A Conversation With Joan Vincent

A Conversation with Joan Vincent
Author(s): David Nugent

Source: Current Anthropology, Vol. 40, No. 4 (August/October 1999), pp. 531-541
Published by: The University of Chicago Press on behalf of Wenner-Gren Foundation for
Anthropological Research
Stable URL: http://www.jstor.org/stable/10.1086/200050 .

Accessed: 20/05/2015 08:58
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The University of Chicago Press and Wenner-Gren Foundation for Anthropological Research are collaborating
with JSTOR to digitize, preserve and extend access to Current Anthropology.
http://www.jstor.org
This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

All use subject to JSTOR Terms and Conditions
Reports
A Conversation with
Joan Vincent1
d av i d n u ge n t
Department of Anthropology, Colby College,
Waterville, Me. 04901-8847, U.S.A. 25 ix 98
DN: From conversations we have had in the past, I
know that experiences in your early life gave you a keen
sense of the importance of national boundaries, national identities, and also of landscape and space.
JV: Yes. I was raised in England, but my fathers family
came from Alsace[-Lorraine], which is between Germany and France. And therefore the question of identities has always been very critical for me. My father was
an antiquarian, and when I was a child he would take
me out in the car when he had long trips to Big Houses
to look at their furniture, talking about the history buried in the landscape. I was born in 1928 and grew up
during the depression and the war. My father worked
for the Free French, and my mother was a firefighter and
air-raid warden. I remember shrapnel coming through
the front door and my father picking it up, not realizing
it would be hot. After the war only the vets got places
in university, so instead I went to teacher training college and then taught secondary school.
DN: And then you went to Africa?
JV: Yes. I had begun to write fiction, and a journal called
The Guide paid me 98 pounds for a storyenough for
an ocean voyage. I ended up in Northern Rhodesia,
where I realized that I would get as much time as I
wanted to travel and see the country by becoming a
schoolteacher.
DN: Was the traveling and teaching enlightening, eyeopening, about events that were happening at the time?
JV: Not really. I was very simple-minded. When I look
back at my diary, life was full of bridge parties and cinema and boyfriends and tennis and the bush. A very colonial life in that respect. I thoroughly enjoyed it. The
other thing I did during my year there was sit the exam
1. 1999 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/99/4004-0005$1.00. This interview has been extracted from the transcript of a significantly
longer conversation between Joan Vincent and David Nugent that
took place on June 5, 1997, in Vincents New York City apartment.
Joan Vincent.
for a university correspondence course at LSE [London

School of Economics] in what was called B.Sc. (Econ.)
Part 1. To prepare for this I had to drive down to Capetown, because I didnt have a library. And there was one
experience Ill never, never forget. I was having to read
political philosophy, which meant I had to read Marx,
and in the University of Capetown library, in a little
room, the Communist Manifesto was chained to a reading desk!
DN: No! Thats astonishing.
JV: I swear it! And in spite of Bob Murphys teasing me
that it was only after 68 I became a Marxist, that was
a very formative experiencethat any book could be
seen to be that important! But most of my swotting was
on 13th-century English castle building, and Marx
didnt fit awfully well. I passed the exam, which meant
that I could go to LSE to pursue economic history or social anthropology. Nobody in my family had a university education, so I didnt know what I was doing. I
went to LSE and said to the porter at the door, I have
passed B.Sc. (Econ.) Part 1. Whom should I see? And
he said, Well, Maurice Freedmans in today. Why dont
you go up and see him? So I went up to the fourth
floor, and Maurice Freedman welcomed me to anthropology. The next day, in my first session, I walked into
531

532 c ur r e n t an t hr o p o l o g y
a seminar about Jinghpaw terminology. And I think my

preference for process over structure was born that day!
DN: Indeed.
JV: I was at LSE for nine months. When I got my degree I
still wasnt career [-oriented]. Someone offered me a job
doing consumer interviewingthat was the way they
saw graduating women students. So instead I applied for
and got a job teaching school for the sultan of Zanzibar.
DN: Nothing closer to home than Zanzibar?
JV: Oh, no. I wanted to travel. All of my family have
traveled. My mother was born in India. All my mothers
sisters were in the Indian Army Education Corps, but a
very lower-middle-class or upper-working-class/subaltern teaching. I have an imperial thread, then, but so
have 80% of the worlds population born when I was.
And all the time I wasnt resistant at all to this way of
life. I was having a whale of a time.
DN: Who were your students in Zanzibar?
JV: I had four and a half years lecturing in history at the
Teacher Training College. My students were all Zanzibaris, of all races. And I got them going out into the island doing oral history. All the time I was there the island was on the verge of independence. There were very
tense elections and also preelection riots [in June 1961]
and horrible massacres, mainly of the Manga Arabs out
in the bush by the mainland African workers on the
clove plantations. That spread into violence in the
towns. For me, it meant two things. First, my friends
there (who were all Zanzibaris) got split between the
parties not by race but by politics. Second, as a European I had to join the B Specials,2 and I was asked to do
research into the relationships between those who
killed and those who were killeda type of Arensbergian inquiry (i.e., who does what to whom and why).
DN: Did that inform your later interests as an anthropologist?
JV: Definitely! It was the sort of research Ive loved to
do ever sincefirst to see what is really going on and
then to look behind the event, to deconstruct it. My
days were spent doing that, and in the evenings I had
the horrible experience of working in the hospital. It
was there I learned what panga [machete] savagery is,
because the typical way of bringing somebody down
was first to strike across the legs, just below the knees,
and when the victim had fallen attack the head. It was
a pretty ghastly nursing experience. Most of the people
who reached the hospital did so too late.
DN: After this experience in Zanzibar, you went to the
University of Chicago for graduate work?
2. A reserve police unit.
JV: Yes. I was accepted by the anthropology department,

but at registration I suddenly had qualms over the
course offerings, which seemed like what Id been doing
at LSE. I just remember Durkheim standing out like a
guard in front of me, so I moved to political science.
After those four years in Africa and on the eve of independence, comparative politics was much more my
thing than reading Durkheim all over again.
DN: When you look back at Chicago, what stands out?
JV: Hannah Arendt and Victor Turner. Hannah Arendt
came and gave a series of guest lectures, and I was in
the crowd.
DN: It doesnt sound as if you had a great deal of contact
with anthropology.
JV: But I did. The problem with Chicago was that the professors one most wanted to work with were away, but I
sat in on everything I could. I didnt disown anthropology;
I just put it into comparative politics. You see, because I
got to know Africa before receiving my degree I was very,
very critical of how Africa was represented in anthropology. Studies of kinship and lineages were instructive, because they were constraining on people, but there were
more important things, like the use of military force and
labor migration. That linked very much with the work
the Manchester school was doing in Central Africa. The
Mancunians were really in the 20th century, and their
anthropology was way ahead of the game. If you think of
what they were actually reporting, you can reanalyze; I
could rework their anthropology, unlike the studies of social structure, in terms of my own experience.
DN: By the fall of 1964 youd finished your Masters
[1964] at Chicago and had started in the Columbia
Ph.D. program in anthropology.
JV: Yes. My first year was entirely four-fields, and on
the exam at years end I was asked to justify the four
fields. I could justify three quite easily. Linguistics, for
me, was very important, because languages are. Archeology was, because it was history. But the physical anthropology I did have problems with.
DN: They werent doing work on race?
JV: Not that I remember.
DN: Was it a shock to encounter the kind of racism that
one found and finds here?
JV: I hate to say it, but I wasnt very sensitive to it until
the long, hot summers. It was when Newark burned
[1967] that it hit me and I saw racism as a national problem. I didnt see it as a personal problem. I didnt see it
as a university problem, though I probably should have.
Class, not race, was then uppermost for me. Growing

Volume 40, Number 4, AugustOctober 1999 533
up there had always been class. And always nationality,

because of my French father.
DN: Did you get some fieldwork training at Columbia?
JV: I learned about fieldwork from the Wednesday Seminar, when Joan Mencher, fresh from India, gave a beautiful talk, with beautiful maps and diagrams. I learned
most from the substantial analysis that somebody did.
One didnt get training in it.
DN: Thats interesting, because you have a very strong
emphasis on fieldwork.
JV: Well, as you know, with my own students what I
have done is simply get out a whole batch of my field
materials, and we go through them together. Its a
hands-on thing for me.
DN: What is distinctive about the way you approach
fieldwork?
JV: The Mancunian word for it is nitty-gritty
down to earth, really. But there is also a gendered dimension to this. Women academics are marginalized a
great deal. They have generally read themselves as a
margin in narratives conceived and written by those
who have discursive power. This also relates to fieldwork. Women in the field, especially women alone in
the field, are very vulnerable. You rarely see yourself as
controlling situations in the way fieldwork is generally
represented.
DN: To what extent did your doctoral committee at Columbia influence the research program that you were
developing in Uganda?
JV: Very little. I dont remember going to see anyone
about anything. I did appreciate [Andrew] Vaydas training in ecology. Its just that I could never close my people off; they were moving too much to mess up the
kind of systems analysis he wanted to talk about. But
his actual involvement in Indonesia, the applied side,
comes out of Columbia. I also got reinforcement there
for wanting my anthropology to be engaged with the
real world, so that when the ethnicity article [1974b]
got written, it wasnt wholly academic. It was because
those blue-collar racists were saying, We all have the
same chances. But in my graduate training Connie
[Arensberg] had more influence than anyone else, in
the sense that he was always saying things like, What
do you really mean by structure?making one define
terms clearly. Also, I liked his sense of needing diagrams, place, space, needing movement of people. I
liked what he was doing very much.
fieldwork at the same time.3 Makerere was very exciting back in those days. Very small groups of students,
all African, but from East Africa, generally. It was the
elite university of East Africa. I planned to work south
of Lake Victoria in Tanzania, in a multiracial, European-African-Asian community. But when I got to
Makerere I realized that that [project] was impractical,
so I went to Teso district.
DN: You had gone to the field with a multiracial research problem to begin with?
JV: Yes. My Africa as opposed to anthropologys Africa.
DN: And the problem that you worked on in Teso was
very similarnot polyracial but certainly polyethnic.
JV: Yes. In fact, having come from Zanzibar, the problem as I perceived it was, Why arent these people at
each others throats? I wrote a very harmony-oriented
dissertation. It was only when George [Bond] knew Id
had it accepted for publication that he suggested a chapter on competition or conflict. George has been a good
friend. But the interesting thing about the actual dissertation I wrote is that it had a very large historical component. I did no additional research between my dissertation [1968], published as African Elite [1971], and
Teso [1982]. In the dissertation it was just [Robert] Murphy saying, What you dont want is thishe cut out
the historical material.
DN: The distinctiveness of your approach really comes
out here. You consciously located your field site on an
industrial frontier of colonial capitalism, you were doing quite detailed historical work, and you were looking
at local, regional, and national contexts.
JV: This was the Mancunian influence. The world is
there in their work. The Africans in their monographs
are members of the Catholic church, of the Communist
party, of industrial workforces.
DN: Lets turn to your years as an assistant professor,
from 68 to 72. This was when your first published
work on Uganda began to come out, some of which
deals with cooperatives and issues of economic development [1969, 1970, 1976d, 1988a].
DN: Lets talk about your fieldwork in Uganda.
JV: Once you were a fellow at the East African Institute

of Social Research you got involved in all the research
problems being addressed there, including this one. But
I have always thought that cooperation hasnt been appreciated enough as a form of politics and had made the
cotton marketing co-op in Gondo part of my field research from the beginning. I was also able to link as-
JV: It came about because in 1966 I got a letter from Colin Leys asking if I would teach political sociology at
Makerere [University College, Kampala], be a fellow at
its East African Institute of Social Research, and do
3. Leys had been a visiting professor at the University of Chicago

when Vincent was earning her Masters degree in political science
there.

pects of the cooperative organization with the instrumental activity of its key figures [1974c].
involved in the antiwar movement, and there was a big

confrontation on campus.
DN: Let me ask you about African Elite, in which you

focus on polyethnicity and the emergence of a rural
elite in Teso. In reading this book one is struck by the
fact that its content is strongly historical materialist
but at the same time there are no references to works
of an explicitly historical materialistic nature. Within a
couple of years of the publication of African Elite, however, in your published work [1974b, 1977b] you refer to
people like E. P. Thompson [1963], Lenin [1964 (1899)],
Engels [1958], and others who are clearly in the Marxian
tradition. Furthermore, when I was in graduate school
I heard a story about interactions that you had as a graduate student with Marvin Harris when he was preparing
The Rise of Anthropological Theory [1968]. According
to this bit of Columbia folklore you were insisting on
the importance of the Scottish political economists in
making sense of Marx. All this suggests a deep familiarity
with political economy from an early date. In light of this,
it seems as if African Elite could have been more explicitly historical materialist than it was. Considering the arguments that you make later on, in works like Anthropology and Politics [1990a] and Engaging Historicism
[1991b], about the various conditions that affect what can
and cannot be written and what is and is not said at any
particular point in time, what were some of the conditions that affected how African Elite was written?
JV: Thats right! Once again [as in Zanzibar], I was

seeing people I highly respected torn between two factions.
JV: It was written as a doctoral thesis, and very quickly,

when I first came back from the field. Given Murphys
cutting out of anything historical, what was published
as African Elite was a synchronic account of the social,
political, and economic relationships where I had done
my fieldwork. I defended in May of 1968, so the thesis
was written prior to the key political developments of
that year. And 68 was a very, very important time. Just
as the feminist movement hit me a few years later, so
did 68. It certainly made me read a whole set of different things, in part because the people around me were.
It also embodied the critical things that were being exposed at the time with respect to the department, the
university, the country, the lot. The exposes had a direct appeal for me because this is the way Ive always
tried to understand somethingby looking behind and
by diagramming and showing this connection with this.
So why the E. P. Thompson and the rest? I think all of
usBill Roseberry, certainlywere moving toward
this. My knowledge of historical materialism came out
of reading Ramparts, from reading what came out in
68, and linking up with the MARHO group,4 and reading the Review of Political Economy.
DN: That must have been a very exciting time. That
was also the period when the Columbia department got
4. The Mid-Atlantic Radical History Organization, a collective of
critical social scientists in the New York City area that was formed
in the 1960s. MARHO originally published the MARHO Newsletter and since 1975 has published the Radical History Review.
DN: Lets talk about The Structuring of Ethnicity

[1974b]. This piece has had a significant impact on the
field, and it is novel and important in a number of ways.
You take a very highly original approach to the whole
topic. You stress the need to make the study of ethnicity part of the comparative study of different forms
of inequality, like caste and class and race. You focus on
the centrality of boundary-making mechanisms, which
may or may not come from the state, but you really
stress the pivotal role of the state in different constructions of inequality. You talk about the relevance of ideology, about who controls ways of understanding and
what is considered true, and you also stress the contingent and contextual nature of ethnicity as a whole. Can
you talk about who your major influences were? I think
weve already heard about some of them.
JV: I think, with respect to ethnicity, my own marginality. I had a very clear sense of being a very English
woman in America. This notion of being a resident
alien anthropologist in the United States was certainly
there when I wrote In the Shadow of the Armory
[1989a]. But I really wrote the ethnicity article out of a
kind of anger that people could say that it was equally
easy for anyone to get to the top in a society like this
ignoring color, class, capitalism, ignoring everything. I
am very impressed by what you saw in the article; I
must read it sometime. I think it was timely, probably.
DN: Yes. I think it was both timely and . . .
JV: Dont you dare say ahead of its time [laughing].5
In the New York Times there were letters galore about
ethnicity, and the whole situation seemed so impossibly unjust. So I wrote it as a political statement.
DN: You started your research project on Ireland at
about the same time that you wrote The Structuring
of Ethnicity, in 73. Why Ireland?
JV: Because Amin had taken over in Uganda. It took
friends who worked in Uganda five or six years to feel
they could go back without absolute distress, and I just
didnt want to get involved in that. I got a Guggenheim6
to investigate the roots of the ongoing political conflict
[see 1989c, 1991a]. With my Uganda work I thought I
was going to the most peaceful, harmonious country in
5. The allusion here is to her review of the work of Alexander
Lesser (1988b), where her emphasis is on showing how Lesser was
a man of his time although today he is frequently regarded as having been ahead of it.
6. A John Simon Guggenheim Memorial Foundation Fellowship for
fieldwork in Northern Ireland (197374).

