Clifford Geertz: Singular Genius of Interpretive Anthropology


  • For comments, encouragement and pointers to references many thanks to Martin van Bruinessen, Martin Doornbos, Linda Herrera, Frans Hüsken, John Roosa, Ratna Saptari, Henk Schulte, Nordholt Pujo Semedi, Ann Stoler, Robert Wessing and the editorial board of Development and Change. Spelling of Indonesian words in pre-1973 quotations has been changed to current usage.


Clifford Geertz (1926–2006), the foremost anthropologist of his generation, was a key figure in the interpretive turn in the social sciences and the re-thinking of boundaries between the social sciences and humanities. In 1970, he became the first professor in the new School of Social Sciences at the Institute for Advanced Study in Princeton, where he created ‘a school of interpretive social science’,1 and devoted himself almost full-time to research and writing for the next thirty-six years. In anthropology, after a period of virtual saturation up to the mid-1980s, his intellectual influence began to wane, but his influence over other disciplines has lasted much longer (Vincent, 1990: 426).

What was the basis of his influence, and what has he left behind? Geertz was a paradoxical scholar in many ways. He drew his inspiration not from the great traditions of anthropology or sociology (with the exception of Max Weber) but from mid-twentieth century linguistic philosophy and literary criticism, creatively transposing these ideas to his own work.2 A master at drawing profound implications from an ethnographic detail, his ethnographic work on Java, Bali and Morocco was criticized as carelessly researched and prone to too-casual generalization. His three major works on Indonesian social history were all completed without once setting foot in an archive. Distrustful of all grand schemes and theories, he placed his first five books firmly in the frame of modernization theory, development studies' grandest evolutionary scheme of all. Awkward in person and in speech —‘a fidgety, scratchy, given-to-mumbling sort of guy’3— he developed a distinctive, flamboyant writing style more self-consciously erudite than that of any other anthropologist. In fact the main source of his intellectual influence was not his ethnographic or historical work but his consummate skill as an essayist, especially in the three volumes of what he called ‘ethnographically informed reflections’ (Geertz 1973, 1983, 2000a). The first of these, The Interpretation of Cultures (1973) is widely seen as marking a major paradigm shift, a Geertzian revolution in the theorization of the concept of culture.

Geertz wrote on a wide range of topics, and any account of his work and legacy must be selective. I have chosen a more or less chronological treatment. After a short account of his early career, this assessment focuses first on his early work on Indonesian religion, politics and agrarian change, next on his general vision on the nature of culture and ethnography, and finally asks some critical questions on the Geertzian vision of culture, politics and social change.


It is important to recall the timing and intellectual and political context of Geertz's formative years, which shaped and constrained his early writing and to which, eventually, he was bound to react. Post-war north American social science, especially as it concerned the study of the ‘new [post-colonial] states’, was driven, and generously funded, by Cold War concerns and anxieties. It was positivistic and functionalist in orientation, and its main theoretical underpinning was modernization theory.

Farmed out by his divorced parents, from the age of three, to a foster mother in the northern California hills, Geertz went to a two-room school in the depths of the Great Depression. He became an anthropologist and an Indonesia specialist through a succession of happenstances. Enlisting at age seventeen in the US Navy (1943–46), his four years of service gave him access, like so many other young men of his generation, to a GI Bill-funded college education.4 He read literature, linguistics and philosophy at Antioch College, and drew inspiration particularly from the ‘New Criticism’ of Kenneth Burke. At first he thought that he would like to be a novelist and a newspaper man, but on graduation he drifted into anthropology with his wife Hildred (who was to become a distinguished anthropologist in her own right). He was lured away from philosophy by his own teachers,5 and into anthropology by a long meeting with Margaret Mead which convinced the Geertzes to accept the offer of graduate fellowships at Harvard's new ‘Department of Social Relations’. Established by Talcott Parsons, the conservative, neo-Weberian structural functionalist sociologist, the Department aimed to further Parson's vision of an integrated, interdisciplinary social science.

Among Geertz's eminent teachers was anthropologist and Russia-watcher Clyde Kluckhohn, the ‘driven, imperious … fierce controversialist, player of favourites, and master money-raiser’ (Geertz, 2002: 4), who involved Geertz in many of his projects. One of these projects plunged him into the heart of anthropology's most contested problematic, ‘culture’, when he was assigned to study and comment on the draft of Kluckhohn's and Alfred Kroeber's Culture: A Critical Review of Concepts and Definitions, which aimed at a definitive compilation of all the prevailing definitions of ‘culture’, finding 171 definitions and sorting them into thirteen categories.6 This dry, inductive exercise must have helped to sow the seeds of his distaste for the vague, all-embracing definition of culture (‘anything human beings might contrive to do, say, imagine, be, or believe’) that held sway in North American anthropology, and his later determination ‘to cut the idea of culture down to size’ (Geertz, 2000b: 12, 13).

