ArticlePDF Available
Ann. Rev. Anthropol. 1986. 15:25-66
Copyright ? 1986 by Annual Reviews Inc. All rights reserved
Linda Marie Fedigan
of Anthropology, University of Alberta, Edmonton, Alberta T6G 2H4,
Imagine three anthropologists:
A primatologist
observes a female chimpanzee fashioning several crude
tools from grass stems, which she will use to fish termites from an un-
nest over many hours, in the presence
of her sometimes
sometimes impatient offspring. An ethnographer
lives with a group
of human
foragers at one of their campsites on the edge of a waterhole, recording
detail the daily patterns
of adult women and men as they go about their
preparing food, caring
for their
children, enjoying
with their
An archaeologist
and a team of bone hunters
fan out across an escarpment slowly descending the years, squatting every
now and then to peer and scratch
at the surface; they walk and look
and listen for the call that will signal a "find."
Probably the primatologist
and ethnographer
would quite properly deny
that the objective of their research was the reconstruction
of the lives of our
earliest human ancestors.
The latter
is there first and foremost
to understand
the lives of these contemporary
human beings before their way of living
disappears entirely, and the former works to explicate the animal
species for
itself, another
life form
in danger
of disappearing
before we can understand
Nonetheless, the information
obtained from all these studies will be gathered
up, if not by the original researchers,
then by others, and woven into a
scientific story of the origins and evolution of early human behavior. For we
have a powerful urge to know our origins-scientists and
public alike-allied
to a strong cultural imperative to justify our present social arrangements
through reference to historical precedents.
And what more significant
guide to
the structure
of our own underlying nature could we discover
than the original blueprint
for human society?
That is why the practice of modeling the life of early humans, although
shunned by many anthropologists,
is nevertheless a scientific game played
with great determination;
its reward
is the right
to propound a view of human
nature. Some of these models are widely disseminated, in high school and
college textbooks, in popularized
scientific writings, in fiction, on film. And
in a society which tends to believe that what is natural
is good, or at least
acceptable, "scientific"
about the original
of human
represent applications of data which even those who disapprove of such
modeling can ill afford to ignore.
In this review, I take one aspect common to models of early hominid
namely, the reconstruction
of sex roles, examining
in particular
the part that
women are seen to have played in human
society and
in the evolution of those
that distinguish
us from our primate
relatives. The title of this
review allows me to examine not only how the perceptions
of women's roles
in human evolution have changed, but also to describe how women lately
have come to play a part
in the very construction of models of their origins.
As anthropologists, we might have expected that women, with their dis-
tinctive life experiences, would have origin stories
to tell that would differ in
significant ways from those of men.
This review begins with an historical overview of the more influential
models, from Darwin's ideas in 1871 (15) to those of Lovejoy in 1981 (71).
Then I attempt to dis-articulate the models for an examination of their
significant parts by discussing
the major
of evidence and/or
analogy for early human social life: the comparative
data from studies of
primate societies; the indirect data from contemporary
human foraging
societies; and the archaeological
evidence drawn from the
material remains of our ancestors. Throughout
these sections, I also make
reference to the cultural
about the appropriate
behavior of men
and women that inform
our theories. The final section suggests how we may
improve our ability to reconstruct an early human
society that is more than a
of current cultural beliefs and practices.
evolution as a concept
was in use by social philosophers
historians long before Darwin's time, it was in scholarly treatises of the
second half of the nineteenth
that the idea of gradual, adaptive
came to be widely applied
to the place of humans and human
societies in the
world. Evolutionary
models became something of a fashion among
European and North American
scholars, including
those interested in explain-
ing the social nature of humans, as well as those more concerned with the
biological nature
of humankind.
Biological Evolution in the Nineteenth Century
After publishing The Origin of the Species in 1859 (16), which set out his
theory of, and evidence for, natural selection, Darwin was left with several
puzzles. Two of these were: the explanation
of secondary
sexual characteris-
tics in a wide range of species, and the extent to which evolutionary theory
could be applied
to human behavior and biology. He set out to explore both of
these topics in his 1871 book, The Descent of Man and Selection in Relation
to Sex (15). The book thus has two intertwining objectives:
the development
of a theory of sexual selection applicable
to the entire animal
kingdom, and
the establishment of the human
species as subject
to the laws of both natural
and sexual selection. In the process of demonstrating
that the characteristic
features of the human
phenotype and the human
way of thinking
and living
show rudimentary
similarities with those of other animals, Darwin also
provided sketches of his own view of early human life. Especially in the
course of discussing the application
of sexual selection theory to humans, he
provided us with a clear picture of how he saw the roles and the in-
of men and women in human
society. First let us look briefly
at Darwin's conclusions on sexual selection and the human
place in nature,
the two platforms
on which he was to build his scenario
of early
human social
Darwin's reasoning was that secondary
sexual characteristics, which nei-
ther are directly
for reproduction
nor for survival,
were the result
two types of interactions
involving the sexes: competition
and choice. Com-
petition, Darwin believed, generally occurred between males for access to
female mates, and
choice, he reasoned,
was exercised
by females from
the male mates available to them. Thus, certain traits in males which en-
hanced their ability to win in competitions and/or to be chosen by females
were sexually selected. It seemed obvious to Darwin that sexual selection had
occurred in humans, because he believed the human male to be more
courageous, energetic, inventive, pugnacious, and sexually assertive than the
female. The human male is also bigger than the human female, because,
Darwin argued, in primitive times men fought to the death for access to
women, and in modern
times his size advantages
are maintained because he
has to work harder than woman for their
joint subsistence. Women are more
more reclusive, and more altruistic than men, traits which occur
because of the lack of selection for the assertive, selfish male traits listed
above, and also because of an extension of "maternal
instincts" toward
members of the group as well as toward infants.
Several authors (e.g. 25, 77, 108) have pointed
out that Darwin
onto the large screen of nature
his own images of appropriate role behavior for
men and women, images which were clearly drawn from upper-class Victo-
rian culture
in Britain
in the 1800s. Not so often pointed
out (but see 125) are
certain inconsistencies in the conceptualization and application of sexual
selection theory itself. For
example, Darwin saw selection as operating almost
entirely on males. Competition
selected for male armaments
(size, strength,
weapons) and choice selected for male ornaments (colors, elaborate head-
dresses, beautiful voices). Females of the species were seen to be, as a general
rule, similar
in appearance
and behavior to juveniles, their traits
the absence of sexual selection. Darwin weakened his principle of female
choice by equivocating about
the actual
of females to exercise choice in
which males would mate. At times he thought
females had the
selective power to bring about elaborate male features such as the peacock's
tail. At other
times he thought
females could do no more than accept the
least distasteful
male available, or accept
the winner
of a previous male-male
competition, a lack of selective power which elsewhere I have likened to
"Hobson's choice" (24).
Having already equivocated about the power of female choice to bring
about sexual selection in animals, Darwin then contradicted
himself when
applying the theory to humans. For Darwin believed that the human
was sexually selected by males. Since this is the opposite of his principle
female choice, it is odd that Darwin argued repeatedly
that men in various
societies around
the world exercise choice among possible female mates on
the basis of the latters'
and behavior. He did seem to believe that
female choice had operated
on human progenitors,
but apparently
at some
point in human evolution he saw the process reversing. The human species
appears to be the only one for which Darwin argued that males presently
exercise both the mechanisms of competition and of choice, although no-
where does he discuss the matter
of how or why the process of intersexual
selection reversed, with choice as well as competition becoming the pre-
rogative of the human male.
Darwin's second objective in The
Descent of Man was to demonstrate
many human
then thought
to be unique, had simple analogs
in other
animal species. Thus, he spent two chapters discussing the evidence for
rudimentary beginnings
of higher
powers in animals: faculties of mind
such as reasoning, imagination, aesthetics, ability to produce material ob-
jects, and religious beliefs. He also argued that humans shared with many
animal species the "social instincts": desire for company, sympathy
for others
in the social group, altruism,
love of praise and fear of blame. However, the
most important
to distinguish early humans from animals
was a
sense of morality. Once humans
had developed the "self-regarding virtues,"
which Darwin
saw as self-control
and awareness
of good and
evil, they began
to develop societies based on higher mental faculties than those of other
animals. He believed that early men developed their tool-making skills to
produce weapons and to become efficient hunters. They also began to
accumulate property which helped to bring about social stratification.
alleviate sexual jealousy and because of their ability to exercise self-control,
marriage practices were instituted which would regulate sexual behavior,
of women. In some societies, powerful men could take more than
one wife.
In sum, Darwin suggested that early humans lived in small hunting
munities made up of monogamous or polygynous units. Before cultural
such as infanticide
were introduced,
which he thought
would coun-
ter the effects of natural and sexual selection, these biological processes
selected for courageous, intelligent, tool-using men. In the absence of an
of how traits are biologically transmitted
to the next genera-
tion, Darwin used a concept he called "equal
transmission of characters" to
explain how women were not left totally behind in the process of human
evolution. In this way, Darwin helped to pioneer what I call the "coat-tails"
theory of human
traits are selected for in males and women evolve
by clinging to the men's coat-tails. This model became, and remains, the
image of human
evolution, though rarely so candidly stated as
by Darwin:
Thus man has ultimately
become superior
to woman. It is indeed fortunate
that the law of
equal transmission
of characters
to both sexes prevails with mammals. Otherwise it is
that man would have become as superior
in mental endowment to woman as the
peacock is in ornamental
plumage to the peahen (15, p. 874).
Social Evolution in the Nineteenth Century
In the nineteenth century, the biological and social sciences were not the
widely separate fields built on often incompatible
that they are
today. There was great
and cross-fertilization
of ideas between those
largely in human
biological nature
and those interested
mainly in
human social evolution. All of the writers discussed in this section were
with Darwin, and most of their
were published
The Origin of the Species but before The Descent of Man (e.g. 3, 72, 74,
81, 82). Thus, some of these scholars such as McLennan are
widely quoted
the human
behavior sections of The Descent of Man, and must have had an
impact on Darwin's view of early human society, even though they did not
the same understanding
of "evolution"
(see below). On the other
hand, a
few of the writings
in this section (21, 105) appeared
Descent of Man and were clearly influenced by the latter. Perhaps it is
because of such cross-fertilization of ideas that many of these multiple
schemes of human social evolution seem to be variations on a single theme.
In essence, the theme of the nineteenth century
social modelers was that all
societies pass through
a series of stages which represent
and social progression from an initial primitive aggregation to the final
civilized state. Furthermore, most of these scholars
believed that contempo-
rary societies of the world are at various stages along the path toward
common goal of civilization, and therefore
they could be used as representa-
tives of landmarks
along the way in a reconstruction of the human social
journey. It is important
to note at this point that such a view of the evolution
of societies is different
from Darwin's view of the evolution of species, and
may in fact have little in common other than a concept
of "change over time."
the principles
of natural and sexual selection are irrelevant to these
models of social evolution, whereas progressive change (only episodically
implicit in Darwin's works) is the leitmotif of nineteenth century social
evolution theory.
there were variations,
a common concern of the authors was the
increasing regulation
of human
as societies progressed
more complex technological, political, economic, and
kinship systems. Since
this concern with the regulation
of sexuality directly
reflected their
of male and female roles, it is upon
this aspect
of the models that
I shall focus.
All of these scholars, except Maine, believed that the original
human societies
were promiscuous. (Interestingly, Darwin, anticipating
opinion, ex-
pressed doubt about this assumption because he believed that no known
society or nondomestic
society is totally promiscuous,
even if
individuals mate with multiple
This initial stage of promiscuity
followed by a universal matrilineal
stage, which in turn was followed by the
present patrilineal stage. During the matrilineal
stage, the only kinship ties
that were recognized
were those of women to their
children, so that what we
would now call "matrifocal" units prevailed. It was because many of these
scholars believed that
or "consanguineous
were occurring
the matrilineal
stage that they concluded the fathers' relationships
to their
offspring would not have been recognized.
