1
The Concept(s) of Culture
WILLIAM H. SEWELL, JR.
The aim of this chapter is to reflect upon the concept-or more properly
the concepts-of culture in contemporary academic discourse.
Trying to clarify what we mean by culture seems both imperative
and impossible at a moment like the present, when the study of culture
is burgeoning in virtually all fields of the human sciences. Although
I glance at the varying uses of "culture" in a number of disciplines,
my reflection is based above all on the extensive debates that
have occurred in anthropology over the past two decades-debates'.,
in which some have questioned the very utility of the concept.' I feel
strongly that it remains as useful, indeed essential, as ever. But given
the cacophony of contemporary discourse about culture, I also believe
that the concept needs some reworking and clarification.s,
The current volatility of the concept of culture sharply contrasts
with the situation in the early 1970s, when I first got interested in a
cultural approach to social history. At that time it was clear that
if you wanted to learn about culture, you turned to the anthropologists.
And while they by no means spoke in a single voice, they
shared a Widespread consensus both about the meaning of culture
and about its centrality to the anthropological enterprise. I began
borrowing the methods and insights of cultural anthropology as a
means of learning more about nineteenth-century French workers.Ji
Cultural analysis, I hoped, would enable me to understand the
If
'I'
meaning of workers' practices that I had been unable toget at by using
quantitative and positivist methods-my standard tool kit as a
practitioner of what was then called "the new social history." 2 I ex­
'~i
perienced the encounter with cultural anthropology as a tum from
a hardheaded, utilitarian, and empiricist materialism-which had
both liberal and marxisant faces-to a wider appreciation of the
35
37]6 Willialll H 5cl/'('II, [r.
range of human possibilities, both in the past and in the present.
Convinced that there was more to life than the relentless pursuit of
wealth, status, and power, I felt that cultural anthropology could
show us how to get ill that "more."}
Anthropology at the time had a virtual monopoly on the concept
of culture. In political science and sociology, culture was associated
with the by then utterly sclerotic Parsonian theoretical synthesis. The
embryonic "cultural studies" movement was still confined to a single
research center in Birmingham. And literary studies were still fixated
on canonical literary texts-s-nlthough the methods of studying them
were being revolutionized by the importation of "french" structurnlist
and poststructuralist theory. Moreover, the mid-rooos to the
mid-1970S marked the glory years uf American cultural anthropology,
which may be said to have reached its apotheosis with the publication
of Clifford Ceertzs phenomenally influentialllltcrp,·etl7tiOlI of
Cult urcs in 197J. 1
Not only did anthropology have no serious rivals
in the study of culture, but the creativity and prestige of cultural anthropology
were at a very high point.
During the 1980s and 1990s, the intellectual ecology of the study
of culture has been transformed by a vast expansion of work on culture-indeed,
a kind of academic culture mania has set in. The new
interest in culture has swept over a wide range of academic disciplines
and specialties. The history of this advance differs in timing
and con tent in each field, but the cumulative effects are undeniable.
In literary studies, which were already being transformed by French
theory in the 1970s, the 1980s marked a turn to a vastly wider range
of texts, quasi-texts, paratexts, and text analogs. If, as Derrida declared,
nothing is extratextual ("il n'y a pas de hors-texte"), literary
critics could direct their theory-driven gaze upon semiotic products
of all kinds-legal documents, political tracts, soap operas, histories,
talk shows, popular romances-and seek out their intertextualities."
Consequently, as such "new historicist" critics as Stephen
Greenblatt and Louis Montrose recognize, literary study is increasinglv
becoming the study of cultures." In history the ea rly and rather
self-conscious bt)!T\1\ving from anthropology has been followed by a
theoretically heterogeneous rush to the study of culture, one madded
as much on literary studies or the work of Michel Foucault ClSon
anthropology. As a consequence, the self-confident "new social hisTile
Conceptts) «[Culmre
tory" of the 1960s and 1970S was succeeded by an equally selfconfident
"new cultural history" in the 1980s/
In the late 19708, an emerging "sociology of culture" began by
applying standard sociological methods to studies of the production
and marketing of cultural artifacts-music, art, drama, and literature.
By the late 1980s, the work of cultural sociologists had broken
out of the study of culture-producing institutions and moved toward
studying the place of meaning in social life more generally, Feminism,
which in the 1970S was concerned above all to document
women's experiences, has increasingly turned to analyzing the discursive
production of gender difference. Since the mid-rosos the
new quasi-discipline of cultural studies has grown explosively in
a variety of different academic niches-for example, in programs or
departments of film studies, literature, performance studies, or communications.
In political science, which is well known for its propensity
to chase headlines, interest in cultural questions has been
revived by the recent prominence of religious fundamentalism, nationalism,
and ethnicity, which look like the most potent sources of
political conflict in the contemporary world. This frenetic rush to the
study of culture has everywhere been bathed, to a greater or lesser
extent, in the pervasive transdisciplinary influence of the French
poststructuralist trinity of Lacan. Derrida, and Foucault.
It is paradoxical that as discourse about culture becomes ever
~
more pervasive and multifarious, anthropology, the discipline that
invented the concept-or at least shaped it into something like its
present form-is somewhat ambivalently backing away from its
long-standing identification with culture as its keyword and central
symbol. For the past decade and a half, anthropology has been rent
by a particularly severe identity crisis, which has been manifested in
anxiety about the discipline's epistemology, rhetoric, methodological
procedures, and political implications," The reasons for the crisis are
many-liberal and radical guilt about anthropology's association
with Euro-American colonialism, the disappearance of the supposedly
"untouched" or "primitive" peoples who were the favored subjects
for classic ethnographies, the rise of "native" ethnographers
who contest the right of European and American scholars to tell the
"truth" about their people, and the general loss of confidence in the
possibility of objectivity that has attended poststructuralism and
39}8 Willinll1 H Sewell,Jr.
postmodernism. As anthropology's most central and distinctive concept,
"culture" has become a suspect term among critical anthropologists-who
cla im that both in academia and in public discourse,
talk about culture tends to essentialize, exoticize, and stereotype
those whose ways of life are being described and to naturalize their
differences from white middle-class Euro-Americans. If Geertz's
phrase "The Interpretation of Cultures" was the watchword of anthropology
in the 1970s, Lila Abu-Lughods "Writing against Culture"
more nearly sums up the mood of the late 1980s and the 1990S.9
As [ohn Brightman points out in his superb commentary on the
recent disputes about culture in anthropology, the anthropological
critics of the 1980s and 1990S have exhibited Widespread "lexical
avoidance behavior," either placing the term "culture" in quotation
marks when it is used, refusing to use "culture" as a noun while continuing
to use it as an adjective (as in "cultural anthropology"), or replacing
it with alternative lexerues such as "habitus," "hegemony,"
or "discourse." HI This emerging anthropological tabu seems to me
mistaken on two counts. First, it is based on the implicit assumption
that anthropology "owns" the lexeme and that it is therefore responsible
for any abuses that might be pcrpet ratcd by others employing
the term. Second, it assumes that anthropological abstention from
the use of the lexerne will magically abolish such abuses. The truth is
that the term has escaped all possibility of control by anthropologists:
whatever lexical practices the anthropologists may adopt, talk
about culture will continue to thrive-in both abusive and acceptable
ways-in a wide range of other academic disciplines and in ordinary
language as well. Moreover, as Brightman again points out,
even the critical anthropologists find it impossible to give up the COIIceptof
culture, as opposed to the lexeme. James Clifford's lament that ~;
,
"culture is a deeply compromised concept that I cannot yet do without"
seems emblematic of the unresolved ambivalence: the concept
is compromised and he hopes in the future to do without it, but because
it continues to perform valuable intellectual work the fateful
act of renunciation is indefinitely deferred."! If, as I believe, Clifford
is right that we G\t1IHlt do without a concept of culture, I think we
should try to shape it into one we can work with. We need to modify,
rearticulato, and revivify the concept, retaining and reshaping
what is useful and discarding what is not.
Tile Concepits) of Culture
WHAT DO WE MEAN BY CULTURE?
Writing in 198], Raymond Williams declared that "culture is one of
the two or three most complicated words in the English language."12
Its complexity has surely not decreased since then. I have neither the
competence nor the inclination to trace out the full range of meanings
of "culture" in contemporary academic discourse. But some attempt
to sort out the different usages of the word seems essential,
and it must begin by distinguishing two fundamentally different
meanings of the term.
In one meaning, culture is a theoretically defined category or as­

