that a pro-
pensity to favour certain cognitive elements might be found on much more
basic levels than suggested by Boyers theory of minimally counterintuitive
items.The process of generating and altering imagined scenarios suggests to
me that imagination must be viewed as a sequential process involving vari-
ous competing levels of attention, intent and input. The metaphoric under-
standing of imagination being played out on an inner screen is attractive,
and some studies, for instance Colin McGinns monograph Mindsight from
2004, focus on imagination primarily as mental imagery. The visual aspect
of imagination is probably very salient because we humans rely on sight for
navigation in the physical world. Since visual brain-areas developed to pri-
macy some time before our forefathers decided to come down from the trees
(Dunbar 1996), this modality is heavily represented in the neural activations
which make up conscious experience. More often than not, however, phi-
losophers such as McGinn (2004) or Sartre (1966) get side-tracked on visual
introspection, and fail to appreciate the generative and semantic nature of
imagery, while grappling with the fact that consciousness does not exhaust all
aspects of the things to which it refers (Edelman 2004). Further, the idea of
pictures in the mind leads to the homunculus-problem, which we do not
have space to address here.
Viewing imagination as a process, or flow of consciousness, which is modi-
fied according to competing inputs, opens up for a number of considerations.
For instance, WM theory shows us that short-term memory systems may
juggle somewhere between three and seven active processes at a time, but
according to Edelmans TNGS, several others will constantly be stepping up
in competition to these. In this sense, unconscious processes are constantly
vying for prevalence in action, emotion and conscious experience, and there-
fore intentionality should only be seen as a spotlight falling upon a very lim-
ited segment of the many neural maps that are active in the brain at any one
time. Thus, many religious inferences may be triggered by inputs like those in

religion and the emergence of human imagination

Barretts now classical stories, but immediate intuitive responses and mental
imageries may quickly be overruled by a more theologically correct inten-
tional spotlight. David Lewis-Williams suggests that consciousness ranges
from the most alert problem-oriented state (in which, I think, theologically
correct analysis is more prevalent), over daydreaming (with less intentional-
ity), to more autistic or inward-turned states like dreaming, and ultimately
hallucination or unconsciousness (Lewis-Williams 2002: 125). On this spec-
trum, intentional directedness and control over content declines while con-
sciousness may remain. This is supported by imaging studies showing that the
firing of neurons does not, in fact, decrease during low-wave sleep. Instead, it
becomes more staccato and seems less integrated in re-entrant processing as
time proceeds, which is wonderful proof that the brain goes on working with-
out intentionality or inputs from the periphery (Edelman & Tononi 2001).

Evolutionary origins of strange ideas

If humans were nave realists there would be no religion. If we thought only

about the immediate inputs to our senses, we would not contemplate crea-
tion, talk about past experience or engage in pretend play. We would imagine
the occasional non-existing predator in the bushes, but that hardly qualifies
as religion in itself. Any idea about life, the universe and everything would
dissolve almost as soon as it entered our primitive minds, and even with
rudimentary autobiographical memory any semi-supernatural mental models
would never survive beyond the lifespan of any single individual. Off-line
thinking and memory thus seems to be the first step towards a cognitive basis
for religious thought, with external representation such as language (or semi-
otic capacity, at any rate) closely at their heels.
Many theories of imagination come handily packed with an evolutionary
perspective. Since we are looking for the origins of religion, cognition and
culture, some of these give valuable hints at a possible emergence of imagina-
tion in human prehistory, which in turn made religious ideas and representa-
tions possible. Roughly 3040 millennia ago, our ancestors started leaving
pictures and other archaeological evidence of cultural sophistication behind,
which suggests that something important happened. Mental domains were
breached and combined into entirely new creations (Boden 1998; Mithen
1998), and tools started surfacing that require more than strokes of luck or
simple peer-learning to create. Our contemporary cousins, the Neanderthals,
never reached this level of innovative achievement, and died out relatively
soon after. A staggering amount of findings from this symbolic explosion
look religious in nature, and so it seems that when creativity entered the
scene, religious ideas and/or the need to express these were close as well.

andreas lieberoth

The prevalent trend in evolutionary psychology links development of

higher-order cognitive processes, which includes meta-representation, decou-
pling and representational memory, to the demands of life in the social niche.
Among these theories, probably the most famous study was done by Robin
Dunbar and Leslie Aiello, who pointed to archaeological evidence of a link
between relative hominid neocortex volume and the size of social groups
(Aiello & Dunber 1993; Dunbar 1996). To Dunbar, our large brains evolved
to manage social relations more efficiently than through mere grooming, and
language was a product of this development. The emergence of either con-
cept or communication might conceivably have led to the other, but the two
were most likely achieved through a process of mutual scaffolding. This is of
particular interest because semiotic capabilities allow cultural traditions to
emerge from shared mental representations (Sperber 2000; Boyer 2001), but
also because semantic concepts according to Edelman play a part in forming
and manipulating mental imagery and a sense of self. In other words, having
a consciously available symbol for, say, thirst or monkey, may help imag-
ining it as a general concept away from peripheral inputs, and relating it to
the self and hypothetical situations. Language is based on combinations of
elements (words) already existing in the mental dictionary, arranged accord-
ing to certain rules of reference and syntax. These two aspects are primarily
represented biologically in Brocas and Verniches areas, which were accompa-
nied in evolution by the development of specialized somatosensorical oral and
facial structures, relevant to speech. Brocas area has homologous counterparts
in other mammals, but interconnection to specialized auditory areas is only
found humans (Dunbar 1996; Gazzaniga et al. 2002).
While Edelman argues that the development of semantic meaning led to
the realization of higher-order consciousness and thus decoupling (Edelman
& Tononi 2001; Edelman 2004), Dan Sperber (2000) believes that meta-
representational ability gave birth to communication, and thereby seman-
tics. In Sperbers view, meta-representation developed not for language, but
as part of a social arms race in which the best conniver or mindreader had
the advantage. Inferential communication was just a lucky side-effect of this
evolutionary competition, and decoupling was already firmly in place to allow
our forefathers to second-guess each others intentions. This is an offshoot of
Alan Leslies theory-theory of human empathy (for a review, see Friedman &
Leslie 2007). As noted, Edelman disagrees with Sperber over the order of devel-
opment. According to his theory, language and semantics are bootstrapped,
and only when concepts and an advanced feeling of self were in place, could
higher-order consciousness, and thereby decoupling, have developed.
The discussion above is well known to many, but the development of
decoupling and dissociation from immediate reality may have had dire con-
sequences to the developing and fragile human psyche. Psychologist and

religion and the emergence of human imagination

anthropologist Daniel Nettle (2001) suggests that dispositions for schizo-

phrenia and some emotional disorders remain widely distributed in the gene-
pool because those very same genes are responsible for human imagination.
Schizoid psychosis is a case of capacity for cognitive fluidity in higher-order
consciousness getting out of control. Breakdown in scope syntax could offer
an adequate explanation for lacking ability to separate personal thoughts
from perceptual inputs, which is often seen in the development of delusion
in schizophrenics (Cosmides & Tooby 2000). Depressed mammals are not
uncommon (especially if removed from their natural habitats by humans),
but one must have a higher-order conscious to form a deluded belief. We can
all experience seemingly real scenes in dreams, but in intentional cognizant
states, the mind is very good at sorting fact from fiction. With this in mind,
it is hardly surprising that delusion often involves culturally salient religious
content, and that many harsh secularists have attributed religion to some
sort of mental deficit in primitive, or just gullible, people. The only problem
with this view is that healthy minds picture supernatural images just as well
as insane ones, but only rarely get carried away by them. Thus the core of the
problem seems to be belief not representation in itself.

Imagining gods, ghosts and dragons

The creation of something new is not accomplished by the intel-

lect but by the play instinct acting from inner necessity. The crea-
tive mind plays with the objects it loves.
(Carl Gustav Jung 1971: 123)

So, what have we learned about imagination? It is a creative and generative

faculty which, albeit not entirely free from the shackles of inputs and culture,
can play around with anything that exists in our episodic or semantic mem-
ory. Thus we can think with-and-about whatever gets called up through
distributed activations in any given situation. Mostly this is a very bland and
rational thought-process, but it can just as well deal with supernatural items.
Imagination is not picky about content, but it has ways of keeping track of
its proper contexts.
It should be fairly obvious that religion requires imagination. A devotee
must be able to picture his objects of devotion, and infer about their rel-
evance and action in his day-to-day existence. This involves mobilizing a lot
of mental matter, perhaps with help from cognitive artefacts such as scrip-
ture or icons, and synthesizing new content from such building blocks, in
the context of everyday reality. Employing imagination does not mean that
religion is irrational, for as we have seen, imagination is every bit as sensible

andreas lieberoth

as other thought processes. We start playing around with our repertoire of

reality building-blocks at a very early age, and even though pretence is often
scaffolded and developed through social interactions, it seems to manifest
spontaneously in all normal humans. Once inside your head, you can think
with-and-about any culturally postulated item, and you can even make up
your own supernatural representations with a bit of effort (see Boyer 2001 for
a further discussion of these).
The apparent inability to process strangeness or create new combinations
in autism, seems like a tale of what to expect if we did not have free use of our
brushes on the imaginary canvas. Not only would supernatural ideas not man-
ifest spontaneously, but we would be unable to imagine them for ourselves,
even if prompted by others. Some evidence suggests that Neanderthals had
little innovation in their 50,00060,000 years of thriving, despite relatively
large brains (Diamond 1992). Even though there are examples of Neanderthal
burials, there is little indication of any religious content. Neanderthals might
simply not have imagined an afterlife or soul. Their simple burials seem to
be concerned with the here and now the content of primary consciousness
which may be an indication that religious thinking is made possible only
by the emergence of our comparatively more sophisticated, higher-order con-
sciousness (Lewis-Williams 2002).
Play often seems very elaborate and staged, and it has been suggested that
exaggerated acting (manner cues) helps children know when something is
just pretence (see Friedman & Leslie in press for a review). This could seem
to point to imagery in ritual or myth as a controlled process of intention-
ally directing the minds eye (Modell 2003; McGinn 2004; Sartre 1966).
Dreams, involuntary associations to highly emotional memories (Modell
2003; McCauley & Lawson 2002), hallucinations under sensory depriva-
tion or altered states of mind (Lewis-Williams 2002; Edelman & Tononi
2001), and the intricate delusions of schizotypic psychotics (Nettle 2001),
however, shows us that imagination is only wholly intentional in very control-
led conditions if ever. Many factors are at play in the contest for conscious
prevalence, and will is only one of them. I have mentioned how individual
mental representations may be shared and influenced by way of social proc-
esses, labelled representational negotiation (Lieberoth 2006, 2007). Religion
is a social and cultural phenomenon, so such negotiations will usually direct
mental representation toward socially and culturally salient imagery and/or a
theologically correct standpoint. This means that fantasy can be guided and
manipulated by public representations under all sorts of conditions not
just in apprentice shamans or ritual action and that language and semantics
would sometimes overdetermine raw qualia. Therefore the scientific commu-
nity of cognitive science should be careful about distancing itself too much
from the school of social constructivism.

religion and the emergence of human imagination

Salient emotions or contextual triggers may also influence which mem-

ory-elements are activated as ingredients in imagination and, thus, in inter-
pretation of events (Modell 2003). This is commonly seen in victims of
post-traumatic stress disorder, but a similar effect may be produced by way
of high sensory pageantry in ritual as described by Whitehouse (2002) and
McCauley & Lawson (2002). In such cases, reuse of routines or symbolic
elements may even target particular prominent memories or meanings, guid-
ing imagination very efficiently. This underlines the point that context is all-
important to imagination. Inputs and chemically active synapses determine
which cards the mind is most likely to deal in any given mental fantasy,
and contextual influences include anything from physical circumstances to
the vicissitudes of social, cultural and linguistic reality. Inputs to our mind
(including other thoughts) influence the competition between re-entrant
neural activations all over a brain, and something will always emerge victo-
rious. Therefore a mind can be primed to link any sign in the world to reli-
gious meaning. Furthermore, this process is self-reinforcing, as new memories
are generated to add to the pool of reference, and strengthening of neural
Living within a social context of culturally postulated supernatural agents
allows these meaningful elements to be activated as elements in imagery.
Further, culture might easily contain situations and artefacts which elicit
the mobilization of such supernatural representations. When higher-order
consciousness appeared in our forefathers, some might have had genuinely
original religious ideas, while others just used their imagination to think with-
and-about those acquired through social interactions and this division of
labour is still seen in religion today. Decoupling does not just allow the mind
to wander away from present reality, but also to impose decoupled content
on that reality.
The potentially playful or generative character of imagination is one of its
defining points. If this was not the case, all supernatural reasoning and tales
would have to be attributed to erroneous reasoning, madness or faulty men-
tal models. The generation and use of ideas like talking trees, dead relatives
intervening on the behalf of the living, and malignantly inclined computers
plotting the end of the world takes more than a hyperactive agency detec-
tion device. All modern humans can have weird ideas or be fascinated by
fiction from time to time, but such notions are usually discarded as fantasy
not adopted in everyday belief. Below, I will briefly explore the boundaries
between fantasy and reality.

andreas lieberoth

Belief, social sensibility and reality boundaries

Alan Leslies theory of mind mechanism (ToMM) theory suggests that the
capacity for pretence is probably yoked to the development of conceptual
understanding of pretence and other mental states (Leslie 1994; Friedman &
Leslie in press). This is consistent with Edelmans insistence that semantics
are a prerequisite for higher-order consciousness. This means, that when the
capacity for pretence6 is fully developed, the child is also capable of infer-
ring about her own and others attitudes and mental states, such as pretence,
(false) belief, and so on. The child, in other words, knows that some things
can lack in truthfulness. If this is the case, then the conceptual understanding,
and scope syntactical isolation of imagined representations, should in fact pro-
tect the young mind from entertaining most unverifiable beliefs. In this sense,
the mechanisms of decoupling are a hindrance to believing too far-fetched
(for instance religious) representations, while at the same time allowing the
child to freely think with-and-about them. The human imagination is both
very pragmatic and playful. Dan Sperber (2000) further suggests that check-
ing for inconsistencies in information might be an important weapon in the
arsenal of self-defence against false communication, so empirical or cultural
structures of plausibility must stand ready to back up any perceptually unveri-
fiable propositions. This supports Boyers notion that supernatural items must
fit pretty well with established modes of thought to be well-received, but still
does not account for religious belief. A socially competent mind with a nor-
mally developed imagination should simply weed out any notion too weird
to fit in the everyday setting of life. This is somewhat reminiscent of the
secularist speculation that only deficient (be they primitive, unenlightened
or diseased) minds would take to superstition, which we know is a theory
that will not stand scrutiny. Religion and religious invention appears much
today as 40 millennia in the past, so the notion that religion is the product
of faulty proto-minds will not do, even if the first religious mind was in fact
less advanced than ours which is quite likely, but ultimately a matter of
Even though systems for decoupling would seem to shield the mind from
religion, one important aspect might hold a final clue to its persistent survival:
the social responsiveness to which meta-representation evolved. Any child
with a fully developed ToMM should be able to understand when others are
serious about, for instance, the concept of witchcraft, which may support the
social transmission of such supernatural items. This could suggest that beliefs
are more memetically contagious than mere ideas. When scope-syntax is
applied to a representation, it includes attitude-understanding, and this may
be a very simple part of the answer to the Mickey Mouse problem: when,
for instance, primary caregivers signify serious belief in a communication

religion and the emergence of human imagination

with-and-about any item, the communication is massively more credible to

the socially respondent mind than, say, a childs cartoon which no adult looks
at twice even if both ideas are equally noticeable in everyday culture and
rich on minimally counterintuitive traits. Laughter and panic is contagious,
and so, it would seem, is seriousness and belief.

Conclusions and research perspectives

All activation of supernatural items employs a bit of creative imagination

because all mental content is constructed by use of mental capacities already
in place. If imagination is, as I suggest, mainly based on the recontextualiza-
tion of existing mental content, in negotiation between competing endog-
enous and outer influences, then semantics and memory systems must play
a crucial part of the process. Man is a creator of meaning and an active con-
tributor to the sociocultural surroundings rather than just a static processor of
the world inside and outside his head. Items must be retrieved from memory,
and new ideas may be stored, possibly with a scope-syntactical tag identifying
them as originated in imagination. This makes supernatural representations
equal to all other items in our mind, but also handicaps them in a way.
Imagination is able, by virtue of itself, to filter any erroneous representa-
tions or ideas away or at least mark them as figments of the imagination.
This is the most prominent chink in the armour of the long-standing implicit
notion that religion and imagination are linked. The notion of the primitive
man mistaking his own ideas for reality is dead wrong, unless the hominid
in question was suffering from seriously schizoid psychosis or had simply yet
to develop a fully functional scope-syntactical barrier between primary con-
sciousness and decoupled thoughts. Religion, however, pervades our modern
world of burger franchises, neuroscience and mobile gaming, and religious
innovation still occurs in the life of all believers. I think that religion prevails
because of factors in our minds stronger than rational thought, namely,
social sensibility. Our brains may very well have evolved to negotiate complex
social relationships, and so social and discursive receptivity will sometimes
overwrite better judgement. In a way, this makes complete evolutionary sense,
since our most prominent racial advantage may be the ability to share knowl-
edge over the span of generations.
Studying imagination shows us how minds think with-and-about religious
items and how these are negotiated between building blocks in the mind and
contextual circumstances which can be both artificially and/or socially con-
structed. Imagination is personal, but highly susceptible to contexts, values
and all sorts of implicit frames of reference that we gather in the course of our
social and symbolically laden lives. If imagination is tied to semantics, then

andreas lieberoth

semantics may tie to fantasy and set the stage for mental scenes with all sorts
of actors including those of our rich religious repertoires.
Imagination could definitely use further formal investigation within cogni-
tive studies, and many would benefit from the knowledge, but it must first be
broken down into underlying attributes, each evaluated as a separate cognitive
function. I hope that the present exploration of the phenomenon in relation
to religion will benefit the study of both, and open up opportunities to do
experimental studies in the area of religion and imagination.


1. Unless understood as a sort of directedness or attention of the minds eye.

2. The mind seems to keep unfinished tasks at the ready for later reactivation, and even
work on them when the conscious is otherwise engaged.
3. For a review, see Barsalou et al. (2003).
4. Deeply fevered or psychedelic experiences may be vivid while including no semanti-
cally meaningful objects. See Lewis-Williams (2002) for more on entoptic phenomena.
Further, it is a complex discussion whether, for instance, somatic or olfactory experience
can be labelled with meaning, in the same way that we ascribe semantics to visual or
linguistic stimuli. It is interesting to note that the sense of smell does not pass to the
neocortex by way of the thalamus, and therefore may induce more direct or emotional
associations than other modalities.
5. From film studies, now used much in role-playing theory: content that is present in
the framework of fiction. For instance, an actor speaking or the sound of gunshots is
diegetic sound, but background music is not.
6. And imagination; Friedman and Leslie (2007) explicitly separate the two, but let us for
the sake of argument assume that they are very close relatives.


Aiello, L. D. & R. I. M. Dunbar 1993. Neocortex Size, Group Size, and the Evolution of
Language. Current Anthropology 34(2): 18493.
Baddeley, A. D. 1986. Working Memory. Oxford: Oxford University Press.
Baars, B. J. & S. Franklin 2003. How Conscious Experience and Working Memory Interact.
Trends in Cognitive Science 7(4): 16672.
Barsalou, L. W., P. M. Niedenthal, A. K. Barbey & J. A Ruppert 2003. Social Embodiment.
Psychology of Learning and Motivation 43: 4393.
Boden, M. A. 1998. What is Creativity?. In Creativity in Human Evolution and Prehistory,
S. Mithen (ed.), 2260. London: Routledge.
Boyer, P. 2001. Religion Explained. London: Vintage.
Byrne, R. M. J. 2005. The Rational Imagination: How People Create Alternatives to Reality.
Cambridge, MA: MIT Press.
Cosmides, L. & J. Tooby 2000. Consider the Source: The Evolution of Adaptations for
Decoupling and Metarepresentation. In Metarepresentation, D. Sperber (ed.), 53116.
New York: Oxford University Press.

religion and the emergence of human imagination

Currie, G. 1995. Image and Mind, Cambridge: Cambridge University Press.

Diamond, J. 1992. The Third Chimpanzee. London: Harper Perennial.
Dunbar, R. 1996. Grooming, Gossip and the Evolution of Language. London: Faber & Faber.
Edelman, G. M. 2004. Wider than the Sky. London: Penguin.
Edelman, G. M. & G. Tononi 2001. Consciousness: How Matter Becomes Imagination. London:
Friedman, O. & A. Leslie 2007. The Conceptual Underpinnings of Pretence: Pretending is
Not Behaving-as-If . Cognition 105(1): 10324.
Gazzaniga, M. S., R. J. Ivry & G. R. Mangun 2002. Cognitive Neuroscience: The Biology of the
Mind. New York: W. W. Norton.
Harris, P. L. & H. J. Leevers 2000. Pretending, Imagery and Self-Awareness in Autism. In
Understanding Other Minds, 2nd edn, S. Baron-Cohen, H. Tager-Flusberg & D. J. Cohen
(eds), 182202. Oxford: Oxford University Press.
Jung, C. G. 1971. The Collected Works of C. G. Jung, vol. 6: Psychological Types. Princeton, NJ:
Princeton University Press.
Leslie, A. 1994. Pretending and Believing: Issues in the Theory of TOMM. Cognition 50:
Lewis-Williams, D. 2002. The Mind in the Cave. London: Thames & Hudson.
Lieberoth, A. 2006. Drawing on the Canvas of Imagination. Paper presented at Playing
Roles, 31 March, Tampere University.
Lieberoth, A. 2007. Technologies of Experience. In J. Donnis, L. Thorup & M. Gade (eds),
Lifelike, 3550. Copenhagen: Knudepunkt/LLR.
McCauley, R. N. & E. T. Lawson 2002. Bringing Ritual to Mind: Psychological Foundations of
Cultural Forms. Cambridge: Cambridge University Press.
McGinn, C. 2004. Mindsight. Cambridge, MA: Harvard University Press.
Miller, G. A., E. Galanter & K. H. Pibram 1960. Plans and the Structure of Behavior. New
York: Henry Holt.
Mithen, S. 1998. A Creative Explosion? In Creativity in Human Evolution and Prehistory,
S.Mithen (ed.), 15593. London: Routledge.
Modell, A. H. 2003. Imagination and the Meaningful Brain. Cambridge, MA: MIT Press.
Nettle, D. 2001. Strong Imagination. New York: Oxford University Press.
Oxford University Press 2000. Oxford Advanced Learners Dictionary, 6th edn. Oxford: Oxford
University Press.
Sartre, J.-P. 1966. The Psychology of Imagination. New York: Washington Square.
Schwebel, D. C., C. C. Rosen & J. L. Singer 1999. Preschoolers Pretend Play and Theory
of Mind: The Role of Jointly Constructed Pretence. British Journal of Developmental
Psychology 17: 33348.
Sperber, D. 2000. Metarepresentation in an Evolutionary Perspective. In Metarepresentations,
D. Sperber (ed.), 11739. New York: Oxford University Press.
Whitehouse, H. 2002. Arguments and Icons. Oxford: Oxford University Press.

The origins of religion, cognition and
culture: the bowerbird syndrome
Luther H. Martin

[T]he emergence of culture remains a mystery to man. It will so

remain as long as he does not succeed in determining, on the bio-
logical level, the modifications in the structure and functioning of
the brain, of which culture was at once the natural result and the
social mode of apprehension. (Claude Lvi-Strauss 1976: 14)

We wished to awaken the feeling of mans sovereignty by showing

his divine birth: this path is now forbidden, since a monkey (ein
Affe) stands at the entrance. (Friedrich Nietzsche [1887] 1971:
4, cited in Foucault 1977: 142)

As long as there are animals to behave and humans to wonder

why, cognitive interpretations and explanations will be offered.
In our view this is not only permissible, it is often enlightening.
Sometimes it is even science. (Jamieson & Bekoff 1996: 756)


The pursuit of origins as an explanatory strategy has a long and generally

unproductive history of increasingly diminishing returns (Masuzawa 2000).
This was especially so during the nineteenth century when this quest, applied
to human history, presumed a social-Darwinian view of evolutionary progress
in which a natural history replaced but continued to function in the tel-
eological manner of an earlier divine providence. Influenced by Friedrich
Nietzsche, who had resoundingly challenged such views at the end of that
century, subsequent commentators like Michel Foucault (1977) and Edward
Sad (1985) have usefully distinguished between such notions of origination,

the origins of religion, cognition and culture

with their ontological and metaphysical assumptions illustrated, for exam-

ple by the locative premise of H. G. Wellss Time Machine (1895) and
those of beginnings as a methodological strategy designed to indicate, clar-
ify, or define a later time, place, or action (Sad 1985: 5). In this sense, to
speak of origins is to speak mythologically (ibid.: 372) whereas to speak of
beginnings is to speak historically (ibid.: xvii). If , as this point had already
been expressed by Durkheim, we are to understand [by origin] the very
first beginning, the question has nothing scientific about it, and should be
resolutely discarded (Durkheim [1915] 1965: 20). We might, therefore,
reformulate the problem here under consideration as the evolutionary his-
tory of religion, cognition, culture. We can consider evolutionary theory to
be, in other words, a historiographical theory of what Sad called humanly
produced, and ceaselessly re-examined beginnings (Sad 1985: xiii, 5). It is
not incidental that Darwin published his Origin of Species (Darwin [1859]
2003) during the mid-nineteenth century, precisely the time when history
was first being established as an academic discipline in European universi-
ties. For Darwins theory is but the historicization of biology. And, although
Darwin was unaware of Gregor Mendels contemporaneous work on genetic
heredity (Mendel [1865] 1965), I assume with contemporary evolutionary
theorists that the object of evolutionary study is not any species or group but
is genetics (e.g. Dawkins 1989).
But which Darwinian mechanism of genetic history is referenced?
enhancement of survival through natural selection, which has been the gen-
eral presumption in such studies? or enhancement of reproduction through
sexual selection, a mechanism that has been somewhat neglected but is
receiving increased attention? or both? This is a significant distinction since
breeding always takes precedence over survival when they come into con-
flict as they often do (Ridley 1993: 20) and as the perennial example of
costly and conspicuous peacock tails instantiates.
And what is the relationship of our evolutionary history to culture? Most
evolutionary psychologists and cognitive scientists consider cultural forma-
tions to be based in our evolutionary history. Language, for example or at
least a disposition towards learning language seems to have been, in some
sense, selected for. But many of these same researchers conclude that religion
is a by-product of the naturally selected mental capacities of Homo sapiens
sapiens. If this is so, what then is the sense of referring both to language and
to religion as cultural as they typically are? We cannot, in other words,
simply assume an evolutionary or cultural parity for such human characteris-
tics as language and religion. In contrast to views of cultural singularity (and
essentialism), Henry Plotkin has observed that culture is not some single
thing. Rather, it comes in different forms (Plotkin 2003: 3). Because of
this aspectual character of culture (Poole 1986: 41415; see also Chapter 3,

luther h. martin

this volume), we shall defer a consideration of religion and focus initially on

the relationship of cognition and clearly evolved cultural traits.
The evolutionary anthropologists Kevin Laland and William Hoppitt note
that culture has been defined by many so narrowly that it is, by fiat, a dif-
ferentiating characteristic only of Homo sapiens (Laland & Hoppitt 2003:
15051) a vestige, perhaps, of our Judeo-Christian bias (e.g. Gen. 1:26-28;
Dunbar 2004: 14). Such restrictive definitions not only exclude inquiries
by biologists and animal behaviourists into the question of whether or not
other species have culture but also preclude an exploration of the evolutionary
roots of human culture and of humans place in nature generally (Laland
& Hoppitt 2003: 150151). As the pioneering neuroscientist Jos Delgado
recognized some forty-five years ago, we should accept the fact that the exist-
ence of man, together with all of his attributes and creationsis actually and
inescapably the result of natural fate. Man did not invent man (Delgado
1969: 6). Consequently, Laland and Hoppitt have proposed a broad defini-
tion of culture as those group-typical behaviour patterns shared by members
of a community that rely on socially learned and transmitted information
(Laland & Hoppitt 2003: 151) a definition of culture shared by a number
of evolutionary anthropologists (Janson & Smith 2003: 57; Mesoudi et al.
2006). This broad, if minimalist, definition recognizes the presence of cul-
tural traits among many hundreds of species of vertebrates (Laland & Hoppitt
2003: 151; Mesoudi et al. 2006). Jared Diamond insists that we must ask
what were [the] precursors among these species [f ]or each of [the
presumed] defining cultural traits [of Homo sapiens] (Diamond 1991: 123).
Although Alex Mesoudi, Andrew Whiten and Laland have cautioned that it
must be determined whether the evolution of such cultural precursors was, in
fact, homologous, that is, an inheritance from a common ancestor (or ances-
tors), or analogous, that is, independently evolved (Mesoudi et al. 2006), Eric
Kandel, among others, has argued that evolution tends to conserve and to
exploit existing mechanisms a strategy necessary for survival (Kandel 2006:
186). At the molecular level, new, specialized molecules to produce a new
adaptive mechanism are not required (ibid.: 234, xiii); at the cellular level,
few if any unique neuronal types [in the human brain], and few if any genes
lack a significant ancestral precedent (Marcus 2006: 1118).
Because of ethical constraints upon the experimental manipulation of
Homo sapiens, a number of evolutionary anthropologists have argued that
experimentally tractable evidence for human culture is lacking, as it is, for
the same reason, limited with respect to our nearest primate relatives. Rather,
evidence for culture among Homo sapiens and primates, they argue, remains
largely anecdotal, based upon observations of populations in situ (but see
Wrangham et al. 1994; Hauser 2000). Rather, these anthropologists argue
that hard scientific evidence for culture exists only among certain species of

the origins of religion, cognition and culture

fish and birds (Janson & Smith 2003: 57; Laland & Hoppitt 2003: 15152;
Mesoudi et al. 2006). With reference to their proposal for a unified science
of cultural evolution, Mesoudi, Whiten and Laland have argued that science
has always proceeded by such simplifying assumptions and by using what
may be comparatively crude but workable methods, in order to make com-
plex systems tractable (Mesoudi et al. 2006). This judgement concerning the
simplifying venturesomeness of scientific ways of knowing characterizes the
cognitive sciences as well.
The study of cognition has been, and continues to be especially among
traditional philosophers of mind the consideration of the (largely) conscious
and reflective ways in which humans think; it is concerned, therefore, with
such issues as aims, assumptions, beliefs, desires and intentions, as well as
with how external aids, such as writing, archives and computers, supplement
these processes. For contemporary cognitive scientists, on the other hand,
cognition generally refers to the evolved, (largely) non-conscious capacities
of and constraints upon the ways by which the brain processes perceptual
input and to reflexively produce just the kinds of representational outputs
that it does. Similar to the conflation of cultural aspects, references to these
reflective and reflexive processes are often used interchangeably, resulting in
theoretical confusions between, for example, phylogenetic and ontogenetic
issues (Donald 2001: 208). It is with the evolutionary framework and basis
of cognition that we are concerned here. For as Nicholas Humphrey observes,
many traditional philosophers may be right in their conclusions but if they
dont look into the evolutionary background of their problems, they dont
know why (Humphrey 2006: 129; see Lakoff & John 1999).
Situating the problems of cognition and culture, and their relationship to
each other, in evolutionary time allows us to focus upon simplified models for
the evolutionary origins of culture and possible extrapolations from them rel-
evant to the relationship between cognition and culture among Homo sapiens
(Schaffner 2001). So, slipping past the usual subject for the investigation of
non-human culture Nietzsches monkey at the door we might nevertheless
avoid the pit of divine origins by beginning our discussion of cognition and
culture with their relationship among birds, a rather more simple species of
vertebrates that are nevertheless in closer evolutionary proximity to mammals,
including to Homo sapiens, than are fish (Macphail 1982).