There was considerable
confusion in these works and in many subsequent
of them between matrilineality, or reckoning of descent
the female line, and matriarchy,
or rule by older women. Bachofen
(3) used ancient myths to argue that women had dominated society in its
earliest stages and were later to lose power. Morgan (82, 83) used his
extensive knowledge of the Iroquois
to argue
that women in the promiscuous
and matrilineal
stages were either
equal to, or dominant
over, men, and were
in control of sexual relations, descent, and property.
He believed that these
forms of female power were lost as societies evolved toward
civilization (see
discussion in 77). However, for the other
(72, 81, 105) the story of
social evolution was not one of the decline of female power during the
evolution of societies, but rather
of the rise of female prestige. In particular,
they argued that
when matrilineality
was overthrown
by patrilineality
and the
monogamous marriage, women were finally and rightfully protected and
by individual men. Women were thus able to give up unseemly
labor and overt sexuality in the public domain and retreat
to their
"natural," socially valued domestic functions.
Some years after the publication
of these works and of Darwin's
Descent of
Man, Engels (21) was to reinterpret
extensive work
on the Iroquois
in order
to argue
that early human promiscuous
and matrilineal
societies did
offer greater
social power and prestige
to women, that
these earliest societies
were in fact socially egalitarian,
the opposite
conclusion to that
of McLennan,
and Spencer. According
to Engels, it was only with the invention
that the accumulation
of property
became important
to men, and
descent systems were instituted
to afford men greater
the disposition of their
of which their wives and children
became a
Respective Fates of These Early Models
Darwin's views on the evolution of human behavior were reinterpreted
Herbert Spencer
to support
his views of appropriate
political action in Britain
in the late 1800s. A coverage of social Darwinism
is beyond
the scope of this
review. However, it is fair to say that whereas Darwin's ideas on the
biological mechanisms
of evolution
the plant
and animal
were to have continual and increasing influence on the life sciences of the
century, his ideas specifically on human
social and racial evolution
were largely dropped
or forgotten.
Even when aspects
of his thoughts
on early
in modem models, his work often is not cited and
not remembered.
The ideas of the social evolutionists suffered a more severe fate than mere
neglect. In the first half of the twentieth century, the Boasian school of
historical particularism
took it as part of their mandate to discredit the
methodology, the data, and the conclusions of the nineteenth
century social
evolutionists. The teleology and ethnocentrism
of these models (that all
societies are progressing
one goal, represented
by European
tions) was particularly
offensive to a discipline founded
on cultural
the nineteenth century modelers had worked largely in the
absence of good ethnographic
data. With the rise of extensive field research
the early twentieth
century (much
of it conducted on nonindustrial societies by
Boas and his students),
it became increasingly
obvious that the earlier
had relied on incomplete and often incorrect data. Although there have
to be some evolutionists
among social anthropologists,
the days of
"modeling" early human societies from a social scientist's perspective
ended at the turn of the century.
Engel's work had a somewhat different
fate. Recall that he had drawn the
opposite conclusion about female status from that of several of his con-
namely that the amount
of labor which women put into subsis-
tence is directly
with their social power and prestige. Although
work has been largely ignored
in Western
it has been taken
by Russian anthropologists,
and by a few American women anthropologists
such as Leacock (59, 60) and Reed (93), who continue to theorize on the
evolution of women's status
in relation
to productive
and in the context
of hypothesized evolution in social organizations.
Overview of Twentieth
For the first half of the twentieth century, sociocultural anthropologists
labored mainly in an effort to collect vital information on nonindustrial
societies before the latter transformed
entirely under the impact of contact
with colonizing or emergent nation-states. Various theories of sociocultural
such as functionalism
and structuralism
emerged, and social evolu-
tion remained
very much out of favor. Physical anthropologists
for their part,
largely under the influence
of Ales Hrdlicka,
also occupied themselves
with data collection, primarily
in the area
of anthropometry.
Although occa-
sional sparks of interest in human social origins appeared
time, it was not until the 1960s that a strong interest was rekindled among
physical and some social anthropologists.
When models of human social evolution and origins began to reappear
widely (28, 65, 96, 101, 102, 113, 115-117), they shared one powerful
"Man the Hunter."
The lines of thought,
drawn from the accumulating
of the first half of the century, by mid-century
to have converged into a strong
focus on one distinguishing
human trait: the
pursuit, killing, and eating of animals with the use of tools. The most
influential and widely quoted
of this new model was undoubtably
& Lancaster's
1968 paper
on the "Evolution of Hunting"
(117). In
it they argued
all those qualities
of human
man so sharply from the other primates. Thus, although
the exact
sequence of events varies in the different versions
that were to follow (e.g. 2,
11, 44, 87, 113), the hunting
model was premised
on the idea that this means
of procuring
food was the catalyst for all of the technological, social, and
of human
beings. Just a short
list of traits believed
to have resulted
from hunting (which was said to be not simply a subsistence
technique, but a way of life) would include: bipedalism, elaborate
tool kits,
development of language, appreciation
of beauty, male aggressiveness and
in killing, division of labor, the nuclear
family, loss of
female estrus, the invention of incest taboos, and bonding between males.
Furthermore, Washburn
and Lancaster argued that the killing of animals
with tools dominated
human history for such a long time that it became the
shaping force of the human psyche for all time, even when men no longer
hunt for a living. This argument was repeated
in so many articles and intro-
ductory textbooks that it took on something akin to the status of a received
the Washburn
and Lancaster
was later to be singled out for
both emulation and criticism, it was by no means the most extreme statement
of the hunting hypothesis. For example, it only traced human hunting patterns
(by which the authors clearly meant
big game hunting)
back to the beginnings
of Homo erectus (dated at 600,000 years B. P. when their article was written),
whereas most others extended the hunting argument
back to "99%" of the
entire 2-3 million years of hominid evolution. The latter
was in spite of the
lack of any paleoanthropological
evidence of hunting at these early dates.
Indeed, if one takes the parsimonious
view that hunting can be said to be a
common activity only when an extensive hunting
technology is found, which
the Oldowan and Acheulian tool industries
clearly were not, then we must
wait until the Upper Paleolithic for the first incontrovertible evidence of
societies. Furthermore,
Washburn and Lancaster
that "gathering"
as a means of procuring
food also set humans apart from
other primates, and that "receptacles
for carrying vegetable products may
have been one of the most fundamental advances
in human evolution"
p. 297). In contrast, Debetz (17) denied the possibility of gathering having
played any role in human
evolution, and most authors of the two influential
of the time, Man the Hunter (65) and The Social Life of Early
Man (115), simply failed to mention means of procuring
food other than
The picture of human sex roles that emerges from the hunting models is
altered in metaphor, but is little changed in essence from that drawn by
Darwin a century earlier. Men are still seen as actively and aggressively
engaged in procuring
food and defending
their families, whereas women are
seen as dependents, who remain close to home to trade their sexual and
reproductive capacities
for protection
Some authors such as
Sahlins (96) retained
Darwin's concern over the control of human sexuality,
which at least implied a consideration
of two sexes. However, many of the
evolution models of the 1960s, premised
as they were on the idea that
is the master behavior
of the human
species" (37, 58), and
assuming that women do not participate
in hunting, effectively omitted the
female half of the human species from any consideration whatsoever.
In retrospect, there are two significant peculiarities
of the book Man the
Hunter. Based on a symposium that gathered together ethnographers
the world
to exchange
information and ideas on foraging societies, this
volume stands as a landmark
for studies of contemporary
foraging peoples
and as a sourcebook for Man the Hunter models. The first peculiarity is that
the participants
were unable to agree on a definition of hunting (see 18, p.
281; 66, p. 4; 97, p. 341), a failure which could not help but weaken any
resultant theorizing, and which inevitably led to later disagreements
generalizations about the importance
of hunting in human foraging patterns
(e.g. 20). The second rather odd aspect of the book concerns
its title and its
ostensive promotion of the hunting model. For it was the very same
collected on modern
and the in-
terpretations made by the collectors for this volume, which were to turn the
minds of many researchers away from hunting as a central humanizing
activity and toward alternative
of human
origins, that is, to the
significance of human
gathering, carrying, and sharing
of mainly vegetable
foods. For example, Lee (61) argued
in a paper
hunters do for
a living," that plant and marine resources
are far more important
than game
animals (i.e. "hunters"
gather for a living); and Deetz (18) cautioned of
hunting that we must not let the label overdescribe the subject. From this
perspective, the papers in Man the Hunter, championing as they did the
explanatory power of hunting, also provided the insights and the data that
were to lead to its undoing.
In 1971, Sally Linton published a paper entitled, "Woman
the Gatherer"
(70), in which she pointed out various shortcomings
and examples of an-
drocentric bias in the Washburn and Lancaster
paper, and then drew on a
variety of sources to develop a model of early hominid females gathering,
and sharing
foods with their
young. It seemed to her that these three
by hominid
females would have been a logical extension
the intense mother-infant bond found in all primates,
and she suggested that
the first cultural inventions were containers to hold the products
of gathering
and the infants. According to Linton, the hunting
of large animals
by males
was a late development, after the matrifocal sharing-family
was well es-
tablished. She argued that the first hunters shared food not with sexual
but with their mothers and siblings who had shared with them. Such
a scenario would obviously set human sex roles on a very different
from the "male as husband and
model that has come down
to us from Darwin. Men would still hunt and women would still gather, but
sexual bonds and sexual exchange would not be the cornerstone of society,
and the activities of women as autonomous
in society would play
for almost the first time a significant
in the story
of how we evolved those
traits that make us uniquely human.
Linton's ideas, only generally
sketched out in her essay, obviously struck a
with a number of women anthropologists,
because several of them (13,
14, 33, 53-55, 77, 108, 109) began to focus simultaneously
on the question
of what women did in early human societies.
Zihlman (109, 123-133) produced
a series of elaborations
on what came to
be called the "Woman the Gatherer"
model, in which she stressed that
obtaining plant food with tools was the "new" or catalytic event in human
evolution. She argued that bipedal locomotion and the invention of carrying
devices first enabled women to walk long distances with babies in slings in
order to exploit the resources of the more open savannah areas, and to carry
these gathered plant foods back to safer familiar areas for shared consumption
with their children. Plants and not meat were the focus of technological and
social innovation for the emerging hominids three million years ago, and
females, ever responsible
for the nourishment of themselves and their
were the providers
and the inventors. It might be said by critics that males
have now become the inconsequential
sex in the story of our origins, because
they may bring
in meat, but these modelers
see it as being of little importance,
and it is shared with the matrifocal
unit to which the males belong rather
with dependent
female sexual partners.
Indeed, in some early versions of the
Woman the Gatherer
model, the male's role was so little described that he
might be said to have evolved clinging to the apron strings of the women. In
more recent versions (133), called simply the "gathering model," the male's
role is elaborated,
but still considered
to be secondary to the part played by
women in unique human inventions.
In retrospect,
it may seem discouraging
that the choice had to be seen as
or gathering,
with either men or women inventing
the cultural
that make us distinct. However, at the time, it must have seemed
necessary to establish that a credible scientific origin story could be con-
structed in which women invented tools, chose mates, developed social
systems, provided for themselves and their offspring, and generally partici-
pated in the evolution of significant human abilities.
