Part 2, Chapter 6

Dancing With Corpses Reconsidered: An Interpretation of Famadihana (In Arivonimamo, Madagascar)

People :

Author : David Graeber

Text :

6 — Dancing With Corpses Reconsidered: An Interpretation of Famadihana (In Arivonimamo, Madagascar)

In September of 1990, I was talking with a woman named Irina about something an ancestor of hers had done some sixty years before. Like all of the andriana or nobles of Betafo (a community to the north of the town of Arivonimamo, in Imerina, Madagascar) she was descended from a certain Andrianambololona, whose body, together with that of his wife and daughter and those of three of his retainers, was buried in a large white tomb in the center of the village of Betafo, a five-minute walk across the rice fields from her house.

This particular ancestor, she was telling me, has long had the custom of appearing to his descendants in dreams to announce when the occupants of the tomb felt cold, and needed to have a famadihana performed: that is, to be taken out and wrapped in new silk shrouds. When this happened in 1931, his descendants quickly got together and organized the ritual; but, in their hurry perhaps, forgot to exhume the bodies of his three retainers, who were buried at the foot of the tomb somewhat apart from the rest. “The afternoon after they’d finished,” she said, “the town suddenly caught fire and burned to the ground. And the next morning he came once more to the person”—the one who had originally had the dream—“and said: ‘if you don’t wrap us all, next time I’ll kill you outright...’ So they got the tombs ready again and rewrapped them.”[91]

This story is a good place to begin an essay about the Merina practice of famadihana, if for no other reason because it shows how high the stakes involved can be. Admittedly, this was the worst disaster of its kind I heard about. Irina was doubtless justified in concluding that her ancestor was unusually “arrogant and cruel.” But stories like this were in no way unusual. Rural communities in Imerina were, I found, largely organized around the memory of ancestors whose presence in the lives of their descendants made itself felt largely in terms of the constraint and violence they were capable of inflicting on them. The dangers surrounding famadihana—and these were said to be great—really only marked them as the culminating moment in an ongoing relationship between memory and violence that was implicit in the organization of everyday life, but was here played out over the actual bodies of the ancestral dead.

The theme of ancestral violence was not one that everyone in Betafo was entirely comfortable with. Older men usually did their best to avoid speaking about such matters at all, at least with me, and instead echoed the themes of formal rhetoric, where ancestors were represented as the benevolent guardians of the moral unity of the community of their descendants. Several people besides Irina told me much the same story about the fire of 1931 as she did. The few older men I asked denied anything of the sort had ever happened.[92]

Generally speaking, points of view like Irina’s have not found their way into the ethnographic literature on Madagascar (there are exceptions: cf. Astuti 1995), so, in part, this essay is meant to fill a gap. More importantly, it’s meant to address the question of why such radically different perspectives should exist within the same community to begin with.

Some Background

The classic interpretation of famadihana is that of Maurice Bloch (1971, 1982), who has argued that, through such rituals, participants create the image of a timeless, idealized ancestral order identified with death and the past; one explicitly set apart from life, fertility, and the mundane contingencies of everyday existence. Most people’s fundamental sense of social identity, he says, is based on membership in descent groups that are still identified with territories from which their families have long since moved away. Hence, groups which no longer exist on the level of daily life have to be reconstituted in death, by reassembling and reordering the bodies of the dead.

My intention in this essay is not so much to take issue with this argument, but to take off from a different point Bloch made in his early writings on famadihana—that these are rituals more than anything else about the connection between memory and violence (1971: 168–169). Bloch’s analysis is based on fieldwork in a part of Imerina that had experienced unusually high levels of out-migration; it also reflects an ongoing theoretical interest in questions of ideology—particularly, in how ritual acts legitimate relations of authority. My own fieldwork was in an area where local descent groups still provided the basic framework of local politics, and I am more interested in immediate questions of action: just what are the dead supposed to do to the living, just what are do the living do to the dead?

Let me begin, however, by explaining precisely what famadihana are, and what they are like.

A century ago, the word famadihana was used to refer any ritual which involved transferring a body from one place of burial to another.[93] According to contemporary accounts (Callet 1908: 272–3; Cousins 1963 [1876]: 79–81; Haile 1891), there were several reasons why this might be done. In dedicating a new tomb, for instance, it was (as it still is) the custom to remove the bodies of one’s immediate ancestors from wherever they had been buried and place them in positions of honor within it. Famadihana might also be held to return the body of someone who had been buried temporarily in some other part of the country. Finally, if for some reason it was considered dangerous or inauspicious to open a person’s ancestral tomb at the time they happened to die, that person would often be buried in a shallow grave at the tomb’s foot, and left there for months or even years, until such time as the astrologer determined it was safe to let them enter. Transferring such bodies was also considered a form of famadihana.

Now, it had long been the Merina custom to wrap dead bodies before burial in one or more lambamena, mantles made of colorfully dyed Malagasy silk. It had also been a common practice, when tombs were opened during funerals or famadihana, to replace the worn-out lamba of those ancestors already in the tomb with new ones. At some point around the end of the nineteenth century, doing so sometimes became an end in itself, and people began to perform famadihana simply for the purpose of renewing their ancestors’ shrouds.[94] And, it would seem, this aspect became more and more important as time went on, to the point that, while the older forms are certainly still practiced, everyone I talked to between 1989 and 1991 from the region of Arivonimamo and, for that matter, elsewhere took it for granted that the rewrapping of ancestral bodies was what famadihana were basically about.

Usually, I was told this should be done once every six or seven years—the exact number is often said to vary from tomb to tomb. Often, also, famadihana are said to be held because some ancestor demanded it, appeared like Andrianambololona appearing in a dream or vision to complain of being cold.

The overwhelming majority of famadihana about which I have information fell into one of two categories. The first were “return famadihana” (Bloch’s phrase: 1971: 146), their sponsors almost always families no longer living in the region who still periodically disinterred their dead for reburial there. When they did so they would almost always take advantage of the occasion to rewrap other ancestors in the same tomb. While some return famadihana were quite elaborate, the most celebrated and important famadihana of any given year were almost always of the second kind: dedicated to one particular ancestor who, dead usually some five to ten years, had never been the object of a famadihana before.[95] Four or five different tombs might be opened at a famadihana of this kind, since it was considered important to honor, as well, each of the ancestor’s own immediate ascendants (mother, father’s mother, mother’s father, and so on), and often these more distant ancestors were buried in different tombs. But the focus was always on the final tomb, from which the ancestor around whom the ceremony was organized was always the very last to emerge.

A Capsule Description

No matter what sort of famadihana, or how many tombs were involved, the sequence of events at each tomb was always more or less the same. It’s fairly easy, then, to construct a generic description. What follows is an outline of this basic sequence of events, roughly modeled on the sort of accounts participants would give me when speaking of such things in the abstract, but mainly drawing on my own observations from the eight or nine famadihana I attended between June 1989 and January 1990, all in the region of Arivonimamo and all but two in Betafo.[96] Most of what I say could be a description of any one of these.

