TEN

Spatial Practices:
Fieldwork, Travel, and the Disciplining
of Anthropology
James Clifford
For George Slocking

The day after the Los Angeles earthquake of 1994, I \'latched a TV intervie\vwith an earth scientist. He said he had been "in the field" that morning
looking for new fault lines. It was only after a minute or so of talk that I realized he had been flying around in a helicopter the whole tin1e. c:ould this

be fieldwork? I was intrigued by his invocation of the field, and so1nchow

unsatisfied.
My dictionary begins its long list of definitions for "field" with one about
open spaces and another that specifics cleared space. The eye is 11ni1npcded,
free to roam. In anthropology Marcel (;riaule pioneered the use of aerial
photography, a method continued, now and again, by others. But if o\crvie\v,
real or imagined, has long been part of fieldwork, there still see1ned In he
an oxymoronic bump in the earth scientist's airborne "field." P<trlicul;1rly in
geology, indeed in all the sciences that value ficlchvork, the pr.1cticc or research "on the ground," observing 1ninute particulars, has been a sine qua
non. The French analogue, trrrain, is unequi\.'ocal. Gentlcn1cn-natur-alists
were supposed to have muddy boots. Fieldwork is Carthbound-inti1natcly
involved in the natural and social landscape.
It was not always so. Henrika Kuklick (chapter~) rc1ninds us that the n1ove
toward professional field research in a range of disciplines, including anthropology, took place at a particular historical n101nenl in the late nineteenth century. A presumption in favor of professional work that was downclose, e1npirical, and interactive was quickly naturali1e<l. Ficld\.\'ork \Vould
put theory to the test; it would ground interpretation.
In this context, flying around in a helicopter seemed a bit absrract. Yet,
on reflection, I had to allow the earth scientist his praclice of going "in10
the field" while never setting foot there. In so111e crucial way, his use of the
term qualified. What mattered was not si1nply the acquisition of fresh e1n-

JAMES CLI FFC>RD

pirical data. A satellite photo could provide that. What ma~e tl~is field:-ork
rhe act of physically going out into a c/,eared place of work. Going out preM
suppos("S a spatial distinction between a ho1ne base and an exterior p~ace of
discovery. A clea1ed space of work assunies that one can keep out d1stractiuv, influences. A ftcld, hy definition, is not overgrown. The earth scientist
1ould 1101 have done his helicopter "fieldwork" on a foggy clay. An archaeologist t:annot excavate a site properly if it is inhabited or built over. An an1111 opolog-ist 1nay frel it necessary to clear his or her field, at least conceprnallY of tourists, 1nissionaries, or govcrn1nent troops. Going out into a
4-1 1; 11'.;.d J>l<tcc of work presupp11ses specific practices of displace1nenl and fo~-11.,.t'd, disciplin<'d attention.
In 1his <haptcr I hope to chtfify <l crucial and au1bivalent anthropological
11g;1ey: tl1e n1le <if tra\'el, pl1ysical displacenlt"lll, and temporary dwelling away
In 1111 !10111t i11 the t1111slitulion of fieldwork. I will discuss fieldwork and travel
in 1hree stctions. The first sketches some recent developments in sociocul1ur;1l anthropoloh'Y showing where classic research practices are under pressun. I sug)~est why fieldwork rcni;1ins a cenLral feature of disciplinary selfdtlinition. rhe second section focuses on fieldwork as an embodied spatial
pr;urit't'. sho\vi11~ how, since the turn of the century, a disciplined profes-.ion<ll hotly h<ts hten <.trticulatcd along a ch<1.nging border with literary and
jour11alis1ic travel practices. In opposition to these purportedly superficial,
sul>'11c1i\'e, a11d hi;lsed fonns of knowledge, anthropological research was ori1ntl'd 1oward 1he production of deep, ndtural knowledge. I argue that the
bonier i.., unstable, const;ullly renegotiated. The third section surveys cur1 1nt c11ntts1<.11ion~ of norntative EuroAinerican travel histories that have long
s1r1 1ctur<d anthropology's research practices. Notions of community insides
.ind cnusidts, ho111es and ahroads, f1t~lds and n1etropoles, are increasingly
cli 41 11rugrd hy postcxotic, deculoniLing: trends. It is much less clear what
n 111111s, 111d<1y, as ;\fcep!ahle fit"ldwork, the range of spatial practices "cleared"
h\' 1hr discipline.
I borrow the phr;,1se "spatial practice" fnun Michel de Certeau's book 1'he
/)u1d 1r,, 11( F,1irryday /,ife (1984). for de <:erteau, "space" is never ontologic.ally
gi\t11. h is cli.scursivtly n1<1pptd and corporeally practiced. An urban ne1ghll11duic1d, for t'x; 1111 ple, nta)' he laid out physically according: to a street plan.
Bui i1 is not a splce uni ii it is practiced by people's active occupation, their
11 10\'t"lllt"nts 1hroL1gh <t1Hl around it. [n this perspective, there is nothing given
a\ 11111 1 a "licld." lt 1n 11s1 he worked, turned into a discrete social space, by embodied pr<tclicts ofintcrat1ive travel. I \viii have n1ore to say, en route, about
the exp;111dtd stnse, ;incl li1nit<1tions, of the term travel as I use it. And I will
he tonct'rned, p1-i 1uarily, wi1h nonns oind ideal-types. In chapter 1, Gupta
;ind Ftr14uson ;trgue 111;,ll rt11Te11t practice potentially draws on a broad range
c1f ttl11u1gn1pl1ic acti\'ilics, sonic of then1 unorthodox by rnodern standards.
But ther confirrn that, sinct" the HJ20S, a recognizable norm has held sway

\VitS

SPATIAi. PRACTICES

in the academic centers of Euro-America. 1 Anthropological fieldwork has

represented something specific among overlapping sociological and elhnographic methods: an especially deep, extended, and interactive research encounter. That, of course, is the ideal. In practice, criteria of ..dcp1h" in lieldwork (length of stay, mode of interaction, repeated visits, grasp oflangu;iges)
have varied widely, as have actual research experiences.
This multiplicity of practices blurs any sharp, rcfCrential 1neaning for
"fieldwork."\r\Thatare we talking about when ,.,.e invoke a11thropologic;1J fieldwork? BefOrc proceeding, I must linger a 1non1ent on this prohlen1 of definition. Elernentary se1nantics distinguishes se\'eral W<l}'S 1nea11ings arc sustained: roughly, by reference, concept, and use. I will draw prinr<trily on 1he
latter two, com1nonly qualified as .. 1nentalist" (Ak1n;~jian et ;11. 19~1:r 198201). Conceptual definitions use a prototype, often a visual iniage, to define
a core against which variants are evaluated. A fa111ous photograph of Malinowski's tent pitched in the midst ofa Trobriand vil1;1ge has long ser\'ed as
a potent mental image of anthropological tield\\/Ork. (Everyone "knows" it,
but how many could describe the actual scene?) There have hC'en other in1ages: visions of personal interaction-for exan1ple, photos of Margaret
Mead leaning intently toward a Balinese rnothcr and bah)' l\foreo\'cr. 01.o; I
have already suggested, the word "field" itself conjures up 1nental in1agcs of
cleared space, cultivation, work, ground. When one spc<.lks of working in the
field, or going into the field, one draws on 1ne111al iinages of a dis1inc1 pl<1ce
with an inside and outside, reached by practices of physical n1ovcn1cnt.
These mental images focus and constrain definitions. For exan1plc, !hey
rnake it strange to say that an anthropologist in his or her office talking on
the phone is doing fieldwork-even if what is actually happening is the disciplined, interactive collection of ethnographic data. In1ages rnaterialize concepts, producing a semantic field that seems sharp at the "cen1cr" and blurred
at the "edges." The same function is served by n1ore abstract concepts. A
range of phenomena are gathered around prototypes; I will, in fleference
to Kuhn (1970: 187). speak of exnn/Jlars.Just :.is :.i robin is taken to he a n1ore
typical bird than a penguin, thus helping to define thr conccpl "hird," so
certain exemplary cases offieklwork anchor heterogeneous experiences. "Exotic" fieldwork pursued over a continuous period of al leasl a year has, for
some time now, set the norm against which other pracliccsarejudgcd. (;ivcn
this exemplar, different practices of cross-cultural research seen1 leM like
"real" fieldwork (Weston chapter 9 of this book}.
Real for whom? The meaning of an expression is ultin1ately delern1incd
by a language community. This usecri1erion opens space for a history and sociology of meanings. But it is con1plicatcd, in 1he present case, by 1he fart that
those people recognized as anthropologists (the rclev.tnl conununity) arc critically defined by having accepted and done something close (or close enough)
to "real fieldwork." The boundaries of the relev.1nt cornn1u11ity have been (and

18N

Sl'1\TIAL PRACTICES

JA~tES CLl'FORD

an increasingly) constiluted by struggles over the term's proper range of

1neanings. This cu1nplication is present, to some extent, in all communityust criteria for 111caning, especially when "essentially-contested concepts"
(< ~<1llic 19G4) ;,ire ell sl<tke. But in the case of anthropologists and "field\vork,"
the loop of mutual constitution is unusually tight. The con1munity does
not siinply use (define) the 1er1n "fieldwork"; it is ntaterially used {defined)
h\' ils sc11scs. A different range of 1neanings would make a different com11;11nity of;uuhropolog:ists, and vice versa. The sociopolitical stakes in these
d(liuitions-issut:s of inclusion and exclusion, center and peripherytH'l"d to hr kept explicit.
DISC! Pl.I NARY HORDERLANDS

( :, 1nsidtr lht pr<~jtct of Karen r...1cC:arthy Hrown, who studied a vodou priestess in Hrooklyn (and acco1npanied her on a visit to llaiti). Brown traveled
i1110 1he held by car, or on the New York subway, fro1n her home in Manhattoul. Iler ethnography was less a practice of intensive dwelling (the "tent

iu the village") and 1nore ;.11nattcr of repealed visiting, collaborative work.

( >r prrhaps \vh;:1t was involved wa.s what Renato Rosaldo once called, in a disl'll'~ion ol\vhat n1akes anthropological ethnography distinctive, "deep hangin~ ou1.'"1 Before working with Alourdes, the subject of her study, Brown had
1nade research trips lo 1-laiti. But when she visited Alourdes for the first time,
she felt a new kind of displace1nent:

()ur 1111strils lillccl with the s1nells of charcoal and roasting ntcat and our ears
h'ilh o\'trl<1ppi11~ episodes of salsa, reggae, and the bouncy n1onolony of what
If ai1ians call jal7.. Ani1nated conversations could be heard in 1-taitian French
C:nolc, Spanish. and 1nore than one lyrical dialect of English. The street was
a rra;.y <1uilt of shops: Chicka-Licka, the Ashanti Bazaar, a storefront Christian
d1urth wilh an iruprobably long and specific na1ne, a Haitian restaurant, and
g(,l;tnka Shan go-one or the apo1hecaries of New Work:I African religions oflt'riui.i; fast-luck ouul gc1-rid1-<1uick powders, High John the Conqueror root,
;uul \'Olive candles 1narked IOr the Sevt"n African Powers. I was no more than
<t ft_w 1niles fnun 1ny ho111e in lower Manhattan, but I felt as if I had taken a
w1ung turn. slip1>ed through a crack between worlds, and emerged on the main
'"'ttI ofa iropical rity. (l~rown 1991: 1)
C:o111parc this "arrival scene" (Pratt i986) with Malinowski's famous
"l111;1gi11<' yourstlf' !'it.'I down [on a Trobriand Island beach]" (Malinowski
t< 1f)1 ). Both rhe-1orically construct a sharply different, tropical "place," a topos
,t;ul topic for the work to li:>llow. But Brown's contemporary version is prestnted wilh a degree t>f in,ny: her lropical city in Brooklyn is sensuously real
a111J i111<1ginary-a11 "illusion," she goes on to call it, projected by an ethnogr.iphic 1raveler iu <l co111plexly hyhrid world-city. Hers is not a neighborhood
(urban village) study. If it has a n1icrocos1nic locus, it isAlourdes's three-story

189

row house in the shadow of the Brooklyn-Queens Exprcssway-ho1ne of t11e

only Haitian family in a black North A1nerica11 neighborhood. lliasporic
"Haiti," in this ethnography, is n1ultiply located. Brown's ethnography is situated less by a discrete place, a field she enters and inhahils fi:>r a ti rue, tha11
by an interpersonal relationship-a 1nixture of obser\ation, dialogue, apprenticeship, and friendship-with Alourdcs. With 1his rela1ionship as its
center, a cultural world of individuals, plct('t'S, 1nen1oriC'.~. and praciiccs is
evoked. Brown frequents this world both in Alourdes's house, where ceremonies and socializing take place, and elsewhf're. Brown's "field" is \\'herever she is with Alourdes. She returns, typic;tlly, to sleep, reflect, write up her
notes, and lead her life at ho1ne in lower J\.1anhattan.
Following established fielrlwork practice, Ilrow11's t'thnography contains
little detail about the everyday lifC in Manhatt<111 intcr.~ptrsed wilh 1hc visits
to Brooklyn. Her field re1nains discrete, ..out there." Auel while the relationship/culture nnder study cannot be neatly sp;Hialized, a different place
is visited intensively. There is a physical, interpcrs(l11al i111cr<1cticltl v.:ith a clistinct, often exotic, world, leading to an experience of initia1ion. \'\'hile the
spatial practice of dwelling, taking up residence in a con111n1nity, is not observed, the ethnographer's 1nove1nent "in" and "out," her coining and going, is systen1atic. One wonders what effects these proxiruities and dis1<u1ces
have on the \V-.ay Brown's research is conceived and represented. I-low, for
example, does she pull back fron1 her research relcttionship in onkr 10 write
about it? This taking of distance has typically been conceived as a "clep;1rture" from the field, a place clearly removed fro1n hon1c (Crapanzano 1977).
'What difference docs it make when one's "inforn1ant" routinely calls one at
home to demand help with a cercn1ony, support in a crisis, a favor? Spatial
practices of travel and te1nporal practices of writing have been crucial to the
definition and representation of a 1t>j>ic-the translation of ongoing experience and entangled relationship into something distanced and rcprcsc11t;1hlc
(Clifford 1990). How did Brown negotiate this 1ransla1ion in <l llcld '''hose
boundaries were so fluid?
A similar but more extren1e challenge for the ~lefinition of .. real" lieldwork is raised by David Edwards in his article "Afghanistan, Ethnography,
and the New World Order." Entering anthropology with hopes of returning to Afghanistan to conduct "a traditional sort of village study in sorne
mountain community," Edwards confronted a war-torn, disperst'd .. field":
"Since 1982, I have carried out fieldwork in a votricly of places, including
the city of Peshawar, Pakistan, and various refugee can1ps scattered around
the Northwest Frontier Province. One sununer, I also traveled inside
Afghanistan to observe the operations of a group of mujaluu/in, an<l 1 have
spent quite a bit of time among Afghan refugees in the Washing1on, ll.C:.,
area. Finally, I have been n1onitori11g the activities of an Afghan co1npu1cr

newsgroup" (Edwards 1994: 345).