pect of social life that must be abstracted out from the complex real­

ity of human existence. Culture in this sense is always contrasted to

some other equally abstract aspect or category of social life that is not

culture, such as economy, politics, or biology. To designate some­

thing as culture or as cultural is to claim it for a particular academic

discipline or subdiScipline-for example, anthropology..or cultural

sociology-or for a particular style or styles of analysis-for ex­

ample, structuralism, ethno-science, componential analysis, decon­

struction, or hermeneutics. Culture in this sense-as an abstract ana­

lytical category-only takes the singular. Whenever we speak of

"cultures," we have moved to the second fundamental meaning.

In that second meaning, culture stands for a concrete and

bounded world of beliefs and practices. Culture in this sense is com­

monly assumed to belong to or to be isomorphic with a "society" or

with some dearly identifiable subsocietal group. We may speak of

"American culture" or "Samoan culture," or of "middle-class cul­

ture" or "ghetto culture."13 The contrast in this usage is not between

.'
culture and not-culture but between one culture and another-between
American, Samoan, French, and Bororo cultures, or between
middle-class and upper-class cultures, or between ghetto and mainstream
cultures.
This distinction between culture as theoretical category and culture
as concrete and bounded body of beliefs and practices is, as far
as I can discern, seldom made. Yet it seems to me crucia I for thinking
clearly about cultural theory. It should be clear, for example, that
Ruth Benedict's concept of cultures as sharply distinct and highly
integrated refers to culture in the second sense, while Claude Levi­
41WiJlirJ11l H SCil'df, Jr.4°
Strauss's notion that cultural meaning is structured by systems of oppositions
is a claim about culture in the first sense, Hence their theories
of "culture" arc, strictly speaking, incommensurate: they refer to
different conceptual universes. Failure to recognize this distinction
between two fundamentally different meanings of the term has real
consequences for contemporary cultural theory; some of the impasses
of theoretical discourse in contemporary anthropology are attributable
precisely to an unrecognized elision of the two, TIlUS, a
dissatisfaction with "Benedictine" etlmographies that present cultures
as uniformly well-bounded and coherent has led to what seem
to me rather confused attacks on "the culture concept" in generalatl,lcks
that [ail to distinguish Benedictine claims about the tight integrillion
of cultures from Levi-Straussian claims about the semiotic
coherence of culture as a system of meaningsJ ' Conversely, anthropologists
who defend the culture concept also tend to conflate the
two meanings, regarding claims that cultures are rent with fissures
or that their boundaries are porous as implying an abandonment of
the concept of culture altogether.
Here, 1 will be concerned primarily with culture in the first
sense-culture as a category of social life. One must have a clear conception
of culture at this abstract level in order to deal with the more
concrete theoretical question of how cultural differences are patterned
and bounded in space and time. Once 1have sketched out my
own ideas about what an adequate abstract theory of culture might
look like, I will return to the question of culture as a bounded universe
of beliefs and practices-to the question of cultures in the
Benedictine sense.
CULTURE AS A CATEGORY OF SOCIAL LIFE
Cultu re as a category of social life has itself been conceptualized in a
number of different ways. Let me begin by specifying some of these
different conceptualizations, moving from those I do not find especially
useful to those 1 find more adequate.
Cultnre t1~ learned bclnnnor. Culture in this sense is the whole body
of f'ri1ctice~, beliefs, inst itut ions, customs, habits, myths, and so on
built up by humans and passed on from gcneration to genemtion.ln
this usage, culture is contrasted to nature: its possession is what distinguishes
LIS from other animals. When anthropologists were strugrile
Conceptis) a/Culture
gling to establish that differences between societies were not based
on biological differences between their populations-that is, on
race-a definition of culture as learned behavior made sense. But
now that racial arguments have virtually disappeared from anthropological
discourse, a concept of culture so broad as this seems impossibly
vague; it provides no particular angle or analytical purchase
on the study of social life.
A narrower and consequently more useful conceptualization of
culture emerged in anthropology during the second quarter of the
twentieth century and has been dominant in the social sciences generally
since World War II. It defines culture not as all learned behavior
but as that category or aspect of learned behavior that is concerned
with meaning. But the concept of culture-as-meaning is in fact
a family of related concepts; meaning may be used to specify a cul'.
tural realm or sphere in at least four distinct ways, each of which is
defined in contrast to somewhat differently conceptualized noncultural
realms or spheres.
Culture as all institutional sphere devoted to the making of meaning.
This conception of culture is based on the assumption that social
formations are composed of clusters of institutions devoted to specialized
activities. These clusters can be assigned to variously defined
institutional spheres-most conventionally, spheres of politics,
economy, society, and culture. Culture is the sphere devoted
specifically to the production, circulation, and use of meanings. The
cultural sphere may in turn be broken down into the subspheres of
which it is composed: say, of art, music, theater, fashion, literature,
religion, media, and education. The study of culture, if culture is
defined in this way, is the study of the activities that take placewithin
these institutionally defined spheres and of the meanings produced
in them.
This conception of culture is particularly prominent in the discourses
of sociology and cultural studies, but it is rarely used in anthropology.
Its roots probably reach back to the strongly evaluative
conception of culture as a sphere of "high" or "uplifting" artistic and
intellectual activity, a meaning that Raymond Wi! Iiams tells us came
into prominence in the nineteenth century.IS Bu t in contemporary
academic discourse, this usage normally lacks such evaluative and
hierarchizing implications. The dominant style of work in American
sociology of culture has been demystifying: its typical approach has
42 43Williall1 H. Setccll, Jr.
been 10 uncover the largely self-aggrandizing, class-interested, rnanipulntive,
or professionalizing institutional dynamics that undergird
prestigious museums, artistic styles, symphony orchestras, or
philosophical schools. And cultural studies, which has taken as its
particular mission the appreciation of cultural forms disdained by
the spokesmen of high culture-rock music, street fashion, crossdressing,
shopping malls, Disneyland, soap operas-employs this
same basic definition of culture. It merely trains its analytical attenlion
on spheres of meaning production ignored by previous analysts
and regarded as debased by elite tastemakers.
TIle problem with such a concept of culture is that it focuses only
on a ceriain range of meanings, produced in a certain range of institutional
locations-c-on self-consciously "cultural" institutions and
on expressive, artistic, and literary systems of meanings. This use of
the concept is to some extent complicit with the Widespread notion
that meanings arc of minimal importance in the other "noncultural"
institutional spheres: that in political or economic spheres, meanings
are merely superstructural excrescences. And since institutions in
political and economic spheres control the great bulk of society's resources,
viewing culture as a distinct sphere of activity milY in the
end simply confirm the Widespread presupposition in the "harder"
social sciences that culture is merely froth on the tides of society. 111e
