[1] Failure to recognize this is the weakness, I think, of much of the existing theoretical literature on the subject. Mary Douglas’ essay on “jokes” for instance starts out as an analysis of joking relations. The result is a brilliant reflection on the nature of humor, but, it seems to me, is of little use in understanding the nature of joking relations in the traditional anthropological sense of the term.

[2] Though cf. Stasch 2002.

[3] Again, I remind the reader that I am using the term “joking” here in a special, technical sense, meaning “along the lines of the sort of construction of human relations typical of joking relations”; hence I do not simply mean “humorous.”

[4] “Sacred” implies “not to be touched” in most European languages as well—a fact which Durkheim made much of—though I do not know how widespread this is elsewhere.

[5] Tikopians for instance identify a man with a canoe by a term Raymond Firth translates as “linked,” the same term that is used for, say, bond-friends (1965: 257–8).

[6] Claude Levi-Strauss (1962) has made the point that totemic systems are not really about identity but analogy, that is, they are not saying clan X are like bears or clan Y like eagles, but that the relation between clan X and clan Y is like the relation between bears and eagles. This is, of course, a very famous argument. However, in a later work (1966), he also noted that such totemic systems usually develop between groups that all share a roughly equal status; and makes the intriguing suggestion that, when one begins to hear that clan X really do resemble bears, it is usually because some element of hierarchy has entered in. If nothing else, this certainly seems to work for the Lau Islands.

[7] And the Maori seem to have been typical of Polynesian societies in this respect.

[8] Again, when I say that joking behavior never seems to accompany gift-giving, I do not mean to suggest that it never accompanies exchange. It certainly does. The most obvious example is in some very common forms of barter; another, somewhat more obscure, can be found in certain forms of inter-village exchange said to be practiced by the Yanomami of Venezuela (Chagnon 1968): one group enters the village of the other making every sort of mock-threat—threats which the latter are expected to ignore with casual aplomb—and then, begins demanding items of property—demands which the latter cannot refuse. Their demands are only limited by their knowledge that their victims will later have the right to come to their village and do the same. The interesting thing here is that we are dealing with a sort of mirror image of Mauss’ formula, not the reciprocal giving, but instead the reciprocal taking of goods. That it should be accompanied by behavior that smacks of joking then should hardly be surprising.

[9] When Shakespeare’s Henry V refers to France as another jewel for his crown, he is expressing perfectly the equivalence of ornaments or insignia and what we like to call “real property,” in terms of signification. Though on this, see also Graeber 1996.

[10] Which often accompany what Marshall Sahlins (1972) has called “generalized reciprocity.”

[11] True, different systems lean more or less heavily to one side or another. The Indian caste system, certainly, presses down very hard on the linear side; the Nuer segmentary system, to take a famous example, lean with equal weight in the opposite direction. But I doubt one can find any society based entirely on one principle and not the other.

[12] “Ownership” in this sense generally had little to do with any kind of rights and duties.

[13] In linguistic terminology, one would say the higher up he is, the more he is an unmarked term: standing for not only “man,” but “household,” “clan,” “tribe,” and so on. This does fit quite nicely with my observations about avoidance and universalism (moving upwards on the taxonomic hierarchy). But it makes tapu a somewhat paradoxical process: the marking of the unmarked.

[14] One could go on from here to speak of legal notions, which described peasants as being “owned by the land” as much as the other way around, or for that matter the etymologies of words still in common use today: the Oxford English Dictionary for instance, has it that the English word “clown” is derived from an Germanic root meaning both “peasant” and “lump of earth” (“clod” has the same derivation).

[15] Burke (1978: 199–204) notes that the metaphor of “letting off steam” began to be employed the moment it was technically possible; before that, the preferred metaphor was letting off pressure in a wine cask. Even at the time, though, many objected that, as safety-valves go, popular festivals made extraordinarily poor ones, considering how many genuine rebellions grew out of them (see Bercé 1976; Burke op cit.; Davis 1980). Bercé provides vivid accounts of preparations for carnival in French cities during the sixteenth and seventeenth centuries, during which the soldiers manning the city walls would systematically turn the cannons on the parapets around so they would face into the town, in case of any serious trouble.

[16] Anyway it strikes me that it can be more potentially revealing, for the analysis of rituals such as Carnival, than, say, Victor Turner’s notion of liminality and communitas (1969)—terms often thrown around so very casually that their use can stifle further discussion more than encourage it.

[17] Each entails its own characteristic notion of exchange: an abusive (or mock-abusive) exchange of substances in one, a benevolent (or mock-benevolent) exchange of properties in the other.

[18] Even before medical science was able to produce arguments of “personal hygiene,” Erasmus was warning children to restrain their manners even in private, because angels could be watching one unawares.

[19] Elias’ idea of the “civilizing process” is pretty unabashedly evolutionist and has been widely criticized as such. Many have also pointed to Elias’ undue attention to courtly circles and his neglect of Puritan and other middle class ideas as a crucial flaw in his analysis.

[20] There are parallel cases which don’t involve a breach of avoidance but other kinds of bodily contact considered too intimate for the relation in which it occurred. In the New Hebrides: “Sodomy between two genealogically related men is regarded as incestuous. However it is not viewed too seriously, as the punishment inflicted is that both parties must kill and exchange two pigs” (Corlette 1935: 486).

[21] For Manus parallels, see Mead 1934: 191, 308.

[22] Obviously, it was unusual for any two individuals to be exactly equivalent in worth at any given time, but they were inherently capable of being so.

[23] Overton clearly did not mean to include women; or for that matter servants. There is some debate as to whether the Levelers even meant to give wage laborers the franchise.

[24] Elias himself notes (1978: 42–50) how thoroughly embedded these ideas had become in the common sense of the middle classes most dedicated to the reform of manners.

[25] The literate class and the courteous class tended always to be one and the same.

[26] It’s not so much that “apprenticeship and service were confused” as Aries puts it (ibid.: 366–367) than that they were never really distinguished to begin with.

[27] It would be interesting to examine the institution of Medieval and Early Modern service in the light of the anthropological literature on initiation, particularly the kind which involves “fictive kinship” of one sort or another. The study of compadrazgo in Latin America provides some obvious parallels: while authors such as Wolf (1966) highlight the way such ties create ties of patronage across class lines, symbolic analyzes (e.g., Gudeman 1971; Bloch and Guggenheim 1981) stress the division between the female domestic, and male public domains—which in Western culture have been generally presented in terms of the spirit and the flesh. I’ve already mentioned that, in Europe, most youths served masters of a marginally higher social class. As for the symbolic aspects, Aries notes that the age of “seven or nine”—the age at which the Italian author of the above-cited account of English habits claims most families sent off their children to the houses of strangers—was “in the old French authors...given as the age when the boys leave the care of the womenfolk to go to school or enter the adult world” (op cit.). The opposition of spirit and flesh—or anyway, something very much like it—was also at play in the very definition of “youth” itself.

[28] An obvious parallel is the career military officer who is never obliged to stand as stiffly or salute as smartly as recruits have to do to him, but is still seen as reflecting in his ordinary bearing a more “military” comportment than they.

[29] I note in passing that the notion of reforming the lower strata was a bit difficult to reconcile with Calvinist doctrine, which encouraged most heads of household to, at least, the strong suspicion that their charges were predestined from the start to go to hell (cf. Hill 1964). But this merely underlines how much the project itself—of defining a social class in terms of a stage in the life-cycle—was inherently contradictory.

[30] Any more than capitalism, about which very similar arguments are often made.

[31] He also seems to assume that all holistic systems must be hierarchical; but this is another issue.

[32] I have elaborated this argument in an earlier work (Graeber 2004: 24–37).

[33] To take one example, a little while ago a book came out called The Consumer Society Reader (Schor & Holt 2000), which contains essays by twenty eight authors ranging from Thorsten Veblen to Tom Frank about consumption and consumerism. Not a single essay offers a definition of either term, or asks why these terms are being used rather than others.

[34] Especially if the band had not yet received a record contract or many professional gigs. If they were able to market some kind of product, it might be considered production again.

[35] Here, I also want to answer some of the questions rather left dangling at the end of my book on value theory (Graeber 2001: 257–261).

[36] In French the word consummation, which is from a different root, eventually displaced consumption. But the idea of taking possession of an object seems to remain; and any number of authors have remarked on the implied parallel between sexual appropriation and eating food.

[37] “Produce” is derived from a Latin word meaning to “bring out” (a usage still preserved in phrases like “the defense produced a witness...” or “he produced a flashlight from under his cloak”) or “to put out” (as from a factory).

[38] Bataille’s argument was that production, which Marx saw as quintessentially human, is also the domain of activity most constrained by practical considerations; consumption, the least so. To discover what is really important to a culture, therefore, one should look not at how they make things, but how they destroy them.

[39] Similar lists appear throughout the Western tradition. Kant also had three—wealth, power, and prestige—interestingly, skipping pleasure.

[40] The sensual pleasures they had in mind seem to have centered as much on having sex as on eating food, on lounging on silk pillows as burning incense or hashish; and by “wealth” both seemed to have in mind, first and foremost, permanent things like mansions, landed estates, and magnificent jewelry than consumables.

[41] One could even argue that Smith’s approach to questions of desire and fulfillment is so one-sided, centering almost entirely on social recognition and immaterial rewards (wealth, in his system, was only really desirable insofar as wealthy people were more likely to be the object of others’ attention and spontaneous sympathetic concern) that it is meant to head off the very possibility of the consumption model that was to develop from his economic work.