As the time for field research approached, Geertz had no special regional interest or preference. Through Kluckhohn's influence in the Massachusetts Institute of Technology's new Centre for International Studies (Cenis), both Clifford and Hildred Geertz were invited to join MIT's new Indonesia Project. In 1952 this project sent the Geertzes and four other Harvard PhD candidates7 to Indonesia's young, revolutionary Gadjah Mada University under an agreement by which they would undertake joint research, each paired with an Indonesian graduate student, in the Central Javanese hill-town of Wonosobo. Not long after their arrival the MIT team left, without their Indonesian counterparts, to carry out field research in and around the East Javanese town of ‘Mojokuto’8 and the research co-operation with Gadjah Mada fell through. Accounts of this breakdown — the first in a long series of troubled relations between foreign PhD researchers and their Indonesian hosts — are divergent. Both Koentjaraningrat (1975: 223) and Geertz (interviewed in 1988 and 1991) note the Indonesian professors' disappointment and suspicion as they found themselves hosts to a leaderless group of American graduate students in what had been negotiated as a much higher-level, professor-to-professor collaboration (Geertz, 1988b: 33–4; and in Olson, 1991: 605).9 The Americans were also alarmed at the large numbers of Indonesian students who planned to join them in fieldwork (Geertz, 1995: 105). Later Geertz laid the blame for the breakdown on the Indonesian professors, who apparently insisted that the American and Indonesian students' fieldwork be conducted from the comfort of an old Dutch resort hotel; local officials would summon people from the surrounding countryside to the hotel to be interviewed in groups, in an ‘extraordinary reincarnation of the pith-helmet procedures of colonial ethnology’ (ibid.). But he also recalls that at the time, ‘an armed gang of leftist rebels controlled much of the countryside’ around Wonosobo (ibid.: 106), which helps to explain the Gadjah Mada professors' reluctance to let their young American guests loose there. Wherever ‘truth’ lies (somewhere in between the various accounts, perhaps) it is the later ‘pith-helmet’ version which will stick, now adopted and further exaggerated by Geertz's intellectual biographer (Inglis, 2000: 3–14).

Under its Director Max Millikan (previously the CIA's Director of Economic Research), Cenis aimed at ‘the production of an alternative to Marxism’ (Rosen, 1985: 29). Other core staff included Walt Rostow, who was busy writing The Stages of Economic Growth: A Non-communist Manifesto (Rostow, 1960). Kluckhohn, who had brokered the Ford Foundation funding for the Indonesia project, was himself an important conduit for CIA area studies funds (Ross, 1998). The Indonesia project as a whole was intended to explore a ‘common thesis’ on the contexts and conditions of Rostovian ‘take-off’ into economic growth, summarized in the words of the project director: ‘A “take-off” is most likely when the growth of an indigenous entrepreneurial class occurs simultaneously with the appearance of a political elite which has the power and the will to provide a policy framework favourable for the exercise of entrepreneurial talents’ (Higgins, 1963: ix). Geertz's task in the team, to study ‘Javanese religion’, was also conceived in relation to this broader scenario of the entrepreneurial roots of economic growth; his thesis proposal was ‘a test of the Weberian hypothesis, that the strongly Muslim sector would be the functional equivalent of the Protestants in the Reformation’ (Geertz in Handler, 1991: 605).

In Mojokuto, the Geertzes lived in the house of a railroad worker on the fringe of town, spending their time mainly in extended interviews with key informants and participant observation. Returning to Cambridge, Geertz was offered a research position at Cenis where he embarked on a great rush of writing on a wide range of topics, overlapping considerably with those assigned to his fellow Mojokuto team members. In 1956, besides completing his 700-page thesis on the religion of Java, Geertz produced two book-length monographs for the project: The Development of the Javanese Economy: A Socio-Cultural Approach (Geertz, 1956a) and The Social Context of Economic Change: An Indonesian Case Study, a social history of Mojokuto (Geertz, 1956b). These monographs attracted little attention at the time, but most of Geertz's many publications in the following decade can be seen as the working out, topic by topic, of ideas first sketched out in them (Smail, 1965).

New fieldwork in Bali (1958) was followed by a teaching job at Berkeley and a year at Stanford's Centre for Advanced Studies. Among the fellows at Stanford, Edward Shils and David Apter from the University of Chicago persuaded Geertz to join them in the new ‘Committee for the Comparative Study of New Nations’, which also sponsored a teaching appointment. In the next decade at Chicago, Geertz produced a flood of landmark publications. Besides numerous articles and chapters on Indonesian religion and rural society, six books emerged in eight years: The Religion of Java (Geertz, 1960), Agricultural Involution (Geertz, 1963a), Peddlers and Princes (Geertz, 1963b), Old Societies and New States (Geertz, 1963c), The Social History of an Indonesian Town (Geertz, 1965), and (after new fieldwork in Morocco in 1963) Islam Observed: Religious Development in Morocco and Indonesia (Geertz, 1968). By the mid-1960s it had become impossible to discuss almost any aspect of Indonesian society, or more general debates on Third World religion or politics, without reference to Geertz's work.


Geertz's best-known books on Indonesia are The Religion of Java (1960), Agricultural Involution (1963a) and Negara: The Theatre State in Nineteenth-Century Bali (1980).10The Religion of Java11 is the first, masterly exemplar of the classic Geertzian style which so many subsequent generations of anthropologists have imitated. The book has no clear objectives and makes no effort to locate itself in relation to previous scholarship or ongoing debates on Indonesian society, or on Islam.12 It is simply an extraordinary, readable and powerful work of descriptive ethnography. The informants ‘speak for themselves’ (or an eye-witness account of an event is described by the ethnographer), and the ethnographer often appears himself, something which earlier generations of anthropologists might not consider quite proper, as the material is relayed verbatim in the standard Geertzian indented excerpts from field notes, as in the following example

I talked to Joyo on the corner the other night about his marvelous grandfather … He said his grandfather was able to disappear magically. Also he could go great distances in a short time. He would walk out of the house and announce to his wife that he was going to Semarang [three or four hundred miles away] and in fifteen minutes he would walk back in … His grandfather was arrested once by the Dutch and taken to Bragang and put into jail because of his ilmu[science]— all his pupils walking along behind him as he was led in. When they returned home, they found him there in the house ready to teach, and it turned out that he was in both places at once … (Geertz, 1960: 89)

Joyo's tall tale, relayed neutrally in simple prose and without further comment or interpretation, is ‘admitted, not as a curiosity, but as something that reasonably passes between two persons’ (Siegel, 1995: 94); ‘I talked to Joyo on the corner the other night’ places the reader right there, listening to Joyo over Geertz's shoulder, and invited to reflect on what he says, making the overall effect of the episode unforgettable (Anderson, 1995: 20). This ‘being there’ approach to field research and writing is a source of the great evocative power of Geertz's work, but also of some of its weaknesses. As critics were quick to note, The Religion of Java embodied quite serious problems in Geertz's depiction of the Javanese social world, and the place of Islam in that world.