East Africa. When that went sour, with Amin, I felt

kind of qualified to work in Northern Ireland. With
Uganda I felt I couldnt be critical, whereas in Northern
Ireland, given its present status as part of the United
Kingdom, I can be as critical as I want to be.
DN: Why County Fermanagh?
JV: Because it was on the borderbecause, given the
1922 treaty, it should have been in the south. It had a
very large Catholic population, and the Boundary Commission proposed dividing Ireland on the basis of religion, but the whole of Fermanagh got placed into the
north. Once I started researching the roots of the present condition it was fascinating, because the IRA had
actually fought against British troops in Fermanagh.
They invaded it in May 1922 and again in the 1950s. It
was contested space.
DN: In many of your works, past and present, you make
it clear that while the great European social theorists of
past times have a great deal to contribute to our field,
the unique contribution that we can make comes
from fieldwork-based analysis, which drives innovative
new theory and at least has the potential to contribute
to comparative understandings in the present. You
develop this argument in Anthropology and Politics
[1990a], in System and Process [1986], and in Engaging Historicism [1991b]. Furthermore, in Engaging
Historicism, you make a distinction between ethnographies that empower the reader and those that dont.
Those that empower the readeryou call them rich
multivocal analysesgive the reader something to
fight back with. And these ethnographies, you argue,
are based on fieldwork. Why does new theory follow
from innovative fieldwork and not vice versa?
JV: Im very concerned about the term theory. I dont
think anthropology is a theoretical discipline. This is
my problem. I am very unhappy with constructed
knowledge, whether its state statisticsthe categories
that the state formsor academic constructivism, and I
suppose my emphasis on fieldwork is a kind of resistance.
Theres somebody out there talking, and one is capturing
something that is not in the systems one comes out of.
My feeling for pacificism, globalism, and anticapitalism
doesnt allow for constructed knowledge.
DN: This sounds, in many ways, like an argument
against hegemony.
JV: Yes, and thats a fairly new word in our discipline.
Its very important to take into account the meaning of
a word in the time it is used.
DN: By the mid-seventies one senses in your work an
increasingly direct engagement with issues of social differentiation, colonialism, capitalism, and class. One
can see this change in, for example, Colonial Chiefs
and the Making of Class: A Case Study from Teso,
Uganda [1977a], Teso in Transformation: Colonial

Penetration in Teso District and its Contemporary Significance [1978b], and Political Consciousness and
Struggle among an African Peasantry [1983]. It is most
clearly reflected, however, in Teso, where you talk
about the formation of a peasantry, a rural proletariat,
a chiefly class, and the development of underdevelopment. If one compares African Elite with Teso, the shift
in focus from polyethnicity to class formation is pretty
clear. Youve already explained that you did not consult
any new sources for Teso and that you originally had
a chapter on history in your dissertation that Murphy
discouraged you from including.
JV: Im not sure it was a chapterit was a third of the
manuscript.
DN: Can you talk about your intellectual development
during the seventies in relationship to the new kinds of
questions you address in Teso?
JV: Much of the reading that I did during this period was
in connection with a course I taught at Barnard. I inherited it from Joan Mencher, a visiting lecturer, who
called it Societies in Transition. I focused on the
transformation of agrarian society, the development
of agrarian capitalism and its concomitant social
changes. This led me to read much more on peasantries,
pure and simple. I read the Latin American material for
the first time, because I needed to know more about
smallholding as a notion and more about haciendas or
estates or whatever youre going to call them. Lenin had
always been there, as a need, from the point of view of
simply where the manure comes from. So had the really
basic things, like Slicher van Bath [1963]. So by then I
was reading pretty broadly on agrarian society.
DN: In what ways did your conceptual frame of reference shift between African Elite and Teso?
JV: Things were coming together analytically in the
1970s, partly because of the Latin American material.
This was a period when the development/underdevelopment argument was being mountedconsider the rethinking that took place along these lines in the seventies. Teso came out in 82, so I probably finished it in
1980. The fact that I was reviewing European work was
also important.
DN: Youre definitely reviewing European work [1974a,
1976a, b, c, 1978a, 1980a, b, c]. There is clearly an engagement between what youre doing in Ireland and in
Africa and Latin America as well.
JV: The next turning point there comes in 85, when I
was asked to produce a review essay on anthropology
and Marxism [1985]. I remember talking to Shirley [Lindenbaum]7 in the corridor at a party and saying to her,
Do you realize what it will do for my reputation if I
7. Lindenbaum was editor of the American Ethnologist at the time.

do it? And this is when I got labeled as a Marxist

having always taught Weber and Durkheim and Pareto
and Marx, Id suddenly become a Marxist.
DN: You began publishing articles based on your research in Ireland in the early to mid-eighties, and in a
number of these works you critically address representations of religion as the primordial, polarizing force of
contemporary Ireland. Could you briefly explain your
views on the limitations of religious-based arguments
to explain the contemporary problems in Ireland?
JV: Yes. Allow me another five hours! Its not a general
statement but a statement about Northern Ireland,
where when you say religion you really mean rightwing Protestantism as expressed in the Orange Order
and when you say Catholicism you mean all who
rebel. And once you adopt that Establishment view of
a religious rightness (and right in both senses of the
term) and a Catholic revolutionary mob, the whole
thing becomes characterized in those terms. When you
look underneath, you go back to the 17th century to colonial domination and the penal code and a whole system of political, economic, and legal constraints on
Catholics [see 1990b]. One of my contributions, I think,
is the fact that Im working in a county which was dominated by an aristocracy which was Orange-oriented but
also anti-Whig, so Ive got this lovely contradiction.
When the famine comes along the British wont take full
responsibility for the relief, so the old establishment steps
in, and then the question of sectarian discrimination
comes in. Out of these developments emerges a nonsectarian, more class-based type of politics.
DN: Fascinating. You have a book in progress on the
culture and politics of the Irish Famine [n.d.a].
JV: Thats coming out this summer. When I say coming out, I mean out of my gut.
DN: What are you doing in the book?
JV: Im historicizing and contextualizing and politicizing the Great Irish Famine in a particular time and
place. I begin in the 19th century, when the British had
introduced an Irish Poor Law that was really intended
to cut away at Irish local government, which in Fermanagh was in the hands of a Protestant minority of landowners. Ostensibly, the Poor Law was meant to relieve
poverty, but the Fermanagh workhouse held only a few
hundred people and something like 8,000 needed assistance. It was very clearly said that the workhouse system wasnt intended to relieve the famine; it was simply a new form of taxation. At the height of the famine
the British withdrew completely from famine relief,
which meant that the whole cost had to be borne by the
local people. Private correspondence between leading
politicians shows that from the beginning their goal
was to create a transitional period between traditional and capitalist agricultureand theyre con-
stantly using the term capitalist agriculture. I provide

a very detailed account of what happened in Fermanagh
during the famine. I describe the county in terms of a
carrying trade during which even potatoes were being
exported (those that were grown on the estates). And I
have a final chapter, Aftermath, in which I cut right
to the bone of who profited. In this I draw on anthropological critiques of development and comparative
work on famines [see also 1995, 1997b,c].
DN: In 1986 you published System and Process in the
Annual Review of Anthropology. In this piece you distinguish two broad approaches to anthropology, systemsand process-oriented, and you suggest that they may be
understood as integral to the internal conceptual contradiction that lies at the heart of theory in general.
JV: Did I say that?
DN: Yes. You go on to say that a shift from system to
process, process to system, is likely to occur when the
second source of conceptualization in anthropology, its
empirical base, the fieldwork situation, changes. And
you suggest that the systems thinking of the 1950s and
the early sixties gave way to process thinking in the sixties and seventies, partly in response to postindependence changes in the Third World. Since the mid-seventies, global changes in the organization of power and
economy of the kind that Harvey [1989], Mandel [1975],
and others have talked about have resulted in significant shifts in the fieldwork situation and in the problems posed by many anthropologists. In light of your argument about the centrality of the fieldwork situation
in general and considering the global shifts that began in the early seventies, how would you characterize
the postcolonial movement in anthropology? Are we
witnessing a kind of a hyperprocessuality that has
emerged in relation to radical shifts in the empirical
base of our discipline?
JV: There we are. The answer is yes. First, I think I need
to specify the postcolonial fieldwork situation, because
really, the fieldwork situation is the fieldworkers situation. Having said that, its very hard to recognize a distinct Third World (or a North distinct from South).
Im not sure whether it was out of ignorance that we
didnt include global shifts and hyperprocessuality earlier, but certainly theyre much more dominant in our
thinking now. Consider the drugs and arms trades, particularly. Second is the issue of representations of the
postcolonial worldparticularly that of a floating diaspora. Im with Gayatri Spivak on this one [see Spivak
1993, 1994]. I think its a rather elitist view of the postcolonial world. By far the great mass of people are still
being pinned down and oppressed as theyve always
been, for example, in Timor. There is great value in the
notion that we should see where we ourselves are coming fromthat we should position ourselves within our
perception of the Third World and seek to get behind
and beyond it.

DN: In 198586, you were a fellow at the Institute for

Advanced Study at Princeton, and the following year
you were a fellow at the National Humanities Center
in North Carolina.
JV: I received both fellowships in the same year. NHC was
willing to wait the year. That was marvelous of them.
DN: Can you talk about your intellectual development
during this period? What were you reading? What were
you finding particularly influential?
JV: The Princeton year was great because Joan Scott was
there. I learned a lot from the gender and feminist historians. The year at NHC was equally valuable. I read
Jameson [1984] for the first time, I read literary criticism and everything that came out of the Duke English
Department, I read in what one would call cultural
studies. By then, anthropology had seen the text, and I
was reading a social history that saw the text as relevant. I was crossing every line possible in my reading,
which was great.
DN: Do you feel that your work shifted directly as a result of being exposed to this literature?
JV: No doubt about it. Stylistically, I loosened up in my
writing. I also got invited to write on law around this
time, which was nice, because June Starr asked me to
write about colonial law in Uganda [1989b].
DN: In a number of your published works you offer a
critique of cultural explanations and a focus on cultural
artifacts like texts to the extent that they are disembodied from the social and political contexts in which they
are embedded and from which they emerge. One can see
this very clearly in Anthropology and Politics, which
will certainly stand for decades as the seminal work on
the history not only of anthropology and politics but of
the field. In fact, it is already being used as such.
JV: So it seems. This, again, relates to my ignorance
about how to write books. Today I would retarget that
book completely.
DN: What would you do now?
JV: Id treat it more as an intellectual history of a discipline which happened to look at politics. The work originated in my having taught political anthropology for
several years. These rubrics are damning. Theyve
closed my mind. What you say about Anthropology and
Politics is very flattering, but I hope that its not the
case. Sally Falk Moores review of the book is excellent.
She brings out that it is an Africanists view, an English
persons view, a social anthropological view, and really
an LSE/Mancunian view.
DN: One of the very interesting things about the book
is that you use the insights and the methodologies that
anthropologists usually apply to other societies and

turn them back on the history of the discipline. And one
thing that emerges is really a quite fundamental challenge to the way in which the history of the field is usually understood. In the process, you recover hidden histories and alternative, possible trajectories for the
discipline that might have beenalmost were, were on
the verge of becomingbut ultimately never were. Your
position and your evidence seriously undermine intellectual histories of the discipline that focus on linear progression of ideas or theories. And of course, these are the most
popular, the ones that seem never to go away.
JV: This is, again, the System and Process argument.
DN: How did Anthropology and Politics come into
being?
JV: In terms of influences, if you go back to somebody
like [J. G. A.] Pocock [1975] in intellectual history, this
is what its all about, which is contextualizing. What I
try to do is point to a counterfactual history. Theres an
element of historical imagination in all of this, and I
find that excitingbut, essentially, the old idea of contextualizing, sociologizing. Ive always loved detective
novels, and it shows in African Elite. No one has ever
followed up on that list of people who are suspects to
see who gets to the top, and I must do an essay explaining why so-and-so got to the top.8 But in the same
way, the detective aspect of Anthropology and Politics
looks at footnotes, acknowledgments, reads the margin
area, and sees even more in them than in the explicit
text, which is a very self-conscious, finished production. There are little hints of this methodology, but it
hasnt been systematized within anthropologyand
this is exactly what people in literary criticism and philosophy are doing. So this purely came out of my reading at North Carolina.
DN: So youre doing the systematizing?
JV: [laughing] Sure. I didnt realize Id used that term.
Youre quite right.
DN: One of the interesting concepts that you developed
in Anthropology and Politics is the idea of the moving
frontier, which you had used in Teso as well but now
defined somewhat more broadly.
JV: Of course, it comes from F. G. Baileys [1957] reference to a moving frontier in colonial India. I then trace
this frontier historically across the United States, and
also the moving frontier of knowledge, which I link
with the capitalist, colonialist picture.
DN: In Anthropology and Politics you have three different meanings for the moving frontier. One is the expan8. The reference here is to the individuals listed on p. 13 of African
Elite, all of whom were competing for political influence during
Vincents fieldwork.