[42] Working here on the assumption that, if one examines any intellectual tradition carefully enough, one could find the materials for a genuinely insightful analysis of such “big questions”—i.e., sufficient perusal of the Buddhist tradition would also have yielded useful results, had I been competent to do it, which I’m not.

[43] For the best collection of essays on Spinoza’s theory of desire, see Yovel 1999. On his theory of imagination, see Gates & Lloyd 1999.

[44] I am especially drawing on the famous “strong reading” of this passage by Alexander Kojéve (1969) that had such an influence on Sartre, and through him, de Beauvoir, Fanon, etc.

[45] In Hegel’s language, they construct themselves as a negation, therefore they seek to negate that negation by negating something else—i.e., by eating it.

[46] Lacan’s “mirror phase” itself actually draws directly on Hegel’s master-slave dialectic (Casey & Woody 1983). I might note too that it’s the Hegel-Kojéve-Sartre connection which is responsible for the habit of writing about “the Other” with a capital “O,” as an inherently unknowable creature.

[47] It would appear that much of Couliano’s work draws on Agamben for inspiration, though Couliano only cites Agamben occasionally, and always to attack him on minor points.

[48] “That is the incapacity of conceiving the incorporeal and the desire to make of it the object of an embrace are two faces of the same coin, of the process in whose course the traditional contemplative vocation of the melancholic reveals itself vulnerable to the violent disturbance of desire menacing it from within” (1993a: 18).

[49] There is a lot of evidence which suggests that levels of clinical depression do in fact rise sharply in consumer-oriented societies. They have certainly been rising steadily in the US for most of the century. I should emphasize, by the way, that while Agamben and Couliano draw exclusively on European sources, these ideas were very likely developed earlier and more extensively in the Islamic world, and what they could in the European sources are ultimately derivative from them. Unfortunately little of this work has been translated, nor the history of Arabic and Persian theories of the imagination discussed in contemporary work in European languages. But I would underline that this is yet another way in which when one refers to the “Western tradition,” one should think of oneself, especially in this period, referring equally or even primarily to Islam.

[50] These images were seen to act on the imagination in ways already developed by the contemporary Art of Memory: see Yates 1964.

[51] Almost always, this also ends up involving a certain degree of fetishization, where the objects end up appearing, from the actors’ perspective, to be the source of the very powers by which they are in fact created; because, from the actors’ position, this might as well be true. Often, too, these objects become imaginary micro-totalities which play a similar role to Lacan’s mirror-objects or similar critiques of the commodity as capturing an illusory sense of wholeness in a society fragmented by capitalism itself (Graeber 1996a; Debord 1994).

[52] Even women, when they wrote love poems, tended to adopt a male point of view.

[53] In other words, rather than asking how is it possible to truly “have” or possess some object or experience, perhaps we should be asking why anyone should develop a desire to do so to begin with.

[54] Supposedly, in early Roman law the paterfamilias did have the power to execute his children, as well as his slaves; both rights, if they really did exist in practice, were stripped away extremely early.

[55] Or more technically, I suppose, synecdoche.

[56] And it has the additional attraction of being almost the only power which kings do not have over their subjects: as one sixteenth century Spanish jurist wrote, in arguing that American cannibalism violated natural law, “no man may possess another so absolutely that he may make use of him as a foodstuff” (in Pagden 1984: 86).

[57] As noted above, much of the Medieval philosophy Agamben and others discuss was probably first developed in the Arabic and Persian literatures and only later adopted in Europe.

[58] As, incidentally, do those people from other cultures who radically reinterpret TV shows, so much beloved of anthropological media theorists (e.g., Graburn 1982).

[59] In the last chapter of Towards an Anthropological Theory of Value (Graeber 2001), subtitled “The Problem of the Fetish, IIIb.” What follows was, in large part, originally written for that chapter but ended up having to be cut for reasons of space. I was tempted to call it “The Problem of the Fetish IIIc,” but decided the joke was too obscure even for my tastes.

[60] Especially in Italy. The most familiar representative for most readers in the Anglophone world is Toni Negri, but most of the ideas presented in Empire are the products of a long tradition involving many other writers and activists.

[61] For Castoriadis, history is no longer a matter of the development or play of productive or class forces, but the work of the “the imaginary, which is creation ex nihilo,” such that change is “the positing of a new type of behavior...the institution of a new social rule, ... the invention of a new object or a new form” that is “an emergence or a production which cannot be deduced on the basis of a previous situation” (Castoriadis 1987: 3, 44).

[62] The tie to the Autonomist school can be seen by looking at the early work of Toni Negri, on constituent power (1999). Essentially, he’s trying to work out exactly the same problems: what is that popular power of creativity that emerges during moments of revolution and how would it possible to institutionalize it?

[63] Actually there’s no particular reason why gold should be a better medium of exchange than beads. Economists of course might make the argument that the supply of gold in the world is inherently limited, while glass beads can be manufactured in endless number. However, there is no way that European merchants of that day could have had the slightest idea how much of the earth’s crust was composed of gold; they saw it as precious because it was got with difficulty from far away, just as Africans did beads.

[64] At least, he does not in the first three, best-known articles (1985, 1987, 1988). He does address West African ideas in two later articles concerned with debt and human sacrifice (1995, 1997). These essays, however, are concerned with a later historical period, and somewhat different sorts of questions.

[65] This would be one reason why Africans have been, from such an early period, comparatively receptive to religions like Christianity and Islam.

[66] Most African cosmologies posit the creator as in one way or another beyond good and evil, as, for instance, an otiose creator who has abandoned the world, or a force of violence beyond all moral accounting whose very arbitrariness demonstrates his local priority to, and hence ability to constitute, any system of human justice.

[67] More precisely, symptoms.

[68] Bohannan interprets these movements as regular features of Tiv social structure. More recently, Nigerian scholars (Tseayo 1975; Makar 1994) have placed them in the colonial context, as a result of British efforts to force a highly egalitarian group into the framework of a state based on indirect rule. In fact, there’s no real way to know whether such movements did occur earlier, but it seems reasonable to assume some such mechanism existed, at least, for as long as Tiv egalitarianism itself did.

[69] MacGaffey (1986) suggests the archetypal BaKongo ritual cycle leads from affliction to sacrifice to retreat to receiving gifts to new status.

[70] The text is as in MacGaffey except I have translated nganga (“curer, keeper of a nkisi”) and its plural (banganga) into English.

[71] Personal communication, March 2000. Just as in Hobbes, this creates some overarching power of violence which can ensure people fulfill their contractual obligations and respect one another’s property rights—which, if we look again at Pietz’s material, becomes especially ironic. Here we have European merchant adventurers swearing oaths and making agreements with Africans over objects they called “fetishes,” at exactly the same time authors like Hobbes were inventing social contract theory back home. But it was apparently the Africans who saw the act as creating a sort of social contract; the Europeans seem to have had other fish to fry.
All this obviously raises the question of whether there is any reason to believe that Hobbes, among others, were aware of what was going on in Africa at the time. In Hobbes’ case at least, I have managed to find no concrete evidence. While Hobbes grew up in a merchant household, in his entire published corpus his only mentions of Africa, as far as I am aware, are via Classical references.

[72] Clearly, what I am suggesting here could be considered a variant of the famous “wealth in people” argument (see for instance Guyer 1993, Guyer & Belinga 1995).

[73] Obviously, this is a bit of a simplification.

[74] According to Bohannan & Bohannan (1968: 233), having a “strong heart” means you have “both courage and attractiveness.”

[75] This is by no means unique to Madagascar. In the BaKongo case, too, royal power was seen as created through the same means as fetishes.

[76] In fact, the word “fetish” derives from a Portuguese term meaning “something made,” or even “artificial.” This is why the term was also used for cosmetics (“make-up”) (Baudrillard 1972: 91). Baudrillard’s conclusion—that fetishes do not make some arbitrary ideology seem natural, but instead inspire a kind of fascination with its very artificiality—while wildly overstated, it seems to me, still has something profoundly insightful about it, and might be compared with my own conclusions at the end of this article.

[77] The phrase is adopted from Frederick Jameson. Jameson’s notion of “cultural revolution” (1981: 95–97), in turn, goes back to a certain strain of Althusserian Marxism: the idea is that as one ruling class is in the gradual process of replacing another, the conflict between them can become a crisis of meaning, as, radically different “conceptualities, habits and life forms, and value systems” exist alongside one another. The Enlightenment, for example, could be seen as one dramatic moment in a long cultural revolution in which those of the old feudal aristocracy were “systematically dismantled” and replaced with those of an emerging bourgeoisie. In the case of the West African coast one is, of course, speaking not of one class replacing another but a confrontation of different cultural worlds.

[78] Mauss for example advised his students that the term “fetish” was useless as a theoretical term and should be eliminated.

[79] This is why, as I’ve suggested (2001: 239–47), Marxists have such a difficulty figuring out what to think about magic.

[80] That is to say there was nothing like the fixed, mythological pantheon one finds among the Greeks, or Babylonians, or Yoruba, where objects of cult could be identified with some enduring figure like Athena, Marduk, or Shango.

[81] As Terry Turner and others have argued at some length (see Graeber 2001: 64–66), all this is pretty much exactly what Piaget was talking about when he described childish “egocentrism”: the inability to understand that one’s own perspective on a situation is not identical to reality itself, but just one of an endless variety of possible perspectives, which in childhood too leads to treating objects as if they had subjective qualities.

[82] Even this is somewhat deceptive language because it implies the production of people and social relations is not itself “material.” In fact, I’ve argued elsewhere (see chapter 3) that the very distinction between “material infrastructure” and “ideological superstructure” is itself a form of idealism.