In this and other works, Geertz distinguished three main aliran (‘streams’) or ‘cultural types which reflect the organization of Javanese culture’: the santri (orthodox, modernist), abangan (Javanist, syncretist) and priyayi (Indic, Hindu-Buddhist) varieties of Javanese Islam (Geertz, 1960: 4). This tripartite horizontal social cleavage is put forward with typical Geertzian panache, as ‘too apparent to be missed by even the most positivist sociologist’ (Geertz, 1965: 124). Both Indonesian and foreign scholars were quick to point to Geertz's conflation of horizontal (religious orientation and world view) and vertical (social class) axes of social distinction, both of which are equally important in Javanese society, and which do not coincide but cut across each other; priyayi refers to a distinction of social class (opposed to wong cilik or common people) rather than of religious culture, and priyayi may be either orthodox or syncretist Muslims (Cruikshank, 1972; Bachtiar, 1973; Hefner, 1987; Koentjaraningrat, 1960). This is an early example of the reluctance to recognize the importance of distinctions of class and power which runs like a red thread through Geertz's work.

Geertz's determination to explore ‘everyday Islam’ through ethnographic study of the practices through which ordinary Muslims experienced their religion (Hefner, 1999: 14), was a healthy reaction to prevailing Orientalist traditions of historical and literary (textual) scholarship on Islam. However, Geertz seems to have accepted uncritically his modernist santri informants' narrow views of what is really Islamic, dismissing the religious practices of abangan and priyayi as ‘Hindu-Buddhist’ and un-Islamic. He thus embraced, perhaps unwittingly, an updated version of colonialism's comfortable, Orientalist vision of Islam as a thin, late-deposited cultural ‘veneer’ floating loosely on the bedrock of ‘real’ eastern culture, and in this way contributed to the relative marginalization of Islam in Indonesian (and more generally Southeast Asian) studies, which later scholarship has had to redress (Hefner, 1999: 11–16).

Agricultural Involution: The Processes of Ecological Change in Indonesia (1963a) set the agenda for much subsequent research on social change in rural Java for the next two decades. The book is not directly based on field research — research findings from ‘Mojokuto’ or other locations are not referred to at all — but on a few, mainly Dutch, publications on Indonesian colonial history. Using statistics from 1920 to show a positive relation between sugarcane production, population density and per-hectare rice yields, and ‘doing history backwards’ to ‘figure out how the situation characteristic of this later period could have been produced’ (Geertz, 1963a: 70), Geertz argued that the Javanese response to colonial pressures on production and population growth had consisted of a combination of labour intensification in subsistence production, and increased complexity on both the technical and the social side of agriculture. Geertz invoked the concept of involution, ‘the overdriving of an established form in such a way that it becomes rigid through an inward overelaboration of detail’ (ibid.: 82)13 (contrasted, therefore, with ‘evolutionary’ or developmental patterns of change) to depict the cultural response of Java and other regions of ‘inner Indonesia’ to Dutch colonial policies, made possible by their ecological setting. Irrigated rice terraces, he argued, can respond to labour intensification almost indefinitely without loss of soil fertility, but with only stable or declining per capita output, and this allowed a symbiotic rotation of the main export and subsistence crops (sugarcane and paddy) from the early nineteenth century until Indonesian independence and beyond.

Agricultural change thus consisted mainly of involutionary ‘technical hairsplitting and unending virtuosity’ in ‘pregermination, transplantation, more thorough land preparation, fastidious planting and weeding, razor-blade harvesting, double-cropping, a more exact regulation of terrace flooding, and the addition of more fields at the edge of volcanoes’ (ibid.: 77–8). The hallmark of Java's involuted rural society of the 1950s was a relative social-economic homogeneity of village communities, encapsulated in the phrase ‘shared poverty’,14 achieved by ‘dividing the economic pie into a steadily increasing number of minute pieces’ through the elaboration of traditional land-tenure and labour relations, ‘mechanisms through which the agricultural product was spread, if not altogether evenly, at least relatively so, throughout the huge human horde which was obliged to subsist on it’ (ibid.: 97); this elaboration being ‘matched and supported by a similar involution in rural family life, social stratification, political organization, religious practice, as well as the “folk-culture” value system … in terms of which it was normatively regulated and ethically justified’ (ibid.: 101–2).15

Agricultural involution acts against development, threatening for example to incorporate any productivity changes through technical change into further stagnation ‘in the absence of any genuine reconstruction of Indonesian civilization’ (ibid.: 146). This ‘reconstruction’ would require the breaking down of involutionary constraints to the emergence of ‘a modern, commercial farming community’ (ibid.: 90), or as Geertz explained in the Social History of an Indonesian Town: ‘one of Java's greatest needs: a “virile yeomanry”’ (Geertz, 1965: 49).