rise of a cultural sociology that limited itself to studying "cultural"
institutions effected a partition of subject matter that was very unfavornble
to the cultural sociologists. Indeed, only the supercession
of this restrictive concept of culture has made possible the explosive
growth of the subfield of cultural sociology in the past decade.
Culture as creativity 01' ascllcy. This usage of culture has grown up
particularly in traditions that posit a powerful "material" determinism-most
notably Marxism and American sociology. Over the past
three decades or so, scholars working within these traditions have
carved out a conception of culture as a realm of creativity that escRpes
from the otherwise pervasive determination of social action by
economic or social structures. In the Marxist tradition, it was probably
E. r.Thompson's !v1I1ki"S (:fHit'Ellglislr Workillg Class that first conceptualized
culture as a realm of agency, and it is particularly English
Marxists-v-for example, Paul Willis in Learning to Labor-who
hilH' elaborated this conccption.!" But the defining opposition on
which this concept of culture rests- culture versus structure-has
The Conceptts) o/Culture
also become pervasive in the vernacular of American sociology. One
clear sign that American anthropologists and sociologists have different
conceptions of culture is that the opposition between culture
and structure-an unquestioned commonplace in contemporary sociological
discourse-is nonsensical in anthropology.
In my opinion, identifying culture with agency and contrasting it
with structure merely perpetuates the same determinist materialism
that "culturalist" Marxists were reacting against in the first place. It
exaggerates both the implacability of socioeconomic determinations
and the free play of symbolic action. Both socioeconomic and cultural
processes are blends of structure and agency. Cultural actionsay,
performing practical jokes or writing poems-is necessarily
constrained by cultural structures, such as existing linguistic, visual,
or ludic conventions. And economic action-such as the manufacture
or repair of automobiles-is impossible without the exercise of
creativity and agency. The particulars of the relationship between
structure and agency may differ in cultural and economic processes,
but assigning either the economic or the cultural exclusively to structure
or to agency is a serious category error.
This brings us to the two concepts of culture that I regard as most
fruitful and that I see as currently struggling for dominance: the concept
of culture as a system of symbols and meanings, which was
hegemonic in the 1960s and 1970S'and the concept of culture as practice,
which has become increasingly prominent in the 1980s and
1990s.
Cultureasasystem ofsymbolsand meanings. This has been the dominant
concept of culture in American anthropology since the 1960s. It
was made famous above all by Clifford Geertz, who used the term
"cultural system" in the titles of some of his most notable essays.!"
The notion was also elaborated by David Schneider, whose writings
had a considerable influence within anthropology but lacked
Geertz's interdisciplinary appeal." Geertz and Schneider derived the
term from Talcott Parsons's usage, according to which the cultural
system, a system of symbols and meanings, was a particular "level of
abstraction" of social relations. It was contrasted to the "social system,"
which was a system of norms and institutions, and to the "personality
system," which was a system of rnotivations.!? Geertz and
Schneider especially wished to distinguish the cultural system from
the social system. To engage in cultural analysis, for them, was to
44 45IVi/Halll H. Sf1l'cll, Jr.
abstract the meaningful aspect of human action out from the flow of
concrete interactions. The point of conceptualizing culture as a system
of symbols and meanings is to disentangle, for the purpose of
analysis, the semiotic influences on action from the other sorts of
influences-demographic, geographical, biological, technological,
economic, and so on-that they are necessarily mixed with in any
concrete sequence of behavior.
Ceertzs and Schneider's post-Parsonian theorizations of cultural
systems were by no means the only available models for symbolic
anthropology in the 1960s and 1970s. The works of Victor Turner,
whose theoretical origins were in the largely Durkheirnian British
school of social anthropology, were also immensely influential."
Claude Levi-Strauss and his many followers provided an entire alternative
model of culture as a system of symbols and meaningsconceptualized,
following Saussure, as signifiers and signifieds.
Moreover, all these anthropological schools were in a sense manifes,
tations of a much broader "linguistic turn" in the human sciencesa
diverse but sweeping attempt to specify the structures of human
symbol systems and to indicate their profound influence on human
behavior. One thinks above all of such French "structuralist" thinkers
as Roland Barthes, Jacques Lacan, or the early Michel Foucault.
What all of these approaches had in common was an insistence on
the systematic nature of cultural meaning and the autonomy of symbol
systems-their distinctness from and irreducibility to other fealures
of social life. They all abstracted a realm of pure signification
out from the complex messiness of social life and specified its internal
coherence and deep logic. Their practice of cultural analysis consequently
tended to be more or less synchronic and formalist.
Culture nspractice. The past decade and a half has witnessed a pervasive
reaction against the concept of culture as a system of symbols
and meanings, which has taken place in various disciplinary 10cations
and intellectual traditions and under many different slogans-for
example, "practice," "resistance," "history," "politics," or
"culture as tool ki L" Analysts working under all these banners object
10 a portrayal of culture as logical, coherent, shared, uniform, and
static. Instead lhey insist that culture is a sphere of practical activity
c;hot through by willful action, power relations, struggle, contradiclion,
and change,
Tile COllcepf(s) of Culture
In anthropology, Sherry Ortner in 1984 remarked on the turn to
politics, history, and agency, suggesting Pierre Bourdieu's key term
"practice" as an appropriate label for this emerging sensibility. Two
years later the publication of James Clifford and George Marcus's collection
Writing Cultureannounced to the public the crisis of anthropology's
culture concept." Since then, criticisms of the concept of
culture as a system of symbols and meanings have flowed .thick and
fast. The most notable work in anthropology has argued for the contradictory,
politically charged, changeable, and fragmented character
of meanings-both meanings produced in the societies being
studied and meanings rendered in anthropological texts. Recent
work in anthropology has in effect recast culture as aperformstive
term. .
Not surprisingly, this emphasis on the performative aspect of culture
is compatible with the work of most cultural historians. Historians
are generally uncomfortable with synchronic concepts. As they
took up the study of culture, they subtly-but usually without comment-altered
the concept by stressing the contradictoriness and
malleability of cultural meanings and by seeking out the mechanisms
by which meanings were transformed. The battles in history
have been over a different issue, pitting those who claim that historical
change should be understood as a purely cultural or discursive
process against those who argue for the significance of economic and
social determinations or for the centrality of concrete "experience" in
understanding it.22
Sociologists, for rather different reasons, have also favored a more
performative conception of culture. Given the hegemony of a
strongly causalist methodology and philosophy of science in contemporary
sociology, cultural sociologists have felt a need to demonstrate
that culture has causal efficacy in order to gain recognition for
,~,
their fledgling subfield. This has led many of them to construct cul­

