Cognitive Archaeology, Evolutionary Psychology

and Cultural Transmission, with Particular

Reference to Religious Ideas
Steven Mithen
University of Reading

ABSTRACT

Archaeologists have drawn on evolutionary theory for more than three decades resulting in a diversity
of approaches that lay claim to being a type of Darwinian archaeology. These are briefly reviewed, and
some suggested to be of limited value to the discipline. The paper argues that a valid evolutionary
approach requires an explicitly cognitive perspective, and suggests that there will be considerable ben-
efit from an integration between cognitive archaeology and evolutionary psychology. The result would
be a non-functionalist Darwinian archaeology with potential to throw new light on cultural develop-
ments during the Palaeolithic and more recent times. To illustrate this, the paper discusses the cultural
transmission of religious ideas.

INTRODUCTION as space shuttles and CD players — from the Oldowan

tools of 2.5 million years ago by the application of the
Charles Darwin's theory of evolution by natural same theory? Archaeologists have struggled with such
selection has achieved remarkable success in explaining questions for at least two decades and tried to make use
patterning and variation in the natural world. It provides of evolutionary theory in a variety of ways. In this paper
the theoretical basis and unifying theme for academics I argue that the only respectable evolutionary approach,
working in all areas of the life sciences, ranging from and the only one that will make a contribution to ex-
molecular biology to community ecology. Few challenge plaining past behaviour, is one that adopts an explicitly
the central tenants of Darwin's theory with debate con- cognitive approach.
cerning the relative role of historical contingency in evo-
lutionary process and the value of adaptationism (e.g.,
DARWINIAN ARCHAEOLOGIES
Gould & Lewontin 1979, Gould 1989, Eldredge 1995,
Dennett 1995, Orr 1996). A recent edited book on the use of evolutionary
Without Darwinian theory, the diversity of the theory in archaeology was appropriately titled Darwin-
natural world would appear bewildering. As bewildering, ian Archaeologies, (Maschner 1996) rather than Darwin-
perhaps, as the diversity of the cultural world of humans. ian Archaeology, as even among those who feel commit-
If we document the range of human cultures in the mod- ted to a Darwinian approach, there is no consensus as to
ern world, and supplement that range with those inferred how evolutionary principles can be applied to human
for the past, we find remarkable variation in human behaviour and changes in material culture. Such plural-
behaviour, notably in the material culture that people ism may be a fashionable feature of post-modernism —
make and use. To what extent can Darwinian theory aid there is no single truth — but it is a crippling attitude to
in the explanation of this diversity? Just as the diversity accept when trying to make progress in our understand-
of life forms today are understood by evolution from ing of the past or present. Some of the approaches cur-
some primordial single celled organism, is it possible to rently being pedalled as Darwinian are fundamentally
explain the evolution of modern material — such things flawed and must be rejected.

67
68 Steven Mithen

The most obvious of these is a group adaptationist rying to say the least. After at least two decades of devel-
approach. It was common during the 1970s to read about opment, such models should have moved beyond the
hunter-gatherer groups being adapted to their natural purely theoretical phase and be providing some respect-
environments. By this was meant that their way of life able case studies. The fact that they have not suggests
appeared appropriate for their environments—it was a way that there may be fundamental flaws in cultural trans-
of life that appeared to use energy and information effi- mission models—as indeed there are. One of these is that
ciently. This was indeed a good type of short hand de- such models lack reference to people as intentional
scription. But the term adaptation has no relevance to agents—an essential element of explanation in the hu-
group behaviour: while the use of such group man sciences. Another flaw will be exposed below.
adaptationist terminology may have provided a veneer One variant of cultural transmission models fo-
of science it provided no explanatory power. An indi- cuses not on material artefacts but on the underlying
vidualistic approach is a prerequisite for any evolution- ideas, which are referred to as memes—an explicit anal-
ary approach in archaeology, or indeed any other do- og}' with genes. Associated with this is the notion that
main of the human sciences (Mithen 1989, 1993). some ideas form 'cultural viruses* that are located in hu-
A second manner in which supposedly Darwinian man minds and might spread through a population (e.g.,
theory was adopted by archaeologists was for the devel- Dawkins 1976, Cullen 1993, Dennett 1995). The notion of
opment of cultural selectionist models (e.g., Dunnell 1980, the 'meme' is notoriously ill-defined. Lake (1997) has dis-
Rindos 1984, O'Brien & Holland 1990). Although such cussed and analysed the concept in detail, exploring the
approaches are variable, their basic premise is that the analogy between memes and genes, and asking whether
material culture that an individual adopts is as much a the idea of memes has any value for models of cultural
part of his/her phenotype as his/her physical appear- transmission and archaeological explanation. He takes a
ance and behaviour. Consequently such material culture sympathetic approach, believing that there may be mile-
traits should be as subject to the process of natural selec- age in the concept when applied to symboling behaviour.