sive frontier of industrial capitalism. Another is what

you call an internal frontier as first electoral representation and then public education were introduced
into British and American society. And then theres a
third meaningthe frontiers of science.
JV: I am most interested in the second. British anthropology was first funded by the British Association for
the Advancement of Science, in the form of studies of
Scotland, Ireland, and Wales. You can see the link between capitalism and imperialism in the need to know
about these dangerous people who were being uprooted from their villages with industrial advance.
DN: One senses in your work a growing engagement
with questions of hegemony [see 1994a].
JV: But hegemony understood as subtlety and nuance.
If you grew up in a class society this is what you look
for. The Gramsci I really like is [his work on] South Italy [1971].
DN: Considering the importance of the broader political environment in your formulation and in light of the
position advanced by Jameson [1984], Harvey [1989],
and Mandel [1975] that capitalism has entered a new
phase of organization as of the early 1970s (which has
corresponding influence in the realm of culture and intellectual movements), how would you characterize your
position with respect to the debate within anthropology
concerning postmodernism as qualitative rupture?
JV: I always have this problem of looking at my bookshelf and seeing these 50 definitions of postmodernism
and asking what exactly anyone is really talking about.
But for me the question goes back to the old distinction
between colonialism and capitalismnot seeing them
as the same thing. And here postmodernism fits very
nicely in terms of intellectual and European history,
whereas what [Mandel and Harvey] are talking about
can be completely divorced from imperial history and
also from European roots. If only people would look for
a few more postmodernisms somewhere else. David,
you work a lot in this area of noncapitalist societies in a
capitalist world, and I feel that thats the emphasis that
allows you to penetrate this type of question. Im hopeless on abstractions; I really do not like abstractions. Its
a limitation, and Im sticking with it.
changing. So any association with place isnt at all automatic when you talk culture. I find social fields, arenas,
and activity fields to be more useful concepts.
DN: Practice theory has emerged in the past decade or
so as an important new theoretical approach.
JV: Yes. As you know, I tend to disagree with that chronology. I think its been around much longer than that. Bourdieu took a great deal from my dear old Mancunians.
DN: And the discussion in practice theory of structure
and agency?
JV: Its funny, because some Irish historians are still using
agency to mean the institution. I used the word at Dublin Castle9 and realized that what they meant by agency
was the relief agency. The use of agency has had unfortunate masking effects. Thats a problem. When you
talked activity or action or even behavior there
was substance attached to it, whereas now agency . . .
DN: Is an abstraction.
JV: Yes.
DN: Dual abstractions [structure and agency].
JV: Yes! I like Bourdieus work very much, but essentially the work which isnt in the practice-theory volume. Anything on Algeria per se really needs to be read
and taken aboard in postcolonial anthropology. And his
later work on Distinction [1984] I just love.
DN: In much of your published material you stress the
importance of carefully detailed fieldwork and a comprehensive review of the archive in the field. This approach to fieldwork has come under attack in many
quarters in recent years, especially by people who are
interested in what they call nontraditional research
problems. To what extent do you feel that fieldwork
methods and fieldwork as you have done it and appreciate it are ill-suited to what some people are regarding
now as the unique research problems and concerns of
the post-Fordist or postmodern era?
DN: In recent years there has been some very important

work done in our field about the relationship between
culture and place, focusing on the idea that what might
be called the postmodern condition has allowed anthropologists to recognize long-standing hidden assumptions regarding culture as tied to place. Do you see an
earlier genealogy within the discipline with which to
connect this more recent work?
JV: I wish youd stop that last sort of characterization

and just tell me precisely what you mean. Ive never
seen observation as simple. And I dont think fieldwork
is a method. Its a situation that one is in, and other people are in, and you are trying to understand the resolutions that they come to with respect to the human conditions they face. And I cant see how that can ever be
out-of-date. In my work on AIDS with George [Bond] in
Uganda [see Vincent and Bond 1988, 1991, 1997a, b;
Bond et al. 1997; Vincent, Bond, and Rijal 1991], I think
it was our training in old-fashioned, traditional fieldwork
JV: I thought it had long been generally understood that

culture was something that people construct and is
9. At the conference International Perspectives on the Great Irish

Famine, Dublin, May 910, 1997 (see Vincent 1997b). A revised
and expanded version of this paper was published as Vincent 1997c.

that allowed us to be attuned to things going on that we

otherwise could not have seen. I dont see how any anthropologist can not study interactions and interrelationships. Otherwise, youre not studying human beings.
DN: Lets talk about your AIDS project. In the late
eighties, you began working on a project focusing on
AIDS in Africa, and youve written several articles, coauthored with George Bond, based on your research
[Vincent and Bond 1991, 1997a, b], one of which was a
report for UNDP [Vincent, Bond, and Rijal 1991]. You
have also coedited a book on AIDS in Africa and the Caribbean with George Bond, John Kreniske, and Ida Susser [Bond et al. 1997]. Here we see a return to a theme
that you pursued early in your published work and problematized in Anthropology and Politicsthe relationship between applied and academic work. How did you
get involved in this project, and where is it going?
JV: The Columbia Presbyterian Hospital was organizing
AIDS relief in Uganda, and they wanted somebody to
write a critical report of their program. I was asked because I was from Uganda, as it were. I accepted because
I wanted to get back to Uganda. George was also invited. We hoped that one day there would be another
consultancy request in Zambia so I could go back there,
but it never came off.
DN: How are you approaching AIDS in this work?
JV: As we would approach anything else. The AIDS epidemic could as well be a messianic movement. It could
be a new form of currency. It has happened to hit a group
of people in a place at a time, and one wants to understand
its impact and the reaction to itunderstand its meaning
to the people. But here there is the added dimension of
having to understand the role of American intervention.
So its very much the actual understanding of the society
that has been hit by the HIV virus. We havent gone in
as medical anthropologists but as anthropologists into a
community that has been hit by crisis.
DN: Do you plan on continuing research in this area?
JV: Yes. George and I have got so many unpublished papers that we ought to continue, but hes terribly busy.
I think for both of us our careers are constantly overtaken by what else we have to do all the time. I have
several times put all of our AIDS articles, unpublished
and published, under chapters, one to ten, and an introduction could be written and we could get it out, but it
would be very out-of-date. And that means that we need
to go back.
DN: In one of your articles on AIDS, Living on the
Edge: Changing Social Structures in the Context of
AIDS [Vincent and Bond 1991], you emphasize the
need to recognize, as an historical and social condition,
the effects of war and its aftermath on a population and
region, even if it means setting aside for the short term
the concepts and the constructs of an earlier academic
era [see also Vincent n.d.c; Vincent and Bond n.d.]. Do

you remember which concepts and constructs you had
in mind here?
JV: I dont, offhand. Its one of my favorite points, and
its antistructuralist, or anti- the Africa that I was
taught, with its emphasis on kinship, community, social order, and stability. I see far more need for recognition of men in motion, social fields, power, violence,
and global relations. You think of labor migration. And
my example is usually from Fortes [1945], where you
suddenly learn that four-fifths of the young men of the
Tallensi village are not there. And actually, that was
something with the AIDS inquiry as well; we found that
one entire questionnaire was given out in one district
when most of the young men werent there. Obviously,
it has been the Uganda experience which has forced on
me the need to focus on war.10 Once you find something
in peoples lives as prominent as labor migration or war,
then you begin to critique the way a monograph was not
written, permitting a counterfactual intellectual history of our discipline.
DN: When I was reading several articles youve done recently on war, Hostilities on the Margin of Knowledge,
Northeast Uganda, 1991 [n.d.b], and War in Uganda,
North and South [1994b], my impression was that you
were beginning to conceptualize war as more centrally
and regularly involved in contemporary social processes
than you had in the past.
JV: Certainly.
DN: This awareness would also be suggested by a volume that youve coedited with George Bond on violence
in African states [Bond and Vincent n.d.], which I believe is forthcoming this year.
JV: Yes. Its interesting, because I saw violence differently
in Northern Ireland. In Northern Ireland you can see
fighting between interest groups and factions, neighborly killings, Seamus Heaney calls it, even if it is also
them versus the state, whereas with the civil war in
Teso its been the impact of war on people who werent
themselves fighting. I think that difference has hit home
[with me] a lot more. Its that impact of war that is capturing me rather more than the politics of war.
DN: In Hostilities on the Margin of Knowledge you
show how the experience of war can leave populations
with powerful lingering memories of abuse and atrocities. And, furthermore, that these memories can act as
a counter to state rhetoric that seeks to sanitize or rationalize the atrocities into a totalizing official history
that legitimizes state power. Is this an example?
JV: Yes. I think you need to speak of regime power
rather than state. The civil war in Teso is not denial of
10. See Vincent 1989c, d and 1990b on hostilities in Ireland and
1994b and n.d.c on war in Uganda.

state at all. At the time of the conflict human rights

commissioners were going around inquiring about the
past atrocities while electoral commissioners were
touring the countryside organizing the next election. So
it isnt the state. Its the regime that could be held responsible for allowing the abuses to occur. One of the
great fears in Uganda was that after the Commission of
Inquiries encouraged the population to itemize the
atrocities there would be no follow-up. It was catharsis
but nothing more. Being behind the scenes as I was
when I accompanied a Columbia University human
rights team to Kampala in 1991, I heard the human
rights commissioner saying, But I only have one typewriter! You get a sense of realism about whats going
on. Im not sure who is served by the inquiry.
DN: Lets talk about reviews. Youve done over 40 reviews, and over 5 review essays. Why so many?
JV: I think when youre busy teaching and giving most
of your energies to teaching and administration in a college, which one has to, its a way of having the world
reach out to you that allows you to reach out to the
world. And Ive reviewed some great books in my time.
I also love reading reviews. I like analyzing reviews. I
like writing an intellectual history using reviews, because I think many of peoples best ideas come in reviewsideas that they never have time to develop. If I
could get a job reviewing for the rest of my life, Id love
it. (current anthropology, please note . . .) Have you
seen the Geertz review I did [1996a]? Im trying now to
analyze the whole context of a writers work. One editor for whom I did this wrote back, particularly, to say
why he liked it [see also 1994c, d ].
DN: I have one closing question. What do you like best
of your own work?
JV: My teaching! Only in teaching do I present myself
as a whole person. But you meant to refer to my written
work, and I honestly dont think I can say. For one
thing, I forget half the things I write. Im not kidding
about that. If you mean what do I enjoy writing, the answer is nothing, because I find it very hard work. I prefer
giving talks to writing. I did find the encyclopedia articles [1996b] strangely challenging, simply because it
seemed such a ludicrous task to have to write 3,000
words on the subject of American anthropology or factions or the Ghost Dance. I can tell you paragraphs I
like. I like the opening of On Law and Hegemonic Moments [1994a]. I liked doing the review of [Geertzs]
After the Fact [1996a] because of what I become aware
of in writing itfor example, how very important the
Chicago Committee on New Nations was for Geertz
from a networking point of view and how he was never
alone in the field but was always with family, students,
or colleagues. That really struck me. It is so alien to my
experience of fieldwork. But I hadnt realized that until
I started writing the review. This is why, for me, reviewing is analysis. Ahead of His Time? [1988b] I
liked because I had a book to review and yet had space

to research the background. Also, it was the first time
I had tried to bring in the textual analysis that I was
learning at NHC. In System and Process I like the one
paragraph where I set up a binary opposition between
systems and process vocabularies. I like the kind of passion that lay behind Structuring of Ethnicity.
DN: Those are the questions that I wanted to ask you,
unless youd like to add anything . . .
JV: I dont think so. I think youve probed as far as I
could think. Its a funny exercise that were involved in,
I must say.
DN: How did it seem to you?
JV: Youve constructed a self that I hadnt recognized as
being as consistent as youve exposed, because my
whole sense is of very diverse sets of specialist reading
audiences. And yet, somehow, youve connected them
quite a bit. The word career throws me. I think trajectory is more appropriate to describe what I have
done. I dont know who launches it and where it ends,
but thats what it is. With your questions youre doing
what Im trying to do with my reviews. Youre trying to
do the analysis. This is what I like about it. I think this
is the only way we can have a history of anthropology.
References Cited
b a i l e y, f. g. 1957. Caste and the economic frontier: A village
in highland Orissa. Manchester: Manchester University Press.
b o n d, g e o r g e c. , j. k r e n i s k e, i. s u s s e r, a n d j. v i n c e n t. Editors. 1997. AIDS in Africa and the Caribbean: Case
studies of an epidemic. Boulder: Westview Press.
b o n d, g e o r g e c. , a n d j o a n v i n c e n t. n.d. Editors. Paths
of violence: Destruction and deconstruction in African states.
Newark: Gordon and Breach. In press.
b o u r d i e u, p i e r r e. 1984. Distinction: A social critique of the
judgement of taste. Cambridge: Harvard University Press.
e n g e l s, f. 1958. The condition of the working class in England in 1844. Translated and edited by W. O. Henderson and
W. H. Chaloner. Oxford: Basil Blackwell.
f o r t e s, m e y e r. 1945. The dynamics of clanship among the
Tallensi. London: Oxford University Press.
g r a m s c i, A. 1971. The prison notebooks. Edited and translated by Quintin Hoare and Geoffrey Nowell Smith. New
York: International.
h a r r i s, m a r v i n. 1968. The rise of anthropological theory.
New York: Crowell.
h a r v e y, d a v i d. 1989. The condition of postmodernity. Oxford: Basil Blackwell.
j a m e s o n, f r e d r i c. 1984. Postmodernism, or The cultural
logic of late capitalism. New Left Review 146:5392.
l e n i n, v. 1964 (1899). The development of capitalism in Russia. Moscow: Progress Publishers.
m a n d e l, e r n e s t. 1975. Late capitalism. London: New Left
Books.
p o c o c k, j. g. a. 1975. The Machiavellian moment: Florentine political thought and the Atlantic republican tradition.
Princeton: Princeton University Press.
s l i c h e r v a n b a t h, b. h. 1963. The agrarian history of Western Europe, a.d. 5001850. Translated by Olive Ordish. London: E. Arnold.
s p i v a k, g a y a t r i c h a k r a v o r t h y. 1993. Outside in the
teaching machine. New York: Routledge.