ture as a collection of variables whose influence on behavior can be

rigorously compared to that of such standard sociological variables

as class, etlmicity, gender, level of education, economic interest, and

the like. As a result, they have moved away from earlier Weberian,

Durkheimian, or Parsonian conceptions of culture as rather vague

and global value orientations to what Ann Swidler has termed a

"tool kit" composed of a "repertoire" of "strategies of action." 21 For

4746 William H. Setrcll, Jr.
many cultural sociologists, then, culture is not a coherent system of
symbols and meanings but a diverse collection of "tools" that, as the
metaphor indicates, are to be understood as means for the performance
of action. Because these tools are discrete, local, and intended
for specific purposes, they can be deployed as explanatory variables
in a way that culture conceived as a translocal, generalized system of
meanings C<11Ul0t.
CULTURE AS SYSTEM AND PRACTICE
Much of the theoretical writing on culture during the past ten years
has assumed that a concept of culture as a system of symbols and
meanings is at odds with a concept of culture as practice. System
and practice approaches have seemed incompatible, I think, because
the most prominent practitioners of the culture-as-systemof-meanings
approach effectively marginalized consideration of
culture-as-practice-if they didn't preclude it altogether.
This can be seen in the work of both Clifford Geertz and David
Schneider. Geertz's analyses usually begin auspiciously enough, in
that he frequently explicates cultural systems in order to resolve a
puzzle arising from concrete practices-a state funeral, trances, a
royal procession, cockfights. But it usually turns out that the issues of
practice are principally a means of moving the essay to the goal of
specifying in a synchronic form the coherence that underlies the exotic
cultural practices in question. And while Geertz marginalized
questions of practice, Schneider, in a kind of reductioadabsurdum,explicitly
excluded them, arguing that the particular task of anthropology
in the academic division of labor was to study "culture as a system
of symbols and meanings in its own right and with reference to
its own structure" and leaving to others-sociologists, historians,
political scientists, or economists-the question of how social action
was structured." A "cultural account," for Schneider, should be limited
to specifying the relations among symbols in a given domain of
meaning-which he tended to render unproblematically as known
and accepted by all members of the society and as possessing a
hi~hh· determinate formal logic."
Nor is the work of Cecrtz and Schneider unusual in its marginalization
of practice. As critics such as James Clifford have arTlie
Concept's) afCulture
gued, conventional modes of writing in cultural anthropology typically
smuggle highly debatable assumptions into ethnographic accounts-for
example, that cultural meanings are normally shared,
fixed, bounded, and deeply felt. To Clifford's critique of ethnographic
rhetoric, I would add a critique of ethnographic method. Anthropologists
working with a conception of culture-as-system have
tended to focus on clusters of symbols and meanings that can be
shown to have a high degree of coherence or systematicity-those of
American kinship or Balinese cockfighting, for instance-and to present
their accounts of these clusters as examples of what the interpretation
of culture in general entails. This practice results in what
sociologists would call sampling on the dependent variable. That is,
anthropologists who belong to this school tend to select symbols and
meanings that cluster neatly into coherent systems and pass over
those that are relatively fragmented or incoherent, thus confirming
the hypothesis that symbols and meanings indeed form tightly coherent
systems.
Given some of these problems in the work of the culture-assystem
school, the recent turn to a concept of culture-as-practice
has been both understandable and fruitful-it has effectively highlighted
many of the earlier school's shortcomings and made up some
of its most glaring analytic deficits. Yet the presumption that a concept
of culture as a system of symbols and meanings is at odds with
a concept of culture as practice seems to me perverse. System and
practice are complementary concepts: each presupposes the other.
Toengage in cultural practice means to utilize existing cultural symbols
to accomplish some end. The employment of a symbol can be
expected to accomplish a particular goal only because the symbols
have more or less determinate meanings-meanings specified by
their systematically structured relations to other symbols. Hence
practice implies system. But it is equally true that the system has
no existence apart from the succession of practices that instantiate,
reproduce. or-most interestingly-transform it. Hence system
implies practice." System and practice constitute an indissoluble
duality or dialectic: the important theoretical question is thus not
whether culture should be conceptualized as practice or as a system
of symbols and meanings, but how to conceptualize the articulation
of system and practice.
48 49William H. Scwcll, Jr.
TJlE AUTONOMY OF CULTURE
Let me begin this task by stating some assumptions about practice. I
assume that human practice, in all social contexts or institutional
spheres, is structured simultaneously both by meanings and by other
aspects of the environment in which they occur-by, for example,
power relations or spatiality or resource distributions. Culture is neither
a particular kind of practice nor practice that takes place in a
particular social location. It is, rather, the semiotic dimension of human
social practice in general. I further assume that these dimensions
of practice mutually shape and constrain each other but also
that they are relatively autonomous from each other.F
TIle autonomy of the cultural dimension of practice can also be
understood by thinking about culture as a system. TIle cultural dimension
of practice is autonomous from other dimensions of practice
in two senses. First, culture has a semiotic structuring principle
that is different from the political, economic, or geographical structuring
principles that also inform practice. Hence, even if an action
were almost entirely determined by, say, overwhelming disparities
in economic resources, those disparities would still have to be rendered
meaningful in action according to a semiotic logic-that is, in
language or in some other form of symbols. For example, an impoverished
worker facing the only manufacturer seeking laborers in that
district will have no choice but to accept the offer. Yetin accepting the
offer she or he is not simply submitting to the employer but entering
into a culturally defined relation as a wageworker. Second, the cultural
dimension is also autonomous in the sense that the meanings
that make it up-although influenced by the context in which they
are employed-are shaped and reshaped by a multitude of other
contexts. The meaning of a symbol always transcends any particular
context, because the symbol is freighted with its usages in a multitude
of other instances of social practice. Thus, our worker enters
into a relationship of "wageworker" that carries certain recognized
mcanings-of deference, but also of independence from the employer
and perhaps of solidarity with other wageworkers. These
meanings are carried over from the other contexts in which the
meaning of wage work is determined-not only from other instances
of hirings but from statutes, legal arguments, strikes, socialist
tracts, and economic treatises. They enter importantly into
The Concepits) ojCII!tII re
defining the local possibilities of action, in this case perhaps granting
the worker greater power to resist the employer than the local circumstances
alone would have dictated.
To understand fully the significance of this second sort of autonamy,
it is important to note that the network of semiotic relations that
make up culture is not isomorphic with the network of economic,
political, geographical, social, or demographic relations that make
up what we usually call a "society." A given symbol-mother, red,
polyester, liberty, wage labor, or dirt-is likely to show up not only
in many different locations in a particular institutional domain
(motherhood in millions of families) but in a variety of different institutional
domains as well (welfare mothers as a potent political
symbol, the mother tongue in linguistic quarrels, the Mother of God
in the Catholic Church). Culture may be thought of as a network of