Subsequent, detailed research by social historians, anthropologists and others has undermined almost every argument in the involution thesis, leaving it as ‘a brilliant hypothesis, brought down by available evidence’ (Brown, 1997: 110). These criticisms questioned both involution's supposed ecological basis (for example, ecological mutuality of sugarcane and rice) and its economic and social features (labour-intensification, and ‘shared poverty’). As Wertheim remarked, Geertz's vision mirrored the blindness of colonial and post-colonial élites, whose idea of the harmonious and homogeneous village community was derived from, and promoted by, the village élite themselves (Wertheim, 1975: 177–214; cf. Utrecht, 1973: 40). There is certainly a striking lack of fit between Geertz's accounts of Javanese homogeneous rural and small-town culture and the many violent political conflicts in the region both before and after his fieldwork. Many of the critics (including this author) had originally admired and even been inspired by his work; but they emerged from fieldwork (or, in the case of the social/economic historians, from the colonial archives of The Hague or Jakarta) with an empirically grounded reading of Javanese society which they simply could not find in the pages of Geertz's book. Irritated by these criticisms and particularly by this author's review article (White, 1983),16 Geertz responded in kind, dismissing the critics as having failed to see what the book was about; failing to see the involution thesis in its relation to his other work, and approaching the study of rural Javanese society with the wrong theoretical and methodological baggage, but still leaving readers feeling that he had side-stepped the main points of criticism (Geertz, 1984).

Peddlers and Princes, also published in 1963, pursued a parallel search for an emerging middle class on the urban side, by comparing small-town entrepreneurial practices in Mojokuto and Tabanan (Bali). It concludes with six propositions on ‘pre-take-off social change’, all of them about the dynamics of ‘innovative economic leadership (entrepreneurship)’ (Geertz, 1963b: 147–53). A third book published that year rounded off Geertz's early contributions to modernization theory, by turning his attention to Third World politics. In Old Societies and New States: the Quest for Modernity in Asia and Africa, Geertz's much-cited essay ‘The Integrative Revolution: Primordial Sentiments and Civil Politics in the New States’ adopts Shils's (1957) notion of ‘primordial ties’ based on (assumed) ties of kin, race, language, region, religion or custom. The conflict between ‘primordial’ and ‘civil’ sentiments threatens governments and nations and requires, in Geertz's view, an ‘integrative revolution’ led by modernizing élites to defuse, dismantle or at least to ‘domesticate’ primordial sentiments and reduce them to civil order (Geertz, 1963d). This general schema was the subject of much debate among political scientists, following an early critique by Doornbos (1972); further critique of these notions as applied to Indonesian political dynamics was to follow in the 1980s and 1990s.


In all his early work Geertz was preoccupied, like his teachers and mentors at Harvard, MIT and Chicago, with the problems of modernization. Agricultural Involution, Peddlers and Princes and ‘the integrative revolution’ all make frequent reference to the modernization theories of Parsons, Rostow and Shils. While firmly wedded to their ideas on development, Geertz was increasingly uncomfortable with their positivistic approach to social science epistemology and method and it was inevitable that he would make an explicit break with this tradition. His Cambridge and Chicago years also saw the emergence of fierce debates on the concept of culture. In the 1960s, as the younger generation tried to make sense of Marxism and Levi-Straussian structuralism, North American anthropology became the site of polarized struggles and stand-offs between the advocates of a (positivist, materialist) ‘science of culture’ (Harris, 1968, 1979) on the one hand, in which human affairs are essentially ‘caused by the ways human beings cope with nature’ and on the other a more humanistic, ‘mentalist’, qualitative approach which saw ‘humankind as spinning ever more complex webs of signification through autonomous processes of the symbolic faculty’ (Wolf, 1984a: 148). All of Geertz's publications of the 1960s show a clear alignment with the ‘symbolic’ approach, although he preferred to give his version of it another name.

Once at Princeton —‘that island of upmarket composure’ (Geertz, 2000c: 53) — and secure in his reputation, Geertz finally freed himself from his uneasy relationship with the modernization theorists, threw down the gauntlet against the functionalist and positivist science represented by his Harvard and Chicago mentors, and argued eloquently, forcefully and often for what he called ‘interpretive anthropology’. The Interpretation of Cultures (Geertz, 1973) is a collection of twelve essays published between 1957 and 1972. While some of these essays already appeared outdated, the book's great and lasting impact — which makes it figure in various lists of the twentieth century's greatest books — came from its introductory and closing essays. In ‘Thick Description: Toward an Interpretive Theory of Culture’, Geertz attacked eclectic notions of culture — singling out his mentor Kluckhohn's work as an example of the ‘conceptual morass’ which they reflect (Geertz, 1973: 4–5)17— and proclaimed a narrowed, semiotic concept of culture, encapsulated in the ever-quoted passage: ‘Believing, with Max Weber, that man is an animal suspended in webs of significance he himself has spun, I take culture to be those webs, and the analysis of it not to be an experimental science in search of law but an interpretive one in search of meaning’ (ibid.: 5).

If culture is webs of meaning, what then is ethnography, the anthropologist's practice of studying it? To explain this, Geertz invoked Gilbert Ryle's distinction between a ‘twitch’ and a ‘wink’, and ‘a not untypical excerpt from my own field journal’: a Moroccan tale of murder, robbery and kidnapping involving a tribal sheikh, the Jew Cohen, the local commandant Captain Dumari, herds of sheep and various other characters. What ethnographers do is what Ryle called ‘thick description’: ‘sorting out the structures of signification … and determining their social ground and import’, the result of which he calls, reflexively, ‘our constructions of other people's constructions of what they and their compatriots are up to’ (ibid.: 9). Doing ethnography, he wrote, ‘is like trying to read (in the sense of “trying to construct a reading of”) a manuscript’ (ibid.: 10).

The closing essay, ‘Deep Play: Notes on the Balinese Cockfight’, exemplifies the new paradigm of how to study culture. It begins again with a fieldwork story, but this time one which gets the writer directly into the action. When in 1958 the Geertzes moved into a Balinese village, their initial efforts to build ‘rapport’ were met with studied indifference: people ‘acted as if we simply did not exist’. Everything changed after some days when the police raided an illegal cockfight they were attending, and they ran away with the other villagers. When the police marched into the compound in which they had taken refuge and demanded to know what they were doing there: ‘our host of some minutes leaped instantly to our defence, producing an impassioned description of who and what we were, so detailed and so accurate that it was my turn, having barely communicated with a living human being save my landlord and the village chief for more than a week, to be astonished’ (ibid.: 412–15). This anecdote actually has little relevance to the intricate analysis of Balinese cockfighting and its interpretive function which follows, but this does not seem to have bothered readers; the rhetorical technique has since become ‘a major (and now perhaps, dully repetitive) strategy of both writing and analysis in ethnographic, historical and literary scholarship’ (Marcus, 1999: 106).