tion as any other trait: those which increase the repro- My own judgement is harsher: the concept of a 'meme'
ductive success of an individual will survive and spread as a unit of replication and selection is simply fallacious.
within a population. Those which do not will disappear
from the pool of material culture variants. As with the HUMAN COGNITION AND HUMAN
group adaptationist arguments, such ideas have an intui- BEHAVIOUR
tive appeal and initially appear effective: we are indeed
all followers of fashion and adopt material culture, ei- The ideas of memes and cultural viruses, along with
ther consciously or unconsciously, that appears most that of cultural transmission in general, focus attention
effective at solving the problems we face in the world. on human minds rather than human behaviour. As such,
The problem with such ideas is that the material culture these models are aligned with a further variant of Dar-
of an individual is not an expression of his/her genotype winian archaeology and one I want to explore within
in a particular environment, as are true phenotypic traits. this paper: an explicitly cognitive approach.
There is no reason why the descendants of a reproduc- The premise of this approach is that the behaviour
tively successful individual will necessarily have the same of individuals, of groups, the character of their material
material culture of their ancestor, as in biological evolu- culture, and ultimately the long term patterns of culture
tion. As with group adaptationism, cultural selectionist change we see in the archaeological record, arise from
models are fatally flawed. the short term decisions of individuals (Mithen 1990).
Models of cultural transmission (e.g., Cavalli-Sforza Some of these decisions, perhaps only a small subset, are
& Feldman 1981, Boyd & Richerson 1985) differ from made consciously and with intent. As such, this approach
those of cultural selectionism by avoiding a functional- explicitly rejects ideas that there are long term, and as yet
ist agenda and attempting to explain how maladaptive undefined, processes of culture change, as argued by
traits may survive within human populations. The basis Murrey and Walker (1988). Long term patterns in the
of these models is that cultural transmission is analo- archaeological record are no more than the accumulated
gous to genetic transmission. The most sophisticated results of many individual decisions andpost-depositional
versions are dual inheritance models, which recognise that processes concerned with site formation. Consequently
to explain variation in human behaviour we need to con- to provide adequate explanations for variation in their
sider both genetic and cultural transmission within a data sets and inferred patterns of past behaviour—whether
single model. But the failure of such models to engage they are concerned with the artefacts of the earliest Homo
with archaeological data and to address archaeological or modern material culture-archaeologists have no choice
problems, let alone contribute to their solution, is wor- but to focus on processes of decision making and learn-
Cognitive Archaeology, Evolutionary Psychology and Cultural Transmission. 69

ing by individuals. Some of these processes are concerned of Pleistocene environments and societies indicates that
with social learning, others with learning from an these were characterised by immense diversity: the iden-
individual's own experience. It is the cognitive processes tification of a discrete set of adaptive problems beyond
for these tasks which have been moulded by evolution- the trivial level of 'gaining food' and 'selecting mates'
ary abilities (including natural selection), not behaviour seems improbable.
itself, let alone material culture. Nevertheless there is much to be welcomed in the
Humans are no different in this respect than other development of evolutionary psychology. It does, after
animals. When biologists study the foraging behaviour all, stress the importance of palaeoanthropology to un-
of animals ranging from blue tits to gorillas, they as- derstanding the modem world—and soon evolutionary
sume that this behaviour arises from decision making psychologists will realise that instead of simply making
processes which were shaped by natural selection. Those vague claims about life in the stone age, they will need to
individuals who could make decisions which reduced draw on the work of archaeologists. The basic idea that
foraging time (and hence exposure to predators) or in- our minds are somehow adapted to a hunter-gatherer
creased the rate of energy gain over expenditure, were lifestyle does appear to have considerable value in un-
reproductively more successful. And consequently the derstanding our behaviour today. Why, for instance, do
decision making processes they used spread within the so many people today crave foods high in fat and sugar
population. There is no reason to adopt anything other which cause such serious health problems and conse-
than this conventional evolutionary approach to humans, quently appears highly maladaptive behaviour? Well,
either those living today or long dead in prehistory. possibly it is simply because these foods were in short
supply in our ancestral environments and our desire for
and pleasure in them was at that time a highly adaptive
EVOLUTIONARY PSYCHOLOGY AND
mental traits. But today, when such foods are in abun-
COGNITIVE ARCHAEOLOGY
dance, such mental traits are now seriously maladaptive
As the realisation that archaeologists need a more as our physiology remains adapted to a low fat diet.