semiotic relations cast across society, a network with a different
shape and different spatiality than institutional, or economic, or political
networks." The meaning of a symbol in a given context may
therefore be subject to redefinition by dynamics entirely foreign to
that institutional domain orspatial location: thus, for example, in the
1950S a particular political meaning of the symbol "red" became so
overpowering that the Cincinnati Reds baseball team felt the need to
change its name to "the Redlegs." This fact is what makes it possible-indeed
virtually guarantees-that the cultural dimension of
practice will have a certain autonomy from its other dimensions.
If culture has a distinct semiotic logic, then by implication it must
in some sense be coherent. But it is important not to exaggerate or
misspecify the coherence of symbol systems. I assume the COherence;
of a cultural system to be semiotic in a roughly Saussurian sense: that
is, that the meaning of a sign or symbol is a function of its network of
oppositions to or distinctions from other signs in the system. This
implies that users of culture will form a semiotic community-in the]
sense that they will recognize the same set of oppositions and therefore
be capable of engaging in mutually meaningful symbolic action.
To use the ubiquitous linguistic analogy, they will be capable of using
the "grammar" of the semiotic system to make understandable
"utterances."
It should be noted, however, that this conception actually implies
only a quite minimal cultural coherence-one might call it a thin coherence.
The fact that members of a semiotic community recognize a
51Willia1ll H. Sciccll, Jr.5°
given set of symbolic oppositions does not determine what sort of
statements or actions they will construct on the basis of their semiotic
competence. Nor does it mean that they form a community in
any fuller sense. They need not ilgree in their moral or emotional
evaluations of given symbols. The semiotic field they share may be
recognized and used by groups and individuals locked in fierce enmity
rather than bound by solidarity, or by people who feel relative
indifference toward each other. The posited existence of cultural coherence
says nothing about whether semiotic fields are big or small,
shallow or Jeep, encompassing?r specialized. It simply requires that
if meaning is to exist at all, there must be systematic relations among
( signs and a group of people who recognize those relations.
That this Saussurian conception implies only a thin cultural coherence
seems consonant with certain deconstructionist arguments.
The entire thrust of deconstruction has been to reveal the instability
of linguistic meaning. It has located this instability in the signifying
mechanism of language itself-s-claiming that because the meaning of
a linguistic sign always depends on a contrast with what the sign is
opposed to or different from, language is inevitably haunted by the
traces of the very terms it excludes. Consequently, the meaning of a
text or an uttcrance can never be fixed; attempts to secure meaning
can only defer, never exclude, a plethora of alternative or opposed
interpretations.
Cultural analysts who-like me-wish to argue that cultural systerns
are powerfully constraining have often drawn back from deconstructionist
arguments in horror. I think this is a major mistake;
indeed, I would maintain that a broadly deconstructionist understanding
of meaning is essential for anyone attempting to theorize
cultural change. Deconstruction does not deny the possibility of coherence.
Rather, it assumes that the coherence inherent in a system of
symbols is thin in the sense I have described: it demonstrates over
and over that what are taken as the certainties or truths of texts or
discourses are in fact disputable and unstable. This seems entirely
compatible with a practice perspective on culture. It assumes that
symbol systems have a (Saussurian) logic but that this logic is openended,
not closed. And it strongly implies that when a given symbol
S\'SIC111 is taken by its users to be unambiguous and highly constraining,
these qualities cannot be accounted for by their semiotic
qualil ios alone but must result from the WilY their semiotic structures
Tile Conceptis) o!Cllfture
are interlocked in practice with other structures-economic, political,
social, spatinl, and so on.>
Thus far in this section I have mainly been considering culture as
system. But what I have said has implications for how we might conceptualize
culture as practice. First, the conception of culture as
semiotic implies a particular notion of cultural practice. To engage in
cultural practice is to make use of a semiotic code to do something in
the world. People who are members of a semiotic community are capable
not only of recognizing statements made in a semiotic code (as
I have pointed out above) but of using the code as well, of putting it
into practice. To use a code means to attach abstractly available symbols
to concrete things or circumstances and thereby to posi t something
about them. I would also argue that to be able to use a code
means more than being able to apply it mechanically in stereotyped
situations-it also means having the ability to elaborate it, to modify
or adapt its rules to novel circumstances.
What things in the world areis never fully determined by the symbolic
net we throw over them-this also depends on their preexisting
physical characteristics, the spatial relations in which they occur,1 the relations of power with which they are invested, their economic
value, and, of course, the different symbolic meanings that may have
been attributed to them by other actors. The world is recalcitrant to
our predications of meaning. Hence, as Marshall Sahlins has pointed
out, every act of symbolic attribution puts the symbols at risk, ma~es
it possible that the meanings of the symbols will be inflected or transformed
by the uncertain consequences of practice. Usually, such at,
tributions result in only tiny inflections of the meaning of symbols.
But on some occasions-for example, when Hawaiian chiefs used
the category of tabu to enforce their monopoly on trade with Western
merchants-novel attributions can have the result of transforming
the meaning of a symbol in historically crucial ways."'
Part of what gives cultural practice its potency is the ability of
actors to play on the multiple meanings of symbols-thereby redefining
situations in ways that they believe will favor their purposes.
Creative cultural action commonly entails the purposeful or
spontaneous importation of meanings from one social location or
context to another. I have recently worked on a telling example of
the importation of meaning. The men and women who captured the
Bastille in July of 1789 were unquestionably characterizable as "the
52 53Wilfiillll H. Scit'cll, fl'.
people" in the common sense of "the mob" or the "urban poor." But
Parisian radicals and members of the french National Assembly
played on the ambiguity of the term to cast those who took the
Bastille also as a concrete instance of the abstract category of "the
people" who were said to be sovereign in radical political theory. Importing
the association between the people and sovereignty from the
context of political theory into that of urban crowd violence had the
not inconsequential effect of ushering the modern concept of revolution
into the world."
CULTURES AS DISTINCT WORLDS OF MEANING
Up to now, I have been considering culture only in its singular and
abstract sense-as a realm of social life defined in contrast to some
other noncultural realm or realms. My main points may be summarized
as follows: culture, I have argued, should be understood as a
dialectic of system and practice, as a dimension of social life autonomous
from other such dimensions both in its logic and in its spatial
configuration, and as a system of symbols possessing a(real but
thin cohercnce)h<lt is continually put at risk in practice and therefore
subject to transformation. Such a theorization, I maintain, makes it
possible to accept the cogency of recent critiques yet retain a workable
and powerful concept of culture that incorporates the achievements
of the cultural anthropology of the 1960s and 1970s.
But it is probably fair to SilY that most recent theoretical work all
culture, particularly in anthropology, is actually concerned primarily
with culture in its pluralizable and more concrete sense-that is,
with cultures as distinct worlds of meaning. Contemporary anthropological
critics' objections to the concept of culture as system and