These two essays sent ‘a tidal wave across the disciplines’, by showing:

how to take a piece of culture — a ritual, a tall tale, a performance, a symbol, or an event — and treat it as ‘text’… Liberated from the rigours of explanation and able to take as a focal text any piece of the social world, great or small, historians …, literary critics … and even policy analysts … were freed to put culture centre-stage. (Swidler, 1996: 300).

Geertz thus located culture and ethnography squarely within the semiotic turn in anthropology, and the general revival of intellectual interest in hermeneutic (meaning-centred) approaches in many other disciplines. But he also distanced himself implicitly from the more rarefied schools of symbolic and cognitive anthropology, and structuralism, that treat culture as purely a symbolic system, and thus ‘run the danger … of locking cultural analysis away from its proper object, the informal logic of actual life’ (Geertz, 1973: 17). In another equally well-quoted passage (later forgotten by many of his followers) he argues: ‘If anthropological interpretation is constructing a reading of what happens, then to divorce it from what happens — from what, in this time or that place, specific people say, what they do, what is done to them, and from the whole business of the world — is to divorce it from its applications and render it vacant’ (ibid.: 18).

For the next decade, while the Indonesia specialists, irritated by his apparent carelessness and disregard for evidence, engaged in a long season of Geertz-bashing, Geertz himself emerged as north American anthropology's superstar (a position formerly occupied by Margaret Mead, who died in 1978) and its chief ambassador to other disciplines, with major influence on the other social sciences, literary studies, philosophy, and beyond (Sewell, 1999: 35).


Negara: The Theatre State in Nineteenth Century Bali (1980), Geertz's most important work on Bali, is the only one of his Indonesia books written after he had abandoned modernization theory, although many of its main ideas had been sketched out in earlier publications (Geertz, 1966a, 1977; Geertz and Geertz, 1975). Negara is also the only one of Geertz's works which makes significant use of the work of Indonesian scholars.18

Like Agricultural Involution, Negara adopts a strategy of ‘doing history backwards’ and is not mainly based on his village fieldwork, although that has informed the chapters on kinship and irrigation organization. Available (mainly Dutch) literature and retrospective interviews with elderly Balinese are invoked to build a much broader thesis about the nature of the ‘classical’ Balinese state. The precolonial Balinese state, Geertz argued, contravened Western theories of politics and power, being first and foremost concerned with spectacle and symbolism, not as means to (political or economic) ends but as ends in themselves:

The expressive nature of the Balinese state was apparent through the whole of its known history, for it was always pointed not toward tyranny, whose systematic concentration of power it was incompetent to effect, and not even very methodically toward government, which it pursued indifferently and hesitantly, but rather toward spectacle, toward ceremony, toward the public dramatization of the ruling obsessions of Balinese culture: social inequality and status pride. It was a theatre state in which the kings and princes were the impresarios, the priests the directors, and the peasants the supporting cast, stage crew, and audience. … Power served pomp, not pomp power. (Geertz, 1980: 13)

This bold vision required Geertz to make the heroic assumption that elderly informants interviewed in the late 1950s could reliably recall conditions of precolonial Bali, more than a half-century earlier, and also that no important structural changes had occurred in Bali from the fifteenth century until the Dutch conquest in 1906–8. It is thus actually a kind of ‘anti-history’, ostensibly situated in time but greatly underplaying historical change (Errington, 2007: 196). As with Agricultural Involution, subsequent detailed work by various authors (in this case, with colonial archives and court manuscripts) has eroded Geertz's claim that Bali's royal elites were not involved in precolonial village rule, irrigation organization, taxation and the like. They suggest, instead, that what Geertz saw as ‘classical’ Bali was, in fact, ‘the very result of 20th century colonial policy and ideology’ (Schulte Nordholt, 1981: 474):

Anthropologists like Geertz and [Stephen] Lansing have, for reasons of their own, overlooked the role of the king as a ruler … The kings were dispossessed of their power to manage and control irrigation, manpower, taxes, and adjudication by the bureaucratic colonial state … In sum, the kings were left with the roles that anthropologists have labeled the ‘theatre state’ and the peasants with the ‘democratic irrigation model’. (Hauser-Schäublin, 2003: 170)


Paradoxically, the moment when history and other disciplines began to take serious notice of Geertzian anthropology was precisely the moment when anthropologists began to question it (Rabinow, 1986: 241–2). William Sewell, a social historian, encapsulates the ‘edgy relationship’ between Geertz and his anthropological colleagues:

The positivists criticize Geertz for abandoning … scientific values in favour of the more ‘glamourous’ or ‘alluring’ qualities of interpretive method. The post-modernists, by contrast, reproach him for not pushing his interpretive method far enough — in particular, for failing to subject his own interpretive ethnographic practice to critical interpretation. The materialists, finally, criticize him for his neglect of history, power, and social conflict. (Sewell, 1999: 35–6)