explicitly cognitive approach to past behaviour emerged A wide range of issues concerned with human
during the last decade, it was matched by the develop- health today can indeed be illuminated by adopting such
ment of a more explicitly evolutionary approach in psy- an evolutionary perspective (Eaton et al. 1988, Nesse &
chology. Indeed a distinct sub-discipline has emerged Williams 1995). Evolutionary psychologists are making
calling itself 'evolutionary psychology' (Cosmides & many advances in our understanding of human behaviour
Tooby 1987, Barkow et al. 1992). Its basic premise is that today in the realms of inter-personal violence and coop-
the human mind-brain is a product of evolution in pre- eration, social interaction and mate choice by adopting
cisely the same way as any other organ of the human this type of approach (e.g., Daly & Wilson 1988, Barkow
body. As such, to understand how it works today, we et al. 1992, Buss 1994). But evolutionary psychologists,
must understand its evolutionary history. concerned with explaining human behaviour in the mod-
Unfortunately, psychologists adopted a very sim- ern world, have much to learn from archaeologists who
plistic attitude to that evolutionary history, proposing can provide the understanding about our hunter-gath-
the notion of the EEA, the 'Environment of Evolution- erer past and the selective pressures on the human mind.
ary Adaptedness' (Symonds 1979, Tooby & Cosmides 1992). Conversely, archaeologists concerned with explain-
This was argued to be the world of Pleistocene hunter-gath- ing past behaviour have much to learn from evolution-
erers. Consequently all human minds are described in their ary psychologists as they can provide the understanding
work as the mind of a Pleistocene hunter-gatherer, whether of decision making and learning that are required in a
that mind is possessed today by a city dweller, peasant cognitive archaeology. One of their fundamental tenets
farmer, industrialist or modem hunter-gatherer. is of particular importance to an evolutionary archaeol-
It is easy to criticise this idea. It is clear from even ogy: that rather than there being a single learning pro-
the most rudimentary study of comparative psychology cess, there are likely to be multiple modules in the hu-
that there can no such thing as a single EEA. It is readily man mind, each attuned to learning and making deci-
evident that parts of our mental apparatus are shared sions about a specific domain of behaviour.
with all primates, others with the great apes, others with
early Homo while others are unique to our species alone.
MENTAL MODULARITY AND DECISION
In other words these evolved in response to a succession
of different environments and problems during the last MAKING ALGORITHMS
100 million years: there is no single, unitary EEA (Foley To explain the notion of mental modularity, Leda
1996). Moreover, even the most rudimentary knowledge Cosmides and John Tooby (1987) have adopted the meta-
70 Steven Mithen
phor of the mind as a 'Swiss army knife' of specialised may be achieved by reducing the time spent foraging, or
mental adaptations. Each of these, they argue, is a men- improving the rate of energetic return, or avoiding the
tal device specifically for solving a problem faced by our risk of shortfall in food supply or by some other proxi-
ancestors in their EEA. They would include devices for mate goal-our task is to identify what types of goals
selecting food stuffs, recognising predators, and choos- were adopted and the constraints under which they were
ing mates. Cosmides and Tooby (1994) provide powerful achieved As I have demonstrated elsewhere, with a case
arguments for why the mind would have such a design, study exploring Mesolithic foraging in southern
rather than having one or a small number of general Scandinavia (Mithen 1990), it is possible for archaeolo-
purpose problem solving mechanisms. In essence, this is gists to build models based on the foraging decisions of
because each of these problems have a different structure individuals and test these against the archaeological
and to be solved efficiently a unique type of decision record. This involves modelling information exchange
making process is required: what one needs to know between individuals, the formation of the archaeological
when making a choice about foodstuffs, and how one record as well as decision making processes of individu-
should process that information, are different from als themselves.
the knowledge required for choosing mates. An indi- We can do such work because there are strong rea-
vidual with a single decision making process applied sons for believing that human minds have evolved mecha-
to these problems would be out competed by one who nisms for solving foraging problems. But of course the
has specialised mental algorithms for these different types modern mind today, and for much of prehistory, does
of problems. not possess a unique problem solving device for each
These mental algorithms facilitate the learning of problem it faces. Hunter-gatherers in the EEA did not
complex tasks because they are content laden. In other need to make choices about buying cars or computers.