Zihlman (127) described
four types of reaction
to the gathering
model: to
accept it wholeheartedly;
to reject it as sex-biased;
to integrate
its parts
existing models; or to ignore it even while taking its salient features. The
latter two reactions are of most interest to this review, since they brought
changes to the scenarios of human evolution.
The response of some of the proponents
of the hunting model was to
the new model on the older hunting scheme, and to emphasize
mixed economy in which early hominid men and women were mutually
interdependent (e.g. 46, 53-55, 60a, 62, 63). In many respects
this has been a
gesture of conciliation and a genuine attempt to modify the models to
new thinking. Many authors now emphasize
the importance
sharing between gathering
women and hunting
or scavenging
men as the key
invention, i.e. the sexual division of labor. Isaac (45-49) has done the
most to develop a model in which food sharing
is the "central
argued that the archaeological
evidence from East Africa demonstrates that
the earliest
food and tools to certain locations
where we now
find their remains. In his view, this is evidence that the unique human social
and economic arrangement of sexual division of labor had already begun to
take place, and the reason they carried food to consistent locations was in
order to share it. He hypothesized that males and females ranged in separate
groups, engaging in specialized activities, and brought food back to a home
base to share, as do contemporary foragers.
for the sake of conciliation, and for what seems at first sight
an anthropologically pleasing "holistic"
models are constructed on
a foundation of assumptions
causal chains and
about human sex roles. It
may not be possible to simply superimpose
one on the other like so many
building blocks without
faults in the logic of the whole. Gould (36),
for example, has said that
models are really about meat sharing,
and both Hayden (41) and Isaac (45, 47) have stated that neither sharing nor
social living would have been particularly advantageous
to foragers living
largely on vegetable
foods. Indeed, Isaac in his later versions (48) still tended
to see meat eating, now scavenged
rather than
as the key factor
in the
of human
intelligence, language, and social patterns.
And since
he saw women as encumbered with children and handicapped
in meat-
obtaining activities (45), females still do not seem to be credited with full
in the "sharing"
model. The recognition
that simple choppers
hand axes would have facilitated
scavenging, but not hunting, has been slow
to find its expression
and implications
in the sharing
model (e.g. 60a). There
has been no "scavenging model"; rather, scavenging has replaced or been
added to hunting, without any concomitant
changes to other aspects of the
model or consideration
of its implications
for sex roles.
The fourth reaction described by Zihlman has been to ignore women's
productive roles (and women anthropologists' models) altogether,
while in-
corporating some of their undeniably salient points. The currently most
quoted model of human
social evolution contains such borrowings
and could
be said to illustrate
this fourth response.
In 1981 Owen Lovejoy published
a paper
entitled "the
of Man"
in which the postulated
sex roles and division of labor
of early hominids
described precisely as Darwin had imagined
them 100 years earlier;
home bases to bear and rear children and were dependent
men to protect
and provision them. The arguments
as to why women had to
remain dependent and sedentary
were new, but otherwise the origin story
remained familiar. Lovejoy's argument
drew from several new and diverse
sources (such as life history theory) and can be summarized as follows. The
earliest hominids were able to become successful as a lineage, especially in
to their ape relatives, by facilitating higher fecundity and lower
infant mortality
rates than the present chimpanzee
life history pattern
of one
infant every four years and only five live offspring in a female's lifetime.
Hominids increased their reproductive success by reducing the mobility of
lactating mothers and inventing the provisioning of the sedentary females by
mobile, bipedal males. Lovejoy's scenario began x.
ith the assertion that
hunting was not the crucial human technological invention, but rather that
gathering was the key innovation. It did so without any reference to the
published Woman the Gatherer models, which had accumulated the major
body of evidence and arguments
for gathering (and against hunting) as the
"master behavior pattern."
Further, Lovejoy attributed
the collecting of plant food items, and all the
ramifications of gathering and shaling in hominid evolution, to the early
hominid men. Since male anthropologists
had shown no previous signs of
wishing to associate
their sex with gathering,
and since all of the ethnographic
evidence points to women as primary gatherers,
this sudden enthusiasm
has been seen as the co-opting
of the gatherer
model (127). The core
assumption of Lovejoy's scheme is that for hominoid females, successful
of reproduction
and productive
activities are
incompatible, and thus men
the impetus for hominid success by inventing the provisioning
vegetable foods to sedentary, monogamous
female mates.
No extended
of Lovejoy's model have yet appeared (but
see 7, 10,
43, 80, 85, 130, and
below); however, his view of early hominid
sex roles is
cited in many recent editions of physical anthropology textbooks
and popular
accounts. Appearing
as it did in an invited article in the prestigious journal
Science, Lovejoy's model could be said to represent
the current
about human evolution.
recent, but much less widely noted, model of human origins (68)
began with a similar question to that of Lovejoy's (how were early humans
able to survive and succeed?), but offered a very different, even opposing
answer. Leibowitz argued that a sexual division of labor was a very late
human invention, and that for much of hominid evolution both males and
females engaged in the same sorts of productive
activities (126). Females
simply combined productive activities with reproductive activities, as do
many contemporary
women. In Leibowitz's view, the key human
was production, by which she means food-getting
with tools, and which was
initially unspecialized and undifferentiated
by age or sex within the group.
She drew an analogy to the manner in which every weaned member of a
monkey or ape social group is an independent foraging unit.
Like Lovejoy, Leibowitz interpreted
the material
evidence to mean that
early hominids were "hovering precariously
on the edge of extinction"
(68, p.
135), and argued
that their major hedge against a marginal replacement
was to invent the practice
of accumulating surplus
food through production.
All individuals in the group participated
in gathering surplus and in the
or exchange. In her view, it was only with the invention of
fire and projectile weaponry at the time of late Homo erectus that a sexual
division of labor began to appear. The sexual division of labor also served as
an instrument for stabilizing and extending both intragroup sharing and
intergroup exchange. However, for most of human history, production alone
(and not a specialization
of roles by age and
sex) was necessary and sufficient
to create the characteristic human patterns.
Leibowitz's idea is noteworthy
for two reasons: it shows again how the
same data can be interpreted
in quite different
ways, and it is one of the very
few attempts (see also 12) to strip away the remaining assumption common to
all models, that sex differences must have been significant in the earliest
stages of human evolution. It seems that one of our own cultural patterns is to
oppose male to female characteristics and to assume and emphasize sex and
gender differences rather than similarities. That human technological and
social success can be attributed
to a specialization
of tasks by sex is an often
repeated assumption
of anthropology,
and some type of sexual division of
labor seems to be universal
in human
societies today, although the importance
it is variable. Yet it can be very enlightening
to think
through what
we have assumed to be the less probable solution. Could
societies have originated without a sexual division of labor beyond that
directly related to insemination, gestation, and lactation? Could some be-
havioral invention, characteristic
of neither males nor females and requiring
equivalent participation,
have been the catalytic
event that
set humans
along their
own distinctive
evolutionary path?
Given that
females are
able to combine foraging
with infant
care, and that women in most societies
at least as much as men to subsistence
in addition to their
tive activities, Leibowitz's scenario may be no more or less data-based and
plausible than the many models that seek to give preeminence
to one of the
other sex in the story of human evolution.
In the following sections, I review these "data
bases" or the sources of
evidence from primatology, ethnography, and paleoanthropology
for the
models just described.
who are trained as anthropologists
not infrequently study
subjects with an eye to casting some light on the behavior and
evolution of our own species. It is reasoned that since humans
are members
the order Primates, the study of our nearest animal relations can help us to
both the ways in which we are similar to other species and the
ways in which we are distinctive. Although many primatologists
are un-
comfortable with inferences drawn from animals to humans, and unhappy
with what they regard
to be facile analogies made in the past, there exists
pressure from colleagues and the public alike to make primate
studies more directly relevant to the study of humans.
Such was the intent
of one of the earliest and most widely publicized
studies of a nonhuman
primate, the baboon. DeVore and Washburn
116a) observed
baboons in East Africa in 1959-60 and constructed
model of early human
life based on baboons. They argued that early homi-
nids, like baboons, differentiated
from other primates by exploiting the
resources of the East African savannah.
Like baboons, humans
would have
become both predators
of savannah
flora and fauna and the prey of the large
savannah carnivores. In order to protect themselves, given their relatively
ineffectual physical abilities as individuals, the model proposes that both
humans and
came to rely upon a social system
of defense. This social
system was said to be based on the bonding
and cooperation
of mature males
organized into a rigid dominance hierarchy
and employing an "army-like"
pattern of "troop"
movement across the dangerous
plains [e.g. "Baboons
move in a carefully structured
defense formation, guarding
the nucleus of
females and infants. Early humans
may have traveled
in similar formation"
(31, p. 94]. According
to the model, and
there are
versions of it (1, 84,
87, 113, 114), human males distinguished
themselves even further as exploit-
ers of the savannah through
the invention
of weapons
and thus
in turn led to unique human
traits like language
and the family. However, it
was argued, this complex of distinct human characteristics
initially was
founded on a social system very like that which DeVore described for
Such a "baboonization"
(87) model of early human life experienced a
that may have surprised
even its authors. Throughout
the 1960s
and 1970s, no textbook or course in introductory anthropology, and no
concluding chapter
on the human species in the animal behavior and evolu-
theory texts, seemed complete without reference to the baboon analo-
gy for early humans. Even an elementary
school social science curriculum,
called "Man, A Course
of Study"
included extensive coverage of the baboon
of this depiction
of baboon social life and of this model for early
human life has come from many quarters,
including primatologists,
gists, and social anthropologists,
and such critiques will not be covered in
detail here (but see 24). For the purposes of this review, the portrayal
primeval sex roles in the baboon model can be said to have been traditional
and consistent with contemporary
role expectations for Western men and
women (73, 77): males were seen as aggressive, competitive, and protective;
females were seen as nurturant,
dependent, and submissive.
Today, with the extensive evidence available
from anatomical,
cal, paleontological, and behavioral
studies, it is widely accepted
that chim-
panzees are the nonhuman
most closely related to humans,
and it may
seem odd to have chosen any other species from which to draw analogies.
However, at the time the baboon model was developed, this presumably was
not so evident, and the ecological analogy between these two distantly related
primates was widely accepted.
In the past decade, many reconstructions
of early hominid
life have drawn
heavily from the accumulating data on the behavior of common chimpanzees
(Pan troglodytes) and pygmy chimpanzees
(Pan paniscus). Following a. line
of argument
established by Darwin nearly a century ago, some of the recent
models suggest that
show rudimentary patterns
of behavior that
also might have been exhibited by our common ape forebears, and which
were greatly
by hominids as the latter
differentiated from the other
hominoids. Some of these patterns of behavior, it is argued,
were ultimately
to become the distinguishing characteristics of the human lineage. Even
though chimpanzees have traveled their separate evolutionary
route for the
past 5 million years, it is believed that their traits
can give us some clues to the
general "ape-like" way of life of our hominoid ancestors, a way of life that
was to set the stage for the human
The rest
of this section is organized
around the behavioral characteristics
of chimpanzees which modelers have
isolated and suggested as possible antecedents for human patterns (see es-
pecially 35, 79).
Social Bonds
The core of chimpanzee
social life (indeed, almost
all mammalian social life)
is the enduring mother-offspring
bond. In most primate species, the male
at puberty
whereas the female remains close to her mother for life.
In chimpanzees, the reverse seems to be the case, with adolescent females
leaving their
mothers and communities at first
estrus, but whether
for mating, or permanently
to live in a new community, is not yet well
Nonetheless, a chimpanzee
mother suckles each infant for around
four years and remains physically close to her offspring until they reach
sexual maturity
at ten to twelve years
of age. Since a female chimpanzee
an infant approximately every four years, she may have two or more de-
offspring traveling
with her at any one time, but usually
only one that
is suckling and
being carried.
sons, and less frequently
her mature
daughters, also travel with her on occasion. Some male-male bonds are
(often between maternal
and estrous females may travel and
forage with male parties,
but the enduring
and primary
social unit is matrifo-
cal, that is, centered
upon and articulated around the ties between a female
and her offspring.