Avian cognition and culture

Birds, it turns out, are not only highly intelligent (e.g. Pepperberg 2002),
but they exhibit many cognitive capacities once claimed as unique to Homo
sapiens. The most recent example of such a cognitive capacity once claimed

luther h. martin

as uniquely human is the learned ability of common starlings (Sturnus vul-

garis) to accurately recognize acoustic patterns defined by a recursive, self-
embedding, context-free grammar and [to] reliably exclude agrammatical
patterns (Gentner et al. 2006: 1204; see also similar findings of syntactic
generation of semantic calls among putty-nosed monkeys, Cercopithecus nic-
titans; Arnold & Zuberbhler 2006). Evolutionary, developmental and cog-
nitive scientists have found considerable advantages in working with birds
because behavioural and cognitive paradigms such as learning can be better
controlled in them than in mammals (Stewart 1991: 324). Further, accord-
ing to avian biologists, an understanding of the avian forebrain is now suf-
ficiently advanced (Dubbeldam 1991) to allow profitable comparison with
mammals (Andrew 1991b: 61; see also Hinde 1961). Although avian brains
are constructed very differently from humans, they nevertheless exhibit func-
tional similarities (Scholtyssek 2006: 53) that makes them complementary as
experimental material (Andrew 1991b: 61). Because domestic chicks possess
the simplest vertebrate brain to show cognitive abilities of interest (Johnson
1997: 103), they have come to be considered the avian equivalent of labora-
tory rats for developmental and cognitive studies (ibid.: 102; Andrew 1991a).

The domestic chick (Gallus gallus domesticus)

At an early stage in their lives, domestic chicks are capable of highly organized
capacities for preferences and motor patterns, from the complex coordination
of neck and leg movements required for the process of hatching to pecking
for food shortly thereafter (Bateson 1991: 113). In addition to an array of
such innate proclivities and predispositions, they also have considerable
capacities for learning, imprinting being the most well known and studied.
In addition to imprinting, however, chicks also quickly learn which foods are
palatable or unpalatable, which patterns of behaviour bring them to social
companions or to sources of heat, and how to use events to predict the occur-
rence of further events that may have significance for them (ibid.). As the
psychologist Euan Macphail concludes in his comprehensive treatment titled
Brain and Intelligence in Vertebrates, there seems to be ample space [in the
forebrains of domestic chicks] for mechanisms devoted to intellectual opera-
tions (Macphail 1982: 190). In other words, the forebrain of the chick is
the site of learning as well as being the location of certain inbuilt, automatic
types of behaviour (Johnson 1997: 103; Macphail 1982: 3742). Because
these complex interactions of developmental predispositions and abilities
with the effects of learning are unusually well understood in the domestic
chick. (Andrew 1991b: 2), psychologists continue to turn to the chick to try
to resolve [their] theoretical issues (Andrew 1991a: 9).

the origins of religion, cognition and culture


In contrast to the simplicity of chicks, avian taxonomists conventionally rank

passerines, or songbirds, which constitute about 60 per cent of the some
9,000 to 10,000 species of birds, as the highest, meaning the most intel-
ligent and by far the most [evolutionarily] successful (Macphail 1982: 169,
171). The somewhat dubious criteria for such rankings sometime include
anatomical complexity, for example, the muscles of the syrinx the organ
used to produce bird song are most highly developed. As Macphail has
concluded, however, any correlation between such somatic criteria and intel-
ligence remains in doubt (ibid.: 169).
There is a large body of evidence that aspects of avian behaviour which
are counted as intelligence, the varied courtship songs and complex preda-
tory warning signals of passerines, for example, may be socially transmitted
(Laland & Hoppitt 2003: 154, with reference to the review of Freeberg
2000). Passerines, in fact, account for 113 out of the 123 species of birds
for which there is such evidence (Frith & Frith 2004: 141). When, for
example, members of one of these, the common starling (Sturnus vulgaris),
are introduced into a new environment, they will acquire the calls of
birds in that local alien environment (ibid.). While the ability for mimetic
performance itself is probably phylogenetic (ibid.), the vocal signalling
employed by different species of passerines seems to be cultural in the
sense of culture defined by Laland and Hoppitt (2003; for recent over-
views of research on birdsongs, see Marler & Slabbekoorn 2004; Zeigler
& Marler 2004).

Bowerbirds (Ptilonorhynchidae)

One species of passerines, the bowerbirds of Australia and New Guinea, have
conventionally been placed among those at the top of the passerine list. I
should like to consider these birds in more detail, not only because their brain
structure and organization are, as with all avians, experimentally complemen-
tary with that of mammals, nor because of their accomplished abilities of
vocal mimicry of other bird species but also of other environmental sounds,
including human-made ones (Frith & Frith 2004: 1414), but because their
complex and elaborately decorated bowers have been described by ornithol-
ogists, evolutionary biologists and anthropologists alike as the most elabo-
rate creation of any animal apart from Homo sapiens (Diamond 1986: 3042;
Miller 2000: 269).
Jared Diamond describes the first bower he encountered, while on an
excursion into the jungles of New Guinea:

luther h. martin

a beautifully woven, circular hut 8 feet in diameter and 4 feet

high, with a doorway large enough for a child to enter and sit
inside. In front of the hut was a lawn of green moss, clean of
debris except for hundreds of natural objects of various colours
that had obviously been placed there intentionally as decorations.
They consisted mainly of flowers and fruits and leaves, but also
some butterfly wings and fungi. Objects of similar colour were
grouped together, such as red fruits next to a group of red leaves.
The largest decorations were a tall pile of black fungi facing the
door, with another pile of orange fungi a few yards further from
the door. All blue objects were grouped inside the hut, red ones
outside, and yellow, purple, black, and a few green ones in other
locations. (Diamond 1991: 156)

Had he not already heard of bowers, he tells us, he would have mistaken
this first one he saw for something man-made, as did nineteenth-century
explorers in New Guinea (ibid.).
In addition to such elaborate if diminutive thatched huts, the bowers of
other species all built of sticks, grasses, plant stems (Frith & Frith 2004:
1) may take the form of tepees or perfectly circular bowls with a cen-
tral conical tower of sticks. Or, they make take the form of open-topped
boxes of four vertical stick walls, or of twin-walled avenues of upright sticks
(ibid.: 81).
The impressive bowers constructed by these small male birds (weighing
approximately 150200 g) serve as extraordinary courtship sites where they
engage in familiar avian vocalization, posturing and dance in order to attract
females. They may regularly attend to and maintain a specific site over con-
secutive seasons. On the other hand, a traditional site may be abandoned
for several seasons and subsequently acquired by another adult male (ibid.:
104). Females, by contrast, build their own functional nests for the incuba-
tion and brooding of nestlings completely unassisted by the males. During
mating season, however, they cruise the elaborate structures of their suitors,
often in groups, in order to assess their qualities and those of their proprietors
(Diamond 1986: 3046; 1991: 156; Frith & Frith 2004: 214). Males, conse-
quently, are very discriminating in decorating their bowers and typically sit on
adjacent perches in order to appraise their newly completed or subsequently
adjusted decorations (Frith & Frith 2004: 122).
The objects, man-made as well as natural, that make up the decorations of
a single structure may exceed several thousand (ibid.: 1). They are not, how-
ever, randomly collected from among available objects in the environment
but are carefully selected by colour and are distinctively placed, by different
populations and individuals, on the basis of what Diamond considers to be

the origins of religion, cognition and culture

learned sets of rules (Diamond 1986: 3044; 1991: 157). These decorating
rules, Diamond emphasizes, are not automatic or instinctive but involve
decision-making and changes of mind (Diamond 1986: 3045).
In addition, the males of some species paint their bowers with a mixture
of pigment made from fruit pericarp, leaf matter or charcoal which are mas-
ticated with saliva (Frith & Frith 2004: 125). They apply this paint with a
wad of vegetable matter held in their bills (ibid.: 1) distinguishing them as
one of the few animals apart from Homo sapiens that practice tool use (ibid.:
12, 126).
The architectural and decorative achievements of bowerbirds may be com-
pared to the kind of socially transmitted behaviour expressed in their vocal
displays. While both forms of behaviour have a phylogenetic origin in the
case of bower building, an ancestral nest building ability, an abundance of
food resources that emancipates males from nest duties and permits promis-
cuity (ibid.: 141, 153, 2048; Freeberg 2000), both forms endure as acquired
and transmitted traits (Frith & Frith 2004: 141; Madden 2001; Diamond
1987, 1991).

Bowerbird culture
Bowerbirds are a rather intricately structured social species in which there is
increasing competition between young males with those of higher status while
nevertheless maintaining a pedagogical relationship with them. Adolescent
bowerbird males must successfully enter this competitive and established
society of bower owning peers if they are to have any significant opportu-
nity of reproducing (Frith & Frith 2004: 128). They visit and observe the
constructions and decorations of accomplished mature males from which
they acquire their own skills that they hone through experience and practice
(ibid.: 1267, 128, 212; Madden 2001). Whereas male bowerbirds raised in
relative isolation are instinctively capable of and are driven to accumulating
and to playing with sticks and decorations, they are not able to build and
decorate a bower with any skill and finesse. Similarly, immature males in the
wild were found to bring inappropriately coloured objects and thick sticks to
their first bower sites. Adolescent males are able to learn the finer points of
bower construction and decoration from their elders without evoking their
aggression because they exhibit dichromatic, female-like plumage during their
first six to seven of their 20 plus year lives (Frith & Frith 2004: 100, 1267,
128, 212; Diamond 1986: 3046; Madden 2001: 833). It is this powerful
effect of bowerbird behaviour as of birds generally on conspecifics that
has most surprised researchers (Andrew 1991b: 10). But is this behaviour
really culture?
Laland and Hoppitt have proposed two straightforward experimental
manipulations to demonstrate that a species exhibits culture. First:

luther h. martin

a sample of individuals from population A are introduced into

population B, and vice versa The observation that the intro-
duced animals adopt the behavior exhibited by members of the
host population [would be] inconsistent with an explanation
in terms of genetic differences between populations and consistent
with an explanation reliant on some form of learning.
(Laland & Hoppitt 2003: 152)

In a second experimental manipulation proposed by Laland and Hoppitt to

demonstrate that a species exhibits culture:

population A is collectively removed from its environment and

replaced in the environment of population B, which in turn is
removed and replaced in the environment of population A. If
the introduced populations come to exhibit the same behavior as
the former residents, this would suggest that the behavior results
from shaping to divergent ecological conditions, and asocial learn-
ing could not be ruled out. However, the observation that the
introduced population exhibits group-typical behavior that differs
significantly from that exhibited by the former residents would be
inconsistent with an explanation based on ecological differences
between environment and consistent with culture. (Ibid.)

The presence of culture among bowerbirds is demonstrated from experiment

B with the simple observation that the different species of bowerbirds, with
their differing styles of production, share the same or very similar environ-
ments (Diamond 1986: 3046). More interesting is the demonstration which
can be elucidated for experiment A in which, a lone vagrant male Spotted
[bowerbird] built a bower within the range of the satin bowerbird) well
beyond his own species range and decorated it with blue items as Satins
typically do but Spotteds do not (Frith & Frith 2004: 117; Neville 1988).
While both of these conditions have been observed in the wild among bower-
birds, neither have been experimentally and systematically manipulated.
Nevertheless, such geographically varying bower styles may [well] be a cul-
turally transmitted trait that is, as Diamond concludes, like human art
styles (Diamond 1986: 3042; Frith & Frith 2004: 124).

Aesthetics and symbolism

Because of the extent of their architectural and decorative culture, aesthetic
sensibilities have been attributed to bowerbirds (Frith & Frith 2004: 1, 13;
Dennett 2006: 397, n. 1). Although the behaviour surrounding bowers is

the origins of religion, cognition and culture

done primarily to attract females for mating, it has often also been observed
in the absence of females. Some observers have suggested, therefore, that
males may enjoy such activities for their own sake (ibid.: 1), that is to say,
they find it aesthetically pleasing (ibid.: 208, 222; Gilbert 1939). For exam-
ple, decorations displaced by researches or those added but disliked by the
birds are immediately removed (Diamond 1991: 157; Frith & Frith 2004:
122). This taste for the beautiful among bowerbirds was already noted by
Darwin ([1871] 2004: 465).
Further, some scientists even speak of bowerbirds exhibiting symbolic traits
(Frith & Frith 2004: 81). Even as there is an inverse relationship between
the colourfulness of a birds plumage and the beauty of its song, so there is
an inverse relationship between ornate plumage and the more elaborate and
decorated bowers (Ridley 1993: 163). Decorated bowers represent, in other
words, secondary sexual characteristics of male bowerbirds that have been
transferred external to the males body (Gilliard 1969: 47, 55). In the judge-
ment of some, this externalization of symbolic traits is a cultural characteristic
shared only with Homo sapiens (Frith & Frith 2004: 81; Diamond 1986). If
accepted, this judgement problematizes the claim that symbolic representa-
tion is a unique trait of human culture (e.g. Donald 1991; Deacon 1997).

Bowerbird cognition and culture

Bowerbirds, specifically, have been recognized for their intelligence (Frith &
Frith 2004: 1, 1113). As Diamond notes, the bowerbird, like any tropical-
forest bird, must quickly master an enormous amount of information con-
cerning local flora or insect fauna if it is to survive (Diamond 1987: 17). They
also require a well-developed spatial memory in order to remember the loca-
tions of supplies for decorations or of neighbouring bowers (Madden 2001:
833). While the bowerbird brain receives the inputs necessary to allow sen-
sory integration, it does not, however, seem to be totally dominated by [that]
sensory input (Macphail 1982: 190). Rather, its brain exhibits the neural
plasticity the ability of nerve cells [in human and non-human brains alike]
to change the strength and even the number of synapses (Kandel 2006: 218)
that is required for learning (Johnson 1997: 103; Macphail 1982: 3742).
Since bower building [and maintenance] is a cognitively complex task
involving elaboration of a variety of brain regions, the Cambridge zoologist
Joah Madden has raised the question of whether total brain volume [and,
thus, cognitive potential] may vary with the complexity of the bower that the
bearer builds (Madden 2001: 833). In fact, bowerbirds do have relatively
larger brains than ecologically similar songbirds of their region, and bower
building species have larger brains than non-bower-building ones (Frith &
Frith 2004: 1, 12, 200; Madden 2001: 836). Madden concluded that this dif-
ference in gross brain size may reflect the range of cognitive process necessary

luther h. martin

for successful bower building (ibid.: 833). On the other hand, Macphail
has concluded that any assumption that correlates brain size with greater
cognitive sophistication is still disputed (Macphail 1982: 169; see also ibid.:

The bowerbird syndrome

I have attempted to show that bowerbirds maintain culture in Laland and

Hoppitts sense of group-typical behavior patterns shared by members of
a community that rely on socially learned and transmitted information
(Laland & Hoppitt 2003: 151). And I have attempted to show that bower-
bird culture is not characterized simply by minimal patterns of reproduction
behaviour nor by the intricate patterns of their birdsong though these are a
central aspect of virtually all avian culture but, additionally, by a complex
and rich architectural and decorative tradition that has been compared with
the most elaborate creation of any animal apart from Homo sapiens. The bow-
ers created by bowerbirds not only employ tool use but seem to be aesthetic
projects which so far exceeds the requirements of reproductive strategy that
some commentators have attributed a human-like symbolic quality to them.
These acquired and transmitted traits seem to have a phylogenetic basis in
the evolutionary history of the bowerbird; they are, in other words, produced
on the basis of, and as an exploitation of, ordinary cognitive capacities of the
bowerbird brain. There is no evidence, however, that the cognitive constraints
characteristic of avian forebrains have been violated by the production and
transmission of bowerbird culture.
Clifford and Dawn Frith concluded their recent comprehensive study of
bowerbirds with the observation that the

Debate over the origins, significance, and function of [the] elab-

orate traits [of bowerbirds] has barely begun. Numerous
hypotheses will undoubtedly attempt to further explain the won-
derful examples of [the] frozen behaviour that bowers represent.
Given the variation and complexity of bowers, and associated
decoration and male behaviour, it is unlikely that any single
explanation will satisfactorily account for their origin(s) and/or
function(s). (Frith & Frith 2004: 222)

The Friths conclusions concerning the origins, significance and function of

bowerbird culture and its relationship to cognition, well describe the notion,
familiar from diagnostic medicine, of a syndrome, a group of observed and
seemingly correlated characteristics usually forming an identifiable pattern but

the origins of religion, cognition and culture

without reference to the mechanisms of causality, which, though assumed, are

not as yet fully known. Consequently, we might well term observations about
the contingent but causally underdetermined relationship between culture
and cognition of bowerbirds, the bowerbird syndrome.
Observations, correlations and questions similar to those that define the
bowerbird syndrome define also the diagnostic task for researchers into the
culture of Homo sapiens. So, slipping again past Nietzsches monkey at the
door, we can return from our discussion of the beginnings of cognition and
culture among bowerbirds to a brief consideration of their relationship in
Homo sapiens.

Cognition and culture among humans (Homo sapiens sapiens)

Robert Hinde has concluded that the dispositions and predispositions that
profoundly affect the kinds of information that birds acquire about the world
and that constrain their learning, subtle though they may be, are unlikely to
be restricted to birds (Horn 1991: 248, with reference to Hinde & Stevenson-
Hinde 1973). Might we then learn something from our avian contemporaries
concerning relationships that might exist between the evolved predisposi-
tions and dispositions and that affect the kinds of information which humans
acquire about their world and that constrain their learning? Has the latter, i.e.
the massive cumulative and extensively more complex character of acquired
human culture, influenced cognition? And if so, how and to what extent?

Human culture

Although Laland and Hoppitt have argued that hard scientific evidence for
culture is experimentally tractable only among fish and birds while only anec-
dotally attested for primates and Homo sapiens, I think it nevertheless rea-
sonable to presume as do they that humans do possess culture and an
exceedingly complex and multi-faceted one as well. And, we might further
presume that there is for Homo sapiens, as for birds and fish, phylogenetic dis-
positions for the production and transmission of those distributed behaviours
and ideas collectively referred to as culture. Since specific human cultures
have emerged independently of one another, however, it would seem that the
actuality of these cultures is an ontogenetic realization (i.e. a social exploita-
tion of those phylogenetic dispositions). Being anecdotal (i.e. ethnographi-
cally and historically attested), studies of the various cultures have tended to
reflect the self-interests of its students no less than those of it producers. These
interests are well illustrated by Geoffrey Millers lovely imaginative recounting
of a male bowerbirds explanation for his own cultural constructions:

luther h. martin

I find this implacable urge for self-expression, for playing with

colour and form for their own sake, quite inexplicable. I can-
not remember when I first developed this raging thirst to present
richly saturated color-fields with a monumental yet minimalist
stage-set, but I felt connected to something beyond myself when
I indulge these passions. When I see a beautiful orchid high in a
tree, I simply must have it for my own. When I see a single shell
out of place in my creation, I must put it right. Birds-of-paradise
may grow lovely feathers, but there is no aesthetic mind at work
there, only a bodys brute instinct. It is a happy coincidence that
females sometimes come to my gallery openings and appreciate
my work, but it would be an insult to suggest that I create in
order to procreate. We live in a post-Freudian, post-modernist era
in which crude sexual meta-narratives are no longer credible as
explanations of our artistic impulses. (Miller 2000: 26970)

As Rgis Debray has emphasized, however, such self-interested cultural ide-

ologies cannot be analysed in ideological [i.e. top-down] terms. Rather,
the hidden dynamic of the action of ideas in history is to seek their mate-
rial forms and sequences of transmission (Debray 1999: 2; also Sperber
1996). If we accept Laland and Hoppitts broader definition of culture,
which recognizes the presence of cultural traits not only among fish and
birds but among many hundreds of species of vertebrates as well (Laland &
Hoppitt 2003: 34), then human culture must also be explained in terms of
the material bases that characterizes its own evolutionary history. We must,
in other words, offer a bottom-up evolutionary history of those human
brain functions that might explain each aspect of that syndrome we call
human culture.

Brains, culture and minds

Mind is a designation often employed to signify the functionally discrete

consequence of interactions between brains and particular cultural environ-
ments. This concern seems to reflect, however, vestiges of Western philo-
sophical dualism, if not metaphysics (Bechtel & McCauley 1999), as well
as the lingering concern by a number of anthropologists sharply to differ-
entiate Homo sapiens from other vertebrates. The question of mind is not
raised, for example, of bowerbirds, nor does it seem relevant to a study of the
relation between their brains and their culture. As Delgado already recog-
nized, mind is an ill-defined group of mental activities that can only be
described and researched in terms of the evolved capacities and constraints of

the origins of religion, cognition and culture

brains (Delgado 1969: 31). Whereas brains are material entities the functions
of which may be explained, minds are not subject to scientific investigation.
It makes sense, therefore, to base a material explanation for the relationship
of cognition and culture with brains (Macphail 1982: 25).

Cultural influence upon brain size

Bowerbirds, as we have seen, have slightly larger brains than some ecologically
similar non-bower building passerines of comparable size (Madden 2001).
Whether culture has contributed to the evolutionary increase of a speciess
brain size and complexity or whether an increase in brain size and complexity
in response to some non-cultural adaptive pressure, diet for example is the
basis for expanded cultural production remains an open question (LeDoux
2002: 72). There is some evidence, however, that human brain size may have
increased as a consequence of rapid and complex cultural developments, even
since the emergence of anatomically modern humans.
It was recently reported that two genes thought to regulate human brain
growth have continued to evolve under natural selection over historical time,
and perhaps are continuing to do so today (Balter 2005, reporting on Evans
et al. 2005 and Mekel-Bobrov et al. 2005). One of these alleles, microcephalin,
is calculated to have appeared around 37,000 years ago with a confidence
level ranging from 14,000 to 60,000 years at about the same time as the
explosion of symbolic behaviour in Europe, and a second, ASPM (or abnor-
mal spindle-like microcophaly-associated), arose some 5800 years ago with a
possible range of 500 to 14,100 years just before cities arose in the Near
East (Balter 2005: 1662). Reminiscent of Macphails scepticism concerning
the correlation of brain size with cultural abilities among birds, however,
Chris Tyler-Smith of the Sanger Institute near Cambridge, England, argues
that possible genetic links to such historical events are highly speculative
(cited by Balter; ibid.: 1662). While somatotopic areas of the brain relevant
to certain skills or behaviours can become enlarged with usage or training
(Kandel 2006: 21618, 306), cognitive potential depends, it would seem,
not upon gross brain size but upon the excess of neural circuitry beyond that
required to maintain physical competence (Dunbar 2004: 68). Consider, for
example, the plight of the Neanderthals whose brain size exceeded that of
anatomically modern humans by more than 400 cm3. As R. E. Passingham
concluded some twenty-five years ago, the modern human brain has exactly
the proportions and structures that might be predicted of a very large pri-
mate brain by extrapolating earlier primate expansions (Donald 1998: 10,
citing Passingham 1982). And Mark Stoneking of the Max Planck Institute
for Evolutionary Anthropology in Leipzig concludes that there is absolutely

luther h. martin

nothing in [recent findings] to relate the signature of selection to any

brain-related phenotype (cited by Balter 2005: 1662).

The influence of culture upon brain function

If not brain size, the remarkable plasticity of vertebrate brains (e.g. Edelman
1992) has given rise to considerations of influences from its environment
upon the brain. However, extensive plasticity, itself an innately determined
characteristic (LeDoux 2002: 9), is predominately a feature of the brain
in utero and in early childhood (ibid.: 96). Plasticity, in other words, refers
primarily to ontogenetic learning and memory (ibid.: 28), to idiosyncratic
differences among members of a species rather than to their commonality
(ibid.: 29).
Recently, a team of population geneticists has suggested, on the other
hand, that some human genes may have continued to evolve in selective
response to changing pressures of the physical environment (e.g. climate,
available food resources, even population density) until the present day. The
candidate alleles which have not reached fixation are those influencing
metabolism, skin pigmentation, reproduction and, perhaps, even brain func-
tions (Voight et al. 2006). It is not yet known whether or not and how the
altered genes that influence brain function actually affect behavioural or psy-
chological traits, nor has it been determined whether or not their adapta-
tion to changing environmental pressures might include analogous responses
to cultural niches.1 Before venturing generalizations based upon such, still
untested, hypotheses, however, the overwhelming numbers of examples where
the human genome did not respond selectively to adverse environmental con-
ditions, whether natural or created, to infectious diseases, for example, or to
recurrent cognitive malfunctions must be explained. As Benjamin Voight et
al. have concluded, gaining a full view of the kinds of selective pressures that
have faced modern humans, and our biological adaptations to these pressures,
remains a challenging problem (Voight et al. 2006: 454; also ibid.: 446).
Further, individual brains must still process input from their environment
however complex and extensive in ways constrained by their evolved bio-
logical morphology and functions (LeDoux 2002: 889). The increasingly
complex and invasively hot character of our contemporary visual environ-
ment (McLuhan & Fiore 1967) has not, after all, altered our evolved physi-
ology of vision. We must distinguish, in other words, between phenotypic
capacities for the processing of sensory input and for developmental poten-
tials, on the one hand, and upon neural plasticity as the ontogenetic basis
for learning and development, on the other. And the potential for both is
specified at the level of genetic expression (Kandel 2006: 202). Even Gerald

the origins of religion, cognition and culture

Edelman, the foremost advocate of neural plasticity, argues that a morphology

of the brain associated with constraining cognitive functions is sufficiently
similar among members of Homo sapiens to be identified and described
(Edelman 1992: 25; Edelman & Tononi 2000: 478).
To paraphrase Reinhold Niebuhrs insight with respect to history, humans
are both the creatures as well as the creators of culture (Niebuhr 1955: 413).
It would seem, nevertheless, that creative but constrained cognitive disposi-
tions associated with cultural production would take explanatory precedence
over any consideration of the possible effects of that culture on cognition
(Donald 1998: 12). Like the decorated bowers first stumbled upon by Jared
Diamond in the jungles of Papua New Guinea, cultural artefacts would them-
selves seem to have little, if any, explanatory efficacy unless and until their
production and their functions are first explained.

Biological and non-biological cognition

The question remains of the relationship between evolved capacities of and

constraints upon brains and what has been termed non-biological cognition.
Whereas biological cognition, whether reflexive or reflective (Lieberman et
al. 2002), is extremely proficient at, for example, pattern recognition, non-
biological cognition excels at information storage, retrieval and transmission.
But do external mechanisms for the manipulation of information influence
biological cognition in any qualitative way? It is, for instance, a relatively
simple and inexpensive operation to upgrade computer memory. To do so
indeed allows the computer to store a greater amount of information and to
process this information in a significantly shorter period of time. However,
this increase in performance capacity in no way alters the architecture of its
central processor. Similarly, the classicist Jocelyn Small has argued that when
the ancient Greeks were confronted with the task of managing and retrieving
data from the massively expanding archives which resulted from the introduc-
tion of literacy they continued to rely on [and to improve the efficiency of ]
their well-known and proven [pre-literary] techniques of [natural] memory
a concern with mnemotechnics that endured until the Renaissance and
beyond (Small 1997: 81, 87, 136, 243). The increased demands of managing
archival information seems, too, to have in no way altered the cognitive ways
by which the Greeks processed information despite the availability of the
newly formalized Greek alphabet to do so (ibid.; on the protracted realiza-
tion of cognitive possibilities resulting from the introduction of literacy, see
Donald 2001: 306). Since technologies for the external management, trans-
mission and retrieval of information traits usually subsumed under the cat-
egory culture (Debray 1999: 2) are produced by human minds, Mesoudi,

luther h. martin

Whiten, and Laland have concluded that they are probably adapted from the
biologically evolved or developmentally acquired cognitive features of those
minds (Mesoudi et al. 2006). Even Merlin Donald, well known for his argu-
ments in support of cultural influence upon human cognition, succinctly
emphasizes that culture reflects the innate cognitive features of the individu-
als that make it up (Donald 1998: 12), however few these features may be
(Donald 2001: 224). Such non-biological technologies of culture as symbolic
inventories or external memory systems could not, he concludes, trigger
new innate mental capacities (Donald 1998: 15). External aids to cognition
seem to have developed, in other words, as exploitations and extensions of
our natural cognitive (and motor) capacities and they remain constrained by
the limits of these evolved capacities. Even as some species of migratory pas-
serines now face extinction because they continue to follow their phenotypic
schedule of migration and arrive, consequently, at their breeding sites follow-
ing the time of maximum food abundance that has been altered by climate
change (Both et al. 2006), and enculturated apes continue to use [their
newly acquired] symbols for traditional primate agendas (Donald 2001:
204), the culture of humans developed, it would seem, according to the phe-
notypical agendas of evolved brains rather than as Frankensteinian automata
that might in some way turn upon and finally seduce the workings of those
evolved processes.

(Tentative) conclusions

I have attempted to argue that Homo sapiens sapiens share with bowerbirds
and with all vertebrates an evolved disposition for the production of those
behaviours we designate collectively as culture. There are, however, many lev-
els of analysis to be differentiated in the attempt to understand and to explain
cultural syndromes, among them (and from the bottom up), the genetic, the
cellular, the neurophysiological, the cognitive, the social, and while explana-
tions at the lower levels are understood to constrain those for the next high-
est respectively, there is, nevertheless, not yet a very clear understanding of
how these levels are related (Hogan 2003: 2023). And there are two funda-
mentally different temporal frames within which to evaluate the beginnings
and the functions of each of these levels of analysis and their relationships
the evolutionary (phylogenetic), in which information is genetically encoded
and transmitted and the historical (ontogenetic) in which group-typical
behavior patterns shared by members of a community are socially learned
and transmitted (Laland & Hoppitt 2003: 151). We might ask ourselves, in
other words, exactly what it is that we want to know. In this pursuit, Robin
Dunbar has noted a common confusion between function and ontogeny

the origins of religion, cognition and culture

the evolutionary goal of an animal, which is always genetic fitness, and the
reason why an animal behaves in the way that it does, which is always some
combination of its genetic inheritance, environmental effects and experiential
learning, including, in humans, cultural transmission (Dunbar 2004: 105).
As the ethologist Paul Bateson suggested with reference to the behaviour
of chicks, our questions must be refined so that we focus on each aspect or
sub-system of our culture, on its development, and on the connections that
form between them (Bateson 1991: 130). We need to ascertain, as best we
can, the precursors of each of the cultural traits in which we are interested
perhaps as an exercise in reverse engineering? When in our ancestry did
these traits approach their modern form and can the early stages of their evo-
lution be traced archaeologically (Diamond 1991: 123)? We need, further,
to define and to differentiate more precisely the concepts that we frequently
employ in our investigations, our notions of cognition, for example, and
of development, intelligence and learning. As well, we need to differentiate
between biological cognition and non-biological information management;
between aspects of cognition and culture that are subject to scientific expla-
nation and those that remain the object of description; and between the uni-
versal (phylogenetic) disposition for and the contingent (ontogenetic) aspects
of culture. And, we need to avoid ideological biases in horizontal (compara-
tive) and in top-down cultural studies while distinguishing such interesting
indeed significant considerations from explanatory, i.e. bottom-up studies.
If such pursuits are to be scientific, we should presume a conceptual integra-
tion (Cosmides et al. 1992) and theoretical consilience (E. O. Wilson 1998)
among the humanities and the social sciences and with the natural sciences,
even though a paradigm for such conceptual integration is not yet fully estab-
lished. And whatever the influence of the cultural upon the cognitive might
prove finally to be, we need to bear in mind that descriptive, non-explanatory
attention to the particularizing character of such influences tends inexorably
to divert us into post-modernist cul-de-sacs of incomparability while loosing
sight of the question, what, exactly, is human about that syndrome we term
human (Ridley 1993: 121).