words, a certain degree of knowledge about the world Consequently learning and making decisions about these
appears to be hard wired into our minds at birth. The problems for which there are no specialised mental mod-
classic example concerns language acquisition. It is sim- ules is notably less efficient, and involves much more
ply impossible that in their relatively short exposure to conscious awareness of the decision making process.
spoken words, young children could learn all the rules of Compare, for instance, how children leam the rules of
grammar that they have mastered by the age of three or language and those of mathematics. The first are acquired
four (Pinker 1994). Somehow, these rules are partly al- effortlessly and with very limited awareness. But when
ready encoded into their minds. Similarly, knowledge children need to learn multiplication tables they struggle,
about the physical and natural worlds appears to be en- needing to be formally taught and to invest great effort,
coded as an 'intuitive physics' and 'intuitive biology', even though this set of rules is remarkably simpler than
while children seem to have a vast intuitive understand- those they use in every spoken utterance they make. Pleis-
ing about other minds, as explored in current research tocene hunter-gatherers are unlikely to have needed to know
on 'theory of mind' (Hirschfeld & Gelman 1994, Mithen their seven times table, and the mind today lacks specialised
1996a, chapter 3). Because such mental algorithms are mental adaptations for acquiring such knowledge.
content laden, they make learning about their relevant The significance of this for a Darwinian archaeol-
domains relatively quick and efficient. ogy is that it has serious consequences for models of so-
One of the cognitive domains for which we should cial learning and cultural transmission. Indeed it exposes
expect hunter-gatherers to have specialised, content rich, a fundamental flaw in such models. Within models such
mental models is that relating to the natural world, and as those of Boyd and Richerson (1985), no reference is
more specifically for making foraging decisions. As I have made to the actual content of what is being learnt There
discussed at length, there are good grounds for believing is an assumption that content has no relevance to the
that early humans possessed a distinct domain of'natu- nature of cultural transmission. Those models which have
ral history intelligence'—a bundle of integrated modules addressed the relative significance of individual or social
used for learning and making decisions about the natu- learning in different conditions (eg., Boyd & Richerson
ral world (Mithen 1993,1996a, 1996b). And even within 1996) have failed to consider how the value of these vary
modern human minds, in which natural history intelli- as to what is actually being learnt. Yet the message from
gence is integrated with that of other previously isolated evolutionary psychology (and indeed that from our ev-
intelligences, those content rich decision making rules eryday experience) is clearly that some things are much
for foraging continue to exist. easier to learn than others. The nature of cultural trans-
A consequence of this is that we should expect in- mission for knowledge about social relationships — some-
dividuals to make foraging decisions which ultimately thing which our minds seem to be well tuned to learning
lead to an increase in their reproductive success. This about—will be very different to that regarding the internet,
Cognitive Archaeology, Evolutionary Psychology and Cultural Transmission 71

a recent topic of knowledge about which communication world, making and using artefacts, and interaction with
and thought are far less efficient because we do not have the natural world, were essentially isolated from each
evolved mental algorithms for this domain of activity. other. This isolation explains the rather odd character of
the early Palaeolithic record in which we see evidence for
very complex and sophisticated behaviour within these
ARTEFACTS AND THE CULTURAL domains, but very simple behaviour at the 'domain in-
TRANSMISSION OF RELIGIOUS IDEAS terfaces' (Mithen 1996a, 1996b). For instance, while Ne-
anderthals clearly possessed great technical skill in pro-
While this might be a flaw in the abstract models ducing artefacts such as levallois points, and to have sur-
of cultural transmission and one reason why they have vived in Pleistocene Europe must have had an intimate
made very limited impact on explaining variation in the knowledge of the natural world, the design of their hunt-
archaeological record, the differential ease with which ing weapons appears remarkably simple. The recent dis-
different types of information can be transmitted does covery of 400,000 year old hunting spears at SchOningen,
in fact reveal a new approach to understanding many Germany (Thieme 1997), further indicates the absence
aspects of cultural variability. To illustrate this, I will of technological innovation during the Lower and Middle
briefly consider the cultural transmission of religious ideas. Palaeolithic as these appear as well designed as anything
Explaining the existence of religious ideas from a produced by the Neanderthals. This lack of innovation
Darwinian perspective on human behaviour is clearly a and the absence of multi-component hunting weapons,
challenging task (Dennett 1997). It is a characteristic fea- notably projectiles, arises, I have argued, from an inabil-
ture of religious behaviour that people act in ways that ity to integrate their knowledge of tool making with that
appear clearly maladaptive for their own reproductive of the natural world.
