kubik jan-ethnograpy of politics

Upload: miguelaillon

Post on 14-Apr-2018

216 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    1/17

    Political EthnographyWhat Irnrnersion Contributesto the Study of Power

    Edited byEDWARD SCHATZForeword by Myron J . Aronoff

    1 The University of Chicago PressChicago and London

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    2/17

    ~") . " , . . . , . ,= > \ . .. y,., )( ,) '-, 1 ,l' e

    1b l n ! . J 'f E L l . ; - ' , . , - : > . E e"~cl !,r .:::,~~e-m'tre-ZoI( D v v - F ' +2 '+ ,15

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    3/17

    ONE

    Ethnography of Politics: Foundations,Applications, Prospects

    JAN KUBIK

    Today's poli tical science is a massive, multistranded research enterprisewhose complexity defies sucdnct characterization. It is therefore difficultif not irnpossible to provide a simple and concise answer to the question,"What is the use of ethnography for political scientists?" It depends on theaim of the research project, the specific ontological assumptions about so-cialfpolitical reality. and the particular conception of ethnography. Since itis impossible to consider al !possibilities in a single chapter, 1 delirnit thescope of my rernarks to two problmatiques that are central to the compara-tivist enterprise: the significance of the cultural aspect of social reality andthe consequences of the recent turn from macro- to rnicro-levels of analysis(Elster 1985; Geddes 2003). I focus discussion on the subfield of compara-tive politics, since it has been the site of most ethnographic work conductedwithin political science. It is also the subfield I know best. IThe relative utility of ethnography is closely related to our understand-

    ing of politics. March and Olsen (1989, 47-48) typify a cornrnonplace,material ist -institu tional understanding by suggesting, "The organizing prin-ciple of a political system is the al!ocation of scarce resources in the faceofconflict of interests." This i s a tirne-honored way of thinking about politics,for, as March and Olsen observe, "a conception ofpolitics as decisin mak-ing and resource al!ocation is at least as old as Plato and Aristotle."Yet March and Olsen are keenly aware of a significant shortcoming of

    this conception: "Although there are exceptons, the modern perspectivein political science has generally given primacy to substantive outcomesand either ignored symbolic actions or seen symbols as part of manipu-lative efforts to control outcornes" (1989,47). Not al! politics can be re-duced to competit ion over material resources: indeed, much of it concerns

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    4/17

    26/ Jan Kubikthe struggle over collective identiry, ineluding often deadly contests over themeaning of symbols signifying this identity. Dirk s, Eley, and Ortner (1994,32) develop this thought further:

    Poliucs isusually conducted as if identity were fixed. The quest ion then be-comes, on what basis, al different t imes in different places, does the nonfix-ity become tempora ri ly f ixed in such a way that individuals and groups canbehave asa particular kind of agency, poli tical or otherwise? How do peoplebecome shaped into acting subjects, understanding themselves in particularways? In effect, politics consists of the e ffort to domesticate the infinitude ofidentiry, It is the atternp t to hegemonize identiry, to order it into a strong pro-grammatic statement. If identity is decente red, pol itics is about the atternptto create a center.

    Such centers emerge and disintegrate as a result of specific actions byconcrete actors who propose, disserninate, and interpret cultural meaningsencoded in a variety of symbolic ways. To study such processes, political sci-entists-at least those who recognize that any attempt to propose and prop-aga te a vi sion of colIective ident iry, and thus any "cultural" ef fort whose a imis endowing human (parti cu la rIy collective) action with rneaning, is pa r ex-cellen ce political-must move beyond the ma teria/ist-insti tutional perspec-tive and employ a symbo lic -cultural approach.' And within such an approach"the researcher should ask whether the theory is consi stent with evidenceabout the meanings the histor ical actors themselves attributed to their ac-tons" (Hall 2003, 394). For researchers who embrace ontologies that in-e lude the "rneaningful" layer of reality, ethnog raphic approaches emerge aspromising tools for studying polit ics.

    If we understand politics as, in some important measure, loc ally pro-duced, we again might turn to ethnography. Indeed, attention to the rn icro-level of analysis constitutes an important trend in today's study of politics(Geddes 2003). Game-theoretic ambition to develop a concise theory ofpolitics (Bates et al. 1998) and a more sociological quest to identify rnicro-or meso-level mechanisms governing social and political life (Tilly 2001)share an assumption that progress in the soc ial sciences is more likely whenour analytic gaze is focused on the details of concrete interactions ratherthan the workings of "large" structures. Thomas P. O'Neill Ir's memorablequip that "all politics is local" captures this idea.

    This turn to the local coincides with renewed interest in observing "ac-tual" human behavior: students are increasingly admonished to focus onthe "real life" interactions of people in "real time," rather than on interac-

    Ethnography of Politics /27tions of variables in abstract theoretical spaces.' Hall captures this perspec-tive:

    The systernatic process analyst then draws observation from the empiricalcases, not only abOUIthe value of the principal causal variables, bUI aboutthe processes linking these variables to the outcomes .... This is not simplya search for "intervening" variables. The point is to see if the mult ipIe ac-tions and statements of th e actors at each stageo f the causal process are con-sistent with the imageo f the world historical process impli ed byeach theory,( 2 00 3, 3 94 )

    This attention to the s ymboli c-cultural and to the local, micro-scale, and "ac-tual" should make political scientists hungry for more ethnography, a re-search tool welI suited for addressing these emergent concerns.

    But we need to pause to consider the intellectual, philosophical, andepistemological origins of the long and tangled traditions of ethnographicinquiry before we can appreciate ethnography's potential value for thestudy of politics.

    The Promise of EthnographyMost writer s posit p articipan t ob servation as th e defining method (or tech-nique) o f ethnography (Bayard de Volo and Schatz 2004, 267; TilIy 2006,410). Below 1 invest igat e the use fulness, i f not indispensability, of ethnog-raphy for studying a reality that is construed as meaningful ("ideal"), pro-cessual (" diachronic"), and interactive. Suffice it to note he r e th at e thnogra-phy's usefulness for studying "constructed" realities has been demonstratedin sociology (where it serves as an auxiliary tool) and anthropology (whereit is the princ ipal tool). and should thus be examined by political scientists,particu la rIy comparativists, who are often admonished to take culture se-riously (Norton 2004; Chabal and Daloz 2006; Harrison and Huntington2000; Rao and Walton 2004).' In sociology, it supplements various inter-pretive techniques (for example, content or textual analysis) in studies thattreat cultures as assemblages of (broadly understood) texts; in anthropol-ogy , it is indispensable for studying culture in action.

    But ethnography obviously can be and has been employed by more pos-itivisticalIy oriented researchers. It is thus imperative to outline the differen-ti al uses of ethnography in positivistic and interpretivist research programs.Let us begin in ethnography's "maternal" discipline, cultural/social anthro-pology. To be sure, culture is not the only object of this discipline, which is

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    5/17

    28 / Jan Kubikcomposed of several, partially separate intellectual traditions, to some ex-tent overlapping with "national" schools." An exhaustive discussion is be-yond my scope, but it is worthwhile to highlight one dis tinction: whilethe "British" have developed social anthropology, the "Arnericans" tend topractice cultura l anthropology. Beyond sernantics lie fundamental ontolog-ical. epistemological, and methodological issues. In a nutshell, while the"British" generally tend to focus their efforts on studying so cia l s tru c tur esand their "political" dimension (initially in non-Westem societies), the"Americans" tend to construe the object of their s tudies as culture is] andthe multiple ways in which culture interacts with power. These differentdefinitions have consequences for the nature of specific projects, their con-ceptualizations, and methodologies. But a t the same time these two tradi-tions have something in common: they both approach politics as an as-pect of social relations that needs to be studied in practice, in s ta tu nasc end i,through extensive fieldwork centered on (preferably long-terrn) participantobservation. What ethnographers observe (via participation) depends on the par-

    ticular school or research tradition. Byand large, while "Britsh" lenses "de-tect" structure, "American" ones are fitted for studying culture. Significantly,both "structure" and "culture" can be and often are these days def ined in aconstructivist manner. It is enough to consider Ciddens's theory of st ructu r-ation or Bourdieu's theory of pr ac ti ce-b o th par exc el/ enc e constructivist con-ceptions of social structure. Moreover, at least since the wave of postrnod-em critiques, we know that "objects" of study do not exist out there, in an"objective reality," ready to be "discovered"; rather, they are coconstitutedby the two (or more) participants in a re se arch int e ra ction.