their insistence on the primacy of practice are not, in my opinion, really
aimed at the concept of system as outlined above-the notion
that the meaning of symbols is determined by their network of relations
with other syrnbols.lRather, the critics' true target is the idea
~ that cultures (in the second, pluralizable sense) form neatly coherent
wholes: that they are logically consistent, highly integrated, consensu.rl.
extremely resistant to change, and clearly bounded. This is how
cultures tended to be represented in the classic ethnographiesf\lead
011 5'11lIOa, Benediet 011 the Zuni, Ma linowski on the 1"1'0rile
Conceptts) ofCulture
briands, Evans-Prichard on the Nuer, or, for that matter, Geertz on
the Balinese. But recent research and thinking about cultural practices,
even in relatively "simple" societies, has turned this classic
model on its head. Itnow appears that we should think of worlds of
meaning as normally being contradictory, loosely integrated, contested,
mutable, and highly permeable. Consequently the very concept
of cultures as coherent and distinct entities is widely disputed:.....J ~
Cultures are contradicioru. Some authors of classic ethnographies
were quite aware of the presence of contradictions in the cultures
they studied. Victor Turner, for example, demonstrated that
red symbolism in certain Ndembu rituals simultaneously signified
the contradictory principles of matrilineal fertility and male bloodletting.
But he emphasized how these potentially contradictory
meanings were brought together and harmonized in ritual performances.F
A current anthropological sensibility would probably emphasize
the fundamental character of the contradictions rather than
their situational resolution in the ritual. It is common for potent cultural
symbols to express contradictions as much as they express coherence.
One need look no farther than the central Christian symbol
of the Trinity, which attempts to unify in one symbolic figure three
sharply distinct and largely incompatible possibilities of Christian
religious experience: authoritative and hierarchical orthodoxy (the
Father), loving egalitarianism and grace (the Son), and ecstatic spontaneity
(the Holy Ghost). Cultural worlds are commonly beset with
internal contradictions.
Cultures are loosely integrated. Classic ethnographies recognized
that societies were composed of different spheres of activity-for
example, kinship, agriculture, hunting, warfare, and religion-and
that each of these component parts had its own specific cultural
forms. Out the classic ethnographers typically saw it as their task
to show how these culturally varied components fit into a wellintegrated
cultural whole. Most contemporary students of culture
would question this emphasis. They are more inclined to stress the
centrifugal cultural tendencies that arise from these disparate
spheres of activity, to stress the inequalities between those relegated
to different activities, and to see whatever "integration" occurs as
based on power or domination rather than on a common ethos. That
most anthropologists now work on complex, stratified, and highly
54 55William H. Sewell,Jr.
differentiated societies, rather than on the "simple" societies that
were the focus of most classic etlulOgraphies, probably enhances this
tendency.
\
Cultures are col/tested. Classic ethnographies commonly assumed,
at least implicitly, that a culture's most important beliefs were consensual,
agreed on by virtually all of a society's members. Contemporary
scholars, with their enhanced awareness of race, class, and
gender, would insist that people who occupy different positions
in a given social order will typically have quite different cultural beliefs
or will have quite different understandings of what might
seem on the surface to be identical beliefs. Consequently, current
scholarship is replete with depictions of "resistance" by subordinated
groups and individuals. Thus James Scott detects "hidden
transcripts" that form the underside of peasants' deference in contemporary
Malaysia and Marshall Sahlins points out that it was
Hawaiian women who most readily violated tabus when Captain
Cook's ships arrived-because the tabu system, which classified
them as profane (/1011) as against the sacred (tabu) men, "did not sit
upon Hawaiian women with the force it had for men." 33 Cultural
consensus, far from being the normal state of things, is a difficult
achievement; and when it does occur it is bound to hide suppressed
conflicts and disagreements.
Cultures are subiec! to constant c/tllllge. Cultural historians, who
work on complex and dynamic societies, have generally assumed
that cultures are quite changeable. But recent anthropological work
on relatively "simple" societies also finds them to be remarkably mutable.
For example, Renato Rosaldos study of remote Ilongot headhunters
in the highlands of Northern Luzon demonstrates that each
generation of Ilongots constructed its own logic of settlement patterns,
kinship alliance, and feuding-logics that gave successive
generations of Ilongots experiences that were probably as varied as
those of successive generations of Americans or Europeans between
the late nineteenth and late twentieth centuries."
Culturc« tlrc wcakly iniundcd. It is extremely unusual for societies j
or their cultural systems to be anything like isolated or sharply
bounded. Even the supposedly simplest societies have had relations
of trade, warfare, conquest, and borrowing of all sorts of cultural
items-technology, religious ideas, political and artistic forms, and
so Oil. But in addition to mutual influences of these sorts, there
Tile Conceptis)of Culture
have long been important social and cultural processes that transcend
societal boundaries-colonialism, missionary religions, interregional
trading associations and economic interdependencies, migratory
diasporas, and, in the current era, multinational corporations
and transnational nongovernmental organizations. Although these
transsocietal processes are certainly more prominent in more recent
history than previously, they are hardly entirely new. Think of the
spread of such "world religions" as Islam, Christianity, Hinduism, or
Buddhism across entire regions of the globe or the development of
extensive territorial empires in the ancient world. I would argue that
social science's once virtually unquestioned model of societies as
clearly bounded entities undergoing endogenous development is as
perverse for the study of culture as for the study of economic history
or political sociology. Systems of meaning do not correspond in any
neat way with national or societal boundaries-which themselves
are not nearly as neat as we sometimes imagine. Anything we might
designate as a "society" or a "nation" will contain, or fail to contain,
a multitude of overlapping and interpenetrating cultural systems,
most of them subsocietal, transsocietal, or both."
Thus all of the assumptions of the classic ethnographic model of
cultures-that cultures are logically consistent, highly integrated,
consensual, resistant to change, and clearly bounded-seem to be
untenable, This could lead to the conclusion that the notion of coherent
cultures is purely illusory; that cultural practice in a given
society is diffuse and decentered; that the local systems of meaning
found in a given population do not themselves form a 'higher-level,
societywide system of meanings. But such a conclusion would, in my
opinion, be hasty. Although I think it is an error simply to assume
that cultures possess an overall coherence or integration, neither can
such coherences be ruled out a priori.
HOW COHERENCE IS POSSIBLE
Recent work on cultural practice has tended to focus on acts of cul­
C'
'I.	 tural resistance, particularly on resistance of a decentered sort- ~
those dispersed everyday acts that thwart conventions, reverse valuations,
or express the dominated's resentment of their domination.w
But it is important to remember that much cultural practice is concentrated
in and around powerful institutional nodes-including
~.
56 57William H. Sewell, Jr.
religions, communications media, business corporations, and, most
spectacularly, states. These institutions, which tend to be relatively
large in scale, centralized, and wealthy, are all cultural actors; their
agents make continuous use of their considerable resources in efforts
to order meanings. Studies of culture need to pay at least as
much attention to such sites of concentrated cultural practice as to
the dispersed sites of resistance that currently predominate in the
literature.37
Even in powerful and would-be totalitarian states, centrally