JAMES CLIFFORD

Muhilocale ethnography (Marcus and Fischer 1986) is increasingly faniiliar; 1nullilocale fieldwork is an oxy1noron. I-low many sites can be studied
inlt'nsively betOre criteria of "depth" are compromised?~ Roger Rouse's fieldwork in l\VO linked sites retains the notion of a single, albeit mobile, com111unily (Rouse 1991 ). Karen McCarthy Brown stays within the "world" of an
individu<tl. But David Edwards's practice is 1nore scattered. Indeed, when he
ht~ins to link his dispersed instances of "Afghan culture," he rnust rely on
t~ 1 irly weak 1he111atic resonances and the corn1non feeling of "a1nbiguity" they
proc.lucr-at ltast for hirn. Whatever the horders ofEdwards's "1nultiply-in1lt11ed" cultural object (I-larding 1994), the range of spatial practices he
~Ld11pls to encounter ii is exen1plary. He writes that he has "carried out fieldwork" in,, city and refugee (an1ps~ he has "traveled" to observe the mujahadi11; he has "spent quite a bit of ti1ne" (hanging out deeply?) with Afghans
i 11 \t\'ashington, l).C.; and he has been "n1onitoring" the exiled co1nputer
iHwsgroup. l'his last elhnographic activity is the least comfortable for Ed'"arcls (19~)1= 349). At lhe ti1ne of writing, he has only been "lurking," not
p(1s1 ing his own 111essages. 1-lis research on the Internet is not yet interactive.
H111 ii is ve1v infonncHive. Edwards intensively listens in on a group of exiled
AIKhcuis-;nale, relatively affluent-worrying together about politics, reli~ii 111s J>ractices, and the nature and boundaries of their community.
rhr experiences of Karen McCarthy Hro\vn and David Edwards sub.-gest
-. 0 1111 of 1he current pressures on anthropological fieldwork seen as a spa1i;1l pracli<e of intcusivc dwelling. The "field" in sociocultural anthropology
has hern consti1u1ed by a "historically specific range of distances, boundaries,
;111d 1nodrs of trave-1" ({:lilford l~HJO: ()4). These arc changing, as the geogr;1phy of distance and difference alters in postcolonial/neocolonial situations, as power relations of research are reconfigured, as new technologies
of transport and co1nnu1nication are deployed, and as .. natives" are recognit<'(I for 1heir specific worldly experiences and histories of dwelling and traveling (Appadurai t988h; C:lifford 1992; Teaiwa 1993; Naray.in i993). What
re1nains of classic anthropological practices in these new situations? How are
1he1101ions of trave I, boundary, coresidence, interaction, inside and outside,
wl1irl1 l1ave de lined the field .and proper fieldwork, being challenged and
reworked in conte1nporary anthrnpolobry?
!\(fore 1;iking up thest <1uestions, we need a clear sense of what do1ninant
11 1;1c1 itts (If lht" "lie Id" ;ire al issut, and what issues of disciplinary definition
r 1 111,,tr; 1i11 current argurncnL'i. Fieldwork norn1;.11ly involves physically leaving
"ho111r" (however I hat is di:flnt"d) to Iravel in and out of some distinclly dif
/t>-r11t selling. Today, the setting can be 1-lighland New Guinea, or it can be
,1 nrighhorhood, house, offict", hospital, church, or lab. It can be defined as
;111111hile socitty, that ofh1ng-clistanC"e truckers, for example-providing one
o;pt1ut-. long hours in 1he cah, 1alking (Agar 1985). Intensive, "deep" inter-

SPATIAL PRACTICES

191

action is required, something canonically guaranteed by the spatial practice

of extended, if temporary, dwelling in a comrnunity. Ficl<hvork can also involve repeated short visits, as in the Arncrican tradition of reservation ethnology. Tearnwork and long-term research (Foster et al. 1979) have lX"cn variously practiced in different local and national 1raditions. Bui con1n1011 to
these practices, anthropological fieldwork re<1uires that one do so1ncthing
inore than pass through. One must do 1nore than conduct in1ervic\vs, 1nakc
surveys, or compose journalistic reports. This rec111ire1ncnt continues 1oclay,
e1nbodied in a flexible range of activities, fron1 corcsidcnce to various forrus
of collaboration and advocacy. The legacy of intensive fieldwork defines
anthropological styles of research, styles critically important for disciplinary
(self-) recognition. 4
There are no natural or intrinsic disciplines. All kncnvledKe is interdisciplinary. Thus, disciplines define and redefine the1nselves interactively and
competitively. They do this by inventing traditions and canons, by consecrating rnethodological nonns and research practices, by appropriating,
translating, silencing, and holding at b:ay ;uUacent pcrspe<"lives. Ac1ive
processes of disciplining operate at various levels, defining "hot" ;111d "cold"
domains of the disciplinary culture, certain areas th;ll change rapidly and
others that are relatively invariant. They articulate, in tactically shihing \vars.
the solid core and the negotiable edge of a recognizable do111ain of knowledge and research practice. Institutionalization channels and slows, but cannot stop, these processes of redefinition, except at peril of sclerosis.
Consider the choices faced today by so1ncone planning 1hc syllabus for
an introductory graduate proseminar in sociocul1ural antJ1rop(>logy.'' (;ivc11
a lin1ited number of weeks, how important is it that novice an1hropologisls
read Radcliffe-Brown? Robert Lowie? Would it be better tu include Mever
Fortes or Kenneth Burke? Levi-Strauss, surely ... but why not also Sinu;nc
de Beauvoir? Franz Boas, of course ... and Frantz Fanon? Margaret Mead
or Marx ... or E. P. Thompson, or Zora Neale I-lurston, or Michel Foucault?
Melville Herskovits perhaps ... and W. E. B. DuBois? St. c:Jair Drake? \o\'ork
on photography and media? Kinship, once a disciplfnary core, is now actively
forgotten in some departn1ents. Anthropological linguistics, s1ill invoked as
one of the canonical '"four fields, n is very unevenly covered. In so1ne programs, one is more likely to read literary theory, colonial history, or cogni
tive science .... Synthetic notions of 'nan, the "culture-bearing ani1nal, .. that
once slitched together a discipline now see1n ;u11iq11ated or perverse. c:an
the disciplinary center hold? In the introductory syllabus, hylnid selection
will eventually be made, attuned to local lrculitions and c11rrtn1 dC"rnands,
with recognizably "anthropological" authors at the re11tcr. (Sonn1in1es tla
.. pure" disciplinary lineage will be cordoned off in a history of anthropology course, required or not.) Anthropolohry reproduces itself while .'.;electively
engaging with relevant interlocutors: fro1n social hislory, frorn cuhural stud-

Hj2

JAMES ClJFI-'<)RD

its, frotn biology, fron1 cognitive science, frorn mi~or~ty and femi~ist sc~ol
<uship, fronl colonial discourse critique, from sem1oucs and media studies,
frorn lilerary and discourse analysis, fron1 sociology, frorn psychology, from
li11g11is1ics, front ecoloh')', from political economy, from ... .
. .
~orin(ultural anthropoloh'Y has always been a fluid, relatively open d1sc1plinc. It lias prided itself on its ability to draw on, enrich, and synthesize other
!irids of study. Writing in 1~164, Eric \Volf opti1nisti.cally defined_ anthropoloh'1' as a "disciplinr bttween disciplines" (Wolf 1964: x). But tlu~ openness
pi;.;cs recurring prohle1ns of self-definiti_on. A1_1d .ra.rtly becaus: Its theo~et
ic.d purview h<ts rernained so bnJad and 1nterd1sc1~hnary, despite recurnng
;ii1cn 1p1s to cut it down 10 size, the discipline has focused on resear.ch practitC'i ~ 1 ., ton', defininv; elc1nents. Fieldwork has played-and co11ttnues to
plav-; 1(tntral disciplining function. In the current conjt~ncture, the ~ang~
uf lopics anthropoloh')' can study and the array of theoretical perspectives It
c; 111 dei>loyare inuncnsc. In these areas the discipline is "hot,"constanrlychan?ing, hyhridii.ing. In the ..colder" do1nain of accepta~le fie~c~wor~; cha..nge is
; 1 js 11111 <urrin~ but 1nore slowly. In 1nost anthropological n11heus, real fieldwork l'Olltiinu.s 10 he actively defended against other ethnographic styles.
rhc t'Xotic exc1nplar-coresidence for extended periods away fro1n
hoTllt', tht "tent in th(' village"-relains considerable authority. But it has,
in pr.utitc, hc('ll dcrcntercd. l'he various spatial practices it authorized, as
wrll ;p; the. relev~u1t criteria for evalu;lling "depth" and "intensity," have
ch;ingc.d ;uul continue to change. C:onte1nporary polit~c~l, cultural, and eco1101nit' rondi1ions bring new pressures and opportunities to anthropology.
Th(' r;iugT of possible venues for elhnographic study has expanded dra111a1 iralh'. and rhe discipline's potential 111e1nbership is n1ore diverse. An1hropolohry's geopolitical location (no longer so securely in the Euro-~er
lt ; 111 .. rt'nter") is challenged. In this context of change and contestauon,
;1racltniic anlhropology struggles to reinvent its traditions in n~w ~ircur:r1st;uut.'i. I .ike the changing societies it studies, the discipline sustains itself tn
liliintd and polict"d h11rderlands, using strategies of hybridization and reaul Ii t' nt ilic;,1t ic111, ;,1.,si1nil<1t ion incl exclusion.
Suggestive hounclary problc1ns e111erge fron1 David Edwards's awkward
tiiue 011 1ht Afghan Internet. \\'hat ifso1neone studied ~he c.ulture of.compiHt'r hacktrs (a perfectly acceptable anthro.pology pr~~ct in many, l~ not
.tll, depar011ClltS) and in the process llCVCr "interfaced Ill the nesh with~
.,j 11 ,i;lt harkrr. Would tl1e 1nonths, even years, spent on the Net b~ fieldw..o~k.
rhc nsc.arch n1igh1 well pass both the length-of-stay and the depth /1n1tr;1tti\'ily tests. (\Ve know 1hat sonic strange and intense .convers~tions can
o<Tllr ovrr the Net.) And cleclronic travel is, o-lfler all, a kind of depayserrumt.
Jr could <tdd up to intensive participant observation in a different con1muuit\' \\'llhotlt one's cv<r physic;,1lly leaving ho1ne. When I've asked anthropologi.,ts whr1hcr this could be fieldwork, they have generally responded

SPATIAL PRA<:TJCES

'93

"maybe," even, in one case, "of course." But when I press 1hc point, asking
whether they would supervise a Ph.I). dissertation based pri1narily on this
kind of disembodied research, they hesitate or say no: it \vould not he currently acceptable fiel(lwork. (;ivcn 1hc traditions of the discipline, a graduate student woul<l be ill advised to follow su<h a course. We c.on1c up again.'it
the institutional-hislorical consuaints that enforce 1hc dis1i11cti1111 hetwte11
fieldwork and a broader 1ange of ethnographic activities. Ficlchvork in anthropology is sedimented with a discipli11ary history. and it lo11linues lo funttion as a rite of passage and n1arker of profcssio11<1lis111.
A boundary that currently preoccupies sociocultural anthropoloh'Y is lha1
which separates it from a heterogeneous collcclion of ac;1d<n1ic pr<1rtirc!'i
often called "cultural studies. "fi This border renegoliatcs, in a new <01Hex1,
sorne of the long-established divisions and crossings or so(iology and anthropology. Qualitative sociology, at least, has its own erhnog1aphic traditions, increasingly relevant to a postexoticist anlhropoloh'Y 7 Hui given fairly
firm institutional identities, in the Uni led States ;11 Jeast, the bonier with !;1r
ciology seems less unruly than that with "cultural studies." rhis new sit<' of
border crossing and policing partly repeats an ongoing, fraughl relationship with "textualism" or "lit crit." The n1ove to .. recaprure" anthropoloh'Ymanifested in disn1issals of the collection WritinK Culture ({:Jifl(lrd and f\-1arcus 1986) and n1ore recently, often incoherc111ly, in s\vrcping rcjrTlions of
"postmodern anthropology"-is by JlO\'I/ routine ill some quarters. But 1he
border with cultural studies 1nay be less n1anagcahle; for it is easirr lt> 1nai11tain a clear separation when the disciplinary Othcr-li1crary-rhttoric.al
theory or textualist semiotics-has no fieldwork con1ponent and at htst an
anecdotal, ..ethnographic" approach to cultural phcnorncna. "C:ultural studies," in its Birmingham tradition as well as in so111c of its soci11logical veins,
possesses a developed ethnographic tradition n111ch closer to anthropological fieldwork. The distinction "\Ve do fieldwork, they do cliscour~e analrsis"
is more difficult to sustain. Some anthropologists h;ne turned lo cultur;1l ~l\ul
ies ethnography for inspiration (Lave ct al. 199~), and inclccd there is n1uch
to learn from its increasingly coniplcx articulalions of class, gcnclC'r, race,
and sexuality. Moreover, what Paul Willis did with the worki11g-cl<1ss .. l;ids"
of Learning to Labour (1977)-hanging out with tl1c1n at school, 1alking with
parents, working alongside them on the shop floor-is c1nnparahle lo good
fieldwork. Its depth of social interaction was surely grcalt"r lhan, say, th<ll
achieved by Evans-Pritchard during his ten 1nonths \Vilh hos1ilc and reluctant Nuer.
Many contemporary anthropological pn~jtcts arc cliffi<11h to clisti11guish
from cultural studies work. For cxa1nplc, Susan J larding is writing an l'tl111ography of Christian fundatnentalis1n in the United States. She has done extensive participant observation in Lynchburg, Virginia, in and around jl'rry
Falwell's church. And of course the tclC'vision 1ninistry of f;1hvcll and others

JAMES CLIFFORD

like hint is very 1nuch her concern-her "field." Indeed, she is interested
not pri1narily in a spatially defined conununity but in what she calls the "disrourse" of the new funda1nentalis1ns. 11 She is concerned with rv programs,
sennons, novels, 1ncdia of all kinds, as well as with conversations and everycl<w hl"havior. f Iarding's 1nixture of participant observation, cultural criticis1n,
and n1C'dia and discourse analysis is characteristic of work in the current
1thnoj.{ltphic hordcr zones. How "anthropological" is it? How different is
Susilll I lanling's frequenting of evangelicals in Lynchburg fro1n Willis's or
Angela I\ le Robbie's snulies of youth cultures in Britain or the earlier work
11f the C:hi<ago School sociologists? There are certainly differences, but they
do not toalesce as a discrete 1neth~d, and there is considerable overlap.
( ltH' i1nportant difference is Harding's insistence that a crucial portion
ofhtr ethnographic work involves livinguJith an evangelical Christian fa1nilv. lndctcl, she reports that this was when she felt she had really "entered
1ht field." Previously, she had stayed in a motel. One might think of this as
a tla . . sic articulation of fieldwork deployed in a new setting. In a sense it is.
Hut i1 is part of a potentially radical <lecentering. For there can be no question of calling the period of iruensive coresidencc in Lynchburg the essence
or core of the pn~ject to which the TV viewing and reading were ancillary.
In I Ianting's project, "fieldwork" was an i1nportant way of finding out how
the new fundarnentalisrn was lived in everyday terms, And whilC it certainly
ht'lptcl define her hybrid pr<~ject as anthropological, it was not a privileged
'iiH" of interactive depth or initiation.
I larding's work is an exan1ple of research that draws on cultural studies,
discourse analysis, and gender and ntedia studies, while maintaining crucial
;tnlhropolo~ic<1l features. It 1narksa current direction for the discipline, one
in which fieldwork ren1ains a necessar}' but no longer privileged method.
Does this rnean that the institutional border between anthropology, cultural
.;tudits, and allied !raditions is open? Far from it. Precisely because the crossings an. so prorniscuuus an cl the overhtps so frequent, actions to reassert identity an nun1n1ed at str;:1tegic sites and 1no1ne11ts. These include the initialofy proctss of graduate certification, and 11101nents when people need to
I it d<11ied a job, funding, or authority. In the everyday disciplining that makes
anthropologists and not cultural studies scholars, the boundary is reasserted
1011ti11el~ l\lost publicly perhaps, when graduate students' .. field" projects
.ut ;ipproved, the distinctive spatial practices that have defined anthropologv h'IHI to he reasserted-often in nonnegotiable ways.
rJH roncept of !he field and the disciplinary pr.tctices associated with it
1011s1it111e a cenlral, ;;11nbiguous legacy for anthropology. Fieldwork has ben 111u a prohlc1n because of its posilivist and colonialist historical associations
( t hr fitlcl as .. lahoral ory," !he field as place of "'discovery" for privileged sojourners). It has ;dso beconte n1ore difficult to circumscribe, given the proliftr;11io11 of cthno~raphic topics and the tin1e-sp;1ce co1npressions (Harvey

SPATIAL PRACTICES

'95

1989) characteristic of postmodern, postcolonial/ncocolonial situations.

