La reproduccin digital de este material es para fines de investigacin y docencia de los cursos

acadmicos que im parte El Colegio de Michoacn (COLM1CH), conforme a lo establecido en:

Lev Federal de Derechos de Autor. Ttulo VI De ias Limitaciones de Derecho de Autor y de ios
Derechos Conexos, Captulo II De ia Limitacin a ios Derechos Patrimoniales, Artculo 148
Apartado III:

Reproduccin de partes de ia obra, para ia crtica e investigacin cientfica, literaria o artstica.

ANTHROPOLOGIES
AND

HISTORIES
Essays in Culture, History,
and Political Economy

William Roseberry

RUTGERS
NEW

U N I V E R S IT Y PRESS

BRUNSWICK

AND

LONDON

BIBLIOTECA

LWS

GONZA

O N E

Balinese Cockfights
and the Seduction
of Anthropology
Few anthropologists in recent years have enjoyed wider in
fluence in the social sciences than Clifford Geertz. Sociologists,
political scientists, and social historians interested in p op ular cul
tu re an d mentalits have tu rn e d increasingly to anthropology, and
the anthropologist most often em braced is Professor Geertz.
A n u m b e r o f factors can be addu ced to account for this
trend. In the first place, G eertzs position at the Institute for
Advanced Study has allowed him to transcend the disciplinary
and subdisciplinary involution that characterizes anthropology
and o th e r social sciences. At the Institute, he is able to attract
scholars from a variety o f disciplines, adopting an antidisciplinary m ood an d focus that is rare in cu rren t academic prac
tice. Second, Geertz is an excellent e th n o g rap h er who writes
with an eloquence and sophistication uncom m on for the social
sciences. His cultural essays can be read with profit by introd uc
tory students o r g rad u ate students in advanced seminars. A nd
his descriptions o f life in Bali or Java or Morocco call to mind
one o f the aspects o f anthropology that has always been so
seductive: the lure o f distant places an d oth er modes o f being.
Thus, in part, the title o f this essay. But the title is intended to
suggest an o th e r aspect o f G eertzs work as well, for there is a
sense in which anthropologists and other social scientists
have been seduced by G eertzs writings on culture.

concentrating on symbols that carry and com m unicate m e an

ings to social actors who have created them. Unfortunately, at
no point does he say w hat he m eans as clearly and rigorously as
does H arris. Instead, he places his definitions in a m ore elegant
an d elusive prose. For exam ple: Believing, with Max Weber,
that m an is an animal suspended in webs o f significance he
him self has spun, I take culture to be those webs . . . (1973b:
5). Or: culture consists o f socially established structures o f
m eaning in term s o f which people do such things as signal
conspiracies an d join them o r perceive insults and answer
them . . . (ibid.: 13). Or: T h e culture o f a people is an ensem
ble o f texts, themselves ensembles, which the anthropologist
strains to read over the shoulders o f those to whom they p ro p
erly belong (1973c: 452). T h e last quote comes from the wellknown essay, D eep Play: Notes on the Balinese Cockfight, to
which m ore attention is devoted here. It was noted earlier that
G eertz seems to be w orking with a concept o f culture as socially
constituted an d socially constituting. We m ust now question
w hether he has realized this promise. This essay com pares
G eertzs claims fo r him self in Thick Description with one o f
his own pieces o f description. Because Geertzs ethnographic
work is volum inous, an d the aims o f this chapter are modest, we
shall concentrate on his essay on Balinese cockfights.1

G eertzs essay is at once an attem pt to show that cultural

products can be treated as texts and an attem pt to interp ret
one such text. T h e m e tap h o r of the text is, o f course, a favor
ite o f the practitioners o f both structuralism and h e rm e n e u
tics, th o u g h G eertz takes his lead from Ricoeur rath er than
Lvi-Strauss. T h e reference to culture as a text, given G eertzs
project, calls fo r an exercise in interpretation. Geertz in te rp re
tation m ust be sum m arized before we can ask some questions
o f it. Notes on the Balinese Cockfight begins with an account
o f the G eertzes difficulties when first arriving in the field,
their response to a police raid on a cockfight, and their final
acceptance, given that response, by the villagers. T h e essay
then moves into a description o f the cockfight itself, including

a discussion o f the psychological identification o f m en and

cocks, the pro ced ures associated with cockfights and wagers,
and so on. Prelim inaries ou t o f the way, Geertz moves toward
an in terp retatio n o f the fight itself. H e begins with Jerem y
B en th am s notion o f d eep play, o r games in which the conse
quences for losers are so devastating that participation in the
games is irrational for all concerned. Noting that the central
wagers in Balinese cockfights seem to correspond to such a
high stakes game, h e then counters:
It is in large part because the marginal disutility o f loss is so
great at the higher levels o f betting that to engage in such
betting is to lay o n e s public self, allusively and metaphorically,
through the m ed iu m o f o n e s cock, on the line. A nd though to
a B entham ite this m ight seem merely to increase the irrational
ity o f the enterprise that m uch further, to the Balinese what it
mainly increases is the m eaningfulness o f it all. And as (to
follow Weber rather than Bentham ) the imposition o f m ean
ing in life is the major and primary condition o f hum an exis
tence, that access o f significance m ore than compensates for
the econom ic costs involved. (1973c: 434)