placed actors are never able to establish anything approaching
cultural uniformity. In fact they rarely attempt to do so. The typical
cultural strategy of dominant actors and institutions is not so much
to establish uniformity as it is to organize difference. They are constantly
engaged in efforts not only to normalize or homogenize but
also to hierarchize, encapsulate, exclude, criminalize, hegemonize,
or marginalize practices and populations that diverge from the sanctioned
ideal. By such means, authoritative actors attempt, with varying
degrees of success, to impose a certain coherence onto the field of
cultural practice.?" Indeed, one of the major reasons for dissident
anthropologists' discomfort with the concept of culture is that it is
so often employed in all of these WClys by various powerful institutional
actors-sometimes, alas, with the help of anthropologists.
The kind of coherence produced by this process of organizing
difference may be far from the tight cultural integration depicted
in classic ethnographies. But when authoritative actors distinguish
between high and low cultural practices or between those of the
majority ethnicity and minorities or between the legal and the criminal
or between the normal and the abnormal, they bring widely varying
practices into semiotic relationship-that is, into definition in
terms of contrasts with one another. Authoritative cultural action,
launched from the centers of power, has the effect of turning what
otherwise might be a babble of cultural voices into a semiotically and
politically ordered field of differences. Such action creates a map of
the"culture" and its variants, one that tells people where they and
their practices fit in the official scheme of things.
The official cultural map may, of course, be criticized and resisted
The Conceptts) ojCulture
implies a recognition of their centrality. Dominant and oppositional
groups interact constantly, each undertaking its initiatives with the
other in mind. Even when they attempt to overcome or undermine
each other, they are mutually shaped by their dialectical dance.
Struggle and resistance, far from demonstrating that cultures lack
coherence, may paradoxically have the effect of simplifying and clarifying
the cultural field.
Moreover, dissenting or oppositional groups work to create and
sustain cultural coherence among their own adherents, and they do
so by many of the same strategies-hierarchization, encapsulation,
exclusion, and the like-that the authorities use. Once again, it is notable
that the concept of culture is as likely to be deployed politically
by dissident groups as by dominant institutions, and with many of
the same exclusionary, normalizing, and marginalizing effects as
when it is deployed by the state. To take an obvious example, dissident
nationalist and ethnic movements nearly always attempt to impose
standards of cultural purity on those deemed members of the
group and to use such standards to distinguish between those who
are and are not group members.
None of this, of course, implies that cultures are always, everywhere,
or unproblematically coherent. It suggests instead that coherence
is variable, contested, ever-changing, and incomplete. Cultural ~
coherence, to the extent tha t it exists, is as much the product of power
and struggles for power as it is of semiotic logic. But it is common for
the operation of power, both the efforts of central institutions and the
acts of organized resistance to such institutions, to subject potential
semiotic sprawl to a certain order: to prescribe (contested) core values,
to impose discipline on dissenters, to describe boundaries and
norms-in short, to give a certain focus to the production and consumption
of meaning. As cultural analysts we must aCknOWledge}
such coherences where they exist and set ourselves the task of explaining
how they are achieved, sustained, and dissolved.
It is no longer possible to assume that the world is divided up into
discrete "societies," each with its corresponding and well-integrated
"culture." I would argue forcefully for the value of the concept of
culture in its nonpluralizable sense, while the utility of the term as
bv those relegated to its margins. But subordinated groups must to
some degrec orient their local systems of meaning to those recognized
<IS dominant; the act of contesting dominant meanings itself
.~:
~-.:i :
pluralizable appears to me more open to legitimate question. Yet I
think that the latter concept of culture also gets at something we
need to retain: a sense of the particular shapes and consistencies of
./
59
58 William H. SC1PC!!, Jr.
I
worlds of meaning in different places and times and a sense that in
spite of conflicts and resistance, these worlds of meaning somehow
hang together. Whether we call these partially coherent landscapes
of meaning "cultures" or something else-worlds of meaning, or
ethnoscapes. or hegemonies-seems to me relatively unimportant
so long as we know that their boundedness is only relative and constantly
shifting. Our job as cultural analysts is to discern what the
shapes and consistencies of local meanings actually are and to determine
how, why, and to what;;te~they ~ng together.
NOTES
I have received valuable comments on this chapter from a number of friends
and colleagues. Even though I heeded their good advice only intermittently,
I would like to thank Anne Kane, David Laitin, Claudio Lornnitz, Sherry
Ortner, William Reddy, Marshall Sahlins, Paul Seeley, Ann Swidler, Lisa
\VedcC'n, members of the Social Theory Workshop at the University of Chicago,
Ihe audience at a Sociology Department Brown 13ilg at the University
of Arizona, and my fellow authors in this volume, who gave me critical comments
at our conference at the University of California at 13erkeley in
April 1996.
I. For a discerning aualysis of this debate, sec Robert Brighllll'lIl, "Forget
Culture: Replacement, Transcendence, Relexification," CuituvatAlllhropology
10 (1995): 5°9-40.
2. One outcome of these efforts was William H. Sewell, [r., Work and
RtTtllll/iml in Frmlcc: TI,c Lmlgllogc orLllh1l'/nllll III" Old Resilllt' I" 1848 (Cambridge,
1<)80).
3. This tum of historians to anthropolt1gy W,1S quite widespread in the
lQ70S, not only in Amcrrca, where it informed the work of such scholars as
Natalie Davis and Robert Darnton, but also in France, England, and Australia,
where antluopology influenced such scholars as Emmanuel Lc Roy
Ladurie, Jacques Le Goff, Keith Thomas, Peter Laslett, and Rhys Isaac.
4· Clifford Ceertz, Tilt' Intcrprciation of Cultures: Selected Essays (New
York, 1(73).
). Jacques Derrida, Of Gmllllllalology, trans. Cayatri Chakravorty Spivak
(Baltirnore, ]976).
6. A good introduction 10 this current of scholarship is II. Aram Veeser,
ed., Tilt' New Hisit,,-ic;slH (New York, 198<)).
7. Lynn Hunt, ed., TTl" New Cultural His/my (Berkeley, 1989).
8. The most celebrated expression of this angst is the collective volume
edited bv James Clifford and George E. Marcus, Wrilillg Culture: Tuc Poetics
'md pt1ri/ics llf [117110<"1'111"1.11 (Berkeley, 1980).
q. Geertz, Jllfcq,rcla/itl/f ofCllltllrcs,andLila Abu-Lughocl, "Writing aguinst
Tire Concepus) OfCulture
Culture." in RecapturingA/llhropology: Workillg;1/ the Present, ed. Richard C.
Fox (Santa Fe, 1991), pp. 137- 62.
10. Brightman, "Forget Culture," p. 510.
11. James Clifford, Tile Predicament of Culture: Twcntietll-Celltllry Ethnography,
Literature,and Art (Cambridge, Mass., 1988), P: 10.
12. Raymond Williams, Keywords: A Vocabulary ofCultureand Society, rev.
ed. (London, 1983), p. 87. See also Raymond Williams, Culture and Sociely:
1780-1950 (New York, 1958).
13. The two types of meanings I have distinguished here can be overlaid,
so that the cultural aspects of the life of a people or a social group are distinguished
from the noncultural aspects of its life. Hence, "Balinese culture"
.,
.. may be contrasted to "Balinese society" or "the Balinese economy." In anthropological
usage, however, "culture" also is commonly used to designate
the whole of the social life of a given people, so that "Balinese culture" becomes
a synonym for "Balinese society" rather than a contrastive term.
14. Ruth Benedict's views are expressed most systematically in Patterns
of Cllltllre (Boston, 1934). Levi-Strauss's views are expressed in many of
his works, including Structural Anthropology, trans. Clair Jacobson and
Brooke Crundfest Schoefp. 2 vols. (New York, 1963), and The Savage Mind
(Chicago, 1966).
15. Williams, Keywords, pp. 90-91.
16. E. P. Thompson, The MI/killg of tire English Workillg Class (London,
963); Paul Willis, LearningtoLabor: HowWorkillgClass KidsGetWorkingClass1