[83] In point of fact, if one does, this can lead to fetishism of a different sort, as in the sort one sees in heirloom valuables in many gift systems, which are seen as embodying or including the personalities of certain former owners.

[84] From a Marxian perspective it might be rather disturbing to see business deals as a prototype for revolutionary activity; but one must bear in mind it comes with the argument that the prototypical form of contract, even between business partners, is communism.

[85] One might contrast the situation here with what obtains in, say, much of rural Brazil, where the situation is quite the opposite, since police, effectively, are only interested in enforcing property rights, and can be expected to ignore mere cases of murder—unless, that is, the victim is a member of the property-owning elite.

[86] The gendarmes’ occasional zeal in pursuing bandits probably did have something to do with a perception that they were the only organized, armed group that had the capacity to form the nucleus of a rebellion—unlikely though that might have been. There had been times, mainly in the nineteenth century, when bandits actually had turned into rebels. But I suspect the concern was rooted in deeper understandings about what a state was all about: under the Merina Kingdom, bandits (referred to in official documents simply as fahavalo, “the enemy”) were, along with witches, the archetypal anti-state, that which legitimate royal authority defined itself against. The connection with witches also helps explain the otherwise puzzling fact that, much though they were unconcerned with Henri’s depredations, Arivonimamo’s gendarmes did leap into action to arrest and interrogate a teenage girl suspected of being behind an outbreak of Ambalavelona, or posesssion by evil ghosts, which affected a whole dormful of students at the state high school in 1979.

[87] Medical services for instance were in theory provided free, but had been effectively privatized by corruption, which, in turn, became universal once government salaries declined to next to nothing.

[88] As Jacques Dez (1975: 54–57) notes in a generally excellent summary; though in the end, he reproduces colonial assumptions by concluding that “the” fokon’olona was “invented” by the late-eighteenth century king Andrianampoinimerina. On the underlying ethos of consensus decision-making see Andriamanjato (1957).

[89] In Europe or North America, this is more true of men than women; in Madagascar it was, if anything, the other way around.

[90] Contemporary archaeologists now believe that significant human settlement in Madagascar was surprisingly late: perhaps from the eighth century CE, and at first seem to have consisted of heterogeneous populations probably of very different origins, Austronesian, African, and perhaps others. During this early period there was even a small Islamic city, Mahilaka, almost certainly Swahili-speaking, engaged in lively trade with East Africa and the Arabian peninsula. Early Malagasy thus had experience of states and world religions from the very beginning; and the moment of “synthesis,” when contemporary Malagasy culture appears to have born, seems to have occurred around the time of the height or perhaps even downfall of Mahilaka. After this, however, it proved surprisingly persistent throughout the island and capable of resisting frequent Islamic attempts to convert and incorporate the island’s population. I strongly suspect that insofar as Malagasy culture emerged as a coherent entity, it was in conscious contrast to everything that was considered “Silamo”—Swahili, Islamic—just as it is maintained in conscious contrast to everything that is “Vazaha” today.

[91] Dia vita ohatran’ny androany antoandro izao ny fonosan-damba, dia injany fa nirehitra ny tanana... Dia may izany. Dia maraina dia iny niavy tamin’ny olona indray hoe ho taperiko mihitsy aza ny ainareo raha ohatra ka tsy mamono lamba fa avelao izahy fonosona... Dia novonona indray ireo fasana ireo dia fonosina indray.

[92] Though it’s only fair to point out the fact that all of these men had been alive at the time, or at least had first heard the story from eyewitnesses. All described the fire in naturalistic terms and denied a famadihana had anything to do with it.

[93] Mamadika is a verb meaning “to turn over,” “to reverse,” or “to betray”; famadihana its nominalization.

[94] For most of the nineteenth century, it was apparently the actual construction of new tombs which was the real focus of mortuary ritual, as it still is in Betsileo according to Kottack (1980: 229). A complete history has not yet been written, but at least in western Imerina, the modern pattern began to take form around the mid-1880s.

[95] Bloch found this to be true as well (1971: 157–8). My sources, by the way, are not only based on direct observation of the ritual season of 1990, but on documents preserved in the offices of the firaisam-pokontany (former canton) covering the years 1985–90, checked against people’s recollections. This latter proved a good way of determining which famadihana people really considered significant or memorable.
Some famadihana did concentrate on a particular tomb rather than on a particular individual, especially if a number of important people belonging to one tomb had recently died. But these were much less common than those dedicated to individuals. The transfer of bodies from temporary graves was, in my own experience, never celebrated as an event in itself; it was mainly children who were buried this way, and their parents generally moved them when their tomb was opened for someone else’s famadihana.

[96] This account is in a number of ways different from Bloch’s (1971: 145–161), which is based on what he observed in the Avaradrano region in the late-1960s. This may be partly due to regional variation, partly a reflection of historical change. Most of these differences however are relatively minor.

[97] On the other hand, in the funerals I attended, people dressed in their most formal and expensive church clothes. The contrast, when a rich zanaka ampielezina is buried in the countryside was in fact quite striking: the village is suddenly full of expensive cars, men in black suits and ties, women in white dresses and gold jewelry carrying elegant parasols. The resentment of the country folk was often palpable.

[98] Everyone agreed the body should properly be carried around seven times, but that this is no longer done. All such details depend on the astrologer’s decision. He may forgo the rounding of the tomb completely and have the bodies carried out directly to the laps of the women.

[99] Famadihana around Arivonimamo, on the other hand, appear relatively staid: I never saw anyone tossing skulls in the air, snatching skeletons from each other, and so on, as others have reported (eg, Ruud 1960: 169).

[100] The next day was one of feasting: pigs slaughtered the day before were cooked in huge vats and ladled out to all; there was music and dancing and almost inevitably, drunken quarrels between rural and urban kin, which could, if the sponsors did not effectively intervene, degenerate into brawls. But the celebrations were only considered part of the famadihana in the broadest sense.

[101] My own experience indicates this is something of a misnomer. Rural people did not even recognize this usage. In fact, there was there was no generic term for “deme” in common use at all. In the nineteenth-century demes were most often referred to as firenena, which is now the term for “nation.”

[102] The andriana of Betafo are descended from military colonists placed there after the Merina kingdom’s conquest of Imamo around 1800. The colonists were from a famous andriana group called the Zanak’ (“children of”) Andrianamboninolona. Not only do Betafo nobles regularly refer to themselves as children of Andrianamboninolona, most think it is he who is buried in the razambe’s tomb.

[103] The descendants of this younger son still predominate in that portion of the deme’s territory. The people of this division, who became Catholics when their eastern kin converted to Protestantism in the last century, remember the deme history quite differently. Many claim at least half of the razambe’s sons for their own division, identifying them with each of the division’s oldest tombs. This kind of contestation is more the rule than the exception, though.

[104] Hence, the internal structure of a tomb is much like that of a deme: there is a single razambe embodying the unity of the group, with a set of ranked children who, in so far as they are remembered, can be appealed to make distinctions between segments.

[105] Women tend to have more options because they can always choose to be buried in their husband’s tombs (or often one of several husbands’ tombs), while only occasionally are husbands buried in their wives’ family tombs. For statistics on actual choices see Bloch 1971: 115; Razafintsalama 1981: 190–200; Vogel 1982: 162).

[106] All the shelves of a new tomb should properly hold at least one body, since, if one is left empty, the spirits of the dead were likely to carry off a child or other family member in order to fill it. So if human bodies were not available, the trunk of a banana tree was usually placed on the empty shelf to substitute for a human being.

[107] Nor were they necessarily ancestors, in the technical sense. All of the bodies in a tomb were razana, whether they had descendants or not.

[108] No famadihana seems complete without at least one argument between the men removing the bodies—who are supposed to carry it out head first—over which end of the razana is its head. Admittedly the bearers are never completely sober (if they had been, they would probably have remembered the way it was facing when the first picked it up) but rarely were there any physical clues to help them.

[109] Some even argue that bodies are combined because it would be too expensive after a while to wrap all of them (and it would be shameful to open a tomb and leave some of them unwrapped). True, infertile razana are rarely given silk but usually polyester; but, if combined with the body of an ancestor who does have descendants, there’s no expense at all.

[110] Such tombs are said to be “full,” though in fact they’re more likely to be largely empty.

[111] Examination, however, usually revealed that these lists represented only a tiny proportion even of those ancestors involved in the ceremony—for each tomb, just one or two razambe and those who had died in the last decade.

[112] A Maori chief, for example, could be said to be intrinsically tapu, meaning sacred in the sense of set apart from the rest of the world. The word fady however is never applied to persons in this way. It is applied primarily to actions. Even when one speaks of, say, a “onion fady” this is usually shorthand for some specific rule of action, like eating or growing one.

[113] I think this is one reason why I found it impossible to come up with anything remotely resembling a coherent list of local fady for the community of Betafo. Everyone agreed that such a list could be written, but no two gave anything like the same account, and many indignantly denied what their neighbors had told me. Spheres of influence were constantly being marked out by who could convince others to accept their view of the local fady, and, Betafo being a place in which authority and group solidarity were in a constant state of flux, opinions about fady tended to being equally shifting and chaotic.

[114] They most often involved pigs and different kinds of onions—particularly garlic, which is called tongolo gasy or “Malagasy onion.” Some of the more erudite held that pigs and onions, being “dirty,” negated the power of magical charms and annoyed the spirits of the dead, and were, for this reason, a frequent subject of fady for users of magic and those who enter sacred places. But it was very rare for taboos to involve such an explicit notion of pollution. I note in passing that typically, restrictions on pork or onions applied only to specific situations—only once or twice did I run into someone never allowed to eat pork or garlic at all. They are, after all, probably the most popular foodstuffs in Imerina, and this would appear to be one of the reasons they were so often the focus of taboo.