Most versions of the Woman the Gatherer
model have used this aspect of
chimpanzee social bonding (and primate
social life in general) to argue that
the matrifocal unit, and not the nuclear family, whether monogamous or
polygynous, was the core of early hominid society. The intensive and exten-
sive mother-offspring
bond of the ape, it is argued, could only have become
more elaborated in a lineage such as hominids,
with their
infants. Following this argument,
the initial social ties of adult males would
have been to their maternal kin and not to their temporary
sexual partners.
Chimpanzees are not as promiscuous as initially reported; still, sexual
bonding is temporary.
Individuals of both sexes avoid mating
with close kin,
and although
males may occasionally act possessively or competitively in a
mating context, most copulations are casual and opportunistic,
with females
exhibiting preferences
for certain males over others. Much has been made of
the "loss of estrus"
in the human female as compared
to other primates
and its
supposed causal relationship
to permanent
pair bonds. However, we now
know that female pygmy chimpanzees
in the wild copulate throughout their
monthly cycles [Badrian
& Badrian
(4)] but are not pair bonded to males,
whereas monogamous nonhuman
species all show pair bonding but
no loss of estrus. Thus the hypothesized correlation
between loss of estrus
monogamy is not supported by the primate
Social Dynamics
social life is complex in that
two levels have been identified:
a large community of individuals who recognize each other and are not
mutually antagonistic
when they meet, and (b) smaller
of individuals
who travel, sleep, and forage together. Their social life is fluid in that the
composition of parties
with only the mother-dependent
offspring unit remaining
constant. Such a pattern
has been referred to as a
"fission-fusion" social organization.
(121) has conceptualized
a chimpanzee
as a coop-
of related
males who overlap
the individual
of individual
units and sometimes behave antagonistically
toward mem-
bers of neighboring
Foraging parties
within the community
believed to fluctuate in size and composition in relation to the changing
abundance and distribution of food resources. This complex, fluid, and
environmentally responsive
social system has been described
by many model-
ers (e.g. 79, 94, 124) as containing
the essential
for early
foragers to adapt their social groups to both the resources and the tech-
nological innovations important
in the human way of life.
Feeding, Food Sharing, and Tool Use
Probably no aspect of chimpanzee behavior has interested anthropologists
more than their dietary
and technological
habits. It is now widely known that
chimpanzees learn to make a variety of simple tools which vary in structure
and function from community to community. Tools are occasionally used in
agonistic contexts, usually by males, but more commonly for food collecting
and processing. Hammerstones are used to crack
open hard
fruits, and
are used to collect insects from underground
nests. The majority of tool
making and tool use is done by females. This is because mature
females con-
sume many
more social insects than do males and thus
much more
ing for termites
their male counterparts. The pursuit and killing of small
animals, primarily
carried out by males, is done without the use of tools.
As exciting as the reports
of tool use and "hunting"
by chimpanzees
the initial descriptions
of food sharing
in these animals. Except for suckling
infants, the basic primate rule of feeding seems to be each individual for
itself. Although it is likely that social groups enhance the abilities of in-
dividuals to find and defend food sources, each nonhuman
past the
age of weaning, male or female, is in all other respects an economically
foraging unit. Provisioning
of dependents
is not a characteristic
of the primate
order, and even minimal sharing
of food [which Isaac (45)
dubbed as "tolerated
is very rare.
The first descriptions of food sharing in the chimpanzee placed such
in the context of meat eating (34, 111). It is mainly adult
males that
kill animals and eat meat, and mainly old, past-prime males who are the
recipients of shared
meat. However, close female kin of the meat possessor
and estrous females also receive more than
expected shares.
These data seem
ideally constituted to construct a model of how human hunting
would lead to male provisioning
of a nuclear
family with meat, and indeed
they have been used to this end (27).
However, further field studies of both the common and pygmy chimpan-
zees and specific investigations
of the nature of food sharing
(52, 78, 104)
have since demonstrated that a great deal of sharing
also occurs with plant
food, particularly large or hard-to-open
fruits. More importantly,
in common
the vast majority
of such sharing (McGrew
within the matrifocal
family. And the provision of food by mothers to their
offspring, either through
cadging of scraps
or through
accounted for almost all cases of plant food sharing
in McGrew's study.
Several authors
(79, 108, 126) have used the information on sex differences
in chimpanzee tool use and food sharing
to reconstruct how these patterns
might have been further elaborated
by transitional
They argue
that gathering,
the catalytic innovation
in hominid
technology, was invented
by females whose digging sticks and unmodified
stone hammers
were refine-
ments on the female ape's tool kit of termite probes and pounding stones.
as a pattern
of accumulating
vegetable food
leads to carrying
and sharing
of foods. Primate
females, with few exceptions,
are adapted
to the burden of carrying
infants, and hominid females would
have invented slings to carry their nonclinging infants and their food supplies.
Extended human sharing of accumulated or surplus foods would have been
founded primarily on the ape pattern of matrifocal sharing.
females are also mobile and
clearly capable
of finding food for
themselves and their dependent young, thus undermining Lovejoy's argument
that a hominoid female could not combine successful reproduction with
subsistence activities. Lovejoy's view of the chimpanzee life history pattern
as leading to a marginal existence, hovering on the edge of extinction, has
been criticized
by several other researchers
(40, 47, 119, 120). They counter-
females produce
and rear
at about the same
rate as the other great apes, and more importantly, at comparable rates to
human females in foraging societies. At Gombe Stream
Reserve, the chim-
panzee females who have been followed over their lifetimes have each raised
several offspring
to maturity,
a replacement
rate that
certainly would not lead
a population
to the brink of extinction if it were also maintained
outside the
protected park. It is clear that the low population
of chimpanzees
in Africa
today is the result of historically
recent human
of the animal and
its habitat.
Still, there clearly exists enough complexity in chimpanzee
behavior and
enough diverse conclusions from the studies of these animals to give rise to
many different
Specifically with reference to sex roles, some of the
models have tended to continue
the emphasis
on males as the main
in the development
of distinctive human
abilities, using as a foundation
the data on male chimpanzee aggressiveness
or male ranging
behavior or male
hunting patterns (94, 107, 111). Whereas
others, using the findings on the
central significance of the female chimpanzee
in social bonding and in food
procurement patterns, have proposed a radical or nontraditional view of
human females as prime movers in the evolution of the essential hominid
traits such as tool use and sharing (13, 70, 126). Finally, some authors
explored the manner in which sex differences in chimpanzee
have set the stage for sexual division of labor in the first hominid societies
(32, 38, 53, 54, 78).
Until the advent of agricultural practices
based on the domestication of plants
and animals no more than 12,000 years ago, peoples around
the world must
have lived as foragers. The archaeological
evidence of lithic artifacts
back some 2 million years indicates that human
have long acquired
and/or processed their food with the assistance of tools. And the evidence
at various, mainly later Pleistocene dates of cut marks on animal
bones, of homebases
with remains
of plant
food collections, of the use of fire,
and of increasingly sophisticated
tools for food collecting and processing all
point to a hunting and gathering
subsistence pattern
at least in late-middle and
upper paleolithic peoples. Reserving for later the issue of whether the earliest
hominds were already
hunter-gatherers, or simply generalized foragers with
tools, or perhaps even tool-less primates who differed little from the ape
forms except in being bipedal, the significant
point for this section is that
vast majority of the cultural remains of paleolithic peoples have been in-
terpreted as resulting from the technological system of hunting and gathering.
And since there are obvious technological similarities between these
archaeological remains and the material culture of contemporary
gatherers, and in some cases ecological/environmental
similarities, some
have turned to the study of modern hunter-gatherers
to shed light
on the reconstruction
of the social patterns
of prehistoric
The logic is
that social structures
to environmental
exigencies and correspond
technological systems. Thus, it is argued
that the basic social forms widely
found in contemporary
hunter-gatherers, especially those dwelling in tropical
zones, probably
occurred as well in paleolithic hunter-gatherers.
A number of anthropologists
have objected to the use of ethnographic
evidence to reconstruct
early human social life, and these objections will be
briefly since they do have a bearing
on the assessment
of the models
themselves. Freeman
(29) has objected
to analogies
ic and modem groups
on both methodological
and theoretical
grounds. First,
he argued that to force archaeological evidence into frames of reference
developed for contemporary
data inevitably distorts and obscures the pre-
historic analysis. It also prevents the development of frameworks based
directly on the prehistoric
material. Secondly, he argued that like environ-
mental stimuli do not necessarily produce like cultural
responses, because
sociocultural systems have tended to regional-and-resource
the course
of human
history. More
recently, in an extensive analysis
the relevance of contemporary hunter-gatherers
to paleolithic societies, Tes-
tart (112) also emphasized
the particular
of each society's history, the
of regional
events, climate, fauna, flora, and the 10,000 years of
separate today's hunter-gatherers
ic antecedents. Nonetheless, Testart's detailed analysis of the ethnographic
evidence led him to conclude that at least some contemporary
can provide insights into prehistoric
A second criticism of the ethnographic analogy rests on ideological
grounds. Berndt (5), for example, has objected to the implication
that the
study of modern
can help us to understand
early human
societies. She suggested that this view is harkening
back to the nineteenth
century social evolutionist and colonialist racist attitudes that aboriginal
peoples are "primitives"
or "survivals";
they are lower on an evolutionary
scale and thus inferior. Schrire
(99) has argued for the !Kung, on somewhat
similar grounds to Berndt for the Australian
aborigines, that we should not
contemporary foraging peoples as "living fossil groups." This objec-
tion is also in part substantive;
following Boas, most social anthropologists
have argued that there are no modem representatives of past cultural
Therefore, many draw the conclusion that any attempt
at reconstruction or
analogy based on contemporary
peoples is not only futile speculation, but also
to those tenets of anthropology
based upon a nonhierarchical
view of
cultural variation.
Others (e.g. 62) reply that to suggest that similar economies and similar
technologies may be associated with similar social structures,
and to construct
hypotheses on the basis of such similarities,
is not to suggest social evolution
in any pejorative sense. Similar reservations about the appropriateness of
to human behavior
have been expressed. A discussion of the
proper use of analogy in natural and social science might be useful in the
context of human evolution theories, but it is beyond
the terms of this review.
the conclusions of such a discussion, it is nevertheless
necessary here to accept the usage and to go on to the question
of which, if
any, of the modem hunting
societies provide
the most appropri-
ate analogies.
The most systematic attempt to answer this question is by the French
Alain Testart
(112). In the process of considering
the issue of
why some hunting
societies persist
in their
rather than adopting sedentary
or pastoral
lifeways, Testart drew up a clas-
sification of six types of hunter-gatherers.