Jobs (New York, 1 981).

17. Clifford Ceertz, "Religion as a Cultural System" and "Ideology as a
Cultural System," in Interpretation of Cultures, pp. 87-125, 193- 233; "Common
Sense as a Cultural System" and "Art as a Cultural System," in Local
KllOwledge: Further Essays ill Interpretioe Anthropology (New York, 1 983), pp.
73-93, 94 -120.
18. David M. Schneider's most influential book is American Kinship: A
Cultural Accoullt (Englewood Cliffs, N.J., 1968). The most systematic statement
of his conception of a cultural system is David M. Schneider, "Notes
toward a Theory of Culture," in Meaningin Anthropology,ed. Keith H. Basso
and Henry A. Selby (Albuquerque, 1976), pp. 197- 22 0 .
1 9. Talcott Parsons, The Social System (Glencoe, Ill., 1959)' Geertz and
Schneider were both students of Talcott Parsons and Clyde Kluckhohn in
the Harvard Department of Social Relations and they taught together during
the 1960s at the University of Chicago, then the epicenter of cultural
anthropology.
20. Victor W. Turner, Tile Forest of Symbols: Aspects of Ndembu Ritual
(Ithaca, N.Y., 1967), TheRitual Process: Structure and Allti-Structure (Chicago,
1 969), and Reoelaiion and Ditnnationill Ndembu Ritual (Ithaca, N.Y.,1975).
21. Sherry Ortner, "Theory in Anthropology Since the Sixties," Comparative
Sludies ill History and Society 26 (1984): 126-66; Pierre Bourdieu, Outline
of a Theory of Practice, trans. Richard Nice (Cambridge, 1977); Clifford and
Marcus, Writing Culture.
60 61Wil1iall/ H. Se1l'(I1, Jr.
22. Joan W. Scott has been at rho center of much of Ihis polemic. "On Language,
Cender, and Working-Class I lis lory," lntcrualional luuor 11 lid Workillg
C/I1SS History, 110. :'11 (spring 1987): I-1J, was a response 10 Gareth Stedman
Jones, LllIlglll1ges 4 Class: SfIIdies ill Ellglisll Workillg Cll1sS Hislory, 18)2 -1982
(Cambridge. 1984). This essay was published with responses in the same issue
by Bryan D. Palmer (14 -21), Christine Stansell (24 -29), and Anson Rabinbach
(30- ]6); Scott then responded to their critiques in no. ]2 ([fall 1987]:
39-45). Scott criticizes John Toews, "Intellectual History after the Linguistic
Turn: The Autonomy of Meaning and Ihe lrreducibility of Experience,"
American Historical Recieio 92 (1987): 879-907, in "Experience," in Feminists
Theorize the Political, ed. Joan W. Scott and Judith Buller (New York, 1992),
pp. 22- 40. Laura Lee DOWl1s criticizes Scolt in "If 'Woman' Is Just an Empty
Category, Then Why Am I Afraid to Walk Alone at Night? Identity Politics
Meets the Postmodern Subject," Compomtii-c Studies ill Societyand Hision; J5
(1993): 4l4-J7; Scott's response, "The Tip of the Volcano" (4]8-4J), and
Downs's reply (,l44 -51) are published in the same volume. Perhaps the most
blistering dcnunciauon of discursive history is Brian D. Palmer, Descent into
Discourse: The Rciiication of Lallgllasraud lite WrilillX t~{ Social Hislory (Philadelphia,
19(0). For a discussion of l11ilny of these issues, see Joyce Appleby,
Lynn Hunt, and Margaret Jacob, Tel/illS ti,e Tl'/lt/I abou! Hislory (New York,
1994). pp. 19!'l-2J7·
2J. Ann Swidler, "Culture in Action: Symbols and Strategies." American
So(iolog;ml Rel'ieil' 51 (1984): 273-86.
24. Schneider, "Noles toward a Theory of CuItUI'('," 214.
25. See, e.g., Schneider, American Kiushu). _,
26. Readers of Marshall Sahlins should find this formulation familiar. See
especially 1511/11.15 o{llislol'll (Chicago, 19R5), pp. 1]6-56.
27. For a fuller exposition of this perspective, see William I-1. Sewell, [r.,
"Toward a I'ost-Materialist Rhetoric for Labor I listory," in Retllil/kil/S Labor
History: Essl1ys 0/1 Discoutsc and Class Analusi«, cd. Lenard R. Berlanstein (Urbana,
111., 1(93), IS -:\8.
28. On the spatial aspect of culture, see Claudio Lomnitz-Adler, "Concepts
for the Study of Regional Culture," American Et/11I0logist 18 (1991):
1Q5-214·
29. This is not, of course, the usual conclusion arrived at by deconstructionists.
who would insist that these "other structures" are no less textual
than semiotic structures and that making sense of them is purely a matter of
intertextuality. This epistemological and perhaps ontological difference between
m v position and that of deconstruction should make it clear that I am
ilppwpriating from deconstruction specific ideas that I find useful rather
than adopting a full-scale deconstructionist position.
_10. Marshall Sahlins, Ilisloricl/l Mclo"'lor~ aud lI·hllhic(/II\('olili/'~ (Ann Arb
or, 1(81), esp, pp, 67-72, and Islands o(l{;slol'tl, pp. \]6-56.
31. \Villiam H. Sewell, [r.. "Political Events as Transformations of Sl ruclures:
lnvcntmg Revolution at the Bastille," "'ht'o/~11 1/11.1 SociclP 26 (19'16):
84 1- 81.
Tlte Concepiis)of Culture
]2. Turner, Forcst ofSym/lols, PP' 41-4J·
33. Sahlins, Historical MelapllOrs, p. 46. See also James Scott, Weapons of lite
Weak: EverydayForms of Peasant Resistance (New Haven, 1985).
34. Renata I. Rosaldo, l/oJlgol HendIJlIII/;llg, 1883-1974: A Siudy iJlSociety
andHislory (Stanford, 1980).
J5. Arjun Appadurars work on recent forms of transnational cultural
forms has been particularly influential. See, e.g., "Global Ethnoscapes: Notes
and Queries for a Transnational Anthropology," in Fox, ed., Recapturing AnIhropology,
PP' 191-210, and Modemi'y at Large: Cultural D;/IICIISiOll5 of Globalization
(Minneapolis, 1996).
36. For a critical discussion of such work, see Sherry B. Ortner, "Resistance
and the Problem of Ethnographic Refusal," Comparaiiue Sit/diesill SodetyaJld
H;story 37 (1995): 173-93·
37. For a fascinating study of state cultural practices, see Lisa Wedeen,
Ambiguilies0fDolllillatiol/: Politics, Rhetoric, audSYlllbols ill COl/telllporary Syria
(Chicago, forthcoming, 1999)·
38. 'I11is characterization seems to me to be roughly consonant with a
Gramscian idea of hegemony; see Antonio Crarnsci, Selections[rom thePrison
No/ebook~, ed. and trans, Quintin Hoare and Geoffrey Nowell Smith (New
York, 1971). For two quill' different Gramscian cultural analyses of politics,
see Stuart Hall, The HardRoad to Rcnctoal: Thotclicrism and tile Crisis of the Lefl
(London, 1988), and David D. Laitin, Hcgel1lol/Y and Culture: Politics and Religiolls
Gral/gcal/lOI/S tile Yorllbtl (Chicago, 1986).
n.