Having decided to hold a famadihana, the sponsors would first consult an astrologer to find the appropriate date and time for the opening of the tombs. Next, they had to inform the local government offices—an old colonial law stipulates that no tomb can be opened without the names of the ancestors to be exhumed being registered and a tax paid for each tomb to be opened. During the two or three months which usually intervene before the ceremony, everyone affiliated with the tomb had to be informed, and money raised to pay for the feast, musicians, and the lambamena themselves. In Arivonimamo some families weave their own lambamena (which are very expensive); if so, work had to begin at least a month or two in advance.

The night before the tomb was to be opened, the sponsor and a few companions mounted the tomb and call out the names of the ancestors to be rewrapped, asking them all to return if they happen to have strayed. This stage is always important in accounts of famadihana, but it’s conducted largely outside the public gaze by a few close kin.

The famadihana proper began the next day with a procession from the sponsor’s home town or village to the tomb. Between the zana-drazana—the “children of the ancestors”—and their guests there were usually at least several hundred people in attendance, dressed in what’s called “Malagasy” style: this, in effect, means that rural people wore their best attire short of formal Sunday clothes, while city people dressed down in something approximating rural dress. This is important because any such gathering will necessarily involve a certain amount of tension between the members of a group who still live on the ancestral lands, called the valala mpiandry fasana or “crickets minding the tombs,” and the zanaka ampielezana or “children spread out,” who are only really connected with their place of origin through their tombs. The emphasis is self-consciously egalitarian and often whole households will make a point of all wearing shirts and dresses made from the same cloth.[97]

The astrologer always led the procession, often accompanied by people carrying photographs of the most important ancestors, and always by a man carrying the Malagasy flag (whose presence confirmed the ceremony’s legal authorization). There were always musicians, and usually women carrying rolled-up papyrus mats close behind.

On arrival, the flag was planted on the roof of the tomb, and men took shovels and began removing the earth that covered its buried stone door. Only the valala mpiandry fasana have the right to dig open the doorway, and, if the sponsor was a zanaka ampielezina, there was often a squabble here, the diggers demanding rum before they’d work. Once the digging began, the atmosphere was festive and informal, though with a certain feeling of anticipation: there was music, and some people danced, others carried shovels and other tools back and forth, took breaks from work, and returned.

Once the door was fully uncovered, some of the diggers splashed it with rum and began to move it aside, as others readied candles or lamps, and then began to descend the stairs leading to the inner vault. As they disappeared inside, the female zanadrazana (their numbers sometimes augmented by some young men or boys) arranged themselves in rows, sitting with legs extended on level ground near the tomb. Usually the men splashed a bit of rum over each of the bodies in the tomb and made a brief invocation asking for its blessing before rolling it from its place onto a papyrus mat. Once they had, three or four of them would carry it up the stairs and, as they emerged, shout out the ancestor’s name as the crowd whooped and shouted its enthusiasm. The music picked up at this point; often other men would join in to help carry the body around the tomb three times, and their abrupt stops and starts would lead to its being twisted and crushed inside the mats.[98]

After being taken around the tomb three times, the ancestors were placed on the laps of the women (who were arranged in hierarchical order from east to west or north to south) and the next phase of the famadihana began. Men and women produced bottles, some full of honey and rum, others cow fat or, occasionally, cologne. There were plastic bags full of honeycombs or pastel-colored “Malagasy” candies, pieces of ginger, and coins. Some moved from body to body, pouring rum and honey over each; others handed the bottles to the seated women, often after taking a sip or swig themselves. Sometimes, a woman would produce a stick of tobacco, put half in her mouth, and the rest inside the tatters of a dead husband’s lamba. Others broke off pieces of honeycomb to place inside the folds of cloth around where the ancestor’s head or chest would be. The same was done with the coins, ginger, and pieces of candy. I’ve been told that some people leave small bottles of rum in the wrappings during one famadihana and drink them during the next, and often I heard about people who take dust from inside the wrappings and smear it on their faces or gums as tooth medicine (though I must say I never saw it done myself), or take a handful of beads from the ancient cloth to preserve for the same reason.

This sequence of giving, taking, and sharing was always referred to as a fangatahana tsodrano or “request for the blessing” of the ancestors—though famadihana as a whole could also be spoken of in this same way. The gifts were called fangataka, “tokens of request,” and participants occasionally called out to the ancestors, beseeching them to give their blessing. It was also always a moment of great emotional intensity. Women, particularly if they had the remains of a close relative on their lap, clearly found what they are doing frightening, sad, and disturbing. Many appeared in something close to a state of shock, barely managing to hold back tears, and in every famadihana I myself attended at least one such woman did break down and cry—others quickly crowded around to do their best to reassure, comfort, or distract her, always reminding her that “this is an act of celebration, not of mourning.”

Next, men divided into teams around each ancestor to begin the actual wrapping. The old lamba were never removed but left in place; nor were the bodies allowed to touch the ground—in fact, the initial stages are done while the body was still on the women’s laps—since it was very important to ensure that at no time will an ancestor touch the earth. Generally, they were rolled first into white sheets, and then one by one into the thicker and more durable lamba. There were almost always at least two layers of cloth all told: mainly silk lambamena for the more important ancestors, polyester for the rest. While women looked on and often gave advice, the actual process of wrapping the bodies, and then tying the resulting bundle together with cords or strips of cloth (there should ideally be seven of these) was always performed by men—who spared no efforts to roll and bind the ancestral bundles as tightly as they could.

This being done, the music picks up once again in volume and tempo and the final, joyous part of the ceremony begins. Mixed groups of men and women carry the bodies, once again carried in their mats, one by one around the tomb, this time stopping and starting and dancing even more vigorously and even violently than they had in the beginning, with all sorts of roughhouse, shouting, whoops, and cries, people generally throwing themselves about in a sort of delirious abandon.[99] The razana are once more twisted and crunched about a great deal over the course of the dancing, which may last around fifteen minutes, until finally being returned to their places inside.

With this, the business is basically finished. If there are more tombs to be opened, a procession will form behind the astrologer once again. If this was the last, the sponsor and some local elders or politicians will mount the head of the tomb to make brief formal speeches summarizing the days events and their significance, and thanking everyone who came. After this the crowd begins to drift off, and a group of men take shovels and begin to pile back the dirt removed from the door to the vault. In theory, it should be the oldest man among the local zanadrazana who removes the first shovelful of earth from the doorway at the beginning of the ceremony; the first returned at the end should, I was told, be done by a young man whose father and mother are both still living.

Later still, often around nightfall when everyone else has long since left, the astrologer and a few assistants will come back to the door of the tomb to make a fanidi-pasana—a “lock to the tomb”—by burying a few magical objects in or around the doorway. These, if placed correctly, should ensure that the ghosts of those “turned over” remain in the tomb and cannot emerge again to trouble the living.[100]

Descent Groups

Merina society is divided up into a number of cognatic descent groups, which in the literature are usually referred to as foko.[101] Bloch calls them “demes” because they tend towards endogamy and are closely identified with ancestral territories. About a third of them claim andriana, or “noble” rank; the rest are hova or “commoner” demes. There’s also a significant portion of the population made up of people descended from nineteenth-century slaves. These mainty, or “black people,” are not organized into demes and don’t usually intermarry with the fotsy or “whites,” though in other respects they share the same social organization.