. 1994. How to teach a culturally different book, in Colonial discourse/postcolonial theory. Edited by Francis Barker,
Peter Hulme, and Margaret Iverson, pp. 12650. Manchester:
Manchester University Press.
t h o m p s o n, e. p. 1963. The making of the English working
class. London: V. Gollancz.
v i n c e n t, j o a n. 1964. The social basis of party conflict in Zanzibar, 19561963. M.A. thesis, Department of Political Science, University of Chicago, Chicago, Ill.
. 1968. Status and leadership in an African town. Ph.D.
diss., Columbia University, New York, N.Y.
. 1969. Anthropology and political development, in Politics and change in developing countries. Edited by Colin Leys,
pp. 3563. London: Cambridge University Press.
. 1970. Local cooperatives and parochial politics in
Uganda. Journal of Commonwealth Political Studies (1):317.
. 1971. African elite: The big men of a small town. New
York: Columbia University Press.
. 1974a. Review of: The Italians of Montreal, by J. Boissevain. American Anthropologist 76:12324.
. 1974b. The structuring of ethnicity. Human Organization
33:37579. (Reprinted in Teacher training manual on multi-ethnic social studies [New York: Box 75, Teachers College, 1978].)
. 1974c. Visibility and vulnerability: The politics of pacemakers. Africa 44:22234.
. 1976a. Review of: The land question and European society, by Frank E. Huggett (London: Thames and Hudson, 1975).
Sociology 3(7):123.
. 1976b. Review of: Peasants against politics, by Suzanne
Berger (Cambridge: Harvard University Press, 1972). American
Anthropologist 78:43940.
. 1976c. Review of: The land and people of nineteenthcentury Cork, by James S. Donnelly Jr. (London: Routledge
and Kegan Paul, 1975). Journal of Interdisciplinary History
7:33435.
. 1976d. Rural competition and the cooperative monopoly: A Ugandan case study, in Popular participation in social
change: Cooperatives, collectives, and nationalized industry.
Edited by J. Nash, J. Dandler, and N. Hopkins, pp. 7197. The
Hague: Mouton/Chicago: Aldine.
. 1977a. Colonial chiefs and the making of a class: A case
study from Teso, Uganda. Africa 47:14059.
. 1977b. Agrarian society as organized flow: Processes of
development past and present. Peasant Studies 4(2):5665.
. 1978a. Review of: Family life and illicit love in earlier
generations, by Peter Laslett (Cambridge: Cambridge University Press, 1977). Sociology 5(2):5859.
. 1978b. Teso in transformation: Colonial penetration in
Teso district and its contemporary significance, in Government and rural development in East Africa: Essays on political penetration. Edited by J. C. Coleman, M. Dornbos, and
L. Cliffe, pp. 5380. The Hague: Martinus Nijhof.
. 1980a. Review of: Farm life Northeast Scotland 1840
1914: The poor mans country, by Ian Carter (Edinburgh: Donald, 1979). Sociology 7(2):47.
. 1980b. Review of: Family and inheritance: Rural society
in Western Europe 12001800, by J. Goody, J. Thirsk, and E. P.
Thompson (Cambridge: Cambridge University Press, 1976).
American Ethnologist 7(3):150.
. 1980c. Review of: The Lanarkshire miners: A social history of their trade unions 17751974, by Alan B. Campbell (Edinburgh: Donald, 1979). Sociology 8(1):15.
. 1982. Teso in transformation: The political economy of
peasant and class in eastern Africa. Berkeley: University of
California Press.
. 1983. Political consciousness and struggle among an African peasantry, in Social anthropology of peasantry. Edited
by Joan P. Mencher, pp. 17790. Bombay: Somaiya
Publications/Atlantic Highlands, N.J.: Humanities Press.
. 1985. Anthropology and Marxism. American Ethnologist
12:13747.
. 1986. System and process, 19741985. Annual Review of
Anthropology 13:33139.
. 1988a. Equality and enterprise: The cooperative movement

in the development economy of Uganda, in Who shares? Cooperatives and rural development. Edited by D. W. Attwood and B. S.
Baviskar, pp. 188210. Delhi, New York: Oxford University Press.
. 1988b. Ahead of his time? Production and reception in the
work of Alexander Lesser. American Ethnologist 15:74351.
. 1989a. In the shadow of the armory: Capitalist production and the construction of ethnic culture. Paper presented at
the annual meeting of the Society for Economic Anthropology,
Mt. Pleasant, Mich.
. 1989b. Contours of change: Agrarian law in colonial
Uganda, in History and power in the study of law. Edited by J.
Starr and J. F. Collier, pp. 15367. Ithaca: Cornell University Press.
. 1989c. Local knowledge and global processes in two border villages: Political violence in County Fermanagh, in Ireland from below: Social change and local communities. Edited
by Chris Curtin and Thomas M. Wilson, pp. 92108. Galway:
Galway University Press.
. 1989d. Review of: Violence in Northern Ireland: Understanding Protestant perspectives, by John F. Galliher and Jerry L.
DeGregory (New York: Gill and Macmillan/Holmes and Meier,
1985). Society for the Anthropology of Europe Bulletin 3:1213.
. 1990a. Anthropology and politics: Visions, traditions,
and trends. Tucson: University of Arizona Press.
. 1990b. Irish border violence and the question of sovereignty, in The anthropology of war and peace. Edited by
Mario D. Zamora, Bjorn B. Erring, and Anthony L. Laruffa,
pp. 5384. Quezon City: Rex.
. 1991a. Fasting to death: The poetics and politics of the
Irish hunger strikes. Paper presented at the annual meeting of
the American Anthropological Association, Chicago, Ill.
. 1991b. Engaging historicism, in Recapturing anthropology. Edited by Richard Fox, pp. 4558. Santa Fe: School of
American Research Press.
. 1994a. On law and hegemonic moments: Looking behind law in early modern Uganda, in Contested states: Law,
hegemony, and resistance. Edited by Susan Hirsch and Mindy
Lazarus-Black. London: Routledge.
. 1994b. War in Uganda, North and South, in Studying
war: Anthropological perspectives. Edited by Stephan P. Reyna
and R. E. Downs. Langhorne, Pa.: Gordon and Breach.
. 1994c. Review of: The ethnographers magic and other essays in the history of anthropology, by G. Stocking. Journal of
the History of the Behavioral Sciences 30:18586.
. 1994d. Review of: Smallholders, householders, and farm
families and the ecology of intensive, sustainable agriculture,
by R. Netting. Man 29:77475.
. 1995. Conacre: A reevaluation of Irish custom, in Articulating hidden histories: Exploring the influence of Eric R.
Wolf. Edited by Jane Schneider and Rayna Rapp, pp. 8293.
Berkeley: University of California Press.
. 1996a. Review of: After the fact: Two countries, four decades, one anthropologist, by C. Geertz (Cambridge: Harvard
University Press, 1995). Studies in International Comparative
Development 31:15862.
. 1996b. American Anthropology, Factions, Genealogical Method, Ghost Dance, Law, Morgan, Plural Society, Political Anthropology, Political Economy, in Encyclopedia of social and cultural anthropology. Edited by Alan
Barnard and Jonathan Spencer, pp. 2528, 22223, 26263,
26566, 33033, 38182, 42526, 42834, 43437. London
and New York: Routledge.
. 1997a. Linen in the Irish Northwest: Unfinished business, or Lived experience in the 1830s, in The warp of Ulsters past. Edited by Marilyn Cohen, pp. 13957. New York:
St. Martins Press.
. 1997b. People, place, and time: An anthropological perspective on the Great Irish Famine. Paper presented at the conference International Perspectives on the Great Irish Famine,
Dublin, May 910.
. 1997c. Interpreting silences: An anthropological perspective on the Great Irish Famine. Eire-Ireland: An Interdisciplinary Journal of Irish Studies 32(23):2139.

. n.d.a. Seeds of revolution: The culture and politics of the

Irish famine. MS.
. n.d.b. Hostilities on the margin of knowledge: War and
peace in postcolonial Teso, in Paths of violence: Destruction
and deconstruction in African states. Edited by George C.
Bond and Joan Vincent. Newark: Gordon and Breach. In press.
. n.d.c. Guerrilla warfare in Northeast Uganda: Explorations in textured violence, in Paths of violence: Destruction
and deconstruction in African states. Edited by George C.
Bond and Joan Vincent. Newark: Gordon and Breach. In press.
v i n c e n t, j o a n, a n d g e o r g e c. b o n d. 1988. Field report
on AIDS in Uganda. HIV Center, New York State Psychiatric
Institute and Columbia University Presbyterian Hospital.
. 1991. Living on the edge: Structural adjustment in the
context of AIDS, in Changing Uganda: The dilemmas of
structural adjustment and revolutionary change. Edited by H.
Bernt-Hansen and Michael Twaddle, pp. 11329. London:
James Currey.
. 1997a. HIV/AIDS in Uganda: The first decade, in AIDS
in Africa and the Caribbean: Case studies of an epidemic. Edited by G. Bond, J. Kreniske, I. Susser, and J. Vincent. Boulder:
Westview Press.
. 1997b. Community-based organization in Uganda: A
youth initiative, in AIDS in Africa and the Caribbean: Case
studies of an epidemic. Edited by G. Bond, J. Kreniske, I. Susser, and J. Vincent. Boulder: Westview Press.
v i n c e n t, j o a n, g e o r g e c. b o n d, a n d a. r i j a l. 1991.
Issues report: HIV/AIDS in African development. Report to
the United Nations Development Program, January.
Painted Slabs from

Steenbokfontein Cave:
The Oldest Known Parietal Art
in Southern Africa1
a n t o ni e t a j e r a rd in o a n d n a t a l i e
s wa n e p o e l
Department of Archaeology, University of Cape
Town, Rondebosch 7701, South Africa/209 Maxwell
Hall, Anthropology Department, Syracuse University,
Syracuse, N.Y. 13244-1090, U.S.A. 12 x 98
In recent years there has been increasing interest in the
possibility of dating rock art. The age of cave paintings
1. 1999 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/99/4004-0009$1.00. Funding
for the excavation and dating of Steenbokfontein Cave material
was received from the Swan Fund, Wenner-Gren Foundation (Grant
Number 5699), and from a Council for Science Development grant
to the Spatial Archaeology Research Unit at the University of Cape
Town. The assistance of all these institutions is gratefully acknowledged. We are particularly grateful to Herman and Kitta
Burger of Steenbokfontein farm for their assistance, friendship, and
hospitality. We are also in debt to Royden Yates for his considerable support in the field and subsequent discussion of the significance of the painted slabs. Thanks are due to Suzanne Behan,
Megan Bryan, Grant Hall, Lee Manning, Jason Orton, and Andrew
Ward for their dedicated work during the recovery of the painted
slabs, as well as to John Parkington, Judy Sealy, and Janette Deacon
for their support. We thank Tony Manhire for helping during the
process of consulting the Spatial Archaeology Research Unit Data
Base and discussion of the painted slab findings and Anne Solomon
and Judy Sealy for commenting on the text.
and rock engravings has always been of concern to archaeologists but has rarely dominated research questions in rock art studies, largely because of our inability
to date the art accurately. Relative dates for painted images have recently been obtained through the analysis
of superimposed images, stylistic changes, and change
in imagery content (Chippindale and Tacon 1993,
Mguni 1997, Russel 1997, Thackeray 1983, Yates,
Manhire, and Parkington 1994). AMS radiocarbon dating has the potential of providing absolute dates for
painted images; because of the small samples required,
it allows carbon-rich pigment samples from rock paintings to be dated directly with minimal impact on the
images (Watchman 1993).
Over the past ten years, AMS dating of parietal art has
been improved and widely employed. As a result of this
effort, painted rock art now appears to be considerably
older than was previously thought. For instance, some
of the oldest rock art imagery in Australia has been
dated, though not without controversy, to between
about 9,000 and 30,000 years ago (Loy et al. 1990,
Watchman and Hatte 1996), whilst the oldest paintings
found in European Palaeolithic caves have been dated
to between approximately 12,000 and 28,000 years b.p.
(Clottes 1997, Clottes et al. 1997). By comparison, very
few AMS dates have been obtained from rock paintings
in southern Africa. A sample from Sonias Cave in the
western Cape (fig. 1) was dated to 500 140 b.p. (Van
der Merwe, Sealy, and Yates 1987), and more recently a
date of 330 b.p. was established for a sample obtained
from Esikolweni Shelter in Kwa-Zulu Natal (Mazel and
Watchman 1997). Additional samples obtained from
paintings in Kwa-Zulu Natal are being processed (Mazel, personal communication).
Dating pigments directly remains problematic, however, in that samples are very easily contaminated (e.g.,
Loy 1994, Nelson 1993, Rosenfeld and Smith 1997,
Watchman 1993) and the method is contingent on the
extraction of organic material, which may not survive.
In view of these problems and the lack of accuracy associated with relative dating methods, the most reliable
information about the age of parietal art may come from
sites where the remains of rock paintings have been incorporated into archaeological deposits and dating of
those deposits can provide minimum dates for the episodes of painting (Thackeray 1983).
In southern Africa, spalls containing traces of paint
have been found in deposits dating as far back as 10,000
years b.p. (Thackeray 1983), and mobiliary art is reported from Apollo 11 Cave, Boomplaas Cave, and Klasies River Mouth Cave 5. Such portable pieces have
come from deposits dating, respectively, to at least
26,000 b.p., 6,400 b.p., and about 3,900 b.p. (Binneman
and Hall 1993, Deacon, Deacon, and Brooker 1976,
Thackeray 1983, Wendt 1976). The most recent find of
mobiliary art in southern Africa was recovered from deposits of Collingham Shelter, and its minimum age has
been established at 1,800 years b.p. (Mazel 1993, 1994).
Clearly discernible images derived from collapsed
cave walls and buried in adjacent archaeological deposits are even more rarely discovered. In this article, we

F ig. 1. Location of sites mentioned in the text. A11,

Apollo 11; SC, Sonias Cave; BP, Boomplaas; KR5,
Klasies River Mouth 5; CM, Collingham Shelter; EW,
Esikolweni; SBF, Steenbokfontein Cave.
report on one such case encountered at Steenbokfontein
Cave in the western Cape. The evidence presented here
suggests that a portion of the cave wall containing
paintings and bare surface cracked and fell from the
cave wall onto the site floor. Subsequently, these rock
slabs became buried and incorporated in the stratigraphic sequence of Steenbokfontein Cave with successive occupations. According to the radiocarbon dates
obtained from the associated archaeological remains,
these painted slabs represent the oldest known examples of parietal art in southern Africa.
the site and the painted rock slabs
Steenbokfontein Cave is located on the coastal margin
of the western Cape (320942S, 1820E). It is situated
in a prominent sandstone outcrop facing west-northwest and overlooks reefs and beaches 1.8 km distant.
The rock paintings on the walls of Steenbokfontein
Cave consist mostly of handprints and human figures,
along with two images of fat-tailed sheep and very recent graffiti. With the exception of the sheep these images are common at other sites in the western Cape
(Manhire, Parkington, and Van Ruijssen 1983, Yates,
Manhire, and Parkington 1994).
Rock art surveys in the western Cape (Manhire 1981:
5758; Manhire, Parkington, and Van Ruijssen 1983)
classify Steenbokfontein Cave as a coastal site because of the substantial amounts of marine shell it contains. Distributional maps show that coastal sites are
located within about 5 km from the present coastline,
which became established by 6,500 years ago (Miller et
al. 1995). Sites located between 5 and about 25 km from
the shoreline are labeled interior Sandveld sites
(Manhire, Parkington, and Van Ruijssen 1983). Preservation of rock paintings at coastal sites is generally
poorer than at interior Sandveld sites (Manhire, personal communication).
Excavations at Steenbokfontein Cave commenced in
1992, and by 1997 7 m2 had been excavated to different
depths (fig. 2) without reaching bedrock. Seven major
stratigraphic layers have been recognized, and the total
volume of material removed amounts to approximately
8.0 m3. Fourteen radiocarbon dates are available from
this sequence, all of which fall within the past 8,500
years (table 1). However, judging by the size of the talus
and the estimated depth of archaeological residue, occupation of this large cave is expected to have spanned
many more thousands of years. In the context of previous archaeological work (Parkington et al. 1988), the
first set of observations from Steenbokfontein Cave has
already proved significant for the reconstruction of precolonial settlement and subsistence patterns in the
western Cape coast area (Jerardino 1996, Jerardino and
Yates 1996).
Immediately above squares I3, J3, and K3 and along the
rear wall of the cave, the morphology and contrast of patination on the rock surface indicate extensive and relatively recent exfoliation. A set of ten matching rock slabs
was recovered from layer 4a in squares I3, J3, and I4 (fig.
2) at depths of 0.2 to 0.5 m below the surface of the deposit
(figs. 3 and 4). It seems likely that these matching slabs
collapsed from the exfoliated area of the wall surface
above and were then covered with archaeological material resulting from subsequent occupations of the site.
During the excavations it became clear that the archaeological strata above the rock slabs were intact, and
no signs of a pit excavated in the past with the purpose
of placing the slabs within it were recognized. We conclude that the rock slabs were found in situ. The deposits in layer 4a are burnt throughout, with textures ranging from extremely consolidated to loose but crusty
sand and ash. Combustion of organic material was
nearly total, as all bone and shell are calcined and charcoal is very scarce (Jerardino and Yates 1996). The intensely burnt deposits of this layer seem to be the result
of fires originating outside the cave or slow-burning un-