Of these six, only two categories,
one largely comprised of North American Indian societies, the other con-
taining notably the !Kung
and Australian
Aborigines, were found by Testart
to have structural features
that would make them good choices as models for
earlier foraging societies. Testart then eliminated North American Indian
societies because of their
recency, geography,
and specialization
for a habitat
unsuitable to agriculture,
leaving as the group
of choice such societies as the
!Kung and the Australian
Aborigines. At the end of his careful and well-
reasoned analysis, Testart returned to the ideological question
by noting that
the choice of a contemporary
society and its application by analogy to
peoples may be informed as much
by subjective
factors as by overt
Lee (62, 63, 66) has directly
to the ethnographic
analogy (see also 122), arguing
that the use of !Kung
data to illuminate
past is not to regard
these people as living fossils. The !Kung have a long
in southern
Africa, over which time regional
events would have had an
on social forms, and
they have not lived in isolation from nonforaging
peoples and ways of life. Nonetheless, Lee believes that by proceeding
cautiously with these caveats in mind, there is much to be learned about a
hunter-gather way of life from studies of contemporary !Kung and other
foragers. In his view, they have a core of features in common which
"represents the basic human adaptation stripped of the accretions and com-
plications brought
about by agriculture, urbanization, advanced technology,
and national and class conflict-all of the 'advances'
of the last few thousand
years" (63, p. 3).
As a result of extracting these "core features," Lee characterized the basic
or generalized hunter-gatherer society as a flexible, bilaterally organized,
nonterritorial group, with a particular emphasis on the genealogical core as
consisting of both related males and related females (62). Earlier, Lee &
DeVore (66) had defined several features generally characteristic of the
hunting and gathering way of life: 1. groups are small and mobile, with
fluctuating membership;
2. food surpluses are not prominent, and mobility
places constraints on the accumulation
of any type of surplus, thus the system
is basically egalitarian; 3. groups are not strongly attached to any one area and
do not ordinarily
maintain exclusive rights
to resources
(i.e. they are nonterri-
torial); and 4. reciprocity and a division of labor lead to an emphasis on
sharing resources.
Leacock (59, 60) has identified several of the same features in her analyses
of present and past hunter-gatherers, although she placed somewhat more
emphasis on egalitarianism and the lack of specialization or hierarchies
to resources. Lee (62) stated that his view of hunter-gatherers and thus
early humans as living in flexible, bilaterally organized groups
is a correction
to the "patrilineal
horde" model first developed by Radcliffe-Browne for
Aborigines (91) and then applied by others (e.g. 28, 101, 102) to
the reconstruction of early human social life.
The Roles of Women
in Hunter-Gatherer
The picture that was painted of the social role of women in much early
and ethnological
work on hunter-gatherers
was of a dependent,
lesser, and even passive social category. Ethnographers, mainly men, studied
social phenomena of greater interest to men and talked mainly to male
informants. The emphasis on hunting, weapons, and warfare ignored the
of women to subsistence and to social dynamics. Theoretical
models (e.g. 69) viewed men as actors and women as objects of sexual
exchange. However, in the last two decades, many new ethnographic
employing female as well as male perspectives have been undertaken
extensive review in 90). Thus, a picture of women as active, competent,
and even self-sufficient members
of hunter-gatherer societies,
with their
own stories to tell, has begun to emerge from the shadows of early
In particular, Lee's (61, 64, 67) continuing analysis of women's contribu-
tion to subsistence in contemporary hunter-gatherer societies has been an
important starting point in a reassessment of the parts women might have
played in early human society. In a survey of 58 foraging societies from
around the world (61), Lee concluded that on average
hunted foods contrib-
uted only 35% of the diet and thus, contrary
to popular conception, men
provided less than
half of the food of "hunting" peoples. Of the 58 societies he
surveyed, 29 (slightly more than half) depend primarily
on gathering, one-
third primarily
on fishing, and only one-sixth primarily
on hunting. Thus, he
concluded that except in the Arctic, where meat is of primary
plant food, shell fish, and fish, collected primarily by women, form the bulk
of the diet. Hence his argument
women generally
are capable of
feeding themselves and are not dependent
on men for subsistence.
These conclusions were taken up enthusiastically by the various modelers
of Woman the Gatherer
and are also occasionally mentioned
in introductory
textbooks, perhaps to temper the emphasis on males as providers in de-
scriptions of hunter-gatherers.
Hiatt (42) extended Lee's analysis to demon-
strate the economic importance
of women in Australian
Aborigine society in
and in tropical
in general. Martin
& Voorhies (77)
and Whyte (118) also extracted
of foraging societies from
Atlas and concluded that women generally
substantially to subsis-
tence. However, the different samples and definitions used in these studies
render detailed
(20), for example, drew a
different sample from the Atlas, using different definitions of hunting, and
came to differing conclusions from those of Lee, Hiatt, Martin
& Voorhies,
and Whyte.
There is a serious problem in any attempted generalizations
from Mur-
dock's Ethnographic
Atlas. The foraging societies described
in the Atlas are
not a random
sample, nor a representative sample, nor a complete
of all the hunter-gatherers
societies that have existed, even in historical times.
Worldwide surveys taken from the Atlas may well be biased toward those
cultural zones which for many reasons have been more often studied, unless
some form of corrective
sampling is attempted.
the quality of the data coded for the various societies is uneven and often
unrefined, as noted by Hayden (41), and ethnographers
often have not
collected the original data with Murdock's ultimate categories in mind.
Also taking a different view from that of Testart,
that North
American foragers are more instructive about the past than those of the Old
World. However, for the purpose of extrapolation
to early human societies
that existed in the African
zones, it seems clear that one would not
want a comparative sample composed predominantly
of North American
temperate and arctic zone dwellers. Tropical and subtropical
zones offer a
greater abundance and diversity of edible plants than do more northerly
latitudes, and various researchers (42, 77, 112) have demonstrated that in
contemporary tropical and subtropical foragers, meat forms a small propor-
tion of the diet, whereas vegetable foods provide a high percentage of the
subsistence base. Testart found that the percentage
of meat in the diet of
hunter-gatherers correlated with latitude, going from a low of 10% near the
equator to 90% in the Arctic.
Finally, the issues are further
muddled by the fact that although women are
primarily associated with gathering plant foods, they do also obtain small
hunt with weapons
for larger ones (22). Men, on the
hand, often help with gathering
or feed themselves on plant matter while
hunting. Dahlberg (14) presented
a short but cogent overview of the results of
various surveys on male and female contributions to subsistence, using
samples and different definitions
of hunting, gathering,
and fishing.
The issue of differential contribution to subsistence has been dealt with at
some length because it is important
in assessing women's status in early
foraging societies. Women's reproductive
roles have never been in question
(except the degree to which they are handicapping);
it is their productive
capabilities that are contentious. Anthopologists
who have followed Engel's
at its most basic (e.g. 59, 60) have long argued
that those women
who actively contribute
to subsistence, and who are not economically de-
pendent but interdependent
with all the other producing members of the
group, will have equivalent status to that of the men. Others (30) have
modified this argument to add that women must not only contribute to
subsistence but also have a measure of personal
control over the disposition
and distribution
of the fruits of their labor in order to achieve power and
prestige equivalent
to that of the men. The ability to control production and
is more difficult to demonstrate, and possibly is less true of
women than the ability to contribute to production. However, if the data
continue to show that women are not economically dependent
on men for
in most hunter-gatherer societies, indeed
that they often produce
more than do the men, then the assumption
of the nonproductive female,
which has been a key element in most reconstructions of our earliest an-
cestors, must be seriously reexamined.
Implications for Recent Models
Since the ethnographic
evidence on contemporary hunter-gatherers
in tropical
and subtropical
zones supports
the economic independence
claimed in the
Woman the Gatherer
model, and since no ethnographic example exists of
women in foraging societies being provisioned by their husbands
with plant foods, it is not surprising
that the male provisioning
model makes
no reference to ethnographic sources. Lovejoy's argument that early human
females would not have been able to carry babies and burdens, and would
have had to remain sedentary in order to reproduce
successfully, is also
contradicted by ethnographic
evidence. It is clear that in most parts of the
foraging women are
assigned the tasks of carrying heavy burdens: food,
children, water, and firewood. Sedentary women simply do not exist in
hunter-gatherer societies. Where quantitative data have been collected, it has
been found that women are away from basecamp for equivalent
amounts of
time and walk equivalent
distances, carrying
and heavier
burdens than
do the men.
Finally, Lovejoy's argument
that the earliest
would have differen-
tiated from the apes by rapidly increased reproduction (becoming "r"
selected) is not supported
by any of the ethnographic (or paleolithic) evi-
dence. Contemporary
foraging women only produce
one child that is raised
every three to four years (a reproductive
rate that is strikingly
similar to the
rate found in female apes), and
may be assumed to arrange
lives around the demands of their productive activities (a unique skill in
humans which, however, never seems to appear on the trait lists). The
record shows that the great population
increase, indicat-
ing heightened fecundity
possibly concomitant
of women,
which Lovejoy postulated
as one of the first and necessary
events in human
evolution, actually
does not occur until much later in human
evolution, when
humans radically altered their subsistence techniques to domesticate their
food resources.
Although she did not use ethonographic sources, Leibowitz's model of an
early human society in which every mature individual could feed itself and
also contribute to the group
a sexual division of labor could have been
supported by the example of one contemporary foraging group, the Tasaday
(26). Although much controversy and too much publicity surrounded the
of this isolated
foraging group
in the Philippines
in the 1960s, there
are many noteworthy aspects
to their
lifeway. The Tasaday
seem to have been
isolated from all but two neighboring groups
of people, similar to themselves,
for at least 600 years, and they practiced a simple but successful way of
living. The small band (24 people) practiced
no sexual division of labor, and
until first contacts with explorers, no hunting or trapping. However, they
collected small animals from riverine areas, without the use of tools, to
foods. The technology
was very simple, food was
easily gathered in a few hours, a short distance from the home base, and
readily shared
the group. As Hayden (41) noted in his worldwide
survey of hunter-gatherer
groups, an unmeasured
but possibly large propor-
tion of food is simply "snacked"
in an ad hoc fashion as people move about
collecting a surplus
to be brought back to camp. In the Tasaday,
all decisions
were made by consensus, with no evidence of an authority structure or
dominant sex. In these respects, the Tasaday would illustrate
model of an egalitarian,
unspecialized, autonomous-yet-sharing,
foraging group.
work with three types of material
evidence about the
early hominids:
osteological remains,
the physical
traces of their various
activities, and the associated
or contextual information on the environment
which they lived. The latter two will be discussed under "archaeological
evidence" in a following section; here I will discuss briefly how these
osteological remains are described and interpreted, focusing on those aspects
that are relevant to sex role reconstruction.
Before describing the data, a few comments on the distinction between
evidence and inferred evidence would be useful. Bones and stones
are a very fragmentary
record of the past, and, like other empirical phe-
nomena, cannot speak for themselves. Thus in some respects, inference and
must occur in every description, at every level. However, to
clarify the distinctions, Isaac & Crader
(49) have suggested three levels of
in paleoanthropology:
first, interpretation
of the empirical
dence (the "finds"); second, interpretation
of the processes that led to this
evidence; and
the formulation
of general
models to explain
evidence. In terms of the earliest
hominids, we can use the fossil material
draw some first-order, descriptive inferences about body size and shape,
and dental
At a second level, we can infer
behavioral and
and the selective pressures
have led to these
phenotypes. Finally, we can construct models that incorporate
our various second-order
inferences into a coherent framework
of explana-
tion. Although interpretation
does occur even in seemingly straightforward
of the fossil remains, which are fragmentary
and often must be
still, as we move from descriptions
of material
remains to
processes and then to models, our inferences are increasingly dependent upon
assumptions to be tested by internal
consistency and plausibility, and de-
creasingly by reference to empirical evidence.
Fossil Evidence
for Early Hominid Sex Role
DESCRIPTIVE DATA In East Africa, approximately
three million years ago,
one or more species of hominid
lived in a savannah-like habitat of grasslands
with pockets of forested and riverine areas. Paleoanthropologists
have classified these creatures as hominids because
they were bipedal
and had
human-like dentition,
but they have not yet agreed
upon the number
of species
living contemporaneously.