Each deme has its history, usually beginning with an account of the origins of its founding ancestor, how he came to the territory on which his descendants now reside, how by his various movements he defined its boundaries, created its villages, named various prominent aspects of its landscape, and so on. In most cases, the stories go on to how he subdivided the territory by giving each of his children (or sometimes, each of his wives) their own village or territory: that of the eldest furthest to the east, with the others ranging westward in order of seniority (cf. Condominas 1960: 199–203; Rasamimanana & Razafindrazaka 1957 [1909]: 9–13, etc.).

If most people could tell you from which of these branches they consider themselves to descend, it’s not because they could trace any genealogical link to the founder. Genealogical memory was extremely shallow: I met very few people who could remember further back than to their grandparent’s generation or, at any rate, to people they personally remembered from their childhood. Nor are deme divisions in most cases any longer identified with clearly bounded territories (if they ever really were). What’s significant is not where one lives, but the location and history of one’s tomb.

Merina villages are surrounded by tombs—usually there’s literally no place one can stand outside without being in sight of one. Ancient tombs, by now little more than grassy mounds of earth, sit next to white-washed stone and cement ones topped with wreathes and stone crosses, and—now and then, if there’s a particularly wealthy family in the neighborhood—brightly painted palatial structures on wide platforms, their doorways shielded by metal lattice gates. Whatever their size though, their granite solidity is meant to contrast with houses, which are never built of stone but usually of mud brick. Clearly, tombs were meant to be symbols of permanence; constant reminders of the enduring presence of the ancestors.

They are also organized into a hierarchy: and it’s this hierarchy of tombs which forms the real physical framework of the deme, and provides the terms of reference by which people can place themselves within it. Most people I knew had only the haziest idea of their deme’s history, but anyone could point out their own tomb, and explain how it fits in.

Andrianambololona, for example, was as I’ve mentioned buried with his wife and daughter in an impressive stone tomb to the east of the village of Betafo.[102] In the western part of the same village were four tombs, each said to hold the bodies one of his five eldest sons, and half an hour’s walk further to the west was a fifth, that of his youngest son who had had a falling out with his seniors and moved away.[103] Each of the deme’s divisions were said to descend from one of these brothers—whose relative rank is remembered even if their names have long since been forgotten. And, while only a handful of the present-day inhabitants actually expected to be buried in one of these ancient tombs, each new tomb that was created was linked to one of them by the affiliation of its founder. In other words, what really knits a deme together is not a human genealogy but a genealogy of tombs. Older tombs are seen as generating younger ones, and the organization as a whole inscribes a pattern of historical memory in the landscape, in a form that makes it seem indelible and permanent.

“Playing with Corpses”

This is not to say that this framework is any sense really permanent and unchanging. In fact it is continually being transformed and redefined through human action. New tombs are constantly being built, old ones emptied and abandoned. Bodies are transferred back and forth; they are broken apart and combined with one another. And in a purely practical sense, it’s this which famadihana can be said to do.

Whatever their outer appearance, Merina tombs are always much the same inside. The doorway always faces west; the door itself is a huge buried slab of stone. Moving it aside, one descends a stair to enter a single large chamber from whose northern, western, and southern walls emerge stone “beds” (farafara) or shelves, set one above the other. Typically there are three shelves on each wall, making nine in all, but people are rarely willing to place bodies on the bottom shelves, so that in most tombs the number available is, effectively, six.

In principle, everyone who has the right to be buried in a given tomb is descended from a single individual, who is referred to as the razambe or “great ancestor” of that tomb. The razambe’s body is always placed either on the highest shelf to the north, or the highest to the east, usually together with his (or her) eldest child. Each of the other children is allotted a different shelf on which they become, as it were, minor razambe, and on which only their descendants have the right to be buried.[104] Sometimes, individual shelves are further subdivided along the same lines. Shelves and spaces on the shelves thus become a form of property: I have even heard of a case of a man in extreme financial trouble who tried to sell his space in a prestigious tomb, though I’m not sure anyone would have dared to buy it from him and eventually his relatives talked him out of the idea.

In practice, however, it’s not only lineal descendants who have access; one can draw on a variety of other connections (marriage, fosterage, blood brotherhood, and so forth), so that most men and almost all women have a fair degree of choice over what tomb they intend to be buried in.[105] However, it was often stressed to me that each tomb has its own regulations concerning who can and cannot be buried in it, and on which shelves. In one, children linked through women are not allowed on the upper shelves. In another, only actual descendants of the razambe are allowed in, not their husbands or wives. These regulations can take many forms but they are always negative in their phrasing—in fact, they are usually referred to as the tomb’s fady or “taboos,” and not distinguished from taboos against, for instance, wearing clothes with buttons that apply inside some tombs, or against the giving of tobacco or participation of slaves during famadihana in others.

The first few times I actually went inside one of these tombs, I was surprised by how few bodies they seemed to contain. Even in very ancient tombs, only two or three of the shelves might hold bodies and, even when there were more, there were frequently only three or four bodies lying on any given one of them—remarkably few, considering some of these tombs had been in continual use for over a hundred years. There were, I found, a number of reasons for this. For one thing, new tombs are constantly being built. On completing a new tomb, it’s customary to take at least one ancestor from one’s former tomb (typically the founder’s grandfather or great-grandfather) to be the new razambe. If one can get all the owners’ permission, a whole shelf’s worth of ancestors might be cleared out, and divided up among those of the new tomb.[106] And since the division of shelves in the old one is considered to have been fixed by ancestral decree—which makes people very reluctant to rearrange the bodies—whole walls of shelves may end up lying empty as a result. For the same reason, demographic vagaries can lead to empty or nearly empty shelves as some branches die out without their space being reapportioned.

A more important reason, however, is that the number of bodies is kept limited by the habit of consolidating them.

Here the reader should understand that these bodies—the Malagasy term razana which is used to refer to them actually means at the same time both “ancestor” and “corpse”—are not really “bodies” at all in any sense suggested by the English word.[107] Certainly they didn’t look anything like human bodies. Mainly they looked like wrapped bundles of red earth.

On death, corpses are always wrapped in one or more lambamena—cloth made of a material that (like the polyester now sometimes used to substitute for it) is valued for its hardness and durability—before being placed in the tomb. New razana have to be left there, undisturbed, for several years until they are considered to be “dry,” by which time little but dust and bones are likely to remain. During a famadihana the bodies are usually subjected to a great deal or rough handling: they’re danced with, pulled and tugged at, wrapped and bound with extreme force, and then danced with again in an even more tumultuous manner before being returned to their shelves. After twenty years and several famadihana, they have been quite literally pulverized: even the skeletons have largely crumbled away, and there’s very little left to serve as a reminder that the thing had once had human form.[108] People say they’ve turned into “dust” (vovoka), and in fact it’s basically impossible to tell what was once body from what was once cloth, both having turned the same brick red color—which, incidentally, is the same as that of the lateritic Malagasy soil.

Bodies can only be combined after their first famadihana—that is to say, after they have been already largely reduced to dust. It’s a relatively simple matter to rewrap two such bodies in the same cloth. In fact, if one doesn’t, ancestors—unless they’re famous ancestors, regularly rewrapped in large numbers of shrouds—tend to become thinner and thinner as time goes on, until in the end they look like mere tubes of cloth that bulge in the middle, the whole bundle the thickness, say, of a person’s arm or leg. On the other hand, razana ikambanana or “combined ancestors,” which for all anyone knows may be made up of the remains of a dozen different individuals, along with all of their old lamba, can often attain a very large size, two or three times that of a living human being.