    Ethnogr aphyas a m et ho d (participant observation), therefore, isnot l irn-ited to the study of culture. Iust as many interpretive studies of politics relyon participant-observation, noninterpretive studies use the same method(for example, studies of organizational structures, informal networks, oreconomic exchanges). At the same time, not all interpretive studies ofpowerand polit ics are ethnographic (Bayard de Volo and Schatz 2004, 267). Infact, most are not. Table 1 illustrates these dis tinctions, based on examplesdrawn mostly from comparative studies of politics.Most research in political science is based on a naturalist ontology of

    the social and does not rely on participant observation; the work of Prze-worski and his collaborators (2000) on the relationship between economicdevelopment and the survival of political regimes is exemplary. Such workis also practiced to great effect in the broadly defined area of "political cul-ture": consider the large-n studies of Inglehart and his collaborators who

    Table 1Ethnography of Politics / 29

    Preferred Ontology of the Social andthe Attendant Epistemology/MethodologyMeaningful/lnterpretivist

    Natural/Positive/Noninterpretivist

    ResearchTechnique

    ParticipantObservation

    NonparticipantObservation

    Aronoff(1989);Kertzer (1996);Wedeen (1999)

    Kubik (1994);Edles (1998);Bonnell (1997);Chabal and Daloz (2006);O'Neill (1999)

    Malinowski (1922);Laitin (1998);Fortes andEvans-Pritchard (1940)Most political science;Przeworski et al. (2000);Petersen (2001)

    survey "values" of the world's population. Also, most work in game theoryis noninterpretive, as it is built on deductively derived models of purport-edly universal motivation mechanisms. Some game-theoretic work is sensi-tive to local contexts and is "ethnographic" in i ts tenor, although it does nottypically use participant observation. (Petersen's work, for example, dealswith past events, as I d iscuss below). The second category features natural-ist/positivis tic works that relyon participant observation but do not provideinterpretive accounts of the social worlds actors live in. Much of classicalBritish social anthropology belongs to this category. Most influentia l worksin comparative politics that rely at least partially on participant observationbelong to the naturalistic, noninterpretive genre, though some-such asLaitin's inf luential 1998 study-are close to the boundary between inter-pretive and noninterpretive types of work.The third type includes works that are interpretive but do not use

    participant observation. Bonnell's (1997) analysis of Soviet posters astools of power, Edles's (1998) work on the symbolic dimension of Span-ish democratization, or my own study (Kubik 1994) of Polish Solidarity'ssymbolic challenge to the hegemonic power of the Communist Party be-long to this rype, Finally, .works belonging to the fourth type combine in-terpretive epistemology with participant observation as the main method.Consider Aronoffs (1989) study of the inner workings ofthe Israeli LaborParty, Kertzer's (1996) detailed reconstruction of the Italian Comrnu-nis t Party cell's operation in a local setting, or Wedeen's (1999) analysisof everyday, counter-hegemonic challenges to Hfiz al-Asad's power inSyria.

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    6/17

    30/ Jan KubikTo summarize: as a methcd." ethnography is used to study culture

    (meaning systems) or other aspects of the broadly conceived social, suchas economy, power (politics), or social structure. lts essence is participantobservation, a discipl ined immersion in the social life o fa given group ofpeople. Ethnography is sometimes erroneously equated with (1) in-depthinterviewing (as opposed to administering surveys); (2) case studies (as op-posed to large-n statistical studies); (3) process tracing (as opposed to find-ing correlations); and (4) interpretat ion of meaning (as opposed to the"naturalistic" study of "objective" social facts). Studies based on these fourmethods are not necessarily ethnographic; they become so when they relyon participant observation ofconsiderable length.'An answer to the question "What is ethnography good for (in the study

    of power and politics)?" depends not only on the definition of ethnogra-ph y, but also on the conception of a discipline (its ontological and episte-mological assumptions) within which it is def ined and practiced. Becausethe track record of ethnographic work ismore robust in anthropology thanin political science, let me examine what it has contributed to each of threebroad traditions of political anthropology: positivistic, interpretive, andpostrnodern."

    Ethnography in Positivistic (Political) AnthropologyIn this section I take stock of the contributions tha t traditional, positivisticethnography has made to the study of power. I do so to emphasize a centralpoint: ethnography can benefit positivistic research agendas at least as wellas it can contribute to interpretive ones. Thus, as Allina-Pisano shows inchapter 3, e thnography has a long tradition of working in a "realist" ve in .Political anthropology is a subdiscipline with a distinguished tradi-

    tion of realist inquiry, Many iIlustrious nineteenth-century scholars (mostprominently Maine, Spencer, Marx, Morgan, and Tylor) studied non-Westem political systems and, par ticularly, their evolution.? They can beseen, therefore, as precursors or early practitioners of the discipline. But themodem field of political anthropology is often said to have emerged withthe publication in 1940 of A frica n P oliti ca l S ys tem s, edited by M. Fortes andE. E. Evans-Pritchard. All studies reported in that volume were based onextensive ethnographic fieldwork, but, by contrast to today's anthropolo-gists-who would emphasize the cultural specificity ofeach case-the ed-itors made an explicit effort to strip all social processes of "their culturalidiorn" and reduce them "to functional terrns" to generate comparisonsand arrivea t generalizations (quoted in Vincent 1990, 258). At the time of

    Ethnography of Politics / 31the volurne's publication, anthropology (including political anthropology)was still predominantly characterized by its focus on non-Westem, "exotic."or at least "marginal" societies. As Vincent (1990, 24) notes, "Not untilthe 1950s, in the face of challenges from other disciplines on the eve oftheir massive intervention in the anthropological domain, did anthropolo-gists make manifest that 'anthropology is characterized by a set of methodsrather than bound by a subject matter' (Bohannan 1967, xv)."

    Etlm ography and he S tu dy o f P ow erLong before Foucault made it fashionable, ethnographers were tr ackingdown the exercise of power within the interstices of official structures, be-hind the veil of various officialdoms, and in ostensibly apolitical spacesand domains. But, perhaps more important, there is no other method thatcan allow researchers to study power in sta tu n ascend i in all settings, for-mal and informal, and-particularly-to reconstruct the in fo r m al w o r ki ngso f f o r m al p ow er s tr u ct tl re s.Ethnography is thus used to map out the multiple layersof power within

    complex bureaucratic structures and to complement ifnot supersede recen-structions of "rnodern" power generated by other methods. For exarnple,Wedel et al. (2005) detail the anthropology of policy making in complex,modem organizations; Abeles considers the workings of the European Par-liament (1992) and the European Commission (2004); Aronoff (1989) ex-plores the Israeli Labor Party; Bailey (1983) unravels the politics of variouscommittees (at parliaments, govemments, universities, and so on): Wedeen(1999) focuses on the intricate mechanisms of resistance in authoritarianSyria;and Gaventa (1980) highlights the politic s of inequality in an Appa-lachian valley.Likewise, ethnography seems indispensable to the study of collective ac-

    tion; no other method can better expose mechanisms of the important,early stages of mobilization (Blee and Currier 2006). Petersen (2001) usesan "ethnohistorical" approach to reconstruct the mechanisms of anti-Nazimobilization in Lithuania in the 1940s. He begins by modeling the socio-log ica l mechanisms and group features that help explain why individualsin certain c ommun ities (a t certain times) rebel against oppressors, whileothers remain pass ive or collaborate. His model specifies such attr ibutes ofthe community as (1) the initial distribution preferences and constraintsconceming the risk of rebellion; (2) the types of norms (of honor or familyobligation, for example) that define the types of subgroups: and (3) thedistribution of those subgroups within the larger communiry, Petersen

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    7/17

    32/ Jan Kubikdivides mobilization into two stages: from passivity to resistance, and fromresistance to rebellion. Then, he considers separately the problern of rebel-lion's sustainability. For each stage or problem he identif ies a different setof mechanisms and shows how they reinforce each other, propelling theprocess forward ( to rebellion) or backward (to collaboration).Since Petersen studies pas t events, he canno t employ participant obser-

    vation, but he collects and interrogates his data as iJ they were generated bysuch a method. The ethnographic tenor of his study does not come, there-fore, from participant observation, but rather from an e thnogr aphic pr ob-lem at iza tio n an d fra m ing of the work. 10 He sets out to study the detai ls ofcommunity organization and uses all available information not only to re-construct actors' views and preferences, but also to map out their actionswithin the local structures that both empowered and constrained thern.and to identify the mechanisms that make mobilization possible. His is aquintessentially ethnohistorcal. indeed ethnographic, project.