The years 1982–4 in particular saw a barrage of critique of many of the key ideas on which Geertz had made his reputation, in the attacks on his Balinese work (Roseberry, 1982), on agricultural involution in Java (summarized in White, 1983), on his conceptions of religion (Asad, 1983) and ideology (Wolf, 1984b), primordialism (Anderson, 1983) and on the interpretive paradigm generally (Shankman, 1984). Geertz, however was, more often than not, conspicuously absent from these and other debates surrounding his work, both in the 1980s and later. He declined to join an otherwise balanced discussion of Shankman's positivist critique in a Current Anthropology debate: ‘the enigmatic silence of Clifford Geertz hovers over this commentary like a cloud’ (Shankman, 1984: 276). When persuaded or provoked to ‘join in the merriment’, as he once put it, his rejoinders to criticism were acerbic —‘this dog is mean, he bites when he is kicked’19— and often side-stepped the main points of criticism.20 From the mid 1980s onwards his writing, still prolific, was largely devoted to literary criticism and reflective commentary: on the method and style of other anthropologists, review essays on a variety of themes aimed at a broad North American intelligentsia in the New York Review of Books and The New Republic, and from the late 1980s onwards, a series of reflections on his own career in books (Geertz, 1995), essays (Geertz, 2000a, 2002), lectures (Geertz, 1999) and interviews (Geertz, 1988b; Handler, 1991; Micheelsen 2002; Olson, 1991).

Culture, Politics and Social Change

Geertz argued that ideology (Geertz, 1964), religion (Geertz, 1966b), common sense (Geertz, 1975), art (Geertz, 1976), and many other aspects of social life and ideas are ‘cultural systems’, and (as we have seen) that culture itself is a general system of symbolic forms, to be understood in their actual practice. But he never explained precisely what he thought ‘symbols’ are, what he really meant by ‘systems’ of symbols, or how they should be analysed (Lieberson, 1984: 11); indeed he proclaimed, late in life: ‘I am an ethnographer … from beginning to end; and I don't do systems’ (Geertz, 2000a: x).

On the Geertzian method, Elizabeth Colson noted in an early review that Geertz's ethnography ‘does not provide a model for other anthropologists or sociologists of lesser talent to follow since he proceeds from an intuitive grasp of what is important and reaches his conclusion with a flourish that conceals the tedium of the procedures’ (Colson, 1975: 637–8). A great deal of debate around Geertzian interpretivism has centred around the lack of clear criteria for evaluating alternative cultural interpretations, or ‘intuitive grasps of what is important’. In the mid-1980s Geertz joined the debate by turning his talents as a literary critic on to his own discipline, most notably in Works and Lives: The Anthropologist as Author (1988a). Taking selected works of Claude Lévi-Strauss, Edward Evans-Pritchard, Bronislaw Malinowski and Ruth Benedict, he exposed their distinctive styles of writing in ‘new criticism’ mode, and asked where lay their ‘capacity to persuade readers … that what they are reading is an authentic account’ (Geertz, 1988a: 143). Arguing in this and other writings that anthropology is ultimately a literary and ‘rhetorical’ vocation, Geertz came perilously close to arguing that the best interpretations of culture and social life are simply the best-written ones, a position taken up by various post-modernist authors and carried to extremes from which he had to distance himself. Geertz thus provided fuel for the debate among postmodernists around ethnography's ‘double crisis’ of representation and of legitimation (Brewer, 2000: Ch. 2; Clifford, 1988; Clifford and Marcus, 1986; Marcus and Fischer, 1986: 45–76).

Despite Geertz's warning against those varieties of hermeneutics which divorce culture from ‘the whole business of the world’, this seems to have been forgotten by many of his followers, who continue to neglect ‘issues of power, interests, economics, and historical change … in favour of simply portraying the native point of view as richly as possible’ (Marcus and Fischer, 1986: 77; cf. Hefner, 1990: xiii). This has led some scholars to question the ability of interpretive frameworks ‘to explicate the things we most wish to understand’ such as contemporary processes of domination, displacement, marginality and violence (Steedly, 1999: 432).

The fact that social historians, who are professionally concerned with questions of transformation over time, should have been so strongly influenced by Geertz's ‘insistently synchronic’ work, is something of a paradox. Geertz's cultural systems do appear somewhat impervious to change, but the Geertzian approach did open the door to modes of historical enquiry and exegesis based on ‘thick description’, at a time when ‘social’ history was gaining in popularity. If historians want to combine this with diachronic analysis, however, they need to adopt a different theory of culture than Geertz's, focusing not on the assumed unity of cultural systems but on multivocality and internal differentiation as sources of change (Sewell, 1999). Both historians and anthropologists have argued that Geertz's vision of cultures as systems of shared symbols (and associated practices) blinded him to questions of social differentiation, social conflict, and associated negotiations and contestations over meanings. Who in society gets to spin the ‘webs of significance’, and who gets caught in the web? These questions bring us into areas of enquiry which ‘thick description’ cannot so easily reach, and raise questions about Geertz's approach to politics.

In general, Geertz appeared indifferent to the trend to place issues of power and history more centrally in anthropology,21 as reflected in Roseberry's (1982) critique of his work on Bali and Asad's (1983) critique of his approach to religion. Geertz's own political stance was more or less that of the liberal North American humanist; he rarely took public stands or joined campaigns: ‘my own general ideological position is largely the same as that of Aron, Shils, Parsons and so forth; … I am in agreement with their plea for a civil, temperate, unheroic politics’ (Geertz, 1973: 200)