[115] Actually, he cursed them not to eat bokana, a variety of caterpillar used in local silk production as well as occasionally as food. Be this as it may, everyone I talked to found the restriction highly amusing and rarely avoided an opportunity to remark on it. The death by gluttony motif is in fact probably the most popular story used to explain group fady (and was a theme most found intrinsically funny in any context). Stories about the origins of marriage restrictions usually traced them back to some incident where the ancestors gambled, cheated, and got mad at each other—as with the gluttony stories, most of the people who told them to me made it clear that, as far as they were concerned, the ancestors were acting like fools.

[116] A few very old and venerable women would try to put a moral slant on this: e.g., the ancestors are merciless in the punishing of evil-doers; most did not.

[117] Parental ozona and ancestral ozona were seen by at least some to depend on one another: one woman told me you should be careful to observe all the ancestral fady lest you lose the ability to curse your own children.

[118] In Imerina, in fact, I could find no popular interest in a cosmological time of origins at all; tany gasy, or “Malagasy times,” which is the time of historical origins in which the ancestors lived and demes were founded, is seen as differing from the present mainly in a political sense.

[119] While most of the demes and deme-territories in the area around Arivonimamo do not seem to have changed in any dramatic way since the last century, archival documents reveal the existence, in Ambohibe (a town near Ambohibeloma, seven or eight kilometers north of Arivonimamo) of an enormously rich man with the appropriate name of Andriampenovola—“lord full of money”—who, throughout the 1880s and 1890s seems to have gone about accumulating descendants through adoptions: each adoptee was guaranteed a portion of his land, slaves, and other property as long as he or she remain on the ancestral territory. When I passed through Ambohibe in 1990 no one remembered the name of the 19^th century deme (the Zanak’Andriandoria) but instead gave Andriampenovola as the name of the local razambe.

[120] While there is again no generic term in Malagasy for such groups, people usually would refer to particular local families after their founders; as, e.g., “the offspring of Ranaivo” (ny terad-Ranaivo). Vogel (1982) for this reason calls such groups teraka, or “offsprings,” but the term would never be so used by a native speaker.

[121] By this I mean that, while there may be other descendants still using the tomb, there are none that live nearby. Often, though, the head would feel responsible for the upkeep of a whole set of tombs he had links to, if no other descendants were to be found who were capable of keeping them up in a respectable fashion.
Otherwise, it was largely left to the heads of the fragmentary families I’ve described to keep up local relations to the less famous tombs in any given area that were still in use. This gave them much of their local social importance, since the other owners tended to be city people, migrants, or children of migrants who depended on them to mediate in dealings with the tomb.

[122] When women talked about leaving their husband they always, I noticed, spoke of “going home to father” and never “to mother.”

[123] Efa maty daholo ny efa lehibe, fa izahay zaza mpandimby fotsiny no sisa.

[124] For example Ramose, technically the male equivalent of Madama, was in practice only used for men of that age who are also schoolteachers; Rangahy, the male parallel for both Ramatoa and Ramatoabe, is much more informal and in practice used much like the English word “guy”; finally, the term Ingahibe is a term of great respect applied only to the one or two oldest men in a given community. In all of this, by the way, I am only speaking of the vernacular Merina I am familiar with from Arivonimamo: I can’t say for sure how far these generalizations hold beyond it.

[125] I only saw women crying and male ancestors being cried over, but I only witnessed four or five incidents first-hand.

[126] Izaho izao ohatra tamin’ny 1989, nanofy izany izaho eto hoe hitako i dadanay—izy izany efa maty io—ary Avaratr’Ambodivona ary—fa misy hazo eo, dia niresaka aminy izahay fa ity dada ity ve mbola tsy maty hoy izy izany; mbola miseho eto indray. Dia omeo tsodrano aho hoy aho fa izaho tsy salama... Dia niresaka eo izahay mianaka: tsy fanao izay Irina hoy izy, dia iny izy dia nidaboka maty tamin’izy nafatotra iny. Dia izaho niakatra tamin’ny tanana misy an’ilay zokinay lahimatoa hafareny tery. Dia izy koa mba nikisaka niala an-tanana izy izany nidaboka an’iny fahafatesan’iny, Ohatran’ny hoe: mahatsiravana mampahatahotra.

[127] Her dream is somewhat complicated, though, by the fact that she had it at a time when her siblings were all quarreling—which probably explains his sudden transformation from benevolent to stern and authoritarian. It’s unclear whether Irina meant to imply her illness was caused by her father’s disapproval or not.

[128] In Betafo, for instance, I heard of the case of several absentee owners living in the capital who, on converting to an Evangelical sect that did not allow them to participate in famadihana, immediately sold off their rice fields in Betafo.

[129] No post-independence Malagasy government has to my knowledge ever erected a statue in the European sense—that is, one bearing some kind of likeness. Public monuments always take the form of standing stones.

[130] About clothing: one elderly man made a great point of this, in speaking of his father, who he resented for not having taken care of him as a child. “He never so much as clothed me, but even so, I clothe him now” (meaning at famadihana). About the lamba: as Gillian Feeley-Harnik reminds us (1989) these are feminine products; ideally, they should be the handiwork of the participants themselves.

[131] Though, admittedly, it’s really just a homonym, and only works in the active voice (mamono), since the two verbs actually come from different roots (fono for wrapping, vono for beating/killing).

[132] Irina’s request for a tsodrano in the dream cited above might be an example of the same thing; it’s unclear from the context whether her father’s annoyance with his squabbling children was the cause of her illness or not.

[133] Generally speaking, every ritual gesture which involved giving something to the ancestors—i.e., pouring rum over the door of the tomb or over the bodies inside, giving gifts when the ancestors are placed on women’s laps, and so on, are all referred to as “requests for tsodrano.” Likewise, anything taken away by the zanadrazana, such as the pieces of mats which are said to bring fertility to women and the tooth medicine mentioned above, can be called “tsodrano.”

[134] Also—though this was a matter of some debate among my own acquaintances—tsodrano could be given to relieve the consequences of tsiny, which is the guilt or blame a person may have due to the detrimental effects their actions have had on others (Andriamanjato 1957). Parents, for instance, might give an errant child who has returned such a tsodrano. This is, of course, in keeping with what I’ve said about people who leave bottles of honey and so forth on top of tombs. It also may relate to the notion that famadihana are meant to counteract tsiny, which was very important to Bloch’s informants though I never heard much about it where I worked.

[135] There’s no room here to enter into ambalavelona, which involves a sorcerer’s manipulation of material from tombs to cause an enemy to become possessed by such an evil ghost, typically driving the victim mad. Similar charms are used to cure ambalavelona, and to drive off Vazimba spirits, which are considered by some to be the final malevolent form which ancestors take when their descendants do not “take care of” them.

[136] I would like to thank Jennifer Cole, Jean Comaroff, Gillian Feeley-Harnik, Michael Lambek, Pier Larson, Nhu Thi Le, Stuart Rockefeller, Marshall Sahlins, Johanna Schoss, and Raymond T. Smith for all sorts of useful comments and suggestions. My fieldwork in Madagascar was funded by a Fullbright/IEE. I should note also that the language and people of Madagascar are referred to as Malagasy, the inhabitants of Imerina are called Merina, and that ody fitia is pronounced OOD fee-TEE.

[137] A Foucauldian approach, for instance, might emphasize how imported disciplines of education or hygiene transformed domestic relations; but this is not my project either. At least among the majority of people in Imerina, colonial disciplines really did not have that direct an impact on the kind of issues I am dealing with here.

[138] The richest source on nineteenth century ody is a catalog assembled by the Norwegian Lutheran missionary, Lars Vig, between 1875 and 1902 (Vig 1969), but there is a fairly abundant literature, including material by Malagasy authors (R.P. Callet 1908: 82–103) and European ones (Dahle 1886–88; Edmunds 1897; Renel 1915), as well as some material extending to the middle of the present century (Bernard-Thierry 1959; Ruud 1969). I have discussed the relation between spirit and object in much greater detail in an earlier work (Graeber 1996).

[139] My information is drawn mainly (though not exclusively) from women—where the nineteenth-century material was presumably drawn almost entirely from men—and this may be the cause of some distortion. But I did talk to dozens of men as well, and never found their basic understanding of ody fitia to be significantly different. If anything, women were more likely to point out that men could, occasionally, employ such medicine—if only because for them the fact was more a matter of immediate practical concern.

[140] Since all the accounts I draw on assume a female victim, when speaking in the abstract, it seems best to follow their usage. Still, the reader should note that Vig at least occasionally indicates that any of these charms could be used by women against men, and cites two charms said to have been used primarily by women. (Vig 1969: 94–97).

[141] Here, as elsewhere, my translation from the French.

[142] A shocking breach of decorum: adult brothers and sisters were never supposed to see each other naked.

[143] “The man could tell her ‘why are you proud towards me?... The custom of our ancestors applies to us all.’” “During the persecution of Christians under Ranavalona I,” he adds, “one accusation made against the Christians was that their wives were chaste...” (Vig 1969: 20). On the other hand, on the subject of sexual relations, Vig was apparently willing to believe almost any scandalous story his more whimsical parishioners could tell him. He interrupts descriptions of charms on two different occasions to relate that in previous times, old men would regularly gather at the village gates to watch young men fight battles over women, with the women going to those who prevailed; adding that in these times men were afraid even to bring their daughters with them to the weekly markets because “if she gave preference to only one, nine or ten rejected suitors would join forces and attack the father’s house the next night” (1969: 26, 88). All of this had of course completely disappeared, he added, with the advent of Christianity some twenty or thirty years before.