    1I

    Ethnography a nd G am e T he or yEthnography is particularly well suited to test the limits of rationality ingame-theoretic models, which otherwise run the ri sk of circular reasoning.As Morrow proposes: "Rational behavior means choosing the best meansto gain a predetermined set of ends. It is an evaluat ion of the consistencyof choices and not the thought process, of implementation of fixed goalsand of the morality of those goals" (1994, 17). But to determine whethersuch consistency exists, researchers often infer both the in tentio n to employcertain means and their ac tual em pl oy m ent f rom observing ex po s t J ac to thevery same action. To determine an actor's rationality, one would have tofirst infer that an actor in tended to employ cer tain means to achieve a givengoal; then observe the actual means ernployed: and finally compare the two.Such tests of rationality are rare, perhaps because the ethnographic methodisseldom used in s tudies reIying on garne-theoretic modeIs.

    Et hnography and the S tu dy o J Soc ia l Str uctu reAny fuIl understanding ofpower requires an understanding of social struc-ture, which is both a product of and a constraint on individual and collec-tive action. Traditional ethnography makes at least three contributions inthis area. First, while focusing on small-scale phenomena, ethnography al-lows the researcher to see the way social structure actualIy works in people'sdaily lives (say, how a position in class structure influences one's life

    Ethnography of Politics / 33choices; see Willis 1981 or Sider 1986). Second, it contributes empiricalmater ial to the study of one of the thomiest problems of social theory: therelationship between structure and agency. By observing people up close,ethnographers can gauge the "structural'Timitations actors face, reconstructthe range of strategic choices they have, observe their actual action, and as-sess it s possibly transformative impact on structure. Third, ethnography isthe best method of studying the complex interplay between (formal) socialstructure and (informal) social organization."Ethnography's contr ibution to the study ofthe relationship between for-

    mal and informal inst itut ions should be carefully appraised by politicalscientists, as this reIationship has been one of the hallmarks of new insti-tutionalism. In a path-breaking and influential study, the economist Doug-lass North shows that economic behavior is shaped by both formal andinformal institutions (1990, 4) .12 Positing that economic performance isdetermined by a complex interplay ofthree factors-demography (humancapital), the stock of knowledge (including everyday beliefs), and the for-mal institutional framework-North (1997, 14) warns that "we know verylittle about this interaction." Participant observation iswelI suited to study-ing the complex interplay of such factors.

    Ethnograph y an d the Study o J S o c i al P ro cessResearch methods typical of political science, such asopinion surveys, peri-odie coIlection ofeconomic statistics (usually aggregated on ayearly basis).or pooled expert opinions on institutional changes (also routinely aggre-gated and reported yearly; see the World Bank or Freedom House) regis-te r th e occ ur rence oJ change ; they do not sp ec ify th e m echanism s oJ change .As political science faces increasing calIs to tum fram macro- to micro-levels and to study actual mechanisms, the value of e thnography shouldbecome apparent. AsTrickett and Oliveri (1997, 149) argue, "ethnographycan capture the dynamic ofchange inways that snapshot surveysusing pre-established dimensions and response categories cannot." Ethnography al-lows researchers to reconstruc t the manner in which large-scale social pro-cesses (say, postcommunist transformations) actuaIly occur and how theyconstrain or empower people in their dai ly l ives. I t is, after all, the repro-duction and transformation of daily lives that are observable, not "struc-tural change."Ethnography can also detect how the rnacro- and micro-dynamics of

    change may be out ofstep. Introducing a collection of ethnographies aboutpostcommunist transformations, Burawoy and Verdery opine:

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    8/17

    34 / Jan KubikOur view of the relation between macro structures and everyday practices isthat the collapse of parry states and adrninistered economies broke downmacro structures, thereby creating space for micro worlds to produce auton-omous effects that may have unexpected influence over the structures thathave been emerging .... It is precisely the sudden irnportance of micro pro-cesses lodged in mornents of transformation that privileges an ethnographicapproach. (1999, 3)

    SignificantIy, they do not merely signal the differential rhythms of macro-and micro-changes; they claim that the often overlooked and/or unintendedrnicro-processes may influence, derail. or even halt macro-changes. Such ob-servations dovetail with recent writing on the mechanisms of social changein historical sociology and historically oriented comparative politics."

    Finally, ethnography is critical for identifying the sources of impendingchange. Norton (2004, 41) observes that s ince change often comes from theperiphery, it is important to "recognize the power of liminal, or marginal.groups .... Because they stand on the boundaries of identity, they are oft encentral to debates over those boundaries."

    Ethnography and the Study of Political EconomyBird-David observes, "A diversity of exchange forms had been reified byanthropologists into either 'gift' or 'cornrnodity' while in the concretenessof sodal lfe=-arnong indigenous people as among Westerners-there aremultiple kinds. These have to be studied, too" (quoted in Herzfeld 2001,111). To study economic t ransactions in isolation from their cultural andsocial contexts entails a risk of serious distortion. Aware of this, most an-thropologists rely on ethnography to advance what has come to be knownas the "substantivist" conception of economy (as distinct from the "formal-ist" view derived f rom neodassical economic theory).The distinction between substantivist and formalist definitions of eco-

    nomic activity was first introduced by Karl Polanyi (1957):

    The two root meanings of "econornic," the substantive and the formal. havenothing in common. The latter derives from logic, the former from fact. Theformal meaning rnplies a ser of ru les referring to choice berween alternativeuses of insufficient rneans. The substantive meaning irnplies neither choicenor insufficiency of means; rnan's livelihood may or may not involve the ne-cessity of choice and, if choice ther e be, it need not be induced by the lirnit-ing effect of "scarciry"of the means. (1957,243)

    Ethnography of Politics / 35In the same essay he observed:

    The human economy, then, is ernbedded and enmeshed ininstitutions, eco-nomic and noneconomic. The inclusion of the noneconomic is v ital. Forre-ligion or governrnent may be as imponant for th e structure and functioningof the economy as rnonetary institutions or the availability of tools and rna-chines thernselves that lighten the toil oflabor. (1957, 250)

    This distinction can be understood ontologically, methodologically, andhistorically. In ontological terms, the substantivist (holistic) approach treatseconomic activity as an "aspect of sociallife rather than a segment of soci-ety" (Plattner 1989, 14). Accordingly, the economic domain is construedas inseparably embedded within other domains; it cannot be fruitfully an-alyzed in isolation. Formalists, by contrast, conceptualize economic activ-ity as a separate domain (segmen t o fsociety) with its own, specific mecha-nisrns, best specified by neoclassical microeconomics.

    Substantivists daim that since economic activity is usually (or always)-even in the capitalist system (see Narotzky 1997; Gudeman 2001)-em-bedded in social. cultural. and polit ical contexts, it has to be studied socio-logically/anthropologically/ethnographically (in its ful l soc ial context)."For their part. methodological formalists attempt to apply the methods ofmodern microeconomic analysis to non-Western societies. They assumethat "indiv idua ls in every culture exercise r ational choice in a rneans-ends,constraints, and opportunities framework" (Plattner 1989, 13). They pro-pose that th e economic domain can be usefully isolated from other do-mains of social life (religious, farnilial. and so on) in every society andpeople can be always studied as "tility maxirnizers."