The most frequent practice was to wrap children in one lamba together with their parents, and husbands together with their wives. I was frequently told that two siblings could never be combined. Apart from this, it’s difficult to generalize, since, as in so many things, different families and tombs have different customs; but almost always, the ancestors combined together are those on the verge of being forgotten—that is, contemporaries of the parents or grandparents of the tomb’s oldest living descendants. Usually, children who died at an early age are the first to be so treated (these are incorporated in their parents); next, adults who died childless, or anyway who no longer have living descendants, and thus no one to provide them with lambamena during future famadihana.[109] These are incorporated with ancestors that do. The names of such minor razana are for the most part quickly forgotten; the same is usually true of wives wrapped together with their husbands, or the occasional husband buried in his wife’s family tomb who’s been combined with her. But in the end, unless the tomb’s owners make a point of marking certain razana with written labels or keep family notebooks—which few do—all but two or three of the most famous older names will inevitably pass from memory. Most older tombs end up containing at least one and often several large bundles referred to only as a razambe ikambanana, or “combined great ancestors,” since none of the current owners have the slightest idea what the name of any of its component ancestors might be.

Since none of these razambe—named or nameless—can ever be removed to another tomb, no tomb, however old, can ever be entirely stripped of bodies. But, as some of the branches of descendants die out and others build new tombs and remove their own immediate ancestors, many tombs reach the point where they are no longer used for burial.[110] Most such tombs will still be opened now and then during elaborate famadihana, and one or two bodies rewrapped. But, at least in my experience, this is usually the occasion of a good deal of confusion, as the zanadrazana inspect the half-dozen or so ancestral bundles left in the tomb, trying to identify their own. And even these connections are not remembered forever. Hillsides are dotted everywhere with the remains of ancient tombs, which often look like nothing more than low mounds with a few worked stones here and there visible through the grass, whose remaining occupants have long since been forgotten. In fact, it might well be that the most prestigious ancient tombs—which are seen as key nodes in the hierarchical framework of the deme—are really just the oldest ones that have managed to avoid being forgotten.

The whole process of pulverizing and then consolidating bodies in famadihana can be seen as the concrete or tangible aspect of a process of genealogical amnesia. The bodies of the ancestors are gradually dissolved away at the same time as their identities are gradually forgotten. In the end, both are destined to be absorbed into that of some more famous razambe. Something of this sort occurs wherever genealogies are important, but, in the Merina case, the whole issue of remembering and forgetting is much more of a tangible problem than it usually tends to be—if only because ancestors are much more tangible objects. If relations with ancestors have to be worked out in so absolutely material a medium, the process of forgetting itself has to be made an active one, rather than something that just happens.

Similarly, while ancestral names played an important role in famadihana—they were called out from the tomb the night before, called out again as the bodies emerge, and usually, listed a fourth time in the speeches which close the ceremony—next to no one made the slightest effort to preserve these names permanently in writing.[111] There’s no reason they couldn’t be: Merina society is a highly literate one. Famadihana are referred to as “memorials” (fahatsiarovona) for the dead, but one of their central ironies is that what they actually accomplish is to make descendants actively complicit in forgetting them.

Cursing and Taboo

Outside of purely ritual contexts like famadihana however, the main way that ancestors manifest themselves in the lives of their descendants is through the imposition of fady or taboos—which, unlike the practices surrounding tombs, have a constant and immediate impact on people’s practical affairs.

A great deal has already been written about Malagasy fady (Standing 1883; Van Gennep 1904; Ruud 1960; Lambek 1992) but rather than review the literature, I’ll limit myself to a few critical points. The first is that the logic of fady is not really the same as, say, that of Polynesian taboo. Fady are not a mark of the sacred or, for that matter, usually of pollution; they are not about the state of persons or things at all so much as about actions which one cannot do. A fady takes the form of a simple statement: “do not do X”; “it is fady to do Y.” It’s the action, not the object or actor which is “tabooed.”[112] To be able to impose such a restriction on others is one of the most basic ways of demonstrating authority over them; to share such a restriction with others is one of the most basic ways of demonstrating solidarity.[113]

There are various sorts of fady. Some are rules of conduct that apply to anyone (or to anyone in a given situation: e.g., “it is fady for a pregnant woman to sit in a doorway”), others apply only to people using or being protected by certain forms of magic, others are imposed by some ancestor and shared by all of his or her descendants. It’s with the latter sort that I’m concerned right now.

The older men who were considered the ultimate authorities on matters of local history and custom almost always described these restrictions to me in moralistic terms: they were the means whereby ancestors maintained the harmony and integrity of the deme. The examples they’d choose were almost always the same: fady against stealing from one’s kindred, against selling deme land to outsiders, or against intermarrying with inferior groups (especially the descendants of slaves). Though always attributed to the local ancestors, the list stayed pretty much the same from group to group. But there were other taboos that really did set one deme apart from another: usually there were certain other demes or divisions of demes of equal status into which its members could not marry; always, a range of animals and plants which they couldn’t raise, or grow, or eat.[114] Tombs often had, as I’ve mentioned, their own sets of fady, usually attributed to their respective razambe; and even living parents had the power to “curse” their descendants never to eat pork, or own white cattle with black markings, or whatever else—thereby creating a taboo.

Many demes had stories about how their most important fady had originally come about. People were in fact much more likely to know these than the more formal deme histories—if only because they were usually much more entertaining as stories. Many were explicitly comic and clearly meant to poke fun at their ancestral protagonists. A Betafo ancestor for instance was supposed to have so gorged himself on pork and garlic that he burst apart and died (whereupon his survivors imposed a taboo on their descendants to prevent them from doing the same). The ancestor of the neighboring Andriamasoandro similarly stuffed himself with the caterpillars he discovered crawling out of trees during a brush fire—but, in some versions, he came to his senses before it was too late and, realizing what a stupid thing he’d been doing, cursed his descendants never to eat caterpillars again.[115]

One might argue that the absurdity is meant to underline the perceived arbitrariness of so many ancestral restrictions. But these were not the only genre of stories concerning fady; there was another which tended to be even more universally known, and for the most immediate practical reasons. These concerned the consequences of their transgression.

The consequences were, with few exceptions, devastating. A rich andriana who married a woman descended from slaves suddenly lost everything he had: now he is a pauper. Someone grew garlic where he shouldn’t have; his crops were destroyed by hail. Someone else tried to remove a body from a tomb in violation of its regulations; he was blasted by lightning and died. There’s no one, young or old, male or female, who could not easily recount a dozen such stories or more. They play an important role in local politics, since there was always a great deal of subtle maneuvering around who can convince others to accept their version of the local taboos. Moreover, it’s almost exclusively these stories that described how the hasina or invisible power of the ancestors actually manifested itself to living people—or, to put this another way, how ancestors continued to act and to play a direct role in their descendant’s daily lives. What’s surprising is that, when they did, it was almost always by attacking them—in fact, by actions which would, had they been carried out by a living person, be instantly condemned as the most reprehensible kind of witchcraft.