Postscript: The origin of religion

One cultural trait that virtually all researchers agree is unique to human cogni-
tion and culture is those behaviours and ideas that we consider under the taxon
religion (e.g. Bering 2001; Dunbar 2004: 197; Dennett 2006: 6; pace Guthrie
2002). As with culture, I will also stipulate a minimalist that is, a necessary
if not sufficient definition for religion (at least as an initial working defini-
tion) as referring to socially transmitted sets of ordinary human behaviours

luther h. martin

or ideologies which are legitimated by claims to the authority of superhuman

agents and religious, consequently to refer to any particular behaviour or
ideology so legitimated. (This definition begs the question of selection and
motivation whereby some but not all behaviours and ideologies so legitimated
are transmitted, but this is another issue). As with the evolutionary origins
of bower building, those for religious claims remain obscure though as we
noted at the beginning of this chapter religion, unlike language, for example,
is generally considered by cognitive scientists to be a by-product of adaptive
cognitive capacities. For this reason, I offer a brief consideration of religion as
a postscript to our thoughts on the origins of cognition and culture.
The persistence of non-falsifiable claims to superhuman agency among
humans seems to have little to do with adaptive strategies of survival quite
the contrary and, while widespread, neither religious behaviours nor ideas
are, in fact, phenotypic. In fact, those who identify themselves as non-
religious or as atheistic make up the third largest world population after
Christians and Muslims (Pearson Education 2005: 5). Those who wish to
argue to the contrary (e.g. that human religiosity may flow from dedicated
cognitive structures configured to form precise non-natural understandings
that motivate intricate adaptive responses to ancestral conditions, in the
words of Bulbulia 2005: 96) must then explain the rather significant presence
and perseverance of such non-religious people. Religious behaviour and ideas
do, however, seem to be a universally learned and transmitted characteristic
of human societies, that is, of bounded sets of relationships between select
individuals. Those who wish to argue that religious societies are themselves
adaptive groups that have evolved over time in ways analogous to biological
species (e.g. D. S. Wilson 2002) must offer a more convincing argument as
to the internal stability of such groups as well as for a relative impermeability
for their boundaries (for an extended critique of group selection with refer-
ence to Wilson, see Atran 2002: 20710, 21619, 22734). Whatever the
origin(s) of religion, it has, however, most certainly been exploited by human
societies in service of their own self-interests but that, like the question of
the social motivation and transmission of religion, is another topic. Since,
however, social rather than environmental features seem to be the main fac-
tors motivating religious behaviour and perseverance, we might suggest that
the origins of religion and their connection to cognition and culture may
well have an explanation in the similar features which motivate sexual selec-
tion (Borgia 1986: 92) as do the elaborate constructions of bowerbirds. For
Darwin, sexual selection refers to the advantage which certain individuals
have over others of the same sex and species solely in respect of reproduction
(Darwin [1871] 2004: 243).
Whatever the adaptive origins for the elaborate constructions of bowers by
males, their material and behavioural displays have become a staple example

the origins of religion, cognition and culture

in support of sexual selection theory by which one sex [is] modified in

relation to the other (e.g. Borgia 1986; Diamond 1991: 1569; Miller 2000:
26770; Frith & Frith 2004: 153, 20922, with reference to Darwin [1859]
2003: 956; [1871] 2004).
Recently, scholars such as Matt Ridley (1993) and Geoffrey Miller (2000)
have explored the implications of sexual choice for the origins of human
behaviour, including the religious (e.g. Miller 2000: 264, 4214). Madden
has suggested that the correlation between brain size in bowerbirds and the
complexity of their material constructions if significant provides some
indirect support for the hypothesis put forward by [Geoffrey] Miller (2000)
that sexual selection may [have driven] the evolution of the human
brain in response to female choice targeting novel, complex male behaviours
(Madden 2001: 837).
Miller concludes that [m]ost of our mental adaptations that patiently
guide our behavior remain intuitively accurate. The mental fireworks
show of courtship, however, can, he suggests, undermine our evolutionary
epistemology by turning our cognitive faculties into ornamental fitness-
advertisements rather than disciples of truth (Miller 2000: 4234). Millers
description of courtship behaviour subverting the intuitive categories of our
evolutionary epistemology resonates with Pascal Boyers characterization of
counterintuitive, or religious, concepts as violations of certain expectations
from [evolved] ontological categories (Boyer 2001: 624). Once introduced,
such violations, according to Boyer, are attention grabbing, highly memorable
and, therefore, readily transmitted. He does not, however, offer any motiva-
tion or explanation for why such categories might be violated in the first
place. Millers view about religious behaviours and ideas having their origins
in the posturings, dance, costumes2 and music of male display or in the coun-
terintuitive imaginings of innovative storytellers on the prowl, might, how-
ever, provide an intentional motivation for such epistemological violations.
To paraphrase the late Saul Bellow, all a man has to do to get a woman is to
say hes a storyteller. Its an aphrodisiac.3
While religious behaviours and ideas may have had their origins in male
display, they neednt, of course, have been transmitted among social groups
for this reason. For, in addition to being attention grabbing and memorable,
such behavioural displays and counterintuitive ideas are also highly entertain-
ing (Goode 2000: 235). The philosopher and musician, Nathan Rothenberg,
for example, has argued that while birds sing to claim their territories and
to attract mates precisely the reasons for which bowerbirds construct their
decorated courts they also sing simply because they love to. His evidence
for this conclusion is like that for culture generally anecdotal. When
invited to play his clarinet at the National Aviary in Pittsburgh, for exam-
ple, Rothenbergs jazz attracted a White-crested Laughing Thrush (Garrulax

luther h. martin

laucolophus) which began to sing along with his playing and actually to
improvise upon his own improvisations (Rothenberg 2005: 112). Thus, as
with the innovative ability of bowerbirds for employing novel decorations
or for producing abnormally elaborate bower designs (Madden 2001: 833),
there seems to be an aesthetic motivation for birds songs. Aesthetics as
expressed in ritual and musical performance and in the production of art (see
Dennett 2006: 1534) is precisely the reason given by a number of my
learned acquaintances for their continuing participation in Christian worship.
Although most of these acquaintances are Anglicans, the relationship between
engaging entertainment, aesthetics and religion has a long history, from the
Dionysian theatre of ancient Greece to the modern theological reflections of
a Friedrich Schleiermacher (e.g. Schleiermacher [1799] 1958: 139) to Max
Webers sociology of religion (Weber 1963: 2425). In this connection, it
is interesting to note that Darwin devoted but three out of the some 700
pages of The Descent of Man to a discussion of religion (Darwin [1871] 2004:
11619). While he mentioned neither aesthetics nor sexual choice in these
three pages, tellingly, they follow immediately upon his discussion of sexual
selection as the basis for a sense of beauty among birds, with bowerbirds
being elicited as one of his primary examples (ibid.: 11416).
Whether the origins of those practices among Homo sapiens that we associ-
ate with religion and the meta-representational claims by which those prac-
tices are legitimated may be accounted for by some adaptive feature of survival
or by sexual choice as it is in bowerbirds is still debated. If, however, there
is an evolutionary story to be told for the origins of religion, it would seem
that sexual selection should receive more consideration than it has.

I would like to thank Judith Grant, William Paden and Donald Wiebe for reading an earlier
draft of this paper making suggestions for its improvements. Any excesses in expression or
extravagances of argument remain, of course, my own.

1. See, however, Cochran et al. (2006), who argue that the unique demography and soci-
ology of the Ashkenazim in medieval Europe led to a social environment selected
for intelligence.
2. It is interesting to note that indigenous peoples of Papua New Guinea and Australia
compete with bowerbirds for the rather scarce feathers from birds of paradise for use in
their ritual activities (Frith & Frith 2004: 7, 120).
3. All a writer has to do to get a woman is to say hes a writer. Its an aphrodisiac (cited
in Time 166(26), 26 December 20052 January 2006, p. 163).

the origins of religion, cognition and culture


Andrew, R. J. 1991a. The Chick in Experiment: Techniques and Tests. In Neural and
Behavioural Plasticity: The Use of the Domestic Chick as a Model, R. J. Andrew (eds), 557.
Oxford: Oxford University Press.
Andrew, R. J. (ed.) 1991b. Neural and Behavioural Plasticity: The Use of the Domestic Chick as
a Model. Oxford: Oxford University Press.
Arnold, K. & K. Zuberbhler 2006. Semantic Combinations in Primate Calls. Nature 441
(8 May): 303.
Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. New York: Oxford
University Press.
Balter, M. 2005. Are Human Brains Still Evolving? Brain Genes Show Signs of Selection.
Science 309(5741): 16623.
Bateson, P. P. G. 1991. Making Sense of Behavioural Development in the Chick. In Neural
and Behavioural Plasticity: The Use of the Domestic Chick as a Model, R. J. Andrew (ed.),
11332. Oxford: Oxford University Press.
Bechtel, W. & R. N. McCauley 1999. Heuristic Identity Theory (or Back to the Future):
The Mind-Body Problem; Against the Background of Research Strategies in Cognitive
Neuroscience. In Proceedings of the Twenty-first Meeting of the Cognitive Science Society,
M.Hahn & S. C. Stones (eds), 6772. Mahwah, NJ: Lawrence Erlbaum Associates.
Bering, J. M. 2001. Theistic Percepts in Other Species: Can Chimpanzees Represent the
Minds of Non-Natural Agents? Journal of Cognition and Culture 1(2): 10737.
Borgia, G. 1986. Sexual Selection in Bowerbirds. Scientific American 254: 92100.
Both, C., S. Boushuis, C. M. Lessells & M. E. Visser 2006. Climate Change and Population
Declines in a Long-distance Migratory Bird. Nature 441 (4 May): 813.
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Bulbulia, J. 2005. Are There Any Religions? An Evolutionary Exploration. Method and Theory
in the Study of Religion 17(2): 71100.
Cochran, G., J. Hardy & H. Harpending 2006. Natural History of Ashkenazi Intelligence.
Journal of Biosocial Science 38(5): 65993.
Cosmides, L., J. Tooby & J. H. Barkow 1992. Introduction: Evolutionary Psychology and
Conceptual Integration. In The Adapted Mind: Evolutionary Psychology and the Generation
of Culture, J. H. Barkow, L. Cosmides & J. Tooby (eds), 315. New York: Oxford University
Darwin, C. [1859] 2003. The Origin of Species by Means of Natural Selection. New York:
Darwin, C. [1871] 2004. The Descent of Man, and Selection in Relation to Sex. New York:
Dawkins, R. 1989. The Selfish Gene, 2nd edn. New York: Oxford University Press.
Deacon, T. W. 1997. The Symbolic Species: The Co-evolution of Language and the Brain. New
York: W. W. Norton.
Debray, R. 1999. What Is Mediology?, M. Irvine (trans.). Le Monde Diplomatique (August):
32. www.georgetown.edu./faculty/irvinem/theory/Debray-What_is_ Mediology.html.
Delgado, J. M. R. 1969. Physical Control of the Mind: Toward a Psychocivilized Society. New
York: Harper & Row.
Dennett, D. C. 2006. Breaking the Spell: Religion as a Natural Phenomenon. New York: Viking.
Diamond, J. 1986. Animal Art: Variation in Bower Decorating Style among Male Bowerbirds
Amblyornis inornatus. Proceedings of the National Academy of Science USA 83(9): 30426.

luther h. martin

Diamond, J. 1987. Learned Specializations of Birds. Nature 330 (5 November): 1617.

Diamond, J. 1991. The Rise and Fall of the Third Chimpanzee. London: Radius.
Donald, M. 1991. Origins of the Modern Mind: Three Stages in the Evolution of Culture and
Cognition. Cambridge, MA: Harvard University Press.
Donald, M. 1998. Hominid Enculturation and Cognitive Evolution. In Cognition and
Material Culture: The Archaeology of Symbolic Storage, C. Renfrew & C. Scarre (eds), 717.
Cambridge: McDonald Institute for Archaeological Research.
Donald, M. 2001. A Mind So Rare: The Evolution of Human Consciousness. New York: W. W.
Dubbeldam, J. L. 1991. The Avian and Mammalian Forebrain: Correspondences and
Differences. In Neural and Behavioural Plasticity: The Use of the Domestic Chick as a Model,
R. J. Andrew (ed.), 6591. Oxford: Oxford University Press.
Dunbar, R. 2004. The Human Story: A New History of Mankinds Evolution. London: Faber
& Faber.
Durkheim, . [1915] 1965. The Elementary Forms of the Religious Life, J. W. Swain (trans.).
New York: Free Press.
Edelman, G. M. 1992. Bright Air, Brilliant Fire: On the Matter of the Mind. New York: Basic
Edelman, G. M. & G. Tononi 2000. A Universe of Consciousness: How Matter Becomes Imagin-
ation. New York: Basic Books.
Evans, P. D, S. L. Gilbert, N. Mekel-Bobrov, E. J. Vallender, L. M. Vaez-Azisi, S. A. Tishkoff,
R. R. Hudson & B. T. Lahn 2005. Microcephalin a Gene Regulating Brain Size, Continues
to Evolve Adaptively in Humans. Science 309(5741): 171720.
Foucault, M. 1977. Nietzsche, Genealogy, History. In Language, Counter-memory, Practice:
Selected Essays and Interviews by Michel Foucault, D. F. Bouchard (ed.), 13964. Ithaca, NY:
Cornell University Press.
Freeberg, T. M. 2000. Culture and Courtship in Vertebrates: A Review of Social Learning
and Transmission of Courtship Systems and Mating Patterns. Behavioural Processes 51:
Frith, C. B. & D. W. Frith 2004. The Bowerbirds Ptilonorhynchidae. Oxford: Oxford University
Gentner, T. Q., K. M. Fenn, D. Margoliash & H. C. Nusbaum 2006. Recursive Syntactic
Pattern Learning by Songbirds. Nature 440 (27 April): 12047.
Gilbert, P. A. 1939. The Bower-Painting Habit of the Satin Bower-bird. Emu 39: 1822.
Gilliard, E. T. 1969. Birds of Paradise and Bower Birds. London: Weidenfeld & Nicolson.
Goode, E. 2000. Paranormal Beliefs: A Sociological Introduction. Prospect Heights, IL: Waveland
Guthrie, S. 2002. Animal Animism: Evolutionary Roots of Religious Cognition. In Current
Approaches in the Cognitive Science of Religion, I. Pyysiinen & V. Anttonen (eds), 3867.
London: Continuum.
Hauser, M. D. 2000. Wild Minds: What Animals Really Think. New York: Henry Holt.
Hinde, R. A. 1961. The Establishment of Parent-Offspring Relations in Birds, with some
Mammalian Analogies. In Current Problems in Animal Behaviour, W. H. Thorpe & O. L.
Zangwill (eds), 17593. Cambridge: Cambridge University Press.
Hinde, R. A. & J. Stevenson-Hinde (eds) 1973. Constraints on Learning. London: Academic
Hogan, P. C. 2003. Cognitive Science, Literature, and the Arts: A Guide for Humanists. New
York: Routledge.
Horn, G. 1991. Imprinting and Recognition Memory: A Review of Neural Mechanisms. In

the origins of religion, cognition and culture

Neural and Behavioural Plasticity: The Use of the Domestic Chick as a Model, R. J. Andrew
(ed.), 21991. Oxford: Oxford University Press.
Humphrey, N. 2006. Seeing Red: A Study in Consciousness. Cambridge, MA: Harvard Uni-
versity Press.
Jamieson, D. & M. Bekoff 1996. On Aims and Methods of Cognitive Ethology. In Readings
in Animal Cognition, M. Bekoff & D. Jamieson (eds), 6578. Cambridge, MA: MIT Press.
Janson, C. H. & E. A. Smith 2003. The Evolution of Culture: New Perspectives and
Evidence. Evolutionary Anthropology 12: 5760.
Johnson, M. H. 1997. Developmental Cognitive Neuroscience. Oxford: Blackwell.
Kandel, E. R. 2006. In Search of Memory: The Emergence of a New Science of Mind. New York:
W. W. Norton.
Lakoff, G. & M. Johnson 1999. Philosophy in the Flesh: The Embodied Mind and Its Challenge
to Western Thought. New York: Basic Books.
Laland, K. N. & W. Hoppitt 2003. Do Animals Have Culture? Evolutionary Anthropology
12: 15059.
LeDoux, J. 2002. Synaptic Self: How Our Brains Become What We Are. New York: Viking
Lvi-Strauss, C. 1976. Structural Anthropology, vol. 2, M. Layton (trans.). New York: Basic Books.
Lieberman, M. D, R. Gaunt, D. T. Gilbert & Y. Trope 2002. Reflexion and Reflection: A
Social Cognitive Neuroscience Approach to Attributional Inference. Advances in Experi-
mental and Social Psychology 34: 199249.
Macphail, E. M. 1982. Brain and Intelligence in Vertebrates. Oxford: Clarendon Press.
Madden, J. R. 2001. Sex, Bowers and Brains. Proceedings of the Royal Society of London B
268: 8338.
Marcus, G. F. 2006. Startling Starlings. Nature 440 (27 April): 111718.
Marler, P. & H. Slabbekoorn 2004. Natures Music: The Science of Birdsong. London: Elsevier.
Masuzawa, T. 2000. Origin. In Guide to the Study of Religion, W. Braun & R. T. McCutcheon
(eds), 20924. London: Cassell.
McLuhan, M. & Q. Fiore 1967. The Medium is the Massage. New York: Bantam Books.
Mekel-Bobrov, N., S. L. Gilbert, P. D. Evans, E. J. Vallender, J. R. Anderson, R. R. Hudson,
S. A. Tiskoff & B. T. Lahn 2005. Ongoing Adaptive Evolution of ASPM, a Brain Size
Determinant in Homo sapiens. Science 309 (5741): 172022.
Mendel, G. [1865] 1965.Versuche ber Pflanzen-Hybriden. Brnn: Im Verlag des Vereines.
Mesoudi, A., A. Whiten & K. N. Laland 2006. Towards a Unified Science of Cultural
Evolution. Behavioral and Brain Sciences 29: 32983.
Miller, G. 2000. The Mating Mind: How Sexual Choice Shaped the Evolution of Human Nature.
New York: Random House.
Neville, B. 1988. The Strange Case of Billy the Bowerbird. Geo 10: 739.
Niebuhr, R. 1955. The Self and the Dramas of History. New York: Charles Scribners Sons.
Nietzsche, F. [1887] 1971. Morgenrthe: Gedanken die moralischen Vorurtheile. Berlin: Walter
de Gruyter (English-language edition: Daybreak: Thoughts on the Prejudices of Morality,
trans. R. J. Hollingdale; Cambridge: Cambridge University Press, 1997).
Passingham, R. E. 1982. The Human Primate. New York: Morrow.
Pearson Education 2005. Atlas of World Religions. Upper Saddle River, NJ: Pearson Prentice
Pepperberg, I. M. 2002. The Alex Studies: Cognitive and Communicative Abilities of Grey Parrots.
Cambridge, MA: Harvard University Press.
Plotkin, H. 2003. The Imagined World Made Real: Towards a Natural Science of Culture. New
Brunswick, NJ: Rutgers University Press.

luther h. martin

Poole, F. J. P. 1986. Metaphors and Maps: Towards Comparison in the Anthropology of

Religion. Journal of the American Academy of Religion 54: 41157.
Ridley, M. 1993. The Red Queen: Sex and the Evolution of Human Nature. New York:
Rothenberg, D. 2005. Why Birds Sing: A Journey through the Mystery of Bird Song. New York:
Basic Books.
Sad, E. W. 1985. Beginnings: Intention and Method. New York: Columbia University Press.
Schaffner, K. F. 2001. Extrapolation of Animal Models: Social Life, Sex, and Super Models.
In Theory and Method in the Neurosciences, P. K. Machamer, R. Grush & P. McLaughlin
(eds), 20030. Pittsburgh, PA: University of Pittsburgh Press.
Schleiermacher, F. [1799] 1958. On Religion: Speeches to Its Cultured Despisers, J. Oman
(trans.). New York: Harper & Brothers.
Scholtyssek, C. 2006. Bird Brains? Hardly. Scientific American Mind 17: 5055.
Small, J. P. 1997. Wax Tablets of the Mind: Cognitive Studies of Memory and Literacy in Classical
Antiquity. London: Routledge.
Sperber, D. 1996. Explaining Culture: A Naturalistic Approach. Oxford: Blackwell Publishers.
Stewart, M. G. 1991. Changes in Dendritic and Synaptic Structure in Chick Forebrain
Consequent on Passive Avoidance Learning. In Neural and Behavioural Plasticity: The Use
of the Domestic Chick as a Model, R. J. Andrew (ed.), 30528. Oxford: Oxford University
Voight, B. F., S. Kudaravalli, X. Wen & J. K. Pritchard 2006. A Map of Recent Positive
Selection in the Human Genome. PLoS Biology 4(3): 44658.
Weber, M. 1963. The Sociology of Religion, E. Fischoff (trans.). Boston, MA: Beacon Press.
Wells, H. G. 1895. The Time Machine: An Invention. London: Heinemann.
Wilson, D. S. 2002. Darwins Cathedral: Evolution, Religion, and the Nature of Society. Chicago,
IL: University of Chicago Press.
Wilson, E. O. 1998. Consilience: The Unity of Knowledge. New York: Alfred A. Knopf.
Wrangham, R, W. McGrew, F. de Waal & P. Heltne (eds) 1994. Chimpanzee Cultures.
Cambridge, MA: Harvard University Press.
Zeigler, P. & P. Marler 2004. Behavioral Neurobiology of Birdsong. New York: New York
Academy of Sciences.

The will to sacrifice: sharing and sociality
in humans, apes and monkeys
Henrik Hgh-Olesen

In our species the reciprocity principle is the basic ethic behind any society we
know: you have to give in order to get, and to reciprocate what you receive.
Such is the rule the expectation or code, if you like universally (Mauss
[1925] 1990; Gouldner 1960; Hgh-Olesen 2006). Or as George Simmel
(1950: 387) formulated it: All contracts among men rest on the schema of
giving and returning the equivalence.
This worlds key religious and moral texts are similarly full of stringent
requests for sharing and examples of radical self-sacrifices, and modern
anthropologys studies of huntergatherers rules of sharing worldwide basi-
cally confirm these ethics. Food and resources are shared on the basis of egali-
tarian and reciprocal principles (Kaplan & Hill 1985; Betzig & Turke 1986;
Testart 1987; Lee 1988; Ingold 1988; Hames 2000; Hill 2002), and they are
shared on a large scale. Hawkes (1991) has shown that around 84 per cent
of a hunters prey is consumed by others than himself and his nearest family,
while 58 per cent of the womens yield is eaten outside the family (Barret et
al. 2002: 82).
Bearing in mind the enormous ecological and cultural diversity that other-
wise characterizes these societies and we have data from Arctic areas (Damas
1972), from the South African bush to the Equatorial rainforest (Hart 1978),
and from the Australian desert (Gould 1967) it is thought-provoking that
common patterns of sharing even exist. Naturally, we cannot go into detail
with the different surveys here; however, a single view into how radically the
gesture of sacrifice is actually practised among these indigenous people will be
included, and Kaplan and Hills extensive studies of food sharing among the
Ache tribe in Paraguay is a good example (Kaplan & Hill 1985; Hill 2002).
When an Ache hunter has killed his prey, he brings it to the outskirts of
the settlement and leaves it. Then he starts hunting again, demanding nei-
ther attention nor particular parts of the prey. The days total bag is prepared
by the women, and then distributed among the tribes families by an older

henrik hgh-olesen

man (never the hunter himself ), who gives each family a share equivalent to
the size of the family. Other tribe members are quick to remind him of those
families who have not received anything yet; however, they always mention
other peoples names, never their own. Only the hunter himself never receives
any meat from his own bag. That is simply taboo! The women prepare the
palm marrow, but they are not allowed to share or eat it themselves either.
All other plant material or insects and larvae they have gathered are shared
differently. Here there is no taboo about eating ones own food, and a family
tends to keep the majority of these types of food for themselves, only giving
smaller shares to the group. However, the majority of what they keep is to be
part of reciprocal exchanges with certain families, with whom over a period
of time they are more likely to share food and services.1
So, in many ways, men primarily hunt to share with each other, and not
to guarantee a regular supply of protein for their own family. And this makes
sense, as we are indeed each others means to common goals. Nobody man-
ages alone. Hunting and so many other things might fail, and those of our
ancestors who were capable of entering into reciprocal sharing and working
relations with their fellows had better possibilities of survival than those who
blindly followed a more short-term, selfish strategy.
Along the same lines, experimental economy, and modern game theory
have produced substantial support for phenomena such as altruistic pun-
ishment and strong reciprocity (Gintis 2000; Fehr et al. 2002; Fehr &
Fishbacker 2004) as well as for indirect reciprocity (Alexander 1987;
Weedekind & Milenski 2000; Nowak & Sigmund 2005). Cheaters, para-
sites and other non-reciprocators, who are unwilling to share, spontaneously
arouse a vindictive sentiment (we become indignant, feel resentment and
anger) in the wronged party as well as in those who witness the injustice, and
leave both with a need to punish the offender.
Similarly, social and generous acts spontaneously inspire good will and
generosity in the beneficiary, as well as in those who just witness the deed.
And not only the benefactors but also the witnesses are left with an inner
urge or need to reciprocate (Weedekind & Milenski 2000). Perhaps because a
good deed exerts a beneficent influence both outside and inside the organism,
as Rilling et al.s (2002) neuropsychological studies have shown. The brains
reward centres simply light up when we choose, for example, to cooperate
rather than to let other people down. Complementary to these reinforcing
emotional feedbacks, other studies indicate that our cognitive architecture
is designed to detect whether a person will cheat or is to be trusted in col-
laborative interactions or not, and whether he will reciprocate or deceive by
the subtle, verbal, and non-verbal body and facial cues that he involuntarily
transmits (Frank 1988; Cosmides 1989; Frank et al. 1993; Sugiyama et al.
2002; Brown et al. 2003).

the will to sacrifice

Undoubtedly, these spontaneous emotional feedbacks and inclinations to

sanction non-reciprocators play a key role in the enforcement of our social
norms universally, but these reactions also substantiate the idea stated else-
where (Hgh-Olesen 2006, 2010) that humans are endowed with a hard-
wired reciprocity programme which regulates the majority of our social and
material exchanges.
Of similar relevance is Gerken et al.s (1975) cross-cultural study from the
USA, Japan and Sweden on male university students attitudes to receiving
donations in an experimental game situation, in which they could receive
financial help from their co-players with different obligations attached to
these donations. Gerken et al. found that there was a curvilinear relationship
between the degree of obligations attached to a donation and the benefici-
arys sympathy for the benefactor. Not just donations with many obligations
attached, but also donations without any strings attached resulted in low sym-
pathy for the benefactor, and highest sympathy was rewarded the benefactor
who wanted reciprocity for his help.
It may seem strange that the subjects in these countries do not simply
receive donations without obligations gladly; however, these significant cor-
relations are, if anything, a testimony of how strongly our social relations
are actually regulated by a reciprocity programme. The unrequited gift or
help breaks the social symmetry between people and is thereby in danger
of releasing a number of negative moods in the beneficiary from inferiority
to suspicion: Something for nothing! Whats he up to? and maintains
the beneficiary in what Homans (1961) has called an unpleasant tension of
That this tension is a real social-psychological factor can be seen, not least,
in a phenomenon such as reparative altruism. A person who has received
help which cannot be returned is more likely to help a stranger later on, than
a person who has not previously received any help, or who could reciprocate
a generous act to the original benefactor (Krebs 1970: 297). Moreover, this
tension seems to be a human emotion only, as it is not found with our closest
relatives among the primates. They show no aversion whatsoever to inequita-
ble exchanges that only benefit themselves (Silk et al. 2005).
But from where do these strong norms of sharing and reciprocity originate?
Is it human nature that manifests itself (the cross-cultural agreements, as well
as the emotional and neurological feedback patterns, indicate that this whole
sharing complex of norms, inclinations and sanctions may rest on evolution-
arily developed exchange programmes and dispositions), or are these universal
standards on the contrary an expression of how it does not come naturally to
us to share, and therefore culturally we have to enforce these norms as a bul-
wark against the selfish impulses, by which we are so dominated according to
biologists such as Huxley ([1894] 1989) and Dawkins (1976)?

henrik hgh-olesen

Comparative studies of our nearest relatives among the other primates

can help give us an idea of what is both particularly human and what may
be common evolutionary inheritance. If the principles on which we base
our sociality and morality of sharing are evolutionarily founded, then similar
fundamentals should also be found in a more or less developed form in e.g.
chimpanzees and bonobos with whom we share 9899 per cent of our genes.