    Historically, th e distinction is sometimes said to separate modern societ-ie s that operate on the market principle f rom other societies wher e the mar-ket prin cip ie either is absent o r is inseparably intertwined with other prin-cipIes (moral, familial, or statist). While many people subscribe to someversion of an evolutionary paradigm, believing that societies move from asubstantivist to a formalist phase. ethnographers have p rob lematized thisdaim. For example, anthropologists who study postcommunist transfor-mations have shown that the "progress" from ex-communist substantiv-ism to neocapitalist formalism at best is uneven, slow, and full of reversals.Moreover, to survive under adverse conditions, people create and main-tain complex networks that can be conceptualized as "econornic" only ata risk of seriou s conceptual stretching. The extensive and empirically con-vincing literature on this phenomenon is found mostly in anthropology

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    9/17

    36/ Jan Kubik(Humphrey 2002; Burawoy and Verdery 1999; Buchowski 1997), but alsosociology (Stark and Bruszt 1998) and poli tical economy (Woodruff 1999;Blyth 2002).These insights about pos tcommunist transformations fol !ow on the

    heels of a long anthropological tradition of recognizing complex relation-ships between economy and culture. Beginning in the 1960s, anthropol-ogists using ethnographic evidence wrote books with titles like Poli t icalEc ono rny and Cultuie o/ ... showing, for exarnple, that while in practiceeconomic relations tend to be complex and multistranded, in ideology (ofcapitalism or communism) they appear as separate and sngle-stranded.In a recent, sophisticated ethnography of family firms in Italy, Yanagisako(2002, l3) performs a s imilar service, showing that even today, "familyandkinship processes, relations, and sentiments are crucial for the productionand reproduction of al! forms of capitalism, whe ther family capitalism ornonfamily capitalisrn" -in spite of a prevailing discourse that normativelyseparates "farnily" from "business" relatons, It is hard to imagine how sucha demystification of the dominant view of economic activity could havebeen accomplished without ethnography. Significantly, this realization hasalready filtered to the World Bank, as a path-breaking volume indicates(Rao and Walton 2004).From the onset of anthropology (political anthropology in particular),ethnography has been successfully employed to locate power in hardly ac-

    cessible or atypical places, beyond the world of formal institutions. And ascultural/social anthropology has come to be defined by its method ratherthan its object, its preoccupation with marginal or peripheral phenornenahas continued. Thus, the major contribution of traditional positivistic an-thropology-via ethnography-is to the study of power and politics ou t-side ofcenters and mainstreams, within a complex interplay with economicand cultural processes, and in locations and crevices where the exercise ofpower or authority is often in visib le to other disciplines.

    Interpretive Ethnography:The Study of Meaning (Culture) in Action

    Wittgenstein should perhaps be declared a patron saint of the ethnographicstudy of meaning. as he emphasized that the meaning(s) ofa sign (word,picture, and sound) isbest determined through studying its use, its employ-men t in social practice. As 1 demonstrated earlier, ethnography and inter-pretation are not necessarily paired; quite often they are noto But their corn-bination al!ows for the reconstruction ofhow culture (the meaning-creating

    Ethnography of PoI itics / 37machine) operates in prac tce, and how the actual product ion and inter-pretat ion of meaning are pract ical activities, often cent ral to both powerstruggles and economic rnaneuvers, and shot through with emotions.There are many defini tions of culture; ethnographers need one that

    goes beyond treating it merely as a symbolic structure. The Comaroffs(1992), for example, speak about meaningful (37) or symbolic practices(35) that constitute culture construed as a "sernantic space, the field ofsigns and practices, in which human beings construct and represent thern-selves and others, and hence their societies and histories. It is not merelyan abstract order of signs, or relations among signs. Nor is i t just the sumof habitual practices. Neither pure langue nor pure parole, it never consti-tutes a closed, entirely coherent system. Quite the contrary: Cul ture alwayscontains within it polyvalent contestable messages, images, and actions"(1992, 27).Contests within this semantic space are by their nature political, as

    they often constitute attempts to achieve legitimacy or to establish collec-tive identities (nation, class, gender, race) and endow them with an aura ofnaturalness. Can societies and their politics, perrneated by such symbolicstruggles, be studied in the same manner as "natural systerns" examined bythe "hard" sciences? Sahlins (2004) argues that Thucydides in Th e Pe lopon -ne sian W ar provided a positive answer to this question, thus setting Westemsocial science on a naturalist course. Since Thucydides, the debate berweennaturalists and antinaturalists has run through almost the entire courseof Westem social reflection. In the second half of the nineteenth centurynaturalism and antinaturalism in the social sciences were defined and de-liberated with great clarity by such German scholars as Heinr ich Rickert,Wilhelm Windelband, and MaxWeber (for overviewssee Palmer 1969; Blei-cher 1980; Bambach 1995).Perhaps the most in fluential among thern, Wilhelm Dilthey (see, for

    example, Dilthey 1976), proposed a sharp distinction between the "posi-tivistic" (Na tun oissenschaft en '; I S and the "hurnanistic" (Geisteswissen schaften)disciplines, defined the "proper" method of the latter asinterpretation (Ver-st ehen ), and sharply differentiated it from explanation, the proper proce-dure of the former. Clifford Ceertz revived this dis tinction in his trend-setting 1973 volume, ashe famously declared: "Believing, with MaxWeber,that man is an animal suspended in webs of s ignificance he himself hasspun, 1take culture to be those webs, and the analysis of i t to be thereforenot an experimental science in search of law but an interpretative one insearch of meaning. It is explication 1am afte r, construing social expressionson their surface enigmatical" (1973, 5).

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    10/17

    38/ Jan KubikToday, while so me argue for a basic unity of the "scientific" method

    (King, Keohane, and Verba 1994; Miller 1987), the question remains: giventhat social act ivity entail s human agent s producing, receiving, and inter-preting meaning systerns, does social research require specifically tailoredtechniques? It is a complex debate that has been recently reviewed (Taylor1971; Yanow and Schwartz-Shea 2006; Gerr ing 2001; Henderson 1993; Fay1996; Lit tle 1991, 1998). Nor is it rnerely a methcdological (how to studysomething) debate; i t i s a lso-if not primarily-epistemological (what arethe conditions of knowability of certain types of objects?) and ontological(how are the objects ofknowledge"actually" constituted?) (Hall 2003).

    Many thinkers have art iculat ed unequivoca l ontological-epistemologicalantinaturalist answers to these questions, the clearest of which may be thatoffered by Thomas and Znaniecki (1918-20). What has come to be knownas the "Thornas theorern" is usually formulated as: "If men define situa-tions as real, they are real in their consequences." Already in the early 1920sboth Thomas and Znaniecki offered an extensive set of comments on thisproposition, emphasizing that social reality is constructed by humans (intheir case, ethnic Peles in Poland and in the diaspora). This essentiallyphenomenological position was later developed by Berger and Luckman(1967 ) and eventually spilled over to o ther disciplines in various forms ofconstructivism. In this way, a strong philosophical posit ion ernerged thatcombines constructivism as an ontological stance with interpretivism as anepistemological-methodological position.

    Political scientists hotly debate whether "subject-dependent realities"(as opposed to "objective realities") are the proper subject matter and thuswhat tools of inquiry are appropriate. 16 Meanwhile, much of sociology andalmost all of cultural/social anthropology are de ri gueur constructivist, asare, basically by def intion, cultural studies, much of feminist work, andart history." Significantly, this ernbrace of constructivism often leads toethnographically inflected inquiry in these fields. Anthropologist RichardShweder (1997), f or example, def ines ethnogr aphy as a "species of quali-tative research" that, by definition, dea ls with "qual ia" whose on tologicalstatus often rernains undertheorized as researchers f ocus on nar row meth-odological rather than ontological/episternologcal issues:

    J propase that the t ension between quantitative and qualitat ive turns less onmethodological i ssues than it does on one's answers toquestions about how10 best study subjectivity and how 10 best study realities that are perspec-tive and context dependent. Basically it is the difference between studyingsomething that exists regardless of your's [sic]or anyone's reactions 10 it, and

    Ethnography of Politics / 39studying things that come into real existente by v irtue of their meaning andthe perspective that is taken on them. (1997, 1 G O )

    Sociologist Andrew Abbott (2001a, 61) argues that constructivism (or"constructionism." as he calls it) is characterized by idealism (attention tohow reality is rnediated by interpretation). diachronism (attention to pro-cess), and interactionism (attention to social in te ract ion as const itutive).It s ee rns that any investig ation guided by Abbott's idealism, diachronism,and in ter actionism is more likely to succeed when the researcher privilegesa rnicro-level (over macro-Ieve l) approach, as actual people need to be ob-se rved in real-lfe situations.