No one would openly suggest that ancestors were anything like witches: as I’ve mentioned, elders in particular tended to picture them as the benevolent guarantors of the unity and moral integrity of the group. On the other hand, many of these old men grew distinctly uncomfortable whenever anything touching on the question of ancestral retribution was brought up—just as they would if any mention was made of witchcraft. Within a community, it seems it’s mainly women who pass these stories on. And most of the women I spoke to didn’t hesitate to express their own opinions about the ancestors’ behavior—in fact, the word that came up most often in talking about them, masiaka, means “savage,” “violent,” or “cruel.”[116] Part of the older men’s reluctance to talk about ancestral violence probably had to do with the fact that they were very close to being ancestors themselves and simply as figures of authority, tended to identify with their position. They themselves wielded the power of ozona, or “cursing,” over their own children. And, insofar as it could be used as a weapon to punish offspring who had proved utterly resistant to advice or admonition, it was the ultimate bastion of parental authority.

I only heard of two or three instances involving people I knew where someone really was cursed, but the possibility of doing so was always being alluded to. By all accounts, such curses always took a negative form: “you will never have any children,” “you will never find prosperity in your life,” or “you will never enter the family tomb.” In other words, whatever the content of a curse, or the means of its enforcement, it never took the form of a direct assault (e.g., inflicting a disease on someone, or causing them to lose all the wealth they did have) but, instead, specified something the victim will never be able to do.

One might say that, while the stories of distant ancestors separate the imposition of restrictions and punishment for their transgression, here the two are merged in a single gesture.[117] But this only underlines what I think is a general principle: that the power to impose restrictions is ultimately continuous with the violence through which those restrictions are enforced.

An Initial Synthesis

One reason that Merina ancestors were felt to be a constraint on the actions of their descendants is that the ultimate aspiration, at least of any man, was to become a prominent ancestor himself. To do so however, he must manage both to overshadow the memory of his own ancestors, and constrain his children—particularly his sons—from either moving away, or overshadowing him in turn.

Ancestors, while they were still alive, were simply people; people who were born, had children, built tombs, and died. This is something which emerges very clearly in deme histories, in which ancestors were never represented as having had powers of action and creativity different from those available to people now.[118] Even when magical powers enter in to these stories (and they only rarely do), they are never magical powers one couldn’t come by in the present day, if one had the skill or was willing to pay for them.

This is very different than a situation in which social divisions are said to have been instituted by, say, divine beings or totemic animals in the mythological past. People still have children and build tombs, and in principle, there’s no reason why they shouldn’t be able to become famous razambe themselves, even—and I met plenty of people willing to entertain this possibility—a razambe on the order of the founder of a deme. This obviously opens up at least the possibility of a rivalry between the living and the dead, since, from any individual point of view, the only reason one can’t achieve such a status is because someone else already has. One might even see the stories about the origins of fady as being statements about the essence of this relationship: because our ancestor took this action (e.g., ate caterpillars), therefore we, his descendants, are never allowed to take that action again. At any rate, the fact that the presence of the ancestors is generally felt through a series of constraints on human action becomes much easier to understand.

In this light, consider the kinship relations which dominate people’s daily lives. Madagascar has been one of those places anthropologists have found troubling in the past owing to the lack of “structure” or rules (Wilson 1977, 1991). Many authors have stressed the degree to which even links of descent are seen as created rather than received (“achieved” rather than “ascribed” is Southall’s usage [1971, 1986]), and underlined the importance of links like fosterage, adoption, blood brotherhood, or other sorts of “friendship” in creating links between people (Vogel 1982; Kottak 1986; Feeley-Harnik 1991, etc.). In Imerina, for instance, property and rights of group membership are conveyed as easily through men and through women, marital residence is flexible and marriage easy to dissolve. Most people have a very wide range of options about how and with whom to live their lives.

At the same time tremendous emphasis is placed on parental authority and the role of elders—of which ozona is only the highest form. In other words, people’s freedom of action is not seen to be much limited or constrained by explicit rules, but very much seen to be constrained by other people, especially those who stand over them in positions of authority. As a result, the social groups which, unlike those organized around tombs, do provide a context for people’s daily affairs and are the stuff of local politics are, in most cases, very much the result of some one ambitious man or woman’s personal project.

A Politics of Movement

Not everyone’s personal project was the same, but there was, I found, a clear idea of what constituted a truly successful career (at least for males), the contours of which can already be made out in folktales written down over a century ago (cf. Dahle 1984 [1878]). The story is always roughly this: the hero leaves home as a teenager or young man to seek his fortune. He succeeds, becoming rich in money, cattle, or slaves. At this point he may return home again, or he might also establish himself somewhere new, but in either case he will acquire land, marry, and sire numerous progeny. His ultimate aim however is not simply this, but, first, to build himself a large and impressive tomb, and second, to prevent his own children from acting as he did. That is to say, he has to provide them with enough land and wealth that at least the larger part of them, along with their own descendants will be content to stay, to keep up the tomb in which he will be remembered as razambe.

It was certainly exceptional for someone to achieve the rank of razambe of an entire deme, but by no means inconceivable.[119] And the deme histories themselves almost always describe the founders as people who had abandoned their ancestral territories in the east to “find themselves a better living” (mitady ravinahitra)—which is exactly the same phrase used to describe what young men do today when they go off to the city, or to Tsiroanomandidy, a former frontier region seventy or eighty kilometers to the west which is still considered a kind of land of opportunity, abounding in cattle and cheap land, in the hopes of striking it rich.

Generational politics, then, comes down largely to a politics of movement, with fathers striving to keep their descendants from leaving, and sons at least dreaming of being able to break away.

I’ve already remarked that, in present-day Imerina, much though a kinship group like a deme is identified with a certain tanindrazana, or ancestral territory, most of its members are likely to live elsewhere. People have been following the same pattern—migrating but keeping their links with the tanindrazana—for well over a century, and it’s fair to say that the majority of the “owners” of any given tomb in, say, Betafo no longer live there. Many reside in the capital, which is only an hour away by van, but there are people living in almost every part of Madagascar as teachers, officials, traders, and the like. In addition, little colonies of people from the Arivonimamo area are scattered throughout Tsiroanomandidy. Since almost all of the men (and a fair number of the women) spend a good deal of their time away from their villages engaged in petty commerce of one sort or another or otherwise looking for money, even those who have relatively little education have access to a larger world.

What this means in the end is that it’s only the wealthiest or most successful farmers who have the means to keep any large proportion of their children around them. The less fortunate see them disappear one by one. Daughters marry away; sons may well do the same, or they may follow their mothers, be adopted by wealthier relatives, or simply head out west or to the capital—intending to stay only long enough to make a little money, really never to return. The result, when combined with that of demographic vagaries, can be quite dramatic: in the space of one or two generations whole villages can disappear; once large and prosperous families can be left with no living descendants in the area at all. On the other hand, the most successful can not only keep most or all of their own sons and even daughters at home, but attach to him- (or occasionally her-) self a whole range of dependents or semi-dependents: poorer brothers or sisters and their children, affines (endogamous marriage is often used to cement such ties), kin through adoption or blood brotherhood, the occasional unrelated servant such as a cattle-herd, and so on. To do so required land, and it is largely to keep their children and dependents around them that most parents divide the lion’s share of their rice fields among their inheritors around the time the latter get married, keeping only a modest portion for themselves.