Sharing behaviour in Pan and monkeys

In nature, chimpanzees (Pan troglodytes) and bonobos (Pan paniscus) sponta-

neously give a pant hoot when they come across a big amount of food. This
scream attracts the others attention to the food, and is as such a type of social
advertisement, which announces that there is plenty of food here.
The possession of food usually leads to democratic begging on all levels
of the social hierarchy: lower animals beg from higher ones and higher ani-
mals from lower, and both high and low may react with temper tantrums if
prolonged begging is not rewarded, which both shows that in these animals
there is an expectation that you should share, as well as a widespread respect
for private property, because although they may get angry if they do not get
anything, they usually do not steal (Goodall 1986; De Waal 1996, 1998).
Usually, sharings are passively responsive, and as such a consequence of
widespread begging, and not spontaneous, active gifts whereby an animal
unsolicited gives another animal some of its possession (McGrew & Feistner
1992). Actual gift giving is known from peaceful New World monkeys like
Tamarins and Marmosets who live in extended family groups in which the
eldest members may offer the infants or siblings attractive food items spon-
taneously, and unsolicited (ibid.; Feistner & McGrew 1989; Feistner & Price
2000). Active gifts also occur in Pan (see Yerkes 1943 and Goodall 1986 for
anecdotal examples); however, they are very rare, and Teleki (1973) observed
only four spontaneous meat handovers in a whole year in the troop of wild
living chimpanzees studied by him.
Interesting studies about sharing hunted prey in chimpanzees show that
meat is often used as a social tool to develop and maintain alliances between
the males (Nishida et al. 1992; Mitani & Watts 2001). The meat is not shared
randomly but strategically and reciprocally with (1) those who took part in the
hunt, (2) those who have previously shared with you or (3) assisted you in hos-
tile confrontations with others. Thus, a strong association is found between
meat-sharing and coalitionary support (Mitani & Watts 2001: 920).
Boesch (2001: 3941), who has documented actual collaboration with role
distribution and coordination during these hunts, confirms several of these
discoveries. In the distribution of the prey, the participants in the hunt have

the will to sacrifice

a clear priority, compared to observers who have just come along. Dominant
animals, who have not participated, have to beg, and receive less meat than
the participants. Thus, the individual roles are closely registered and assessed,
and a factor such as time spent on hunt does not mean more meat. Ones
participation is only rewarded if it actually leads to the capture. So we are
far away from the Parable of the Vineyard (Matthew 20:1-16) in the New
Testament, as well as from the Ache peoples egalitarian distributions, and
instead we have an objective calculation according to effort.
The norm of reciprocity we found to such an increased extent in us, will
similarly be found in a number of different primates from the two types of
Pan to Old World and New World monkeys, and it seems fair to conclude
that for primates generally, we have both naturalistic and experimental evi-
dence that (1) grooming, (2) helping in conflict situations and (3) shar-
ing food are reciprocated (Packer 1977; Seyfarth 1980; De Waal & Luttrell
1988; De Waal 1989, 1997; Brosnan & De Waal 2002). And it probably
does not stop here because as De Waal (1996: 153) says: Once a quid pro
quo mind set has taken hold, the currency of exchange becomes secondary.
Reciprocity begins to permeate all aspects of social life. Recently Brosnan
& De Waal (2003) have also found a sense of fairness in brown Capuchin
monkeys, who reacted negatively and refused to cooperate when they were
offered a lesser reward than the one a fellow monkey had just received for
the same effort. Also this equal for equal principle may very well have its
origins in the basic symmetry of justice, which the reciprocity structure
In chimpanzees an even higher type of reciprocity is found, which De Waal
calls calculated reciprocity, and this is based on mental bookkeeping where
not only friendliness and help, but also hostility and resistance, are registered,
remembered and reciprocated over time according to the principles One
good turn deserves another and An eye for an eye, a tooth for a tooth (De
Waal & Luttrell 1988; Brosnan & De Waal 2002).
Only a few actual experimental studies of sharing exist. Older studies by
Yerkes and Yerkes (1935) and Nissen and Crawford (1936) show that chim-
panzees can give food sources to others, but only if they are begged for. At the
same time, the food given away was typically the less attractive. From Silks
studies (1979) of eighteen chimpanzees at the Stanford Outdoor Primate
Facility, we know that sharing most frequently took place between mother
and child, less frequently between adult males and infants, and even less
between adult females and others offspring. Mothers typically share the food
sources, which the offspring cannot get itself, while others primarily share
what is easiest for them to get again. Common for these studies is, however,
that we more or less miss operationalized categories of sharing, as well as com-
parable standards for how much is shared and how.

henrik hgh-olesen

To my knowledge, we only have such information from one chimpanzee

troop at the Yerkes Primate Research Centre consisting of nineteen individu-
als (one male, eight females and ten infants) and one bonobo troop in the
San Diego Zoo consisting of ten animals (one male, two females and seven
infants). Both studies were carried out by De Waal (see De Waal 1989, 1992,
1998 for descriptions), and the method is in all its simplicity exemplary: you
introduce a food source into the animals living quarters, which one animal
can monopolize (e.g. a bunch of attractive, leafy branches tied with a string)
and then you register the exchanges according to a number of operationalized
categories. In 50.4 per cent of the cases the interactions among the chimpan-
zees between possessor and interested party led to sharing, and this number
corresponds with Telekis (1973) observations of meat sharing among wild
living chimpanzees where 54 per cent of the interactions resulted in sharing.
As for the bonobos, approaches and begging led to sharing in 63.4 per cent
of the cases.
Naturally, these data from only two troops are limited; but when Hgh-
Olesen (2004) in a number of pilot studies applied the same procedure and
basically identical interaction categories on a chimpanzee troop of thirteen
individuals (three males, six females and four infants aged 13 years), a chim-
panzee troop of seven individuals (one male, five females and one 2-year-old),
as well as a troop of bonobos of nine individuals (two males, four females
and three infants aged 34 years), the patterns described above were funda-
mentally confirmed. However, when comparative data from nineteen Java
macaques (Macaca fasicularis), sixteen Barbary macaques (Macaca sylvanus)
and a small mandrill troop (Papio sphinx) of four individuals were collected
through the same monopolization procedure, this pattern was dissolved (see
Table 9.1).
When a limited food source that can be monopolized is introduced to an
animal group, the subsequent repertoire of behaviour, regardless of species,
mainly consists of aggressive and competitive impulses. The data from the
mandrill and macaque experiments fully confirm this picture; however, in
Pan there is clearly something more going on. Competitions do take place,
but once a fixed ownership has been established, the troops members change
in the chimpanzees and the bonobos from a competitive mood to a friendly,
appealing pattern of interaction, in which they peacefully try to beg their
way to the food, rather than aggressively fight for it. At the same time, the
possessors displayed considerable tolerance of sharing, which is seen in that
up to 66 per cent of the beggings led to sharing, and out of this no less than
52 per cent of the chimpanzees and 44 per cent of the bonobos exchanges
(the difference is not significant) occurred within the passive sharing category,
in which they eat together of the same food in close proximity of each other.

the will to sacrifice

However, it is perhaps most noticeable that all animals beg! Even when the
possessor is a low-status animal, the dominant animals in these troops, includ-
ing the alpha animals, neatly line up in the row of beggars. For although the
alpha animals have the power to take the prey from the possessor, they appar-
ently have no right to do so! And this is thought-provoking because it indi-
cates that in these hierarchical animals, during food and possession situations,
we trace a set of more complex rules than the simple, physical relations of
dominance: power simply does not equal right and therefore theft is rare. It
occurred more during the experimental conditions than under normal daily
co-existence; but even here it was rare. By combining the naturalistic and
experimental observations, we found that only in 23 per cent of all food
interactions did theft or attempted theft occur, which always led to agitation,
fights and scuffles between those involved. These figures also confirm that
in Pan generally a considerable respect for private property can be traced,
and that they, like us, to a large extent respect the first come, first served
principle, when determining an issue of rights or property, and even penal-
ize breaches of norms. Actual gifts, where an animal gives another animal a
food source spontaneously (and unsolicited), are basically as rare as theft, and
they were only observed in 24 per cent of the cases during the naturalistic
In contrast, with the macaques and the mandrills that live socially in des-
potism, power does equal right. Among these animals there is an absolute
hierarchy, which leaves no room for actual sharing or close contact between
those interested and the possessor. Lower ranking animals must make way for
the higher ranking that can always threaten their way to possession, and then
aggressively monopolize the food source. Consequently, the total number of
Pan sharings far outnumbered the total number of monkey sharings (Pan n =
127, monkeys n = 56; 2(1) = 19.8472, p 0.0001). The close peering and
passive sharings observed in the Pan groups were completely absent. Instead
accepted takings, where an animal jumps in and collects scraps while the
dominant animal is away or busy, was the dominant pattern, significantly
increased in relation to the Pan average (Pan n = 23, monkeys n = 56; 2(1) =
35.9882, p 0.0001), and yet another sign of these monkeys limited ability
to practice close sharing.
The number of observations in these pilot studies is limited, but a more
comprehensive study (n = 80 experiments; Hgh-Olesen in preparation)
involving one group of chimpanzees and one group of bonobos is currently
under completion. In this study the table manners are put further on trial
due to the fact that in half of these experiments the leaves are daubed with
syrup, which makes them even more appetizing. However, this does not
increase the number of thefts, but apparently the number of beggings!

Table 9.1 Sharing in six groups and five species of primates (Hgh-Olesen 2004).

Category Naturalistic observations Experimental observations

Chimps Bonobo Chimps Chimps Bonobo Mandrill Java Barbary
(Givskud) (Givskud) (Aalborg)
Number of interactions 299 98 90 36 74 69 26 48
1 Active sharing 4% 2% 0% 0% 0% 0% 0% 0%
2 Passive sharingb 26.8% 28.6% 52.2% 41.7% 44.6% 0% 0% 0%
3 Accepted taking 5.7% 15.3% 13.3% 11.1% 9.5% 30.4% 50% 45.8%
4 Sexual bargaind 1% 12.2%
5 0 begging ignored 61.5% 52% 26.7% 47.2% 29.7% 69.6% 34.6% 52.1%
6 Theft (inc. attempted)f 2% 1% 7.8% 0% 4.1% 0% 15.4% 2.1%
Total sharing 36.5% 46.9% 65.6% 52.8% 66.2% 30.4% 50% 45.8%
Active, unsolicited transfers.
Peaceful co-feeding after initial begging.
Collects scraps (Pan, macaque, mandrill) or gently removes some of the others possessions without these initiatives being sanctioned (Pan only).
Sexual contact and genital stimulation is offered in return for food.
Begging is ignored and results in no sharing.
An animal aggressively removes (or tries to remove) anothers possession under protest from the possessor.
the will to sacrifice

Them and us: shared fundamentals and human particulars

For humans the shared meal is a central symbol, which signals peace, broth-
erhood and reciprocity worldwide. As long as we can share food and eat
together, the dialectics of recognition are intact, and therefore it is also under-
lined again and again in the New Testament how Jesus would eat with pub-
licans, Pharisees and sinners.
With Pan, they are also able to share and eat peacefully together, and at
the same time there is a considerable closeness, as well as (them being hier-
archical creatures), an amazing equality among the troops members while
they share. They are most definitely capable of playing by the rules and to
a great extent suppress or sacrifice if you like the more selfish and aggres-
sive impulses in favour of a tolerant sharing behaviour. Central aspects of the
sharing and interaction patterns found in our species can consequently be
found in Pan. As with us a norm of sharing exists! Begging in itself is a tes-
timony of this, and furthermore, up to 66 per cent of the beggings result in
sharing, as shown. Like us, Pan is capable of practising closely co-ordinated
or democratic patterns of sharing, whereby more, together, peacefully eat
of the same food. As with us, we find in Pan a widespread respect for pri-
vate property, and the first come, first served principle is also here a widely
accepted guideline, when a new property relation is to be established, and as
with us, the troops internal relations are to a very high extent regulated by
principles of reciprocity. Even the impulse to react emotionally in breaches of
norms towards those who will either steal, or not share after prolonged and
correct begging, can be found in Pan.
These similarities are again substantial indications that central parts of our
own sociality, sharing norms and actual sharing behaviour may be based on
a foundation of evolutionarily developed dispositions, as well as, how these
basic fundamentals are ingredients in a common higher primate register,
which probably already existed in the forefather we had in common with the
Pan species 56 million years ago. Naturally, it is possible that these concord-
ant common traits developed independently of each other at three different
points in time, in three different species; however, it is not very likely. Instead
the theory of a common evolutionary inheritance is after all the simplest
At the same time these common traits are lost when we climb further
down the primate tree, and with the macaques and the mandrills the pattern
disappears, as shown. Passive sharing, begging and protest against monopo-
lization are basically absent in these hierarchical species. Obviously, other
norms dictate what is a reasonable distribution of the resources, and whereas
with Pan and us you can expect a share of the prey, this expectation is not
present in the macaques and the mandrills (see Table 9.2).

henrik hgh-olesen

Table 9.2 Sociality and sharing: qualitative comparisons.

Human Pan Macaque/mandrill

Extensive sharing norms and Sharing norm and No sharing norm, limited
sharing capability sharing capability sharing capability

Emotional reactions at breach of Like humans Dominance and

norms and theft submission

Respect for private property first Like humans Despotism with fear of
come, first served principle punishment

Close, co-ordinated sharing Like humans Monopolization and

patterns despotism with large
possession space

Power right Like humans Power = right

Complex reciprocity Simple reciprocity Few cases of reciprocity

Higher primate register

The branch of Anthropoid apes grew out of the primate tree about 30
million years ago, when our line separated from that of Old World mon-
keys, such as mandrills and macaques. We then shared about 25 million years
of common evolution with the chimpanzees and bonobos before our ways
parted, and therefore it is no surprise that Pan anatomically, socially and
mentally resembles us more than our more distant relatives. However, at
the same time, in the 56 million years which seem to have passed since we
shared a common forefather with our Pan relatives, we have thus, through
about 200,000 generations of selections, had the opportunity to develop both
new traits as well as unique versions of the common sharing dispositions, and
in this connection certain things do separate us.
What is first noticeable is that the actual capability of sharing seems sig-
nificantly increased in our species in relation to what we find in the other
primates. At the same time we find in the moral and religious codices we have
universally developed, that an even very radical demand for sharing is formu-
lated. From the indigenous peoples egalitarian distribution and taboo against
eating ones own prey to the Parable of the Vineyards radical insistence that
those who arrived in the eleventh hour, would receive just as much as those
who had worked the whole day, we come across the same increased demand
for sharing. With Pan the first impulse is primarily to monopolize the food
source and secure it from the other interested parties. A prey killed alone and

the will to sacrifice

without the others knowledge will therefore not be shared but eaten on the
spot (Knauft 1991), whereas our impulse or norm is clearly to actively share
what we have with others. If I feel like having a mint, I have to offer it to
others first. Nor is the prey killed by a hunter, eaten on the spot; rather it is
brought to the common home base in order to be shared. Finally, it is pre-
dominantly passive, responsive sharing, which Pan practises. They have to be
compelled to do it, and only in 24 per cent of the cases do they themselves
take the initiative to give a present.
In us the sacrifice initiative is clearly placed upon the individual and the
norm, as shown, that you have to give to receive. At the same time we are
so emotionally obligated to this norm that we practise among other things
altruistic reinforcement and punishment on our own account, by reward-
ing the good example or punishing people who do not share properly, even
when they are not cheating us personally, just as in the ultimatum game we
reject and punish the offer which is much too low, although strictly speaking,
something ought to be better than nothing.
The last-mentioned behaviour is completely unthinkable in Pan, where
not even a crumb from the others teeth would be rejected! There is also
much that indicates that although the basic patterns of reciprocity are the
same among them and us; we do take the principle to its maximum, unlike
Pan. Reciprocity comes in simple and complex forms, but with the sacrificial
gesture and its universal norm of active sharing, humans climb the ladder of
higher sociality, moving from simple responsive tit for tat reciprocity to
complex and proactive golden rule reciprocity. Just as we prefer gifts, which
can and must be reciprocated, rather than donations without any obligations
attached, because, as shown, they breach the social symmetry between people
and place us in a tension of obligation, which we find unpleasant, but which
undoubtedly does not bother the members of the two Pan species!
These differences also indicate the absence of several of the secondary emo-
tions like pride, shame and guilt, which are so predominant in us, and which
Parker (1998) called self-conscious emotions, because they, unlike the pri-
mary emotions like joy, fear and anger, require increased self-consciousness,
cognitive reflection and an ability to assess ones own behaviour in relation to
a common social standard.
In short, there is more sharing, reciprocity, sanctions and so on in humans,
and although this with certain exceptions is merely more of the kind
found in Pan, the question still arises: why this radicalisation of the require-
ment of sharing, and, during the evolution of the species, which material con-
ditions have demanded and prompted such an increased willingness to make
a sacrifice? The limited space here does not allow a long explanation, and we
have with this surplus come across something specifically human, which a
multitude of concurrent selection pressures may very well have developed.

henrik hgh-olesen

At the risk of over-simplifying matters, there are, however, a few issues that I
will briefly emphasize.
One of the things that separate us from our primate relatives is the enor-
mous resources, which the males in our species invest in the well-being of
the family generally, and in offspring and parental care specifically. The con-
tact to the mother/child entity is far closer among us than in Pan, where the
males do not become involved in parental care or closeness to the offspring
to any great extent; in any case, due to the females promiscuity, they can
never know whether one infant or another is theirs, so such an involvement
would also be a dubious investment, evolutionarily speaking, and therefore
the motivation for systematically providing and sharing is understandably
smaller here.
However, as a consequence of our rapidly increased brain volume, already
around Homo habilis and Homo erectus, we are forced into neotenys increas-
ingly shorter gestation periods, to be able to even give birth to children with
big heads necessary for the increased brain volume. Then this unfinished and
helpless offspring brings out the family and its specialized work distribu-
tion as an evolutionary necessity. A female with such a helpless and vulnerable
offspring has an extreme need for stable provision and protection. Thus, she
is more inclined to select the males that have friendly, caring and generous
traits, and who after mating seem willing to spend time and resources on
her and her offspring. Therefore, between two powerful males, she prefers
the generous one who shares and shows care for the offspring. At the same
time, in order to catch these attractive males she must also sacrifice herself by,
for example, controlling the more selfish and sexually promiscuous impulses
that she may have, so that through her special attention and exclusive sexual
devotion, the male can be reasonably guaranteed that it is not somebody elses
genes he is spending his resources on. Hereby a powerful, self-perpetuating
circuit has started, which through time must have promoted both willingness
to make sacrifices and capability of sharing and reciprocity. Even the worlds
women of today are still focusing on their potential partners financial capa-
bilities, and whether they are good providers, who can guarantee the offspring
its necessities, while this worlds men, like Busss (1989) survey showed, are
more preoccupied with whether their partners are beautiful and sexually
faithful, as the chances for their own healthy offspring are also greatest here.
Likewise, a number of conditions, with their origins in human social life
and its need for work and role divisions, have undoubtedly both demanded
and promoted the capability of sharing and reciprocity. To complete the hunt,
they have, for instance, needed courage, strength and speed, but also weap-
ons, planning and coordination, and the one who was fast and brave was not
necessarily also the best craftsman, track finder or organizer, and therefore
they have distributed the tasks, specialized, and basically depended on each

the will to sacrifice

other. However, all these distributions, which together can lead to the target
of survival, do only function under the mutual sacrifices auspices: I need to
trust that you will do your part, while I do mine, otherwise it wont work.
The individuals (and groups) who have been part of this mutuality, have had
better evolutionary chances than those who were incapable of consecrating
the sacrifice (Hgh-Olesen 2010)!
Thus it is no coincidence that the sacrifice is a ritualized central structure
in religious practices worldwide and as such a sacred universal. As I have
tried to show elsewhere (Hgh-Olesen 2006), the sacrifice is a ritualized gift
accompanied by a prayer, or a wish: The wish that through this offering, the
powers can be committed to reciprocate and return one good favour with
another. Do ut des as it is put in the Lex Romana, or Ddmi se, dehi me
as the Vedas formulate it: I give to you, so that you shall give to me. This
is the sacred formula behind the sacrifice throughout the world, and this
formula is sacred to us, because it concentrates the bare necessities and sym-
bolically highlights the natural quid pro quo relations that have to prevail
among humans, if a society is to exist at all! And seen from this angle, sacrifice
is much more than a particular religious concept or practice: It is a psycho-
logical key factor in the sociality of humans.
So have we finally proven that humans are by nature good and generous,
and not selfish, as biologists like Huxley and Dawkins have so far assumed
it? Not at all: we are exactly as selfish and self-sacrificing as it has evolution-
arily paid off for us to be. Alternatively, what we do see is that several of our
basic social and moral principles seem to be evolutionarily established, just
as Darwin ([1874] 1998) assumed, in that they are based on some social
exchange and reciprocity dynamics, which the evolutionary selection pressure
has developed and placed as common ground in the higher primates. And
this is in many ways a more balanced version of Dawkinss (1976) assump-
tion that anything that has evolved by natural selection should be selfish,
but then not more than that.
There is still some way to go. The data we operate with for people and pri-
mates are limited. Not least research on children would throw more light on
these matters. What would actually happen if we exposed groups of nursery,
kindergarten and school children to the monopolization experiment, and then
for a thirty-minute period registered all interaction? Would the first impulse
here be monopolization? Would the number of active sharings increase in
relation to Pan and with increasing age, and would the total number of shar-
ings exceed what we discovered with chimpanzees and bonobos? A number
of obvious experiments call for an experimenter.

henrik hgh-olesen


1. Hames (2000) has in studies of the Yanomamis found the general tendency that the
bigger a tribe is, the more the food sharing strategy changes from simple equal sharing
between everybody to reciprocal altruism, whereby a few families (45) mutually share
more among each other than they do with other families.


Alexander, R. D. 1987. The Biology of Moral Systems. New York: Aldine.

Barret, L., R. Dunbar & J. Lucett 2002. Human Evolutionary Psychology. New York: Palgrave.
Betzig, L. L. & P. W. Turke 1986. Foodsharing on Ifaluk. Current Anthropology 27(4):
Boesch, C. 2001. Cooperative Hunting Roles among Tai chimpanzees. Human Nature 13:
Brosnan, S. F. & F. B. M. De Waal 2002. A Proximate Perspective on Reciprocal Altruism.
Human Nature 13(1): 12952.
Brosnan, S. F. & F. B. M. De Waal 2003. Monkeys Reject Unequal Pay. Nature 425: 2979.
Brown, W. M., B. Palameta & C. Moore 2003. Are there Nonverbal Cues to Commitment?
An Exploratory Study Using the Zero-Acquaintance Video Presentation Paradigm.
Evolutionary Psychology 1: 4269.
Buss, D. M. 1989. Sex Differences in Human Mate Preferences: Evolutionary Hypotheses
Test in 37 Cultures. Behavioral and Brain Sciences 12: 149.
Cosmides, L. 1989. The Logic of Social Exchange. Cognition 31: 187276.
Damas, D. 1972. Central Eskimo Systems of Food Sharing. Ethology 11: 22040.
Darwin, C. [1874] 1998. The Descent of Man, and Selection in Relation to Sex, 2nd edn. New
York: Promethus Books.
Dawkins, R. 1976. The Selfish Gene. Oxford: Oxford University Press.
De Waal, F. B. M. 1989. Food Sharing and Reciprocal Obligations among Chimpanzees.
Journal of Human Evolution 18: 43359.
De Waal, F. B. M. 1992. Appeasement, Celebration and Food Sharing in the Two Pan Species.
In Topics in Primatology, vol I: Human Origins, T. Nishida, W. C. McGrew, P.Marler, M. P.
Pickford & F. B. M. de Waal (eds), 3751. Tokyo: University of Tokyo Press.
De Waal, F. B. M. 1996. Good Natured: The Origins of Right and Wrong in Humans and Other
Animals. Cambridge, MA: Harvard University Press.
De Waal, F. B. M. 1997. The Chimpanzees Service Economy: Food for Grooming. Evolution
and Human Behavior 18: 37586.
De Waal, F. B. M. 1998. Chimpanzee Politics. Baltimore, MD: Johns Hopkins University Press.
De Waal, F. B. M. & L. M. Luttrell 1988. Mechanisms of Social Reciprocity. Ethology and
Sociobiology 9: 10118.
Fehr, E. & U. Fischbacher 2004. Third-Party Punishment and Social Norms. Evolution and
Human Behavior 25: 6387.
Fehr, E., U. Fischbacher & S. Gchter 2002. Strong Reciprocity, Human Cooperation and
the Enforcement of Social Norms. Human Nature 13: 125.
Feistner, A. T. C. & W. C. McGrew 1989. Food-Sharing in Primates: A Critical Review. In
Perspectives in Primate Biology 3, P. K. Seth & S. Seth (eds), 2136. New Delhi: Today and
tomorrows primers.

the will to sacrifice

Feistner, A. T. C. & E. C. Price 2000. Food-Sharing in Black Lion Tamarins. American Journal
of Primatology 52: 4754.
Frank, R. H. 1988. Passions within Reason: The Strategic Role of the Emotions. New York: W.W.
Frank, R. H., L. Gilovicht & D. T. Regan 1993. The Evolution of One-Shot Cooperation:
An Experiment. Ethology and Sociobiology 14: 24756.
Gerken, K. S., P. Ellsworth, C. Maslach & M. Seipel 1975. Obligation, Donor Resources
and Reactions to Aid in Three Cultures. Journal of Personal and Social Psychology 31(3):
Gintis, H. 2000. Strong Reciprocity and Human Sociality. Journal of Theoretical Biology
206: 16979.
Goodall, J. L. 1986. The Chimpanzees of Gombe. Cambridge, MA: Harvard University Press.
Gould, R. A. 1967. Notes on Hunting, Butchering and Sharing of Game among the
Ngatatjare and Their Neighbours in the West Australian Desert. Kroeber Anthropological
Society Papers 36: 4166.
Gouldner, A. Q. 1960. The Norm of Reciprocity. American Sociological Review 25(2): 16178.
Hames, R. 2000. Reciprocal Altruism in Yanomamo Food Exchange. In Adaptation and
Human Behavior, L. Cronk, N. Chagnon & W. Irons (eds), 397417. New York: Aldine
de Gruyter.
Hart, J. A. 1978. From Subsistence to Market: A Case Study of the Nbuti Net Hunters.
Human Ecology 6: 32453.
Hawkes, K. 1991. Showing Off: Tests of Another Hypothesis about Mens Foraging Goals.
Ethology and Sociobiology 11: 34161.
Hill, K. 2002. Altruistic Cooperation during Foraging by the Ache. Human Nature 13(1):
Hgh-Olesen, H. 2004. Offerets Grundform: Deling og Socialitet hos Mennesket og de
vrige Primater. Bulletin fra Forum for Antropologisk Psykologi 14: 536.
Hgh-Olesen, H. 2006. The Sacrifice and the Reciprocity-Programme in Religious Rituals
and in Mans Everyday Interactions. Journal of Cognition and Culture 6(34): 499519.
Hgh-Olesen, H. 2010. Human Morality and Sociality: Evolutionary and Comparative
Perspectives. New York: Palgrave Macmillan.
Homans, G. C. 1961. Social Behavior: Its Elementary Forms. New York: Harcourt, Brace &
Huxley, T. H. [1894] 1989. Evolution and Ethics. Princeton, NJ: Princeton University Press.
Ingold, T. 1988. Notes of the Foraging Mode of Production. In Hunters and Gatherers, vol. 1:
History, Evolution and Social Change, T. Ingold, D. Riches & J. Woodburn (eds), 26985.
Oxford: Berg.
Kaplan, H. L. & K. Hill 1985. Foodsharing among Ache Foragers. Current Anthropology
26(2): 22340.
Knauft, B. M. 1991. Violence and Sociality in Human Evolution. Current Anthropology
32(4): 391410.
Krebs, D. L. 1970. Altruism: An Examination of the Concept. Psychological Bulletin 73(4):
Lee, R. B. 1988. Reflections on Primitive Communism. In Hunters and Gatherers, vol. 1:
History, Evolution and Social Change, T. Ingold, D. Riches & J. Woodburn (eds), 25268.
Oxford: Berg.
Mauss, M. [1925] 1990. The Gift: The Form and Reason for Exchange in Archaic Societies.
London: Routledge.
McGrew, W. C. & A. T. C. Feistner 1992. Two Nonhuman Primate Models for the Evolution

henrik hgh-olesen

of Human Food Sharing. In The Adapted Mind, J. H. Barkow & L. Cosmides (eds),
22943. New York: Oxford University Press.
Mitani, J. C. & D. P. Watts 2001. Why Do Chimpanzees Hunt and Share Meat?. Animal
Behaviour 6: 91524.
Nishida, T., T. Hasegawa, H. Hayaki, Y. Takahata & S. Uehara 1992. Meatsharing as
a Coalition Strategy by an Alphamale Chimp. In Topics in Primatology, vol I: Human
Origins, T. Nishida, W. C. McGrew, P. Marler, M. P. Pickford & F. B. M. de Waal (eds),
15974. Tokyo: University of Tokyo Press.
Nissen, H. W. & M. P. Crawford 1936. A Preliminary Study of Food-Sharing Behavior in
Young Chimpanzees. Journal of Comparative Psychology 22: 383419.
Nowak, M. A. & K. Sigmund 2005. Evolution of Indirect Reciprocity. Nature 437(27):
Packer, C. 1977. Reciprocal Altruism in Papio Anubis. Nature 265: 4413.
Parker, S. T. 1998. A Social Selection Model for the Evolution and Adaptive Significance of
Self-Conscious Emotions. In Self-Awareness: Its Nature and Development, M. Ferrari &
R.Sternberg (eds), 10834. New York: Guilford Press.
Rilling, J. K., D. A. Gutman, T. R. Zeh, G. Pagoni, G. S. Berns & D. Kilts 2002. A Neutral
Basis for Social Cooperation. Neuron 35: 395405.
Seyfarth, R. 1980. The Distribution of Grooming and Related Behaviors among Adult Female
Vervet Monkeys. Animal Behavior 28: 798813.
Silk, J. B. 1979. Feeding, Foraging and Food Sharing Behavior of Immature Chimpanzees.
Folia Primatologica 31: 12342.
Silk, J. B., S. F. Brosnan, J. Vonk, J. Henrich, D. J. Povinelli, A. S. Richardson, S. P. Lambeth,
J. Mascaro & S. J. Schapiro 2005. Chimpanzees are Indifferent to the Welfare of Unrelated
Group Members. Nature 437(27): 13579.
Simmel, G. 1950. The Sociology of George Simmel. Glencoe, IL: Free Press.
Sugiyama, L. S., J. Tooby & L. Cosmides 2002. Cross-Cultural Evidence of Cognitive
Adaptations for Social Exchange among the Shiwar of Equadorian Amazonia. Proceedings
of the National Academy of Sciences 99: 11, 53742.
Teleki, G. 1973. The Predatory Behavior of Wild Chimpanzees. Lewisburg, PA: Bucknell Uni-
versity Press.
Testart, A. 1987. Game Sharing Systems and Kinship Systems among HunterGatherers.
Man 22: 287304.
Weedekind, C. & M. Milenski 2000. Cooperation through Image Scoring in Humans.
Science 288: 85052.
Yerkes, R. M. 1943. Chimpanzees: A Laboratory Colony. New Haven, CT: Yale University Press.
Yerkes, R. M. & A. W. Yerkes 1935. Social Behavior in Infrahuman Primates. In A Handbook
of Social Psychology, G. Lindley (eds), 9731033. Worcester, MA: Clark University Press.

Apetales: exploring the deep
roots of religious cognition
Tom Sjblom

Everything starts somewhere, although many physicists disagree.

But people have always been dimly aware of the problem with the
start of things. They wonder aloud how a snowplough driver gets
to work, or how the makers of dictionaries look up the spelling
of the words. Yet there is the constant desire to find some point
in the twisting, knotting, raveling nets of space-time on which a
metaphorical finger can be put to indicate that here, here, is the
point where it all began (Terry Pratchett 1997: 11)

Prelude: understanding origins

We humans seem to have an innate drive striving to understand the world

around us. This is not something restricted to scholars and the academic
world but as demonstrated by the quote from Terry Pratchett it is some-
thing our cognition is primed to do (e.g. Guthrie 1993; Holyoak & Thagard
1995; Strauss & Quinn 1997). That we humans strive to understand is not in
doubt but as we all know what exactly is meant by understanding is a hotly
debated issue, and it seems to mean different things for different peoples in dif-
ferent contexts (e.g. Outhwaite 1975; Gothni 2005; Dennett 2006: 25864).
The debate might as well go on ad infinitum, and my intention is certainly not
to tackle it here. Instead, I will draw attention to the rather trivial point that
depending on what we mean by, for example, understanding, religion as an
instance of human behaviour might sometimes have important consequences
in terms of how feasible our approach is in scientific terms. Understanding this
is essential in order to fully appreciate the recent scholarly efforts to revisit the
issue of the origins of religion a task which is commonly held to be unsci-
entific and not worthy of serious scholarship because the evidence whatever
that might be has vanished into the mists of prehistory and is, therefore,

tom sjblom

not accessible to modern scholars. This makes it impossible to argue persua-

sively for any theory about the origins of religion, because by necessity they
are always based on pure speculation which cannot be refuted on the basis of
evidence (e.g. Evans-Pritchard 1965). In the words of Garry Trompf:

It can be asserted with confidence that no one can know for cer-
tain how what we commonly term religion first began. It may
not be trite to infer that it is as old as humanity itself that men
and women are naturally religious but then no one is currently
in a position to decide with any precision when and where the
first humans actually came into being. (Trompf 2005: 21)

The classical evolutionary theories of the origins of religion were criticized

mainly because of their progressivist understanding of how history proceeds.
Thus, scholarly interest had been very much focused on distinguishing dif-
ferent stages in the evolution of mental capacities and the place of religion
in this process. The main line of evidence followed the Spencerian view of
human evolution, where an entity always develops from a simpler to a more
complex form. The critics pointed out, correctly, that evidence of this type
of development was not available in the ethnographic data. Chronologically
earlier religious traditions could be as complex as more recent ones, and the
complexity of a religious tradition did not necessarily correlate with the com-
plexity of the social environment in which it was nested (see Widengren
1946; Capps 1995: 945). As implied by Trompf above, if we are to discuss
the origins of religion, we must turn our attention to those features that all
religious traditions share with each other, not to those that make them dis-
tinct. After all, it is the origins of religion as a universal aspect of human exist-
ence and social behaviour that we are striving to explain, and this capacity
must be shared by all humans alike.
Trompf goes on to argue that discussing the origins of religion is in effect
closely connected with the discussion of the origins of anatomically modern
Homo sapiens and the origins of behavioural modernity. Indeed, the present
archaeological evidence clearly demonstrates that where there are signs of
human culture there are also signs of religion (e.g. Mithen 1996: 171210;
Hayden 2003: 2045; Trompf 2005: 175258). Therefore, the origins of
religion cannot really be separated from the origins of human capacity for
culture. Indeed, understanding this has been the common feature and driv-
ing force of the recent naturalistic discussions of the origins of religion (see
Guthrie 1993; Mithen 1996; Boyer 2001; Atran 2002; Lewis-Williams 2002:
180203; McClenon 2002; Dennett 2006).
Trompf does express his doubts over whether we are ever capable of pin-
ning down the birth of behavioural modernity any more than the origins of


religion. However, his doubts seem to be based on the typical but mistaken
idea shared by many historians and archaeologists that the only empirical
evidence that we can use for discussing prehistoric humans are archaeological
artefacts and other evidence of cultural nature. Luckily, this is not the case.
After all, when we are discussing religion, we are dealing with the mental
aspects of human culture. This means that in addition to the cultural artefacts
we can and ought to include among the evidence the biological aspects
of human behaviour our cognitive architecture and what we know of its
evolution. This also means that instead of striving to sort out a historical era
when religion and behavioural modernity as a whole came into existence, the
focus should be on explaining what cognitive conditions make religion as we
know it possible. This was actually understood already by the classical theo-
rists of the origins of religion, or at least by some of them like Robert Marett,
who pointed out that the search for the origins of religion is not a search of a
historic era, but a search of psychological primitives (Marett 1909: x).