    Interpretivism in cultural anthropology had its heyday in the 1970s andthe early 1980s. The masters of interpretive or symbolic anthropology-Clifford Ceertz, Victor Turner, Marshal l Sahlins, Edmund Leach, DavidSchneider, and Mary Douglass-proposed rich and multifaceted theoreticalframeworks for studying the complex relationship between the politicaland the symbolic; the essence of this relationship was captured by the titleof Abner Cohen's serninal Two-D im ensional M an: An Essay on th e An thro-po log y o/ Po wer an d Sym bo lism in Complex Societ y (1974). Through carefulconceptualization and detailed ethnographic f ie ldwork, these (and manyother) scholars showed that culture is not an irnrnutable, monolithic ter-rain composed of str uctured configurations of symbols and signs, ava il ablefor (contemplativ e) in ter pretation; it is rather a complex set of signifyingpractices through which humans collectively create th e worlds they inhabitand within which they compete for power and material advantages.

    Since the 1980s interpretivism has changed, mostly under the impactof the postmodern challenge, but-it has retained its viability as a researchprogram not only in cultural or political anthropology (for exarnple, Kwon2006), but also in other disciplines, including political sociology (Ashforth2005) and comparative polit ic s (Wagner -Pac if ic i 1986; Fernandes 1997;Edles 1998; Chabal and Daloz 2006).

    Ethnograph y o f N ation- Bui ldingMuch of the competition for symbolic power and cultural hegemony re-mains inscrutable for such standard methods of political science as surveysor institutional analysis. Interpretive ethnography offers a solution. Take, forexample, the study of nation-building. one of the central preoccupations oftoday's comparative politics (Smith 2004). It is hard to imagine a methodother than ethnography that would highlight and clearly dernonstrate that

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    11/17

    40/ Jan Kubiknational-Ievel meaning-fonnation and similar local-leve! processes are often~ncon~ruous and, if related, their relations are cornplex.'! Herzfeld (1997)invesugates the relationships between national and local levels o f identiryformation and shows that there is no single logic that would apply to thefonnation of "nat ional identity" in al! locations where this process takesplace. Gagnon (2004) shows how ethnic identities of "Serbs" and "Croats"are forrned, re-forrned, and defonned through a series of mobilizations anddernobilzations whose local and national rhythms vary considerably. In-terpret ive e thnography can capture this variation.

    lnterpretive Ethnography and DemocratizationInterpretive ethnography has also made critical contributions to the studyof democratization. For exarnple, by showing that "democracy" is inter-preted and thus practiced in many different ways that depend on local cul-tural contexts (Wedeen 2004; Paley 1002), ethnographically inclined re-searchers help us understand why democracy-promotion projects built ondecontextualized, universalistic assumption s are beleaguered by often in-soluble problems.

    Ashforth's (2005) nuanced and mul ti layered ethnography shows thatthe fragile legitimacy of South African democracy is seriously threatenedby the persis tence of wtchcraft.'" As the country undergoes rapid polit icaland economic change, many people, particularly in poorer areas such asSoweto, feel increasingly insecure and unsure why the benefits of post-Apartheid development are sparse and slow in coming. They al so feel jeal-ous of those whose life fortunes have improved. To deal with insecuriry anda growing sense of injustice, they look for explanations offered by their ownculture, in which personal misfor tunes are attributed to evil acts of witches.This culture also suggests a rernedy: affected individuals or communitiesneed to enlist the help of traditional healers, But the healers' authority chal-lenges the ef forts of the new, democratically elected, and "modern" govern-ment to achieve legitimacy. As Ashforth puts it:

    Belief in witchcraft presents severe chalIenges for the project of democraticgovernment within a modern state. A democratic regime cannot acknowledgethe legitimacy of "informal" efforts to seek justice in the face of witchcraft,but if authorities prevent communities from securing their own forms of jus-ricewhile refusing 10 address the underIying problem of occult violence, theyopen themselves to the charge that they are either ignoring the dangers facingthe community or in league with evilforces themselves. (2005, 314)

    Ethnography of Politics /41Ashforth's study dernonstrates that the "top down" politicallogic of democ-ratizing projects often clashes with the "bottorn up." usually local. culturalmechanisms that dictate the meaning of democratization for the "target"populations. Insensitivity to such localized cultural understandings oftenderails or def orrn s even the most promising democr atization p rojects.

    lnterpretive Ethnography and the Politics of Collective MemoryKwon (2006) provides another example of the cJash of between national-and local-level logics in a study examining collect ive memory and i ts impacton regime consolidation in Vietnam. Again, ethnography p roves indispens-able. The Vietnarnese, whose society had been ripped apart by devastatingwars, have recently engaged in the rebuilding of their country 's social andcultural fabric, In a society whose edifice rests on a base of rnultigenera-tional kinship regulated by elaborate ritual s, this is a particularly demand-ing task. The ritua l recons truction of lineages that were destroyed by "baddeaths" that occurred " in the streets" (while the proper loca tion for dying is"at horne"] is fraught with difficulties. Death outside of the cul turally legit-imated locations disrupts the viability of family units grounded in the cultof ancestors. Given tha t family unit s a re the building blocks of social order,a society that lost hundreds of thousands of its mernbers in ritually uncon-trolled conditions, eithe r as sold ie rs on bat tl ef ie lds or as victims of killingfields, is unbalanced. Rebu rying under the "proper" conditions can "re-fit"ancestors into their "rightful" rituallocations and thus restore social order.

    This is exactly where the state's politics and the culture of lineage andcommunity clash. The communist authorities of Vietnarn. although reluc-tantly supportive of society's self-healing efforts, are al so interested in thevictims of the for eign invasions. Reburying can becorne a political ritualthat prornotes the regime's interests: the dead are splendid candidates forhero worship. But the culturallogics of hero worship, championed by theregirne, and cults of the ancestors, needed by the families, are at odds. En-gaging in the latter can underrnine the regirne's cJaims to legitimacy andweaken the nation-building potential of official heroisrn.

    Kwon contributes to the literature on the rel at ionship berween the pol i-tics of memory fonnation and the struggle for politicallegitimacy (see, forexample, Davis 2005) and confinns that the study of this relationship is s e-riously flawed when cu ltu ral mechanisms involved in this process are ab-stracted away. But he insi st s tha t collectve mernory is always forrnulatedaccording to specific cultural rules in concre te soc ial locations and con-structed on several levels, at different scales. o ften simultaneously. The

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    12/17

    42/ Jan Kubikpolitical and cultural logic that governs this construction at the nationallevel, where it contributes to the regirne's self-legitimizing efforts, can beundermined or annulled by the local, regional. or familial processes of col-lective memory formation.

    In te rp retiv e Et hn ogr aph y an d ComparisonsWhat about the relationship between (ethnographic) interpretation andcomparison, a chief task of comparative politic s? Within traditional posi-tivistic anthropology, participant observation was seen as a reliable and un-problematic tool for collecting data that were directly fit for comparisonsand generalizations, but the interpretive turn undercut this methodologicaloptimismo As Holy aptly puts it, while in positivistic anthropology "gen-eralization was seen as problematic, description was not" (1987,4), "sub-jective" or "interpretative" anthropology problematized description as itmoved from "the theory of social facts as things to a theory of them as con-structions" (5). Constructivism made comparisons dubious, and the word"cornparison has completely disappeared from the vocabulary of rnethod-ological discourse" (6-7).Such a conclusion may sound ominous to the practitioners of compar-

    ative politics, but in all fairness many of them have arrived at a similar po-sition (Bowen and Petersen 1999; Smith 2004; Wedeen 2004; Chabal andDaloz 2006), prefigured in the work of a philosopher who once took a goodcriticallook at their practices (MacIntyre 1978). Comparativists, whose fieldhas been stretched-in Collier'sseminal formulation (1993)-betweencase studies and large-n studies, used to search for a "scientific" s alvation inthe direction of an "ever larger n." These days. however, many of them optfor a method of "srnall scale controlled cornparison," which promises that"through a focus on process and mechanism within the detailed study ofthe cases, much of the complexity of pol itical li fe can be addressed whilemaintaining an ability to generalize" (Bowen and Petersen 1999, 11).To summarize: interpretive ethnography based on participant observa-

    tion of semiotic practicesdelivers important and original bodies ofknowledgefor research programs founded on three commitments: (1) con st ruc tivism ]inter pretivism (interpret-not just explain-actions that are "rneaningful"to actors). (2) ontologi ca l rea lism and an attendant epistemology (focus onactual actions of real people, rather than variables VO and (3) m icro -sc ale(observe actual. "srnall-scale" settings and reconstruct relevant mecha-nisms ) .2 1 These three commitments undergird a research agenda that is in-

    Ethnography of Politics / 43dispensable in a world that stubbornly refuses to be rationalized and ho-mogenized and in which the politics of identity ispervasive.