success. People have destroyed their own material pos- The substantial development in cultural behaviour
sessions, sacrificed their own offspring, devoted them- that we see in the archaeological record which begins at
selves to celibacy in the cause of worshipping deities. c. 100,000 years ago, and becomes dramatic after 50,000
Explaining why and how ideas about religious entities years ago derives from a new ability by Homo to inte-
could arise in humans minds is one problem; explaining grate their intelligences, a capacity I have termed cogni-
how such ideas can be so persistent is perhaps even more tive fluidity. This appears to be restricted to modern hu-
challenging. It seems remarkable that during the last two mans, although some traces of cognitive fluidity may be
decades there has been such a rise of fundamentalism present within the minds of the last Neanderthals (Mithen
throughout the world, such that more than 50% of Ameri- 1996a: 209-210). This change in the nature of intelligence
cans apparently believe that humans were divinely cre- is, I suggest, related to changes in the nature of language
ated, when during that period the fossil, archaeological and consciousness (see also the idea of off-line thinking
and genetic evidence for human evolution has become in Bickerton [1996]). Such cognitive fluidity had enor-
so detailed. mous adaptive benefits. By being able to integrate tech-
My own views about the origin of religious ideas nical and natural history knowledge, tools could be de-
have been described elsewhere (Mithen 1996a). In brief signed to markedly improve the efficiency of hunting,
terms, I have argued that these are essentially a spandrel plant gathering and food processing; by being able to
of a change in mental architecture (cf. Gould & Lewontin integrate technical and social intelligence, artefacts could
1979) that arose during the last 100,000 years of human be designed to mediate social relationships providing new
evolution. A spandrel is a 'necessary architectural by-prod- means to manipulate other individuals to ones' advantage.
uct'. Gould and Lewontin argued that many aspects of Yet other consequences of such cognitive fluidity
animal biology and behaviour are spandrels—byproducts have no clear adaptive benefits and can be thought of as
of other features that in themselves lack direct functionalspandrels, inevitable by-products of such adaptations. For
value, drawing an analogy between these and spandrels instance by integrating social and natural history intelli-
found in architectural design (see discussion of this by gence beliefs could arise that entities exist which are half
Dennett [1995] and Houston [1997]). To understand how human and half animal, as clearly evident in the first
religious ideas evolved as a spandrel, we must first con- representational art. And by integrating technical and
sider the mental architecture of early humans, those whose social intelligence, inert objects could be attributed with
minds appear to have lacked any religious concepts. ideas, feelings and intentions, 'living' entities could exist
The early human mincMhat of archaic H. sapiens, which did not need to feed, which were not born and
Neanderthals, H. eredus-was structured, I have argued, could not die. This mixing up of natural categories is the
on the basis of four 'intelligences', or bundles of mental essence of a supernatural being (Guthrie 1993, Boyer
modules. Three of these, those concerning the social 1994a, 1994b, 1996; Mithen 1996c).
72 Steven Mithcn

The human mind does not, therefore, have an transmitted in a context of ritual-the rote repetition of
evolved module/domain for supernatural beings or in- movement and utterances in sequences that must be con-
deed any types of religious knowledge. This inhibits the formed to precisely. Such ritual is essential: religious ideas
cultural transmission of religious ideas. When an indi- cannot be transmitted in an informal manner if people
vidual is told about a social relationship between two are to share religious concepts because there is no evolved
other humans, that information is embedded into con- domain of religious ideas within the human mind. With-
tent rich modules about human social relationships al- out ritual, religious ideas might exist within individual
lowing many inferences to be drawn from a limited minds but a religious institution, based upon shared reli-
amount of information (Boyer 1994a). If we are simply gious concepts, would be impossible.
told that those individuals are a boy and a girl and that A second means by which the cultural transmis-
they are 'in love' we can accurately guess how they are sion of religious ideas is achieved is even more funda-
likely to be interacting with each other, what they will be mental for our understanding of human behaviour: the
doing, their feelings, how they will react in a host of use of material culture. The last century of Palaeolithic
circumstances. But if we were told about someone who archaeology has supported Durkheim's (1915: 307) as-
was 'in love' with a supernatural being, no such infer- sertion that "the principle forms of art seem to have been
ences could be drawn. Perhaps we are told that super- born out of religious ideas" (here I reject the idea that
natural being is invisible but exists in all places, that he there are 'art objects' prior to those of the Upper
once took human form and walked on water, that he Palaeolithic, finding the arguments of Bednarik [1995]
died and then came alive again. Well, such details are of for 'concept mediated marks in the Lower Palaeolithic'
little help. What can it mean to love such a being? How unconvincing [Mithen 1996d]). Throughout human his-
is a person supposed to communicate with or behave tory religious behaviour has involved visual symbols. Why
towards such a being? should there be such a close connection between the two?