[144] Vig actually describes this ceremony as a cure for a condition known as kasoa, the one form of “amorous madness” whose effects he says can indeed lead to permanent insanity. People still talk about kasoa, but nowadays it is classed with ambalavelona as a form of aggressive witchcraft—both are ways of driving a victim insane by causing them to be possessed by an evil ghost.

[145] I might note that this is, in fact, the premise used in the Malay love magic from around the same period described by Skeat (1900: 566–580), which was also considered to be similar to sorcery for this very reason.

[146] The rhetoric here is particularly colorful:
Manara-mody, rao-dia, fehitratra: these are diseases that come with you when arrive here in town. That’s why so many young men die on returning home from traveling. And that’s why people say “life is the slave of wealth”: people know that the distant land is dangerous; but they have to get what others have got, though the pursuit of wealth has to be difficult. “If my eight bones aren’t broken! a road others have gone down, yet I can’t go? Other people’s children have all got rich, I perhaps am the child of an idiot? Do other people’s children know how to do something I don’t know how to do?” So he gets the money on loan, and when he goes trading, he’s bewitched; and when he arrives back, he dies; and his wife and children are enslaved because of the debt he owed to the people here in town (Callet 1908: 106).

[147] The condition, called ambalavelona, was often said to have been used by rejected suitors out of spite. It involved many of the symptoms Vig described—the great strength of the woman, the raving, the fits, the throwing off of clothes in public—but it was never thought of as inspiring love. It was merely a means of revenge.

[148] Other medicines referred to as “kinds of ody fitia” included tsy tia mainty (“to despise,” or literally “hate blackly”), which causes enmity to rise up between lovers or spouses, and manara-mody (“follows one home”), the ody used by coastal women to kill their Merina lovers and which apparently killed the man with three cows. Manara mody most considered a form of witchcraft pure and simple, and some denied it was a kind of ody fitia at all. In every story of its use I heard about, though, it was used in conjunction with other forms of ody fitia.

[149] Even women whose lovers abandoned them on learning they were pregnant would not publicly admit to having used fanainga lavitra to bring them back. The only people I found willing to (quietly) admit they had employed it was a married couple who had used it to recover a teenage daughter who had run away from home.

[150] Something along these lines appears to have happened throughout most of Madagascar during the colonial period; tromba cults, in which the spirits of ancient kings began to possess the living and demand ritual propitiation, brought royal service even to parts of the island which had never been ruled by kings at all. See Althabe (1969). Here too, fanompoana rendered to ancient kings became a principle by which people could assert their cultural autonomy in the face of colonial rule.

[151] I have written of the importance of taboo in my piece on famadihana (Graeber 1995), and in more detail about negative authority in chapter 3 of Lost People (2007).

[152] While the class of truly large-scale slave holders was always relatively small, perhaps only the poorest fifth of Merina households had no access to slave labor whatever. Most of the figures that follow are derived from documents preserved in the IIICC and EE sections of the Malagasy National Archives.

[153] Owners would usually accept a portion of their earnings and expect their attendance at certain critical moments, such as harvest, when labor was in particular demand. But it was often difficult to enforce even these requirements, and some masters appear to have been forced to pay wages to their own slaves. The situation was further complicated by the fact that by the end of the century partible inheritance had ensured that many slaves, perhaps most, had several different masters. For some contemporary accounts, see Sewell (1876), Cousins (1896), Piolet (1896).

[154] See the debate in the Anti-Slavery Reporter, February–March 1883.

[155] The transformation affected women as well as men; some observers note that many wealthy women had to learn to do manual labor for the first time in their lives after the liberation of their slaves (Pearse 1899: 263–64).

[156] I do not believe that, in all the time I was in Imerina, I ever heard the term used with anything but negative connotations. Apart from its political meaning, the only other phrase in which I heard it employed was the expression fanompoana sampy (“serving the idols”), adopted by the missionaries to translate the English “heathenism.” The expression is only used as a term of denigration; no one, no matter how nominal their Christianity, would ever apply it to themselves.

[157] This was, I should remark, much less true of the educated, urban elite than it was with rural people, white or black. Members of the former class would often speak quite casually about their “ancestors’ slaves” (andevon-drazana), clearly seeing their existence as a token of their former glory. Rural people, on the other hand, when they did discuss the matter openly, made it equally clear that they saw it as evidence of their ancestors’ misdeeds. Quite a number who claimed noble descent confided in me that they believed their own present-day poverty was a judgment rendered on their ancestors for having kept other Malagasy as slaves (Graeber 2007).

[158] Interestingly, so do the descendants of their former slaves, who were also considered more loyal to the colonial regime, and more amenable than other Merina not only to wage-labor, but to taking part in the hierarchically organized institutions identified with it. For instance, “black people” served in greatly disproportionate numbers in the military and police, as well as converting in large numbers to Catholicism. Significantly, too, I found them to be much more accepting of the use of kalo and even certain varieties of ody fitia than other Merina.

[159] Other sources on nineteenth-century witches include James Sibree (1880: 202), Bessie Graham (1883: 62–3), and, again, Vig (1969: 112–24).

[160] “So the brigands and thieves who present themselves to me to be catechized throw away their charms, being persuaded that they will lead them back to their careers as brigands without their being able to resist” Vig (1969: 123–24).

[161] Tsy mahazaka an’ilay herin’ilay fanafody. Unfortunately there is very little literature on witchcraft from the colonial period itself. The one main exception I know of is Mary Danielli’s “The Witches of Madagascar” (1947). Danielli’s information comes from exactly half way between my two periods, and offers what seems a unique synthesis between the two sets of ideas: there are ody fitia which simply cause love and devotion, Danielli’s informants told her, and these women do not become witches; but some love medicine has punitive effects, driving its victims mad or making them violently ill, and it is women who acquire this type of medicine who end up becoming possessed and “going out at night.” This seems to be a transitional moment. I never heard anyone say anything of the sort in 1989–1991.

[162] Maurice Bloch (1982, 1986). In addition to Elinor Ochs’ work, Pier Larson (1995) has contributed important insights into differences in male and female speech.

[163] Manjaka (“to rule,” nominalized as fanjakana, “government”) is in fact the reciprocal of manompo (to serve, nominalized as fanompoana). “To enthralled” is manandevo.

[164] Pretty much all major European languages have a term paralleling the English “to oppress.” A fairly superficial examination of dictionaries, and consultations with a few fluent or native speakers, and leaving out those languages using characters or diacritics too difficult to reproduce (such as say Thai or Arabic), adds Albanian (studjoj rëndshëm, shtyp), Basque (zapalketa), Biblical Hebrew (tahan, lit. “to grind down, to oppress”), Chinese (yà min), Coptic (tmtm, xa0x0), Finnish (ahdistaa), Ganda (zitoowererwa), Gurarani (jopy), Hawaiian (kaumaha, koikoi), Hittite (siyyaizzi, siyezzi, siyait), Japanese (osaetsukeru, yokuatsu-sur), Malay-Indonesian (tekan, mameras, tindas, tindih), Mongolian (darulal(ta)/daruldug-a), Nepali (thichnu), Nuer (mieet), Paiwan (q/m/ezetj), Persian (sarkoob, lit. “head pressed down”), Quechua (ñitiy), Sanskrit (avapídita), Shona (udzvinyiriri), Somali (cadaadid), Tamil (nerukku/nerukkam, and other constructions from the root neri, also Dravidian arepuni, arepini, areyuni, arevun, “to grind down or oppress”), Tswana (patikèga), Turkish (baski, ezmek), Tuscarora (turiye), Vietnamese (dé nång, su dàn áp), and Zulu (cindezela). The apparent exceptions are interesting in themselves: Native North American and Australian languages, for example, do not seem generally to have terms glossed “oppression” of any sort. Nor do most spoken by traditionally stateless peoples. African languages are a mix: in Africa words translated “oppression” in dictionaries appear about equally likely to come from terms for injustice or humiliation than “pressure downward.”

[165] I am, of course, hardly the first to discuss these dilemmas. For some analogous reflections from a feminist perspective, see Hodgson 1999 and Jackson 1995. Others have made similar points regarding postmodern forms of relativism: so, Maschia-Lees, Sharpe, and Cohen (1989: 27) cite Nancy Cott’s remark that a feminist approach, motivated by a political project to oppose the oppression of women, is difficult to maintain if one deconstructs the very category of “oppression”—or even “women.”

[166] This also raises the perhaps even more thorny problem of who “we” are, but I will leave this to be addressed, at least briefly, in the essay “There Never Was a West,” below.

[167] It follows that it might be possible to argue the Nuer lack any equivalent to our institutions of religion or the family, but it would not be possible to say they lack any institutional conception of authority, because otherwise, “the Nuer” would not exist.

[168] I note the role of gender in all this is ambiguous: while, as I say, in most matters of seniority between siblings, gender should not really weigh in at all, in reality it almost always does. In this case, elder sisters may well have their younger brothers carry things for them, but in formal occasions at least they would be unlikely to speak for them, at least unless they happened to be very good speakers, or very assertive, and no senior male were available.

[169] It’s a term, then, that could be used either for sending someone to be one’s spokesman, or to send someone to carry one’s things. In the nineteenth century, for instance, royal representatives were always referred to as the King’s iraka, here meaning “spokesmen,” who carried their words. Sometimes these were literally messengers, but the same term was used for those delegated to make decisions in the King’s name.
It was also the only real way in the language of the time in which people freely talked about relations of command, of ordering people around. The word baiko, which literally means “command,” existed at the time but mainly referred to military commands; since the latter were largely given in foreign languages, it meant “foreign words” by extension.