F i g. 2. Site plan of Steenbokfontein Cave.

attended fires left behind after visits were completed
(Jerardino 1996:5354). On the basis of the colour of the
burnt bones in layer 4a, combustion of the organic remains must have reached temperatures as high as 600
800C (Brain and Sillen 1988). It is very likely that these
high temperatures exerted significant structural stress
on the nearby cave wall, leading to the formation of
cracks and the subsequent separation of large painted
fragments from the wall surface.
Three radiocarbon dates based on marine shell samples establish a minimum age of around 3,600 years ago
for the collapse of the rock slabs (fig. 4, table 1): 3,510
50 b.p. from a sample immediately above the collapsed

pile of slabs, 3,635 30 b.p. from a sample retrieved
from amongst three rock slabs (one of which was
painted) and small spalls, and 3,640 60 b.p. from a
sample recovered from one of the bottommost stratigraphic units of layer 4a. A charcoal date of 3,990 60
b.p. has also been obtained for the underlying layer 4b.
These dates are consistent with one another and the
stratigraphic sequence from which they were obtained
(fig. 3, table 1). Parietal rock art at Steenbokfontein
Cave thus dates to at least 3,600 years ago, the oldest
yet known for southern Africa.
table 1
Radiocarbon Dates for Steenbokfontein Cave
Layer
1
1
2
3a
3b
3b
4a
4a
4a
4b
5
5
5
5
Unit
Square
Material
TWIG
HAST
HK23
MT01
KTL2
OCLE
GRFR
AMRK
LFSX
SHN1
MS02
OMTS
SYBS
SYSH
K4
K3
K3
I3
K4
K4
J3
I3
J3
K3
J3
K3
J3
J3
Charcoal
Charcoal
Bone
Wood
Wood
Charcoal
Marine shell
Marine shell
Marine shell
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Date
2,200
2,200
2,360
2,510
2,490
2,690
3,510
3,635
3,640
3,990
4,620
6,070
7,950
8,370
60
50
45
50
50
60
50
30
60
60
70
80
80
80
13 C
Lab. No.
20.7
22.5
19.9
19.1
20.7
23.0
1.9
0.03
1.0
20.9
22.0
21.0
19.9
21.0
Pta-6136
Pta-6424
Pta-6498
Pta-7015
Pta-6505
Pta-6134
Pta-6794
Pta-7020
Pta-6805
Pta-6420
Pta-7323
Pta-6808
Pta-7326
Pta-7327
note: Marine shell dates are corrected for the apparent age of sea water (400 years). Names
of units are acronyms.

F ig. 3. Section drawing of excavations at Steenbokfontein Cave. Not all radiocarbon dates obtained have
been included in this figure. Excavations at the bottom of square J3 extend over 0.9 m2.
The surfaces of two of the collapsed slabs are clearly

painted, whereas the surfaces of the remaining eight
pieces show extensive fire damage, making it impossible to identify any painted images (fig. 5). The surviving
images on the slabs are, however, better preserved than
the ones currently observable on the cave walls.
Three human figures painted in red can be distinguished on one painted slab (fig. 6, left). The figure in
the middle appears to be clad in a white kaross, a type
of cloak. The other two figures are mostly represented
by pairs of legs and hips. On the second painted slab,
the imagery consists of the legs and hips of four human
figures painted in red (fig. 6, right). The second figure
from the right shows details in white pigment: lines alternating with rows of dots around knees and ankles.
The upper portion of these human figures cannot be
seen on the above matching rock slab (fig. 5), as damage
resulting from fires has obliterated them. The surfaces
of both of these slabs have an unevenly spread film of
red paint different from that used in the images. This
smeared paint is particularly dense around and below
the figure to the far left of the left slab. Stylistically, the
slab images clearly belong to the rest of the rock painting tradition of the western Cape.
discussion and conclusions
The painted slabs recovered from Steenbokfontein Cave
are interesting in three respects: with regard to the dating of southern African rock art, in relation to both their
preservation and that of rock art imagery generally at

coastal and interior Sandveld sites, and in terms of their
content and how it relates to the body of rock art found
in the Sandveld area in general. The last two aspects are
interrelated.
On the basis of dated contexts in which mobiliary art
has been found, archaeologists have known for some
time that rock art was being produced in southern Africa as early as 26,000 years ago. More recently, AMS
dates have helped to establish the age for parietal art
more accurately without having to assume contemporaneity with mobiliary art. These studies have yielded
dates around 300 and 500 years b.p. for parietal art. The
dates obtained for the painted slabs from Steenbokfontein Cave push the age of parietal art in southern Africa
back to at least 3,600 years b.p. As further AMS dates
are obtained and excavation programmes at painted
cave sites are continued in southern Africa, we can expect that the difference in age between parietal and mobiliary art may be narrowed.
The date of 3,600 b.p. represents the minimum age of
the wall paintings at Steenbokfontein Cave, as it indicates when the slabs fell off the wall as opposed to when
the imagery was painted. It is difficult to advance an estimated date for the episode of painting. However, the
imagery present on the slabs could not have been
painted very long before they collapsed from the cave
wall. The presence of white paint is an indication of
this, as this colour is well known to be fugitive within
the greater corpus of western Cape rock art and else-

F i g. 4. Photograph (left) and drawing (right) of J3/I3 section, including dates. The surface above the rocks is
not the original cave floor. The string appearing in the photograph was placed during excavation of square I3
in order to avoid a section collapse. The dark portion of the ranging rod represents 50 cm.
where (Pager 1971:322; Vinnicombe 1976:139, 164;
Yates, Golson, and Hall 1985:70; Yates, Manhire, and
Parkington 1994:38).
As described above, the rock paintings surviving in
the cave today are dominated by handprints, other abstract finger paintings (finger dots and grids), and a few
blurred human figures. This set of images bears little
resemblance to the imagery present in the buried rock
slabs, and its content appears less diverse when compared with the variety of motifs present in interior Sandveld sites (Manhire 1981). This has two important implications. First, the fact that handprints are not part of
the imagery present in the ca. 3,600-year-old painted
slabs but are the most frequent motif in the cave today
may support the suggestion that handprints are a late
addition to the rock painting tradition of the western
Cape (Yates, Manhire, and Parkington 1994). Secondly,
an impoverished rock art assemblage is not unique to
Steenbokfontein Cave but characteristic of coastal sites
in general compared with the interior Sandveld sites
(Manhire 1981). With the recovery of the painted slabs
there is reason to suspect that this contrast may have

been absent in the past or, if present, less marked. This
difference could well be the result of the harsher environmental conditions (a combination of wind, sun, and
salt-rich mist) prevailing year-round on the coast.
Differential preservation of rock art imagery as a result of environmental degradation could well be implicated as a causative factor in some of the perceived differences in the content of rock art imagery between the
coast and the interior Sandveld (Manhire, Parkington,
and Van Rijssen 1983). Whilst lines of 15 or more human figures are present in the parietal art of a number of
interior Sandveld sites (Manhire, Parkington, and Van
Rijssen 1983:31), only 4 of the 32 coastal sites where
rock art is present have lines of human figures, and 4
figures, barely distinguishable, is the maximum at any
such site (Spatial Archaeology Research Unit Data Base,
University of Cape Town). In contrast, a line of at least
7 human figures can be distinguished where imagery
has been preserved on the fallen rock slabs from Steenbokfontein Cave (fig. 6), suggesting that in the past long

F ig. 5. Visual reconstruction of matching slabs.

Shaded fragments indicate slabs with paintings
preserved on the surface.
lines of human figures may not have been so uncommon on the coast.
Kaross-clad figures are recorded at coastal sites, although none of them are represented with white paint.
Two sites within a range of 3.5 km show karosses in
yellow or red, and 6 sites about 4.5 km inland show
them faded away, as evidenced by the remaining and adjacent red paint that defines the rest of the human figures (so-called kaross-absent figures). Kaross-clad figures in general are, however, represented in at least 12
sites of the interior Sandveld (Spatial Archaeology Research Unit Data Base, University of Cape Town). Thus,
there seems to be a gradient of increasing number of
rock art sites containing kaross-clad figures as the distance from the coast increases. The existence of this
gradient and the presence of a kaross-clad figure on one
of the painted slabs seems to suggest that kaross figures
may well have been more common at coastal sites in
the past. We thus suspect that the reason kaross-clad
figures are so rare in coastal sites and more common in
the interior Sandveld today is the harsher environmental conditions prevalent on the coast and the resulting
deterioration of paintings and pigments.
Interest in the dating of rock art can be expected to
continue. The establishment of a chronological context
for rock art will allow us to link it with the historical
processes reconstructed as a result of excavations
(Campbell and Mardaga-Campbell 1993; Mazel 1993,
1996; Mazel and Watchman 1997; Russel 1997; Walker
1994; Watchman and Hatte 1996; Yates, Manhire, and
Parkington 1994). Our ongoing excavation programme
at Steenbokfontein Cave is aimed in that direction.
F ig. 6. Drawing of two of the matching slabs from Steenbokfontein Cave where paintings were preserved. All
figures were painted in red except that the second figure from the left has a white cloak and the ankles and
knees of the second figure from the far right have rows of white lines and dots. Wavy line along top margin of
right slab and island depicted on same slab show extent of visible encrusting precipitate. Dashed line shows
area where smeared red paint is most dense. Scale in centimetres.

References Cited
b i n n e m a n, j. , a n d s. h a l l. 1993. The context of four
painted stones from the south-eastern and eastern Cape. Southern African Field Archaeology 2:8995.
b r a i n, c. k. , a n d a. s i l l e n. 1988. Evidence from the
Swartkrans cave for the earliest use of fire. Nature 336:46466.
c a m p b e l l, j. b. , a n d m. m a r d a g a - c a m p b e l l. 1993.
From micro- to nanostratigraphy: Linking vertical and horizontal dating of archaeological deposits with the direct dating
of rock art at The Walkunders, Chillagoe (North Queensland,
Australia), in Time and space: Dating and spatial considerations in rock art research. Edited by Jack Steinbring, Alan
Watchman, Paul Faulstich, and Paul S. C. Tacon, pp. 5763.
Australian Rock Art Research Association Occasional Publication 8.
c h i p p e n d a l e, c. , a n d p. s. c. t a c o n. 1993. Two old
painted panels from Kakadu: Variation and sequence in Arnhem Land rock art, in Time and space: Dating and spatial
considerations in rock art research. Edited by Jack Steinbring,
Alan Watchman, Paul Faulstich, and Paul S. C. Tacon, pp. 32
56. Australian Rock Art Research Association Occasional Publication 8.
c l o t t e s, j. 1997. New laboratory techniques and their impact in Palaeolithic cave art, in Beyond art: Pleistocene image
and symbol. Edited by Margaret W. Conkey, Olga Soffer, Deborah Stratmann, and Nina G. Jablonski, pp. 3752. San Francisco: Allen Press.
c l o t t e s, j. , j. c o u r t i n, j. c o l l i n a - g i r a r d, m. a r n o l d, a n d h. v a l l a d a s. 1997. News from Cosquer Cave:
Climatic studies, recording, sampling, dates. Antiquity 71:
32126.
d e a c o n, h. j. , j. d e a c o n, a n d m. b r o o k e r. 1976. Four
painted stones from Boomplaas Cave, Oudtshoorn district.
South African Archaeological Bulletin 31:14145.
j e r a r d i n o, a. 1996. Changing social landscapes of the western Cape coast of southern Africa over the last 4500 years.
Ph.D. diss., University of Cape Town, Cape Town, South Africa.
j e r a r d i n o, a. , a n d r. y a t e s. 1996. Preliminary results
from excavations at Steenbokfontein Cave: Implications for
past and future research. South African Archaeological Bulletin 51:716.
l o y, t. h. 1994. Direct dating of rock art at Laurie Creek (NT),
Australia: A reply to Nelson. Antiquity 68:14748.
l o y, t. h. , r. j o n e s, d. n e l s o n, b. m e e h a n, j. v o g e l,
j. s a n t h o n, a n d r. c o s g r a v e. 1990. Accelerator Rc dating of human blood proteins in pigments from late Pleistocene
art sites in Australia. Antiquity 64:11016.
m a n h i r e, a. 1981. The rock art of the Sandveld. B.Sc. Hons.
diss., University of Cape Town, Cape Town, South Africa.
m a n h i r e, a. , j. p a r k i n g t o n, a n d b. v a n r i j s s e n.
1983. A distributional approach to the interpretation of rock
art in the south-western Cape. South African Archaeological
Society, Goodwin Series 4:2933.
m a z e l, a. 1993. Rock art and Natal Drakensberg huntergatherer history: A reply to Dowson. Antiquity 67:88992.
. 1994. Dating the Collingham Shelter rock paintings. Pictogram 6:3335.
. 1996. Maqonqo Shelter: The excavation of Holocene deposits in the eastern Biggarsberg, Thukela Basin, South Africa.
Natal Museum Journal of Humanities 8:139.
m a z e l, a. d. , a n d a. l. w a t c h m a n. 1997. Accelerator radiocarbon dating of Natal Drakensberg paintings: Results and
implications. Antiquity 71:44549.
m g u n i, s. 1997. The evaluation of the superpositioning sequence of painted images to infer relative chronology: Diepkloof Kraal Shelter as a case study. B.A. Hons. diss., University
of Cape Town, Cape Town, South Africa.
m i l l e r, d. e. , r. j. y a t e s, a. j e r a r d i n o, a n d j. e.
p a r k i n g t o n. 1995. Late Holocene coastal change in the
southwestern Cape, South Africa. Quaternary International
29/30:310.
n e l s o n, d. 1993. Second thoughts on a rock art date. Antiquity 67:89395.
p a g e r, h. 1971. Ndedema. Graz: Akademische Druck- und Verlagsanstalt.