    Postrnodern/Multi-Sited/Clobal EthnographyIt is paradoxical that when some political scientists have begun turningtoward ethnography, anthropologists have thoroughly reevaluated, andoften scathingly critiqued, the method that trad itionally is the r ai so n d ' treof their discipline." With the postmodern tu rn, ethnography has been facedwith new tasks; its role in the study of power therefore needs to be exam-ined afresh. As I will suggest, a "reforrned" ethnography remains just as rel-evant to the study of power as ever,The reexamination of ethnography's value has been propelled by a

    double engine of pos tm od erni sm and global ization . The former cast doubtson the epistemological adequacy of social-scientific methods, particularlytheir claims to "objectivity," "detachrnent." and the possibility of "accurate"representations ofreality; the last of t hese accelerated a reconceptualizationof the object of study by discarding earlier conceptions of "ethnographic"realities as isolated and self-enclosed systems.Debate about the relative (de)merits of "ethnographc" representations

    of reality has been particularly heated since the publication of Writing Cul-tu r e: The Poe ti c s and Po li tics o/ E th no graph y (1986), whose authors examineethnography as a literary genre belonging to a broader category of "scien-tific" writing. Crapanzano, for example, criticized Geertz for being toodomineering and not allowing enough space in his narrative for the na-tives' unfiltered voices: "Despite his phenomenological-hermeneutical pre-tensions, there is in fact in 'Deep Play' no understanding of the native fromthe native's point of view. There is only the constructed understanding ofthe constructed native' s constructed point of view. Geertz offers no speci-fiable evidence for his attributions of intention, his assertions of subjec-tivity, his declaration of experience" (1986, 74). Postmodernists challengethe authorial authority of the imerpreter, the hero of the interpretive turnoThey posit that ethnographic texts should be polyvocal, allowing the "na-tives" to speak in their own voices and to represent (textually, narratively)themselves." While very few scholars would relinquish control over theirtexts, the idea that the work should allow the reader to "hear" the natives'own conceptualizations of reality is not alien to many practitioners of to-day's comparative politics (Laitin 1998; Wedeen 1999; Ashforth 2005). Itis, however, clear that the postrnodern critiques of e thnography as a genre

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    13/17

    44/ Jan Kubikof wring have not fundamentaIly changed the way most social sciencenarratives are composed. Nor have they diminished interest in ethnographyas a research method, although globalization has caIled for its overhaul.

    Globalized World: From Structures to Flows and NetworksThe image of Structure in today's social sdence has lost its once formidableluster as it has been challenged bythe images of networks (CasteIls 1996),assemblages (Sassen 2006), flows, and scapes (Appadurai 1990). This hasput a new set of demands on the methodology of empirical investlgaton.What should we do, for example, with a me-honored, grand binary op-position: system versus life-world (Habermas), which echoes Marx's owndistinction between base and superstructure? Marcus astutely observes that"the distinction between lifeworlds of subjects and the system does nothold, and the point of ethnography within the purview of its always local,close-up perspective is to discover rrew paths of connection and associa-tion by which traditional ethnographic concems with agency, symbols, andeveryday practces can continue to be expressed on a differently configuredspatial canvas" (1998, 82). Such a perspective produces an image of so-cial reality as a flat plane, composed as a mosaic of pieces ofvarious sizes,complexly interconnected, and subjected to increasingly rapid recombina-tions; the older, Marx-inspired vision of a hierarchicaIly ordered reality (a"causally" weighty base at the bottom and a somewhat less consequentialsuperstructure at the top) is pass. One consequence of this remapping isa cal! for ethnography to focus on complex interactions of economic, so-cial, political, and cultural processes, without a ptiori privileging causallyany ofthem.

    Power in the Globalized WorldIn a path-breaking formulation, Foucault proposed that the study ofpowerneeds to focus on its exercise or actualization( s) "in the complexities ofeveryday practice" (Herzfeld 2001, 122). This premise has long guided eth-nographers and has produced eye-opening results; the postmodern situa-t ion has added a layer ofcomplexity. Having been asked to look for powerliterally everywhere, ethnographers now facethe task of tracking itdown in-side extensive, often hidden networks that connect actors through increas-ingly globalized webs of influence, dependence, and assistance. As localitieshave become increasingly discourse-based and "virtual," actors can escape(at least partial!y) the exerciseofpower by their direct "local" superiors by

    Ethnography of Politics / 45engaging in Intemet-empowered, transnational networks (see Tarrow 2005for an overview). Ethnographic studies of such networks are not easy, butthey are much needed.

    ldentity and GlobalizatiollWelearned some time ago that the s tabili ty/permanence of identity mustnot be a priori assurned, but empiricaIly determined. It is therefore helpfulto conceptualize politics in such a manner that the struggle for identity be-comes as central to it as the struggle for scarce resources." This new con-cept is most powerfully articulated by Dirks, Eley,and Ortner, who construepolitcs as a struggle to stabili ze coIlective identity around a symbolicaIlyestablished center (1994, 32). Sometimes such a center is formed; sorne-times it is not. It has also becorne clear that "resistance in the struggle toestablish identity dces not rest on some nostalgic bedrock of tradition orcommunity, but arises inventively out of the sarne deconstructive condi-tions that threaten to pul! it apart or destabilize what has been achieved"(Marcus 1998, 74). The postmodem turn makes the task of studying suchprocesses o/ invention and stabilization even more demanding: the forrnationof identity needs to be caught in statu nascendi, as various flows intersect ina single local e and/or are traced down through several locations/locales.Again, it is hard to imagine a method better suited for such a task than eth-nography, perhaps pursued at multiple sites, as 1discuss below.

    Ethnography and the Global-Local DynamicIn this increasingly globalized world, researchers face a methodologicalchaIlenge: ethnography, des igned to study the structuring of social life,power, and the formation of identity in "small" locations, has preved inad-equate for studying the world of (fast) global flows of goods, services, andinformation; migrating populations; and shifting meanings. Asthe Coma-roffs observe:

    The economies of signs and practi ces have to be situated in the intimacy ofthe local contexts that gave them life. At the same time, they require to beinser ted into the translocal processes of which they were part ab in icio : pro-cesses-commodificat ion, colonization, proletarianization, and the l ike-composed of a plethora of acts, facts and utterances whose very descriptiondemands that we frame them in terms of one or another Theory of History,(2003, 161)

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    14/17

    46/ Jan KubikThe "terrain" ethnographers are supposed to investigate needs to be recen-ceptualized so they can situate the object of their study (a small cornrnu-nity) within a translocal field of political, economic, and cultural forces.The work to address this need began in the 1970s. Paradoxically as somepolitical scientists began moving their discipline from macro (structural hs-torical studies) to micra (garne-theoretic work on the individual calculationand the growing interest in the small-scale mechanisms of politics), 25 an-thropology in the hands of many of its leading practitioners was alreadytraveling in the opposite direction.By and large, this movement from rnicro- to rnacro-level in anthropol-ogy has had two major phases:" (1) the wo rld-system phas e, inspired mostlyby Wallerstein (1974) and Frank (1969), and arguably culminating withWolf (1982); and (2)the glob alization /postmod em phase, epitomized mostdistinctly by the critical works of Clifford and Marcus (1986) and Marcus(1998). The theoretical tenor of the first phase was decisively rnaterialist,(neo- )Marxist, while the second wave was primarily culturalist, as its practi-tioners pushed the interpretive turn in the social sciences to its limits (andperhaps beyond). During this phase, anthropologists set ou t to dernon-strate how local structures and cultures are influenced and shaped by largerstructures, such as s tates, class structures, and the worId sys tem of (mostlyeconomic) interdependencies.The second phase was marked by the theoretical implosion ofthe wholereper toire of such "traditional" concepts as the binaries of micro/rnacro,system/worldview, or center/periphery, and the concept of a clearly boundcultura l whole. Accordingly, ethnography and fieldwork had to be rein-vented again. In Trouillot's words: "The problem is not f ieldwork per se, butthe taking for granted oflocalities upon which the fetishization of a certainkind of fieldwork was built and the relationship between ... supposedlyisolated localities and supposedly distinct cultures" (2003, 125). To answerthis challenge, methodologists and practitioners propose ethnography thatis postmod ern, mu lti-sited , and glob al.