Local Families and Their Tombs

It was these groups, named after and organized around one prominent individual (again following Bloch [1971: 81–86], I’ll call them “local families”), that provided the real framework for everyday existence.[120] At any given time, a community was basically seen as an agglomeration of such local families, which often held together under a nominal head for a decade or more after the original founder had died. The largest quarter of Betafo, for instance, was made up of three of them: in only one of which was the founder still alive. Between them, they accounted for fifteen of the quarter’s twenty households. The remaining five households were all, in some sense, marginal or fragmentary—most were composed of a single elderly man or woman living with an unmarried adult child, or else, with a small number of younger children or grandchildren. Such families were typically quite poor, and made no particular claim to a public voice.

Local families had a strong tendency to become small social universes unto themselves, working their fields cooperatively even after the patrimony had already been divided, fostering each other’s children, sharing meals freely, and generally allowing people and things to circulate in a far more intimate manner than they would with other neighbors, who were always potential sorcerers. Often, their founders would break away from larger settlements entirely and build little hamlets of their own on a stretch of hillside overlooking their paddy fields.

While different members would have different options, most of the members of any local family would normally expect to be buried in its founder’s tomb. Often, this was one he himself had created; if not, it was usually because he had already succeeded in establishing himself as the exclusive effective owner of one of the most famous ancient tombs—with whose razambe he might ultimately, in the eyes of the neighborhood, be substantially confused.[121]

Now, in either case, this might seem to contradict the notion that the head of the family aspired to be remembered after his death as a local razambe, since in founding a new tomb he would have had to bring in the body of one of his own ancestors to be its razambe, and in the case of the ancient tomb, one is often dealing with an ancestor so famous (one of the children of the deme’s founder for instance) that the identity of the owner could hardly help but be overshadowed. But, in practice, there are a number of different ways things could work themselves out. Even a razambe can be forgotten, or can end up absorbed into some more famous successor (with whom he is often physically merged). I frequently discovered, on inspection of old documents, that the ancestor generally assumed to be the razambe of some tomb was in fact not its oldest ancestor at all, but rather the man who built it. There was a complex politics going on here; one whose very existence was never openly admitted. Everyone spoke of the need to remember and honor their parents after they were dead, but they also knew that the ultimate fame of a father almost necessarily meant the eventual oblivion of his sons (and vice versa). At the same time, much of the daily authority living people had, in their own communities, was derived from that of a more venerable ancestor—most often, in fact, a father—who was no longer alive.

It’s true that in any community there were some people who really didn’t seem to be promoting their own immortality so much as seeking it vicariously through others. But this, according to my own experience, was a strategy mainly adopted by prominent women, and rarely, if ever, by men. Many widows promoted the prestige and memory of a late husband (as daughters often would for their fathers) as razambe of a tomb, thinking little of their own name and reputation in comparison. But women had a very different position in the politics of local families than men, and the relation between fathers and daughters was not at all like the difficult and contradictory relation between fathers and sons. It was, in fact, held to be a particularly close one. I often heard people explain preferences for cousin marriage or other forms of endogamy, for instance, as the result of paternal sentimentality: fathers just weren’t willing to see their daughters move away. A women who has moved always knows that her father would be happy, if he were at all able, to welcome her back should she wish to leave her current husband—all the more so if she has children to bring with her, who will add to the number of his local descendants. To most women, then, a father’s house was a potential refuge; and this was doubtless one factor contributing to the universal assertion that girls naturally form their closest emotional attachments with their fathers, just as boys always tend to remain primarily loyal to their mothers, in childhood as well as later on in life.[122]

The Adults Are All Dead

When I first arrived in Arivonimamo in 1990, I began making the rounds of nearby villages to gather local histories. That each deme should have a history worthy of being told was taken for granted by everyone I talked to. Often, however, it was very difficult to find any one person considered worthy of telling it. Recounting oral histories—at least to outsiders—was felt to be the role of the Ray aman-dReny, or elders of the community; but next to no one, whatever their age, would be so presumptuous as to lay claim to this status themselves. So, if I asked a small group of people if there was anyone who might be able to tell me something about local history, the usual response was to start naming people that were dead. “Well, you could talk to Ingahibe Raoely...except he died six years ago. Then there was Ramatoa Rasoa, but she died just last summer.” In the end, someone would usually come up with the name of a living person, but only after repeating the same stock phrases: “the grown-ups here are all dead; all that’s left are we children who’ve succeeded them.”[123]

This was, I think, more than just a colorful figure of speech. In many ways, the course of a man or woman’s life really is defined in such a way that no one was considered to be unequivocally an adult until they are quite well along in years.

It is reflected, for one thing, by the age categories people used to refer to one another in common speech. While there were a great number of terms for infants and children, there were next to none which formally discriminated between the stages of life that come afterwards. If speaking about someone in a more or less respectful fashion, one normally used titles which were to some extent based on age: a young married woman for instance may be referred to as a Madama, one in her forties or fifties as a Ramatoa, an elder as Ramatoa be (men however have only a much less systematic and less formal set of titles).[124] If speaking of someone younger than oneself—even if it be a fifty-year-old discussing a forty-year-old—or, if speaking in informal or slightly disparaging terms of someone of about one’s own age, one generally employed terms that would best be translated into English as “that kid,” “that boy,” or “that girl” (zaza, ankizy, ankizilahy, bandy, baoikely, ankizivavy, sipa, ikala, ikalakely, idala, etc.). Anyone whose parents were still alive was talked about in such terms constantly, and there was a general feeling that one had not really reached full social maturity until one was at least a grandparent.

Considering that the position of Ray aman-dReny is even more exalted, it should hardly be surprising that there were very few of them around. The simple fact was that anyone fortunate enough to attain such eminence was not likely to enjoy it very long.

What’s more, one effect of the organization described above was to eliminate the vast majority even of the old people in a given community from consideration as elders. The heads of fragmentary families, like the dependent elders within local families, might have been respected for their age, but were never really considered proper Ray aman-dReny whatever their years. Women could in principle be elders, but in practice they were almost never considered so. Other men were disqualified by questions of character. Of the roughly 118 people who lived in Betafo, for instance, there was, in the end, only one man who everyone agreed could be considered an elder—and this was hardly unusual. If anything, Betafo was slightly unusual for having one elder whose status was absolutely uncontested.

On the other hand, anyone who did reach this status attained a social significance so great that it lingered on for many years after they died. Their names were always cropping up in conversation. Time after time, I would note people speaking of Rakoto’s field or Rabe’s house, only later to find out that the Rakoto or Rabe in question had been dead for over a decade. Often, it turned out that the groups these men had founded still existed, perhaps headed now by the man’s widow, or an elder son of less intrinsic authority. As I’ve mentioned, a large share of the local families that exist in any community were of this sort—often, the final division of rights in land and houses among its members had not yet been made. In other cases, the local family had largely broken up, only a few former members were still scattered about the area, and the speaker either was not quite sure who owned the land now, or considered the owner insignificant.

It’s hard to escape the conclusion that the people I talked to about gathering oral histories were really not too far off the mark. The socially determined life-course was so protracted it outruns most biological life spans entirely, and as a result, the Ray aman-dReny of most Merina communities were largely made up of people who were dead.

Gender and the Politics of Memory

Now, I mentioned at the beginning of the essay that the most prominent famadihana of any year were generally dedicated to a single person who had died some four or five years before. In my experience, this person was always the head of a local family whose memory still dominated the lives of his descendants in the way I’ve been describing.