p a r k i n g t o n, j. e. , c. p o g g e n p o e l, w. b u c h a n a n,
t. r o b e y, a. m a n h i r e, a n d j. s e a l y. 1988. Holocene
coastal settlement patterns in the western Cape, in The archaeology of prehistoric coastlines. Edited by Geoff Bailey and
John Parkington, pp. 2241. Cambridge: Cambridge University
Press.
r o s e n f e l d, a. , a n d c. s m i t h. 1997. Recent developments
in radiocarbon and stylistic methods of dating rock art. Antiquity 71:40511.
r u s s e l, t. 1997. Sequencing rock paintings: The application of
the Harris matrix to rock art at Main Caves North, Giants
Castle Game Reserve, Kwa-Zulu Natal. B.Sc. Hons. diss., University of Cape Town, Cape Town, South Africa.
t h a c k e r a y, a. i. 1983. Dating the rock art of southern Africa. South African Archaeological Society, Goodwin Series
4:2126.
v a n d e r m e r w e, n. , j. s e a l y, a n d r. y a t e s. 1987. First
accelerator carbon-14 date for pigment from a rock painting.
South African Journal of Science 83:5657.
v i n n i c o m b e, p. 1976. People of the eland. Pietermaritzburg:
University of Natal Press.
w a l k e r, n. 1994. Painting and ceremonial activity in the
Later Stone Age of the Matopos, Zimbabwe, in Contested images: Diversity in southern African rock art research. Edited
by Thomas Dowson and David Lewis-Williams, pp. 11930. Johannesburg: Witwatersrand University Press.
w a t c h m a n, a. 1993. Perspectives and potentials for absolute
dating prehistoric rock paintings. Antiquity 67:5865.
w a t c h m a n, a. , a n d e. h a t t e. 1996. A nano-approach to
the study of rock art: The Walkunder, Chillagoe, North
Queensland, Australia. Rock Art Research 13:8592.
w e n d t, w. 1976. Art mobilier from the Apollo 11 Cave,
South West Africa: Africas oldest dated works of art. South African Archaeological Bulletin 31:511.
y a t e s, r. , j. g o l s o n, a n d m. h a l l. 1985. Trance performance: The rock art of Boontjieskloof and Sevilla. South African Archaeological Bulletin 40:7080.
y a t e s, r. , a. m a n h i r e, a n d j. p a r k i n g t o n. 1994.
Rock paintings and history in the south-western Cape, in
Contested images: Diversity in southern African rock art research. Edited by Thomas Dowson and David Lewis-Williams,
pp. 2960. Johannesburg: Witwatersrand University Press.
The Accuracy of AttractiveBody-Size Judgment1

a l e x a n d ra a. b r e w is
Department of Anthropology, University of Georgia,
Athens, Ga. 30602-1619, U.S.A. (abrewis@arches.
uga.edu). 12 x 98
Young women in contemporary industrialized societies
often or even characteristically engage in potentially
dangerous behaviors designed to attain and maintain
(extreme) slimness, including restrained eating and re1. 1999 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/99/4004-0006$1.00. The role
of Steven T. McGarvey and the students who have assisted with
data collection and entry on the Samoan body-image project over
the past several years is gratefully acknowledged. Joan Lawrence
prepared the illustrations. I thank John S. Allen and Ben Blount for
feedback and helpful conversations.

fusal to eat, induced vomiting, overexercising, and using drugs for appetite suppression (Walsh and Devlin
1998). Many studies have shown that idealization of
slim bodies and motivation to achieve them is associated with low self-esteem and distorted perceptions of
individuals own body images. The conventional wisdom is that these body-image distortions are encouraged or perpetuated by thinness-depicting and thinnesspromoting (and fat-stigmatizing) media (e.g., Harrison
and Cantor 1997, Monteath and McCabe 1997, Raphael
and Lacey 1992, Stephens, Hill, and Hanson 1994, Shaw
and Waller 1995). Certainly, individual-level studies of
women in industrialized societies have demonstrated
that greater exposure to thinness-depicting media is associated with greater body-image distortion (Harrison
and Cantor 1997, Tiggemann and Pickering 1996, Stice
et al. 1994).
This pattern of body-image distortion is considerably
more pronounced and more common in women than in
men, to the point that it is considered a characteristically female phenomenon.2 Young women are inaccurate in their estimates of their current size, on average
imagining themselves to be larger than they are and
more distant from an identified ideal.3 Men are more accurate in their own body-size estimates. Women are
also less accurate than men in predicting the body sizes
found most attractive by their opposite-sex peers. This
pattern was first observed in publication in 1985, when
Fallon and Rozin demonstrated that American undergraduate women could not accurately predict the body
shape that their male peers found most attractive, instead selecting an exaggeratedly slimmer figure, thereby
decreasing satisfaction with their own bodies. In contrast, men on average selected a body image that corresponded more closely to womens ideals and an ideal
size that reinforced their own body self-satisfaction (Fallon and Rozin 1985). This pattern has since
been replicated with further samples of undergraduate
women, parents of students, partners, adolescents,
and older women4 (Cohn and Adler 1992, Cohn et al.
1987, Furnham and Radley 1989, Huon, Morris, and
Brown 1990, Lamb et al. 1993, Rozin and Fallon 1988;
see also Dwyer et al. 1964; cf. Furnham, Hester, and
Weir 1990).
This widely cited finding has been used to demonstrate two distinct and contrary propositions about the
causes and context of the body-image distortion. First,
2. In the United States, women are exposed to and internalize the
cultural value of slimness more than men (Garner et al. 1980) and
find cultural ideals of beauty more crucial in developing and maintaining self-esteem. They are more motivated than men to conform
to body-size ideals and less satisfied with their bodies, weigh themselves more, perceive weight gain and largeness more negatively,
and experience higher levels of low body self-esteem and exercise
and eating disorders (Dwyer et al. 1969, Parker et al. 1995, Polivy,
Garner, and Garfinkel 1986).
3. The vast majority of this research has been conducted with U.S.
undergraduates, and therefore the conclusions are generally relevant only to populations of young, predominantly white, educated
middle-class women.
4. Body ideals (and body sizes) become larger with age, but the distortion persists (Stevens and Tiggemann 1998).
it is cited as evidence of the way in which popular media misinform women about their bodies; media images
of ultraslim women are internalized by women and result in distorted self-images and conceptions of others
judgments of their bodies5 (e.g., Martin and Gentry
1997, Ogden and Mundray 1996, Shaw and Waller
1995). Second, in the context of evolutionary approaches to human behavior, the pattern is interpreted
to represent evolved sex differences in reproductive
strategy. For example, it has been used as evidence that
men and women have different evolved notions of the
ideal female figure (Singh 1993), that it is each sexs
own evolved preferences rather than the preferences of
the other sex that is guiding mating strategy (Ridley
1993:302), and that men have an evolved propensity to
desire curvaceous women (Buss 1994:56).
These contrary propositions have yet to be tested adequately against data. The implication of the first proposition is that groups exposed to increased media distortion of female figures should demonstrate greater
distortion of womens estimates of male preferences.
The implication of the evolutionary proposition is that
the relative female propensity for distorting male preferences should be consistent across socioecological contexts whatever the local notion of ideal body size.6 To
date, only one set of data from a non-Western setting
has been published, so it is difficult to discern whether
the pattern of female misjudgment shows significant
socioecological variation or is consistent across populations. A study of Arab women shows that they accurately predict the (slim) figure size that male peers judge
the most attractive, even though the women prefer on
average to be smaller than they consider themselves to
be (Ford, Dolan, and Evans 1990).7 There are no examples from traditionally fat-positive societies to compare
with the slim-valuing Arab case, and there are no studies of the relative accuracy of male versus female perceptions in nonindustrial settings.
Here I report the results of a study of the accuracy of
attractive-body-size judgment in Samoans in two countries. The Samoan case allows comparison of genetically similar Samoans living in distinct environments
(e.g., Baker, Hanna, and Baker 1986), one richer in media
exposure than the other. It provides a counterpoint to
the U.S. and Arab data because Samoan society has traditionally placed high social value on big bodies. Samoan populations have some of the largest average body
sizes yet recorded for human groups (McGarvey 1994).
5. Fallon and Rozin (1985) also initially suggested that women

could actually be accurately judging male preferences but judging
the preferences of their ideal men rather than their peers. This was
not supported by subsequent tests (Huon, Morris, and Brown 1990).
6. Ideal or attractive body build is one of the most highly ecologically and culturally varied aspects of female attractiveness (Brown
and Konner 1987, Buss 1994, Sobal and Stunkard 1989). Though
highly variable between groups, cultural values about attractive
bodies tend to be strongly, consistently, and widely held within
them.
7. A Nigerian study seems to suggest that women accurately judged
body sizes preferred by their spouses but unfortunately does not
describe the relevant data (Akande 1993).

F i g. 1. Portions of womens and mens body-image rating scales.

design and data collection
The data presented here are derived from samples of Samoans living in two ecologically distinct environmentsthe more traditional and rural island setting of
Samoa (formerly Western Samoa) and the migrant setting of urban Auckland, New Zealand. There are close
genetic, kin, and social relationships between these two
populations. Samoans in Auckland constitute a minority group within a large, urban, media-rich environment
thatlike other industrialized societiesgenerally idealizes and socially and economically rewards slim bodies. Exposure to thinness-depicting media is considerably greater in Auckland than in Samoa. Samoan
women express slimmer body ideals in Auckland than
Samoa, despite having larger bodies (Brewis et al. 1998).
Measures of perception of attractive bodies were made
for 84 women and 77 men in both rural and periurban
villages in Samoa and 41 female and 24 male Samoans
living in urban Auckland. All participants were between 25 and 55 years of age, and the two field sites
were similar in average age.8
To measure the accuracy of judgments of attractive
bodies, this study used separate drawn scales of male
and female body outlines. The instrument depicted a series of ten figures increasing in size on a continuous
scale (see fig. 1). The tool was adapted from that used
by Fallon and Rozin (1985) and Ford, Dolan, and Evans
(1990), but the figures were made to represent recognizably Samoan body proportions and hairstyles.9 On the
8. The data reported here are part of a larger study on the ecology
of body image in Samoans in three countries, and the sampling
strategy is described by Brewis et al. (1998).
9. Using this tool, sex differences in accuracy of judgment were evident in samples of anthropology undergraduates at the University
of Georgia and the University of Auckland collected in 199798
scale depicting their own sex, participants were asked

to locate their own current size, the size they would
most like to be (ideal), the average size for their agegroup, the size most attractive to the opposite sex, and
the upper and lower limits of acceptable body size for
opposite-sex peers seeking mates. On the scale depicting the opposite sex, participants identified the average size for their age-group, the size they found most
attractive, and the upper and lower limits of body size
acceptable to them in a mate. One-tailed and matchedpair t-tests were used to compare means by sex and
fieldsite, and sex-fieldsite interactions were assessed by
analysis of variance. Testing for age association was
done on the basis of age-groups 2539 and 4055 years.
results
The average values of body perceptions reported by each
sex at each field site are shown in figures 2 and 3. Samoans of both sexes in both countries identified on average
body ideals significantly smaller than the sizes they perceived themselves to be. However, in all cases Samoans
accurately predicted the size the opposite sex found attractive, both in general and in their own age-groups (all
p 0.05). Further, men and women were very accurate
in predicting average size of bodies in their age-groups
as perceived by their own and the other sex. The only
notable sex difference emerged in views of the acceptability of larger bodies in Auckland, where body size is
also on average significantly larger than in Samoa. In an
estimate of the largest acceptable body size, women
(N 258). Women predicted that men would prefer a slimmer figure than they would ( p 0.05), while men did not make the corresponding prediction.

F i g. 2. Average body-image responses for womens bodies in Auckland and Samoa.

showed a tendency to tolerate a larger mens body than
men thought they did, whereas men set a lower limit
for tolerable body size than women anticipated they
would.
A misjudgment measure was calculated for each individual by subtracting that individuals own estimate of
the size preferred by the other sex from the mean value
of most attractive body size reported by that other sex.
Both men and women in Auckland had higher average
misjudgment values (11.9 downward for women, 13.8
upward for men) than in Samoa (1.8 downward for men,
2.5 downward for women), but this difference was not
statistically significant.
conclusion
Samoan men and women identify an ideal body that is
significantly smaller than their perceived current size.
However, Samoans both in more traditional, rural
Samoa and in urban, media-exposed Auckland accurately judge both male and female body-size preferences.
This observation is in contrast to the pattern observed
in U.S. samples and in concord with the Arab case.
The U.S., Samoan, and Arab cases taken together
demonstrate no monolithic pattern of womens consistently misjudging male preferences or of mens consistently being more accurate than women. Rather, a more
complex socioecology of sex and population differences
in attractive-body misjudgment is suggested. Some accounting for this cross-population variation can be suggested with reference to the proximal contexts of Samoan social life and marriage and Samoan views of the
value of bodies. Body form is less important as a vehicle
for social, economic, conjugal, or reproductive success
for Samoan women than it is, for example, for female
undergraduates in the United States. Further, in relatively collectivistic Samoan society, where status and
family are vital social themes and therefore important
considerations in marriage decisions, attractiveness has
less salience in these decisions. Further, almost all Samoan women marry, and they expect to marry regardless of their physical attributes and to marry well because of family advantages and opportunities rather
than because of those attributes.
Samoan women and men in more media-exposed
Auckland and in less media-exposed Samoa are equally
accurate in their assessments of their peers attractivebody preferences. This suggests that the media model is

F i g. 3. Average body-image responses for mens bodies in Auckland and Samoa.
insufficient to explain population variation in patterning in womens ability to estimate mens ideals of
womens bodies. While media imagery and reinforcement of popular cultural ideas must play a role in promoting and exaggerating womens misconceptions
about male preferences, this role may be a product or a
co-result rather than a single necessary causative factor
in the patterning of body-attractiveness misjudgment in
samples of Western women.
References Cited
a k a n d e, a. 1993. Sex differences in preferences for ideal female body shape. Health Care for Women International 14:
24959.
b a k e r, p., j. h a n n a, a n d t. b a k e r. Editors. 1986. The
changing Samoans: Behavior and health in transition. New
York: Oxford University Press.
b r e w i s, a., s. m c g a r v e y, j. j o n e s, a n d b. s w i n b u r n.
1998. Perceptions of body size in Pacific Islanders. International Journal of Obesity 22:18589.
b r o w n, p., a n d m. k o n n e r. 1987. An anthropological perspective on obesity. Annals of the New York Academy of Sciences 499:6672.
b u s s, d. 1994. The evolution of human desire: Strategies of human mating. New York: Basic Books.
c o h n, l., a n d n. a d l e r. 1992. Male and female perceptions
of ideal body shapes: Distorted views among Caucasian college
students. Psychology of Women Quarterly 16:6979.
c o h n, l., n. a d l e r, c. i r w i n, s. m i l l s t e i n, s. k e g e l e s, a n d g. s t o n e. 1987. Body-figure preferences in male
and female adolescents. Journal of Abnormal Psychology 96:
27679.
d w y e r, j., j. f e l d m a n, c. s e l t z e r, a n d j. m a y e r.
1969. Adolescent attitudes toward weight and appearance. Journal of Nutrition Education 1:1519.
f a l l o n, a., a n d p. r o z i n. 1985. Sex differences in perceptions of desirable body shape. Journal of Abnormal Psychology
94:1025.
f o r d, k., b. d o l a n, a n d c. e v a n s. 1990. Cultural factors
in eating disorders: A study of body shape preferences of Arab
students. Journal of Psychosomatic Research 34:5017.
f u r n h a m, a., c. h e s t e r, a n d c. w e i r. 1990. Sex differences in preferences for specific female body shapes. Sex Roles
22:74354.
f u r n h a m, a., a n d s. r a d l e y. 1989. Sex differences in perception of male and female body shapes. Personality and Individual Differences 10:65362.
h a r r i s o n, k., a n d j. c a n t o r. 1997. The relationship between media consumption and eating disorders. Journal of
Communication 47:4067.