So the cultural transmission of religious ideas is Well, as Leach (1976) argued, we transform religious ideas
difficult when compared to ideas about an evolved do- into material form so that we can perform operations on
main of human behaviour, such as social interaction or them which are beyond the capacity of the mind. The
(for hunter-gatherers) animal behaviour. Boyer (1994a) evolutionary understanding of the emergence of religious
has discussed this and noted that those religious ideas ideas that I have summarised above explains why this is
which survive the rigours of cultural transmission are necessary: religious symbols, and more particularly the
those which have a link to a domain of intuitive knowl- images of religious beings, serve to anchor religious ideas
edge. As he has recognised, while concepts of supernatu- within the mind. Ideas about social relationships, the natu-
ral beings have, by definition, elements which are 'super ral world, and stone artefacts did not need anchoring in
natural', such as abilities to be omnipotent, invisible or hunter-gatherer minds as each of these related directly to
ever-lasting, they also frequently have human like fea- an evolved domain of mental architecture which made them
tures, such as suffering jealously and desires. The gods of easy to leam, understand and transmit. Religious ideas had
ancients Greece provide a typical example — supernatu- no such domain and the archaeological record of modem
ral beings who quarrel and deceive each other in a very humans is replete with religious symbols.
human-like manner. By having these human like quali-
ties, concepts of supernatural beings can be more easily
transmitted and understood, than if all features were SUMMARY
supernatural and unable to be grasped by any domain of
intuitive knowledge. I have chosen to focus on religious ideas and
As archaeologists we can never reconstruct the spe- behaviour in this paper because this domain of human
cific ideas that past people held about their religious be- activity is the one which most clearly lacks an adapta-
ings, although we may be able to invoke what appear to tionist explanation but is one which can only be under-
be universal features of religious beings to suggest what stood from an evolutionary perspective. The ability for
ideas may have been present (Mithen 1997). But the dif- humans to create ideas about supernatural beings is a
ficulty of transmitting religious ideas also has enormous product of the emergence of cognitive fluidity between
implications for the archaeological record which cannot 100,000 and 40,000 years ago (Mithen 1996a). This cog-
be fully understood without understanding the human nitive fluidity, the ability to integrate ideas and ways of
mind as a product of evolution. There are two cultural thinking from what had been isolated domains of
means which are widely used to facilitate the transmis- thought, evolved due to the selective advantage it gave
sion of religious ideas and which have major impacts on those individuals who could do things such as design
the archaeological record. First, religious ideas are often better hunting weapons, or use material culture to medi-
Cognitive Archaeology, Evolutionary Psychology and Cultural Transmission 73

ate social relationships. But by having a cognitively fluid chaeology is for an explicity individualistic and cogni-
mind, a whole host of new ways of thought lacking in tive stance to be adopted The human mind is a product
any functional significance were also made possible — of evolution and as a consequence for those domains of
such as believing in beings which were half human/half behaviour which were of significance during the evolu-
beast. Transmission of such ideas required cultural sup- tion of the mind, we should expect people to make deci-
port, such as visual symbols and ritual to anchor them sions which lead to behaviour that increases their repro-
in the mind as they have no evolved domain of their ductive success. This provides a means for building models
own. Consequently, unless we understand the human of human behaviour which can be used to explain the
mind as a product of evolution we will be unable to un- variability and patterning in the archaeological record.
derstand religious ideas and their transmission. And of Other domains of behaviour and thought are spandrels
course religious ideas are just one type of thought made of our evolved mental architecture and will defy adapta-
possible by cognitive fluidity. Another is pursuit of pure tionist explanations. Nevertheless their existence can only
science—investing time in discovering laws of mathemat- be understood from an evolutionary perspective and cul-
ics or the origins of the universe or of modern humans- tural transmission theory may be able to explain the per-
activities of no adaptive benefit but which can be under- sistence of ideas with no functional utility.
stood as a spandrel of our evolved mental architecture.