    Po stmodem EthnographyTrouillot argues that the conceptualization of ethnography as a method fo-cused on the study of small, relatively homogeneous communities has pre-vented anthropologists from achieving the proper understanding of therelationship(s) berween the broader wo rld (however it is conceptualized)and specific loc aci on. This blinded them to the phenomena that constitute

    Ethnography of Politics / 47the bread and butter of today's social science: the expansion of capitalismand its "local" consequences, colonialism and postcolonialism, migrations,and globalization( s). Postmodern sensitivity demands that location be con-strued not as a relatively bounded and separate whole, but as a place wherevarious flows intersect.According to Marcus, postmodern ethnography needs to be sharply dis-

    tinguished from what he calls positivistic or "realist" ethnography. Specif-ically, three dimensions of ethnographic inquiry need to be reconceptual-ized: the spatial, the temporal, and perspective or voice (1998,62).First, the concept of community "in the classic sense of shared values,shared identity, and thus shared culture" (Marcus 1998,62) needs to be re-placed with the concept of "rnulti-Iocale, dispersed identity" (1998, 63),constructed, often simultaneously, by often mutual!y independent flows ofcultural material, complex political configurations, and economic relations.Marcus argues, "It is the burden of the modernist ethnography to capturedistinctive identity formations in al! their migrations and dispersions"(1998,63).Second, to "(postjmodernize" the temporal dimension of ethnography,

    Marcus asks that we replace dominant "Western historical rnetanarratives"that have routinely served as a historical background for many scholarswith local histories and carefully reconstructed loca! collective memories.He contends, "The past that ispresent in any site isbui!t up from memory,the fundamental medium ofethnohistory" ( 6 4 ).21Finally, the traditional ethnographic perspectve, heavily indebted to the

    concept of structure (social or sernotc), needs to be replaced with the con-cept of "voice." For Marcus: "Voices are not seen as products of local struc-tures, based on community and tradition, alone, or as privileged sources ofperspectve, but rathe r asproducts of the complex sets ofassociations and ex-periences which compose thern" (66). As 1understand this postulate, today'sincreasingly mobile people need to be studied as mernbers of (severa!) net-works and participants in (several) f lows,rather than as products and produc-ers ofclearly identifiable structures. Table 2 summarizes Marcus's distinctions.Most political scientists and anthropologists who accept at least somepostmodern insights about the constructed and increasingly fluid makeup

    of identity are aware that from time to time ethnographic studies revealthe emergence of rather "solid" entities: dispersed and incongruous iden-tities becorne communities, localized collective memories captivate us asmetanarratives, and voice freezes into structure. They warn, however, thatthis needs to be empirically demonstrated rather than a p r io r i theorized.

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    15/17

    48/ Jan KubikTable 2Three Dimensions of theEthnographic Subject Realist Ethnography Modemist EthnographySpace (Relatively c1osed)

    community of sharedvalues (culture)

    Multi-Iocale,dispersed identity

    Temporal "western" metanarrativeas a background story Localized collective memory

    Perspective Structu re ( social or serniotic] Voice

    Multi-Sited EthnographyForTrouillot, multi-sited ethnography is an improvement, since it is "a par-tial answer to the ethnographic trilogy'( one observer, one time, one place)"(2003, 125). But, how would one conduct a multi-sited ethnography? Mar-cus proposes seven strategies: (1) follow the people; (2) follow the thing;(3) follow the metaphor; (4) fol low the plot, story, or allegory; (5) followthe !ifeor biography; (6) fol low the conflict; and (7) conduct a strategicallysituated (single-site) ethnography (1998,89-99). There is no room here tocharacteri ze them al!. Instead I briefly ilIustrate their usefulness by recon-sidering one of the key tasks of ethnography: tracking down power in un-usual and marginal places (for exarnple, among the subaItern).Marcus suggests that the primary task of cultural/political anthropol-

    ogy is the reconstruction of a complex dialectic of res is ta nc e a nd acc ommo-dation as marginal or subalte rn people try to come to terms with the pres-sures of political and/or economic globalization. This frarnework, in hismind, has served rea!ist ethnography well, at least s ince the first "wzve" ofgloba!ization of anthropology in the mid-1970s. The postmodern worId,however, calls for its retooling. Most important, the mult i-si ted ethnogra-phy that traces multiple loci of action undermines the binary conceptual-ization of the dominant versus the marginal (subaltern). It not only caIlsfor a more systematic focus on the powerful (and on ethnographic studiesof what they actually do and think), but, more irnportant, it prods the re-searcher to look for as many sites of power and counter-power exercise aspossible (see also GledhilI 2000). As a result, "questions of resistance, al-though not forgotten, are often subordinated to different sorts of questionsabout the shape of systemic processes thernselves and complicities withthese processes among variously positioned subjects" (Marcus 1998, 85).

    Ethnography of Politics / 49

    Global Ethnography" ew" ethnography, in addi tion to being postmodern (in Marcus's sense)and multi-sited, isalso supposed to be global. Building on his earlier workon the extended case study method (1998), Burawoy (2000, 26-28) artic-ulates four methodological guideIines: (1) "the extension of the observerto the worId of the participant" (the essence of participant observation);(2) "extensin of the observations over time and space" (following sub-jects through complex and increasingly global networks); (3) "extend-ing from micro processes to macro forces" (relying on a theoreticaIly in-formed model of the external forces whose contingent character needs tobe grasped); and (4) "extensin of theory" (avoiding the "straitjacketing"power of theory by the incessant, mutually correcting dialogue of theoryand data).Studying politics in the worId inhabited by increasingly mobile and glo-

    balized populations calls for new concepts and methods. Ethnography thatis multi-sited, global, and sensitive to postmodern concerns is an intriguingnew tool. But it seerns that it is particularIy effective when it combines newresearch concerns with the tested techniques of traditional positivistic andinterpretive ethnographies.

    ConclusionThe usefulness of ethnography for comparative politics and political sciencein general cannot be assessed without realizing that there is no single eth-nography, but several different types of ethnography. In this chapter 1out-lined three types: positivistic, interpretive, and postmodern (includingmuIti-sited). Each is associated with a different ontology of the social, andeach can help political scientists in different tasks.Positivistic ethnography is indispensable for studying ol lerl ooke d (infor-

    mal dimensions of) power (Abeles 2004); hidd en (faces of) power (Lukes1974; Gaventa 1980); inaccess ibl e (mechanisms of) power, for example inearIy stages of protest mobilization (Bayard de Volo and Schatz 2004,269);ostensibly in co nspi cuous resi s tance to power (Scott 1990); ambiguous (effectsof) power exercise (Wedeen 1999); and cultura l con stru ction of agent s andsubjects of power (Mahmood 2005).lnterpretive ethnography is crucial for exposing the relations between

    power and meaning in concrete situations. lts s ignificancefor political anal-ysis has becorne clearer as a growing number of political scientists-partic-ularIy in comparative politics-work within a constructivist paradigm and

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    16/17

    50 j Jan Kubikdesign their research programs around such principies as on to logical realism ,co nstruc ti vism lint e rpretiv ism , and m icro-focu s on "small scale" mechanisms.Postmodern ethnography is central forcapturing the dynamics of power

    and identity in an increasingly interconnected and globalized world. Multi-sited ethnography, attentive to the novel (gradually more virtual) ways ofconstructing collective identities and focused, in ter alia, on the increasinglytransnational and translocal nature of political and economic transactions,is a promising addition to the methodological armamentarium of today'ssocial science.