Geertz's avoidance of any serious discussion of the Indonesian mass murders of 1965–6, and what they mean for our understanding of Indonesian politics, is both puzzling and revealing. The bulk of the killings occurred in Geertz's two fieldwork regions of Central/East Java and Bali. Some weeks after the crushing of a bungled leftist coup attempt in Jakarta (which killed twelve persons in total), an orchestrated anti-Communist backlash resulted in the massacre of hundreds of thousands by the army and army-trained militias. In the years that followed Geertz alluded only in passing to the killings, ‘with a weary resignation, in which striking insights are encompassed in turns of phrase full of the kind of detachment and wryness that has angered his younger critics’ (Marcus, 1999: 107, fn 15). In a 1973 postscript to his (pre-1965) article on primordialism, Geertz described the killings as having been done ‘in Java at least, mainly along … primordial lines — pious Moslems killing Indic syncretists’ (Geertz, 1973: 282). On Bali, where the killing was relatively more severe than in any other region, Geertz wrote in the landmark ‘cockfight’ article: ‘if one looks at Bali … also through the medium of its cockfights, the fact that the massacre occurred seems, if no less appalling, less like a contradiction to the laws of nature’ (ibid.: 452). Such ‘cultural’ accounts of the massacres became untenable as the years passed and more became known: ‘it is clear that the military bears the largest share of responsibility and the killings represented bureaucratic, planned violence rather than popular, spontaneous violence’ (Roosa, 2006: 28). On Bali, Geoffrey Robinson's historical account of political conflict along class, caste and ideological lines offers a powerful counterpoint to aliran- and ‘primordiality’-centred views (Robinson, 1995; Sidel, 1997).

Three decades later, in After the Fact, Geertz describes the massacres in and around Mojokuto, mainly through a two-page excerpt from the 1971 account of a retired Nationalist Party leader: ‘In the beginning, things could have gone either way. Each side was trying to kill the other side first, and when the Communists saw that the Muslims had the upper hand, they just gave up’ (Geertz, 1995: 10). In relaying the old nationalist's account without critical comment — the only version that Indonesians were permitted to parrot under the new regime — Geertz appears virtually to endorse the official view of the Indonesian army and the CIA, that the slaughter of Communists was a matter of self-defence and therefore justified (Reyna, 1998). Accounts of such contested matters require the ethnographer to be more than ventriloquist, to go beyond ‘Joyo-on-the-corner’ modes of exposition, and to have a point of view.

Dancing with Words

By his own account Geertz wrote all his work by hand, very slowly —‘about a paragraph a day’— and never wrote drafts: ‘I write from the beginning to the end, and when it's finished it's done … I would not advise that other people write this way’ (in Olson, 1991: 5). The result is a unique, dense prose: erudite, meandering, seductive, often playful and self-deprecating, difficult to ignore: ‘Almost everyone initially gets side-tracked by the visibility and distinctiveness of Geertz's writing style, which is like Cyrano de Bergerac's nose. It is conspicuous, it is spectacular, but … it is best just to ignore it, for the sake of getting on with a discussion of his ideas’ (Shweder, 2005: 2).

From about the mid-1970s onwards, when Geertz was more or less finished with fieldwork and became a full-blown essayist, something seemed to happen to his writing style as the self-conscious flourishes threatened to take over the substance (Anderson, 1995: 20). Local Knowledge, the second volume of essays written between 1974 and 1982 (Geertz, 1983), did not develop his central position on culture and ‘interpretive anthropology’ but consisted mostly of general reflections on dissimilar topics, in prose which, ‘for all its irreverence and occasional brilliance, is too often digressive and clotted with metaphor, tending to a self-conscious virtuosity that is at first immensely attractive but in the end forced and almost oppressive. Drugged into a kind of exhilaration by his racing sentences … one sooner or later realizes that one has no clear idea of his view’ (Lieberson, 1984: 11).

A few years later Works and Lives (Geertz, 1988a) provoked Edmund Leach to vent his irritation at Geertz's habit of tacking long, comma-riddled lists (of names, places, events, schools of thought) on to every point: ‘every point of argument is reinforced as if it needed to be supported by a thesaurus. The resulting garrulity quickly becomes intolerable’ (Leach, 1989: 139). Many younger readers, however, were simply captivated by Geertz's style. It is hardly fair to blame Geertz (or Foucault, or other authors with distinctive writing styles) for those who try, with less success, to imitate them, but the ‘new fashion of incomprehensible writings by less talented followers’ (Schulte Nordholt, 2007: 34) is one of his less fortunate legacies.22 In 2002, at a special session of the American Anthropological Association honouring his life and work, Geertz responded to various good-natured jibes about his ‘shaggy sentences, list-laden discursivity, and prozy maze’. He took explicit issue with Shweder's recommendation to separate the style from the ideas: ‘Having toiled over [my style] for many years, I am quite aware of its deficiencies … I … question whether style and substance are so easily separable … I do think that much of what I have to say inheres in how I say it …’, and added some advice to younger anthropologists: ‘It pays, really pays, in hard, tenurable cash … to be readable. Better over the top than under the pile’ (Geertz, 2005: 109, 112)


Despite his skills in ethnographic field work, Geertz was an awkward and somewhat reclusive person, and many found his erudition and encyclopaedic knowledge intimidating, although he got on well with students, preferring to teach undergraduates. Certainly not a born teacher, he could also inspire, as recalled by Robert Darnton who taught a Princeton undergraduate course with him for many years:

He talked too fast and mumbled into his beard so badly that the students found it difficult to understand him … The rumpled, disheveled figure at the far end of the table frequently said nothing, apparently lost in its own thoughts. Then suddenly it would explode in talk. The words would tumble out in a torrent, and we would sit back amazed … When his eyes lit up, and the words poured out, he infected students with the excitement of the chase. (Darnton, 2007: 32–3)

The excitement which Geertz's work provoked, and his greatest legacy, lie not so much in research ‘findings’ as in his vision of culture and ethnography, and in the evocative power of his writing. For four decades, he kept Indonesia specialists and anthropologists inspired and irritated in equal measure by his research, and a much wider group of intellectuals inspired by his vision: for many social scientists Geertz literally ‘changed the way we study culture’ (Swidler, 1996: 299).