h u o n, g., s. m o r r i s, a n d l. b r o w n. 1990. Differences between male and female preferences for body size. Australian
Psychologist 25:31417.
l a m b, c., l. j a c k s o n, p. c a s s i d a y, a n d d. p r i e s t.
1993. Body figure preference of men and women: A comparison of two generations. Sex Roles 28:34558.
m c g a r v e y, s. 1994. The thrifty-gene concept and adiposity
studies in biological anthropology. Journal of the Polynesian
Society 103:2942.
m a r t i n, m., a n d j. g e n t r y. 1997. Stuck in the model trap:
The effects of beautiful models in ads on female pre-adolescents and adolescents. Journal of Advertising 26:1933.
m o n t e a t h, s., a n d m. m c c a b e. 1997. The influence of societal factors on female body image. Journal of Social Psychology 137:70827.
o g d e n, j., a n d k. m u n d r a y. 1996. The effect of media on
body satisfaction: The role of gender and size. European Eating
Disorders Review 4:17182.
p a r k e r, s., m. n i c h t e r, m. n i c h t e r, n. v u k o v i c, c.
s i m s, a n d c. r i t e n b a u g h. 1995. Body image and weight
concerns among African American and white adolescent females: Differences that make a difference. Human Organization 54:10314.
p o l i v y, j., d. g a r n e r, a n d p. g a r f i n k e l. 1986. Causes
and consequences of the current preference for thin female physiques, in Physical appearance, stigma, and social behavior.
Edited by P. Herman, M. Zanna, and E. T. Higgins. Hillsdale,
N.J.: Lawrence Erlbaum Associates.
r a p h a e l, f., a n d j. l a c e y. 1992. Sociocultural aspects of
eating disorders. Annals of Medicine 24:29396.
r i d l e y, m. 1993. The red queen: Sex and the evolution of human nature. New York: Macmillan.
r o z i n, p., a n d a. f a l l o n. 1988. Body image, attitudes to
weight, and misperceptions of figure judgments of the opposite
sex: A comparison of men and women in two generations. Journal of Abnormal Psychology 97:34245.
s h a w, j., a n d g. w a l l e r. 1995. The medias impact on
body image. Eating Disorders 3:11523.
s i n g h, d. 1993. Body shape and womens attractiveness: The
critical role of waist-to-hip ratio. Human Nature 4:297321.
s o b a l, j., a n d a. s t u n k a r d. 1989. Socioeconomic status
and obesity: A review of the literature. Psychological Bulletin
105:26075.
s t e p h e n s, d., r. h i l l, a n d c. h a n s o n. 1994. The beauty
myth and female consumers. Journal of Consumer Affairs 28:
13753.
s t e v e n s, c., a n d m. t i g g e m a n n. 1988. Womens body
preferences across the lifespan. Journal of Genetic Psychology
159:94102.
s t i c e e., e. s c h u p a k, h. s h a w, a n d r. s t e i n. 1994. Relation of media exposure to eating disorder symptomology: An
examination of mediating mechanisms. Journal of Abnormal
Psychology 103:83640.
t i g g e m a n n, m., a n d a. p i c k e r i n g. 1996. Role of television in adolescent womens body dissatisfication and drive for
thinness. International Journal of Eating Disorders 20:199
203.
w a l s h, b., a n d m. d e v l i n. 1998. Eating disorders: Progress
and problems. Science 280:138790.
First Estimates of Heritability in

the Age of Menopause1
j oc e l y n s c o tt p e c c e i
Department of Anthropology, University of
California, Los Angeles, 341 Haines Hall, 405 Hilgard
Ave., Los Angeles, Calif. 90095-1553, U.S.A.
(jpeccei@ucla.edu). 23 xi 98
Two independent and complementary studies have recently produced the first estimates of heritability (h2) in
the age of menopause (Peccei 1998, Snieder, MacGregor,
and Spector 1998). Snieder, MacGregor, and Spector
(1998), using recall data from a large British twin sample, have suggested that 63% of the variation is genetic,
and the study reported here, using prospectively collected data from American mothers and daughters participating in the Tremin Trust Menstrual Reproductive
History (TTMRH) project, indicates that 4050% of
variation in the age of menopause may be genetic. In
the absence of conclusive evidence of an upward shift
in the mean age of menopause (McKinlay, Bifano, and
McKinlay 1985; Flint 1978; Gray 1976; Amundsen and
Diers 1973, 1970; Post 1971), these heritability estimates suggest that the mean age is maintained at its
contemporary value by some degree of stabilizing selection and therefore that there must be some cost to prolonged reproduction in human females even now. In
particular, these heritability estimates are interesting
because they change the terms of the menopause debate. Whereas previous work has focused on the reproductive ecology of hunter-gatherers in an effort to understand the environment of evolutionary adaptedness
and the origin of menopause, the discovery of such substantial positive heritabilities in contemporary populations strongly suggests that there must have been a cost
to prolonged reproduction even in agricultural societies
over the past 10,000 years. These heritability estimates
may also have clinical value. Here I describe my statistical analyses, discuss the reliability of my results, and
explore the evolutionary implications of these findings.
Although there is a strong central tendency in the age
of menopause, with medians clustering around 50 years
in developed countries, considerable variation exists
both within and between populations (see Wood 1994).
Medians range from 43 years in Central Africa to 51.4
years among Caucasian Americans. Several studies
have examined the contribution of various environmen1. 1999 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/99/4004-0007$1.00. I thank
R. Boyd for guidance during this project. I also thank J. K. H. Lu,
T. Plummer, D. Read, R. Gould, and R. D. Peccei for helpful discussions and suggestions and an anonymous referee for useful comments. I thank A. Voda for permitting me access to the Tremin
Trust Menstrual Reproductive History dataset and B. LaSalle and
K. Smith for their assistance. Finally, I gratefully acknowledge the
financial support of the American Federation for Aging Research.

tal and life-history factors to variation in the age of

menopause (Treloar 1974, Torgerson et al. 1994, Parazzini, Negri, and LaVecchia 1992, Whelan et al. 1990,
Brambilla and McKinlay 1989, Stanford et al. 1987,
McKinlay, Bifano, and McKinlay 1985). As with menarche, nutritional status has been a prime suspect. However, the large-scale multivariate studies which have attempted to control for confounding variables such as
socioeconomic status, ethnicity, marital status, and
parity have failed to show a nutritional effect (Wood
1994). The only well-established risk factors are cigarette smoking, which lowers the median age of menopause by approximately 1.5 years (Torgerson et al. 1994,
McKinlay, Bifano, and McKinlay 1985), and a family
history of premature ovarian failure (40 years)
(Cramer and Xu 1996, Cramer, Xu, and Harlow 1995,
Mattison et al. 1984, Coulam, Stringfellow, and Hoefnagel 1983, Smith, Fraser, and Noel 1979).
As a life-history trait long considered virtually unique
to human females (Pavelka and Fedigan 1991, Caro et
al. 1995), menopause constitutes an evolutionary puzzle posing two distinct questions: how it originated and
what is maintaining it. Although a gradual decline in
age-specific fertility is associated with the generalized
process of aging in almost all mammals, including all
nonhuman primates, only human females and one species of toothed whales experience a total cessation of
reproductive capacity well before the end of their maximum or average life span (Caro et al. 1995, Marsh and
Kasuya 1984). Evolutionary theory predicts that menopause originated because women with premature reproductive senescence enjoyed an inclusive-fitness advantage from increasing the fitness of existing progeny. Yet,
despite much theoretical and empirical investigation,
the inclusive-fitness value of menopause remains difficult to establish in contemporary populations (Hill and
Hurtado 1991, Rogers 1993, Hawkes, OConnell, and
Blurton Jones 1997). This suggests that whatever is responsible for the origin of menopause may have little to
do with the maintenance of the trait and underlines the
necessity of separating the origin and maintenance
questions (Peccei 1995).
Furthermore, the first step to understanding the
maintenance of menopause is not to try to establish the
inclusive-fitness advantage of the trait itself but rather
to establish the nature of the selection working on the
age of menopause in a contemporary population. To do
this, it is necessary to estimate the magnitude of the genetic component contributing to the range of variation
in the age of menopauseits heritability. Although
over the past two decades several researchers have investigated the genetic basis of very early menopause
(Cramer and Xu 1996, Cramer, Xu, and Harlow 1995,
Mattison et al. 1984, Coulam, Stringfellow, and Hoefnagel 1983, Smith, Fraser, and Noel 1979), only one study
has considered, among other factors, the hereditary contribution to the range of variation in the age of menopause, finding a strong intergenerational association
based on daughters recall of mothers ages at meno-
pause (Torgerson et al. 1994). The complementary studies of Peccei (1998) and Snieder, MacGregor, and Spector (1998) were the first designed specifically to
estimate heritability in the age of menopause. A largely
retrospective twin study, the latter benefited from a
large sample size. In contrast, the intergenerational
study reported here benefited from the precision and reliability of prospective data.
Although such a study may seem long overdue, it has
proved difficult to find prospectively recorded menstrual histories from a sufficiently large number of related individuals. In sharp contrast to menarche, the occurrence of menopause is an unrecognized event,
verifiable only in retrospect by some arbitrary but welldefined criterion.2 For both intergenerational and sib
studies, maintaining sufficient subject participation
and compliance to produce an adequate sample of prospective data represents a considerable challenge.
Thanks to the dedication of Alan E. Treloar and his associates and successors in the TTMRH program, prospective data on the menstrual histories and completed
reproductive careers of two generations of related
women now exist (Treloar 1974, 1981). The data for the
heritability estimates presented here come from 117 dyads made up of mothers among the first recruits of the
mid-1930s and their daughters.
An estimate of heritability is obtained from the ratio
of the phenotypic correlation between two variables
and the genetic correlation that would result if the trait
were totally inherited, which is equivalent to the coefficient of relationship between the individuals concerned (Falconer 1989). For mother/daughter dyads,
h2 rP /1/2. The simplest way to obtain the phenotypic
correlation is through standard regression analysis,
where r is calculated from the slope of the regression
and the two variances [r ( 2x / 2y) 1/2]. To qualify for
this, all subjects must have experienced the event in
question. Here this means that both mother and daughter must have gone through natural menopausethat
is, menopause not mediated by use of exogenous hormones during the perimenopause or resulting from hysterectomy. At present there are 21 natural-menopause
dyads in the TTMRH database. The mother/daughter
correlation from standard regression is r DM 0.250
0.22, where the interval estimate is based on the assumption of normality. This yields a heritability estimate of h 2 0.5 0.44 (p 0.127). To check the standard errors of the correlation coefficient, I employed
bootstrap analysis (Efron and Tibshrani 1986). With
2,000 random samples of 21 dyads, the median correla2. The current World Health Organization criterion for menopause
is 12 months without a menstrual bleeding. The TTMRH project
uses a more conservative standard whereby the occurrence of
menopause is established only after a subject has recorded no
bleeding during the entire year covered by her latest calendar card.
Menopause the eventas opposed to the period of transition
known as the climacteric, which typically lasts about five years
(Soules and Bremner 1982)is then designated as the first day of
terminal amenorrhea (Treloar 1981).

table 1
Estimates of Heritability in the Age of Menopause from Standard Regression Analysis of Uncensored Dyads
and SA/MLE and Linear Regression Analysis with Censored Data
Method
Standard regression
SA/MLE
Censored regression
Mother
Daughter
Mean (Variance)
Mean (Variance)
Slope
()
Correlation
(r DM)
Heritability
(h2)
Significance
(p)
21
117
117
51.49 (9.28)
51.91 (9.46)
51.91 (9.46)
48.19 (11.43)
53.94 (22.79)
0.277 0.25
0.287 0.202
0.548 0.505
0.250 0.22
0.185 0.123/0.136
0.353 0.325
0.50 0.44
0.370 0.246/0.272
0.706 0.650
0.127
0.078
0.140
tion was rDM 0.250 and the mean was r 0.23

0.273, suggesting that the standard regression estimate
is reasonable.
There are an additional 96 censored dyads, in which
all of the mothers experienced natural menopause but
the daughters were still pre- or perimenopausal at the
end of the period of observation or, alternatively, began
taking hormones at some recorded time prior to the cutoff date. Not only is there a good deal of valuable information in these censored dyads, but neglecting them
could bias the results and thus weaken any conclusions (Elandt-Johnson and Johnson 1980; Barlow 1996:
86). Survival-analysis/maximum-likelihood-estimation
(SA/MLE) procedures were adopted to analyze the combined dataset including the uncensored and censored
dyads (N 117).
In MLE, some parametric failure distribution is selected which appears to fit the uncorrelated uncensored
data (Elandt-Johnson and Johnson 1980, Miller 1981).
One then finds the parameter values which maximize
the datas fit to the model. The reliability of the MLE
procedure depends on how well the chosen model fits
the real data. The advantage with MLE is that one can
estimate the probability that the event in question has
not occurred and write the likelihood function as the
probability of failure and survival (SA/MLE), thereby incorporating censored data. One can also use SA/MLE to
find the best fit for multivariate data (Johnson and Kotz
1972). For this analysis, a bivariate Gaussian was chosen for the joint probability function, with the correlation coefficient r being one of the models parameters.3
The SA/MLE correlation rDM 0.185 0.123/0.136
is lower than the standard regression correlation, but
with smaller errorsthe resulting heritability estimate
of h2 0.370 0.246/0.272 has a higher level of significance (p 0.078) (Elandt-Johnson and Johnson 1980,
Miller 1981, Barlow 1996).
The 117 uncensored and censored dyads were also analyzed by a linear regression model extended to include
data singly censored on the right. In this model the
3. A bivariate Gaussian is appropriate if regression of one variable
on the other is linear and the two variables singly or together are
normally distributed. With the uncensored data, both conditions
are satisfied.
slope and the intercept estimators are weighted linear

combinations of the uncensored data points, the
weights being derived from Millers (1976) modified version of the nonparametric Kaplan-Meier maximum
likelihood estimator of a distribution function F. As
stated earlier, with standard linear regression, the correlation coefficient r is calculated from the slope and the
two variances, r ( 2x / 2y) 1/2. This is not possible with
linear regression extended to include censored data, because the variance of the censored variable is unknown,
and in contrast to the situation with SA/MLE, estimating this variance is not part of the process because the
maximum likelihood estimator is nonparametric. However, one can use the variance for the censored variable
from the SA/MLE analysis to estimate r from the slope
of the censored regression model. The slope of the regression using Millers modified Kaplan-Meier estimator to derive the weights of the uncensored points is
0.548 0.505. With the variance of 9.46 for mothers
and the SA/MLE variance of 22.79 for daughters, rDM
0.353 0.325, yielding a heritability estimate of h2
0.706 0.650 ( p 0.140).4
The results of all analyses are summarized in table 1
and figure 1. The hypothesis I am testing is that the true
heritability is zero, against the alternative that it is positive, calling for a one-sided test. The SA/MLE is within
the 90% confidence limit (p 0.078), while both regression estimates, because of their relatively larger errors,
are not. Still, the results of all three analyses are consistent, with positive heritabilities close to the lower 90%
confidence limit and central values which suggest that
heritability in the age of menopause is relatively large
for a life-history trait (Stearns 1992, Roff 1997).
My results are also consistent with those of Snieder,
MacGregor, and Spector (1998), who estimate heritability in the age of menopause to be 0.63, with a 95% confidence interval of 0.530.71. Differing from my prospective mother/daughter study with a relatively small
sample (21 uncensored plus 96 censored-daughter dy4. Although the extended regression results are consistent with
those obtained by standard regression and SA/MLE, they are less
trustworthy, because the weighting approximation adopted and the
use of the borrowed censored variance contribute a systematic error that is difficult to estimate.