Other types of ideas can be transmitted without
REFERENCES CITED
cultural support as they directly relate to a domain of
intuitive knowledge. If we are dealing with hunter-gath- Barkow.J.H., Cosmides, L & Tooby, J.
1992 The Adapted Mind. Oxford- Oxford University Press.
erers, ideas and information about social relationships,
the natural world and technology are of this nature and Bednarik, R.G.
1995 Concept mediated marking in the Lower Palaeolithic. Cur-
their minds have content rich, mental modules for mak- rent Anthropology 36: 605-634.
ing decisions within these domains. We would expect such Bickerton, D.
decisions to make adaptive sense and ultimately to im- 1996 Human Behaviour and Language. London: University College
prove the reproductive success of the individuals involved. Press.
This is fortunate for archaeologists as data relating to Boyer, P.
past subsistence behaviour is often the most readily avail- 1994a The Naturalness of Religious Ideas, A Cognitive Theory of Reli-
gion. Berkeley: University of California Press.
able in the archaeological record and can provide a means
1994b Cognitive constraints on cultural representations: Natural
to test models for decision making in foraging strategies. ontologies and religious ideas. In Mapping the Mind; Domain
All people, either in the past or present, those who Specificity in Cognition and Culture (LA Hirschfeld & S A
lived by hunting and gathering, by agriculture or by in- Gelman, eds.), pp. 391-411. Cambridge: Cambridge Univer-
sity Press.
dustry, live in a social world and have an evolved do-
1996 What makes anthropomorphism natural: intuitive ontology
main of social intelligence. And consequently we should and cultural representation. Journal of the Royal Anthropologi-
also expect decisions about social relationships, such as cal Institute!: 83-97.
the choice of friends and lovers, to also make evolution- Boyd, R. & Richerson, P.
ary sense. This seems to be born out by the cross-cultural 1985 Culture and the Evolutionary Process. Chicago: Chicago Uni-
studies of Buss (1994) on mate choice, and on homicide versity Press.
victims by Daly and Wilson (1988). Of course, evolu- 1996 Why culture is common, but cultural evolution is rare. In
Evolution of Social Behaviour Patterns in Primates and Man (W.G.
tionary sense is not necessarily adaptive sense: our deci- RuncimanJ. Maynard-Smith &R.I.M. Dunbar, eds.), pp. 77-
sion making apparatus evolved for life in small scale, 83. Proceedings of the British Academy, Vol. 88.
highly mobile hunter-gatherer societies. We use the same Buss, D.
apparatus when living in our urban societies today and 1994 The Evolution of Desire: Strategies of Human Mating. London:
consequently may make decisions of little adaptive sense Harper Collins.

but which can only be explained by an evolutionary un- Cavalli-Sforza, L.L. & Feldman, M.W.
derstanding of the human mind. 1981 Cultural Transmission and Evolution: A Quantitative Approach.
Princeton: Princeton University Press.
In summary, evolutionary theory is of vital im-
Cosmides, L. & Tooby, J.
portance to archaeologists, and indeed to any one study- 1987 From evolution to behaviour: evolutionary psychology as the
ing human behaviour. But notions of group adaptation, missing link. In The Latest on the Best: Essays on Evolution and
cultural selection, memes and cultural viruses should be Optimality (J. Dupre^ ed.), pp. 277-306. Cambridge: Cambridge
University Press.
rejected. Models of cultural transmission must take into
1994 Origins of domain specificity. The evolution of functional
account the fact that some types of information are easier
organization. In Mapping the Mind: Domain Specificity in Cog-
to learn and transmit than others. The most profitable nition and Culture (LA. Hirschfeld & S.A. Gelman, eds.), pp.
manner in which evolutionary theory can be used in ar- 85-116. Cambridge Cambridge University Press.
74 Steven Mithen
Cullen, B. Leach, E.
1993 The Darwinian Resurgence and the cultural virus theory. 1976 Culture and Communication. Cambridge Cambridge Univer-
Cambridge Archaeological Journal 3:179-202. sity Press.
Daly, M. & Wilson, M. Mithen, S.
1988 Homicide. New York; Aldine de Gruyter. 1989 Evolutionary theory and post-processual archaeology. Antiq-
Dawkins, R. uity 63: 483494.
1976. The Selfish Gene. Oxford; Oxford University Press. 1990 Thoughtful Foragers: A Study of Prehistoric Decision Making. Cam-
bridge: Cambridge University Press.