Geertz's legacy is not straightforward; it ‘endures by way of, and through, the critique that succeeded him’ (Marcus, 2001: 167). His polemical stands for interpretation against ‘explanation’, for description against theory, and against all general theory, were important at the time, to make the point. But they are now largely red herrings (Swidler, 1996: 302). To express these contrasts in ‘either/or’ terms is to misunderstand the nature of both the natural sciences and of the humanities. Hermeneutics is not the only legitimate approach to social analysis; good anthropological (and sociological, and historical) research combines and feeds on the dynamic tensions between particularistic ‘thick descriptions’ and comparative enquiry into larger realities,23 and it is the engagement of general theory and localized research that makes anthropology a ‘discipline’ (Hefner, 1990: xii): the most humanistic of the sciences, and the most scientific of the humanities.


  • 1

    Geertz, interviewed in Handler (1991: 610)

  • 2

    In particular, Gilbert Ryle and Ludwig Wittgenstein among the philosophers, and Kenneth Burke in the ‘new’ literary criticism.

  • 3

    In the words of David Apter, a close friend since the late 1940s (Apter, 2007: 111).

  • 4

    Roosevelt's (1944) GI Bill of Rights gave some eight million Second World War veterans access to job training or higher education, and was covering the costs of half the nation's expanding college student population by 1947 (Branch, 2007: 40).

  • 5

    Thomas Geiger, the disenchanted Deweyite philosopher —‘the greatest teacher I have ever known’— warned him that philosophy had fallen into the hands of ‘Thomists, technicians and … the University of Chicago’ (Geertz, 2000c: 53).

  • 6

    Geertz (2000a: 12), referring to Kroeber and Kluckhohn (1952)

  • 7

    Alice Dewey, Donald Fagg, Robert Jay and Edward Ryan.

  • 8

    Literally ‘Middletown’, the sub-district capital Pare in the heart of the important sugarcane producing region of Kediri, East Java.

  • 9

    At the last minute the team leader Professor Douglas Oliver had decided not to join the team.

  • 10

    Peddlers and Princes (1963b), a comparative study of entrepreneurship cast in the language of modernization and ‘take-off’ theory, has not stood the test of time; The Social History of an Indonesian Town (1965) , in my view one of Geertz's best-written books, attracted little notice outside of Indonesian studies.

  • 11

    The book is a largely unrevised version of Geertz's 1956 PhD thesis (supervised by Cora Dubois), minus the introductory and concluding chapters.

  • 12

    The book has literally less than two dozen citations and no bibliography, and makes only passing reference to a few previous studies on Indonesian society and two general works on ‘Mohamedanism’.

  • 13

    The term ‘involution’ is borrowed from Kroeber's student Alexander Goldenweiser (1936) who had applied it in the 1930s to Maori woodcarving.

  • 14

    Geertz claimed his use of the term ‘shared poverty’ was an allusion to Winston Churchill's characterization of communism (Geertz, 1984: 527), though many have considered it more likely borrowed from Boeke.

  • 15

    Geertz never attempted a more detailed analysis of involutionary change in Javanese culture generally, although he had intended to do so: ‘The one thing I do regret about Agricultural Involution is that I didn't, as I had originally planned, connect up my ecological argument with a general analysis of Javanese culture, but I just felt at the time that if I tried to do that I would be writing the damn thing for a quarter of a century' (Geertz, letter to the author, 15 April 1983).

  • 16

    The review did strain the bounds of academic decorum by making indirect reference to Geertz's work on Indonesia as ‘imperialist software’ (White, 1983: 30, citing Wertheim, 1973 and Utrecht, 1973).

  • 17

    The reference is to Kluckhohn's (1949)Mirror for Man, which defined culture in eleven different ways in the space of a few pages.

  • 18

    While Geertz was revered among Indonesian scholars, he seems to have had a very low opinion of their work. In 1972, in an influential consultancy report for the Ford Foundation on the state of the social sciences in Indonesia, Geertz characterized Indonesian intellectual life as ‘centralised, over-organised, spasmodic, practical, and strongly influenced by economists’, and social science training as ‘bookish, speculative, and … philosophical or even doctrinal’ (Geertz, 1974: 365, 369). Those who have worked in Indonesia during this period will recognize the truth of much of Geertz's assessment. On the other hand, he seems to have retained from his Yogyakarta days twenty years previously the ‘pith-helmet’ view of Indonesian academics avoiding ‘real’ field research, in favour of ‘a brief “study trip” in search of written records, or an interview or two, a generalized summary of the accessible “literature” on the subject, or a fish-net type of fact-gathering survey’ (ibid.: 369). This sweeping dismissal did no justice to the pioneering village-level research of many Indonesian academics and their students during the 1950s and 1960s, including path-breaking efforts at what is now called ‘participatory action research’ (White, 2005: 113–20), which Geertz does not appear to have read, never mentioning them in his own work.

  • 19

    Geertz, letter to the author, 3 January 1985, referring to his (1984) response to critiques of Agricultural Involution.

  • 20

    See for example his response to critiques of The Theatre State (Geertz, 2003) and of his views on post-colonial and post-socialist states (Geertz, 2004).

  • 21

    A key mover in this trend was Eric Wolf (1923–99), Geertz's near-contemporary and another GI Bill beneficiary (Wolf, 1982, 1999; see Gledhill, 2005)

  • 22

    This problem peaks in the tortured, manneristic prose of Fred Inglis, Geertz's sycophantic intellectual biographer in the ‘Key Contemporary Thinkers’ series (Inglis, 2000).

  • 23

    As argued powerfully, and practised, by Eric Wolf (see note 22) and Michael Burawoy (1998; also Burawoy et al. 1991 and 2000).

Ben White is Professor of Rural Sociology at the Institute of Social Studies, PO Box 29776, 2502 LT The Hague, The Netherlands (e-mail: and Professor in Social Sciences, University of Amsterdam. His main research interests focus on agrarian change in Indonesia (where he has been involved in research since the early 1970s), and anthropology/sociology of childhood.