[170] As if to underline the point, Sibree continues the above-quoted passage by adding: “There is a great respect paid to seniority among the Malagasy; so that if two slaves who are brothers are going on a journey, any burden must be carried by the younger one, so far at least as his strength will allow” (ibid., 183). The obvious assumption is that, if two brothers who are not slaves go on a journey, there would be no question of either having to carry anything.

[171] An umbrella: an imported luxury, identified with Western styles of comportment, is the only exception.

[172] The notion of “emblematic labor” might be compared to Barth’s idea of ethnic “diacritics” (1969), where one or two apparently minor features can become the reference to distinguish otherwise overlapping or similar social groups. The situation in eighteenth-century Imerina rather recalls Hocart’s definition of caste (1968, 1970: 102–127; Quigley 1993), where each caste’s nature is determined by the labor they do for the king. The Merina system is sometimes described in fact as a “caste” system (see Bloch 1977).

[173] One group of former andriana, of somewhat ambiguous status, did have the special privilege of providing one silk shroud on such occasions. Another group of similar ambiguous status had the privilege of actually “carrying” the royal body to be placed in the tomb—the most exalted form of carrying, but still one not relegated to a group considered royal kin. These are the closest one has to exceptions.

[174] Oral traditions I gathered around Arivonimamo insisted the Andrianamboninolona, the andriana order ranked immediately above them, were famous as blacksmiths.

[175] One might hazard the following formulation: the production of objects and words are the domain of andriana; carrying and construction that of the hova; to the Mainty Enin-Dreny, in their capacity as royal warriors, is relegated the sphere of destruction.

[176] Sources sometimes substitute “digging red earth” (mihady tanimena), in an obvious allusion to the task of “digging red earth” for royal tombs, mentioned above.

[177] This follows the same order as the list given by Standing (1887: 358), though I left out Standing’s fifth category (building and maintaining roads and bridges) since it does not appear in any Malagasy-language account. For evocations of the standard list in nineteenth-century legal cases, see National Archives IIICC 365 f3: 111–112; IIICC37 f2 (Ambohitrimanjaka 1893). For standard lists of exemptions in the Tantara ny Andriana, a collection of Malagasy histories, see Callet 1908: 411 (Andriamamilaza), and 545 (Antehiroka). See also, entries in the Firaketana (an early twentieth-century Malagasy encyclopedia—Ravelojaona, Randzavola, Rajaona 1937) for Ambohibato, Ambohimalaza, Ambohimirimo, Andriana, and Antsahadinta.

[178] They were referred to in royal documents as alinjinera, or “engineers.”

[179] Traditionally these things are gendered: women carry objects on the head or hips; men on the back or shoulders.

[180] In fact, as I have argued at length elsewhere (Graeber 1995), these ceremonies ultimately have the effect of infantilizing the ancestors and treating them, in turn, like small children. I should also note that my discussion of mutual obligations of “carrying” owe most of their insight to discussions of the subject with Jennifer Cole, whose work with the Betsimisaraka people of Ambodiharina brought out these issues much more clearly than my own.

[181] Lambek’s book The Weight of the Past (2002) contains a detailed analysis of parallel idioms in a rather different social and political context among the Sakalava of Madagascar’s west coast.

[182] Not that the more familiar sort of symbolism was entirely absent (see Bloch 1986). A common expression was “the king is father to the people but the people are both father and mother to the king.”

[183] Domenichini argues that such groups had a ziva or “joking relation” with the crown. See Hebert 1958.

[184] In the royal case, even baggage being carried for the Queen in a sense participated in the Queen’s presence or anyway esteem. Royal carriers, even those carrying jars of water to the palace, were proceeded by a man bearing a spear warning all on the roads before it to make way, step to the side, and remove their hats as a gesture of respect just as they would if the Queen herself were passing.

[185] Literally they did not “remember themselves” (tsy mahatsiaro tena).

[186] This was true whether one was “carried” by dead kings, evil ghosts, or the power of one’s own magic—as were many women who became possessed by their love magic and ended up running around as witches during the night. Generally speaking, the term tsindriana was applied to forces that were essentially benevolent or at least neutral in nature; entina was used almost exclusively for forces that were intrinsically dangerous or malevolent in nature
The reluctance to speak of being “carried” by, say, ancestors or royal spirits seems to derive from a feeling (which I have described at length elsewhere) that to entirely efface or overwhelm the agency of another person, to replace it with one’s own, is a morally dubious way of exercising power.

[187] R: Dia avy hatrany, dia marary andoha tampoka ilay olona, dia very saina avy eo izy. Dia miteniteny foana, toa sahala amin’ny misy olona faharoa ao aminy.
Ka misy zavatra mampahatahotra ny marary. Voa manakenda azy. Voa mampijaly azy. Sahala amin’ny miady ambiby masiaka iray izy, sahala amin’ny bibilava iray. Arakaraky ny fiseho ilay fanahy ratsy, izay atao hoe, olona faharoa ao aminy.
C: Hitan’ny maso ve izany?
R: Hitan’ny masony izany. Hitan’ilay olona. Nohitany ilay bibilava. Niady amin’ireo heny, izay manimba azy, manakenda azy.

[188] R: Tsy ny tompon’ny tena intsony ilay olona voalohany, fa ny olona faharoa no manjaka.
DG: Fa ny olona faharoa dia...
R: Io no adaladala, io no miteniteny foana, io no mandrovitra akanjo...
DG: Fa tena misy olona faharoa sa misy, misy...
R: Fanahy ratsy.
DG: Fanahin’ny olona maty ve?
R: Fanahin’ny maty io, ka mampahatahotra azy. Miseho toy bibilava, miseho toy olona masiaka, miseho toy ny angatra...
C: Izay no mampatanjaka azy io?
R: Izay no mampatanjaka azy io—fa ankizivavy iray voan’ny Ambalavelona no manana ny herin’ny lehilahy dimy. Manana hery manokana.

[189] There is surprisingly little written about Zanadrano in the contemporary ethnographic literature on the highlands: nothing really in English, very little in French, and that largely about shrines and pilgrimage sites rather than ordinary curing practice: e.g., Cabanes 1972; Radimilahy, Andriamampianina, Blanchy, Rakotoarisoa & Razafimahazo 2006.

[190] Often there is a whole network of ody to be dealt with: the “mother ody” may be buried in the fields or yard, with various “children” planted around the house itself. And, often, also sisika—little bits of wood, bone, tooth, or what-have-you—buried in the patient herself, underneath the skin.

[191] One medium for instance would pray, gazing into a mirror placed beside a book and candle in his cabinet, waiting for the spirit to come over him. His wife explained that, as he stared, the face of the andriana would gradually replace his own. When his own features had been entirely effaced, he would be entirely possessed (tsindriana) and begin to speak. Similarly, in ambalavelona, victims often were terrified of mirrors, seeing monsters and snakes in them instead of their own image.
Several mediums were eager to hear my tape-recordings of their sessions, claiming they had never had an opportunity to hear what their spirits sounded like.

[192] Actually “holy” is not a very good translation for masina in most contexts but it will serve for present purposes. For the distinction of masina and mahery, see Bloch 1986a.

[193] Often, too, there is a final ceremony called the famoizina or faditra, in which some object representing the condition is finally cast away or buried, so that it cannot return.

[194] Actually, mediums tend to be reluctant to actually apply the term entina, “carried,” to any basically benevolent spirit; but the description is otherwise the same.

[195] And this is rather unlike better known forms of possession practiced elsewhere in Madagascar, such as tromba.

[196] And are in fact seen as such by the descendants of their former owners: see Graeber 2007.

[197] At least in public. Of course Raombana, the Queen’s personal secretary, expressed nothing but hatred for her in his elaborate history from which the earlier quote about the Ranavalona’s pleasure expeditions was actually taken. But his history was written in English so no one at court could read it. So it’s not as if such a position was unimaginable.

[198] Though here it is useful to consult Scott’s Domination and the Arts of Resistance (1992) on how often the cult of the king and denunciation of “evil councilors” is simply the most obvious practical strategy for peasant farmers to take, and may bear no real relation to what people were likely to say to, for instance, their drinking friends.

[199] Even at the birth of twins, it must be noted who emerged from the womb first to establish who is zoky and who is zandry. For there not to be rank between siblings is inconceivable.

[200] Larson not only finds no evidence for a “hidden transcript” that flat out rejected the basic terms of royal ideology (2000: 256–57), he insists no such hidden transcript existed. With all due respect for Larson’s exemplary scholarship, I don’t understand on what basis anyone could claim to know for certain what Malagasy peasants were not saying behind closed doors.

[201] There is a similar egalitarian message in mortuary ritual. At famadihana, everyone is supposed to dress equally modestly, and if possible, more or less the same. Distinctions are to be effaced in order to emphasize equality in common descent. During the nineteenth century, mortuary ritual focused on the collective dragging of granite stones, much like the dragging of trees for royal houses, to construct tombs.

[202] One must bear in mind that, during most of this period, the Queen was in fact a figurehead.

[203] For a somewhat analogous argument, see Bloch’s excellent “Hierarchy and Equality in Merina Kinship” (1986b).

[204] Or to be more accurate, between pain and the imagination. Pain, she argues, is sensation without an object; imagination, object without sensation.