The elements of behavioural modernity

Reformulating the quest for the origins of religion into an examination of the
cognitive traits that enable religious behaviour is only the first step. As Niko
Tinbergen has pointed out, behaviour (and cognitive traits) does not fossilize,
so there must be some other type of evidence at our disposal. This evidence is
comparative by nature, and it is collected from a variety of sources (Tinbergen
1963: 4278). For example, there is no reason to think that religious behav-
iour of people today would be based on different cognitive traits than that of
our early ancestors, after all, the birth of behavioural modernity seems to be
connected with the evolution of big brains and, almost by definition, the size
of the brain and its cognitive capacities are what we share with the earliest
of modern humans (e.g. Campbell 1998: 24858; Allman 2000). Thus, it is
more than appropriate to use the evidence from cognitive research among con-
temporary subjects for interpreting data from earlier (pre-)historical periods.
In addition, we can also learn much from comparing our own cognitive
traits with those of our closest related species (Tinbergen 1963: 428). Our
closest relatives, other human species, are all extinct, but we have been able
to learn much of ourselves also by comparing our own behaviour to other
living primates and especially to different species of apes (see e.g. Mithen
1996, 2005; Donald 2001; de Waal 2005; Stringer & Andrews 2005).1 It
is on these grounds, by combining the evidence from contemporary human
societies with that collected from the group behaviour of closely related spe-
cies and interpreting this all in a cognitive framework, that archaeologists
and palaeontologists have been able to create lists of the essential features

tom sjblom

characterizing the mind of behaviourally modern humans (see Henshilwood

& Marean 2003: 62751). As always, some controversy prevails over which
features should be counted in, but one of the more influential suggestions
is the one presented by Sally McBrearty and Allison Brooks. Their list is
based on a close examination of the archaeological evidence from the African
Middle Stone Age and the European Upper Palaeolithic and on the basis of
this they argue that the essential cognitive features are:

Abstract thinking or the ability to act with reference to abstract concepts

not limited in time and space.
Planning depth or the ability to formulate strategies based on past expe-
rience and to act upon them in a group context.
Behavioural, economic and technological innovativeness.
Symbolic behaviour or the ability to represent objects, people and
abstract concepts with arbitrary symbols, vocal or visual, and to reify
such symbols in cultural practice (McBrearty & Brooks 2000: 492).

Having been able to sort out the cognitive features that seem to be necessary
for modern behaviour, including religion, it now remains for us to sort out
how exactly they make it happen. Note that any form of social behaviour usu-
ally requires more than one of the above mentioned cognitive features to be
active. Thus, when we want to explain, for example, religious behaviour, we
could approach the topic by discussing the type of abstract thinking used for
religious representations (e.g. Boyer 1994, 2001), or maybe through behav-
ioural innovativeness (e.g. Donald 1991: 20168; Mithen 1996: 21122).
However, in what follows, I will focus on the cognitive origins of symbolic
behaviour. I do this because symbolic behaviour might actually be the most
fundamental of these traits (Chase & Dibble 1987: 2636; T. Deacon 1997;
Henshilwood & Marean 2003: 6357). Also, it is this aspect of religious
behaviour which is probably best presented in the archaeological record. At
least it is symbolic behaviour that has gained most attention by scholars dis-
cussing the origins of religion from an archaeological perspective. Naturally,
by doing this, I am aware that the different cognitive features given by
McBrearty and Brooks are not mutually exclusive, and the cognitive systems
responsible for them sometimes overlap.
It has been a standard strategy among scholars discussing the origins of
cultural behaviour to talk about a creative or cultural explosion or a relatively
short temporal period when quite dramatic changes occurred in human social
behaviour and planning capacity, at least if we are to believe the archaeologi-
cal evidence available to us. This creative period is usually placed sometime
after 60,000 years ago in the Upper Palaeolithic period (see Pfeiffer 1982;
Mithen 1996: 171210; Klein & Edgar 2002). This has led many to presume


that the Upper Palaeolithic period (e.g. c.60,000c.10,000 years ago) is the
formative period for behavioural modernity and, thus, should be the place for
the metaphorical finger pointing to the origins of symbolic behaviour.
As the name of their article suggests, McBrearty and Brooks strive to argue
to my mind accurately that while the Upper Palaeolithic can be taken
to present the last formative phase of modern human behaviour, it is a grave
mistake to think that the cognitive make-up responsible for this, or at least
most of the makeup, would not have been present in the human mind long
before.2 The curious thing with the so-called creative explosion is that until it
arose, all the major developments in human behaviour appear to have evolved
in concert with developments in human anatomy. No evidence for the latter
can be connected with the Upper Palaeolithic era. It appears to be a purely
behavioural revolution for, in terms of anatomy, modern humans came into
being much earlier at least over 200,000 years ago (Donald 1991; Klein &
Edgar 2002: 21; Wood 2005: 100115). In terms of the cognitive origins of
symbolic thought, and religious behaviour, the Herto skulls belong certainly
among the most interesting in this respect. My intention in this article is to
discuss aspects of the origins of religious behaviour by exploring the impact
of this new data in terms of how it could be reconciled with what the present
theories of the origins of religious behaviour argue in the matter.

A new dawn in Herto

One of the more important signs for symbolic behaviour with obvious con-
nections to the origins of religion is the presence of burials with grave goods
and/or mortuary practices with symbolic elements in them. A connection
with religion and burials can naturally be contested and there is an ongoing
debate over what actually counts as a burial, especially a burial with sym-
bolic meaning (e.g. Walker & Shipman 1996: 2856; Mithen 1999: 14769;
Pearson 1999; Taylor 2002). For example, the Spanish palaeoanthropologist
Juan Luis Arsuaga has argued on the basis of his discoveries in Atapuerca that
archaic modern humans (Homo heidelbergensis) engaged in funerary behav-
iour already 350,000 years ago (Arsuaga 2003: 2735). There is also an ongo-
ing debate over whether the signs of the use of ochre and other pigments,
the earliest appearance of which dates to around 300,000 years ago, should
be interpreted as the birth of symbolic culture. This claim is made mainly
because such earth pigments do not seem to have any direct utilitarian use in
later traditions. In the Upper Palaeolithic, for example, ochre was frequently
used in funeral behaviour (see Watts 1999: 11346). However, the earliest
finding which most commentators today accept as showing traces of symbolic
behaviour is the so-called Herto skulls.

tom sjblom

The Herto skulls were discovered in 1997 from Middle Awash in Ethiopia
by Tim White and his colleagues, and the discovery was reported in a series of
articles in the journal Nature in 2003 (see Clark et al. 2003: 74752; Stringer
2003: 6925; White et al. 2003: 7427). The skulls belong to anatomically
modern humans, two adults and one child, which were found together with
some artefacts. Two things make these skulls interesting. First, radioisotopi-
cally the fossils have been dated to between 160,000 and 154,000 years ago.
In addition, according to the excavators, their morphological features com-
bine elements from both archaic African fossils predating them and later
anatomically modern humans of the Late Pleistocene era. White and his col-
leagues therefore describe them as a sample threshold population, which is
on the verge of anatomical modernity but not yet fully human (White et al.
2003: 745; but see Trinkaus 2005: 21314).
The second interesting feature of these skulls is that they had clearly been
exposed to cultural modification. The least modified of them is the skull of a
male individual (BOU-VP-16/1). It shows one superficial vertical cut mark
on the corner of its right parietal, and a second shorter mark on its right
temporal. The second adult cranium, also belonging to a male, is highly frag-
mented and the lack of recovered dental, facial or basicranial parts indicates
that it may have been embedded as a calotte. The bone fragments show also
evidence of intensive bone modification, some of which are probably due
to defleshing of the skull. This could point towards utilitarian cannibalism
instead of ritual behaviour, but according to the research team, the abun-
dance of superficial markings present in the cranium are usually not found in
hominid and faunal remains processed for consumption. Thus, evidence for
mortuary practices seems to be the most likely explanation here (Clark et al.
2003: 751; Stringer 2003: 692).
The third skull belongs to a juvenile, and it displays also a series of deflesh-
ing cut marks but lacks any superficial scoring marks. The location, dimen-
sions and directions of the cut marks indicate that the defleshing manipulation
occurred after the removal of the mandible, indicating that defleshing was
intentional and deliberate. Even more interestingly, the skull lacks the entire
occipital region surrounding the foramen magnum, and the edges of this
broken region are smooth and polished, as are the parietal surfaces. This indi-
cates that the skull has been polished post-mortem. White points out that the
marks most resemble those seen on skulls handled in rituals in New Guinea.
On the basis of this, the research team argues that the polishing and inten-
tional scraping modifications evident on these skulls indicate that we are deal-
ing with some kind of mortuary practice extending well beyond the death of
the individual (Clark et al. 2003: 751;Gibbons 2003: 1641).
Steven Mithen has argued that one of the unambiguous signs for religious
activity that should be seen from the material is that religious beliefs were


not only held by individuals but also transmitted and shared by the commu-
nity. According to him, Neanderthal burials or whatever the archaic modern
humans were doing in Atapuerca show no sign of this (Mithen 1999: 1656).
Mithen means by this that while the Neanderthals clearly had some level of
capacity for symbolic thought, there is no indication that they used symbol-
ism to organize behaviour. It is this latter which is thought to be the key
criterion for behavioural modernity (Henshilwood & Marean 2003: 655).
Merlin Donald has posited a three-stage model for the evolution of mod-
ern behaviour (Donald 1991). According to him, symbolic thought is present
already in the first phase but fully modern behaviour is the product only of
the third phase, where the use of external symbolic storage allows material
culture to intervene directly with social behaviour. Christopher Henshilwood
and Curtis Marean have, therefore, argued that it is the presence of external
memory storages (EMS) that signal the presence of behavioural modernity
(Henshilwood & Marean 2003: 635). Not much can be said about the mean-
ing and contents of the religious traditions connected with the Herto skulls
(Clark et al. 2003: 751). For example, there is no way on the basis of the
skulls alone that we can determine whether the mortuary practices in Herto
also involved a belief in the supernatural. What the data suggests, however,
is that these humans used EMS and were likely to be capable of symbolically
organizing their social behaviour.

Strategies for explaining human cognitive evolution

The importance of the Herto skulls for our present understanding of the ori-
gins of symbolic behaviour is that if the data has been interpreted correctly
then the now still predominant paradigm to link the origins of symbolic
behaviour with the above-mentioned cultural explosion seems to be mistaken.
At least some human populations in Africa appear to have been capable of
fully symbolic behaviour long before that time. Most current authors believe
that behavioural modernity is a consequence of language. After all, language is
the one human capacity which seems to make it so easy to move in our mental
representations from the present to the past and the future and even to imag-
ined times and places, and to share our representations of our mental travels
with others (see Corballis & Lea 1999; Stringer & Andrews 2005: 130).
Trying to keep the anatomical and behavioural evolution of our species in
concert, Richard Klein provides the strictest model by arguing that behav-
ioural modernity came as a package through a neural mutation occurring
around 50,000 years ago providing our species with the capacity for fully
modern language (Klein 1995: 16798; Klein & Edgar 2002: 21, 146). The
evidence from Herto seems to discredit Kleins model. If symbolic behaviour

tom sjblom

is enabled through language use and our ancestors used symbols 160,000
years ago, the capacity for language must have been in place before that time.
Indeed, most authors today agree that human speech must have been in place
by the appearance of Homo sapiens 500,000 to 200,000 years ago (see e.g.
Donald 1991: 200268; Lieberman 2000: 14256; Dunbar 2004: 124). This
is why some authors have suggested that the focus of research should be
placed on the Earlier Upper Pleistocene era, or the boundary between the
Acheulian and Middle Stone Age about 250,000 years ago. The problem with
this solution is that while modern humans may have appeared at this time,
more extensive material evidence of behavioural modernity exist only from
much later stages (e.g. Foley 1998: 33947; Henshilwood & Marean 2003:
630). The same difficulty applies to efforts placing the origins of behavioural
modernity in Late Middle Pleistocene Africa around 150,000 and 200,000
years ago (e.g. H. J. Deacon 2001: 21726; Deacon & Deacon 1999). As
McBrearty and Brooks demonstrate in their article, many traits of behavioural
modernity are present at this time in African populations. However, actual
sites are very few and the evidence they provide should therefore be used care-
fully (McBrearty & Brooks 2000: 453563).
These models are all built on the assumption of a punctuated event in
which modern human behaviour originated as a package. As Christopher
Henshilwood and Curtis Marean have pointed out, however, the originating
process can also be viewed in terms of gradualism (Henshilwood & Marean
2003: 63031). According to this model, the birth of the modern mind cor-
responds to the technological changes observed through the African Middle
Stone Age, starting around 250,000 years ago and proceeding until the begin-
ning of the Later Stone Age, conventionally dated around 40,000 years ago.
According to this view there was no sudden cultural revolution, but the fea-
tures now thought to be the foundational elements of modern cognition
evolved gradually in the course of time. It now seems that there probably
wasnt any cultural revolution in the late Palaeolithic but that what we have is
simply the end-product of a long evolutionary process of cognitive and cul-
tural adaptations and their by-products. The discovery of the Herto skulls,
together with several other recent discoveries mainly from Africa, seems to
reinforce the gradualist model, although we still have to sort out the details
of how it all actually happened.
The basic claim of the gradualist model is that the cognitive capacity for
modern human behaviour was largely intact with the emergence of Homo
sapiens. The debate here is about how to explain the apparent lack of evidence
for behavioural modernity between that time and the Upper Palaeolithic era.
For example, Steven Mithen argues that the driving force here was the evolu-
tion of language. According to this model, early modern humans did not yet
possess language as we know it, but a kind of proto-language, which he refers


to as Hmmmmm (Mithen 2005: 253).3 Through a process of segmentation,

humans then started to break up the holistic phrases of the proto-language
into more flexible units that could be recombined into new phrases (see Wray
1998: 4767; 2000: 285302). According to Mithen, it is this flexibility in
language use that gives birth to the cognitive fluidity present in human cog-
nition, which in turn is the enabling trait for symbolically organized social
behaviour (Mithen 2005: 24665).
The most likely instigator of this process was changes in the natural and
social environments of early modern human communities. For example,
resource intensification or the need to increase productivity per unit of land
coupled with a decrease in the efficiency of production is customarily men-
tioned in this context (Henshilwood & Marean 2003: 6323; Mithen 2005:
2578). Naturally, the role played by language in the process is still very
much unclear. Whether Mithen is right or wrong, the point I want to make
here is that if the driving force behind the origins of behavioural modernity is
at this point largely environmental and social by nature, then it is quite possi-
ble that sporadic outbursts of behavioural modernity appear in our data long
before the latter actually becomes common and widespread among human
populations (Mithen 2005: 2612). Thus, what we might have in Herto are
signs of this kind of local outburst a kind of sporadic religion or an early
predecessor of the symbolic behaviour that came to dominate human social
life 100,000 years later.

A unique species and the third chimpanzee

Underlying the gradualist models of the origins of behavioural modernity is

the principal of cumulative cultural evolution (CCE; see Tomasello 1999:
3740). According to CCE, some cultural traditions accumulate the modi-
fications made by different individuals over time so that they become more
complex (ibid.: 37). It has been pointed out that we have evidence from
primate studies that enculturated apes (i.e. apes that have extensive human
contact) do perform tasks that are thought to be cognitively more demand-
ing than those performed by apes living in the wild. This difference cannot
be based on any biological difference between these two groups but must be
due to the social environment they are exposed to. This suggests that the apes
do not use their full cognitive potential in the wild and that there is room
for behavioural development within the present state of biological cognitive
evolution (ibid.: 346: Donald 2001: 13748).
There is no reason to think that human cognitive evolution would have
proceeded differently. Indeed, as discussed above, the present data suggests
that this is how it went with our species as well. The cognitive potential of

tom sjblom

the human mind was in place with the rise of modern humans, but it took
time and pressure from environmental challenges until the potential in ques-
tion was put into use. What makes humans different from other primates
is that, in contrast to them, our cultural traditions are cumulative. That
is, we transmit our innovations socially, across succeeding generations and
thus prevent backward slippage (Tomasello 1999: 5). This is the bio-cultural
mechanism enabling the accumulation of behavioural modernity among our
As Peter Richerson and Robert Boyd have recently argued, it now seems
clear that culture transformed human cognitive evolution (Richerson & Boyd
2005). We live in the cognitive niche, and the driving force of cognitive evo-
lution is our efforts to increase our fitness in terms of survival. Our cognitive
capacities have evolved as they are in order to enable us to cope with our
environment and to construct and expand our own niche (see Odling-Smee et
al. 2003). As most of the behavioural modernity is manifested through what
we refer to as culture, it is no wonder that naturalistic explanations of the
origins of behavioural modernity have focused on tracing the cognitive traits
that make us unique among other unique species in the world (Foley 1987;
Alexander 1990; Donald 2001). Cognitive fluidity and the ratchet effect are
gadgets which function as the final cognitive traits triggering the possibility
of fully modern behaviour. However, this is not the whole story.
Nothing in our behaviour makes sense without realizing that we are a
biological species and that as such we have a long evolutionary history (see
Foley 1987: xxi; Richerson & Boyd 2005: 23757). The cognitive capaci-
ties we have are products of a long evolutionary process that can be fol-
lowed all the way back to the dawn of time (see Dawkins 2004). Evolution
has no foresight; it proceeds by selecting from what already is available and
builds on that. Before there was behavioural modernity, there were preceding
modes of behaviour, and these preceding modes or at least the cognitive
systems producing them are still in operation in our modern minds (e.g.
Dawkins 1986; Mayr 2001). Indeed, we continue to share about 98 per cent
of our genetic program with our closest primate relatives, the chimpanzees
and the bonobos; until about six or seven million years ago, we also shared the
same ancestry, so we also share the same cognitive basis until that time (see
Diamond 1992: 1531; Stringer & Andrews 2005: 80181).
How our primate background our innate ape affects our modern
behaviour, and cognitive representations is the task dealt by an approach that
I following Jared Diamond refer to as the third chimpanzee approach to
human behaviour (e.g. Diamond 1992; De Waal 2002, 2005). This approach
deals with the deep roots of human cognition the cognitive pre-adaptations
and how they are expressed in our modern minds. One can of course ask how
far back in evolution we should go in order to explain the cognitive basis of


behavioural modernity. Is it really necessary for us to explore the deep primate

roots of our cognition in order to understand social behaviour? The answer
to this is that sometimes it is and sometimes it isnt. Its all about finding the
proper level of explanation to the question at hand, and sometimes the best
level for explaining modern behaviour is found in the deep roots of human
cognition. In what follows, I will try to demonstrate this by examining the
origins of what has been referred to as the cult of the serpent or snake sym-
bolism in religious representations (see Mundukur 1983).

The old enemy, or why snakes are good symbols

Animals have an important place among human symbolic representations

(see Shepard 1978; Baker 1993). Indeed, according to Stanley Walens, animal
representations are the most fundamental symbols available:

Wherever they appear, animal symbols are used to convey the

deepest and most abstruse dimensions of human existence. They
are symbols of core values and categories, representations of the
most fundamental ideas and images of a culture. As core sym-
bols, they are multivalent, complex, antimonic, used simultane-
ously to capture and display many different images and meanings
at many different levels. As core symbols, they also serve to link
other domains of symbolic discourse, creating juxtapositions and
contrasts of images from which people derive meaning and from
which they generate narrative forms. The natural realm of animals
is an important part of the way in which people project their
knowledge and experience through symbolic discourse.
(Walens 1987: 291)

The importance of animal symbolism has customarily been explained by cul-

tural transmission. According to this view the cultural hypothesis of animal
symbolism we use them because we inherit the symbols from our ances-
tors. For example, snakes are a common symbol for evil in modern Christian
thinking. According to the cultural hypothesis, the snake has become the
symbol for evil because it is equated with Satan in the Bible and in early
Christian mythology. Snake symbolism in the Bible, again, has its origins in
the common Near Eastern mythological motif of ophidian chaos monsters
that battled with creator gods (see Fontenrose 1959; Beal 2002: 8081).
Undoubtedly we can, and often do, inherit our symbols from earlier gen-
erations, but this cannot be the whole story because it does not explain why
a certain animal has been chosen as symbol in the first place. This takes us

tom sjblom

to the second possibility, which is referred to here as the ecological hypothesis.

The ecological hypothesis is most famously defended by the anthropologist
Claude Lvi-Strauss. According to him, animals emerge in human symbolic
communication simply because they play an important role in our empirical
environment (Lvi-Strauss 1969: 89).
At the outset, the ecological hypothesis seems to gain strong support from
ethnographic data. Animals used in religious symbolism vary widely, and local
fauna is in most cases preferred over other types of animal representations.
Thus felids, for example, are important as symbols in northeastern and east-
ern Asia, the home of the tiger; in the Near East and Africa, the home of the
lion; and in Meso- and South America, the home of the jaguar. Bears, on the
other hand, are essentially Northern Hemispheric animals typical of Europe,
Asia and America, and it is to this geographical area that bear symbolism is
confined to a large extent. Even more restricted is the symbolic importance
of bats. While different species of bats can be found around the world, bats
have very little symbolic significance outside Mesoamerica and a few other
parts of the tropical and subtropical New World, where particularly sizeable
species of bats live (Mundukur 1983: 10913, 156).4
In this framework, the symbolic link between the snake and evil in the
Near Eastern traditions can be explained through the fact that many ven-
omous snakes inhabit this area, and even in modern times, snake bites are
one of the leading sources of animal-caused deaths. Thus, it would be only
natural that Near Eastern human populations should feel fear and respect
towards snakes and transmit this fear to their symbols as well (ibid.: 44).
Nevertheless, what the ecological hypothesis cannot explain is how snakes
retain their symbolic power and emotional salience even in cultures and envi-
ronments where snakes do not belong to the local fauna, or where they do
not pose a great danger to humans even if they can be found (see Stutesman
2005).5 Moreover, people appear to be more afraid of snakes than they are
of threats that are much more immediate to them. For example, in a survey
conducted by Adah Maurer among 500 American schoolchildren, the snake
was the most unpopular of all animals. The children did not show any fear
towards street traffic or germs, both of which they had been taught to be care-
ful about. Instead 30 per cent of them feared snakes and almost 26 per cent
lions (Maurer 1965: 26577).
This suggests that the ecological hypothesis provides only a proximate
explanation to the question. In order to reveal the ultimate cause for snake
fear, we must turn to what can be called the evolutionary hypothesis. According
to this view, we have an innate fear of snakes which evolved when snakes still
presented a major threat to our species. Lynne Isbell points out that snakes
were the first of the modern predators of primates. Accordingly, the pres-
sure from predation has been so strong that it has influenced the cognitive


evolution of our species (Isbell 2006: 135). Our and other primate minds
have gained a preparedness to create strong emotional responses towards
snakes so that we can avoid their attacks without having to reflect on our
acts (see Mundukur 1983: 21822; Cook et al. 1985: 591610; Jones 2002:
445; Hauser 2000: 138; hman & Mineka 2003: 56). In other words, we
seem to have an innate fear of snakes. According to the evolutionary hypoth-
esis, it is this innate fear of snakes that make them useful as symbols.

Awing snakes

Fear is often conceived in negative terms, but from an evolutionary perspec-

tive, it is a highly constructive emotion that enables us to guide our attention
and respond swiftly to environmental stimuli (Fredrickson 1998: 300319;
Levenson 1999: 494). My claim here is that snakes are used in human sym-
bolic communication not to evoke fear but to evoke attention. Different
fear-related stimuli, like snakes, do not have to have anything to do with
theological doctrines or religious beliefs per se. What such stimuli are designed
to do is to evoke our attention and make us respond to what we are experi-
encing and learning. The automatic responses evoked by such stimuli make
them very usable for this purpose, as this means that we do not even have
to be exposed to snakes or snake-like images in the empirical world in order
to fear them. It is enough to listen to narratives about snakes or simply hear
the word snake to trigger the fear module (e.g. Christophe & Rim 1997:
3754; Curci & Bellelli 2004: 881900). This is enough to create an emotion
evoking a representation of a snake in our minds. As emotional relevance is
so important in the transmission of religious traditions, it is no wonder then
that they prefer to communicate through narratives rather than abstractions.
Moreover, stimulation of the fear module does not always lead us to actu-
ally feel frightened. It alerts us to a possible danger, but if we quickly cognize
that there is no immediate danger, we tend to feel awe instead of fear. Very
little empirical research has been done on awe. According to Richard Lazarus,
it can be described as an ambiguous negative state, rather than an emotion,
that combines fright and amazement. However, he points out that we seem
to speak of awe in more than one sense, and because of its ambiguity, it often
connects with positive emotional states as well (Lazarus 1991: 83, 2389,
295). The environmental cues for triggering awe are also heterogeneous,
including experiences of natural beauty and exemplary or exceptional human
actions or abilities. Nevertheless, in its most basic form, awe seems to be
connected with conditions where we strive to believe and trust in something
that our basic instincts tell us to fear. Indeed, this is why awe appears to be
so often discussed in a religious context as the proper and desirable religious

tom sjblom

emotion signalling the presence of counterintuitive agents (see James [1901]

1960: 8086; Lazarus 1991: 83; Haidt 2003: 863).
Interestingly, snakes have their part to play in the evocation of awe, and
Balaji Mundukur argues that it is in this aspect that they should be under-
stood in religious communication (Mundukur 1983: 5573). For example, in
pre-Christian times, the Baltic peoples were generally sympathetic to snakes
and even venerated some of them as house-snakes, symbols of prosperity and
human fertility. Nevertheless, snakes seem to have been feared, and it was a
taboo to refer to them by their true name. A similar pattern can be found in
Egypt, where the patronage of Nehebkau, an ophidian god, was sometimes
sought. Mythologically speaking, Nehebkau was essentially a sinister divin-
ity, considered to be one of the forces of chaos. The respect he received was
thus based on awe evoked by his great powers rather than his good and loving
character (ibid.: 6064).
An even more illuminating example of the awe-inducing behaviour con-
nected with snakes and the activation of the fear module is the present-day
Serpent Handlers, or the Church of God with Signs Following, which is an
umbrella term for a loosely organized group of small Pentecostal churches
from the southeastern United States (henceforth holiness churches). These
groups differ widely from each other in a variety of theological issues, but what
they all share is the handling of serpents (see e.g. Burton 1993; Kimbrough
1995).6 This practice derives from a strict literal reading of Mark 16, where
it is written And these signs shall follow them that believe; in my name shall
they cast out devils; they shall speak with tongues; they shall take up serpents;
and if they drink any deadly thing, it shall not hurt them; they shall lay hands
on the sick and they shall recover (Mk 16:17-18).7
A typical snake-handling meeting usually consists of singing and preach-
ing. The snakes are held in boxes in front of the church by the altar. The idea
is that only those who are called forth by God, who feel that they have been
anointed by his power, open the box lids and lift the snakes high into the
air, or take them from other participants. It is possible for one participant to
hold more than one snake at the same time and let them slither around their
bodies. The experience of anointment is described by the snake handlers in
various ways, but it certainly entails an altered state of consciousness. Indeed,
when God is felt to move upon the body, the serpent handler loses personal
concern for danger and is fully prepared to do whatever God wants him to
do. In addition to snake handling, such acts of faith might include drinking
poison or exposing oneself to fire.8
Because of its extremeness, serpent handling has gained the attention
of both popular media and the academic community. As pointed out by
Williamson and Pollio, most analyses of snake handling have been treating
it as being pathological (Williamson & Pollio 1999: 203). For example, in


his classic study, anthropologist Weston La Barre, following his training in

psychoanalysis, argued that snake-handling is an attempt to resolve the sex-
ual conflicts of childhood through ritual activity (La Barre 1962). The same
theme of fighting against repressed oedipal conflict and guilt arising from
it has been the preferred psychological explanation for snake-handling ever
since (see also Hood & Kimbrough 1995: 31122).
The psychoanalytical interpretation of serpent handling derives from the
assumption that the snake is a phallic (and sometimes also a vaginal) sym-
bol, and it presupposes that our sexuality is our most fundamental area of
concern. Indeed, most symbols seem to be somehow related to sexuality in
psychoanalytical theory (see Mundukur 1983: 26270; Hood & Kimbrough
1995: 31517). As pointed out by Mundukur, however, snake symbolism in
reality is not easily connected with any one theme, and deducing a hidden
meaning of cultic practices or religious symbols only based on the image used
is nothing more than pure guess-work (Mundukur 1983: 270). Indeed, a
much simpler psychological explanation for snake-handling is that overcom-
ing the innate preparedness we have for fearing snakes by handling them in
a ritual context, ignites in the participants an emotional trust in the spiritual
power of God. As suggested by Williamson and Pollio, snake-handling may
also provide church members with an emotional assurance that they have the
fate that is required to follow Gods will.
The overcoming of fear thus leads to an experience of joy unspeakable and
full of glory, as one of the members of the holiness churches describes it (see
Williamson & Pollio 1999: 214). Indeed, what seems to happen on the emo-
tional level is that the initial fright turns to amazement when one realizes that
the snake you are holding is not harming you. As the cause of this is thought
to be God, the awe aroused by the achievement is directed at him. Finally, the
state of awe is transferred into joy which provides a positive emotional mark
on the experience as a whole, and encourages participants to take part in the
snake handling rituals in the future as well.


In this chapter, I have defended the gradualist hypothesis of the origins of

behavioural modernity. Religion is one of the manifestations of behavioural
modernity, and the origins of the former thus provide insights for explain-
ing the origins of the latter. Joining a growing number of scholars, I chose
symbolically organized behaviour as the fundamental trait for enabling reli-
gious behaviour. Based on the evidence from Herto and some other African
sites, I argued that the cognitive capacity for religion was in place with the
emergence of anatomically modern humans some time between 500,000 and

tom sjblom

200,000 years ago. The reason why we do not have immediate signs of reli-
gious activity from that time is that cultural and environmental cues were
required to activate the capacities available. This was a cumulative process that
came to fulfillment only after 60,000 years ago, although sporadic outbursts
of symbolic behaviour, like in Herto, can be found from even earlier times.
Most of the models for the origins of symbolic behaviour have focused on
what is considered to be the triggering trait of behavioural modernity. Thus,
Steven Mithen argues for cognitive fluidity as the key for behavioural moder-
nity and Michael Tomasello defends his case for the ratchet effect. Indeed, I
recognize the importance of both of these. However, tracing and defining
such triggering traits for symbolic behaviour is like describing how a wall
protects people standing behind it by focusing only on describing the foun-
dation stone. Symbolic behaviour is much more than cognitive fluidity and
transmitting traditions. It is true that without cognitive fluidity we prob-
ably would not be capable of creating and understanding symbols. Still, in
order for a symbol to be effective, it is not enough that we understand it to
be a symbol: it must also be capable of producing emotional feedback in us.
Symbols work because they evoke emotional signals in us; without emotions
or affect they would have no meaning to us and without meaning they would
not be symbols (Greenspan & Shanker 2004: 25).
In other words, symbolic behaviour is based on emotional communication,
and it is to the origins of emotional communication that we must turn to in
order to explain this aspect of the origins of religious behaviour. Emotional
communication evolved early in our ancestors. Indeed, it is one of the traits
that seem to link us with other animal species as there seem to be some
striking similarities in the emotional communication between us and other
species (see LeDoux 1998: 10712; Darwin 1999; Greenspan & Shanker
2004; Mithen 2005). In the second part of this chapter, I demonstrated how
knowledge of the origins of our emotional responses can sometimes prove to
be highly informative for our understanding of symbolic communication.
For example, snakes are among the most common symbolic motifs in reli-
gious communication, and they are found even in traditions and geographical
areas where snakes do not present any great threat to the local human popula-
tions. Different cultural and ecological explanations provided to explain our
fascination with snakes have all proven to be problematic. However, turning
to evolutionary explanations and the role played by the snake as the main
predator of our primate ancestors seems to provide the missing link required
to understand snake symbolism. Fear and awe have long been considered to
be the paradigmatic religious emotions (see Proudfoot 1985; Guthrie 1993;
Fuller 2006: 1458). My case study with snakes supports this general idea,
although it also shows that the situation is much more complicated than what
has been noted before. I hope to be able to deal with the arising issues in my


future research, but here I want to simply argue that emotional communica-
tion is at the heart of religious behaviour, and if we want to discuss the origins
of religion it is with the origins of emotional communication that we must
deal. In order to do this, we must be able and ready to explore the deep roots
of our primate cognition.