Geertz th e n looks to two aspects o f significance in the cock

fight. Both are related to the hierarchical organization o f Bali
nese society. H e first observes that the cockfight is a simulation
o f the social m atrix , or, following Goffman, a status blood
b ath (ibid.: 436). To explore this, Geertz m entions the fo ur d e
scent groups that organize factions in the village an d examines
the rules involved in betting against the cocks owned by m em
bers o f o th e r descent groups, o th e r villages, rivals, and so on.
A lthough he has not yet re fe rre d to the cockfight as a text, as
Geertz moves tow ard the second aspect of significance, he be
gins to re fe r to it as an art fo rm (ibid.: 443). As an art form, it
displays fu n d am en tal passions in Balinese society that are hid
den from view in o rd inary daily life and com portm ent. As an
atomistic inversion o f the way Balinese normally present th e m
selves to themselves, the cockfight relates to the status hierarchy
in an o th er sense no longer as a status-based organization o f

the cockfight bu t as a com m entary on the existence o f status

differences in the first place. T h e cockfight is a Balinese re a d
ing o f Balinese experience, a story they tell themselves about
themselves (ibid.: 448). W hat they tell themselves is that be
neath the external veneer o f collective calm an d grace lies a n
o th e r nature. A t both the social and individual level, there is
an o th er Bali an d a n o th e r sort o f Balinese. A nd w hat they tell
themselves they tell in a text that consists o f a chicken hacking
an o th er mindlessly to bits (ibid.: 449).
A fter this basic interp retatio n o f the Balinese cockfight in
terms o f status organization an d commentary, Geertz closes
with a discussion o f culture as an ensem ble o f texts. H e notes
that their in terp reta tio n is difficult and that such an app roach is
not the only way that symbolic form s can be sociologically h a n
dled. Functionalism lives, an d so does psychologism. But to
reg ard such form s as saying som ething o f som ething, an d say
ing it to somebody, is at least to o pen up the possibility o f an
analysis th a t attends to their substance rath er than to reductive
form ulas professing to account for th e m (ibid.: 453).
A ccepting this criticism o f reductive formulas, we m ust ques
tion w h eth er G eertzs analysis has sociologically h andled the
Balinese cockfight o r paid sufficient attention to its substance.
In w hat follows, no fu n d am en tal reinterp retatio n o f the Bali
nese cockfight is attem pted . Such a reinterp retation is the task
o f a w riter m ore fam iliar with Bali and Indonesia than is the
present one. This essay simply points to a few elem ents present
in G eertzs essay b u t om itted from the interpretive exercise that
should fo rm a p a rt o f a cultural an d sociological interpretation
o f the cockfight. A lth ou gh Geertz m ight reg ard reference to
these elem ents as a form o f functionalist reductionism , no at
tem pt is m ade h e re to account for or explain the existence o f
the cockfight. Rather, by pointing to o th e r aspects o f Balinese
society a n d history with which the cockfight may be involved,
this essay calls into question the m e tap h o r o f culture as text (cf.
Keesing 1987).
Accepting for a m o m ent that m etaphor, we m ight briefly
tu rn to th ree aspects o f Balinese society not included in the
interpretation. T h e first has to do with the role o f women. In a