What will anthropology make of this prohle1n? Ti1ne will !ell. Fiel<h\'ork, a
research practice predicated on interactive depth and spatio1lizecl difference,
is being "reworked" (Gupta and Ferguson's tern1 in chapter 1 ), fo1 it is one
of the few relatively clear rnarks of disciplinary distinction l<ft. But ho'v wide
can the range of sanctioned practices he? And how "decentercd" (Gupta and
Ferguson) can fieldwork become before it is just one of a range of ethnographic and historical methods that the discipline uses in concert 'vilh other
disciplines?
Anthropology has always been n1ore than fieldwork, hllt fielthvork has
been something an anthropologist should have done, 1nore or less well, at
least once.Y Will this change? Perhaps it should. P(rhaps fieldwork \\'ill become n1crely a research tool rather than an essential disposition or professional marker. Time will tell. At present, ho\\'ever, fielchvotk re1nains crilically important-a disciplining process and an arnhiguous legacy.
THE FIELDWORK HABITUS

The institutionalization of fieldwork in the late nineteenth and e;irly t\1tentieth centuries can be understood within a larger history of "travel." (I use
the tcrrn in an expanded sense, of which I will say 111orc in a 11101nc111.)
Among Westerners traveling and dwelling abroad, !he ;;n1thropological
fieldworker was a latecomer. Explorers, rnissionarics, colonial officers,
traders, colonists, and natural scientific researchers were well-established figures before the emergence of the on-the-ground an!hropological professional. Prior to Boas, Malinowski, Mead, Firth, and their colleagues, the anthropological scholar usually remained at hon1e, processing ethnographic
information sent by "men on the spot" who were drawn fron1arnong1he sojourners just mentioned. If metropolitan scholars ventured out, it \\'as on
survey and museum-collecting expeclilions. \.Vhate\'Cr exceptions there 1nay
have been to this pattern, interactive clep1h and co1-csidence were not yet
professional require1nents.
\tVhen intensive fieldwork began to be cha1npioncd by !he Ro;:isians and
Malinowskians, an effort was required to clislinguish the kind of knowledge
produced by this 1nethod from that acquired hy o!her long-term resident<;
in the areas studied. At least three "disciplinary Others" were held at ar1n's
length: the missionary, the colonial officer, and the travel writer (journ01list
or literary exoticist). Much could he said about an!hropology's fraugh! re
lations with these three professional alter egos \\'host' purportedly c11natc11r,
interventionist, subjective accounts of indigenous life would be "killed by
science," as Malinowski put it (1961: 11 ). w My focus, here, is li1ni1ccl to the
border with literary and journalistic travel. As a 1ncthodological principle, J
do not presuppose the discipline's self-definitions, whether positiv< ("'\ve ha\e

JAMES CLIFFORD

special research practice and understanding of human culture") or negative (''we are not inissionaries, colonial officers, or travel writers"). Rather, I
assu1ne that these definitions must be actively produced, negotiated, and
n~negotiated through changing historical relationships. It is often easier to
say clearly what one is nol than what one is. In the early years of modern anthropology, while the discipline was still establishing its distinctive research
tradition and authoritative cxen1plars, negative definitions were critical. And
in tin1es of uncertain identity (such as the present), definition may be
<tC'hievecl rnost effectively by narning clear outsides rather than by attempting
to reduce always diverse and hybrid insides to a stable unity. A more or less
pcrinanent process of disciplining at the edges sustains recognizable borders in entangled borderlands ..
1\1Hhropological research travelers have, of course, regularly depended
1111 n1issionaries (for granun<trs, transportation, introductions, and in certain
<ast's for a dcepc:r translation of language and custo1n than can be acquired
in a one- or two-year visit). The fieldworker's professional difference from
rhe n1issionary, based on real discrepancies of agenda and attitude, has had
to he a.c;serted against equally real areas of overlap and dependency. So, too,
\\:ith colonial (and neocolonial) regimes: ethnographers typically have as~trted their ain1 to understand not govern, to collaborate not exploit. But
tluy l1ave navigated in the do1ninant society, often enjoying \'o'hite skin privilege <uul a physical safety in the field guaranteed by a history of prior puniti\c txpeclitions and policing (Schneider i995: 139). Scientific fieldwork separated itself fro1n colonial regimes by claiming to be apolitical. This
distit1r1ion is currently being questioned and renegotiated in the_ wake of
;u11icolonial 1nove1nents, which have tended not to recognize the distance
fn1111 C(111texts of dornination and privilege that anthropologists have claimed.
rhe travel w-ritcr's transient and literary approach, sharply rejected in the
distiplining of fieldwork, has continued to tempt and contaminate the scientific practices of cultural description. Anthropologists are, typically, people
who leave and write. Seen in a long historical perspective, fieldwork is a dis1i1utivc cluster of travel practices (largely but not exclusively Western). Travel
and travel diM:oursc should not be reduced to the relatively recent tradition
11f litC"rary tra\el, a narrowed conception that e1nerged in the late nineteenth
;uul early twentieth centuries. This notion of "travel" was articulated against
;111 rn1r1ging ethnogr<iphy (and other fonns of "scientific" field research) on
the one hand, and against tourisrn {a practice defined as incapable of producing strious knowledge) on the other. The spatial and textual practices of
what rnight no\Y he called "sophistic;.Hed travel"-a phrase taken fro1n New
li-11k r;nv.~ supple1ntnt'i catering to the "independent" traveler 11 -function
,.,ithin an elile, 111d highly differentiated, tourist sector defined by the state111rnt .. \Ve are ,.ot tourist~.,. (Jean-Didier Urbain in L 'idiot du voyage [ 1991]
ha~ thoroughly analy1ed 1his discursive fonnation. See also Buzzard 1993.)

;1

SPAl'IAI. PRA<:TJCES

197

The literary tradition of"sophisticatcd travel," whose disappeardncc has been

lamented by critics such as Daniel Boorstin and Paul Fussell, is reinvented
by a long list of contemporary wri1crs-Panl Theroux, Shirley I lazzard, Bruce
Chatwin,Jan Morris, and Ronald Wright, a1nong others.'2
"Travel," as I use it, is an inclusive terrn embracing a range of 1nore or less
voluntaristic practices of leaving "home" to go to sorne "01hcr" place. 'l'he
displacement takes place for the purpose of gain-1natcrial, spiritual, scientific. It involves obtaining knowledge or having an "experience" (exciting, edifying, pleasurable, estranging, broadc11ing). Tile long history of tra\'el
that includes the spatial practices of ..fieldwork" is predo1ninantly Westerndorninated, strongly male, and upper 1niddle class. Good critical and historical work. is now appearing in this coniparative dornain, paying atlention
to political, economic, and regional contexts, as well as to the detern1inations and subversions of gender, class, culture, race, and individual ps)'chology (Huln1e 1986; Porter 1991; Mills 1991; Pratt t992).
Before the separation of genres associated with the c1nergence of 1nodern fieldwork, travel and travel writing covered a broad spectrun1. In eighteenth-century Europe, a ricit de voyage or "travel book" n1ight include exploration, adventure, natural science, espionage, co1111nercial prospecting,
evangelism, cosmology, philosophy, and ethnography. By rhe 192<~. howc\'er,
the research practices and writtc11 reports ,,f.;1nt l1n,1u>lf>gisls had been rnl1cl1
more clearly set apart. No longer scientific travelers or explortrs, anl11ropologists were defined as fieldworkers, a change shared with other sciences
(see chapter 2). The field was a distinctive cluster of acadcn1ic research practices, traditions, and representational rules. But while cornpc1ing pr-..1ctices
and rhetorics were actively held at bay in the process, the ne\vly cleared disciplinary space could never be entirely free of con1an1ination. Its borders
would have to be rebuilt, shifted, and reworked. Indeed, one w;iy to understand the current "experimentalisn1" of clhnographic \vriting is as a renegotiation of the boundary with "travel writing," which was agonis1ic;1lly dtfined in the late nineteenth century.
"Literariness," held at a distance in the fig111:c of the travel writer, has n_turned to ethnography in the fonn of strong clainis ;.1bo111 the prefiguration and rhetorical com1nunication of "data." 'fhc facts do not speak f(>r
themselves; they are em plotted rather than collected, produced in worldly
relationships rather than observed in controlled en\iron1nents. 1" 'fhis growing awareness of the poetical and political contingency of ficlchvork-an
awareness forced on anthropologists by postwar anticolonial challenges 10
Euro-American centrality-is reflected in a rnore concrete tcxlual sense of
the ethnographer's location. Ele1nents of the .. literary" tr<t\cl narrative that
were excluded from ethnographies (or marginalized in their prefaces) no\V
appear 1nore pro1ninently. These include the researcher's routes into and
through .. the field"; time in the capitoil city, rC"gist('ring thr surrounding

JAf\.iES CLIFFORD

national/transnational context; technologies of transport (getting there as

wc.11 as being there); and interactions with named, idiosyncratic individuals,
ralhtr than otnunyrnous, representative inforn1ants.
In an earlier discussion, I have worked to decenter the field as a naturali1ed practice of duJeJling by proposing a crosscutting metaphor-fieldwork
as lratlf{ encuunlers (C:lifford 1992). To decenter or interrupt fieldwork-asdwclling is not to reject or refute it. Fieldwork has always been a mix of in. . 1i1utionalized pr.tcticesof dwelling anrl traveling. But in the disciplinaryide;1\i1ation of .. the field," spatial practices of 1noving to and from, in and out,
J>assing through, have tended to be subsurned by those of dwelling (rapport,
initioition, fa1niliarity). This is changing. Ironically, now that much anthropologi<al lieldwork is ronducied (like Karen McCarthy Hrown's) close to
hornc, the rnatcriality of tr;,1vel in and out of the field becomes 1nore appart'nl, indeed bcco1nes constitutive of the o~ject/site of study. Fieldwork in
cities 1nusl distinguish itself frorn other forms of interclass, interracial travel
and ;1ppreciation, n1arking a difference frorn established traditions of urban
social work ;:tnd liberal ..slununing." The home of the research traveler exisls in a politicited prior relation ro that of the people under study (or, in
ronternporary parlance, the people "worked with"), These latter may themsclvc.s travel regularly to and from the ho1ne base of the researcher, if only
for e1nploy1nent. (The "ethnographic," cross-cultural knowledge ofa maid
or service worker is considerable.) These parallel, sometiines intersecting,
spatiopolitical relations have also been present in "exotic" anthropological
research, particularly when colonial or neocolonial flows of armies, comn1odities, labor, or education n1aterially link the poles of fieldwork travel.
But i1nages of distance, rather than of interconnection and contact, have
Iended to naturalile the field as an Other place. The socially established routes
constitutive of field relatiuns are harder to ignore when the research is conducted nearby or when airplanes and telephones compress space.
Fieldwork thus .. takes place" in worldly, contingent relations of travel, not
in controlled sites of research. Saying this does not simply dissolve the bound;1ry ht>tween conrernporary fieldwork and travel (or journalistic) work.
rhert are i111por1ant generic and institutional distinctions. The injunction
10 dv1ot>ll intensively, to learn local languages, to produce a "deep" interpre1atio11 is a difference that rnakes a difference. But the border between the
1wo nl;11ivc.ly rc.cent 1raditio11s of literary travel and aca<len1ic fieldwork is
hc.iug re11t'go1iated. l11clcccl, 1he exaniple offered above by David Edwards's
111ulliple siles of encolnHer brings fielchvork (dangerously, some may feel)
rlost 10 travel. "fhis r.approche1nent takes a different forrn in Anna Tsing's
inn<1v;.uivc. eth11ography /11 lhe lltfllm uf the /)iamond Queen (1993). Tsing conducls lieldwork. in a classic .. exotic" site, the Meratus Mountains of South Kali111antan, Indonesia. \.\'hi le preserving disciplinary practices of intensive loc<1\ inlt"raction. ht'r wri1ing syste1natically crosses the border of ethnographic

SPATIAi. l'RA(:TH:Es

199

analysis and travel narration. l-ler account historicizes both her O\'/n and her
subjects' practices of dwelling and traveling. She deri,ed h('r kno\vlcclgc froin
.specific encounlers between diffCrently cosn1opolita11, gendtrcd i11diYirl11als, not cultural types. (See, particularly, Part rwo: ..A Scicn<e of 'I'ravel.'')
Her field site in what she calls an "out of the \Vay place" is n<ver taken for
granted as a natural or traditional environrncnt. It is a cont<ict space pro
duced by local, national, and transnational forces nf\vhich her r(search tr;:nel
is a part .
Edwards and Tsing exernplify exotic fieldwork <ll 1he edges of changing
academic practice. In both, differently spalialited, \\'<' see the increased
prominence of practices and tropes co1n1nonly asso<ia1td \Vilh rr;l\<I and
travel writing. 11 These are curre11tly visible in n1uch a111 hn1pological eth 1u 1graphy, figuring different versions of the routed/rooted researcher, the .. positioned su~ject" (Rosaldo 1989a: 7). Signs of the 1i1nes includ<' a 1rend 1o
ward use of the first-person singular pronoun in accounts of field\\ork,
presented as stories rather than as observations and interpretations. Often
the field journal (private and closer to the "sul~jec1ive" acToi1111s of travel \\Tiling) leaks into the "objective" field data. I arn not describing a linear ino,c1nent from collection to narration, objcctiYe lo sul~jectivc, irnpersonal 10 personal, coresidence to travel encounter. It is a question not of a progression
fro1n ethnography to travel writing but rather of ;1 shifting balance and a
renegotiation of key relation~ that have conslituterl the two prac1ices and
discourses.
In tracking anthropology's changing relations with !ravel, we 1nay find it ust'ful to think of the "field" as a habitus ralher than as a place, a clusler of ru1bodieddispositions and practices. The work of fcn1inist scholars has played a
crucial role in specifying the social body of the ethnographer, lvhile crilicizing the limitations of androcentric "gender-neutral" work ancl opening
up major new areas of unrlerstanding. 1 ~ Sirnilarly, anticolonial pressure's, coin
nial discours~ analysis, and critical race theory have clecr111rrrd the pre
dominantly Western, and white, traditional fieldworkC'r. S<'C'll in light ofthrse
interventions, the fieldwork hahitus of the Malinowskian generation appC'ars
as the articulation of specific, disciplined practices.
This normative "body" was not that of a lr;tvclcr. As ii drew on older traditions of scientific travel, it did so in .sharpened opposition 10 ron1antic, "literary," or subjective strands. The body legitimated by n1odern flel<hvork was
not a sensorium moving through extended space, across ho1ders. Ir was not
on an expedition or a survey. Rather, it wa."i ;:l hody circ11lct1ing and working
(one might almost say ..cornmuting") wilhin a dcli1ni1ed space. The local n1ap
predominated over the tour or itinerary as a technology of physical location.
Being there was more i1nportant than gelling there (and leaving there). The
fieldworker was a homebody abroad, not a cosn1opolitan ''isitor. I a111, of