Dennett, D.
19 95 Darwins Dangerous Idea. New York: Simon and Schuster. 1993 Individuals, groups and the Palaeolithic record: A reply to
1997 Appraising Grace: What evolutionary good is God? The Sci- Clark. Proceedings of the Prehistoric Society 59: 393-398.
ences Jan/Feb: 39-45. 1996a. The Prehistory Of The Mind: A Starch For The Origins Of Art,
Dunnell, R. Science And Religion, London: Thames and Hudson.
1980 Evolutionary theory and archaeology. inAdvances in Archaeo- 1996b Domain specific intelligence and the Neanderthal mind In
logical Method and Theory, voL 3 (M.B. Schiffer, ed.), pp. 35- The Early Human Mind (P. Mellars & K-Gibson, eds.), pp.
99. New York: Academic Press. 217-229. Cambridge: McDonald Institute for Archaeological
Research.
Durkheim, E.
1915 The Elementary Forms of Religious Life. London: Allen and 1996c Putting anthropomorphism into evolutionary context (a com-
Unwin. ment on Boyer \996).Journal ofthe RoyalAnthropological Insti-
tute 2:717-719.
Eaton, B. et aL
1988 Stone Agers in the fast lane. American Journal ofMedicine 84: 1996d Early Palaeolithic 'concept mediated marks', mental modu-
739-49. larity and the origins of art Current Anthropology 37-. 666-670.
Eldredge, N. 1997 The Supernatural Beings of Prehistory and the external stor-
1995 Reinventing Darwin.. New York: John Wiley and Sons. age o f religious ideas. In The Archaeology of External Symbolic
Storage. (C. Renfrew & C. Scarre, eds.). Cambridge: McDonald
Fodor.J.
Institute for Archaeological Research (in press).
1996 Deconstructing Dennett's Darwin. Mind and Language 11:
246-262. Murrey, T. & Walker, M.J.
1988 Like WHAT? A practical question of archaeological infer-
Foley, R.
ence and archaeological meaningfulness. Journal ofAnthropo-
1996 An evolutionary and chronological framework for human
logical Archaeology 1: 248-287.
social behaviour. In Evolution of Social Behaviour Patterns in
Primates and Man (W.G. Runciman, J. Maynard-Smith & Nesse, R. & Williams, G.
R.I.M. Dunbar, eds), pp. 95-143. London: Proceedings of the 1995 Evolution and Healing: The New Science ofDarwinian Medicine.
British Academy, Vol. 88. London: Phoenix.
Gould, SJ. O'Brien, M. & Holland, T.
1989 Wonderful Lift- The Burgess Shale and the Nature of History. 1990 Variation, natural selection and the archaeological record In
Norton: New York. Advances in Archaeological Method and Theory, vol. 2 (M.B.
Schiffer, ed.), pp. 31-79. Tuscon: University of Arizona Press.
Gould, S.J. & Lewontin, R.
1979 The spandrels of San Marco and the panglossian paradigm: Orr, H A
A critique of the adaptationist programme. Proceedings of the 1996 Dennett's Dangerous Idea. Evolution 50: 467-72.
Royal Society B205: 581-98. Pinker, S.
Guthrie, S. 1994 The Language Instinct. New York; William Morrow and Com-
1993 Faces in the Qouds: A New Theory ofReligion. Oxford: Oxford pany, Inc.
University Press. Rindos, D.
Houston, A. 19 84 The Origins of Agriculture An Evolutionary Perspective. New York:
1997 Are the spandrels of San Marco really panglossian pendentives? Academic Press.
Ecology and Evolution 12: 125. Symonds, D.
Hirschfeld, L A & Gelman, S.A. 1979 The Evolution ofHuman Sexuality. Oxford Oxford University
1994 Mapping the Mind. Cambridge Cambridge University Press. Press.
Lake, M. Thieme, H.
1997 Digging for memes: The role of material objects in cultural 1997 Lower Palaeolithic hunting spears from Germany. Nature 385:
evolution . In The Archaeology of External Symbolic Storage (C. 807-810.
Renfrew & C. Scarre, eds.). Cambridge: McDonald Institute Tooby, J. & Cosmides, L.
for Archaeological Research (in press).
1992 The psychological foundations of culture In TheAdaptedMind
Maschner, H. (ed) (J.H. Barkow, L Cosmides &J. Tooby, eds.), pp. 19-136. New
1996 Darwinian Archaeologies. New York: Plenum Press. York: Oxford University Press.