For the sake of simplicity, I will refer to the very
earliest hominids collectively as australopithecines.
These hominids were
small in frame and short in stature, with brains no larger than those of
contemporary chimpanzees. Their dentition
was distinctive from that of the
pongids, with characteristically
small, incisiformed
canines in both males and
females and large thickly enameled molars. Relative dimensions of their
limbs and aspects
of their fingers and toes indicate that they continued
to have
some grasping
ability and may still have spent some time moving around in
the trees, but their lower limbs and the shape of the pelves indicate they
walked on two legs on the ground. Beyond these rather minimal
statements (and even they are not without contention and exceptions), we
move quickly into the realms
of either
detail or of second-order
Because this review focuses on the reconstruction
of sex
roles, comments on the processes that led to the general australopithecine
will be limited to those aspects that seem most relevant to female
and male patterns.
DIET Diet plays a major
role in models of human
evolution and thus many
have turned
to an analysis
of tooth
shape, and more
wear patterns,
in order
to infer what the earliest hominids
might have eaten.
One recent consensus appears
to be that their tooth morphology indicates
omnivory, with no clear specializations
for meat-shearing
or seed-grinding
bone-gnawing (e.g. 39, 75, 76). Studies
of tooth wear, masticatory
ture, and "microscratches"
on the surfaces
of teeth indicate that
is variation between species (especially between later robust and gracile
these early
hominids were eating a variety
of foods, some
of them soft fruits and
tough, fibrous, and hard to chew. There does not
seem to be consensus
on whether these foods were generally gritty, indicating
that they were mainly tubers and roots dug from the earth, or clean, which
might suggest that they were fruits and other products
from trees (cf 48 to
128). More importantly
for this review, it is not yet possible to determine
from tooth
wear or from chemical
of bones what
of plants
and meat occurred
in these early human diets.
One feature of human dentition that has long intrigued physical an-
is that canines are
small in males as well as in females.
In most primate species, and in the fossil forms that are believed to be
to the australopithecines,
in males than
females. It
is usually suggested
male canines in primates
are selected for, either
as part of a male protective role against predators
or as part of sexual
selection, resulting from male-male competition.
For hominids, the tradition-
al explanation, which forms part of the hunting model, is that male canine size
was reduced after the invention of weapons removed the need for canines as
defensive tools (see 116). This explanation has become dated with the grow-
ing recognition that the reduction in hominid canines began long before the
of tools in the paleontological record. On the other hand, follow-
ing the principle of female choice, but contradicting
Darwin's contention of
male choice in humans, some versions of the gathering
model (124) have
suggested that female proto-hominids may have selected males with smaller
canines as preferred
mates because the latter
represented less of an aggressive
threat to them and their offspring. Finally, it may simply be that smaller
teeth (incisors and canines) and larger, thicker
posterior teeth (pre-
molars and molars) were adaptations
to produce flat, durable surfaces for
chewing the fibrous foods that comprised
the omnivorous diet of the earliest
of the fossil record
that would bear directly
on models of sex roles concerns
the degree of sexual
in the earliest hominids. Although degree of sexual dimorphism
does not correlate perfectly with sex roles, dominance
relations, or mating
systems in the other primate
or mammalian
species (92), nonetheless mono-
gamous primates
tend to be monomorphic
and behaviorally
by sex, whereas highly dimorphic species tend to be polygynous and male-
there is little agreement
on whether or to what degree the
earliest hominids
were dimorphic.
It is not a simple matter to sex fragmentary
fossil hominids, especially when only a few individuals are
known of a given
"type" or species. The gracile and robust australopithecine
material from
South and East Africa was sometimes interpreted
as representing
the females
(gracile) and males (robust) of one species (e.g. 9). However, most would
now agree that separate gracile and robust
species existed in South and East
Africa, and it is not clear what the degree of sexual dimorphism
would have
been within these species.
The problem
of distinguishing species differences from sex differences has
now reemerged
with the very earliest
material from
Hadar, presently
dated at 2.9 to 3.2 million years ago. The famous "Lucy"
and "First
fossils are interpreted by finders Johanson & White (50) as one highly
variable, sexually dimorphic species, Australopithecus
afarensis, whereas
they are interpreted by some of the French members of the team (100, 110)
and by Zihlman
as two separate species. All agree that there is a great
deal of
size variation
in the fossil hominids
so much so that
argued from the limited published measurements
that if these hominids do
represent only one species, they would be more dimorphic
than any known
primate (131). Other researchers, however, continue to discuss and analyze
the Hadar material as one highly variable and dimorphic
species (e.g. 80,
BIRTH, GROWTH, AND DEATH The evidence for relatively
small brain size
in the earliest hominids suggests that although the shape of the pelvis had
altered to accommodate bipedal walking, the process of giving birth
was not
yet the problem
for these females that it was to become for their
descendents. However, the earliest hominid infants would have had feet
more for bipedal walking and less for grasping, and thus may have
needed support
from their mothers, a problem that pongid mothers do not
have to accommodate. Even if their infants were more
precocial than those of
modem humans, and even if the long arms
and stronger
hands of an afarensis
infant would have helped them to cling (especially if their
mothers were hairy,
not recorded),
it is probable
females would have
had to find some way to support poorly grasping
infants. Or perhaps mothers
would have had to restrict
long-distance traveling. Mann's study
of early
hominid dental
development (75) indicates that at least some australopithecine
(perhaps only those of two million rather than 3-4 million years ago)
over roughly comparable periods
to human
today, and thus
more slowly than modem apes. This would also have presented
with an increased
burden, and females would either have had to space birth
intervals widely apart to accommodate dependent children as do modem
or they would have had to find some method to care for
more than one dependent
child at a time. Mann's analysis
of dental
of age at death in australopithecines suggests that life spans were short,
perhaps no more than 25 years. Again, this is the kind of first and second-
order evidence that can be used to construct quite different models. For
example, it can be used to support
the arguments
Lovejoy and Leibowitz
made for a highly stressed, even threatened,
hominid population, or more
conservatively, as simply evidence for a demographic pattem similar to that
of most modem hunter-gatherer
societies and thus not a significant feature.
Interpretations: Modeling Sex Roles and Social
Bonds from Fossil Material
Pilbeam (57, 88) has said that despite their claim to be based on fossil
evidence, most paleontological models of human evolution are relatively
"fossil-free." This is perhaps best demonstrated
through reference to the
following list of the traits commonly focused upon by modelers of early
hominid evolution.
PHYSICAL TRAITS: Upright posture and bipedal walking;
Reduced anterior tooth size and enlarged cheek
brain size;
Increasing hand-eye and fine motor coordination.
TRAITS: Open-country, savannah
Heavy predator
Terrestrial diet.
Hunting and scavenging;
TRAITS: Language;
SOCIAL DYNAMICS: Food sharing;
Division of labor;
"Loss" of estrus in females;
The husband-father
Altricial infants and long dependency periods.
Many theorists have drawn up attribute lists such as the one above, in
which traits found in contemporary
beings, but not in modem apes,
often are projected
back along the hominid record to an assumed very early
at the time of the divergence
hominids and
apes. Howev-
er, it is important
to recognize that fossil evidence for these traits, having
in the earliest
hominids of 3-4 million years ago, only exists for the
first two of the 19 attributes listed. In addition, we have archaeological
evidence for tool use some two million years ago and for aspects of a
scavenging/hunting and gathering subsistence pattern (homebases with
hearths, projectiles) only much more recently, in the middle to upper
Paleolithic. For the majority of these assumed early hominid traits (e.g.
self-awareness, loss of estrus), it is unlikely that we will ever find material
evidence, and thus, as Pilbeam
has argued,
most stories
of human
origins are
by the fossil data, which are used instead to support or
embellish preexisting frameworks of explanation.
For example, the analyses of teeth of the earliest hominids indicate that
they were omnivorous, but they do not make it possible to determine what
of plants and meat occurred
in the diet. Therefore,
a scavenging
or a hunting or a gathering model could claim some support from tooth
and wear patterns,
and theorists
have offered widely different
of the reduction in canine size. Likewise, until the question
sexual dimorphism versus species differences is resolved for the earliest
it is possible to argue for any type of hominid mating
system, unconstrained by the apparent relationship between phenotype and
mating patterns in primates, that is between extreme dimorphism and polygy-
ny or between monomorphism and monogamy. Indeed, Lovejoy appeared to
accept the one, highly dimorphic species argument for Australopithecus
afarensis, while arguing at the same time that they were monogamous.
Finally, the evidence that early hominid infants had poorly grasping feet and
possibly were dependent
for long periods
of time, which may have presented
early hominid mothers with a special problem, can be interpreted in two
opposing manners. First, as in the homebase model, it is possible to argue that
the females became less mobile and more dependent upon males to provision
and protect them, or second, as in the gathering model, we can argue that
females resolved this problem themselves through technological inventions
which in turn led to innovations with wider applicability.
Archaeological Evidence
for Early Human Social Life
The assumption
that the earliest hominids practiced
a way of living that was
somewhere along a direct line between the generalized lifeway of the chim-
panzee and that of the contemporary hunter-gatherer
is best exemplified by
the earlier
work of Glynn Isaac (44-46, 49). Several of Isaac's papers began
by listing the traits that distinguish
Homo sapiens, from the
common chimpanzee, Pan troglodytes, and attempting
to identify the time
periods in the archaeological and paleontological record when these dis-
tinctively human traits first appeared.
As Isaac himself acknowledged
in his
recent papers (e.g. 48), there was a strong tendency
to extrapolate
the modern
traits as far back into the record as possible. The earliest hominids were
credited with complex social, intellectual, and technological abilities, not
quite at the level of modern hunter-gatherers,
but recognizably "human"
More recently, there has been a reassessment of the archaeological
dence for, and interpretations of, early hominid behavior (6, 48, 89, 103,
128). Although
the study of human evolution often has been characterized
heated debate, not all of it enlightening, these recent attempts
to test fun-
damental archaeological assumptions and to develop alternative ways of
explaining the material evidence have been, in Isaac's own words, "liberat-
ing" and an "exciting exercise of alternating leaps of imagination with
rigorous testing" (46, p. 66). Most of these new problem-oriented
studies and
experimental investigations
of the processes that produce archaeological
mains are beyond the range of this review, but the recognition that early
hominids may have been very different
in lifeway from modem humans has
also been liberating
from the perspective
of sex role reconstruction.
And the
most important aspect of this minor paradigmatic
revolution for women's
roles concerns the new interpretations
of bone-and-artifact
associations, or
what were traditionally
known as "home bases."
Isaac, it will be recalled, had developed a "sharing" model which was
founded on the fact that in the early East African
sites of around
two million
years ago, tools are found in dense patches in association with the bony
remains of many animal species. Both stones and bones appear to have been
to "central
locations." Beginning with this one piece of material
evidence, Isaac suggested that humans carried food and possessions to con-
sistent locations as part
of a social system involving home bases, division of
labor, hunting
and gathering, substantial
meat eating, food sharing,
and food
As Potts (89) has noted, Isaac's model could as appropriately
have been entitled a "home base" model as a "sharing" model, since all the
other social characteristics are constructed
upon the initial interpretation
stone-bone associations
are evidence of "social and industrial foci in the lives
of the early hominid tool-makers
to which food was brought
for collective
(48, p. 24).
Several researchers (e.g. 6, 89) have now challenged the home base
Binford (6) analyzed some of the published evidence from
to argue
that the "so-called"
living sites or home bases were in
fact the remains of carnivore activities. Isaac
(48) countered
the published
data sets on which Binford worked were declared by their author (M. D.
to be incomplete
Binford had not accounted
for the fact that the bone assemblages
come from patches
in which thousands
of humans artifacts
(tools) also occur. Thus Potts's (89) detailed, first-hand
analysis of the Olduvai
Gorge and Koobi Fora stone-bone concentrations
to be very influential.