2lH'

JAMES CLJFF()RlJ

course, speaking broadly of disciplinary nor1ns and textual figures, not of

1l1e actual historical experiences of field anthropologists. In varying degrees,
1hcse diverged fron1 the nonns while being constrained by thern.
E111otinns, a necessary part of the controlled e111pathy of participant obser\'alion, were not accorded pri1nary expression. They could not be the chief
soutTe of public judgrnents about the co1nnu1nities under study. This was
partic_ularly true of negative assess111ents. The 1noraljudgn1ents and curses
of the tr.1.vel writer, based on soci;.1! frustrations, physical discon1forts, and
prl'judices, as well as on principled criticisn1, were excluded or downplayed.
An understanding rapport and 111easured affection were favored. Expressions
of ov<'l"t enthusias111 and love were circun1scribe<l. Ang-er, frustratio11,ju<lg-111r111s on i11dh1idu;:1Js, desire, and a1nbivalence went into private diaries. The
~caudal provoked, in ~<HHC quarters, hy the publication of Malinowski's in1i111att diary ( 19fi7) was related to the gli1npse it gave of <t less temperate,
111ore race- and sex-co11scious, subject/ body in the field. Early public trans~ressiuns of tl1e profCssion<1l habit us include works by Leiris (1934, written
;1s a field journal}, Uowen (195,1, in novel forn1). and.Jean Briggs (1970, in
\\l1ih personal ellHltions pt'rhaps for the first tin1e were central to an ethno~raphic 11101u>graph).
If t1notio11s tended to he 1nargi11alizcd, so, for the 1nost part, did the restarcher"s ex1>critnces of gc11cler, race, and sex. Gender, while occasionally
featured (particularly in the "rnarkcd," female case), was not publicly recog11i1.ed as constituting the research process in a systematic way. Margaret
l\ltacl, fof exainple, did at tin1es tonduct research and write "as a wo1nan,"
(n1ssi11g defined 'von1en 's and 1nen 's spheres, but her disciplinary persona
w;1.~ that of a scieulilically authoritative cultural observer, of unmarked gender 011ul hy clefauilt "1nale . ., I fer 1nore "subjective,'' "soft" stylistic experiments
aud popular writings did not bring her credit within the disciplinary fraternity, wht're she adopted a n1orc "ol~jective," "hard" voice. Lutkehaus (1~195)
pro\ides a to11tcxtual account of rhcsc historically gendered locations and
~fe<1d's shifting pers11na. ~tale researchers of Mead'sgeneration did not cond1ut rese;1nh "as llll'll" ~11no11g locally defined wo1nen and 1nen. Many pur11oncdly holislic "culturc1l" accounts were, in fact, based on intensive work
\\'ilh 111<11 only. ()\'Crall, the constraints and possibilities attached lo the re~<arch<r\ gender wt.re not salient features of the field hahitus.
rl1t sa1ne went f11r rac:. l lerc, sociocullural anthropolo!,ry's i1nportant the11fttical ;111d e1npirical critique ,,f racialist essences doubtless influenced the
pn,fessi<>lliil l1al>illls ... R.tce" was 111,t the s<>cial /historical formation of con!(1111>j>rary <rit ical ract theoris1s (for cxa1nple, On1i and Winant 1986; Gilroy
1~187), hut a hiolo~iral tssenrt 'vhose "natural" detcnuin;1tions were contested by the contextual dettnninations of ..cullure." Anthropologists, the
tulture-bearing .scholi.irs, rHeded to decenter and cross over putatively es-~c111 ial r;1cial li11es. r11(_'ir interacti\'c ;111d intensive understanding of cultural

SPATIAL PRA<:TICES

201

furn1ations gave thern a powerful tool against racial reductions. But in attacking a natural phenon1enon, they did not confront race as a ltiMorical
fOrrnation that located their sul~jccts politically ancl that sin1ultancously constrained and ernpowerc<l their own research (I larriso11 1991: :\). 11 ' Occasicn1ally, this positioning could he gli1npsed-f(H' cx;unplc, in Evans-Pri1chard's
introduction to '/'he Nuer (1940); but it W<IS not p;irt of the explicit body, the
professional habitu.s, of the fieldworker.
By contrast, travel 'vriters often noticed color and spoke fio1n a racializcd
position. Of course, they were not necessarily critical of lhe rela1ions invoked
-often quite the reverse! The point is not to celebrate a rclali\'cly greater
awarenes.<; of race-and gender-in tra\'el writing, hut to show how, in contrast, the hahitus of the ethnographer downplayed these historical detcrrninations. IIowever rnarked it 'vas by gender, rare, rastc, or tJass pri\ilege,
ethnography needed to transcend such loca1io11s in order to articulate a
deeper, cultural understanding. 'fhis artic11latio11 was based 011 powerful t<chniques, including at least the f'ollowi111r extended coresiclcntl; sy.stc1na1ic
observation and recording of data; effective interlocution in <H least one lo
cal language; a specific 1nix of alliance, ctnnplici1y, friendship, rcspet, coercion, and ironic toleration leading to "rapport"; a hern1cnc11tic attention
to deep or implicit structures and 111ea11ings. Tl1csc tcch11iq11es were design eel
to produce (and often did produce, \Vithin lhc horit.ons I a1n trying to drlirnit) n1ore contexual, less reductive understandings of local lifclays th;111
did the passing observations of the traveler.
Some writers who could be classified as tra\'clcrs st;1yed for extended periods abroad, spoke local languages, and had co111plcx views of indigc11011 ..,
(as well as of creole/colonial) life. So1ne classified as ethnographers srayed
relatively short times, spoke langu;1ges badly, and did not interact intenshely.
The range of actual social relations, con1nH1nicati\'C lethniques, and spatial
practices deployed between the poles of fieldwork and travel is a to111i111n11n,
not a sharp border. There has been considerable overlap. 17 But in spite of,
or r.tther because of, this hordcr con1plexity, the di~ursivc I ins1itu1ional lines
had to be clearly drawn. This need sustained pressures \\'hich, o\Tr ti1nc, got th
ered empirical experiences closer to lhe two poles. In 1his process, the .. superficiality" of the traveler and travel writer was opposed to the "depth" of
the fieldworker. But one rnight also say, pro\'ocati\cly, lhat the fonner's
"promiscuity" was disciplined in favor of the "fiunily values" often invoked
in ethnographic prefaces: fieldwork as a pructs..;; of getting along 'vi1h others, of adoption, initiation, learning loc,1! nor111s-11n1ch ;.1s a child lc~1rns.
The habitus of modern fieldwork, delint'cl <tgainst thal of tr;;1vcl, has proscribed interactive 1noclcs long ;1ssociatecl with tra\'tI cxperirntT. Perhaps
the 1nos1 absolute continuing taboo is on sexual Ji;1iso11s. Fieldworkers could
love but not desire the ..objects" of their <Hl<"ntion. On the continuurn of
possible relations, sexual entangle1nents were defined as dangerous, too

Continued on Part 2

Part 2
202

JAMES CLIFFORD

close. Participant observation, a delicate management of distance and proxin1ity, should not include entanglements in which the ability to maintain perspc<:tive rnight be lost. Sexual relations could not be avowed sources of research knowledge. Nor could going into trance or taking hallucinogens,
1hough the taboo there has been sornewhat less strict, a certain amount of
cxperintentation., so1ncti1nes being justifiable in the nan1e of participant
ohservation. Sexual experimentation \Vas, however, out of bounds. A discipli1ad, participant-observer ..went along"' \Vith indigenous life, selectively.
At its inception, though, the taboo on sex rnay have been less against "g<>in~ rrative" or losing critical distance than against .. going traveling," violating a professional habitus. In I ravel practices and texts, having sex, heter<>scxual and hornosexual, with locaI people was common. Indeed in certain
travel circuits, such as the nineteenth-century voyage en Orient, it was quasiohligatory. 111 A popular writer such as Pierre Loti consecrated his authority,
his a(-cess to the mysterious and ferninizcd Other, through stories of sexual
<"nrounter. In fieldwork accounts, however, such stories have been virtually
nonexistent. Only recently, and still rarely, has the taboo been broken (Rabi now 1977; Cesara 1982). Why should sharing beds be a less appropriate
'itllllTe of fieldwork knowledge than sharing food? There may, of course, be
111any practical reasons for sexual restraint in the field, just as certain places
and ~u:tivities may be off-litnits to the tactful (and locally dependent) s<>.inurner. Hut they are not off-lirnits in all places and at all times. Practical con_<;l r<iinls, which vary wiclcly, cannot account for the disciplinary taboo on sex
ill fiekhvork. l!t
Enough has been said, perhaps, to nlake the central point: a disciplinary
hahitus has been sustained around the e1nbodied activity of fieldwork: an
ungendered, unraced, sexually inactive subject interacts intensively (on
hern1eneutic/scientific levels, at the very least) with interlocutors. If actual
experiences in the fteld have diverged from the norm, if the taboos have
so1ne1i1ncs heen hroken, and if the disciplinary habitus is now publicly contested, its nor1native power re1nains.
r\1H)lht'r con11non tnn:el pr<lCtice befOre 1900, cross-dressing, was suppressed
or channeled in the disciplining of modern fieldwork's professional "body."
This is a far-reaching topic, ancl I n1us1 limit n1yself lo preliminary remarks.
J);u1i<'I llefert ( 1984) has written suggestively on lhe history of "clothing" in
r11des c1fEuropea11 tr:otvt>I ohseTvation prior to the nineteenth century. Asu~
s1a111ial, integral link was once assu1ned between the person and his or her
c111nvarcl appearanc-e-liabituJ, in Dcfert's prernodern usage. 20 In a deep sense
ii w;1s understood 1h;1t "clothes n1ake the n1an" ("/,'habit fait le 1noine"). lnteq>rt'l<Hions of habillls, not to he confused with habits (clothes) or with the
later concept of culture, were ;1 necessary part of travel interactions. This in,-h1dtd 1he con1111u11icative n1anipulation of appearances-what might be

SPATIAL PRACTICES

203

called, somewhat anachronistically, cultural cross-dressing. By the nineteenth century, in Defert's account, habitus had been reduced to habits, to
surface coverings and adornment<;; co.stume had ernerged as a deformation
of the richer coustume (a term which combined the ideas of costume and
custom).
Clothes would becoine just one of 1nany clernents in a taxono1ny of observations made by scientific travelers, components of an e1nerging cultural
explanation. Defert perceives this transition in GCrando's scientific advice
to travelers and explorers, published in 1800. Explorers have often merely
described the clothes of indigenous peoples, he wrote. You should go further and inquire why they may or may not be willing to give up their traditional clothing for ours, and how they conceive the origin of their customs
(Defert 1984: 39). Herc, the interpretive grid of habitus is replaced (and made
to seem superficial) by a deeper conception of identity and difference. Travel
relations had long been organized by complex and highly codified protocols, "surface" semiotics and transactions. The interpretation and manipulation of clothing, gesture, and appearance were integral to these practices.
Seen as the outcome of this tradition, nineteenth-century culttiral cross-dressing was more than just dress-up. A serious, comrnunicative play with appearances and a site of crossover, it articulated a less absolute or essential
notion of difference than that instituted by relativist notions of culture with
their concepts of nativeness inscribed in language, tradition, place, ecology.
and-more or less implicitly-race. The experiences of a Richard Burton
or an Isabelle Eberhardt passing as "Orientals," and even the rnore blatantly
theatrical costuming of Flaubert in Egypt or Lotion shore leave, partake of
a complex tradition of travel practices held at arm's length by a 1nodernizing ethnography. 21
Seen from the perspective of fieldwork (intensive, interactive, based in
language learning), cross-dressing could appear only as superficial dress-up,
a kind of touristic slumming. In this view, the practices of an ethnographer
like Frank Hamilton Cushing, who adopted Zun~ dress (and even, it has been
suggested, produced "authentic" indigenous arlifaCL"i), would be somewhat
embarrassing. His intensive, interactive research was not quite "rnodern fieldwork." A similar sense of embarrassment is experienced today by many viewers of Timothy A"iche's film A Man (',a//M Bee, devoted to Napoleon Chagnon 's
research among the Yanomami. 1 am thinking particularly of the opening
shot, which zooms in slowly on a painted, scantily clad figure in a fighting
pose who turns out, finally, to be the anthropologist. Whatever the intent of
this opening, satiric or otherwise (it's not entirely clear). the impression remains that this is not a ..professional" way to appear. A certain excess is registered, perhaps too easily written off as egotis1n. Liza Dalby's book Gei.tha
(1983), which includes photographs of the anthropologist being transformed through makeup and wearing full geisha attire, is more acceptable,

JAMES CLIFFORD

since the adoption of a geisha "habitus" (in Dcfert's older sense-a mode
of being, manifested through clothes, gesture, and appearance) is a central
issue in her participant observation and written ethnography. Yet the photographs of Dalby looking ahnost exactly like a "real" geisha break with established ethnographic conventions.
At another pole are the photographs published by Malinowski (in Coral
(;ardrns and "f'JieJr ftrfagic [ 193_~]) of hi1nself in the field. He is dressed entirely
in white, surrounded by black bodies, sharply distinguished by posture and
<lltitude. This is a 1nan insistently not about to "go native." Such a self-prestntation is akin to the gestures of colonial Europeans who dressed formally
for dinner in sweltering climates so as not to feel they were slipping "over
1he edge." (The miraculous stafchccl collars of Conrad's accountant in Heart
of !Jarkne.~s are a paradign1 case in colonial literature.) But ethnographers
have not, typically, been so fonnal, and I would suggest that their fieldwork
habit us was 1nore of an intermediate formation, predicated on not theatrically standing out fro1n local life (not asserting their difference or authori1y by wearing 1nilitary unifor111s, pith helrnets, and the like), while rernaining clearly 1narkerl by white skin, proximity to cameras, notepads, and other
nonnative accotllreinents. 22 Most professional fieldworkers did not try to disappear into the field by indulging in "superficial" travel practices of masquerade. rheir embodied distinction suggested connections at deeper,
henncneutic levels, understandings forged through language, coresidence,
and n1/Ju.ral knowledge.
l\iorc than a few telling glirnpses of the anthropologist's habitus, overlapping and distinct fro1n that of the traveler, are provided by Levi-Strauss
in rn_~J"s rropiqttes (1973). "In Septemher 1950," he writes, "I happened to
hnd 1nyselfin a Mogh village in the Chittagong hill tracts." After several days,
ht asct'nds to the local te111ple, whose gong has punctuated his days, along
with the sound of ..childish voices intoning the Burmese alphabet." All is innocenct" and order. "We had taken off our shoes to clin1b the hillock, and
1hc fine, damp clay felt sofl under our bare feet." At the entry to the simple,
he;u1tif11l tcn1ple, built on stilts like the village houses, the visitors perform
Mprescribed ablutions," which after the cli1nb through the mud seem "quite
na1ural and devoid of ;u1y re-ligious significance,"
A p< <1ct"ful, baTn-likt' atn1ospht'1-e pervaded the place and there was a srncll of
hI)' in the air. The sirnple and spacious roo1n which was like a hollowed-out
haysta<.k, the co11r1cous behaviour of the two priests standing next to their beds
with straw 1nattrcs.'i<.'S, lhe 1ouching care with which they had brought together
cir 111ade lhe iusrn1111rnt'iofworship-all 1hcse things helped to bring me closer
1h;u1 I had t-vrr hten IK'forc to 1ny idea of what a shrine should be like. '"You
nrrd not do whal I a1n doing." 1ny con1p<tnion said to me as he prostrated himsrlf on the ground four ti111es before the altar, and I followed his advice. Howf'\'fT, I did so lt'ss through self-ronsciousncss than discretion: he knew that I did
0