1. Tinbergen points out that the limitation to closely allied forms is necessary because it
is only here that conclusions about common descent can be drawn with any degree of
probability (Tinbergen 1963: 428). For example, Jared Diamond has argued that even
art a type of social behaviour often thought to be uniquely human must have some
animal precursors as demonstrated by the behaviour of the representatives of some other
species like bowerbirds, chimpanzees and elephants (Diamond 1992: 16879). One
reason for explaining the behaviour in question in both bowerbirds and primates, at
least, is that in both birds and primates, a visual specialization evolved that makes their
cognition rely on the visual system at the cost of, for example, olfaction. However, using
the behaviour of bowerbirds and elephants to explain cognitive adaptations in humans
is risky, as although numerous cases of convergence in unrelated taxa clearly do exist,
it is often very difficult to say if these analogies are due to similar causes. Therefore, we
must be very careful (as Diamond is) to what extent we can use the art of bowerbirds as
explanatory evidence for why and how we humans produce art (see Isbell 2006: 1619).
2. The situation is analogous with the notion of the environment of evolutionary adapted-
ness (EEA) used by evolutionary psychologists. EEA refers to those aspects of ancestral
environment that were relevant to the evolution, development and functioning of an
organisms adaptations (Hagen 2005: 152). As the changes in the modern world are
so fast in comparison to biological evolution that it has no real selective force in this
respect, the EEA in the human case is often placed with the hunter-gatherer life-style
of the Pleistocene era often claimed to cover over 99 per cent of the evolutionary his-
tory of our species. As Kevin Laland and Gillian Brown point out, this should not be
taken to mean that the human cognitive make-up originates in its entirety from the
Pleistocene era. No new environment is totally novel, and many aspects of the modern
world resemble those of the Pleistocene, and aspects of the Pleistocene environment
resemble those of earlier eras. Thus, many of our behavioural and cognitive adaptations
may have evolved much earlier, but they remain intact in our biological and cognitive
make-up even today, if they do not turn out to be fitness-reducing in the present envi-
ronment (see Laland & Brown 2002: 1612, 17782; Hagen 2005: 1534).
3. This stands for holistic, manipulative, multi-modal, musical and mimetic.
4. This does not mean that bats are totally without symbolic significance outside this area.
For example, in the Bible, bats are referred to as beings of the Netherworld and denizens
of the dark caves, where people abandoning God could turn to in their worship (Is.
2:20-21). In later tradition they were actually referred to as Devils birds and equated
with demons (Lawrence 1993: 32631). Still, relatively speaking bats have only a minor
role to play in Christian symbolism.
5. Only birds can compete with snakes when it comes to the extent of their geographical
distribution. Snakes can be found everywhere between the Arctic Circle and the south-
ernmost tips of the Southern Hemisphere Antarctica excluded. However, notable

tom sjblom

places where snakes are totally absent or very rare include Ireland, Madagascar, New
Zealand, most of the West Indies, Philippines, Vancouver Island (British Columbia)
and Eastern Samoa (see Mundukur 1983: 4042). Also, it should be kept in mind that
most snakes are not poisonous and in many areas like in Northern Europe snakes
do not pose a great threat for humans, even though on a worldwide scale they belong to
the most dangerous animal threats to humans (Mundukur 1983: 4055; White 2000:
6. The snake-handling sects all derive from the example set by the Pentecostal preacher
George Went Hensley in 1909 or 1910. The first holiness church was established by
Hensley probably in the 1920s, and sister churches were soon established throughout
the Appalachian rural areas. The exact number of church members today is unknown
due to the autonomy of each individual group, but the estimates range usually between
1000 and 2000 members (Melton 1996: 636).
7. As pointed out, however, by Hood and Kimbrough, it is a common mistake to think
that holiness churches rely only on this one text from Mark to legitimate their practice.
In fact they do refer to several other passages from the Bible as well, although the text
from Mark is the most important among them (see Hood & Kimbrough 1995: 313).
8. Several members of the holiness churches have been bitten numerous times. Hensley
himself reportedly survived hundreds of bites before the fatal one in 1955. During the
twentieth century, 75 people have been reported to have died from handling snakes
during religious services, and five other deaths followed from drinking strychnine (see
Hood & Kimbrough 1995: 316, 319; Williamson & Pollio 1999: 213).

Alexander, R. 1990. How Did Humans Evolve? Reflections on the Uniquely Unique Species. Ann
Arbor, MI: University of Michigan Press.
Allman, J. 2000. Evolving Brains. New York: Scientific American Library.
Arsuaga, J. L. 2003. The Neanderthal Necklace: In Search of the First Thinkers. Chichester: John
Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. New York: Oxford
University Press.
Baker, S. 1993. Picturing the Beast: Animals, Identity and Representation. Urbana, IL: University
of Illinois Press.
Beal, T. K. 2002. Religion and its Monsters. New York: Routledge.
Boyer, P. 1994. The Naturalness of Religious Ideas: A Cognitive Theory of Religion. Berkeley, CA:
University of California Press.
Boyer, P. 2001. Religion explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Burton, T. 1993. Serpent Handling Believers. Knoxville, TN: University of Tennessee Press.
Campbell, B. 1998. Human Evolution, 4th edn. New York: Aldine de Gruyter.
Capps, W. H. 1995. Religious Studies: The Making of a Discipline. Minneapolis, MN: Fortress
Chase, P. C. & H. L. Dibble 1987. Middle Palaeolithic Symbolism: A Review of Current
Evidence and Interpretation. Journal of Anthropological Archaeology 6: 26396.
Christophe, V. & B. Rim 1997. Exposure to the Social Sharing of Emotion: Emotional
Impact, Listener, Responses and Secondary Social Sharing. European Journal of Social
Psychology 27: 3754.


Clark, J. D. et al. 2003. Stratigraphic, Chronological and Behavioural Contexts of Pleistocene

Homo Sapiens from Middle Awash, Ethiopia. Nature 423: 74752.
Cook, M., S. Mineka, B. Wolkenstein & K. Laitsch 1985. Observational Conditioning of
Snake Fear in Unrelated Rhesus Monkeys. Journal of Abnormal Psychology 94(4): 591610.
Corballis, M. C. & S. E. G. Lea (ed.) 1999. The Descent of Mind: Psychological Perspectives on
Hominid Evolution. Oxford: Oxford University Press.
Curci, A. & G. Bellelli 2004. Cognitive and Social Consequences of Exposure to Emotional
Narratives: Two Studies on Secondary Social Sharing of Emotions. Cognition and Emotion
18(7): 881900.
Darwin, C. 1999. The Expression of the Emotions in Man and Animals. London: Fontana.
Dawkins, R. 1986. The Blind Watchmaker. London: Penguin.
Dawkins, R. 2004. The Ancestors Tale. London: Weidenfeld & Nicolson.
Deacon, H. J. 2001. Modern Human Emergence: An African Archaeological Perspective.
In Humanity from African Naissance to Coming Millennia: Colloquia in Human Biology and
Palaeoanthropology, P. V. Thomas, M. A. Raath, J. Maggi-Cecchi & G. A. Doyle (eds),
21726. Florence: University of Florence Press.
Deacon, H. J. & J. Deacon 1999. Human Beginnings in South Africa. Cape Town: David Philp.
Deacon, T. 1997. The Symbolic Species: The Co-evolution of Language and the Brain. New York:
W. W. Norton.
Dennett, D. C. 2006. Breaking the Spell: Religion as a Natural Phenomenon. London: Allen Lane.
De Waal, F. 2002. Tree of Origin: What Primate Behavior Can Tell Us About Human Social
Evolution. Cambridge, MA: Harvard University Press.
De Waal, F. 2005. Our Inner Ape: The Best and Worst of Human Nature. London: Granta Books.
Diamond, J. 1992. The Third Chimpanzee: The Evolution and Future of the Human Animal.
New York: Harper Perennial.
Donald, M. 1991. Origins of the Modern Mind: Three Stages in the Evolution of Culture and
Cognition. Cambridge, MA: Harvard University Press.
Donald, M. 2001. A Mind So Rare: The Evolution of Human Consciousness. New York: W. W.
Dunbar, R. 2004. The Human Story: A New History of Mankinds Evolution. London: Faber
& Faber.
Evans-Pritchard, E. E. 1965. Theories of Primitive Religion. Oxford: Clarendon Press.
Foley, R. 1987. Another Unique Species: Patterns in Human Evolutionary Ecology. New York:
Foley, R. 1998. The Context of Human Genetic Evolution. Genome Research 8: 33947.
Fontenrose, J. 1959. Python: A Study of Delphic Myth and its Origins. Berkeley, CA: University
of California Press.
Fredrickson, B. L. 1998. What Good are Positive Emotions?. Review of General Psychology
2(3): 30019.
Fuller, R. C. 2006. Wonder: From Emotion to Spirituality. Chapel Hill, NC: The University of
North Carolina Press.
Gibbons, A. 2003. Oldest Members of Homo Sapiens Discovered in Africa. Science 300:
Gothni, R. (ed.) 2005. How to Do Comparative Religion? Three Ways, Many Goals. Berlin:
Walter de Gruyter.
Greenspan, S. I. & S. G. Shanker 2004. The First Idea: How Symbols, Language, and Intelligence
Evolved from Our Primate Ancestors to Modern Humans. Cambridge, MA: Da Capo Press.
Guthrie, S. 1993. Faces in the Clouds: A New Theory of Religion. New York: Oxford University

tom sjblom

Hagen, E. H. 2005. Controversial Issues in Evolutionary Psychology. In The Handbook of

Evolutionary Psychology, D. M. Buss (ed.), 14573. Hoboken, NJ: John Wiley.
Haidt, J. 2003. The Moral Emotions. In Handbook of Affective Sciences, R. J. Davidson, Klaus
R. Scherer & H. Hill Goldsmith (eds), 85270. Oxford: Oxford University Press.
Hauser, M. D. 2000. Wild Minds: What Animals Really Think. New York: Henry Holt.
Hayden, B. 2003. Shamans, Sorcerers and Saints: A Prehistory of Religion. Washington, DC:
Smithsonian Books.
Henshilwood, C. S. & C. W. Marean 2003. The Origin of Modern Behavior: Critique of the
Models and Their Test Implications. Current Anthropology 44(5): 62751.
Holyoak, K. & P. Thagard 1995. Mental Leaps: Analogy in Creative Thought. Cambridge, MA:
MIT Press.
Hood, R. W. Jr. & D. L. Kimbrough 1995. Serpent-Handling Holiness Sects: Theoretical
Considerations. Journal for the Scientific Study of Religion 34(3): 31122.
Isbell, L. A. 2006. Snakes as Agents of Evolutionary Change in Primate Brains. Journal of
Human Evolution 51: 135.
James, W. [1901] 1960. The Varieties of Religious Experience. London: Collins.
Jones, D. E. 2002. An Instinct for Dragons. New York: Routledge.
Kimbrough, D. L. 1995. Taking Up the Serpents: Snake Handling Believers of Eastern Kentucky.
Chapel Hill, NC: University of North Carolina Press.
Klein, R. 1995. Anatomy, Behavior, and Modern Human Origins. Journal of World Prehistory
9(2): 16798.
Klein, R. G. & B. Edgar 2002. The Dawn of Human Culture. New York: John Wiley.
La Barre, W. 1962. They Shall Take Up Serpents: Psychology of the Southern Snake-Handling Cult.
Prospect Heights, IL: Waveland.
Laland, K. & G. Brown 2002. Sense and Nonsense: Evolutionary Perspectives on Human
Behaviour. Oxford: Oxford University Press.
Lawrence, E. A. 1993. The Sacred Bee, The Filthy Pig, and the Bat out of Hell: Animal
Symbolism as Cognitive Biophilia. In The Biophilia Hypothesis, S. R. Kellert & E. O.
Wilson (eds), 30141. Washington, DC: Island Press.
Lazarus, R. S. 1991. Emotion and Adaptation. New York: Oxford University Press.
LeDoux, J. 1998. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. New
York: Phoenix.
Levenson, R. W. 1999. The Intrapersonal Functions of Emotion. Cognition and Emotion
13(5): 481504.
Lvi-Strauss, C. 1969. Totemism. Boston, MA: Beacon Press.
Lewis-Williams, D. 2002. The Mind in the Cave: Consciousness and the Origins of Art. London:
Thames & Hudson.
Lieberman, P. 2000. Human Language and Our Reptilian Brain. Cambridge, MA: Harvard
University Press.
Marett, R. R. 1909. The Threshold of Religion. London: Methuen.
Maurer, A. 1965. What Children Fear. Journal of Genetic Psychology 106: 25577.
Mayr, E. 2001. What Evolution Is. London: Phoenix.
McBrearty, S. & A. Brooks 2000. The Revolution That Wasnt: A New Interpretation of the
Origin of Modern Human Behaviour. Journal of Human Evolution 39: 45363.
McClenon, J. 2002. Wondrous Healing: Shamanism, Human Evolution, and the Origin of
Religion. DeKalb, IL: Northern Illinois University Press.
Melton, L. G. 1996. Encyclopedia of American Religions, 5th edn. Washington, DC: Gale
Mithen, S. 1996. The Prehistory of the Mind: A Search for the Origins of Art, Religion and Science.
London: Phoenix.


Mithen, S. 1999. Symbolism and the Supernatural. In The Evolution of Culture, R. Dunbar,
C. Knight & C. Power (eds), 14769. Edinburgh: Edinburgh University Press.
Mithen, S. 2005. The Singing Neanderthals: The Origins of Music, Language, Mind and Body.
London: Weidenfeld & Nicolson.
Mundukur, B. 1983. The Cult of the Serpent: An Interdisciplinary Survey of its Manifestations
and Origins. Albany: SUNY Press.
Odling-Smee, F. J., K. N. Laland & M. W. Feldman 2003. Niche Construction: The Neglected
Process in Evolution. Princeton, NJ: Princeton University Press.
hman, A. & S. Mineka 2003. The Malicious Serpent: Snakes as a Prototypical Stimulus for
an Evolved Module of Fear. Current Directions in Psychological Science 12(1): 59.
Outhwaite, W. 1975. Understanding Social Life: The Method Called Verstehen. London: George
Allen & Unwin.
Pearson, M. P. 1999. The Archaeology of Death and Burial. Stroud: Sutton Publishing.
Pfeiffer, J. 1982. The Creative Explosion: An Inquiry into the Origins of Art and Religion. New
York: Harper & Row.
Pratchett, T. 1997. Hogfather. London: Corgi.
Proudfoot, W. 1985. Religious Experience. Berkeley, CA: University of California Press.
Richerson, P. J. & R. Boyd 2005. Not by Genes Alone: How Culture Transformed Human Evolu-
tion. Chicago, IL: University of Chicago Press.
Shepard, P. 1978. Thinking Animals: Animals and the Development of Human Intelligence.
Athens, GA: University of Georgia Press.
Strauss, C. & N. Quinn 1997. A Cognitive Theory of Cultural Meaning. Cambridge: Cambridge
University Press.
Stringer, C. 2003. Out of Ethiopia. Nature 423: 6925.
Stringer, C. & P. Andrews (eds) 2005. The Complete World of Human Evolution. London:
Thames & Hudson.
Stutesman, D. 2005. Snake. London: Reaktion Books.
Taylor, T. 2002. The Buried Soul: How Humans Invented Death. London: Fourth Estate.
Tinbergen, N. 1963. On Aims and Methods of Ethology. Zeitschrift Fr Tierpsychologie
20(4): 41033.
Tomasello, M. 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard
University Press.
Trinkaus, E. 2005. Early Modern Humans. Annual Review of Anthropology 34: 20730.
Trompf, G. 2005. In Search of Origins: The Beginnings of Religions in Western Theory and
Archaeological Practice. Elgin: New Dawn.
Walens, S. 1987. Animals. In The Encyclopedia of Religion 1, M. Eliade (ed.), 2916. New
York: Macmillan.
Walker, A. & P. Shipman 1996. The Wisdom of the Bones: In Search of Human Origins. New
York: Vintage Books.
Watts, I. 1999. The Origin of Symbolic Culture. In The Evolution of Culture, R. Dunbar
(ed.), C. Knight & C. Power 11346. Edinburgh: Edinburgh University Press.
White, J. 2000. Bites and Stings from Venomous Animals: A Global Overview. Therapeutic
Drug Monitoring 22(1): 658.
White, T. D., B Asfaw, D. DeGusta, H. Gilbert, G. D. Richards, G. Suwa & F. C. Howell
2003. Pleistocene Homo Sapiens from Middle Awash, Ethiopia. Nature 423: 7427.
Widengren, G. 1946. Religionens ursprung. Stockholm: Svenska Kyrkans Diakonistyrelsens
Williamson, W. P. & H. R. Pollio 1999. The Phenomenology of Religious Serpent Handling:
A Rationale and Thematic Study of Extemporaneous Sermons. Journal for the Scientific
Study of Religions 38(2): 20318.

tom sjblom

Wood, B. 2005. Human Evolution: A Very Short Introduction. Oxford: Oxford University Press.
Wray, A. 1998. Protolanguage as a Holistic System for Social Interaction. Language and
Communication 18: 4767.
Wray, A. 2000. Holistic Utterances in Protolanguage: The Link from Primates to Humans.
In The Evolutionary Emergence of Language: Social Function and the Origins of Linguistic
Form, C. Knight, M. Studdert-Kennedy & J. R. Hurford (eds), 285302. Cambridge:
Cambridge University Press.

Cognition and meaning
Jeppe Sinding Jensen


There are two general views of religion shared by the greater part of the public
in the modern world. The first is that, despite modernity, science, technology
and so on, religion can be understood as something fundamentally human,
social and meaningful although it may contain dubious claims to truths about
the contents of the world. So conceived, religions are basically clusters of
meaning, analogous to other forms of cultural expressions. The second view
which is very prevalent in public discourse on the subject of religion is to
think of religion along more psychological lines: Religion and religiosity in
particular, is a matter of feelings and conviction, a means of explaining exist-
ence or of orienting ones life. Consequently, religion is understood either
as something related to meaning on the one hand or as a psychological or
mental phenomenon on the other (cf. e.g. Jensen 1999, 2004).
The following remarks are intended as a kind of road map to the difficul-
ties of such trains of thought. The first problematic idea is that it is linguisti-
cally unacceptable to treat meaning as a mental phenomenon but that is not
the sort of problem that disturbs the general public or violates popular epis-
temology. Linguists generally treat meaning as a non-mental phenomenon,
and to understand this demands a certain degree of philosophical understand-
ing, because meaning is not in the head, it is somewhere else.1 The second
difficulty is that most theoretical approaches to the study of religion based
on cognitive science perceive the mental (or cognitive) as processes unrelated
to issues concerning meaning. This involves a number of misunderstandings
which are of profound importance for the future of the study of religion
including the cognitive science-based ones. How it has come to be this way
is what I will discuss below.
Most of the work in the cognitive science of religion does not share
these ideas about the relations between culture and meaning. Many of the

jeppe sinding jensen

contributors to the new cognitive science of religion (let us call them cogni-
tivists) have worked to eliminate the standard culturalist paradigm, which
has generally been operating on the social and cultural levels. It has been
widely practised by scholars of religion and anthropologists who thought that
it was on the levels of culture or society that they could best interpret, explain
or justify human behaviour.2 The resulting incongruity has led to a number
of problems which are profoundly epistemic, and which I will explain below
because they are important for the argument.
The main dogma in a large part of cognitive approaches to religion is cul-
tural reductionism or cultural eliminativism. These are theoretical positions
which explicitly contend that neither culture nor meaning are anything but
(the catchword of reductionism) aggregate material phenomena, i.e. mental
processes which take place in the individual or the more radical position
that culture does not exist (the slogan of eliminativism). Culture is thus
effortlessly reduced and dissolved into individual psychology or biology.3
Significantly, their interests do not focus on conscious mental processes. For
instance, the arguments or reasons which the agents themselves have or pro-
vide are basically assumed to be justifications produced after the fact. Their
interests concentrate on the unconscious cognitive processing of informa-
tion and the related mechanisms and on what they produce. It is therefore
an inside-out programme concerning the cognitive levels.4 Cultural levels,
on the contrary, include questions of language, norms and values and these
are of little interest to cognitivists. The culturalist assumption that justi-
fications provided by the sources (agents) can be analysed and understood
in terms of reasons and their place and function in semantic universes is
elegantly neglected. Cognitivists thus seem to take for granted that an agents
own explanations of thoughts and actions cannot be examined scientifically
(i.e. in controlled experiments). Thus, it is not adequately scientific to study
reasons or justifications.5 This understanding of explanation and causal-
ity results from a view of scientific practice which is highly influenced by
the natural sciences. In some cases it appears as almost scientistic, that is,
an excessive admiration of the natural sciences. This attitude may then lead
to a view of explanations as legitimate only when they are reductionist or
causal and when the methodological requirement is to explain upwards
from lower levels to higher ones (e.g. from a neurological level to a cognitive
level). Higher-level phenomena, such as for instance conscious mental activ-
ity, are not considered to be able to explain anything at lower levels because
these high-level phenomena are nothing but physical brain processes, and
thoughts are considered as being unable to influence the material founda-
tions from which they originate. This conception is not new; in philosophy
it traditionally goes under the name of epiphenomenalism. In brief, the
epiphenomenalist position holds that socio-cultural products such as reli-

cognition and meaning

gion, ideology and various uses of symbolism are created by physical brain
processes, but the socio-cultural products cannot have a reciprocal influence
on those more basic processes. Religious beliefs, for example, do not have
any causal or scientifically verifiable (or falsifiable) influence on our cognitive
processes. This frequently results in the opinion that even if religion, culture,
economy and other social facts can be discussed in nominal or heuristic terms
as abstractions or conceptual generalizations, then they cannot have causal
effects on anything: an abstraction, such as religion or culture, cannot be
a cause and is thus of no scientific interest to most cognitivists.

The truncated perspective

The Canadian psychologist Merlin Donald (among others) characterizes

those pursuing this type of cognitive research as hardliners, who share an
uncompromising belief in the irrelevance of the conscious mind and the illu-
sory nature of free will (Donald 2001: 1). The change in limitations in the
research perspectives in early cognitive research away from a more holistic
idea of connections between culture and psyche and on towards a more indi-
vidualistic and computer-like model of human psychology were the results of
conscious methodological and theoretical choices. Even at a relatively early
point in the history of cognitive science research (since the 1950s), many
scientists found it essential to limit the perspectives to matters that could be
handled in what they considered a genuinely scientific manner. The steady
growth of computer technology and research into artificial intelligence offered
computer-like models of human cognitive activity. The cognitivist hardliners
created humans in the image of computers. It has since turned out that this is
not only a restricted, but also a truncated vision.6 It is not a new story, and it
is interesting to recall that already in 1973, Clifford Geertz had criticized the
now widespread mentalization of cultural and social levels in his discussion
of the concept of culture:

The cognitivist fallacy that culture consists (to quote another

spokesman for the movement, Stephen Tyler) of mental phenom-
ena which can [he means should] be analyzed by formal meth-
ods similar to those of mathematics and logic is as destructive
of an effective use of the concept as are the behaviourist and ideal-
ist fallacies to which it is a misdrawn correction.
(C. Geertz 1973: 12)

Nevertheless, there were some anthropologists who continued to work from

cognitive orientations, but they have not been included (in any significant

jeppe sinding jensen

way) in work by the most dedicated hardliners in the cognitive science

of religion such as that by, for example, Pascal Boyer, Scott Atran and Ilkka
Pyysiinen. Although the work of such anthropologists as Roy DAndrade,
Roger Keesing, Edwin Hutchins and Bradd Shore has been known, this ear-
lier research was more or less ignored by the new cognitivists in the study of
religion. Other examples of forgotten research are those presented in Models
in Language and Thought (Holland & Quinn 1987) and in the later Modes
of Thought: Explorations in Culture and Cognition (Olson & Torrance 1996).
These were attempts at integrating the social sciences and the humanities with
psychology and human biology. But these ambitions have not been incorpo-
rated into the theoretical and methodological cosmos of the hardliners either.
They exclude almost anything published before 1990. The earlier work was
broader and one can reasonably ask why this development took place.
Taking a closer look, one may suspect that the unspoken reason for not
including this earlier work was that it was not validated by experimental work
and so it was dismissed as mere description or anecdotal evidence.7 The
new cognitive hardliners working on religion base their work more on experi-
mental and especially developmental psychology. That explains why there is
so much concern with babies, toddlers and pre-schoolers before one gets to
anything that touches upon the subject of religion. For exactly the same rea-
son, the pre-cultural cognitive mechanisms and capacities are essential for
this type of research. Culture and everything related to it becomes noise.
The basic idea is to find only that which is evolved, inherited and innate (i.e.
given at birth) and develops without cultural contamination through social-
ization, language and so on.8 The cognitive hardliners in the study of religion
are thus mostly quite acute naturalists whose concerns are limited to those
aspects of human cognitive functioning which they assume to be present at
birth and biologically conditioned to develop ontogenetically. The method-
ology is highly individualistic and naturalistic. Everything which cannot be
studied using the methods of natural science will have to be excluded from
the realm of scientific research and so eliminated as scientifically irrelevant.
In my opinion, it seems that this orientation leads to a rather special view of
human nature and not least because these very same scholars are actually
obliged to use exactly the same means of communication, etc. which they
argue cannot be scientifically investigated.

Cultural eliminativism as method

Cultural eliminativists generally suggest that meaning is an ontologically

mystical notion. Meaning as it is generally understood can very well be
described as metaphysical as it (i.e. the notion) does not refer to anything

cognition and meaning

that can be measured or weighed. It is difficult to see how meaning could

have extensions in space or time, and it certainly does not obey the laws of
gravity. Hence as the cultural eliminativists are thoroughly naturalistic in their
epistemological position and scientific realists in their understanding of sci-
ence, it is evident that it is difficult for them to pose meaningful questions
about meaning. Furthermore, the kind of questions involved can hardly be
studied experimentally. Since the advancement of scientific theory depends
upon control, replication, prediction and laws, refutation or falsification,
it seems an impossibility to develop scientific theories about meaning. The
understandings of science and methodology which have been adopted so far
thus demand avoiding the discussion of meaning by systematically excluding
the possibility that it can at all be studied. In this case, ontology and episte-
mology are clearly the results of the chosen method. In my view, the oppo-
site ought to be the case in science, for if the methods determine ontology,
epistemology and the results of the research, then methodology has become
dogma. And so, the chosen discourse turns to immunization strategies which
conveniently avoid the possibility of falsification.
The rationale behind eliminativism can be difficult to grasp. It seems that
there is an effort at the development of a theory, a development which may
itself be demonstrated in a sociology of science perspective. A central figure
among the cognitivist hardliners is the French anthropologist Dan Sperber
who was once a structuralist. In 1975 he published Rethinking Symbolism, a
remarkable book in which symbolism and symbols are redefined as mental
processes instead of public and intersubjective socio-cultural phenomena. In
Sperbers empiricist theory of epistemology, symbolisation is some crea-
tive, but erroneous brain process which takes place in the absence of a clear
and identifiable reference for meanings (signifiants). To this, one must add
Sperbers view of interpretation in anthropological methodology, because he
views interpretation as nothing but an extension of the semantics of the
informants data, and, therefore, interpretation cannot contribute to a sci-
entific understanding as an explanation could. Sperbers strict naturalism is
likewise recognizable in his theory of culture which maintains, for instance,
that there is no meaning in a text, there is nothing but brain processes and
the clutter of printers ink. Ideas spread from brain to brain, and thus he
created the concept of imaginary epidemiology to account for such proc-
esses. Such are the empirical and scientific backgrounds which allow him to
proclaim that he is a true materialist in Explaining Culture (Sperber 1996).
Sperbers theoretical approach has been successful among some in the UK,
France and the US, where his ideas about anthropology as science in a com-
bination of naturalism, empiricism and individualistic methods have been
accepted as useful. If there is anything at all left of meaning in Sperbers
theory, it is confined to a mentalistic model. Cognitive, psychological models

jeppe sinding jensen

begin with the basic dogma that meaning, interpretation and understanding
are internal and mostly unconscious cognitive states and functions in the
individual. As an example: in his dismissal of Clifford Geertz as an anti-
psychologist, Pyysiinen produces the following statement about meaning
and symbolism:

As meaning does not reside in isolated objects, such as works of

art, texts, or buildings, but is the property of a conscious cogni-
tive process, symbolism is best viewed as a cognitive mechanism
that participates in the construction of knowledge as well as in the
functioning of memory. It is not an instrument of social commu-
nication or a property of phenomena that can be considered apart
from this mechanism. (Pyysiinen 2003: 41)

It could be interesting to apply this rather imperialistic attitude to other fields

in the humanities and social sciences: the history of art or macroeconomics
as exclusively cognitive mechanisms? Or, should we best study philosophy as
the scrutiny of the cognitive processes in philosophers brains?