Potts came to a different, but nonetheless startling,
conclusion from both
Binford and Isaac about the processes which formed the bone-stone tool
assemblages. He argued
that the animal bones at these sites were marked both
by carnivore teeth and by stone tools, including tooth marks from gnawing
and cutmarks
by slicing, scraping,
with stone. Somehow,
both early hominids and large carnivores were active at these locations, in
some cases upon the same parts of the carcass, even the same bones.
However, it is not whole carcasses of animals that are represented
and the
bones were not completely processed for meat and marrow, suggesting that
were abandoning
of the available
food. Final-
ly, the incredible density of bones at some of the sites and the patterns
indicate bone accumulation
5-10 years. All four
to Potts, argue against
a home base interpretation
of the sites. The
presence of large carnivores
would certainly
have restricted the activities of
early hominids
at such locations, and
would never
have been
established in such unsafe places. Modem hunter-gatherers
carry whole or
nearly whole carcasses back to camp, not restricted
portions, and they in-
tensively modify the bones of animal food. Finally, hunter-gatherers rarely
occupy a campsite for a long period
of time, and seldom reoccupy an old site.
Thus Potts concluded that it is not possible to assume that the behaviors
associated with home bases (sharing, division of labor) occurred at the early
sites in Olduvai.
How then can one explain the presence of hominids at these sites? Potts
argued that the sites represent
stone tool caches and meat-processing
tions. Because animal carcasses attract
many meat-eaters, the hominids were
forced to transport parts
of the animal
away from the original location where it
was obtained either by scavenging or hunting. These portions
of meat were
taken to the nearest stone tool cache in the foraging area, where raw stone,
manufactured tools, and bones remained from
previous visits. Even chimpan-
zees are known to take food, in this case vegetable food, to consistent
locations where tools have been left for processing (8). It is hypothesized that
the hominids
the meat quickly with the stone tools in the cache and
abandoned the site before direct confrontation
occurred with the carnivores
who were attracted to the remains. Thus, over the years, many remains of
partially processed, gnawed bones and large numbers of stone tools were
in one location. Such sites could represent
the antecedents of
home bases, but Potts believes that until hominids
the controlled
use of
fire to make home bases safe from carnivores, and the first evidence of
controlled use of fire is much more recent in the record, they may well have
continued to sleep in trees and to range widely during
the day as do the other
primate species.
One implication
of this new understanding
of bone-and-artifact associations
for early hominid sex roles is clear: if there is not evidence for home bases
where the sick and the dependent
waited for the well and
the productive,
we can finally free our minds of the image of dawn-age
women and
waiting at campsites
for the return of their
Even though
the sharing model and many other anthropological
scenarios appear
to be
a division of labor in which women return to camp with vegetables and
men with meat, it has almost always been assumed
that women would have
been more tied to the campsites. Women and homes have been inextricably
linked in our cultural
imagery, and thus the shaking
loose from the home base
focus for early hominid social life may allow our imaginations
to turn to
Given the necessarily
limited evidence of social life and the correspondingly
large role played by speculation
in the endeavor to reconstruct
early hominid
-society, it seems appropriate
to ask if it is worth doing at all. As I have
pointed out, many primatologists
and anthropologists oppose such modeling,
often for different reasons than
the one offered by Evans-Pritchard
some years
ago (23), that it is a waste of time to speculate upon unanswerable questions.
Yet origin myths exist in all societies, leading
me to suspect that
humans have
"wasted their time" in just this manner ever since self-awareness
became one
of the hominid characteristics. Indeed, some scholars have argued that
storytelling itself is a defining human trait (see 56); that "our need for
chronological and causal connection defines and limits all of us-helps to
make us what we are" (98, p. 207).
it is hard to imagine
other sciences such as physics attempting
to restrict themselves only to nonspeculative, empirically answerable
tions. Pilbeam has argued that some unanswerable questions in
to be asked because
they help to direct
efforts and channel thinking into fruitful pathways. The problem comes in
knowing which unanswerable questions
to ask" (88, p. 268). Elsewhere in the
same article (and see 57) Pilbeam made it clear that, in his opinion, recon-
structions of early
hominid behavior
would be much
through greater
reference to the actual fossil and archaeological
data. Because contemporary
apes are not necessarily like fossil apes, and because the hominoid fossil
record in any case is virtually nonexistent, Pilbeam has concluded that a
comparative approach
is not likely to yield fruitful theories. It should be
added to Pilbeam's point about these models being "fossil-free" that few
even ostensibly comparative ones, take complete or accurate
account of the primate and ethnographic
data that are available. Zihlman's
most recent publications
on the gathering
model attempt
to account for more
of the data from all three sources than any other model I have seen, and yet
her interpretation
of early hominid life has received no more attention from
the paleoanthropologists
than other less "data-based"
Thus one answer that has been offered to the question
of how theories of
early hominid behavior can be improved
is by giving them a firmer
foundation. However, it is clear that the data-bases of human
evolution will
always remain
limited, and as Isaac has noted, the really important
any model cannot be addressed "purely by recovering bones, stones, and
pollen from layered prehistoric deposits" (48, p. 248). Isaac believed that
there are two related routes to a fuller understanding
of the dynamics of
human evolution. The first is an emphasis on problem-oriented
and ex-
perimental studies of the processes that might have led to characteristic
archaeological remains by making use of analogous modem activities and
The second is that
should be expressed
as a series
of falsifiable, alternative
hypotheses, and tests should involve attempts
intuitively favored hypotheses. His suggestion was that reconstruc-
tions would be better
served by each researcher
a series of alterna-
tive models, rather
than promoting
and defending a single model.
Both Pilbeam's and Isaac's suggestions reflect the view that greater
scientific rigor will solve, or help to solve, the problems of subjectivity
models of human evolution. But another
analyst (56) has suggested that
it will
probably be impossible to remove the subjective or "storytelling"
from evolutionary
accounts. Landau argued
that many scientific theories are
essentially narratives,
that is, the creative piecing together of an organized
and plausible
of events by application
of the imagination
to standard
forms. Particularly
with its description
of the events of
human evolution is, in her view, a form of storytelling, open to narrative
analysis (see also 86). Landau
did not address the question
of the part
by "fiction"
in human
evolution models which would be one implication
of a
literary analysis to which many scientists would object. But surely any
modeler would agree that it takes creativity as well as data to create a
plausible account of human evolution. And neither creative nor scientific
minds function in a cultural
vacuum. Landau's narrative analysis attempts
make some of the implicit structural
guidelines of any human
origins model
explicit. I recount her approach
here with the suggestion that the process of
to bring
and structures into the open will allow
us, if not to eliminate or even agree upon them, at least to become critically
aware of their potential presence and influence.
took a structural
that looks for common elements in the
versions of the human evolution story. For example, she identified
four major events or episodes that are consistently emphasized by paleoan-
a shift from trees to the ground
posture (bipedalism);
the development
of the brain, intelligence, and
language (encephalization);
and the development
of technology, morals, and
society (culture). She suggested
that the question
of which episode came first
has been a major
source of debate since Darwin, but in all versions the same
episodes are recognized. She then argued that the diverse theories of what
in human
evolution actually
follow a common narrative structure.
This structure takes the form of a "hero story"
in which the protagonist
hominid) starts from humble origins on a journey in which he will be both
tested by environmental
stresses (savannah
etc) and by his own
weaknesses (bipedalism,
lack of biological armaments),
by power-
ful agents (intelligence, technological
inventions, social cooperation)
until he
is able to transform
himself into a truly human hominid, the hero's final
which always ends the story.
this approach
to human evolution accounts not as a criti-
cism but as a demonstration
to scientists that they are interpreters
of text as
well as of nature,
and as a potentially
useful tool in comparing
structural and
conceptual differences between theories. If she is correct that human evolu-
tion theories follow a common narrative
and adhere to a recogniz-
able literary
model (the hero's tale), which can be traced back through many
centuries of European
storytelling, this approach
may give us some insight
into why women generally
play a subordinate
role in these stories. For clearly
the tale of the hero is about men and not heroines;
women function in such
stories either as secondary
characters (mothers,
sisters) related
to the hero, or
as potentially desirable sexual partners, often in need of rescue. If the
Western raconteurs
of human evolution had been raised in
different narrative
for example learning as children the enduring
Chinese legend of the woman warrior,
the female troubleslayer
who rides into
adventures carrying her infant in a sling inside her armor, then perhaps
women would not have been so consistently restricted to the merely
roles in our origin stories.
I have argued
that one recurring
theme in the human evolution accounts,
from Darwin
to Lovejoy, is that
early men were the achievers, the producers,
and technological innovators; whereas early women were limited by the
of bearing and rearing
children. Or as Sacks (95) has
put it: men make culture and women make babies, two mutually
activities. Anthropologists
have long applied
sets of dichotomous
the roles of men and women in human
society: public/domestic,
However, a number
of women anthropologists
(e.g. 60, 95) have begun to challenge these dichotomies as being largely a
of the Western
belief in the opposition
of the sexes that
been mistakenly generalized
into a universal
and "natural"
These dichotomies are also present
as hidden assumptions
in most models of
origins, and
yet we do not know how generally
they express
the human
today, much less in the past. For example, foraging
societies do not
have secluded family units or households within the band, nor are women
confined to campsites. Thus a discussion of public/male versus domestic/
female spheres
has not been a particularly
to understand-
ing the lives of these people. If Potts (89) is correct
that our early ancestors
lived without home bases, the domestic, "house-bound"
vision of early
women becomes singularly inappropriate.
A similar inapplicability may exist for the productive/reproductive
chotomy. Does a foraging woman or a foraging society functionally com-
lives and activities into these two supposedly
realms, or is this merely an abstract
and possibly ethnocentric
tion of how lives should be arranged?
Is it necessary to assume, as does
Lovejoy for example, that the human female's energy is so limited that
activities must necessarily
be detrimental
to reproduction,
that the
behaviors involved in subsistence and child rearing are incompatible and
mutually exclusive? One of the peculiar human phenomena that an-
thropologists have identified is that it is possible for people to widely and
passionately hold cultural
beliefs that are in direct contradiction
to their social
actions. I suggest that this is the case with our own cultural
belief that the
people who are reproductively engaged cannot be productively active, a tenet
clearly belied by the sexual makeup of the workforce
in our society today.
When Spencer first articulated this "ideal"
in Victorian
England (called by
historians the "cult of female domesticity"), women of the working classes
were widely employed in industry
at the same time as they were reproducing
at a rate alarming
to the social Darwinists.
in all social systems except
those based on intensive agriculture
or some forms of pastoralism,
the same
women who bear and rear offspring always have contributed
actively to
subsistence, and in many societies they are even more
for produc-
tion than the men. The assumption
of female domesticity
has functioned
as a
pillar in the construction of most theories of human social evolution, and yet
its accuracy and applicability
have never been openly debated.
Theories of human origins do function as symbolic statements about and
indeed prescriptions
for human nature. By making the assumptions
of any
more explicit, one can test or debate
them rather
to act
as though
between models reflect only varying descriptions
of the
material evidence. When the evidence changes, as when gathering replaces
in economic importance,
but the implications
for men and women are
seen to remain fundamentally
the same, as when Man the Hunter
Man the Provider, it is clear that powerful cultural sex role expectations
inform these reconstructions even more strongly
does material evidence.
Some readers
may find it hard to accept that cultural
beliefs and narrative
traditions play a significant role in scientific models of human evolution.