SPATIAL PRACTICES

205

not share his beliefs, and I would have been afraid of debasing the rilual gestures by letting him think I considered the1n as 1nerc COtl\'entions: bul for once.
I would have felt no e1nbarrass1nent in pcrfonning thcnt. Between 1his fonn of
religion and myself, there was no likelihood of 1nisundcrs1anding. It w;_1s not a
question of bowing clown in front of idols or of adodn~ a suppo.~td supernalural order, but only of paying hon1age 10 1he drdsivc wisd(Hll 1ha1 a thi11kcr, or
the society that created his legend, had e\'olvcd lW('nly-livc f'("llllffics ht'f()I'(' and
to which rny civilization could conlrihutc 011ly h)' confinning it. (197;r 410-.111)

Going barefoot could hardly be a casual gesture fo1 LCvi-Strauss: hut here,
along with ritual cleansing prior to entering the shrine, it seen1s si1nply natural. Everything dra\vs him into syn1pathy and participation. But he 1narks
a line at the physical act of prostration. 'fhe line expresses a specific diJnftion, that of a visitor who looks beyond "rnere conventions" or going along
with appearances to a deeper level of respect based on historical kno\vlcdgc
and cultural con1prehension. The anthropologist's oullhentic bo\v to Bucidhism is a mental one.
LCvi-Strauss is te1npted, retrospectively at least, to pro:<.tnllc hirnselfin the
hill temple. Another anthropologist 1night \veil have done so. My point in
noticing this line between physical and hcrn1cneutic ;u1s of co111uction is
not to clai1n that Levi-Strauss draws it in a pl;1cc typic<1I of;111thropologi.~ts. I
do want to suggest, however, that a siinilar line will he clra\\111 son1cwhrrt',
sometime, in the maintenance of a professional ficlchvork hahitus. Lt,iStrauss is clearly not one of those Western spiritual tra\'elcrs \Vho st~journ in
Buddhist temples, shaving their heads and \Vcaring saffron robes. And in this
he represents the traditional ethnographic nonn. One could, of course, irnagine a Buddhist anthropologist bcco1ning alrnost indistinguishable, in bo1h
practice and appearance, front other adepts during a period of fieldwork in
a temple. And this would be a li1nit case for the discipline, to be uta1ed wilh
suspicion in the absence of other clearly visihle si~ns of profCssional disrrftion (etymologically: a separation).2::i
Today, in many locations, indigenous penplt.\ cthnogri-lphcr.<;, and tourisl.o;
all wear T-shirts and shorts. Elsewhere, distinctions of dress arc 111orc salient.
In highland Guaten1ala it 1nay be a ncces.<iily of dccoru1n, a sign of respect
or solidarity, to wear a long skirt or an embroidered shirt in puhlic. But this
is hardly cross-dressing. Can, should, an anthropologist \Vear a turban,
yarmulke,jallabryya, huipi~ or veil? Local conventions vary. But \vhatever tac
tics are adopted, they are en1ployed from a position of assurnecl n11Jural discretion. Moreover, as ethnographers \vork increasingly in their own societies,
the issues I have been discussing in an exoticisl frarne bccon1c confused, 1he
lines of separation less self-evident. En1borliccl professional practices of .. the
field"-gendered, raced, sexualized locations and crosso\crs, fonns of self~
presentation, and regulated patterns of access, cleparlure, a11<I rc1t1r11-are
renegotiated.

206

SPATIAi. PRACTICES

JAMES CLIFFORD

REROUTING THE FIELD

I havt' 1ricd to identify so111e of the sedi1ne11ted practices lhrough and against
which newly diverse ethnographic project~ struggle fo1 recognition within
an1hropology. Established practices co1ne under pressure as the range of sites
that can he 1rcatecl ethnographically multiplies (the acadernic border with
"tultural studies") and as differently positioned, politically invested scholars
entrr the field (the: challenge of a "postcolonial anthropology"). The latter
d<"\'tlc1pn1e11t has f~tr-reaching in1plications for disciplinary reinvention.
Fieldwork defint?d through spatial practices of travel and dwelling, through
tht" disciplintd, e1nhoclicd intt;ractions of participant observation, is being
nrcitlled by "indigenous," "postcolonial," "diasporic," "border," "n1inority,"
acti\'ist," and ..co1nn1unity-ha.sed" scholars. The terrns overlap, designating
c1n1111l<'x sites ( >f identification, not discrete identities.
Kirin Narayo.111 (1~193} questions the opposition of native and nonnative,
insider ;incl outsider anlhropologists. This binary, she argues, stems from a
discredited, hierarchical colonial structure. Drawing on her own ethnography in different part-; of India, where she feels varying degrees of affiliation
and disrance, Nar;;1yan shows how "native" researchers are con1plexly and
111ultiply lo(<llcd vis-;l-vis their work sites and interlocutors. Identifications
crosstlll, co1111>le1nent, ancl trouble each other. "Native" ;;tnthropologistslike all <tnthropologists, Narayan argues-"belong to sever.ii communities
si 1nultaneously (not least of all the conununity we were born into and the
co1n111unity of professional acadcn1ics)" (Narayan 1993: 24). Once the structuring: opposition between "native" and "outside" anlhropologist is displaced, the relations of cultural inside an<l outside, home and away, same
;uul different that have organized the spatial practices of fieldwork must be
rt'thcntght. flow does the disciplinary injunction that fieldwork involve some
~ort of /nn1t'L-a practice of physical displacement that defines a site or objtt1 of in1ensive research-constrain the range of practices opened up hy
Naravan and others?
ln,Narayan's ilnalysis, fieldwork begins and ends in displacernent, enacted
<ltToss con!'.litutive borders-fraught, an1orous edges. There is no simple,
undivided, "native" position. Once this is recognized, however, the hybriditv she en1hrares needs specification: What are its li1niL~ and conditions of
n;ovt111t111? ()ne can he rnore or less hybrid, native, or ..diasporic" (a term
rh:.tt 1><rhaps htst c;ipturt's Naray<1n 'sown complex locations) for detennin;llt' historical reasons. Indeed, the title of "native" or "indigenous" anthropologist 1nig-hl he rt'l;;lincd lo design;;Ue a person whose research travel
11;1ds out and h;itk fro1n " horne base, "travel" understood as a detour
through a university or other site rhat provides analytic or comparative per~p<'cli\'e on the place of dwelling/research. Here, the usual spatialization
of hon1e and abroad would he reversed. Moreo\'er, for many fieldworkers,

neither the university nor the field provides a stable hasc; rather, both serve
as juxtaposed sites in a 1nobile co1nparativc projr(t. A continuu1n, not an
opposition, separates the explorations, detours, and rc111rns of the indig-<nous or native scholar fron1 those of the diasporir or postcolonial.~ 1 Thus,
the requiren1ent that anthropological fieldwork invoJ\'C .WJllU' kind of lra\'t')
need not 1narginalize those forn1erly callt'd "na1iv<"s . ., l"hc roots and routes,
the varieties of "travel," need to he 11101-c broadly understood.
Recent work by Mary Hchns (198H), l)avid Scott (1~)89), 1\n1ita\' (;hosh
(1992), Epeli Hau'ofa ct al. (19~t~), Ten:sia Tcai\v;1 (1~19:\), B<"n Finney
(1994), and Aihwa Ong (1995), a1nong: others, has rei11forct'd ;i growing
awareness of discrepant travel routcs-lraditions of 1110\'l'llH'nt and inlerconnection not definitively orien1cd hy the "\Ve.st" a11cl an expanding tuJtural-economic world systenl. 'rhese routes follow "traditional" <llHI "1nodern" paths, within and across contc1nporary transna1 ion al and i11tcrrcg:ional
circuits. A recognition of these paths n1akcs space for tr;uel (and fieldwork)
that does not originate in the n1etropolcs ofEuro-A111crica or their ottlposts.
If, as is likely, so1ne form or travel or displaccnH_'llt rcn1ai11s a constinuing ele1ncnt in professional fieldwork, reworking !he "ficld'' 11111st nHa111nultiplying the range of acceptable routes and pracli<Ts.
An attention to the varieties of "travel" also helps claril): ho\v, in the pasl,
cleared spaces of scicntifi< work hilV<' ht't'll co11stit111ecl through lhe suppression of cosnH>politan experiences, especially Ihose of the people under
study. Generally speaking, the localization of "11a1ivcs" 1ncoint that i11te11.~ivc
interactive research was done in spatially dclitnitcd fields and not, for t'X
ample, in hotels or capital cities, on ships, in 111ission schools or 1111i,crsi1 ics,
in kitchens and factories, in refugee can1ps, in di;tsporic 11cighborhoocls, on
pilgrimage buses, or at a variety of cross-cultural silcs of rncounlt"r.~'' As~'
Western travel practice, fieldwork was grounded by a his1oric;1I \'ision, \vha1
Gayatri Spivak calls a "worlding," in which one .section of hu111<111i1y was restless and expansive, the rest rooted and inunobilc. lndigcuous authori1ics
were reduced to native inforn1ants. The 1nargin;11i.lation of lra\'el practices,
those of researchers and hosts, contributed to a donu.\liration of licld\\ork,
an ideal of interactive dwelling lh.at, ho\vcvt'f lc1npo1oiry, tould not he sttn
as merely passing: through. That anthropoloi.,ry's intcrloc111ors often saw thi11gs
differently did not, until recently, dis111rb the discipline's S<"lf~i111;ige.:i 1 i
Alternate for1ns of travel/lield\\'Ork, whC'thcr inclig-t11ous or diasporic,
grapple with rnany problems siniilar to those of coll\'Clllional research: problems of strangeness, privilege, 1niscon1prchensi1>11, slen111yping, <u1d p11li1 ical negotiation of the c11cou11ter. (;bosh is tsprrially 1uncha111 on the potentially violent 111isco1nprchensions and stcrcorypc.~ in1<gr;1l 111 his research
as a doktar al lli11di a1nong M11slin1s. Eptli J lau'ofa speaks for an interconnected "Oceania," but he does so as a To11ga11 lhing in F~ji. a lo(:alinn not
forgotten by his diverse Islander a11dic1ucs. At lht so1111e 1inH', lhr routes and

JAMES <:LIFFORD

t'ntounters of ethnographers such as Ghosh or 1-lau'ofa arc different fro1n

!host of traclitional fieldwork S<~journcrs. Their cultural co1nparisons need
not presuppose a Western, univrrsity ho1ne, a "central" site of theoretical actun1ulation. And while their research encounters may involve hierarchical
rtl<uions, they need not presuppose "white" privilege. Their work n1ay or may
not crucially depend on colonial and neocolonial circuits of information
access, and power. For exatnple, Hau'ofa publishes in Tonga and Fiji and
wants to articulate an old/new ..Oceania." In this he differs from Ghosh
who puhlishes, crucially though not exclusively, in the West The language(s)
fh<' tthno~1-.1phy uses, the audiences it a<ldresses, the circuits of academic
and 1ncdia prestige it appeal~ to, 1nay he discrepant frorn, though seldon1
unconnected with, the co1nnn1nicative structures of global political econ01ny. A case in point: A Nnv ()rf'<Utia, hy llau'ofa ct al. (!!)!).'~).was delivered
to 1ne by hand.'0 Published in Suva, the hook would not have reached n1c
through 1ny l\(1rn1al reading: networks. Can a work centered and routed like
1hisone intervent" in Euro-An1erican anthropological contexts? \Vhatare the
i11stitu1ional boirriers? The power to detern1ine audiences, publications, and
translations is very unevenly distributed, as Talal Asad has often reminded
us (Asad 1981>).
rhe oxy1noronic 1ern1 "indigenous anthropologist," coined at the beginning: of the ongoing postcolonial/ neocolc>11ial recentering of the discipline,
is no longer adequate to characterize a wide range of scholars studying in
tlltir ho1ne societies. Difrlcult issues arise. How exactlywill .. home" he defined?
If, <ts I assun1e, no ;,dvrrnt authority can be accorded to "native" ethnographi(s and histories, what constitutes their differential authority? How do they
s11pplc1nent and criticize long-established perspectives? And under what conditions will local knowledge enunciated by locals be recognized as "anthropnlc1g:ical kno\vledg:e"? What kinds of displace1nent, comparison, or taking of
c1isltnce" are required for fan1ily knowledge and folk history to he recognilrd as."ieriotl'i e1hn0h'riif>hy or cultural theory by the disciplinary center.>
Anthropoloh'Y potenlially includes a ca~t of diverse dwellers and travelers
\:hose cli~pl<1c~n1en.t or travrl in .. fieldwork" differs fron1 the traditional spat 1al prac11re of 1he field. The \\lest itselfbeco1nes an object of study from vari1 t11sly distanced and entangled locations. Going ..out" to the lield now someti111<"s 1neans going ..hack," the ethnography becoming a "notebook of a
return to the native land." In the case of a diasporic scholar, the .. return"
1nay ht" to a place Jl<'Vt'r known personally hut to which she or he a1nbival_t11tly, powerfully .. belongs." lft>turning to a field will not be the same as gtr
111K oul to a field. l>iffere11t sul~jcctive distdnces and affiliations are at stake.
A growing aW<ll'eness of the~ clifTerenct!s has emerged within Euro-A1neric;u1 an1hropolot-,'Y during rt>cent decades. In an important discussion, David
Sroll na111ed S<>n1e of the historical locatio11s constraining an emergent "postroloniality" in anthropoloJ.,ry:

SPATIAi. l'RAC:TIC:ES

By raising in difTerent ways the problc1n of "l>late" and the non-Wtstern an-

thropologist, both Talal Asad (19!-h~) and A~j1111 Appad11n-1i (19KKa) have sug
gested that to undennine the asynunetry in anthropological praclirt 1nany
1nore such anthropologisL'i should study Wcsl(rn socictirs. This, 10 be sure, is
a step in the right direction inasn1uch as it suhvC'l'IS thr perv<isivr notion 1h;l1
the non-\Vestcrn su~jecl can speak only within the 1crn1.~ of his/ hl'r own culture. Moreover, it privileges in son1e clcgrre the possibility of lacking h;ick and
forth between cultural spaces. At the sa1nc ti111c, it would scent to fix and repeal the colonially establish('d tf"rritorial bo1111dari<'s within which 1he post
colonial is encouraged to 1nove: centerI periJ>h<ry-and typically, tht" cen1tr
of neocolonial governance and the periphery of 11rigin. European ;1ncl A1nerican anthropologists continue to go wl1cre lh<y pl(asr, while the p11stnloni;1I
stays ho1ne or else goes \Vest. ()ne wonders whether thtn 1night not he a 11101-e
engaging problc1natic to be encountered wlure the postnilonial i11ttllectual
fiu1n Papua New Guinea g()(__'S, not tu Philadl'~Jhict hut 111 Boin h.:1~01 Kingston
01 Accra. (Scott 1989: Ho)
Escape from the polarizing historical f{>rce field of lhe "\Vcsl" is no easy 1natter, as Scott's subsequent discussion of (;hosh 1nakes cle.ir. But Scoll also argues that the cross-cultural "tacking" of anthropologists should not be reduced to movements between centers and peripheries in a \vorld sysletn.
Contemporary ethnography, including Scott's own fron1 Jan1aica via Ntw
York lo Sri Lanka, is necessarily "traveling in the West" ({;hosh, quoted hy
Scott 1989: 82). It is also traveling in and against, through the West.
Ethnography is no longer a nor1native practice of outsiders visiting or
studying insiders but, in Narayan's words, of attending to "shifting identities
in relationship with the people and issues a11 anthropologist seeks to rcpr<'sent" (Narayan 1993: 30). How identities are negotiated relationally, in determined historical contexts, is thus a process constituting both the sul~jtcts
and objects of ethnography. Much emerging work now tnakes these complex
relational processes explicit. Paulla Ebron (1994. t<>rthcon1ing), for CX<llll
pie, conducts research on Mandinka praise-singers hoth in \Vest Africa :uul
in the United States, where they find apprt~ciative audiences. I Irr elhnog:raphy is multiply located and-as she clearly Shows-cnlanglcd in the 1r<1\'
cling culture circuits of world 1nusic and 1ourism. Ehron's ethnography alsc1
works in tension with a history of clo1ninant Weslern inv<"ntions of Afric.ashe cites Mudimbe (1988)-and more or less rcnnantici1.C'd African A1nerican projections formed in reaction to historiC's of racistn. Ehron 1110\cs anltlng:
these intersecting contexts. "Africa" cannot he held ..out 1herc." It is an c1npowering and problematic part of her own African Arnerican tradition as
well as a relay-not an origin-in a contint1ing diasporic hislory of transils
and returns. This history implicates her acaclc1nic ethnography, whose sile
is the relational negotiation of ..sul~ject'i in diflCrcnce," a spac<" where praisesingers, tourists, and anthropologists clai1n and negotiate cultural n1eaning:s.
Her field includes the airports where thC'se travelrrs cross.

210

JA~tES

CLIFF<)RD

"Indigenous," "postcolonial," "diasporic," or "minority" attachments are

lrtc111t111ly at is."ille in the way anthropological "fields" are negotiated. Scholars such as Rosaklo (HJH!)a), Kondo (1990), Behar (1993), and Lim6n
(1994), to cite 011ly a few, define the spatial practices of their fieldwork in
ttTllls of a polil ics of locations, of tactically shifting insides and out<iides, afliliations and distances. Their anthropological "distance" is challenged,
blurred, relationally rccons1ruc1cd. Often. they express their complex situaled k1u>\vledges by lextual strategies in which the embodied, narrating, travrling s<:holar-1heorist is pron1inent. But this choice should be seen as a criti(;i] i111trvc11tion a).{ainst <liscn1bodied, neutral authority, not as an en1erging
nonn. rherc is no narrative forrn or way of writing inhcre1nly suited to a
politi<s of local ion. Others worl<.ing within and against a still predorninantly
\\'lst<"rn a11thropoloh')' 1nay choose to adopt a 1nore irnpersonal, den1ystifyi11g, indttd ol~jectivc rhetoric. David Scott and Talal Asad are strong exainplt's. rhcir discourses are, nevertheless, openly that of politically co1nmit1cd, situated scholars, not neutral observers. A very wide range of rhetorics
a11<l narrati\'es-personal and i1nperson<ll, objective and subjective, e1nhodicd and dise1nho<lied-are ;1vailable to the located scholar-traveler. The
0111~ tactic excluded, as Donna I faraway f1as said, is the "God Trick" (Harawar 1988).
!\lost of the anthropologists cited in the previous section have <lone so 1ne1hing like tradition;1I liel<hvork: studying ..out" or "down." This has contrihutecl to their survival, ind<~cd success, within the acade1ny, even as they
'vork to critici1.c and open it up. The licensing function of having done "real"
fieldwork-intensive and displaced fro1n 1t1c universi1y-ren1ains strong. Indet'd, <thnography that takes place "'ithin diasjwric affiliations may be more
(';1sil~, accepted than research whose attac}Hnents, however arnbivalent, are
i11(li).{enous or '1"tir1t'. (Recall that these locations fall on an overlapping con1i11ut11n, 11c1t ( ) I \ eitl1er side ofa binary opposition.) Diasporic (dis)locations
h;1ve tra\'el and distance built into thc1n, usually including metropolitan
.'i1>acrs. Native (re) 1',cations, while they i11c:lude travel, are centered in a way
th;1t 111akt"s the ntttropole and the university peripheral. I have suggested
1h;11 clisplace1nent, Scott's "larking" het\veen cultural spaces, ren1ains a constitulive featl1re ,,f <1ntl1ropological flelchvork. Can this displacement be extt'nclt'd to includt' tr;1vrl to ;ind through the university? <:an the university
il.'itlf hr seen a~ a kind of field site-a place of cultural juxtaposition, estra11ge1ntnt, rile of passage, tr:ansit, and learning? Mary John (1989) opens
st1ch ;1 possibility in her prescient discussion of a co1npro1nised, emergent
'\u11hropolohry in n.'\'erst:'" for postcolonial fe1ninists: a coerced and desired
travt'I in .. lhe \Vest," and an unstable coexistence of roles-anthropologist,
a11d 11<1tivc info11nant. I low docs travel through the university reposition the
.. 11011ive" place \Vhcre the anthropologist rnaintains connections of residence,

SPATIAL PRACTICES

211

kinship, or political affiliation that go beyond visiting, however intensive?

Angie Chabram explores this repositioning in her provocative sketch of a
Chicana/a "oppositional ethnography" ((:habra111 1990). llere, "minority"
and "native'' trajectories may overlap: rooted in the "community" (hov,1eve-r
defined) and routed through acaden1ia.
When ethnography has pri1narily served the interests of con1nn1nity
memory and mobilization and only secondarily the needs of ron1para1ivc
knowledge or science, it has tended to be relegated to the less prestigious
categories of "applied anthropolohry," "oral history." "folklore," "political j1n1rnalism," or "local history." But as fieldwork becomes diflerently rooted and
routed in so1nc of the ways I have been tracking, rnany schola1s 1nay lake a
renewed interest in applied research, oral his1ory, and folklore, stripped IHI\\'
of their sometiincs paternalistic traditions. l'hc oral historyI co1n1n11nity llHlbilization work of the El Barrio Pr(~ect at 1hc New York <:entro de Estudios
Puertorrique1los is a frequently cited exan11Jle (see l~e11n1ayor 1q91; (;onion
1993). Dara Culhane Speck's .An Hrrorinjudf(t'llU'1Jl (1987) carefully fuses co1nmunity memory, historical scholarship, and current political ad\'Ocacy. Esther Newton's subtle articulation of 1nargi11s, as loyal lesbian participant-observer, outsider/insider in a predominant!)' gay n1alc conununity, proch1<-cs
an exemplary fusion of local history and ctiltural criticis1n (Newton 19~}:\a).
Epeli Hau'ofa's research in Tonga is another case in point (as distinct front
his cxoticist work in Trinidad or his studies in Papua Ne\\' (;uinC'a, where he
was a different kind of "Pacific" outsider). Returning to do 1csearch in his
native Tonga, Hau'ofa writes in more tharl one language and style both to
analyze and influence local responses to Westerni1.ation. I-le 1naintains a stylistic distinction between writing for the discipline, writing <1s politital inl('J"vention, and writing as satiric fiction (Hau'ofa 1982). But the discourses are
clearly connected in his view, and others 111ight be n1ore inclined than he lo
blur them.
To do "profe-ssional" anthropology.one tnust 111aintain connections "'ith
university centers and their circuit<; of pul>lication and soriality. llo\\" close
must these connections be? How central? \\'hen dots one begin lo lose di~
ciplinary identity at the margins? These <1uestions ha\'C always fatcd scholars working for governments, corpor:.:1tions, acti\'ist social organi1.ations, and
local communities. They continue to trouble, and cliscipline, the work c,f the
differently located anthropologists I have been discussing. lvforeo\'cr, the university itself is not a single site. Though it 1nay ha\'e \.Vestcrn roots, it is hybridized and transculturated in non-Western places. Its ties to nation, lo "development," to region, to post-, neo-, and anticolonial politics can make it a
significantly different base of anthropological operations, as 1-lussein Fahin1 's
pioneering collection Indigenous Anlhro/H1lof!J in Non-\Ve.~tern (:Ounlril's (1982)
makes clear. In principle at least, universities are sites of con1parative theory,
of communication and critical argument a1nongscholars. The ethnographic

,,,

JAMES Cl.IFFl>RD

or ethnohistorical interpretations of non-university authorities are seldom

recognized as fully scholarly discourse; rather, they tend to be seen as local,
arnateurish knowledge. In anthropology, the research that produces such
k110\vledgc, howt~ver intensive and interactive, is not fieldwork.
rhe disciplin.1ry "Other"who perhaps most epito1nizes the border at is!'illt' here is the figure of the Local historian. This supposedly partisan chronicler and keeper of 1he co1n1n11nity's records is even harder to integrate with
conventional fieldwork than the e1nerging figures of the diasporic postcolonial, the oppositional n1inority scholar, or even the traveling native.
r.1i111ed hy a prcloi11111ed inunobiliry ancl by assu111ptions of a1natcurisrn and
hoo.'ittrisin, lhe local historian, like the ;1ctivist or culture-worker, lacks the
required professional ..distance." As \Ve have seen, this distance has been nat11ralized in spatial practices of lhe "field," a circu1nscribed place one enters
and lcilVCs. Movc1ncnt in and out has been considered essential to the inttrprcti\c procc~. the 1nanagen1ent of depth and discretion, absorption and
"'tht view fron1 afar" (LCvi-Strauss 198_r,).
rhe disciplinary border that keeps locally based authorities in the position of inforn1an1s is, however, being renegotiated. Where and ho\V the
boundary is redrawn-which spatial practices will be accornmodated by the
('\olving 1n.1dition of anthropological fieldwork and which will be excluded1Tn1ains to be seen, But in this context it ntay be useful to ask how the legacy
ol"I itldwork-a.s-tr.tvel helps to account l<>r an issue raised during recent prcsidt'ntial sessions on diversity at the A111erican Anthropological Association:
lhc fact that North A1nerican 111inorities are entering the field in relatively
s111all nu1nber.s. Anthropology has difficulty reconciling goals of analytic disl<ulct \~ith the aspirations of organic intellectuals. Has the discipline adeq11a1tly confr<nllcd the proble1n of doing sanctioned, .. real" fieldwork in a
con1nn1nityone wanL'i uut to leave? Deparu1re, taking distance, has long been
crucial to the sp~llioil practice of fieldwork. How can the discipline make room
for rc.st>arch that i'i i1npor1antly about return, reterritorialization, belonging
-attach1ncnts that go beyond gaining rapport as a re.search strategy? Robert
Ahartz (19~).I) provides a revealing discussion of these issues, showing how
dil h.rcnt kinds of <'Onununity involve1ne11t in the course of research are valuc.cl and devalued by the discipline in ways that tend to reproduce a white
l1egr1nony.
rhe <lcllni tion of "ho1ne" is fundarne111ally at issue here. In local/global
situations whtr< displact"n1cnt appears increasingly to be the norm, how is
tolltr1ivc dwt.'lling: sus1ained and reinvented? (See Bammer 1992.) Binary
oppositions hctWt't'll ho1nt and abroad, staying and moving, need to be thorou~hly <}ll{"Slioned (K.aplan 199.J). These oppositions have often been naturali1.ed along lines of gender (fe1nale, do1nestic space versus 1nale travel),
<lass (the activt, alienated bo11rgeoisie versus the stagnant, soulful poor), and
r;1re/ruhure (n1odern, rootless \\'esterners versus traditional, rooted "na-

SPA'flAI. l'RACTU:Es

21)

tives"). The fieldwork injunction to go elsewhere construes "home" as a site

of origin, of sameness. Feminist theory and gay/lesbian studies have, perhaps most sharply, showed ho1ne to be a site ofunrestful differences. Moreover, in the face of global forces that coerce clisplaccn1cnt and travel, st;ying (or making) home can be a political act, a forrn of resistance. I-lonte is
not, in any event, a site of immobility. These few indications, of which 1nuch
more could he said, should be enough to question anthropological assumptions of fieldwork as travel, going out in search of 11ijfm'nrf'. To a degree
these assumptions continue to apply in practices of "repatriated" fieldwork
(Marcus and Fischer 1986) and of"studying up" (Nader 1972). The field ren1ains somewhere else, albeit within one's own linguistic or national context.
An unsettling discussion of "ho1nc" with reference to anthropological
practice is provided by Kan1ala Viswesw.tran (1994). She argues thott fe1ninisl
ethnography, part of an ongoing struggle to decolonize anthropology, needs
to recognize the "failure" that is inevitably bound up \Vi th the pn>jcct of crosscultural translation in power-charged situations. Precisely at "those 1non1e11ts
when a project is faced with its own i1npossibility" (98), ethnography can
struggle for accountability, a sense of its own positioning. Building on (;<tyatri Spivak's formulation of every cultural/political sul~ject's "sanctiontd
ignorances," Visweswaran argues that by openly confronting failure, fc1ninist ethnography discovers both li1nits and possibilities. A1nong the latter arc
critical moven1ents "hon1eward." In a section titled "l--lo1ncwork, Not Fieldwork," she develops a concept of ethnographic work not based on the
home/field dichotomy. "Hon1ework" is not defined as the opposite of c.xoticist fieldwork; it is not a matter of literally staying ho1nc or studying one's
own community. "Ho1ne," for Visweswaran, is a person's location in determining discourses and institutions-cutting across locations of race, gender, class, sexuality, culture. "llo1nework" is a critical confrontation with the
often invisible processes of learning (the French wordfurmalion i.s apt here)
that shape us as su~jects. Playing on the pedagogical senses of the lernt,
Visweswaran proposes "ho1nework" as a discipline of unlearning as much as
of learning. "Home" is a locus of critical str\1ggle that both empowers and
limits the subject wherever she or he conducts forn1al research. By deconstructing the home/field opposition, Visweswaran clears space for unorthodox routings and rootings of ethnographic work.
In a related, but not identical vein, Gupta and Ferguson (chapter 1) urge
an anthropology focused on "shifting loralions rather than bountlf'd jiPlds."
Theirs is a reformist rather than a deconstructive prc~ect. While rejecting
the tradition of spatially restricted research, they preserve certain practices
long associated with fieldwork. Anthropology still studies "Others" intensively and interactively. It provides, they rt"ntind us, one of 1he fe,.,. Weslern
academic sites where unfamiliar, 111arginalized, nonclite peoples are seriously attended to. Long-term i1n1nersion, interest in inforn1al knowledge