[205] Or really, to own up to doing so. After all, no one developing a theory of ritual writes as if ritual is a phenomenon that exists only in Africa and parts of Eurasia, but not in, say, South America. Analytical terms are always universal. As anthropologists discovered in the 1970s when they began deconstructing away every familiar term from “marriage” to “religion,” once you have done so, you have very little left to talk about, except perhaps some abstract theories of structures of the mind—which then turned out to be ridiculously simplistic.

[206] I have been referring to “cultural relativism” in a broad sense. In fact, there are various kinds and degrees of such relativism. Mark Whitaker (1996) distinguishes three: (1) conventional cultural relativism, which holds that any human action can only be understood in its cultural context; (2) epistemological (or cognitive) relativism, which holds that different systems of knowledge are fundamentally incommensurable; and (3) ethical relativism, which insists that cross-cultural judgments are therefore impossible. Each clearly builds on the others. When I speak of “classical relativism” I am really speaking of the rather haphazard mix of the three that seems to emerge when anthropologists find themselves arguing with those they consider universalists.

[207] Since scholars have a tendency to read sentences like that in strangely reductionist ways, allow me underline: I said “first and foremost” about authority. Not “only.” Obviously it should be about everything else as well.

[208] I also should point out that I’m aware I’m being a bit hypocritical here by indulging in some of the same sort of sectarian reasoning I’m otherwise critiquing: there are schools of Marxism which are far more open-minded and tolerant, and democratically organized. There are anarchist groups which are insanely sectarian; Bakunin himself was hardly a model for democracy by any standards.

[209] One might think of this as the Tom Franks version of history.

[210] This paper was originally presented at a conference in Paris between the 12^th and 14^th of June 2003, entitled Perspectives d’une théorie sociologique générale à l’ère de la mondialisation (Perspectives on a General Sociological Theory in the Era of Globalization), sponsored by Alain Caillé of the MAUSS group. The conference was intended as a kind of summit of social theorists, attended by such luminaries as Margaret Archer, Raymond Boudon, Shmuel Eisenstadt, Bruno Latour, Hans Joas, Anne Rawls, Saskia Sassen, and Alain Touraine. Inviting me was very much an act of generosity on Caillé’s part and I still greatly appreciate it. The essays from the conference were later published in Revue du MAUSS Semestrielle, all without footnotes or bibliography. I have decided here to preserve it in its original published form. The first part of the title is the one chosen for the piece by the French editors (Alain Caillé and Stephane Dufois). The second half is the original question the organizers posed to all participants in the conference.

[211] But not those that speak Spanish or Portuguese. It is not clear if Huntington has passed judgment on the Boers.

[212] It was utterly unremarkable, for example, for a Ming court official to be a Taoist in his youth, become a Confucian in his middle years, and a Buddhist on retirement. It is hard to find parallels in the West even today.

[213] Some of his statements are so outrageous (for example, the apparent claim that, unlike the West, traditions like Islam, Buddhism, and Confucianism do not claim universal truths, or that, unlike Islam, the Western tradition is based on an obsession with law) that one wonders how any serious scholar could possibly make them.

[214] Actually, one often finds some of the authors who would otherwise be most hostile to Huntington going even further, and arguing that love, for example, is a “Western concept” and therefore cannot be used when speaking of people in Indonesia or Brazil.

[215] Or a French person to read Posidonius’ account of ancient Gaul and identify with the perspective of an ancient Greek (a person, who if he had actually met him, he would probably first think was some sort of Arab).

[216] This is why Classical Greek philosophers are so suspicious of democracy, incidentally: because, they claimed, it doesn’t teach goodness.

[217] This conclusion is in world-systems terms hardly unprecedented: what I am describing corresponds to what David Wilkinson (1987) for example calls the “Central Civilization.”

[218] One reason this is often overlooked is that Hegel was among the first to use “the West” in its modern sense, and Marx often followed him in this. However, this usage was, at the time, extremely unusual.

[219] One should probably throw in a small proviso here: Orientalism allowed colonial powers to make a distinction between rival civilizations, which were seen as hopelessly decadent and corrupt, and “savages,” who insofar as they were not seen as hopelessly racially inferior, could be considered possible objects of a “civilizing mission.” Hence Britain might have largely abandoned attempts to reform Indian institutions in the 1860s, but it took up the exact same rhetoric later in Africa. Africa was thus in some ways relegated to the “savage slot” that had been the place of the West—that is, had been before Europeans decided they were themselves “Westerners.”

[220] “Though the first English planters in this country had usually a government and a discipline in their families and had a sufficient severity in it, yet, as if the climate had taught us to Indianize, the relaxation of it is now such that it is wholly laid aside, and a foolish indulgence to children is become an epidemical miscarriage of the country, and like to be attended with many evil consequences” (op. cit.).

[221] Usually, one can pick out pro-democratic voices here and there, but they tend to be in a distinct minority. In ancient Greece, for instance, there would appear to be precisely three known authors who considered themselves democrats: Hippodamus, Protagoras, and Democritus. None of their works, however, have survived so their views are only known by citations in anti-democratic sources.

[222] It’s interesting to think about Athens itself in this regard. The results are admittedly a bit confusing: it was by far the most cosmopolitan of Greek cities (though foreigners were not allowed to vote), and historians have yet to come to consensus over whether it can be considered a state. The latter largely depends on whether one takes a Marxian or Weberian perspective: there was clearly a ruling class, if a very large one, but there was almost nothing in the way of an administrative apparatus.

[223] Obviously the Chinese state was profoundly different in some ways as well: first of all it was a universalistic empire. But, Tooker to the contrary, one can borrow an idea without embracing every element.

[224] Rather than pretend to be an expert on early-twentieth-century Indian scholarship, I’ll just reproduce Muhlenberger’s footnote: “K.P. Jayaswal, Hindu Polity: A Constitutional History of India in Hindu Times 2^nd and enl. edn. (Bangalore, 1943), published first in article form in 1911–13; D.R. Bhandarkar, Lectures on the Ancient History of India on the Period from 650 to 325 B.C., The Carmichael Lectures, 1918 (Calcutta, 1919); R.C. Majumdar. Corporate Life in Ancient India, (orig. written in 1918; cited here from the 3^rd ed., Calcutta, 1969, as Corporate Life).”

[225] I say “almost.” Early Buddhism was quite sympathetic: particularly the Buddha himself. The Brahamanical tradition however is as one might expect uniformly hostile.

[226] Most were in fact published in a journal called Symbols.

[227] One is tempted to say this leaves us to choose between two theories for the origin of Huntington’s “Western civilization,” one neoliberal, one crypto-fascist. But this would probably be unfair. At least the authors here do treat the broad zone that later includes Islam as part of a “Western” bloc to which they attribute the origin of Western ideas of freedom: though it is hard to do otherwise, since virtually nothing is known of what was happening in Europe during this early period. Probably the most fascinating contribution is Gregory Possehl’s essay on Harappan civilization, the first urban civilizaion in India, which, as far as is presently known, seems to have lacked kingship and any sort of centralized state. The obvious question is what this has to say about the existence of early Indian “democracies” or “republics.” Could it be, for instance, that the first two thousand years of South Asian history was really the story of the gradual erosion of more egalitarian political forms?

[228] I am drawing here on a conversation with Nolasco Mamani, who, among other things, was the Aymara representative at the UN, in London during the European Social Forum 2004.

[229] I’m adopting here the name most commonly employed by participants in North America. Most firmly reject the term “anti-globalization.” I have in the past proposed simply “globalization movement,” but some find this confusing. In Europe, the terms “alternative-” or “alter-globalization” are often used, but these have yet to be widely adopted in the US.

[230] Obviously, this assumes that the groups in question are broadly on the same page; if a group were overtly racist or sexist no one would ask about their internal decision-making process. The point is that questions of process are far more important than the kind of sectarian affiliations that had so dominated radical politics in the past: i.e., Anarcho-Syndicalists versus Social Ecologists, or Platformists, etc. Sometimes these factors do enter in. But, even then, the objections are likely to be raised in process terms.

[231] That policy can be summed up by the New York Times’ senior news editor, Bill Borders, who, when challenged by FAIR, a media watchdog group, to explain why the Times provided almost no coverage of 2000’s inauguration protests (the second largest inaugural protests in American history), replied that they did not consider the protests themselves to be a news story, but “a staged event,” “designed to be covered,” and therefore “not genuine news” (FAIR 2001). FAIR, needless to say, replied by asking in what sense the inaugural parade itself was any different.

[232] One effect of the peculiar definition of violence adopted by the American media is that Gandhian tactics do not, generally speaking, work in the US. One of the aims of nonviolent civil disobedience is to reveal the inherent violence of the state, to demonstrate that it is prepared to brutalize even dissidents who could not possibly be the source of physical harm. Since the 1960s, however, the US media has simply refused to represent authorized police activity of any sort as violent. In the several years immediately preceding Seattle, for instance, forest activists on the West Coast had developed lockdown techniques by which they immobilized their arms in concrete-reinforced PVC tubing, making them at once obviously harmless and very difficult to remove. It was a classic Gandhian strategy. The police response was to develop what can only be described as torture techniques: rubbing pepper spray in the eyes of incapacitated activists. When even that didn’t cause a media furor (in fact, courts upheld the practice), many concluded Gandhian tactics simply didn’t work in America. It is significant that a large number of the Black Bloc anarchists in Seattle, who rejected the lockdown strategy and opted for more mobile and aggressive tactics, were precisely forest activists who had been involved in tree-sits and lockdowns in the past.

[233] Those with puppets have been attacked and arrested frequently as well, but, to my knowledge, the corporate media has never reported this.