However, I would argue that the theories
reviewed in this paper do combine
the realms of science and of storytelling.
If this is so, we can begin the useful
exercise of learning
to analyze how the two realms interact and overlap in a
given model and how we can evaluate the model according to the criteria
to each realm. To paraphrase
Kermode (51), if we cannot free
ourselves of subjectivity, then we must attempt
to make sense of it. People
will not stop wanting to hear origin stories and scientists will not cease to
write scholarly
tales. But we can become aware
of the symbolic content of our
stories, for much as our theories are not independent
of our beliefs, so our
is not independent
of our theories of human
society. In these origin
tales we try to coax the material
evidence into telling us about the past, but the
we weave about the past also tells us about the present.
1. Scientific models of early human social life are not simply plausible
inferences from the material evidence, but also function as statements of
human nature. Such models rely heavily upon speculation, which often is
culturally informed. Some social and natural scientists
doubt the value of such
2. Darwin developed the theory of sexual selection to account for second-
ary sex differences. Applying it to human traits, he argued that men were
selected for courage, intelligence, and technological abilities, whereas
women were selected for generalized maternal attributes, only acquiring
intelligence secondarily through males.
3. Nineteenth century social theorists believed all societies go through
common stages: promiscuity, matrilineality, and patrilineality. However,
they disagreed on whether or not women benefit from increasing social
control of human sexuality. The rejection of social Darwinism and the
collection of systematic
ethnographic data led to the abandonment of theoriz-
ing about human social evolution in anthropology
until the 1960s.
4. Man the Hunter, a model drawn from primate, ethnographic, and
archaeological evidence, became the dominant theory of the 1960-1980
period. Although
from Darwin's scenario
in evidence and concepts,
Man the Hunter represents
a continuation
of his belief that only male traits
were selected and that women play an insignificant
in human evolution.
5. A reappraisal
of the primate, ethnographic,
and material data led some
anthropologists, most of them women, to propose a "countermodel"
Woman the Gatherer, in which gathering, sharing, and tool use were de-
scribed as female inventions,
to the evolution of humans.
Both hunting
models and gathering models appeal to similar sources of evidence, yet
present opposing and mutually
exclusive accounts of human social evolution
and thus of human nature.
6. The most widely discussed current
theory, Lovejoy's male provisioning
model, makes male gathering
and the provision of sedentary, highly fecund,
and monogamous females the central adaptation
in human evolution, and is
premised upon the supposed failure of pongid reproductive
life history pat-
7. Appeals to theorists to tie their models more closely to empirical
evidence and to account for more of the evidence led to modifications,
especially to the gathering
model and to some extent to the hunting model,
where the importance
of scavenging was recognized. However, the resulting
models have received no more attention than more speculative ones.
8. Isaac advocated more testing of experimental models with multiple
hypotheses to increase
the rigor of the models and to reduce the advocacy of
modelers. Yet it seems unlikely that increased scientific data, or rigor, would
solve the problem
of subjective interpretations.
Most of the important
that define sex-role differentiation
are intangible.
9. If theories
of human
evolution are seen as narratives
as well as scientific
discourse, the literary
of the structure
of origin stories
might allow us
to distinguish the subjective from the empirical, the art from the science.
Reconstructions of the past are in some respects also reflections of the
This paper was written with the help of a Release-Time Grant from the
Faculty of Arts, University of Alberta, and with the research and editorial
of Larry Fedigan. I thank C. Lanigan
and N. Collinge for help with
the manuscript, and NSERC (#A7723) for continuing support of my re-
Literature Cited
1. Ardrey, R. 1961. African Genesis. New
York: Dell
2. Ardrey, R. 1976. The Hunting
esis. New York: Atheneum
3. Bachofen, J. 1861. Das Mutterrecht.
Basel: Benno Schwabe
4. Badrian, A., Badrian, N. 1984. Social
organization of Pan paniscus in the
Lomako Forest, Zaire. In The Pygmy
and Behavior, pp. 325-45
5. Berndt, C. H. 1981. Interpretations
in Aboriginal
See Ref.
13, pp. 153-203
6. Binford, L. R. 1981. Bones, Ancient
Men and Modern Myths. New York:
7. Blanc, M. 1983. La bipedie, une affaire
de sexe? Recherche 141:240-43
8. Boesch, C., Boesch, H. 1981. Sex dif-
ferences in the use of natural hammers
by wild chimpanzees. J. Hum. Evol.
9. Brace, C. L. 1973. Sexual dimorphism
in human evolution. Yearb. Phys. An-
thropol. 16:31-49
0. Cann, R. L., Wilson, A. C. 1982. Mod-
els of human evolution. Science
[1. Crook, J. H. 1980. The Evolution of
Human Consciousness. Oxford: Clar-
[2. Cucchiari, S. 1981. The gender revolu-
tion and the transition from bisexual
horde to patrilocal band. In Sexual
Meanings. The Cultural
Gender and Sexuality, ed. S. B. Ortner,
H. Whitehead, pp. 31-79. Cambridge:
Univ. Press
[3. Dahlberg, F., ed. 1981. Woman the
Gatherer. New Haven: Yale Univ. Press
14. Dahlberg, F. 1981. Introduction. See
Ref. 13, pp. 1-33
15. Darwin, C. 1936 (orig. 1871). The Des-
ent of Man and Selection in Relation to
Sex. New York: Modem Library
16. Darwin, C. 1936 (orig. 1859). The Ori-
gin of Species. New York: Modem Li-
17. Debetz, G. F. 1961. The social life of
paleolithic man as seen through the
work of Soviet anthropologists.
See Ref.
115, pp. 137-75
18. Deetz, J. 1968. Hunters
in archeological
perspective. See Ref. 65, pp. 281-85
19. DeVore, I., Washburn,
S. L. 1963. Ba-
boon ecology and human evolution. In
African Ecology and Human Evolution,
ed. F. C. Howell, F. Bourliere, pp. 335-
67. Chicago: Aldine
20. Ember, C. 1978. Myths about hunter-
Ethnology 17:439-48
21. Engels, F. 1972 (orig. 1884). The Ori-
gin of the Family, Private Property
the State. In the Light of the Researches
of Lewis Henry Morgan. New York: Int.
22. Estioko-Griffin, A., Griffin, P. B. 1981.
Woman the hunter: the Agta. See Ref.
13, pp. 121-51
23. Evans-Pritchard,
E. E. 1965. Theories
of Primitive Religion. Oxford: Clar-
24. Fedigan, L. M. 1982. Primate Para-
digms. Sex Roles and Social Bonds.
25. Fee, E. 1976. Science and the woman
problem: Historical
perspectives. In Sex
Differences, Social and Biological Per-
spectives, ed. M. S. Teutelbaum, pp.
175-223. Garden
City: Anchor
26. Fernandez,
C. A., Lynch, F. S. J. 1972.
The Tasaday: Cave dwelling food
of South Cotabato,
Philip. Soc. Rev. 20:279-330
27. Fisher, H. E. 1982. The Sex Contract.
The Evolution
of Human Behavior. New
York: Quill
8. Fox, J. R. 1967. Kinship
and Marriage.
Baltimore: Penguin
Z9. Freeman, L. Jr. 1968. A theoretical
for interpreting
materials. See Ref. 65, pp. 262-67
30. Friedl, E. 1975. Women
and Men. An
View. New York:
& Winston
31. Friedl, J., Pfeiffer, J. E. 1977. An-
thropology. The Study of People. New
York: College Press
32. Galdikas, B. M. F., Teleki, G. 1981.
Variations in subsistence activities of
female and male pongids: New per-
spectives on the origins of hominid
division. Curr. Anthropol. 22:241-320
33. Gayle, F., ed. 1978. Women's
Role in
Aboriginal Society. Canberra: Aust.
Inst. Aboriginal Stud.
34. Goodall, J. 1968. The behavior
of free-
living chimpanzees of the Gombe
Reserve. Anim. Behav. Monogr.
35. Goodall, J. 1976. Continuities
human behavior.
In Hu-
man Origins. Louis Leakey
and the East
African Evidence, ed. G. Isaac, E. R.
McCown, pp. 81-95. Menlo Park, CA:
36. Gould, R. A. 1981. Comparative
gy of food-sharing in Australia and
Northwest California. See Ref. 39, pp.
37. Hamburg,
D. A. 1968. Primate
and the evolution of aggression. See
Ref. 65, pp. 339-41
38. Hamburg,
D. A., McCown, E. R., eds.
1979. The Great Apes, Perspectives on
Human Evolution, Vol. 5. Menlo Park,
CA: Cummings
39. Harding, R. S. O., Teleki, G., eds.
1981. Omnivorous
Primates. Gathering
and Hunting in Human Evolution. New
York: Columbia Univ. Press
40. Harley, D. 1982. Models of human
evolution (letter). Science 211:296
41. Hayden, B. 1981. Subsistence and eco-
logical adaptations of modern hunter-
gatherers. See Ref. 39, pp. 344 421
42. Hiatt, B. 1978. Woman the gatherer.
See Ref. 33, pp. 4-15
43. Hill, K. 1982. Hunting and human
evolution. J. Hum. Evol. 11:521-44
44. Isaac, G. 1971. The diet of early man:
Aspects of archeological
evidence from
lower and middle Pleistocene sites in
Africa. World Archeol. 2:278-99
45. Isaac, G. 1978. The food-sharing be-
havior of protohuman hominids. Sci.
Am. 238:90-106
46. Isaac, G. 1980. Casting the net wide: A
review of archeological
evidence for ear-
ly hominid
and ecological rela-
tions. In Current Argument on Early
Man, ed. L. K. Konigsson, pp. 226-51.
Oxford: Pergamon
47. Isaac, G. 1982. Models of human
tion (letter). Science 211:295
48. Isaac, G. 1984. The archeology of hu-
man origins: Studies of the lower Pleis-
tocene in East Africa 1971-1981. In
Advances in World
Archaeology, ed. F.
A. E. Close, pp. 1-87. Orlan-
do: Academic
49. Isaac, G., Crader, D. 1981. To what
extent were early hominids
An archeological perspective. See Ref.
39, pp. 37-103
50. Johanson,
D. C., White, T. D. 1979. A
systematic assessment of early homi-
nids. Science 203:321-30
51. Kermode,
F. 1967. The Sense of an End-
ing. Oxford: Oxford Univ. Press
52. Kuroda, S. 1984. Interaction
over food
among pygmy chimpanzees.
See Ref. 4,
pp. 301-23
53. Lancaster,
J. B. 1975. Primate
and the Emergence of Human Culture.
New York: Holt, Rinehart
& Winston
54. Lancaster,
J. B. 1976. Sex roles in pri-
mate societies. See Ref. 25, pp. 22-61
55. Lancaster, J. B. 1978. Carrying and
sharing in human
evolution. Hum. Nat.
56. Landau, M. 1984. Human evolution as
Am. Sci. 72:262-68
57. Landau, M., Pilbeam, D., Richard, A.
1982. Human origins a century after
Darwin. Bioscience 32:507-12
58. Laughlin, W. S. 1968. Hunting. An in-
tegrating biobehavior system and its
evolutionary importance. See Ref. 65,
pp. 304-20
59. Leacock, E. B. 1972. Introduction.
Ref. 21, pp. 7-67
60. Leacock, E. B. 1978. Women's status
society: Implications
for so-
cial evolution. Curr. Anthropol. 19:
60a. Leakey, R. E. 1981. The Making of
Mankind. New York: Dutton
61. Lee, R. B. 1968. What hunters
do for a
living, or, how to make out on scarce
resources. See Ref. 65, pp. 30-48
62. Lee, R. B. 1974. Male-female residence