Modules and modularity

The theory of the cognitive hardliners is also based on theses about modu-
larity, which is the idea that the brain or the cognitive apparatus consists of
a series of modules reflecting the evolutionary development of our species.9
These modules each have their own specific functions, each being uncon-
scious and independent of the others.10 In the specific case of humanity,
it so happened that our mental plasticity means that we can cross the
domains of the modules and their functions so that we can imagine things
which are counter-intuitive, such as omnipresent, omniscient, immortal
gods who cannot be perceived. Religion is thus simply a case of the brain
having run amok.11 Most of these hypotheses are founded on theoretical
developments about cognitive representations and the existence and function
of the modules which date to the 1980s. However, most of this modularity
discussion has been refuted by a series of scholars such as Merlin Donald
(1991, 2001), Annette Karmilloff-Smith (1992), Jeffrey Elman et al. (2001)
and Irene Appelbaum (1999).12 There are of course many meaningful ways
of performing scientific reductions but not much is left of society and cul-
ture if it all boils down to the functioning of the central nervous system.
Ilkka Pyysiinen, one of the cognitivists, has responded to the criticism by
pleading for the upholding of the culture concept; that is, not culture as a
something, but merely of the concept of culture as a useful abstraction. The

cognition and meaning

article was accepted in the Journal of Cognition and Culture, but one of the
editors called it crazy since hardliners do not recognize anything as objec-
tive reality except on the mental (i.e. cognitive) level, and then only in a
strictly individual understanding (Pyysiinen 2002, reprinted in Pyysiinen
2004).13 It was for the same reasons that Pascal Boyer was able to claim that
we humans do not share things, not culture and certainly not the money
in their pockets (Boyer 2001: 35).14
What is the goal of this programme? It is at least partially, but also impor-
tantly, intended to demonstrate that humans are not simply tabula rasa crea-
tures where the biology is a mere foundation upon which you can build
whatever you want as some earlier constructivist approaches to the human
and social sciences have maintained. Human cultural and social constructions
are limited and determined by their biological and thus cognitive apparatus.
Linked to the theory of innateness and modularity is the idea that humans are
not simple passive receivers of the environment, but that they actively process
their perceptions cognitively. Yet this interpretation has been recognized by
developmental and social psychologists and others for a long time as psycho-
logical constructionism. On the other hand, however, the basic methodology
of cognitivism means that such constructionism cannot be grasped on a col-
lective socio-cultural level but only at the individual level.
Thus, the hardliner version of this theory faces serious difficulties in accom-
modating explicit cultural knowledge. And it is precisely here that religion
and scholars of religion (in particular) come into the picture. Since all the
conclusions of the cognitive approach are based on the assumption that you
cannot say anything about anything except for the non-cultural, the entire
cognitive project remains relatively unattractive for, for example, philologists
and archaeologists and scholars of religion and of particular religious tradi-
tions, say Buddhism or Islam.
There are thus good reasons for believing that the cognitivist paradigm as it
is currently presented has some drawbacks as a tool for scholars of religion. Its
relevance would be enhanced if socio-cultural realities could also be attributed
explanatory validity. Meaning should thus be included as a relevant element
in the explanatory framework of the enlarged cognitive programme. Upon
closer inspection, it turns out that the hardliner cognitivists among the schol-
ars of religion have chosen to associate themselves with exactly that specific
research agenda which denies the levels of the social, cultural and linguistic
any meaning.15 One could just as easily create cognition and brain related
research approaches in the study of religion which include the social, cultural
and linguistic levels. That is one of the reasons that we in the Aarhus School
work with psychological and neurological researchers who are interested in
the links between the individual ontological level and the socio-cultural levels
in the research unit Religion, Cognition and Culture.16

jeppe sinding jensen

A joint project

My own dissatisfaction with the theoretical and scientific limits of the hard-
liner paradigm is an open secret. On several occasions I have played the dev-
ils advocate (e.g. Jensen 2002). I believe, along with the philosopher Andy
Clark, that it is now time to put the world back together again (Clark 2001).
There are connections between the body, the brain, consciousness, language,
culture and religion which are interesting; the last (religion) being particu-
larly relevant for some of us. In discussing the relationships between artefacts,
Clark advocates a much broader understanding, in the shape of:

[a] quite liberal notion of the scope of computation and cognitive

processes one that explicitly allows for the spread of such proc-
esses across brain, body, world, and artifact. Paramount among
such artifacts are the various manifestations of public language.
Language is in many ways the ultimate artifact: so ubiquitous it is
almost invisible, so intimate it is not clear whether it is a kind of
tool or a dimension of the user. (Ibid.: 218)

Something similar could be said about religion. Like language, religion is

among those means by which humans transcend their animal existence.
Precisely this is stressed by Merlin Donald:

We have evolved into hybrid minds, quite unlike any others,

and the reason for our uniqueness does not lie in our brains,
which are unexceptional in their basic design. It lies in the fact
that we have evolved such a deep dependency on our collective
storage systems, which hold the key to self-assembly. The ultimate
irony of human existence is that we are supreme individualists,
whose individualism depends almost entirely on culture for its
realization. It came at the price of giving up isolationalism, or cog-
nitive solipsism, of all other species and entering into a collectivity
of mind. (Donald 1991: 12)

It is here that the thing called meaning appears among the decisive fac-
tors in the formation of this collectivity of mind. On the one hand are the
cognitive hardliners and their individualistic methodology based upon the
empiricist idea that studies of cognitive activity understood as instinctive,
intuitive, repertories and domains activating modules in direct contact with
the environment will exhaust the explanatory possibilities. Opposed to these
views are the more philosophically founded insights concerning links in our
experience. Even such an empirically and natural science-oriented philoso-
pher as Quine says:

cognition and meaning

The totality of our so-called knowledge or beliefs, from the most

casual matters of geography and history to the profoundest laws of
atomic physics or even of pure mathematics and logic, is a man-
made fabric which impinges on experience only along the edges.
(Quine 1980: 42)

There is also another sense in which we are dependent upon everything which
others have discovered before us even if each generation believes that it rep-
resents the smartest ever. We are not much taken individually and by ourselves.
To take one example, Michael Tomasellos research into developmental psy-
chology demonstrates the degree to which the cultural context influences chil-
drens development and learning. Most of that knowledge which we have of
the world (taken as a whole) is mediated through others in a process which he
calls the ratchet effect (Tomasello 1999: 38). As Tomasello points out, and
here he includes phenomena related to the study of religion among the factors:

Beyond fundamental skills of primate cognition, therefore, chil-

drens domain-specific knowledge and expertise depend almost
totally on the accumulated knowledge of their cultures and its
transmission to them via linguistic and other symbols, includ-
ing both writing and pictures. The amount of knowledge that any
individual organism can gain by simply observing the world on its
own is extremely limited even nonliterate cultures have impor-
tant domains of knowledge that are almost exclusively in symbolic
format, and so they can only be transmitted symbolically most
clearly knowledge concerning things removed in space and time
such as characteristics of distant relatives and ancestors, myths and
some religious rituals, some knowledge of local flora and fauna,
and so forth. (Ibid.: 165)17

It is therefore by virtue of our linguistic and symbolizing abilities and social-

ity that we are what we are. And none of this would work if meaning did
not exist. Ideas about meaning can be created in many ways, and there are
an awful lot of them. Yet there is still not a single one which has persua-
sively and absolutely proved how thought (cognition) and speech (culture)
are related. This is by no means due to a lack of hypotheses, but they are still
just hypotheses. Some are admittedly more convincing than others.18 It might
then be appropriate to turn to cognitive semantics or cognitive semiotics. It
would appear that these approaches really do promise breakthroughs along
the future route. (e.g. Bundgrd et al. 2003) But the hardliner cognitivists
in the study of religion may not be inclined to take up the offer. In fact, the
discussion of meaning meets resistance. Those pleading for the importance

jeppe sinding jensen

of meaning in human cognition may be tagged as culturalists or exponents

of the the standard social science model.
However, there is a great deal of research that favours developments of
intensified theoretical connections between cognition and culture. These may
take different forms and come with different labels, such as social cognition,
distributed cognition, or collective intentionality. Ultimately, empirical
studies seem to suggest that the ontological development of human cogni-
tion is far more dependent upon non-cognitive factors than has hitherto been
allowed.19 This is because humans have a more complex social life than any
other species. And it is primarily via language that human sociality is created,
preserved and passed on to the coming generation. The recognition of the
importance of culture for human development, individually and collectively,
is currently gaining momentum:

Until recently, the relevance of culture for brain development,

structure and function remained unrecognized. This is changing,
however, with the emergence of cultural neuroscience, a cross-
disciplinary field of study that integrates cross-cultural psychology,
cognitive neuroscience and, in one formulation, molecular biol-
ogy to study how neural development, structure and function vary
from one cultural group to the next. (Duque et al. 2010: 138)

Human intentionality and being-in-language

Cognitive research approaches would thus be far more relevant to the study
of religion if they could accommodate culture and meaning once again. The
primary theoretical obstacle is that research which sticks to an individualistic
methodology cannot be capable of including issues of meaning since these are
not individual, private phenomena, but rather social and public. One can do
worse than recalling what Clifford Geertz wrote more than a generation ago:

The generalized attack on privacy theories of meaning is, since

early Husserl and late Wittgenstein, so much a part of modern
thought that it need not be developed once more here. What is
necessary is to see to it that the news reaches anthropology.
(C. Geertz 1973: 12)20

And as Geertz went on to say: Culture is public because meaning is.

However, this conception of meaning as public still goes against the intui-
tions of many scholars of religion who share the assumptions of the cogni-
tive approach when it comes to placing meaning in the (private) head. The

cognition and meaning

intuition is that When I say so, it is because I mean so, in my mind The
fact that this approach to meaning is not theoretical so much as intuitive is
the other thing that is difficult to grasp.
The difficulty for naturalists and empiricists is that meaning appears to be
a metaphysical manifestation simultaneously taking place in brains and in
symbols and thus meaning appears to belong to both the mental and physical
worlds. However, the solution could be to produce a world 3, a world of
meaning. It is an old idea, primarily promoted by Karl Popper (cf. e.g. Jensen
1999, 2004). The world of meaning consists of products derived from our
intentionality and of our general relationship to the world. The trick is, as I
wish to propose in this attempt, to simplify the problem and see how humans
as a species can produce complex communicative equivalents to an animals
scent tracks. Right from the simplest objectifications of our observations (e.g.
a lion!) up to the most complicated and abstract collective experiences (e.g.
the philosophy of Hegel), humans are able to express it all in those systems
of signs which they possess, and it is precisely the objective character of the
systems of signs which guarantees that everyone collectively has potential
access to intention.
Intention does not merely point to relations between physical objects but
also, and so much more, to imaginary, mental collective conceptions: inten-
tions direct intentions in vast networks. The World Wide Web is nothing
new there has been something of the kind for thousands of years (precisely
how long is a matter of debate). The primatologist Nicholas Humphrey (2002)
states that humans have such large brains in order to understand one another
and this certainly sounds like a convincing hypothesis. But brains also have
effects on each other and signs and symbols are important means through
which this communication takes place. As Mark Turner expresses this:

If we use the old metaphoric conception of the brain as an agent

who deals with language or as a container that for a moment
holds language while examining it for storage or discard, then
it is natural to think of the biology of the brain as unchanged by
its dealings with language. But if we use instead the conception
of the brain as an active and plastic biological system, we are led
to consider a rather different range of hypotheses: The brain is
changed importantly by experience with language: language is an
instrument used by separate brains to exert biological influences
on each other, creating through biological action at a distance a
virtual brain distributed in the individual brains of all the partici-
pants in the culture; early experience with language affects cogni-
tive operations that go beyond language.
(Turner 1996: 15960)

jeppe sinding jensen

Thus, meaning can be understood as a collective means of preserving inten-

tion. It starts off at the simplest level and then develops into the infinitely
complex and abstract. And furthermore, we can understand intentions of
others and adopt them as our own. We may, for example, chose (or be forced
to) let the intentions of a religious system of reference replace and override
our own intuitive intentions. However, not just anything can be preserved as
an intention in any one way there are constraints and rules, domains and
functions. On this issue, the cognitivists are right; a great deal of the human
intentional machinery is in fact dependent upon basic physical and mental
Again, so much in humans is conditioned by their history, background and
culture. All human practice invariably takes place within a space of reasons
(as the philosopher Wilfrid Sellars termed it) which is a part of our collective
world. Human practice is intimately related to what Ludwig Wittgenstein
summarized in the notion of form of life. This inspiration was later adopted
by the philosopher of language John McDowell in his criticism of the idea
that human rationality should simply be given with the individual cognitive
equipment. He calls it the myth of the endogenous Given:

Now I think we should be suspicious of the thought that we can

simply credit human individuals with this equipment, without
benefit of anything like my appeal to initiation into a shared
language and thereby into a tradition. I think the idea that this
cognitive equipment needs no such background is just another
outcropping of Givenness. We must take subjectivity and the
concept of objectivity to emerge together, out of initiation into
the space of reasons. (McDowell 1996: 1856)

The individually subjective, being the cognitive object, and the socially
objective, being the collective social constructions are the mutual results
of such initiations into the space of reasons. As this also involves the study
of religion and the meaning of religion in cognitive contexts, I will take the
liberty of citing myself:

What this means in terms of religion is that it lends theoretical

plausibility to the more old-fashioned idealist (but intuitively plau-
sible) view that religions somehow condition the ways in which
we think: that they as semantic engines are co-responsible for
the ways in which we process information and construct mean-
ing. In most traditional societies, culture, religion, and language
have been learned (installed) simultaneously so that meanings
are multi-meshed in the classificatory architecture.
(Jensen 2004: 246)

cognition and meaning

This may sound like an old-fashioned, almost mystical reification, claiming

that religion does something, but it does so as religions contain and consist
of specific programmes for thinking, speaking and acting. Such programmes
have prescribed forms for intentionality and practice which the participants
may accept on how to think, speak and act. Religion and the history of
religions provide an abundance of cases illustrating what is meant by collec-
tive intentionality. None of this would be possible without the exchanges of
meaning on the level of language: subjectivity and objectivity arise at the same
time through the initiations into the spaces of reasons. In the course of the
development of cultural history, one of the arenas of human practice where
this has happened explicitly and profusely is that arena which we, for the sake
of simplicity, call religion (Jensen 2003a).
Possibly, there are aspects of human cognition without meaning or lan-
guage and these might well interest cognitive scientists. There might possibly
also be meaning and language without cognition but this is very unlikely and
difficult to imagine. However, many disciplines in the human and social sci-
ences are outright sceptical of cognitive approaches to their subject materials
and seem to think that meaning and language do not relate to cognition.
I am convinced that for scholars of religion it is ultimately more useful to
accept that religious traditions are born, live and die, because humans are
concerned with their lives and this they do with their bodies, minds, emo-
tions and as members of collectives and societies. For all that to be possible,
cognition and meaning must come together in real life. And so they must
also in research.


1. An excellent introduction which addresses issues of religion and meaning directly is

Frankenberry and Penner (1999).
2. One could get the impression that cognitivism is widespread and especially so in the
US. This is not entirely correct; there are only a few academic contexts where cognition
plays a role in teaching. The impression simply reflects the fact that US-resident scholars
have dominated in this new paradigm.
3. Both forms of reduction can be recognized in the cognitive scepticism of socio-cultural
phenomena. The milder can be recognized by the slogan that x is nothing but y while
eliminatory reduction states that x does not exist. Often, however, category mistakes
are involved, and abstractions exist because they are tools for thinking. One could
illustrate the problem with analogous statements such as that literature is nothing but
books, or that traffic does not exist there are only individual vehicles, etc.
4. For a brief introduction to the paradigm, cf. my review of Boyer (2001) in Jensen
(2003b, 2009).
5. There is a long and involved philosophical debate about whether reasons can be causes,
see e.g. Evnine (1991: 2557). Lawson & McCauley also touch upon the problem in
their Rethinking Religion (1990).

jeppe sinding jensen

6. This story has already been covered in an excellent fashion by DAndrade (1995) and
A. W. Geertz (2004).
7. According to the hardliners, descriptive work in anthropology, sociology and psychology,
etc. is epistemologically weak as it is based upon the observations in non-experimental
contexts without documentation of controlled experiments allowing for replication and
prediction. Some of these scholars using experiments even refer to Popper to validate
their idea that observational research is problematic since it is theory-laden whereas
they regard controlled experiments with tests of hypotheses as genuine and value free
8. In class, students are surprised when confronted with these eliminativist hypotheses. A
recent eliminativist idea (aired at a conference) is that socialization does not exist
because in reality you can only deal with the psychological development of the indi-
vidual. The level of abstraction is not high.
9. There is no absolute agreement on these issues as no one actually knows how it all hangs
together. One should just be aware that a great deal of what appears to be objective,
visible reality is actually quite hypothetical. At times one could get the impression that
the heuristic models are understood as epistemological and perhaps even ontological.
10. The standard definition of a module: a specialized, independent, mental organ, which
develops through evolution for processing specific information of particular relevance
for a species. See, for instance the critical discussion in Elman et al. (2001). Tomasello
(1999: 206) has struck tellingly in expressing his doubts about modularity: Thus I do
not see the point of trying to modularize human cognition, and the many different pro-
posals for what the module menu looks like attest to the practical difficulties of doing
this as well.
11. One of Boyers later contributions to the discussion is entitled Out of Africa: Lessons
from a By-Product of Evolution (Boyer 2003).
12. Raymond Tallis delivered a very critical contribution to the discussion in the chapter
The Poverty of Neurophilosophy (Tallis 1999: 12754).
13. Pyysiinen goes to considerable lengths, e.g. Cultures do not exist as real abstract
entities; nor are they mere names. They also do not exist as fixed and given wholes
in the mind. Cultures are abstract wholes like sets or collections in mathematics and
logic; they are produced by the minds ability to create and understand abstractions
(Pyysiinen 2004: 228).
14. Pascal Boyers view of culture is also eliminativist, because culture does not have a
robust ontology and so, he argues, it is confusing to say that people share a culture, as
if culture were common property. We may have strictly identical amounts of money in
our respective wallets without sharing any of it(Boyer 2003: 356). This seems like a
convincing argument, but it merely demonstrates how Boyers individualistic method-
ology prevents him from recognizing the fact that we must share an economy to have
money in the first place
15. One can get a glimpse of the actual dimensions of the discipline of cognitive science
by taking a look at Bechtel and Graham (1999).
16. In addition to Merlin Donald and Michael Tomasello, Daniel J. Siegel, Joseph LeDoux
and Louis Cozolino also deserve mention. Most of the socio-psychological research
builds on such assumptions. This is only new in that with new techniques scientists can
actually see that the brain is a plastic organ and that the processes of socialisation have
real effects, see e.g. recent contribution by Margaret Wilson (2010).
17. Cf. Victor Turners now classic anthropological reflections on the use of ritual objects
and presentation of sacra in initiation rituals where the grotesque figures also bear
meaning in an analogous (non-language based) code (Turner 1967).

cognition and meaning

18. I am much in favour of the view of philosopher John R. Searle (e.g. Searle 1995, 2010).
However, in Philosophy in the Flesh, George Lakoff & Mark Johnson (1999) have pre-
sented a grandiose cognitive theory on these connections, where the entire history of
philosophy is being dissolved by the cognitive science of philosophy. Even if some
part of language semantics does have some bodily foundation, their theory does suffer
from the general weakness of cognitivist subjectivism: that they simply cannot (or do
not want to) heed the linguistic turn in philosophy. The social level is still lacking in
their model. Instead, they present what might be termed the neural theory of language
19. For example, research into social cognition as pursued in the Foresight Cognitive
Systems Project at the Institute of Cognitive Neuroscience in London (www.foresight.
gov.uk/cognitive.html). DeLoache (2004) presents an account from the perspectives of
developmental psychology. Interesting results and perspectives are presented be Vogeley
and Roepstorff (2009), Wexler (2006) and Zahn et al. (2009).
20. Something suggests that the history of the discipline deserves a larger share in our schol-
arly collective consciousness.


Appelbaum, I. 1999. Modularity. In A Companion to Cognitive Science, W. Bechtel &

G.Graham (eds), 62535. Oxford: Blackwell.
Bechtel, W. & G. Graham (eds) 1999. A Companion to Cognitive Science. Oxford: Blackwell.
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Boyer, P. 2003. Out of Africa: Lessons from a By-Product of Evolution. In Religion as a
Human Capacity, T. Light & B. Wilson (eds), 27144. Leiden: Brill.
Bundgrd, P., J. Egholm & M. Skov (eds) 2003. Kognitiv semiotik: En antologi om sprog,
betydning og erkendelse. Copenhagen: Haase.
Clark, A. 2001. Being There: Putting Brain, Body, and the World Together Again. Cambridge,
MA: MIT Press.
DAndrade, R. 1995. The Development of Cognitive Anthropology. Cambridge: Cambridge Uni-
versity Press.
DeLoache, J. S. 2004. Becoming Symbol-Minded. Trends in Cognitive Science 8(2): 6670.
Donald, M. 1991. Origins of the Modern Mind: Three Stages in the Evolution of Culture and
Cognition. Cambridge, MA: Harvard University Press.
Donald, M. 2001. A Mind so Rare: The Evolution of Human Consciousness. New York: W. W.
Duque, J. F. D., R. Turner, E. D. Lewis & G. Egan 2010. Neuroanthropology: A Humanistic
Science for the Study of the CultureBrain Nexus. Social Cognitive and Affective
Neuroscience 5(23): 13847.
Elman, J. et al. (eds.) 2001. Rethinking Innateness: A Connectionist Perspective on Development.
Cambridge, MA: MIT Press.
Evnine, S. 1991. Donald Davidson. Cambridge: Polity Press.
Frankenberry, N. & H. H. Penner (eds) 1999. Language, Truth and Religious Belief: Studies in
Twentieth-Century Theory and Method in the Study of Religion. Atlanta, GA: Scholars Press.
Geertz, A. W. 2004. Cognitive Approaches to the Study of Religion. In New Approaches to
the Study of Religion, vol. 2: Textual, Comparative, Sociological and Cognitive Approaches,
P.Antes, A. W. Geertz & R. R. Warne (eds), 34799. Berlin: Walter de Gruyter.

jeppe sinding jensen

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

Holland, D. & D. Quinn (eds) 1987. Cultural Models in Language and Thought. Cambridge:
University Press.
Humphrey, N. 2002. The Inner Eye. Social Intelligence in Evolution. Oxford: Oxford University
Jensen, J. S. 1999. On a semantic definition of religion. In The Pragmatics of Defining
Religion: Contexts, Concepts and Contests, J. G. Platvoet & A. L. Molendijk (eds), 40931.
Leiden: Brill.
Jensen, J. S. 2002. The Complex Worlds of Religion: Connecting Cultural and Cognit-
ive Analysis. In Current Approaches in the Cognitive Science of Religion, I. Pyysiinen &
V.Anttonen (eds), 20325. London: Continuum.
Jensen, J. S. 2003a. The Study of Religion in a New Key: Theoretical and Philosophical Soundings
in the Comparative and General Study of Religion. Aarhus: Aarhus University Press.
Jensen, J. S. 2003b. Pascal Boyer: Den ganske historie om al religion (nogensinde).
Religionsvidenskabeligt Tidsskrift 43: 5369.
Jensen, J. S. 2004. Meaning and Religion: On Semantics in the Study of Religion. In New
Approaches to the Study of Religion, Vol. 2: Textual, Comparative, Sociological and Cognitive
Approaches, P. Antes, A. W. Geertz & R. R. Warne (eds), 21952. Berlin: Walter de Gruyter.
Jensen, J. S. 2009. Religion as the Unintended Product of Brain Functions in the Standard
Cognitive Science of Religion Model. In Contemporary Theories of Religion, M. Stausberg
(ed.), 12955. London: Routledge.
Karmiloff-Smith, A. 1992. Beyond Modularity: A Developmental Perspective on Cognitive Science.
Cambridge, MA: MIT Press.
Lakoff, G. & M. Johnson 1999. Philosophy in the Flesh: The Embodied Mind and its Challenge
to Western Thought. New York: Basic Books.
Lawson, E. T. & R. N. McCauley 1990. Rethinking Religion: Connecting Cognition and Culture.
Cambridge: Cambridge University Press.
McDowell, J. 1996. Mind and World. Cambridge, MA: Harvard University Press.
Olson, D. R. & N. Torrance (eds.) 1996. Modes of Thought: Explorations in Culture and Cogni-
tion. Cambridge: Cambridge University Press.
Pyysiinen, I. 2002. Ontology of Culture and the Study of Human Behavior. Journal of
Cognition and Culture 2(3): 16782.
Pyysiinen, I. 2003. How Religion Works: Towards a New Cognitive Science of Religion. Leiden:
Pyysiinen, I. 2004. Magic, Miracles, and Religion: A Scientists Perspective. Walnut Creek, CA:
AltaMira Press.
Quine, W. V. O. 1980. Two Dogmas of Empiricism. In his From a Logical Point of View, 2nd
edn, 2046. Cambridge, MA: Harvard University Press.
Searle, J. R. 1995. The Construction of Social Reality. New York: Free Press.
Searle, J. R. 2010. Making the Social World: The Structure of Human Civilization. Oxford:
Oxford University Press.
Sperber, D. 1975. Rethinking Symbolism. Cambridge: Cambridge University Press.
Sperber, D. 1996. Explaining Culture: A Naturalistic Approach. Oxford: Basil Blackwell.
Tallis, R. 1999. On the Edge of Certainty: Philosophical Explorations. Manchester: Manchester
University Press.
Tomasello, M. 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard
University Press.
Turner, M. 1996. The Literary Mind: The Origins of Thought and Language. New York: Oxford
University Press.

cognition and meaning

Turner, V. W. 1967. The Forest of Symbols: Aspects of Ndembu Ritual. Ithaca, NY: Cornell
University Press.
Vogeley, K. & A. Roepstorff 2009. Contextualising Culture and Social Cognition. Trends in
Cognitive Science 12(12): 51116.
Wexler, B. E. 2006. Brain and Culture: Neurobiology, Ideology, and Social Change. Cambridge,
MA: MIT Press.
Wilson, M. 2010. The Re-tooled Mind: How Culture Re-engineers Cognition. Social
Cognitive and Affective Neuroscience 5(23): 18087.
Zahn R, J. Moll, M. Paiva, G. Garrido, F. Krueger, E. D. Huey & J. Grafman 2009. The
Neural Basis of Human Social Values: Evidence from Functional MRI. Cerebral Cortex
19: 27683.

Wittgenstein and the naturalness
of religious belief
Mark Addis

Wittgenstein and the cognitive science of religion share the broad objective of
distinguishing between science and religion but vary in their methodological
approaches. These divergences raise the question of the extent to which they
are in accord about substantive questions concerning religion. The cognitive
science of religion aims to differentiate between science and religion through
investigation into their cognitive foundations and cultural expressions using a
number of methodological precepts (Boyer 1994). Insight into many features
of religious cognition can be gained through the use of the approaches and
findings of cognitive science. Concept choice should be guided by explana-
tory scientific theories. Repeated patterns of individual and communal reli-
gious behaviour are to be clearly explicated so that testable theories which
elucidate this can be formulated. A widespread conception in religious stud-
ies which the cognitive science of religion opposes is that religion requires
special methods of study because it and especially religious experiences are
deemed to be unique. In this conception, scholars (e.g. Cannon 1996; Paden
1992) argue that the distinctive position of religion puts a restriction in prin-
ciple upon the capacity of scientific theories to explain it. It is argued that
customary explanatory practices in the natural and social sciences will fail to
provide appropriate insight into the phenomena of religion and that their
effectiveness is limited. An assumption which underpins this perspective is
that religion is concerned with the non-natural, but sustained reasoning is
infrequently provided for claims of this kind. Indeed some commentators (for
instance Lawson & McCauley 1990, 1993) claim that this presupposition
indicates that the field of religious studies quite frequently contains concealed
assumptions about religion.
A core tenet of the cognitive science of religion is that all religious thought
and action overwhelmingly depends upon the utilization of perfectly ordi-
nary forms of cognition available to all normally equipped people. Religious
representations and conduct depend on humdrum cognitive abilities which

wittgenstein and the naturalness of religious belief

naturally develop in people. It follows from this that providing an account

of religious belief and practices does not necessitate the use of special meth-
ods or the claim that there are faculties which are distinctively religious. In
opposition to common presuppositions in both religious studies and anthro-
pology the cognitive science of religion position is that acquiring under-
standing about religion does not necessarily require extensive and detailed
scrutiny of particular cultures. Theories in the cognitive science of religion
are grounded in philosophical naturalism but the type of explanatory frame-
work adopted varies. Some accounts (such as Boyer 2001) subscribe to com-
prehensive semantic and explanatory reduction but others take a stance of
semantic holism and explanatory pluralism (such as Lawson & McCauley
1990; McCauley 1996).
Wittgenstein stressed the distinctive nature of religious practice and the
potential for misunderstanding if it is seen as a competitor to science. His
remarks about religion emphasized the particular character of religious lan-
guage and how it can be misunderstood if its presuppositions are regarded
as alternatives to scientific ones. Religion was an important concern for
Wittgenstein although he was not religious in a conventional kind of way
(Malcolm 1993: 723). The influence of his remarks about religion upon
religious studies is quite disproportional to their quantity. Wittgensteins writ-
ings about religion are very limited and many of them are located in brief
collections of remarks, other peoples notes of his lectures, and records of
fragments of his thought. The primary sources for the later period are the
Remarks on Frazers Golden Bough (Wittgenstein 1979), Lectures on
Religious Belief (Wittgenstein 1970) and occasional comments in Culture
and Value (Wittgenstein 1980a). Wittgenstein did not explain how his reli-
gious and philosophical ideas were connected, so his writings about religion
might be misunderstood if they are taken to be parts of a philosophy of reli-
gion that was never properly developed. All these difficulties contribute to
the significant interpretative controversies over the remarks. A fundamental
claim of Wittgensteins philosophical methodology was that philosophy is not
a kind of science which aims at developing theories (and thus philosophical
naturalism of any kind is fundamentally misguided). He applied his distinc-
tion between philosophy and science when he claimed that psychological
investigations were not relevant to philosophy. Wittgenstein claimed that the
creation and evaluation of empirical models of how the mind operates is the
work of psychologists. He would have thought that the causal explanations of
cognitive science of religion were solely the province of psychology.
Intellectualism in the anthropology of religion is the position that although
religion may serve other purposes, its primary concern is the provision of
theories which explain the world (Horton 1970, 1993). As such religion and
science have identical explanatory objectives and the difference between them

mark addis

lies in the kind of language which they employ. The anthropologist James
Frazer represented religious beliefs as mistaken hypotheses and rituals as prim-
itive attempts to achieve what science does. Wittgenstein claimed that his
perspective was fundamentally mistaken. He argued that religious beliefs are
not analogous to scientific theories and therefore the application of scientific
thinking is erroneous. Religious beliefs should not be judged by the same
evidential standards as those in scientific theories and to do so would show
a profound misunderstanding of the character of these beliefs. Wittgenstein
commented that:

I would say, they are certainly not reasonable, thats obvious.

Unreasonable implies, with everyone, rebuke. I want to say: they
dont treat this as a matter of reasonability. Anyone who reads the
Epistles will find it said: not only that it is not reasonable, but that
it is folly. Not only is it not reasonable, but it doesnt pretend to be.
(Wittgenstein 1970: 58)

What is unreasonable here is the defence or criticism of religion using the

assumption that religious beliefs can be corroborated or falsified by evidence.
Wittgensteins attitude to the disparities between religion and science has very
broad affinities with the strand of thought in the cognitive science of religion
which rejects intellectualism in the anthropology of religion. The cognitive
science of religion argues for the view that the differentiation between religion
and science is not merely one of language but of actual cognitive process.
Some of the cognitive tasks required for high quality science are compara-
tively unusual and also ones that people often find hard to perform. In con-
trast, the cognitive processes underpinning religion are at least superficially
harder to explain and are simultaneously much more pervasive.