JAMES CLIFFORD

and embodied practices, an injunction to listen are all elements of the fieldwork tradition they v.ilue and hope to preserve. Moreover, Gupta and Ferguson's nolion of shifting locations suggests that even when the ethnographer is positioned as an insider, a .. native" in her or his community, some
taking of distances and translating differences will be part of the research,
analysis, and writing. No one can be an insider to all sectors of a community. I-low the shifting locations are rnanaged, how affiliation, discretion,
and critical perspective are sustained, have been and will remain matters
of t<lctical i1nprovisation as 1nuch as of formal methodology. Thus, whatever
ccnnes to be recognized as a refonned fieldwork will entail David Scott's
"tacking between cultural spaces," though not necessarily or solely along
colonial or neocolonial axes of center and periphery.
Moreover, the constitutive displace1nents need not be between "cultural"
spaces, at least not as the term is conventionally defined in spatial terms. An
ethnography focused on shifting locations would assume only that the borders negotiated and crossed were salient to a co-constructed project in a specific "cunt.act zone" (Pratt 1992). This would mean not that the borders in
question were invented or unreal, but only that they were not absolute and
could be crosscut by other borders or affiliations also potentially relevant to
lhe project. These other constitutive locations might become central in other
historical and political co1~junc1ures or in a differently focused project. One
cannot represent "in depth" all salient differences and affinities. For example. a 1niddle-class researcher studying a1nong working people may find class
to he a critical location, even if his or her research topic is explicitly focused
elsewhere-say, on gender relations in secondary schools. In this case, race
n1ight or 1night not be a site of crucial difference or affinity.
A project will always "succeed" on certain axes and "fail" (in Visweswaran's
constitutive sense) on others. Thus, we should not confuse a more or less
conscious research strategy of shifting locations with being located (often antagonistically) in the ethnographic encounter. For an Indian Hindu working in Egypt, religion 1nay be imposed as a prime differentiating factor, asserting its salience for a research project on agricultural techniques, in spite
of the author's desires (Ghosh 1992). Moreover, the process need not be antagonistic. A student of his or her own community may be located firmly and
loving!}' as .. farnily,,. thus putting real restrictions on what can be probed and
revealed. A gay <)r lesbian ethnographer 1nay be constrdined to highlight or
downplay sexual location, depending on the political context of research.
Or an anthropologist from Peru may find himself or herself negotiating a
national boundary when working in Mexico, but a racial one in the United
Si ates. The ex~unples could be 1nultiplied.
None of these locations is optional. They are imposed by historical and
political circun1stances. And because locations are multiple, conjunctural,
and cros.OR:utting, there can be no guarantee of shared perspective, experi-

SPATIAi. PRACTICES

ence, or solidarity. l build here on a nondis111issive critique of identity politics that has been compellingly stated by June.Jordan (198!)) and developed
by many others (for exa1nple, Reagon t983; ~1ohotnty 1987). In ethnography, what was previously understood in tcr1ns or rajJ/HJrl-a kind of achie\cd
friendship, kinship, e1npathy-now appears as so1ncthing closer to atlianrr
lrnilding. The relevant question is less "What fund;unenlally uniles or separates us?" and 1nore "What can we do for each other i11 the present conjuncture?" What, fro1n our si111ilaritics and diffen_JH'.cs, can
hend toge! htr.
hook up, articulate? (See l-lall 1986: r}2-r>rl; 1-lannvay 199~: ;\ofl-;\1;,.) A11d
when identification becu1nes too close, htnv can a disarticulation of <lgt'ndas be managed, in the context of alliance, without resorting to clairns to
objective distance and tactics of definitive departure? (For a sensitivt' account
of these issues in the context of lesbian ethnography, M'C Le\vin 199!).)
A stress on shifting locations and tactical affiliations explicitly recogni1_es
ethnography's political din1ensions, di111ensions th;1t can be hidden by presumptions of scientific neutrality and h11111an rapporl. Hui "political" in whal
senses? There are no guaranteed or n1orally unassailable positions. In the
present context-a shift fro1n rapport to alliance, fron1 representation to
articulation-rigid prescriptions of advocacy have a lendency to enu1)~-r. An
older politics of neutrality with ils goal of ulti1natc discngage1nent 1n;1y si1nply be reversed-a binary starkly evident in the juxtaposition of eloquent,
opposing essays by Roy <l'Andraclc and Nancy Schepcr-1 h1ghes in a 19~t!i fortun of Current AnthropohJf!J. The place for a politics of skcpli<is1n and <Titique (nol to be confused with dispassion or neutrality). for engaged disloyalty, or for what Richard lIandler ( 1985, quoting Sapir) calls "des1rurtivc
analysis," seems endangered. An alliance 111odel l<avcs little roo1n for work
in a politicized situation that pleases none oflhc contcst;inls. I a111 not suggesting that such research is superior or n1ore ol~ject iv<'. It, tcu>, is parlial and
located. And it should not be excluded fio1n the range of sin1ated research
practices now con lending for the 11an1e "anthropology."

'''l'

These are just some of the dilernmas f~1cing anthropological cthnogroiphy as

its roots and routes, its different pallerns of affiliation and displacc1ncnt, arc
reworked in late twentieth-century contcxls. \r\-'hat 1e1nains of firU/Jt)(1rk? V\11a1,
if anything, is left of the injunction to travel, tu get out of the house, to .. enter the field," to dwell, to interact intensively in '' (relatively) u11fa1niliar rontext? A research practice defined by ..shifting: loc;llions," without the prescription of physical displacc1nent and extended fo1tc-to-face e1lCot11Her,
could after all describe the work of a literary crilic, attcntiv<, as n1anr ar<' lo
day, to the politics and cultural contexts of dif!Crcnt tcxlutl readings. Or,
once freed of the notion of a "fieldn as a spatialized site of rese;_irch, could
an anthropologist investigate the shifting loc,1tiuns of her or hi.., own life?
Could "homework" be autobiography?

216

JA~1ES

<:LIFFORD

Herc we cross a blurred border that the discipline is struggling to define.

1\utobiography can, of course, be quite "sociological"; it can move systematically between personal experience and general concerns. A certain de~ree of autobiography is now widely accepted as relevant lo self-critical projects ofci1ltural ~lnalysis. But h1,w n1uch? Where is the line to be drawn? \\1hen
is self-analysis dis1nissecl as "1nere" autobiography? (One son1eti1nes hears
rather 1nodcst an1011nts of personal revcl<Hion in ethnographics described
as solipsisrn or "n~1vcl-gazing.") \Vriting <tll ethnography of one's subjective
"P4t(l' as a kind of co1nplex con11n11nity, a site of shifting locations, could be
<ltf<11<lcd as a valid contribulio11 to anthropological work. It would not, I
I hink, he widely n.cogniLed as fully or characteristically anthro/Jological in the
way that \\'ork in an txtcrnalized jif'ld still is. One could hardly count on being ;iwarded a Ph.D. or finding ajoh in an anthropology clcparunent for autuhiographi<al r('sc;1rch. rhc lt'gacy of the field in anthropolohiy requires,
;II least, that "firsthand" n_search involve extended face-to-htce interactions
with 1nc1nbcrs of a c11nununity. Practices of clisplaceinent and encounter still
play a delining role. \'Vithout these, what are under discussion are not new
\'trsions of fiel<lwork hut a range of quite different practices.
In this chapter, I have tried to show hcnv definite spatial practices, patterns of dwelling and traveling, have constituted fieldwork in anthropology.
I ha\'C' argued that lhe dis<iplining of fieldwork, of its sites, routes, temporalities, and cn1bodied practices, has been critical in 1naintaining the idenlity of sociocultural anthropoloh'Y c:urrently contested and under renegotiation, fieldwork re1nains a 1nark of disciplinary distinction. The rnost
disputed element~ of traditional fieldwork are, perhaps, il~ i1~junction to leave
honH' and its inscription within relations of travel that have depended on
colonial, race-, class-, and gender-based definitions of center and periphery,
r<1s1nt>polilan and local. The linked require1nent 1hat anthropological fieldwork be intensi\'e and interactive is Jes.~ controversial, although criteria for
n1easuring: ..clrpth" are 1nore debatable than ever. Why not sin1ply purge the
discipline's t"Xoticist travel legacy while SLlstaining its intensive and interac1ivr Myles of research? In a tllopian 1nodc one nlight argue for such a solu1ion, and indce-d thiu~s seen1 lo he rnoving in this genC'r<tl direction. A radical course is urged by L>eborah D'An1ico-Sa1nuels in an essay Ihat anticipates
111any of the critiques previously referred to. She questions traditional spatial and n1ethodological definilions of the .. field," concluding rigorously:
...fht field is everywhere'' (1991: 83). But if the field is ever}'Where, it is
1H>\~here. \Ve should nc1t he s11rprised if institutional traditions and interests
r<'sisl such radical dissolutions of fieldwork. Thus, sorne forn1s of travel, of
disciplined displace1nent in and out of one's "community" (seldom a single
pl.ice, in any event). will prohably re111ain the norm. And this disciplinary
"travel" will require at least a serious sojourn in the university. I conclude,
pro\ocatively, in this hazardous future tense.

Sl'ATIAl. PRAC:TJ<:ES

217

Travel, redefined and broadened, will re1nain constitutive of fieldwork,

at least in the near ter1n. This will be necessary for institutional and malt'rial reasons. Anthropology must preserve' not only its disciplinary identity
but also its credibility with scientific institutions and funding- sources. GivC'n
a shared genealogy with other natural science ancl social siencc research
practices, it is no accident that the field has, at ti1nes, been called anthropology's .. laboratory." (;riteria of objectivity associated with a dctalched, outside perspective are strongly represented in 1he acadC'1nic and g:ovC'n1n1C'nt
milieus that control resources. Thus, socicirultural antl1n1poh1f..,')' will ren1ain
under pressure 10 certify the scientific cr("clentials of an intC"raclive, inter
subjective rnethodology. Researchers \Viii he constrained to 1;1ke a cert;iin "distance" front the communities Ihey study. Of course, cri1ical distance can he
defended without appealing to ulti1nate grounds of authority in scientific
objectivity. At issue is how distance is 1nanifts1ed in research practices. In
the past, physically leaving the "field"-to "write up" resc<1rch results in the
presumably n1ore critical, o~jective, or <.tt least co1nparativc t11viron1nent of
the university-was seen to be an important guarantee of acadc1nic inclrpendence. As we have seen, this spati;1lization of "inside" and "outside" locations no longer enjoys the credibility it once did. Will oullhropology flnd
ways to take seriously new forn1s of"field" research that diverge fron1 earlier
models oft1niversity-centcred travel, spHial discontinuity. ;ind 11hin1ate clis
engagement?
As anthropology moves, haltingly, in postcxoticisl, postcolo11ial dirC'ctious,
a diversification of professional norms is under \Vay. 'fhe process, accC'lC'r
ated by political and intellectual critiques, is rei11f(>rced by material constraints. In many contexts, given falling levels of funding, sociocultural titld
work will increasingly have lo be conducted "on lhe cheap." For gradualr
students, relatively expensive long-term S<~journs abroad 1nay he out of the
question, and even a year of full-time research in a U.S. connnunity can be
too expensive. While traditional fieldwork will cenainly 1nain1a1n its prestige.
the discipline nlay come to resemble nH>re ~losely the "nation;1I" anthropologies of many European and non-Western countries, with short, repeated
visits the norm and fully supporled research years rare. It is i1nportant to recall that professional fieldwork in the Malinowskian tnold depended 1naterially on the mobilization of funding for a new "scientific., practice (Stocking t992b). "Subway ethnography," like Karen Mc(:arthy Bnn~ns (discussed
above), will be increasingly co1n1non. But <'Ven as visiting and "deep hanging out" replace extended coresidence and the rent-in-the-village 111odel, legacies of exoticist fieldwork influence the professional hahitus
the "field"
-now conceived less as a discrete, other place than as a set of e1nbodied research practices, as patterns of discretion, of professional distance, of coin

or

ing and going.21-1

..
I have located fieldwork in a long, increasingly contested trad111on of

srATIAL PRACTICES

JAMES CLJFl-"ORD

\'Vestern travel practices. I have suggested, too, that othe1 travel traditions

and diasporic routes can help renovate methodologies of displacement,

l<.'ading to 1neta1norphoses of the "field." '"fravel" denotes more or less voluntary practices of leaving fa1niliar ground in search of difference, wisdom,
power, adventure, an altered perspective. 'fhese experiences and desires cannol be li1nited to privileged 111ale Westerners-although that elite has powtrfuly de lined the tcrn1s of travel orienting modern anthropology. Travel
nreds to be rethought in different traditions and historical predicaments.
~fcirto\"cr, \\'hen criticizing specific legacies of travel, one should not co1ne
to rest in an uncritical localis1n, the inverse of exoticis1n. l'here is truth in
the cliche "travel broadens."\!9 Of course, the experience offers no guarantted r(suhs. But, often, gettiilg away lets uncontrollable, unexpected things
h;1ppen Cl'sing 1994a). An anthropologist friend, .Joan Larco1n, once told
111t ruefully and gratefully: "Fieldwork gc1ve me so1ne experiences I didn't
1hi11k I deserved." I re1nen1hcr thinking that a discipline requiring this of its
aclPpts 1nust he onto so1ncthing. Is it possible to validate such experiences of
displaccn1ent without reference to a 1nystified, professional "rite of passage"?
S<~journing so1newhere else, learning a language, putting oneself in odd
situations and tryi11g to figure the1n out can be a good way to learn something new, siinultaneously about oneself and about the people and places
one visits. This conunonplace truth has long encouraged people to engage
with cultures beyond their own. It underlies what still seems most valuable
in the linked/ distinct traditions of travel and ethnography. Intensive field"ork does not produce privileged or co1nplete understandings. Nor does the
cultural knowledge of indigenous authorities, of "insiders." We are differently situated as dwellers and travelers in our cleared "fields" of knowledge.
Is 1his n1uhiplicity of locations 111erely another symplon1 of postmodern
frag1nentation? Can it be collectively fashioned into something more substantial? Can anthropology be reinvented as a forum for variously routed
fit'ldworks-<l site where different contextual knowledges engage in critical
dialogue ;:tnd respectful pole1nic? Can anthropology foster a critique of cullural don1inanc.:e that exlends to its own protocols of research? The answer
is unclear: powerful, newly flexible, centralizing forces remain. The legacies
of the "field" are strong in the discipline and deeply, perhaps productively,
a111biguous. I have focused on son1e defining spatial practices that must be
turned to new ends if a 1nultiply centered anthropolob'Y is to emerge.
NOTES

Th;inks to the following pc-ople for criliral readings: Judith Aisse11,Ja1nes Fergu
son, Akhil (~upla, Susan I larding, Michelle Kisliuk, Ann Kingsol\'er, William Ladu.;aw, and ();1\'itl Schneider.
1. For tht t111ergtnct of this fieldwork norn1 and its Mmagic," see George Stocki111.(s cla.~<;ir accou1H (Stocking 1~19~a: chapter 1). My discussion here is largely lin1-