    Notes1 would lik e 10 ehan k MO le Ar anoff fo r co mm des hip , in spiraeion , and p acience . Thi s piece owe s aloi 10 /lim. Many e/wnks ea Mi chal Buchow sk i, Am y Unc h, Us a We dee n, Dv ora Yanow , and m yswdencs fo r e xtrem ely usef ul c01 l1111ent son earlier dt ajts of ch is eex t. E dw ard Schacz 's guida nc eand inspir ed edie ing were in dispensable .l. For examples of ethnographic work from other subfields, see the chapters in this vol-

    ume by Pachirat, Walsh, and Yanow, as well as Shdaimah, Stahl, and Schram.2. Aronoff (1991); Chabal and Daloz (2006); Edles (1998); Wedeen (1999, 2002);

    Iohnson (2003). For an innovative marriage of formal modeling and interpretivejsymbolic analysis in the field ofinternational relations, see O'Neill (1999).

    3. Geddes (2003, 23) puts it well: "Although mult iple regression is an excellent tool fortesling hypotheses, it is not always a good image to have i n mind when trying to ex -plain something complicated, because it focuses attention on the idenlification ofcausal factors rather than on how the causal factors work."

    4. For excellent overviews of up-to-date conceptualizations of culture, see Swidler(1986); Sewell (1999); Wedeen (2002).

    5. One might .identify specific characterislics of at least four schools: German, French,British, and American (Barth et al. 2005). But even ifsuch a nation-based typology isrejected as simplistic, three distinct trajectories can be d istinguished (Barnard 2000;Vincent 1990). Bamard proposes to isolate: (1) the "French" sociological traditionrunning from Montesquieu via Saint-Sirnon, Cornte, and Spencer to Durkheim andMauss; (2) the "British" tradition, founded by Ferguson and Smith and running,im er alia, via Maine, Morgan, Tylor, and Frazer to Malinowski and Radcliffe-Brownand their students; and (3) the "American" tradition, also having Montesquieu as itsfounder, but running through Hurnboldt, Grimm, and Bastian 10 Boas (the most in-fluentialthinker here), Kroeber, and Lowie (Barnard 2000).6. Ethnography is also a genre of writing, as I discuss in the section on postmodemethnography.

    7. In his introduction, Schatz argues that some work that uses nonparticipant observa-tion may nonetheless have an ethnographic se nsibility, insofar as it focuses on "in-sider" perspectives. Given my equation of ethnography with participant observation,I leave this possibility 10 the side.

    8. These three traditions emerged in the indicated order, but currently all of them arepracticed.

    9. SeeVincent (1990, 33-77).

    Ethnography of Politics j 5 1lO. This is akin to what Schatz in this volurne's introduction calls an "ethnographic sen-

    sibility."11. The distinction between social structure and social organization has been rnost fa-

    mously introduced and analyzed by Rayrnond Firth (1951). On Firth see Vincent(1990,331) and Bamard (2000,125).

    12. A similar idea is formulated in the work of historical institutionalists (Thelenand Steinmo 1992, 2-3) and in so me game-theoretic analyses (Fearon and Laitin1996).

    13. See Pierson (2004); Mahoney and Rueschemeyer (2003); Ekiert and Hanson (2003);Abbott (2001b).

    14. In more general terms, substantivist approaches tend to conceptualize human re-lations as multistranded and complex, while the formalist approaches emphasizesingle-stranded relations, depending on the issues that are studied.

    15. Anthony Giddens succinctly characterizes positivism: "In nineteenth-century socialphilosophy and social theory positivism was in the ascendant, if positivism is taken10 mean two things. First, a conviction that all 'knowledge, or all that is to count as'knowledge,' is capable of being expressed in terms wh ich refer in an immediate wayto some reality, or aspects of reality that can be apprehended through the senses.Second, a faith that the methods and logical structure of science, as epitornized byclassical physics, can be applied to the study of social phenornena" (1976, 130).

    16. Constructivism has become an accepted and vibrant oplion in many branches ofpolitical science. See, for example, Blyth (2002) in political economy or Adler (2002)and Klotz and Lynch (2007) in intemational relations.

    17. AsTobin (1999, 1)writes, "Paintings, as is the cas e with all cultural production. arenot merely reflections of larger social and economic forces; they participate in theproduction of meanng, in the dynamic construction of identities, and in the struc-turing within discursive fields ofparticular positionalities."

    18. Yet not long ago Wilson observed that "the dialogical relationships between thecrealion and re-creation of national and cultural identities and the same processes atlocallevels have too long escaped critical ethnographic investigation" (1990, 160).

    19. For a discussion of the political impact of witchcraft elsewhere in Africa, see MichaelSchatzberg's contribution 10 this volume.

    20. Following Hall (2003), Iunderstans realism as an ontological stance, assuming thatsocial reality is constructed out of actions of real people, not operations of variables.The attendant methodology requires that, as a minimum, values of variables are"translated" into actions of actual people in "real" contexts.

    21. Moreover, ethnography is the best method to observe meaning-production andmeaning-decoding via nonverbal mechanisms that are often ernployed to cornmu-nicate and exercise power but also to challenge it through counter-hegemonic dis-plays and performances. As Trickett and Oliveri demonstrate, an "ethnographic ap-proach can help discem the meaning of practices that people do not or cannot fullydescribe through verbal rneans" (1997,149).

    22. Consult, in particular, Clifford and Marcus (1986) and Marcus and Fischer (1999).23. AsTyler explains, postmodern ethnography privileges "discourse" over "text," it fore-

    grounds dialogue as opposed 10 monologue, and emphasizes the cooperative andcoJlaborative nature of the ethnographic situation in contrast to the ideology of thetranscendental observer (1986, 126). While the questions of textual representationofsocial reality have been seriously debated in history (White 1978), sociology, andanthropology (Clifford and Marcus 1986), polilical scientists seern far less preoccu-

  • 7/27/2019 Kubik Jan-Ethnograpy of Politics

    17/17

    52/ Jan Kubikpied with studying the impact of the manner in which this reality is represented intheir texts on their knowledge (though see Patterson and Monroe 1998).

    24. For an enlightening discussion, see Brubaker and Cooper (2000) and the HarvardProject on "Identity as a Variable": http://www.wcfia.harvard.edu/miscjinitiative/identityf.

    25. See, for exarnple, Geddes (2003); Lait in (2004); Little (1998).26. Vincent (1990, 388-402); Marcus (1998, 79-80); Trouillot (2003, 120-21).27. This call 10 rethink the relationship berween history and memory ha s been recently

    heeded by sociology, history, and cultural studies; there has been a renaissance ofwriting on historical memory.

    TWO

    How to Tell an Axe Murderer: An Essay onEthnography, Truth, and Lies

    JESSICA ALLINA-PISANO

    In August 2000 and again in 2005, the sci enti fi c journal of reco rd Natureran articles about "polit ical scientists." The sc ientists in question were evo-lutionary biologists who had stepped outside their role as researchers tospeak out aga inst e ffor ts by proponents of the intelligent design move-ment, who wanted their theory of creation taught in public schools along-side Darwinism (Gewin 2005). Around the same time, while teaching aseminar on nationalism at a liberal arts college in the northeastern UnitedStates, I encountered a dean who urged me to "allow the debate" as myundergraduate students contested not how or why, but whether the ThirdReich had killed millions of Iews.' The intelligent design rnovement's at-tempt to cast its clairns, based in b iblical revelation, as a competing theoryto be considered in sci ence classes alongside research-based consensus.' to-gether with the putative "debate" posited in the Holocaust denial literature,highlights a more general problem in the production and transmission ofknowledge: amid a societal commitment to epistemological pluralisrn, howought we to ad judicate truth claims about the worldi"

    In a w orld of epistemological pluralism, truth claims may have no onto-logical status: all claims and perspectives can be treated as ip so Jac to equallyvalido As a result, they may be adjudicated in the political arena, subject tobald contests of power rather th an analytical acumen or spec iali zed knowl -edge evaluated within a common framework of accepted standards of re-search. There is arguably a danger in such instances that populism of themornent, or of a specific segment of society, may overcome knowledge thathas been accrued, tested, and improved by communities of scholars overdecades or generat ions.The political context of these conflicts between particular truth clairns,

    while anchored in a specific historical mornent, points to a more general