footnote early in the article, Geertz notes that while th ere is

little a p p a re n t public sexual differentiation in Bali, the cock
fight is one o f the few activities from which women are ex
cluded (1973c: 417418). This a p p a re n t anomaly may make
sense in term s o f G eertzs interpretation. As with status differ
ences, so with sexual differences. T h e cockfight, and betting on
the cockfight, are the activities o f m en, serving as com m entaries
on the public denial o f difference. But sex cannot be subsum ed
so simply within status. T h e sexual exclusion becomes m ore
interesting w hen we learn in an o th er footnote that the Balinese
countryside was in teg rated by rotating m arket systems that
would encom pass several villages an d that cockfights were held
on m arket days n e a r the m arkets and were sometimes o rg a
nized by petty m erchants. T rade has followed the cock for
centuries in ru ral Bali, an d the sport has been one o f the main
agencies o f the islands m onetization (ibid.: 432). F urtherm ore,
in yet a n o th e r footnote in his m ore recent Negara, Geertz tells
us that the traditional m arkets, which were staffed almost e n
tirely by w om en, were held in the m orning, and that the cock
fights were held on the same afternoo n as the m arket (1980:
199).
Aside from sexual differentiation and the connection with
markets, G eertz also notes th ro u g h o u t the early p art o f the
essay (1973c: 414, 418, 424, 425) that the cockfight was an
im p o rtan t activity in precolonial Balinese states (that is, before
the early tw entieth century), that it was held in a ring in the
center o f the village, that it was taxed and was a significant
source o f public rev en u e .2 Further, we learn that the cockfight
was outlaw ed by the D utch and later by Indonesia, that it is now
held in semisecret in h id d e n corners o f the village, and that the
Balinese re g a rd th e island as taking the shape o f a small,
p ro u d cock, poised, neck extended, back taut, tail raised, in
eternal challenge to large, feckless shapeless Java (ibid.: 418).
Surely these m atters req u ire some interpretive attention. At the
very least they suggest th at the cockfight is intimately related
(though no t reducible) to political processes o f state form ation
and colonialism. T h ey also suggest that the cockfight has gone
th ro u g h a significant change in the past eighty years, that if it is

a text, it is a text that is being written as p art o f a p ro fo u n d

social, political, an d cultural process.
This, finally, brings us to the third point, which is less an
aspect om itted from the in terp retation th an one th at is not
sufficiently explicated. Geertz refers to the cockfight as a status
bloodbath an d tells us that as a com m entary on status, the
cockfight tells the Balinese that such differences are a m atter
o f life an d d e a th an d a p rofoundly serious business (ibid.:
447). Yet, in this essay at least, we learn very little about caste
and status as m aterial social process and the connection that
process does or does not have with cockfighting. In Negara,
Geertz tu rns his attention to elaborate crem ation ceremonies
and sees them as an aggressive assertion o f status (1980: 117).
C om parable in spirit to the potlatch, the crem ation is conspicu
ous consum ption, Balinese style (ibid.: 117) an d is one o f vari
ous rituals th at elaborately tell the Balinese that status is all
(ibid.: 102). In this case, we are dealing in p art with political
com petition am o ng high-caste lords an d princes. B ut lords are
also com m unicating to their com m oners that the hierarchy is
divinely ordained. Status in Bali has to do with inherited caste
but also with positions achieved in life th ro u g h various form s o f
political m an eu v er most clearly am ong lords but also am ong
low-caste Sudras. With so m uch maneuver, and with so many
cultural texts relating to status, some attention should be paid
to the d iffere n t messages o f these texts an d to their construc
tion in the context o f status form ation as a historical process.
T hese th re e problem s lead to a basic point. T h e cockfight
has gone th ro u g h a process o f creation that cannot be sepa
rated from Balinese history. H ere we confront the m ajor inade
quacy o f the text as a m e tap h o r for culture. A text is written; it
is not w riting.3 To see culture as an ensem ble o f texts o r an art
form is to rem ove culture from the process o f its creation.4 I f
culture is a text, it is not everyones text. Beyond the obvious
fact th at it m eans d iffere n t things to d ifferent people or d iffer
ent sorts o f people, we m ust ask who is (or are) doing the
writing. Or, to break with the m etaphor, who is doing the
acting, the creating o f the cultural forms we interpret. This is
a key question, fo r exam ple, in the transform ation o f the cock

fight after the arrival o f the Dutch. In a recent essay, Geertz

has pointed to the separation o f the text from its creation as
one o f the strengths o f the metaphor. R eferring to Ricoeurs
notion o f inscription, o r the separation in the text o f the said
from the saying, G eertz concludes:
T h e great virtue o f the extension o f the notion o f text beyond
things written on paper or carved into stone is that it trains
attention on precisely this phenom enon: on how the inscrip
tion o f action is brought about, what its vehicles are and how
they work, and on what the fixation o f m eaning from the flow
o f events history from what happened, thought from think
ing, culture from behavior implies for sociological interpre
tation. (1983: 31)

T h e re a d e r should not assume that I am calling for the red u c

tion o f culture to action (see C h ap ter 2). Geertz correctly points
to m eanings that persist beyond events, symbols that outlast and
transcend the intentions o f their creators. But neither should
culture be separated from action; otherwise we are caught in
yet an o th er o f anthro po log ys antinomies. Unfortunately, the
text as m e ta p h o r effects precisely this separation.