[234] I owe the phrase to Ilana Gershon.

[235] Similar themes recur in many interviews with radical puppeteers. This is from Mattyboy of the Spiral Q Puppet Theater in Philadelphia: “OK, I’m 23. I’ve lost 13 friends to AIDS. This is wartime, it’s a plague. This is the only way for me to deal with it. With puppets I create my own mythology. I bring them back as gods and goddesses” (Freid 1997). One illustrated volume on Bread & Puppet is actually called Rehearsing with Gods: Photographs and Essays on the Bread & Puppet Theater (Simon & Estrin 2004).

[236] The Pagan Bloc has been a regular fixture in large-scale actions since Seattle, and, unlike the Quakers and other Christian proponents of civil disobedience, was willing, ultimately, to recognize Black Bloc practice as a form of nonviolence and even to form a tacit alliance with them.

[237] Videographers documented police commanders on the first day reassuring activists that the Seattle police “had never attacked nonviolent protesters and never would.” Within hours the same commanders had completely reversed course.

[238] The best source I’ve found on these events is in Boski (2002).

[239] Blocking a street is in fact technically not even a crime, but an “infraction” or “violation”: the legal equivalent of jaywalking, or a parking ticket. If one violates such ordinances for nonpolitical purposes one can normally expect to receive some kind of ticket, but certainly, not to be taken to a station or to spend the night in jail.

[240] New York Times, June 6, Corrections, A2. The original story was significantly entitled, “Detroit Defends Get-Tough Stance” (Christian 2000). The correction reads: “An article on Sunday about plans for protests in Detroit and in Windsor, Ontario, against an inter-American meeting being held in Windsor through today referred incorrectly to the protests last November at the World Trade Organization meeting in Seattle. The Seattle protests were primarily peaceful. The authorities there said that any objects thrown were aimed at property, not people. No protesters were accused of throwing objects, including rocks and Molotov cocktails, at delegates or police.”

[241] This document was transcribed and widely circulated on activist listservs at the time. According to one story in the Miami Herald (Fleischman 2003), it derived from “retired DEA agent Tom Cash, 63, now senior managing director for Kroll Inc., an international security and business consulting firm.” Cash in turn claimed to derive his information from “police intelligence” sources.

[242] A number of Molotovs were thrown during the FTAA summit in Québec City, apparently all by Québec City residents. But francophone Canada has a very different tradition of militancy.

[243] One has to wonder where they actually get these things. A typical example from my own experience comes from the World Economic Forum protests in New York in early 2002. Police at one point attacked a group of protesters, who were part of a crowd waiting to begin a permitted march, when they observed them distributing large plexiglass posters that were designed to double as shields. Several were dragged off and arrested. Police later circulated several different stories explaining the reasons for the attack, but the one they eventually fixed on was a claim that the arrestees were preparing to attack the nearby Plaza Hotel. They claimed to have discovered “lead pipes and jars full of urine” on their persons—though in this case they did not actually produce the evidence. This is a case on which I have some first-hand knowledge, since I knew the arrestees and had been standing a few feet away from them when it happened. They were, in fact, undergraduate students from a small New England liberal arts college who had agreed to have their preparations and training before the march video-taped by a team of reporters from ABC Nightline (the reporters, unfortunately, were not actually there at the time). A less likely group of thugs would have been hard to imagine. Needless to say, they were startled and confused to discover police were claiming that they had come to the march equipped with jars of urine. In such cases, claims that urine or excrement were involved is considered, by activists, instant and absolute proof that the police had planted the evidence.

[244] There is also no clear evidence that 1960s protesters spat on soldiers any more than early feminists actually burned bras. At least, no one has managed to come up with a contemporary reference to such an act. The story seems to have emerged in the late 1970s or early 1980s, and, as the recent documentary Sir! No Sir! nicely demonstrates, the only veteran who has publicly claimed this happened to him is likely to be lying.

[245] I have been unable to trace who first publicly announced such claims, though my memory from the time was that they were voiced almost simultaneously from Mayor Riordan of Los Angeles and a Philadelphia Democratic Party official, during the preparations for those cities’ respective primaries. The claim was obviously also meant to appeal to conservative stereotypes of liberals as members of a “cultural elite”—but it had surprisingly wide influence. As Stevphen Shukaitis (2005) has pointed out, it has been reproduced even by sympathetic voices in the NGO community. While I have not conducted systematic surveys of the socio-economic background of anarchists in the course of my own research, I can rely on six years of personal experience to say that, in fact, “trust fund babies” in the movement are extremely rare. Any major city is likely to have one or two, often prominent simply because of their access to resources, but I myself know at least two or three anarchists from military families for every one equipped with a trust fund.

[246] One common fear is that wooden dowels used in their construction could be detached and used as cudgels, or to break windows.

[247] “In a Zogby America survey of 1,004 adults, 32.9% said they were proud of the protesters, while another 31.2% said they were wary. Another 13.2% said they were sympathetic and 15.7% irritated and 6.9% said they were unsure.” Considering the almost uniform hostility of the coverage, the fact that a third of the audience were nonetheless “proud,” and that less than one in six were sure their reaction was negative, is quite remarkable.

[248] Probably the destruction of productive capacity as well, which must be endlessly renewed.

[249] It might be significant here that the United States’ main exports to the rest of the world consist of Hollywood action movies and personal computers. If you think about it, they form a kind of complementary pair to the brick-through-window/giant puppet set I’ve been describing. Or, rather, the brick/puppet set might be considered a kind of subversive, desublimated reflection of them—the first involving peans to property destruction, the second, the endless ability to create new, but ephemeral, insubstantial imagery in the place of older, more permanent forms.

[250] Some of this history is retold, and the story brought forward to Reclaim the Streets and the current carnivals against capitalism. See Grindon (2006).

[251] For one good example of such reflections, see Wise Fool Puppet Intervention (n.d.). Wise Fool traces its art more back to Medieval mystery plays than festivals, but it provides a nice historical perspective.

[252] Where they will normally turn on shows which take the perspective of the same police in charge of getting them off the streets to begin with. More on this later.

[253] See Bitner (1990) for a good summary of police sociology’s understanding of these constraints and the general issue of “discretion.” Since most Americans assume that police are normally engaged in preventing or investigating crimes, they assume that police conduct is freighted with endless bureaucratic restraints. In fact, one of the great discoveries of police sociology is that police spend a surprisingly small percentage of their time on criminal matters.

[254] Bittner’s phrase. See also Neocleus (2000).

[255] Consider here the fact that “police negotiators” are generally employed in hostage situations. In other words, in order to actually get the police to negotiate, one has to literally be holding a gun to someone’s head. In such situations police can hardly be expected to honor their promises: in fact, they could well argue they are morally obliged not to.

[256] Organizers at Genoa uniformly spoke of their shock during the actions when all the police commanders, whose cell phone numbers they had assembled, suddenly refused to answer calls from activists.

[257] I have yet to hear of a passing pedestrian or other member of “the public” who was injured by even the rowdiest anarchist tactics; in any large-scale action, large numbers of passing pedestrians are likely to end up gassed, injured, or arrested by police.

[258] I have developed these themes in much greater detail elsewhere (Graeber 2006).

[259] Peter Kropotkin (1909, 1924), still probably the most famous anarchist thinker to have developed an explicit ethical theory, argued that all morality is founded on the imagination. Most contemporary anarchists would appear to follow him on this, at least implicitly.

[260] Particularly his Imaginary Institution of Society (1987). Again, this is a theme that I can only fully develop elsewhere, but one could describe the history of left-wing thought since the end of the eighteenth century as revolving around the assumption that creativity and imagination were the fundamental ontological principles. This is obvious in the case of Romanticism, but equally true of Marx—who insisted, in his famous comparison of architects and bees that it was precisely the role of imagination in production that made humans different from animals. Marx, in turn, was elaborating on perspectives already current in the worker’s movement of his day. This helps explain, I think, the notorious affinity that avant-garde artists have always felt with revolutionary politics. Rightwing thought has always tended to accuse the Left of naiveté in refusing to take account of the importance of the “means of destruction,” arguing that ignoring the fundamental role of violence in defining human relations can only end up producing pernicious effects.

[261] One might draw an analogy here to the collapse of levels typical of consensus decision-making. One way to think of consensus process is an attempt to merge the process of deliberation with the process of enforcement. If one does not have a separate mechanism of coercion that can force a minority to comply with a majority decision, majority voting is clearly unadvisable—the process of finding consensus is meant to produce outcomes that do not need a separate mechanism of enforcement, because compliance has already been guaranteed within the process of decision-making itself.

[262] I am referring here, of course, to Carl Schmitt and Walter Benjamin, and more recently, to Toni Negri and Giorgio Agamben.

[263] The T-shirt of the Arts in Action collective, which actually makes many of these puppets, features a quote from Brecht: “We see art not as a mirror to hold up to reality but as a hammer with which to shape it.”

[264] It is interesting to observe that there is a longstanding tradition in American thought that sees creativity as inherently anti-social, and therefore, demonic. It emerges particularly strongly in racial ideologies. This however is properly the subject for another essay.

[265] The fact that almost all the principal figures involved in the repression of protest in America ended up as “security consultants” in Baghdad after the American conquest of Iraq seems rather telling here. Of course, they rapidly discovered their usual tactics were not particularly effective against opponents who really were violent—capable, for example, of dealing with IMF and World Bank officials by actually blowing them up.

[266] Clint Eastwood, of course, in his shift from Spaghetti Western to Dirty Harry, was the very avatar of the transformation. The moment cop movies rose to prominence, cowboy movies effectively disappeared.