During the winter of 1990, there were four famadihana in Betafo: two return famadihana sponsored by people from Tsiroanomandidy, and two dedicated to important elders. The first was dedicated to a remarkably successful man. A local official under the colonial regime, he had managed to keep all thirteen of his children from leaving his village of Ambaribe. The famadihana, sponsored by his elder sons in conjunction with his widow, involved four tombs and perhaps a thousand celebrants.

The second famadihana, sponsored by the widow and seven surviving children of a man named Rakotondrazaka, also involved four tombs and almost as many participants. It might be useful to go into this example in some more detail to get some idea of what a “famous” famadihana might actually involve. Rakotondrazaka had died in 1982 at the age of sixty-six, two years after having finished a tomb of his own. He also had managed to keep almost all of his descendants around him.

The first tomb to be opened was really not a tomb at all, but a grave containing a single nameless skeleton. The occupant was presumed to be a man killed by the poison ordeal in precolonial times and, for that reason, denied entrance to his family tomb as a witch. He had, however, appeared in a dream to Rakotondrazaka’s eldest son many years before and had ever since been included in the family’s famadihana. The second tomb was that of Rakotondrazaka’s father. Rakotondrazaka had, in fact, removed his father’s body from it when he dedicated his own tomb in 1980, but four of his father’s ancestors remained there. They were rewrapped, and some unrelated people took advantage of the occasion to move in two bodies that had been buried in temporary graves nearby. The third was his mother’s tomb. Here, there were a large number of bodies to be wrapped—most of them not directly related to his mother at all, but the ancestors of a rich but childless woman who had passed on her land to him on condition that he take care of them. The final tomb was Rakotondrazaka’s own, where all of the five razana were taken out, the honoree himself, again, the last of all.

Even at the first and most ancient tombs, there was always a certain feeling of fear and anxiety as the ancestors first emerged, a certain air of triumph and rejoicing as they were returned. But the bodies themselves no longer revealed even the suggestion of a human form, and were not considered particularly pitiful or frightening. Nor were they the bodies of people any of the participants had actually known: in fact, the vast majority knew nothing about them, not even their names.

By the time the last tomb had been opened, and the final ancestor was about to emerge, on the other hand, the tension had built up to the point where it many of the zanadrazana clearly found it almost unbearable. Some of the young men carrying out the bodies appeared—despite the rum they’d been drinking all day to build up their courage—so overwhelmed by what they were doing that their faces were those of people in physical pain, as they forced themselves to carry out their parts in the ritual. Others seemed in to have fallen into an almost trancelike state, stumblingly oblivious to what was going on around them. When the final ancestor did emerge—immediately to be surrounded by a press of descendants who flooded him with rum and other offerings—the emotional pitch reached a climax; few were the women who didn’t at least choke back sobs when the ancestors were first placed on their laps, and one or two would always dramatically break down and cry.

As the process of giving gifts and wrapping the ancestors continued, people gradually regained their composure and, by the end, more or less everyone took part in a mood of celebration. But, after the ceremony, women always tended to remark on who had cried—particularly if they had. “It’s because you still remember the person so vividly,” I was told on several occasions, always in more or less the same words, “and then you see just how little is still left.”

I once asked one of these women why, if famadihana were supposed to be such a happy occasion, there was always someone who burst into tears. She looked at me a bit quizzically, and, observing that these were often people who had often just had their father’s corpse placed across their laps, asked “well, how would you feel?” Not wanting to give anyone the idea that we foreigners were lacking in normal human sentiments, I hurriedly assured her that just about anyone in the world would probably have much the same reaction, were they to find themselves in a similar situation. Only later did it occur to me that I could have added that that is precisely the reason the rest of us never put people’s dead fathers on their laps in the first place. If they do in Imerina, it can only be because the memory of the living individual—or at least of some of them—remains so powerful and so persuasive a presence in the minds of their descendants that only a confrontation as dramatic as this can really bring home to them the fact that he’s dead.

That famadihana are largely concerned with transforming the memories of the living is a point that’s already been made by Maurice Bloch (1971: 168–169). This, he adds, makes them quite different from the secondary burials made famous by Robert Hertz (1907), which are primarily concerned with freeing the souls of the dead person from its lingering existence halfway between this world and the next. But, in a way, the two are not so very different: here too the dead could be said linger on in a kind of suspended half-life in the memories of their contemporaries.

In the ritual, it was women’s memories that were most prominently brought into play, while the ancestors who are the real emotional focus of the ritual were almost always male.[125] This was quite in keeping with the emotional bonds that were felt to—and often clearly did—exist between fathers and daughters (as also between mothers and sons). In fact, these attachments colored both women’s and men’s attitudes towards ancestors more generally. One older woman from Betafo told me, for instance, that if she dreamed of her mother, it was always to warn of ill news, while if some particularly fortunate event was about to occur, her grandfather would appear instead. The same pattern—female ancestors boding ill, and males boding well—appeared constantly in accounts of women’s dreams. In one of the more dramatic ones I heard, a woman said to have been neglecting her children saw her mother’s sisters appear to her inside the tomb to say that, if she didn’t change her ways, they would soon be taking her to join them. On the other hand, Rakotondrazaka’s daughter Irina—the woman who had broken down in tears over his body during the famadihana described above—later told me that her father regularly appeared in her dreams as a kind of guardian spirit, protecting her from danger and giving her advice.

This latter was unusual. While astrologers or magicians almost always claimed to have some “ancestral” adviser who appeared to them in dreams or visions, it was rarely said to be their own ancestor—and, even in those few cases where it was, never someone they had known personally. Irina was not an astrologer or magician of any sort, but she had been very close to her father while he was alive. Her father’s favorite child, and only daughter, she had never married or left the village in which she was born, and had seven children (all by different men) that her father had helped her take care of. And, while I know a good deal less about men’s dreams than about women’s—since men were less inclined to tell me about such things—my impression is that the terms were typically reversed. Fathers appeared mainly to chide their sons when they’d been quarreling with each other or had otherwise strayed. This was certainly true of Rakotondrazaka, anyway: Irina told me that aside from being her personal guardian, her father had also made a deathbed promise to his six sons that he would continue, even after his death, to counsel and admonish them when there was a quarrel in the family.

A remarkable feature of dreams about people recently dead was the way that images of real living people become mixed up with images of death. This was true whatever the reason for their appearance. Frequently, as in the following dream, one reported to me by Irina, they appeared in or near their tombs; or the images recounted—particularly when they were chiding the living—shifted back and forth between those of living human beings and those of frightening corpses.

I dreamed that I saw my father in 1989 (this was when he was already dead) to the north of Ambodivona. There were some trees and we were talking among them [and I asked myself] “is this daddy here, not yet dead?”