"9

ited to Euro-American trends. I join Gupta and Ft."rguson (chapter 1 of this book)
in admitting my"sanctioned ignorance" (Spivak 1988;Jnhn 1989) of1nany non-\Veslern anthropological contexts and practices. And t'vcn within 1lu. contrsted hu1 powerful disciplinary "center," rny discussion is pri111arily ftiruscd 011 Nc1nh A1nerica ;111d,
to some extent, England. If the issues rais1;1d extend heyond 1hc:i;c nnHcxts, !hey do
so whh reservations I an1 not yet able to discus.-; sys1e1natically.
2. Renato Rosaldo 1nade this n1n1111e11t at thr "Antl1rop11l11J..'Y ;u1cl '1l1e Fielcl'M nHJference on 18 April 1994. The context was a coniparison 11f cthnoJ.\Taphy by postexotic anthropologists and cultural studies scholars, a discussion of wh;it, in the ;ihst'IHT
of extended coresidence, guarantees interac1i\'e ..<l(pth."
3. In his recent survey of cntt.'rging "n1uhi-sitcd e1h11ography." c;rorge ~1anuo;;
(1995: 100) confronts this question and aq~ues that such rlhnographirs are -inevitablythe product of knowledge h<tses of varying intensi1ies ancl q11ali1iC's." I le adtls:
"It is perhaps anthropologists' apprecitttion of the difficuhy of doing inttnsi\e
ethnography at any site and the sa1isfaclion th;tl cotnts fro1n such work in 1hc pas1
when it is done well that would give then1 pause when the ethnogr:1pher btcon1ts
mobile and still clailns to have done good fieldwork." ()verall, Marcus'!ii iinportant
attempt to grasp an emergent phenon1eno11 hypas!iies the question ofjip/dwr1rk. lie
simply calls all the new mobile prac1ices elhrw;~raJ1h_v. a 1nanifcstly interdisciplinary
orientation, albeit retaining certain recogni1.ahle anthnpol1gic:1l featuns: 11pclose perspectives, cross-cultural translations, language learning, atten1ion lo e\"t'r~
day practices, and the like.
4. Criteria of adequate fieldwork have te1Hkd lo be enforced through 1aci1 nn1sensus rather than explicit rules. A prolCssional culture recog11i1.es "good" ethnography and ethnographers in ways that can appear nhscure, even arbitrary, 10 <1n otlf
sider. I am not concerned, however, with dis1inguishing research of different qualit~
or with showing how such distinctions function professionally. This would require a
history and sociology of the discipline that I an1 nol qualified 10 supply.
5. A single offering that would auen1pt 10 i1itegrate current work in physical :u1thropology and archaeology is barely conceivable. Most departnun1s suslain separate tracks with hopes-more or less serious-of cros.o;.fertili1a1io11.
6. It is a fraught border that, in the United States al leao;1, can lake 1hc for111 ol
turf wars. On the anthropological side, there has been recurrenl gnnnblin~ about
misuse of the culture concept and superficial eth~ography. Moreo\er, sonic l"1nba1tled anthropologists have been tempted 10 disn1is.o; cuhural snuties asjust rnorc trench
"postmodern ism ... This renex is currently \'isihle in negative reac1io1ls lo 1he new tclitorial policies of Barbara and Dennis Tedlock al the discipline's flagshipjournal American Anthropologilt. A tnotion to censure the jo11rnal"s "posunodern turn Mwas i111roduced (and defeated) at the American Anthropological Associalion 'sannual 1nec1ing.
Expressing a more ambivalent sense of the fra11gh1 border, Dale Eickehnan (quo1ed
by the Chronick of Higher Education) finds a rece11t "pho10-!iitudded article on 1he marketing of religious kitsch in Cairo [10 be] something 'radically new' for 1he journal,
work that 'recaptures some of the territory appropriated hy cultur.-1.I s1uclies'" (Zalewski 1995: i6). Handler (199:.i,) ghes a judicio11s account of tht cuhur.-1.I studies border, from the anthropological side.
7. One thinks of the Chicago School. More recently, one 1uight nuntion IJoward
Becker's work (e.g., 1986) or work by Van Maancn (1988), Hurawoy C"t 411. (1991),

220

JAMES (;LIFFORD

and Wellman (1995), all of which expliciliy address anthropological debates about
t1hnographic authority. Anthrupology has until relatively recently been distinguished
fnun sociology by a research object (the pritnitive, the tribal, the rural, the subalterntspecially non-Western and premodern). Michele Duchet (1984) has traced the
en1ergence of anthropology's special object to eighteenth-century anthropologysociology, which divided up the globe according to a series of familiar dichotomies:
\\;th/without history, archaic/modern, literate/nonliterate, distant/nearby. Each opposition has, by now, been empirically blurred, politically challenged, and theoretically rleconstructed.
K. My comments here are based on conversations with Susan Harding. Indeed, her
hybrid research practice was my starting point for reconsidering the "field" in anthropoloKY. Puhlica1ions fron1 her work in progres.'> include Harding 1987, 1990, and 1993.
q. As J send this chapter to press, I sadly note the death of my colleague David
Sc-J~neider, who never tired of re1ninding me that fieldwork was not the sine qua
non 0 anthropology. 1lis general position, a critique of my work among others,
has just appeared in Schneider on Schneider (1995: esp. chapter 10), a mordant, hilarious, intemperate hook of in1erviews. Schneider argues that famous anthropologists are distinguished by ideas and theoretical innovations rather than by good
licldwork. Ethnography, as he sees it, is a process of generating reliable facts that
h'IHI 10 confirm preconceived ideas or are irrelevant to the work's final conclusions.
Fieldwork is 1he e1npirical alibi for a questionable positivism. He dismisses claims
1l1<tt field research involves a distinctive or particularly valuable form of interactive
ll'arning. But under pressure frorn his interlocutor, Richard Handler, Schneider
rt.>lrea1s fro1n his n1ore ex1re1ne points. For example, he accepts that good ethnogr;iphy <utd theory are not strictly separable in the forging of reputations and recogni1.ts that anthropologists do (1nisguidedly) place a special, defining emphasis
on fit"lclwork. I-le also concedes that work in the field can produce new ideas and
lhallt"nge presuppositions. He does not co1nment, however, on how approved
tthnography functions in nonnative ways within the discipline. Schneider's charJcleristicallyvehe1nen1 strictures are a corrective lo the focus of this paper. And his
linal position seems to be that if fieldwork is indeed a distinctive mark of socioruhural anlhropology, it should nut be fetishized. I agree. I do not agree that an1hropology is (rt"ad, should be) "the study of cuhure." That, too, is a problematic
clisciplinary life raft. I will 1niss David's loyal provocations and certainly do not claim
tht l;.1s1 word in this argu111enL
10. See also l..uwie's 1-listory of Ethnological Theory (1937), which begins by sharply
tli!i>linguishing anthropological ethnography from exoticist, "literary" trdvel. See Mary
Louise Pratt's critique (1986) of this discursive move.
11. 'fhe "Sophisticated Traveler" supplements, which feature travel essays bywellknown writt"rs, <tre-along with the weekly Sunday travel section-major sources of
ach-enistn1t"nt rc\'cnue. An introd1H'.tion by New York Times editors A. M. Rosenthal
-.u1d Arthur (;db to 1he first of a series of anthologies based on the supplements claims
an t"<jllivalence between sensitive journalis1n and literary travel writing (Rosenthal
;uul (;elb 1~1l-t4).
1~. Many "good" hookstorf's now consecrate the tourist/traveler distinction by
1naintaining wellstocked, 1~aTale sections for guideOOoks and travel books.

SPATIAL PRACTICES

221

13. Van Maanen (1988) provides a balanced account of new approaches to

ethnographic writing and their consequences for fieldwork-anthropological and
sociological. The title of his book, Tales of the Fiel.d, is indicative of n1y present then1e:
travelers, not scientists, tell tales.
14. Marcus (1995: 105-110) replaces the i1nageof ethnographic dwelling with th;u
of "following." Multisited ethnography ranges widely, and on routes that oflen cannot be prefigured.
15. The literature is now very extensive. Golde (1986), Moore (1988), Bell, <:aplan, and Karim (1993), and Behar and Gordon (1995) indicate the current range of
feminist agendas. The latter work appeared just as this chapter was being sent to press,
and thus has been used sparingly.
16. An exception is the neglected work of Roth L..1.ndes. Sally Cole's illun1inating
account (1995), which arrived too late to be integrated in this chapter, confirms, I
think, Landes's generdl approach. Landes gave sustained attention lo "r.1.ce," resisting its subsumption under "culture." She gave prominence to issues of en1bodin1en1
and sexuality in fieldwork, which she presented in relational. personal tern1s. She
broke the disciplinary taboo on sexual liaisons in the field. The Cit_y of lVomni ( 1994
[ 1947] ), her work on candombli in Bahia, was clis1nis.<;ed by powerful gatekeepers.
according to Cole, as a "travelogue" (tainted also by association with the devalued
genres ofjournalism and folklore). The work of another casualty of professionali1a
tion, Zora Neale Hurston, was n1arginalized in si1nilar ways, seen as too subjellive,
literary, or folkloric. Hurston 's reception was (and still is) con1pounded by essentialist
notions of racial identity that have construed her negalivcly ;1.s a liniitcd native cth
nographer and positively as a conduit for black cultural authenticity. Such receptions,
academic and nonacademic, elide the different worlds and afliliations of rare, gender, and class negotiated by her work on the rural South, the llarlem Ren<tissance,
and Columbia University. The Hurston literature and debates are now quite extensive. Hernandez (1995) provides a valuable discussion.
i 7. Boon 1977 is a prescient historical exploration. Anthropologists are beginning to write self-consciously about and in this borderland (Crick 198::.; Boon 1992;
Dubois 1995).
18. On Flaubert's sexualized Orientalist !ravels see Behdad (1994). One 1nigh1
also mention Bali as a site for gay sex tourism before l~).tO.
19. One of the consequences, perhaps, of this _taboo on physical sex has been to
restrict discussion of the uerotics" of fieldwork. Newlon ( 199:{c) provides an antidote.
20. I have previously been using the tenn habitu1 in 1he generally recogniltct
social-scientific sense made familiar by Bourdieu (1977). This nolion sees the social
inscribed in the body: a repertoire of practices rather than rules, a disposi1ion to pla)'
the social game. It makes conceptions of social and cultural structure more processual: embodied and practiced. Unlike Defert'susage, ii presupposes modern notions
of society and cuhure. The older sense of habit us sees subjec1ivity as a matter of con
crete, meaningful gestures, appearances, physical dispositions, and apparel wi1hout
reference to these determining structures, which became hegemonic only in the late
nineteenth century.
21. The case of Isabelle Eberhardt is con1plica1ed by 1he coincidence of gender
and cultural cross-dressing. See Ali Behdad's acute discussion (t994).

222

JAMES CLIFFORD

22. In her gently reflexive ethnography Storytelltrs, Saints, and Scoundrels (1989),
Kirin Narayan provides a photo of herself in the field. The focus of her research was
the apartment of Swarniji, a Guru storyteller in India, and the photo shows several
wo1nen seated on the aparunent lloor. None of them is Narayan, though were she
.~tated anong thern, she would not with her sari and "Indian" features be easily distinguished from the other South Asian wo1ne11. The caption reads: "Listening at1enlively from the women's side of the room. The bag and camera cover mark my
presence. "The accoutrements of her trade occupy the ethnographer's discrete place.
Indeed, throughout the book, Narayan 's tape recorder is an explicit topic of discussion for Swamiji and his followers.
2:\- Appearances are powerful. Dorrine Kondo {1986: 74) begins her important
txplt1ration of the processes of dissolution and reconstitution of the self in fieldwork
tnnn111terswith a disturbingglin1pse of her own image, reflected in a Tokyo butcher's
display case as she shops for her Japanese "family." For an instant she is indistinKUishahle in every particular-clothes, body, gesture-from a typical young housewife, "a wo1nan walking with a characteristically Japanese bend in the knees and sliding of the feet. Suddenly I clutched the handle of the stroller to steady myself as a
wa\'t" of dizziness washed over me .... Fear that perhaps I would never emerge from
this world into which I was immersed inserted itself into my mind and stubbornly reft1sC"d 10 leave, until I resolved to ntove into a new apartment, to distance myselrfrom
rnyJapanese home and nyJapanese existence." In the border-crossings of fieldwork,
,. holistic ..experience'" is mobilized, and at risk. Kondo argues that this etnbodied
rxpe1ience needs to be brought into explicit ethnographic representation.
2,1. On 1he nonidentical. imbricated, relationship of indigenous, diasporic, and
J)<1stcolonial locations, see Clifford (1994).
2~. I a1n, of course, referring to normative patterns and pressures. Much fieldwork has, in fact, been done outside the (metonymic) "village" or "field site." In an1hropolngy, this is pt'nnitted, as long as the work is seen to be peripheral to a cen1ral sitr of intensive encounter. In other fieldwork traditions-for example, those of
tlicitation and transcription in linguistics-hotels and even universities can be pri1nary "field" sites. Such practices have been actively discouraged in anthropology.
26. Thus, rnany anthropologists were stung-or bemused-by attacks such as Deloria's in Cu.slt'T /Jittl for Your Sins (1969). The predatory visitor he dt'scribed, little betlt"r than a tourist, seen1ed a caricature. Anthropologists were being hostilely ..located,"
roughly shaken out of a self-confirming persona.
'l7. Thanks to Teresia Teaiwa. For her own very complex "native" location, see
j11;111nen1ariebarker and Tt"aiwa (19H4).
2H. In the short run, no1ions of "real fieldwork, shaped by canonical exemplars,
will co111inue to relegate emergent practices to what Weston (chapter 9) calls "virtual
tthnography," not qui le fit'ldwork. Uut the enforcing of fieldwork norms is uneven,
;1111! 10 a degree otlways has bc.en. l-low hierarchies of practice in a diversified/
fragnunted discipline are sustained and reformed remains to be seen.
:l9. I hold to this even in the face of n1y colleague Chris C..onnery's mots on Paul
Tht>roux: "'fravel narrows!"
R

REFERENCES

Abu-Lughod, Lila
1991
Writing against Culture. In Recapturing Anthropology: Working in
the Present. Richard G. Fox, ed. Pp. 137-162. Santa Fe, N.Mex.:
School of American Research Press.
---,ed.
1 993

Screening Politics in a World of Nations. Public Culture 5 (spring):

465-604.
Adas, Michael

1989

Machines as the Measure of Man: Science, Technology, and Ideologies of Western Dominance. Ithaca, N.Y.: Cornell University Press.

Adler, Judith
1988
(On youth travel). Annals of Tourism Research.
African Studies Newsletter
1969
2(6-7): (November-December). Entire issue.
Agar, Michael
1985

Independents Declared: The Dilemna of Independent Trucking.

Washington D.C.: Smithsonian Institution Pres!li.
Agee, James, and Walker Evans
1976[ 1939] Let Us Now Praise Famous Men. New York: Ballantine.
Akmajian, Adrian, et al.
1993
Linguistics: An Introduction to Language and Communication. Cambridge, Mass.: MIT Press.
Alarc6n, Norma
1990
The Theoretical Subject(s) of This Bridge Called My Back and Anglo-American Feminism. In Milking Face, Making Soul: Haciendo
Caras. Gloria Anzaldi1a, ed. Pp. 356-369. San Francisco: Spinsters
/Aunt Lute.
Allen, R.
1995
Mali Adventure, from Timbuctu to Dogon Country. The New York
Times Magazine, part 2 (17 September): 24-27, 36, 38.
22)