T h e em phasis on cultural creation brings out two aspects of

culture that are missing from G eertzs work. T h e first is the
presence o f social an d cultural differentiation, even within an
apparently u n ifo rm text. Reference to differentiation is, in
part, reference to the connections between culture and rela
tions o f pow er a n d dom ination, as implied in the previous com
ments on state an d status. Some m ight think that to refer to
culture and pow er is to reduce culture to power, to treat values
as glosses on p ro p erty relations (Geertz 1973c: 449) or to ru n
on about the exploitation o f the masses (1973b: 22). But there
are reductions, an d then there are reductions. A nd the denial
of such connections is but one o f many classical reductions in
Am erican anthropology. T h e second aspect that is missing is a
concept o f culture as m aterial social process. W ithout a sense of

culture as m aterial process or creationas writing as well as

what is w ritten we once again have a conception o f culture as
p ro d u ct b u t no t as p ro d u ctio n .5 T h e reference to culture as
m aterial social process is not intend ed to take us back to the
anthropological m aterialism o f Marvin Harris. Ind eed , the criti
cism I have directed at Clifford Geertz is similar to the criticism
I directed at M arvin H arris: both treat culture as p rod uct but
not as production. T h e re the similarity ends, of course. But
both have rem oved culture from the process o f cultural cre
ation a n d have th e refo re m ade possible the constant re p ro d u c
tion o f an antinom y between the material an d the ideal.
T h e resolution o f the antinomy, and the concept o f culture
that em erges fro m that resolution, m ust be materialist. B ut the
materialism invoked in this essay is far rem oved from the re d u c
tive scientism th a t has come to dom inate materialism in A m eri
can anthropology. Rather, what is n eed ed is som ething close to
the cultural m aterialism o f Raym ond Williams (1977; cf.
1980; 1982), wtfo notes th at the problem with mechanical m ate
rialism is n o t th a t it is too materialist b u t th at it is not materialist
enough. It treats culture an d o th e r aspects o f a p resum ed su
p e rstru c tu re simply as ideas. It th erefo re makes room for,
indeed requires, idealist critiques th at share the ideational defi
nition b u t deny the m aterial connection or, as in the case o f
Geertz, th a t reject the ideational definition in favor o f one that
sees a socially constructed text that is, nonetheless, rem oved
from th e social process by which the text is created. In contrast,
Williams suggests that cultural creation is itself a form o f m ate
rial production, th at the abstract distinction between material
base an d ideal su p erstru ctu re dissolves in the face o f a material
social process th ro u g h which both m aterial an d ideal are
constantly created an d recreated.
Yet Williams does not leave his analysis at this elem entary
assertion. H e also pays attention to the socially constructed
m eanings th a t inform action. H e does this in part by means o f a
revaluation o f the idea o f tradition, defining it as a reflection
u p o n an d selection from a peoples history (1961; 1977). T h e
process o f selection is political and is tied to relations o f d om in a
tion and subordination, so that Williams can talk o f a dom in ant

culture, o r hegem ony, as a selective tradition. A lthough this

do m in an t cu ltu re is related to an d supports an o rd e r o f inequal
ity, Williams does no t view it simply as a ruling-class ideology
im posed u p o n the dom inated. Rather, as a selection from and
interp retatio n o f a peo p les history, it touches aspects o f the
lived reality o r experience o f the do m inant and dom inated
alike. It is, in sho rt an d in part, m eaningful. B ut Williams also
notes th at no o rd e r o f dom ination is total. T h e re are always
relationships a n d m eanings th at are excluded. T herefo re, alter
native m eanings, alternative values, alternative versions o f a
peoples history are available as a potential challenge to the
dom inant. W h eth er such alternative versions are constructed
dep ends u p o n the n a tu re o f the cultural and historical material
available, th e process o f class form ation and division, and the
possibilities an d obstacles presen ted in the political process. Wil
liamss concept o f culture, then, is tied to a process o f class
form ation b u t is no t red u ced to that process. D om inant and
em erg en t cultures are fo rm ed in a class-based social world, but
they are no t necessarily co n g ru en t with class divisions.
T h e them es o f culture as m aterial social process and o f cul
tural creation as (in part) political action are fu rth e r developed in
an article by Peter Taylor and H e rm a n n Rebel (1981; cf. Rebel
1988). In a m asterful analysis o f culture in history, the authors
concentrate on fo u r texts fo u r o f the G rim m s folk tales that
deal with com m on them es o f inheritance, disinheritance, family
dissolution, a n d m igration. A fter criticizing psychological in ter
pretations, they place the tales in the late-eighteenth- and earlynineteen th-century context in which they were collected. They
then take two innovative methodological steps that are o f great
im portance fo r the concept o f culture. First, they ask who is
telling the tales a n d in w hat context. T hey also note that while the
tales are traditional, they are not timeless; that is, the form and
content o f the tales may change in the telling. T h e question o f
who is telling the tales a n d in what context therefore becomes
im portant. Taking a fo rm o f culture as a text, the authors take
the first step tow ard an analysis o f text as writing, as material
social process. Second, they assume that the peasant women who
are telling the tales fo rm a peasant intelligentsia that is trying to