F i g. 1. Correlation coefficients with standard errors estimated by standard regression, SA/MLE, and censored
regression. Arrows point to lower confidence limit at 0.1.
ads), the maximum likelihood model of Snieder and his

colleagues is based on present status and recalled ages
of menopause from a large twin sample (265 uncensored
plus 260 singly censored twin pairs).5
With the heritabilities reported here and the observed
amount of phenotypic variation in the age of menopause, there is the potential for directional selection if
it led to an increase in fitness. For example, assuming
that prolonged fertility involves only a fitness benefit
and that the rate of fertility increase is constant, with
a mean of 52 years, variance of 9.5, and h2 0.4, the
mean age of menopause could increase 0.07 years in one
generation (the response to selection R h2[ 2M /M ]). In
2,000 years, the mean age of menopause could increase
7 years. Yet the evidence, though not conclusive, suggests that no secular trend of any kind has occurred over
the past 150 years (McKinlay, Bifano, and McKinlay
1985, Gray 1976, Flint 1978). Indeed, it seems plausible
that the mean age of menopause has not experienced directional selection for at least the past 2,000 years
(Amundsen and Diers 1973, 1970; Post 1971).
There are three possible explanations for the apparent
lack of a secular trend: (1) The age of menopause is under stabilizing selection, which means that in contemporary populations there must be some cost to prolonged fertility functioning in the presence of a positive
correlation between age of menopause and fitness.
(2) There is no selection on the mean age of menopause.
(3) There is an upward trend, but the available data are
not good enough to show it.
With regard to stabilizing selection, at present we do
not know the cost of prolonging fertility or whether the
onset of menopause and fitness are positively correlated. In looking for the positive correlation, researchers
5. The larger heritability estimate of Snieder, MacGregor, and Spector (1998) would usually be attributable in part to the nonadditive
genetic effects of dominance and epistasis and/or the confounding
effects of shared environment, all of which are expected in sib analyses. However, according to Snieder et al. (p. 877), shared environment and dominance genetic effects did not contribute significantly to the explanation of the data.
often use parity as a proxy for fitness when what is really required is number of children who survive to reproductive age orbetter yetnumber of grandchildren. Even so, the data are ambiguous with regard to the
existence and meaning of a positive association between age of menopause and parity in contracepting
populations (Parazzini, Negri, and LaVecchia 1992,
Whelan et al. 1990, Brambilla and McKinlay 1989, Stanford et al. 1987, McKinlay, Bifano, and McKinlay 1985).
It is unlikely that a positive correlation would be easier
to find in natural-fertility populations, there being a remarkable similarity in female reproductive physiology
among widely differing populations both historical and
extant (Wood 1989, Bongaarts 1980). This typical reproductive pattern includes rapidly declining fertility prior
to menopause, such that a downward shift in the age of
menopause does not substantially reduce the probability of childbearing (Wood and Weinstein 1988). It appears that any potential fitness advantage from greater
parity is disassociated from the age of menopause, because it stems primarily from adjustments to interbirth
intervals during the peak reproductive years.
If parity per se is not a good proxy for fitness, it is also
clear that the cost of prolonging reproductive life span
is not some dose-related phenomenon associated with
parity such as worsening maternal depletion. In extant
and historical natural-fertility populations, total fertility rates (TFR) vary enormously, as do nutritional status
and mortality rates (Wood 1994). For the well-nourished
Hutterites TFR is 10, for the Gainj and the !Kung TFRs
are slightly over 4, and for a poorly nourished mid-19thcentury English population TFR is about 7 (Wood 1994).
Yet mean age at last birth is around 40 years (Wood
1994). The large cross-populational differences in TFR
are the result of significantly differing mean interbirth
intervals, which are mandated by considerable socioecological and survivorship differences (Wood 1994). In
addition, in empirical tests of the inclusive-fitness advantage of menopause, investigators have found that
neither increased maternal and offspring mortality nor
the opportunity costs of prolonged fertility to existing

Coulam, Stringfellow, and Hoefnagel 1983, Mattison et

al. 1984). Despite the usual caveats that heritability estimates pertain only to the sample population, I believe
that a heritability estimate with confidence limits such
as are reported here has better predictive potential than
the multivariate studies with respect to informing
women when they may expect to undergo menopause.
Figure 2 shows daughters expected age of menopause
based on mothers age at menopause and the SA/MLE
heritability estimate.
References Cited
F ig. 2. Daughters predicted age of menopause and
one-sigma confidence limits, based on mothers age
at menopause and SA/MLE h 2 0.185 0.130 (solid
lines) and mean age and one-sigma interval (dashed
lines).
progeny are sufficient to maintain menopause (Hill and

Hurtado 1991, Hawkes, OConnell, and Blurton Jones
1997).
Still, the heritability estimates presented here suggest
that the mean age of menopause is under some degree
of stabilizing selection, which means that there is a cost
to prolonging menstruating life span even now. Moreover, this costwhatever it may becan logically be
assumed to have existed at least since the beginning of
agriculture and animal husbandry, which suggests that
there is nothing peculiar to the foraging lifestyle now
or in the environment of evolutionary adaptedness that
makes premature reproductive senescence adaptive for
human females. If this is true, we need to look at female
reproductive physiology and the history of survivorship
probabilities rather than specific subsistence behaviors.
Clearly, at present the most useful area of investigation is verifying or discounting the existence of an upward secular trend in the mean age of menopause. The
absence of any significant trend would be reasonable evidence of stabilizing selection, although there remains
the alternative possibility that the age of menopause is
and has been under no selection at all, implying that the
timing of menstrual cessation is a nonissue. However,
recalling the asynchrony between the sharp fertility decline and the termination of menstruation in human females, I suggest that premature termination of fertility
could be an adaptation even without stabilizing selection on the mean age of menopause.
Another function of a heritability estimate is its predictive value (Falconer 1989) and hence its clinical potential. Indeed, most of the multivariate studies of the
factors influencing the age of menopause claim to have
generalizable clinical value relevant to family planning
and risk of diseases associated with estrogen withdrawal (e.g., osteoporosis, cardiovascular disease, and
ovarian cancer) or prolonged estrogen exposure (e.g., endometrial and breast cancer) (Cramer and Xu 1996,
Cramer, Xu, and Harlow 1995, Torgerson et al. 1994,
a m u n d s e n, d. w., a n d c. j. d i e r s. 1973. The age of menopause in medieval Europe. Human Biology 45:60512.
. 1970. The age of menopause in Classical Greece and
Rome. Human Biology 42:7986.
b a r l o w, r. j. 1996. Statistics. Chichester, U.K.: Wiley.
b o n g a a r t s, j o h n. 1980. Does malnutrition affect fecundity?
A summary of evidence. Science 208:56469.
b r a m b i l l a, d. j., a n d s. m c k i n l a y. 1989. A prospective
study of factors affecting age at menopause. Journal of Clinical
Epidemiology 42:103139.
c a r o, t. m., d. w. s e l l e n, a. p a r i s h, r. f r a n k, d. m.
b r o w n, e. v o l a n d, a n d m. b o r g e r h o f f m u l d e r .
1995. Termination of reproduction in nonhuman and human
primate females. International Journal of Primatology 16:205
20.
c o u l a m, c. b., s. s t r i n g f e l l o w, a n d d. h o e f n a g e l.
1983. Evidence for a genetic factor in the etiology of premature
ovarian failure. Fertility and Sterility 40:69395.
c r a m e r, d. w., a n d h. x u. 1996. Predicting age at menopause. Maturitas 23:31926.
c r a m e r, d. w., h. x u, a n d b. l. h a r l o w. 1995. Family
history as a predictor of early menopause. Fertility and Sterility 64:74045.
e f r o n, b., a n d r. t i b s h r a n i. 1986. Bootstrap methods for
standard errors, confidence intervals, and other measures of statistical accuracy. Statistical Science 1:5477.
e l a n d t - j o h n s o n, r. c., a n d n. l. j o h n s o n . 1980. Survival models and data analysis. New York: Wiley.
f a l c o n e r, d. s. 1989. 3d edition. Introduction to quantitative genetics. New York: Longman Scientific.
f l i n t, m. 1978. Is there a secular trend in age of menopause?
Maturitas 1:13339.
g r a y, r. h. 1976. The menopause: Epidemiological and demographic considerations, in The menopause: A guide to current
research and practice. Edited by R. J. Beard, pp. 2540. Lancaster, U.K.: MTP Press.
h a w k e s, k., j. f. o c o n n e l l, a n d n. g. b l u r t o n
j o n e s . 1997. Hadza womens time allocation, offspring provisioning, and the evolution of long postmenopausal life spans.
current anthropology 38:55177.
h i l l, k., a n d a. m. h u r t a d o. 1991. The evolution of premature reproductive senescence and menopause in human females: An evaluation of the grandmother hypothesis. Human Nature 2:31350.
j o h n s o n, n. l., a n d s. k o t z. 1972. Bivariate and trivariate normal distribution, in Distributions in statistics: Continuous multivariate distributions, pp. 84131. New York: Wiley.
m c k i n l a y, s. m., n. l. b i f a n o, a n d j. b. m c k i n l a y.
1985. Smoking and age at menopause in women. Annals of Internal Medicine 103:35056.
m a r s h, h., a n d t. k a s u y a. 1984. Changes in the ovaries
of the short-finned pilot whale, Globicephala macrorhychus,
with age and reproductive activity, in Reproduction in
whales, dolphins, and porpoises. Edited by W. F. Perrin et al.,
pp. 31135. Cambridge: International Whaling Commission.
m a t t i s o n, d. r., m. i. e v a n s, w. b. s c h w i m m e r, b. j.

w h i t e, b. j e n s e n, a n d j. d. s c h u l m a n. 1984. Familial
premature ovarian failure. American Journal of Human Genetics 36:134148.
m i l l e r, r u p e r t g., j r . 1976. Least-squares regression with
censored data. Biometrika 63:44964.
. 1981. Survival analysis. New York: Wiley.
p a r a z z i n i, f., e. n e g r i, a n d c. l a v e c c h i a. 1992. Reproductive and general lifestyle determinants of age at menopause. Maturitas 15:14149.
p a v e l k a, m. s. m., a n d l. m. f e d i g a n. 1991. Menopause:
A comparative life-history perspective. Yearbook of Physical
Anthropology 34:1338.
p e c c e i, j o c e l y n s c o t t. 1995. The origin and evolution of
menopause: The altriciality-lifespan hypothesis. Ethology and
Sociobiology 16:42549.
. 1998. Heritability in the age of menopause and the genetic correlation between the ages of menarche and menopause. Ph.D. diss., University of California, Los Angeles, Calif.
p o s t, j. b. 1971. Ages at menarche and menopause: Some mediaeval authorities. Population Studies 25:8387.
r o f f, d e r e k a. 1997. Evolutionary quantitative genetics. New
York: Chapman Hall.
r o g e r s, a. r. 1993. Why menopause? Evolutionary Ecology
7:40620.
s m i t h, a., i. a. f r a s e r, a n d m. n o e l. 1979. Three siblings with premature gonadal failure. Fertility and Sterility 32:
52830.
s n i e d e r, h., a. j. m a c g r e g o r, a n d t. d. s p e c t o r.
1998. Genes control the cessation of a womans reproductive
life: A twin study of hysterectomy and age at menopause. Journal of Clinical Endocrinology and Metabolism 83:187580.
s o u l e s, m. r., a n d w. j. b r e m n e r. 1982. The menopause
and climacteric: Endocrinologic basis and associated symptomatology. American Geriatrics Society 30:54761.
s t a n f o r d, j. l., p. h a a r t g e, l. a. b r i n t o n, r. n.
h o o v e r, a n d r. b r o o k m e y e r. 1987. Factors influencing
the age at natural menopause. Journal of Chronic Disease 40:
9951002.
s t e a r n s, s t e p h e n c. 1992. The evolution of life histories.
New York: Oxford University Press.
t o r g e r s o n, d. j., a. a v e n e l l, i. t. r u s s e l l, a n d
d. m. r e i d. 1994. Factors associated with age at menopause
in women aged 4549. Maturitas 19:8392.
t r e l o a r, a l a n e. 1974. Menarche, menopause, and intervening fecundability. Human Biology 46:89107.
. 1981. Menstrual cyclicity and the pre-menopause. Maturitas 3:24964.
w h e l a n, e. a., d. p. s a n d l e r, d. r. m c c o n n a u g h e y,
a n d c. r. w e i n b e r g. 1990. Menstrual and reproductive
characteristics and age at natural menopause. American Journal of Epidemiology 131:62532.
w o o d, j a m e s w. 1989. Fecundity and natural fertility in humans. Oxford Reviews of Reproductive Biology 11:61109.
. 1994. Dynamics of human reproduction: Biology, biometry, demography. New York: Aldine de Gruyter.
w o o d, j a m e s w., a n d m a x i n e w e i n s t e i n. 1988. A
model of age-specific fecundability. Population Studies 42:85
113.


A Conversation With Joan Vincent

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Conversation With Joan Vincent

Uploaded by

Copyright:

Available Formats

A Conversation with Joan Vincent

Author(s): David Nugent

Stable URL: http://www.jstor.org/stable/10.1086/200050 .

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

for a university correspondence course at LSE [London

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

a seminar about Jinghpaw terminology. And I think my

JV: Yes. I was accepted by the anthropology department,

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 533

up there had always been class. And always nationality,

DN: Lets talk about your fieldwork in Uganda.

JV: Once you were a fellow at the East African Institute

3. Leys had been a visiting professor at the University of Chicago

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

involved in the antiwar movement, and there was a big

DN: Let me ask you about African Elite, in which you

JV: Thats right! Once again [as in Zanzibar], I was

JV: It was written as a doctoral thesis, and very quickly,

DN: Lets talk about The Structuring of Ethnicity

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 535

East Africa. When that went sour, with Amin, I felt

Uganda [1977a], Teso in Transformation: Colonial

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

do it? And this is when I got labeled as a Marxist

stantly using the term capitalist agriculture. I provide

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 537

DN: In 198586, you were a fellow at the Institute for

anthropologists usually apply to other societies and

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

sive frontier of industrial capitalism. Another is what

DN: In recent years there has been some very important

JV: I wish youd stop that last sort of characterization

JV: I thought it had long been generally understood that

9. At the conference International Perspectives on the Great Irish

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 539

that allowed us to be attuned to things going on that we

era [see also Vincent n.d.c; Vincent and Bond n.d.]. Do

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

state at all. At the time of the conflict human rights

liked because I had a book to review and yet had space

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 541

. 1988a. Equality and enterprise: The cooperative movement

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

. n.d.a. Seeds of revolution: The culture and politics of the

Painted Slabs from

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 543

F ig. 1. Location of sites mentioned in the text. A11,

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

F i g. 2. Site plan of Steenbokfontein Cave.

50 b.p. from a sample immediately above the collapsed

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 545

The surfaces of two of the collapsed slabs are clearly

preservation and that of rock art imagery generally at

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

there is reason to suspect that this contrast may have

This content downloaded from 146.164.3.22 on Wed, 20 May 2015 08:58:13 AM

Volume 40, Number 4, AugustOctober 1999 547