Defining religion

Wittgenstein and the cognitive science of religion share the view that theology
differs from religion. The cognitive science of religion claims that theology
utilizes many of the same kinds of thinking, especially deductive inference,
as science. The individual or collective practice of theology is not required
for either the emergence or persistence of religion (although it is frequently
influential when it does occur). Religion can and does thrive in the absence
of theology (Wiebe 1991). For example, see Barth (1975) on the Baktaman
of New Guinea being entirely indifferent to theology. Wittgenstein thought
that no theological doctrine in itself had the power to change peoples lives.
He held that attempts to demonstrate the truth of religious claims, such as

wittgenstein and the naturalness of religious belief

that of the existence of God, were absurd and would not establish an attitude
of commitment to religious beliefs. Wittgensteins remarks on religious belief
have been interpreted through the perspective of On Certainty. Some scholars
claim that he adhered to the notion of religious belief being the consequence
of a kind of life. For example, Wittgenstein stated:

Here believing obviously plays much more this role: suppose

we said that a certain picture might play the role of constantly
admonishing me, or I always think of it. Here there would be an
enormous difference between those people for whom the picture is
constantly in the foreground, and others who just didnt use it at all.
(Wittgenstein 1970: 56)

This stance accords with the perspective indicated by a significant number

of his remarks on religious belief. Wittgensteins comments appear to suggest
that religious beliefs are associated with the use of religious concepts and the
related attitudes that their employment implies. This viewpoint is most clearly
expressed in his well-known remark in Culture and Value: It strikes me that
a religious belief could only be something like a passionate commitment to
a system of reference (Wittgenstein 1980a: 64). Wittgensteins point was
that religious belief is an alignment of an entire life that does not depend
on changing scientific results or philosophical and theological doctrines. He
argued that religious beliefs are partly distinguished by their unshakeabil-
ity. However, if this view is understood as a general claim, it is false because
unshakeability does not characterize most religious belief.
The diverse range of features that comprise what are usually regarded as
paradigmatic cases of religion indicates how problematic it is to formulate
necessary and sufficient conditions for the application of the concept (Boyer
2001: 610). Progress cannot easily be made by identifying a characteristic
which only religions possess as this appears to be just as difficult as distin-
guishing one which all and only religions have. Wittgenstein took religion
to be a genuine category and due to the influence of Jamess The Varieties of
Religious Experience ([1901] 1960), he appreciated the widely differing ways
in which religious beliefs were expressed. A Wittgensteinian approach would
be to regard the differences between religions in family resemblance terms. It
is possible that Wittgenstein would have accepted the idea that religion is an
essentially contested concept (Gallie 1956). In contrast to this, the cognitive
science of religion view is that religion is not a satisfactory scientific explana-
tory category. For example, Pyysiinen remarks (2001: 4): it is doubtful
whether a scientific category of religion can be constructed, because the cat-
egory includes so many different kinds of phenomena that the cohesiveness
of the category cannot be accounted for by any one theory.

mark addis

An explanation of this diversity is offered by Boyer (2001) who attributes

it to divergences in the subsets of the cognitive dispositions and their pro-
clivities that are utilized by religion. Both Pyysiinen and Boyer are sug-
gesting that the notion of religion actually only indicates something about
the stance human beings tend to have on the distinct socio-cultural systems
that are termed religious before they are subjected to theoretical examina-
tion. What humans call religions prior to theoretical investigation encom-
passes a collection of properties that is too extensive to allow the concept of
religion to have particularly illuminating explanatory power. The analysis of
cognitive processes suggests that religions amalgamate emotional, conceptual
and behavioural patterns which occur in a wide range of other situations. A
group of theories in the cognitive science of religion argue for the hypothesis
that religious phenomena may possibly require the attribute of having some
involvement with modestly counterintuitive representations. There appears to
be evidence that this hypothesis encapsulates all the paradigmatic instances of
religion in a way that has both explanatory and predicative force. Given this,
a notion of religion which is cognitively grounded and fairly free from ethno-
centrism can be developed. However, a serious limitation of the hypothesis in
its current form is that the counter intuitiveness of representations is a very
weak condition for religion (as there are many non-religious representations
which have this feature).
Despite the cognitive science of religion view that religion is not a satis-
factory scientific explanatory category, it still identifies shared characteristics
of religions. The essential claim is that every religion turns to agents and
their actions as the key factors for understanding both the social and natural
world. These factors are not affected by the theologies that those who man-
age organized religion postulate. The usual practice of religion fits within a
structure which subscribes to culturally postulated superhuman agents and
their causal powers in conjunction with the facility of the human standard
theory of mind to account for their actions and states of mind. The knowl-
edge that gods are agents allows inferences about their values, preferences,
mental states, and actions. Myths and rituals utilize essentially the same
cognitive processes as narratives and plays do. Virtually all the best known
religious writings are myths. The explanations which religion offers are nor-
mally surrounded by or inferred from myths that have narrative form and
account for the state of the social and natural world through the invoca-
tion of the actions and states of mind of culturally postulated superhuman
agents. According to religion, rituals have either been prescribed or modelled
by these agents. Appropriately performed rituals either alter or maintain
states of affairs in ways that can be specified. The conduct of rituals supplies
humans with a method for creating order in and control over the social and
natural world.

wittgenstein and the naturalness of religious belief

There is controversy over whether Wittgensteins comments about magic

in Remarks on Frazers Golden Bough tend to support expressivism. The
expressivist view is that Wittgenstein sought to criticize Frazers idea that reli-
gious beliefs were mistaken hypotheses and rituals were primitive attempts
to achieve what science does by emphasizing how belief and ritual were
essentially expressive in nature. From this perspective, religion should not
be regarded as looking to characterize and control the supernatural. Instead
religion is the expression of attitudes and emotions and as such is not a mis-
taken attempt at genuine science but a natural manifestation of what it is to
be human. The cognitive science of religion account of religious agents, nar-
ratives and rituals indicates that the expressivist interpretation misdescribes
religion. Even if Wittgenstein should not be regarded as favouring expressiv-
ism, there is good reason to think that he misdescribed religion (despite the
fact that some commentators, such as D. Z. Phillips, insist that he did not).
It is not all clear that religious belief solely consists of a passionate commit-
ment a system of reference. If Wittgenstein was guilty of misrepresentation,
this indicates that his professed philosophical methodology of description was
in conflict with his actual practice when considering religion.

Forms of life

The concept of a form of life is a much contested area and has been the
subject of abundant discussion. Although Wittgenstein used the expression
form(s) of life a very small number of times, it is plausible to think that he
employed the phrase in two distinct ways. One use is to summarize the bio-
logical aspects of human nature in the sense of the common human way of
acting, namely that which is particularly and universally human. The other
employment is to refer to the cultural facets of human nature in the sense
of stressing the differences between societies. These aspects are broadly con-
cerned with practices in ways which encompass both anthropology and soci-
ology. Both strands of form(s) of life rest upon the very general facts of nature
which are the background stabilities of the natural world. For Wittgenstein,
these very general facts of nature impose limitations upon which concepts
are natural or unnatural to nearly all humans (Wittgenstein 1980b: 708;
1984: 230).
Some scholars have used the notion of a form of life to develop an interpre-
tation of Wittgensteins remarks on religion which is fundamentally opposed
to the cognitive science of religion perspective. The fideist interpretation
employs this concept to explicitly argue for the view that religion requires
special methods of study. Religious concepts and language are deemed only
to be intelligible to participants in a religious form of life, and their full

mark addis

comprehension is inseparable from understanding the associated form of life.

Non-participants in a religious form of life cannot grasp religious concepts
or language. Religious beliefs are distinct language games since they are not
connected to what lies outside religion and their justification stops. The very
thought of analysing the nature and epistemic grounds of this belief indicates
a lack of appropriate understanding about the character of religion. If there
are distinct religious language games which are only comprehensible to par-
ticipants in the relevant form of life, it follows that non-participants cannot
criticize these language games. A major criticism of fideism focuses upon the
implications of there being distinct religious language games for how the
notion of the form of life is perceived. The problem is the excessive com-
partmentalisation of the forms of social life involved in claiming that there
are these distinct language games. This difficulty in conjunction with other
cogent objections to fideism gives strong grounds for holding that this inter-
pretation of Wittgensteins remarks about religion is mistaken.
The cognitive science of religion characterizes thoughts and actions as nat-
ural if they possess attributes which rely upon what Boyer (1994) termed
non-cultural foundations. This conception of natural is similar to that aris-
ing out of Wittgensteins biological form of life as this deals with what is
natural to all humans. Humans have many shared intuitive beliefs, cognitive
and behavioural patterns. The creation and persistence of particular sorts of
patterns frequently needs little cultural support. There is considerable evi-
dence indicating that facets of religious cognition are not heavily dependent
on cultural factors. Religion is present in every human culture with certain
ideas and forms being repeated across time in a wide range of geographical
and cultural settings. Although particular religions can disappear religion does
not and new religions routinely appear in human populations (Sperber 1996).
The cognitive science of religion claims that understanding or retaining
religion does not necessitate the acquisition or possession of any kind of spe-
cialized intellectual abilities. The key facility required for religion is that of
distinguishing agents and their actions from other entities and events in the
social and natural world. Agent detection abilities occur quite early in human
development as do basic action representation systems. Representations of
events which involve the causal powers of agents are significantly dissimilar
from those that do not. Humans find it natural to invoke culturally pos-
tulated superhuman agents and their actions for explanatory purposes. The
human ability to represent agents and their actions prepares people for the
creation, understanding, remembering and transmission of myths and ritu-
als. For instance, the cognitive representation of ritual actions depends upon
the basic action representation system to which is added the representation
of culturally postulated superhuman agents. This description of religious cog-
nitive capacities and tendencies is consistent with Wittgensteins idea about

wittgenstein and the naturalness of religious belief

certain things being natural given a particular biological form of life. When
he discussed the biological aspects of human nature Wittgenstein showed
very little interest in exploring the ramifications of the idea that there are cer-
tain cognitive capacities shared by all normal people that underpin religion
because according to his philosophical methodology this was the concern of

World pictures

A fundamental idea in On Certainty is that knowledge is comprised of two

broad classes of propositions. There is a core of propositions that forms the
basis of inquiry and which is surrounded by empirical propositions that are
the results of investigation. Wittgenstein emphasized that world picture prop-
ositions differ in kind from empirical ones. Although they take the form of
empirical propositions, world picture propositions do not operate as empiri-
cally testable propositions. These core propositions constitute a world picture,
and the latter is a system of propositions which support one another. A system
of propositions is learnt gradually, some of which are certain and indubitable
while others may be doubted to a greater or lesser degree. Propositions which
are certain do not have this status because they are intrinsically obvious or
especially convincing. It is rather that they are certain by virtue of those that
shift around them (Wittgenstein 1969: 144). The propositions adopted and
whether a particular proposition is plausible depends upon the world picture.
The world picture provides the framework for inquiry. Wittgenstein remarked
that a world picture is the inherited background against which I distinguish
between true and false (ibid.: 94). World pictures are not based on reason,
but they are not unreasonable. The world picture itself cannot be doubted. It
is the background against which other knowledge is acquired and the context
in which claims to know are meaningful. World picture propositions function
as rules of testing, but are not tested themselves. A difference in world picture
propositions leads to a difference in what is counted as evidence. For example,
Wittgenstein suggested that he operated with a different concept of evidence
about the Last Judgement and that he found the kind of evidence the believer
appeals to impossible to accept (Wittgenstein 1970: 61).
For Wittgenstein the cognitive capacities involved in learning and process-
ing world picture propositions are a matter for psychology. World picture
propositions are not explicitly learnt, but it is possible to subsequently find out
which propositions these are (Wittgenstein 1969: 279, 152). Wittgenstein
claimed that learning was based on accepting the authority of a community
and that adults as well as children must take much knowledge on trust (ibid.:
15960, 170, 5089). He remarked:

mark addis

As children we learn facts; e.g. that every human being has a

brain, and we take them on trust. I believe that there is an island,
Australia, of such-and-such a shape, and so on and so on; I believe
that I had great-grandparents, that the people who gave them-
selves out as my parents really were my parents, etc. This belief
may never have been expressed; even the thought that it was so,
never thought. (Ibid.: 159)

I really want to say that a language-game is only possible if one trusts

something (I did not say can trust something). (Ibid.: 509)

Accepting much knowledge on trust allows humans to undertake investiga-

tions which modify some of their beliefs (ibid.: 161).
Religion is part of a form of life which is based on a world picture. It
seems at least possible that Wittgensteins account of world pictures could
be associated with the cognitive science of religion idea that the acquisition
of knowledge required for participation in religious practice has far greater
affinities with acquisition of a natural language than the gaining of the abili-
ties and knowledge necessary to undertake scientific work. What underpins
this idea is the view that acquiring religious knowledge frequently occurs
without explicit teaching, as people are born into religious and linguistic
communities. This notion of acquiring knowledge from the community in
which one is born is arguably usefully expressed by the conception of a
world picture. For instance, it helps to explain why the apparent irrationality
of many religious beliefs is not noticed by a good number of their adher-
ents. Wittgenstein argued that the accumulated knowledge of a culture is a
collective human achievement (ibid.: 288, 298). Cognitive science theo-
ries about the transmission of culture can be used to supplement his ideas
about the processes by which religious beliefs are acquired and transformed
in world pictures. These theories claim that culture results from the causal
interaction of mental and public representations. Beliefs in humans which
have emanated from communication show remarkable similarities across
individuals. The majority of cultural change stems from alterations in the
distributions of communicated beliefs in societies, but it should be observed
that these beliefs are only part of the whole set of cultural representations
(Sperber 1996: 25). An important element in explaining the distributions
of communicated beliefs is the idea that cultural representations have a pro-
pensity to be transmitted if they have certain properties. These properties
include being quickly recognized, easily communicable, readily memorable,
balancing the ability to attract attention with that of underwriting cost-free
inferences, and having the facility to motivate humans to devote effort to
their transmission. There is cross-cultural evidence which demonstrates that

wittgenstein and the naturalness of religious belief

the modestly counterintuitive representations of religions are swiftly and pre-

cisely recollected (Boyer & Ramble 2001).
The biological aspects of human nature help to explain the noticeable simi-
larities between many world pictures. However, sufficiently different world
pictures can be incommensurable. For instance, Wittgenstein commented:

We come to an island and we find beliefs there, and certain beliefs

we are inclined to call religious Entirely different connections
would make them into religious beliefs, and there can easily be
imagined transitions where we wouldnt know for our life whether
to call them religious beliefs or scientific beliefs.
(Wittgenstein 1970: 58)

In this case the distinctions between science and religion could be very dif-
ferent to the ones found in contemporary western societies. Wittgenstein
argued that certain world picture propositions could not be revised or rejected
because doing so would dismantle the world picture. A consequence of this
position is that Quines notion of a web of belief cannot be used to resolve
conflicting world pictures as certain propositions cannot be modified or aban-
doned. Wittgenstein was not a relativist so he would not have favoured the
view that each world picture is as much or as little justified as any other
(Wittgenstein 1969: 60812). He commented: I said I would combat
the other man but wouldnt I give him reasons? Certainly, but how far do
they go? At the end of reasons comes persuasion. (Think what happens when
missionaries convert natives.) (ibid.: 612).
Some world pictures are incommensurable, so ordinary argument based on
shared premises will not succeed and only persuasion is possible. Wittgenstein
suggested that a decision to combat a different world picture should be
thought about carefully. If other world pictures are merely regarded as infe-
rior versions of the one which is held, then the tendency will be to combat
them (and in doing so fail to understand them).
The account that has been favoured here interprets Wittgensteins remarks
on religious belief, anthropology and certainty as articulating the view that
religious discourse is not the result of some theoretical attempt at explanation
but instead arises from a commitment to a particular world picture. Some of
his views have affinities with certain aspects of the cognitive science of religion
but he firmly rejected any kind of commitment to philosophical naturalism.
The contribution which Wittgenstein can make to the development of this
research agenda is most likely to be in the area of the relationship between
religious beliefs and world pictures.

mark addis


Barth, F. 1975. Ritual and Knowledge among the Baktaman of New Guinea. New Haven, CT:
Yale University Press.
Boyer, P. 1994. The Naturalness of Religious Ideas: A Cognitive Theory of Religion. Berkeley, CA:
University of California Press.
Boyer, P. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York:
Basic Books.
Boyer, P. & C. Ramble 2001. Cognitive Templates for Religious Concepts: Cross-Cultural
Evidence for Recall of Counter-Intuitive Representations. Cognitive Science 25(4): 53564.
Cannon, D. 1996. Six Ways of Being Religious: A Framework for Comparative Religion. Belmont,
CA: Wadsworth.
Gallie, W. B. 1956. Essentially Contested Concepts. Proceedings of the Aristotelian Society
56: 16798.
Horton, R. 1970. African Traditional Thought and Western Science. In Rationality,
B.Wilson (eds), 13171. Oxford: Blackwell.
Horton, R. 1993. Patterns of Thought in Africa and the West: Essays on Magic, Religion, and
Science. Cambridge: Cambridge University Press.
James, W. [1901] 1960. The Varieties of Religious Experience. London: Collins.
Lawson, T. E. & R. N. McCauley 1990. Rethinking Religion: Connecting Cognition and Culture.
Cambridge: Cambridge University Press.
Lawson, T. E. & R. N. McCauley 1993. Crisis of Conscience, Riddle of Identity: Making
Space for a Cognitive Approach to Religious Phenomena. Journal of the American Academy
of Religion 61(2): 20123.
Malcolm, N. 1993. Wittgenstein: A Religious Point of View. London: Routledge.
McCauley, R. N. 1996. Explanatory Pluralism and the Coevolution of Theories in Science.
In The Churchlands and Their Critics, R. N. McCauley (ed.), 1747. Oxford: Blackwell.
Paden, W. 1992. Interpreting the Sacred. Boston, MA: Beacon Press.
Pyysiinen, I. 2001. How Religion Works: Towards a New Cognitive Science of Religion. Leiden:
Sperber, D. 1996. Explaining Culture: A Naturalistic Approach. Oxford: Blackwell.
Wiebe, D. 1991. The Irony of Theology and the Nature of Religious Thought. Montreal: McGill-
Queens University Press.
Wittgenstein, L. 1969. On Certainty. Oxford: Blackwell.
Wittgenstein, L. 1970. Lectures and Conversations on Aesthetics, Psychology and Religious Beliefs.
Oxford: Blackwell.
Wittgenstein, L. 1979. Remarks on Frazers Golden Bough. In Wittgenstein Sources and
Perspectives, G. Luckhardt (ed.), 6182. New York: Harvester Press.
Wittgenstein, L. 1980a. Culture and Value, Oxford: Blackwell.
Wittgenstein, L. 1980b. Remarks on the Philosophy of Psychology, vol. 2. Oxford: Blackwell.
Wittgenstein, L. 1984. Philosophical Investigations. Oxford: Blackwell.

Peekaboo! and object permanence:
on the play of concealment and
appearance in cognition and religion
Thomas Hoffmann

Wet auto-da-fs and cognitive readings

From the past two hundred years of scholarly endeavour in the study of reli-
gion, we have all become acquainted with a profusion of innovative hypoth-
eses concerning the origin and essence of religion. In the course of theoretical,
methodological, even political, progress, most of these ideas were dismissed
as unwarranted historically, psychologically, ethically, gender-wise and so
forth. However, academic turns and tides unfortunately often assume the
rhetorical form of what could be called wet auto-da-fs (i.e. throwing a lot
of babies out with the bathwater). Although I take cognitive studies to belong
to one of the most promising approaches in the academic study of religion
today, I am also a staunch believer that it is usually worthwhile to re-read and
recycle its precursors. To be sure, not just for the sake of erudite namedrop-
ping, genealogical awe, or an Olympic vista of alterations and continuities,
but for the sake of genuine re-exploration. As Victor Turner put it:

It is not a theorists whole system which so illuminates, but his

scattered ideas, his flashes of insight taken out of systemic con-
text and applied to scattered data. Such ideas have a virtue of
their own and may generate new hypotheses. They even show
how scattered facts may be systematically connected! Randomly
distributed through some monstrous logical system, they resemble
nourishing raisins in a cellular mass of inedible dough. The intui-
tions, not the tissue of logic connecting them, are what tend to
survive. (Turner 1974: 23)

Thus, the following tentative ideas on the interrelations of religion, origins,

evolution, cognition and play are beholden at least indirectly to the early
theories of the playwright, poet and philosopher Friedrich Schiller (1759

thomas hoffmann

1805; Schiller [1795] 1981), the philosopher Herbert Spencer (18201903;

Spencer 1873), and some lesser known neo-Darwinian philosophers and psy-
chologists like Karl Groos (18611946; Groos 1898) and Granvill Stanley
Hall (18441924), as well as the physical educator Luther Gulick (1865
1918; Gulick 1898). Furthermore, an early cognitive pioneer, namely Jean
Piaget (18961980; Piaget & Inhelder 1971), proved not only absolutely
indispensable but also a felicitous match to the most contemporary of my
inspirational sources, the cognitive linguist George Lakoff and philosopher
Mark Johnson. Lakoff and Johnsons theories regarding basic image schemas
imprinted in the mind through embodied action seems to support Piagets
initial conclusions on a phenomenon he designated as object permanence
(explanation follows).
As a final point and critical remark, I believe that in general the scientific
cognitive study of religion should be more committed to specific, special-
ized, and context-sensitive readings of their common materials.1 Otherwise,
I fear, it will deteriorate into mediocre science studies performed by people
from the humanities with only limited training in the natural sciences and
psychology that, anyway, is a critique I sometimes encounter among scien-
tists. Even so, the current state of affairs is still in a nascent phase and a recent
issue of Poetics Today2 dedicated to the topic Literature and the Cognitive
Revolution raises my spirits because the editors call for a cognitive histori-
cism. In light of the indisputable importance of canonization, canons and
stylized (whether innovative or conventional) discourse-registers in numerous
religious texts, an approach more firmly embedded in cognitive linguistics
is certainly warranted, along with the conventional philological disciplines.
After all, strong readings are usually what convince us when experimenting
with a promising theory.

Looking for a hypothesis

Without further ado I will proceed from the classical and, admittedly, not
unproblematic hypothesis that research in child developmental psychology
presents us with one of the most empirically viable starting points regarding
evolutionary cognition.3 Perhaps some infantile cognitive faculties, which
only beset us for a short time during the infancy of our species, held much
more sway over prehistorical man, and have subsequently left us with religio-
cultural imprinting4 or recapitulatory schema. The physiological correlations
between the phylogeny and ontogeny of human beings (including their early
existence as embryonic creatures) might not be the only pertinent guide
it might be that cognitive correlations are relevant as well. The latter idea,
originally espoused by German biologist and philosopher Ernst von Haeckel

peekaboo! and object permanence

(18341919), came to be called the Recapitulation Theory, but it was Gulick

and Hall5 who developed and adapted the theory to child development the-
ory: according to this theory the development of children recapitulates evo-
lutionary developmental phases as they grow up (Gulick 1898; Hall 1920).
Particularly, they thought that several playful motor-activities in children
betrayed a complex of reflections of earlier evolutionary experiential patterns,
such as the animal stage (climbing, swinging), the savage stage (hunting, tag,
hide-and-seek), the nomad stage (keeping pets), the agricultural/patriarchal
stage (dolls, digging in sand) and the tribal stage (team games). Gulick also
viewed modern sports like hard running, accurate throwing and club hitting
as playful recapitulations of early hunting skills.
While such confident details of early quasi-Lamarckian recapitulatory the-
ories are now rejected as nave and crude, exaggerated and even quasi-racist in
their conception of evolution, their influence on later developmental theories
is unquestionable.6 More generally, they bear a strong resemblance to more
accepted evolutionary paradigms, such as various versions of dual inheritance
theory, which argue that humans and their religious systems are the highly
functional, yet often excessive and disastrous, products of both biological and
socio-cultural evolution. In the cognitive study of religion, scholars such as
Walter Burkert (1996), Justin Barrett (2000) and Pascal Boyer (2002), have
clearly proceeded along the dual inheritance track even though some of
them might not subscribe to that description.7
Another progeny of the dual inheritance family has been adduced in cog-
nitive semantics, the somewhat loose term for a branch of the overall field of
cognitive linguistics (e.g. Croft & Cruse 2004). Cognitive semantics usually
revolves around issues of lexical semantics, category formation and conceptual
metaphors (e.g. Lakoff 1987) rather than morphology, grammar or neurol-
ogy. Emphasis is placed on basic embodied image schemas that emerge, on
the one hand, by dint of our evolutionary kinaesthetic engagement with basic
natural givens like, for instance, gravity, the visibility and malleability of
objects, and our creative bent for putting these image schemas into analogical
and metaphorical operation in language, on the other. The modern notion
of schema is of Kantian origin, but its contemporary re-worked notion as so-
called image schemas has been most successfully espoused by linguist George
Lakoff (1987) and philosopher Mark Johnson (1987), who both sought to
demonstrate that conceptualization and linguistic expression whether literal
or metaphorical reflect dynamic cognitive patterning of recurrent bodily
experience. In an early formulation of Johnson:

The view I am proposing is this: in order for us to have meaningful

connected experiences that we can comprehend and reason about,
there must be pattern and order to our actions, perceptions, and

thomas hoffmann

conceptions. A schema is a recurrent pattern, shape, and regularity

in, or of, these ongoing ordering activities. These patterns emerge
as meaningful structures for us chiefly at the level of our bodily
movements through space, our manipulations of objects, and our
perceptual interactions. (Johnson 1987: 29)

Since then, the notion and scope of image schemas has been refined and
grounded. Cognitive linguist Tim Rohrer sums up:

(a) Image schemata are recurrent patterns of bodily experience;

(b) Image schemata are image-like in that they preserve the
topological structure of the whole perceptual experience;
(c) Image schemata operate dynamically in and across time;
(d) Image schemata are structures which link sensorimotor expe-
rience to conceptualization and language;
(e) Image schemata are instantiated as activation patterns (or
contours) in topologic and topological neural maps;
(f ) Image schemata afford normal pattern completions that can
serve as a basis for inference. (Rohrer 2005: 173)

As stated in (b), although the schemas work dynamically, they can be seized
and represented in simple diagrams using elements like arrows and circles.
One schema that shall prove most noteworthy in this context is that of the
containment (Johnson 1987: 21ff.; Lakoff & Johnson 1999). This contain-
ment schema rehearses in a very general and flexible way the most archetypi-
cal and commonplace experience of putting objects into and taking them out
of bounded areas or alternatively visually tracking objects into or out of
some bounded areas, i.e. a container. Diagrammatically, containment in its
most unadorned and minimalist state can be depicted as a dot inside a circle:
like .8
When probing further into the containment schema, dynamic sub-schemas
of concealment and exposure emerge,9 which correlate neatly with conven-
tional religious terminology pace any deconstructive critique such as mys-
tery (from Greek myein, to close ones eyes or mouth or conceal), secrecy
(from Latin secernere to set apart), occultation (from Latin occultus, to con-
ceal) versus revelation (from Latin revelatus, to draw away), apocalypse (lit.
un-hide, to unveil), and epiphany (from Greek epiphaneia, appearance).
In short, all terms conceivable as variations of the containment schema.

peekaboo! and object permanence

Man-child, cognition and play:

speculations on religious links

When dealing with the origins and evolutionary tracks of religion, specula-
tion simply goes with the territory. There is no point in denying the specula-
tive fringes, and, for my part, they centre on play and, as it were, man-child.
Here it seems reasonable that we consult child psychology and its growing
theoretical and empirical body on play. Today child developmental psychol-
ogy clearly offers the most comprehensive and data-loaded field regarding
play.10 For several decades in developmental psychology, it has been sensus
communis that play constitutes a significant factor in the development of
tool use and problem solving (Bruner et al. 1976), of language and cognition
(Vygotsky 1967), of self-concept (Mead 1934), and most important in the
present context in the development of symbolization, pretence and make-
believe (Piaget 1962; Singer 1973; Bateson 1955). In the scientific study of
religion, play and not only the gleeful variety but the serious and intense as
well has been granted an important role too. An early academic landmark
regarding play, culture and religion is Johan Huizingass Homo Ludens from
1938, in which the reader is presented with a cornucopia of play and play-like
notions and practices while making a case for their intricate ties to such key
cultural institutions as sex, jurisprudence, war, art, philosophy and of course
religion. Since then, the subject has only received sporadic attention, even
though the highly influential Turner published From Ritual to Theatre: The
Human Seriousness of Play in 1982.11

Annihilated and resurrected

Trying to circumscribe what kind of infantile cognitive faculty religion

exploits and plays with, we must attend to the very first year of our life, more
exactly between our fourth and twelfth months. From the numerous clini-
cal observations12 that in their first year infants proceed from being unable
to maintain so-called object permanence to being able to maintain so-called
object permanence, it could be hypothesized that prehistorical man under-
went this cognitive development on a much more protracted timescale and
that it might have had some important bearings on the origin of religious
concepts and practices. Now, what precisely is object permanence?
Object permanence is a standard term in cognitive psychology, coined by
Jean Piaget, denoting the ability to make an intrinsic connection between
the very same object that is on display and then concealed or vice versa. Prior
to full-blown object permanence-ability the world of these infants consti-
tutes a visual field of objects that is constantly being, in the words of Piaget,

thomas hoffmann

annihilated and resurrected (Piaget 1954: 11). In such a visual field, an

object represents a mere image which re-enters the void as soon as it van-
ishes, and emerges from it for no apparent reason (ibid.). More precisely, a
leap of conceptualization occurs in the infants ability to mentally grasp and
preserve the continued existence of objects out of sight during the period
between two-and-a-half and twelve months. Piagets experiments led him to
the following conclusions (among many others):

That infants seem to forget objects not present to their senses during
their first three months
That, at four months, they seem to understand that objects continue
to exist even though the objects are out of sight. However, they are still
not able to act on this knowledge and they seem to forget the location
of the object quickly
At about eight months, they begin to search for hidden objects but they
still quickly forget about their location and instead become captivated
by their own movements
From then on their ability to consistently preserve object permanence
improves well into the second year of life.

What makes the infant achieve object permanence seems to be closely related
to the development and fine-tuning of the sensorimotor skills, in particular
the use of hands and probably also gaze.13 In other words, when the infant is
able to act on and literally handle its proximate objects, object permanence is
then likely to evolve.
To be sure, there is a long and winding road from Piagetian sensorim-
otor substages to hypothetical religious scenarios, but if we keep in mind
Piagets words on annihilation and unfathomable resurrection, we begin to
get the drift of a typical theistic and ritualistic scenario perhaps this is no

Prehistoric object permanence

Needless to say it would be unwise to pose prehistorical man as nothing

more than an adult child, a childish precursor to our present apogee posi-
tion. This certainly smacks of outdated research on primitive religions and
their childish practitioners (e.g. Levy-Bruhl 1922). Indeed, it would be
difficult to conceive of an evolutionary scenario in which prehistorical man
would stop chasing an animal fleeing into the thicket or would stop digging
for hidden roots and the like. Contrary to the infant, prehistorical man must
have been able to reason, or rather imagine, that things are not only what

peekaboo! and object permanence

they seem.14 That things in the broadest sense of the word (including agents)
have a life of their own irrespective of ones own narrow-sighted egocentric
perceptions. Like prehistorical man, we intuitively know (based on count-
less experiences) that objects, among many other attributes, by and large
have substance, are external to ourselves, maintain their identity in spite of
being moved and continue to exist even if they are hidden from us. In other
words: instead of adhering to the infantile psychology of out-of-sight, out-
of-mind, prehistorical man was able to entertain and elaborate the idea that
things are still there existing behind any visual barriers or in the form of
traces.15 This still being there must be viewed on an axiological axis with
one end associated with the lustful and cheerful, and the other end associated
with the uncanny and risky. Both ends of the axis can be affiliated with the
traditional argot of religions. As to the uncanny end, this has been associ-
ated with the so-called hyperactive agency detection device (aka HADD), a
concept developed in the cognitive science of religion (e.g. Guthrie 1993;
Atran 2002; Barrett 2000). And just as HADD is believed to be functional
for man in a huntergatherer milieu with predators and venomous animals
but still puts its strains and boons on us as part of our evolutionary heritage,
so might object permanence also be a part of that heritage too. Perhaps pre-
historical man entertained a much more heightened attention to the almost
mystical experience of first seeing an object, then not-seeing the object,
then seeing it again, and then finally this being the decisive cognitive leap
seeing it with the minds eye.

The first part of the hypothesis

Hence the initial part of the hypothesis is that religion in the form of beliefs
in supernatural agents and objects draws on this infantile cognitive phase
and subsequently elaborated and warped it in innumerable cultural-ritual
ways in order to recollect and play on/with this canonical cognitive experi-
ence. The example par excellence would be theistic religions that to put it
very simply propagate beliefs in personalized gods that appear at crucial
moments and then disappear again. To rehearse a few of the standard terms,
we can list revelation, vision, epiphany, occultation and avatar.
It is important to notice that this kind of theistic pattern of appearance
and disappearance often takes the form of some kind of veiling, meaning that
the god is believed to be present but hidden behind or inside something or
at least leaves some kind of vestige or sign. What these religions provide are
not only particular belief systems in the form of linguistic propositions, but
also ritual scenarios that instil, rehearse, and revise these seemingly simple
moments of appearance and disappearance.16

thomas hoffmann

The image schema of containment tallies well with this widespread reli-
gious topos and is furthermore able to anchor it to human beings most per-
vasive features of bodily experience. As Johnson notes:

We are intimately aware of our bodies as three-dimensional con-

tainers into which we put certain things and out of which other
things emerge From the beginning, we experience constant
physical containment in our surroundings We move in and out
of rooms, clothes, vehicles [though probably not for archaic man],
and numerous kinds of bounded spaces. We manipulate objects,
placing them in containers. (Johnson 198