intervene in the social process. T h a t is, the tales are com m entar
ies on w hat is h a p p en in g to them an d their families that call for
particular form s o f action to alter th e situation. This is a crucial
methodological step in the construction o f a concept o f culture
not simply as a p ro d u ct b u t also as production, not simply as
socially constituted but also as socially constituting. Given this
fram ework, the authors then em bark on a detailed symbolic
analysis o f the tales and, finally, suggest that the tales were at
tem pts by peasant w om en to resp o n d to the disruption o f fam i
lies and the d raftin g o f their disinherited sons. T h e suggested
response: inheriting dau g h ters should reno unce their in heri
tance, move from the region, m arry elsewhere, and offer a re f
uge for their fleeing brothers. Taylor and Rebel show that such a
response is in accord with dem ograp hic evidence from lateeighteenth-century Hesse, although it cannot yet be d em o n
strated w heth er the process they suggest actually occurred.
Nonetheless, the autho rs have pro d u ced a cultural analysis that
goes significantly fu rth e r th an does G eertzs in his Notes on the
Balinese Cockfight. To ask o f any cultural text, be it a cockfight
or a folk tale, who is talking, who is being talked to, what is being
talked about, an d w hat fo rm o f action is being called for, is to
move cultural analysis to a new level that rend ers the old antino
mies o f m aterialism an d idealism irrelevant.6
It m ight be arg u ed that this is precisely what Geertz does. As
one o f o u r most able eth n o g rap h ers, he is one o f the few a n th ro
pologists who can provide detailed ecological, economic, and
political inform ation at the same time that he engages in sophisti
cated symbolic analysis. His exam ination o f the th eater state in
nineteenth-century Bali is an exam ple o f this: we find treatm ents
o f political a n d social structure at ham let, irrigation system, and
tem ple levels, o f caste divisions, o f trade, and of the rituals of
hierarchy. T h a t G eertz sees all o f these as necessary for a cultural
argum ent, an d th at he sees his inclusion o f these elements as
ren d erin g an idealist charge absurd, is clear from his conclu
sion to Negara. A lthough all the elem ents are presented and
connected in a fashion, they are never fully joined. C ulture as
text is rem oved fro m the historical process that shapes it an d that
it in tu rn shapes. W hen we are told that in Bali culture came

from the top dow n . . . while pow er welled u p from the bottom
(1980: 85), the im age makes perfect sense given the analysis of
state structure that precedes it. B ut the image implies separation,
a rem oval o f culture from the wellings-up o f action, interaction,
power, an d praxis.
We re tu rn , then, to the com parison of G eertzs promise with
his practice. A lthough this essay already contains m ore q u ota
tions than it can easily bear, it closes with yet another. T h e
quotation re tu rn s us to the prom ising approach to culture ex
pressed in T hick D escription, an d it is a statem ent o f connec
tion ra th e r th an separation. T h e passage establishes a standard
for cultural in terp retatio n that is in accord with the premises of
this essay. T h a t it also serves as a standard in term s of which
G eertzs cultural analysis can be criticized should be apparent.
I f anthropological interpretation is constructing a reading o f
what happens, then to divorce it from what happens from
what, in this time or that place, specific people say, what they
do, what is d o n e to them , from the whole vast business o f the
world is to divorce it from its applications and render it
vacant. A g o o d interpretation o f anything a poem , a person,
a history, a ritual, an institution, a society takes us into the
heart o f that o f which it is an interpretation. W hen it does not
do that, but leads us instead som ew here else into an admira
tion o f its ow n elegance, o f its authors cleverness, or o f the
beauties o f Euclidean order it may have its intrinsic charms;
but it is som ething else than what the task at hand . . . calls for.
(1973b: 18)

In te rp re tatio n canno t be separated from what people say, what

they do, w hat is do n e to them , because culture cannot be so
separated. As long as anthropologists are seduced by the intrin
sic charm s o f a textual analysis th at takes such separation as a
point o f honor, they will continue to do som ething other than
what the task at h a n d calls for.