Then, “give me your blessing,” I said (because I wasn’t well)... So we were talking, when he said “you shouldn’t do such things, Irina,” and, right there, he plunged back into being dead and bound. Later, I went up to the village where my older brother was; and he too just took off out of the village, and plunged into death like that. It was, like, disgusting—and frightening![126]

On first seeing her father, she wonders whether he isn’t really alive. She asks for his blessing—but then he suddenly chides her, and turns into a corpse. The image changes from that of a living person to an ancestor, bound head and foot by the ropes used to fasten on his shrouds. What happens in Irina’s dream—she confronts a vivid memory of an ancestor, asks for his blessing, and then suddenly sees him transformed into a dead, bound corpse—is just what happens to women in famadihana: except that in famadihana, of course, it’s living men who bring about the confrontation by calling out the names of the dead, and thus evoking memories of living persons, before placing those persons’ decomposed bodies on the women’s laps.[127]

The memories evoked by names are tied to physical objects—objects which, as I’ve already described, are then gradually dissolved at the same time as the names themselves are gradually forgotten. The process as a whole can be thought of as one of effacing the individual identities of the dead—or, of all but the very small number who are or will become a tomb’s razambe.

I’ve already described this process as a kind of active form of genealogical amnesia, in which the living begin to combine the remains of ancestors about to pass from memory with others whose names thought more likely to endure. But few endure for long. In examining the names that were actually called out during famadihana at particular tombs, I found that, aside from one or two razambe, almost all of them were those of people who had died within the last ten or fifteen years. In other words, most names continued to be memorialized only so long as memories of the bearers themselves were likely to remain vivid in the minds of any number of the living; or, perhaps more to the point, as long as the social relations to which those memories relate still have some reality in people’s daily lives.

Names like that of Rakotondrazaka however remain enormously important—so much so that local society can be said to be largely organized around them. Local families continued to be referred to by the names of their founders as long as they hold together and, as I’ve said, these same names were regularly invoked when talking about the ownership of houses, rice fields, and tombs long after their bearers had died. The expression most often used to refer to ancestors on one’s father’s side literally means “name of the father” (anaran-dray). It was also used to refer to what might be called “ancestral property”: houses, tombs and rice fields passed on through the male line (as was anaran-dreny, or “name of the mother” for the female line). A number of scholars have remarked on the oddness of this expression, since Malagasy society does not use patronymics or, for that matter, matronymics of any kind (Razafintsalama 1981; Gueunier 1982: 237n2). Why then should the most important elements of one’s inheritance be identified with one of the few aspects of a father or mother’s social identity that was not inherited?

It seems to me that, by using this expression, one underlines the fact that such property does not entirely belong to the person holding it. Sometimes, this is quite literally true: if a group of brothers and sisters postpone the formal division of their parents’ property, land and houses can remain for years legally registered in the dead ancestor’s name. In fact, I was told that one reason why children might decide to hold a famadihana in honor of the founder of their local family was to ask for their “blessing”—or tsodrano—before dividing up such a joint estate. Even when descendants do hold legal title, possession is not without its obligations, because if one holds a rice field inherited from a given ancestor, then one is responsible for providing lambamena and otherwise contributing to the expenses whenever that ancestor is involved in famadihana—an obligation which endures as long as does the memory of the ancestor.[128] But, here again, the logic of the ritual leads back to the theme of the dissolution of identity: several people told me that the reason why it was necessary to combine razana together was to keep such expenses down.

I certainly never heard anyone put it to me quite this way, but one might think of famadihana as a process of transferring ancestral names from an attachment to property to an attachment to stones. Standing stones have always been the archetypal form of memorial in Malagasy culture.[129] In a sense, tombs were themselves memorial stones; in former times they were always crowned by a stela which was said to stand directly over the head of the razambe (then called the tompon’ny fasana or “owner of the tomb” [Jully 1896]) and which received any sacrifices offered to him. In contemporary tombs, the stele have become crosses, but the implication and positioning remain the same. The stone, in effect, represents the tombs as a whole, and both are ultimately to be identified with a single ancestor, whose name in turn would be attached only to it, and not to any property shared by living people.

Famadihana as Reversal

The difference between men’s and women’s attitudes, at any rate, would explain the very different roles they’re given in the ritual—particularly at the critical moment when the confrontation between ancestral bodies and human memories takes place. Women carry ancestors on their laps. The expression used for this is miampofo, which literally means “to nurse a child sitting on one’s lap,” and the candy, honey, trifling sums of money, and so on are just the sort of thing one gives as treats to small children. Even the fact that the zanadrazana clothe the ancestors, and carry them rolled like infants in a blanket-like lamba, could be seen in a way treating them like children—which, assuming the ancestors are here being thought of as symbolically male, reverses the relation between fathers and daughters, turning it into a relationship between mothers and sons.[130]

The men’s part, on the other hand, is to carry the ancestors, to wrap them, to bind them, and to lead the dancing with which they are returned to the tomb at the end of the ceremony. In effect, what this means is that it’s the male role to destroy them, since it’s the combination these actions—none of which are carried out at all gingerly—which result in the dry body being broken apart and turned to dust. One woman told me this was the reason it was men who have to bind the ancestors: the binding has to be done with such “outrageous” (mahatsiravana) force that only men are strong enough to do it.

The word usually used for “wrapping the ancestors,” mamono razana, sounds suspiciously like the phrase for “attacking” or “killing” them.[131] The word famadihana itself, for that matter, can also mean both “reversal” and “betrayal.” Admittedly, I never heard any participant remark on the parallel. But it could certainly be argued that the male role in the ceremonies involves a reversal of roles which goes even beyond that of the living attacking the dead. What is being inflicted on the ancestors is precisely what the ancestors inflict on the living: a form of constraint continuous with a form of violence. This is perfectly summed up in the act of binding the bodies—each cord is yanked so forcefully that the very bones are crushed. There’s also a particular emphasis on the politics of movement. Just as any father or grandfather would strive to keep his male descendants from moving away, so the process of the famadihana is largely one of containing the dead ancestors in space: after being called to return from their wanderings to the tomb, at the start of the ritual, they’re removed, bound tightly with ropes, and locked back into the tomb with magic charms.

Ancestral Blessings

I have been arguing so far that rural society in Imerina was largely organized around the identities of a handful of prominent elders who had succeeded in assembling descendants around them, or at least in keeping them from moving away. The memory of such elders tends to retain enormous social force long after they themselves have died—so much so that to overcome it requires a ritual of profound trauma and violence, in which the relation between ancestors and descendants is turned completely on its head. By transforming their dead ancestors into children, the living can turn back on them the very forms of constraint and violence that constitute ancestral authority and, in doing so set off a process by which the memory of the ancestors themselves will be largely effaced.

This is not an interpretation a participant would be likely to offer, or even to agree with. When discussing famadihana in the abstract, almost everyone tended to avoid references to violence, and instead lay great stress on the tsodrano, or blessing that ancestors convey to their descendants. Usually, they would echo the themes of famadihana orations: that the living wish to honor the dead and so secure their blessing—a blessing which will ensure the continued health, prosperity, and fertility of themselves and their families. While older men and figures of authority were particularly inclined to emphasize these themes, this was a notion familiar to everyone: the formal expression meaning “to ask for a tsodrano” (mangataka tsodrano sy ranombavaka) was the one piece of ritual language even the most ignorant person was guaranteed to know, and the term was constantly invoked in ritual contexts, or in any other context in which a certain formality of speech was felt appropriate.

The notion that the ancestors remembered in famadihana provide positive benefits for their descendants appears, on the face of it, to be in complete contradiction with my own interpretation. But, on closer examination, one finds these positive benefits are very hard to pin down. The “health, prosperity, and fertility” provided by the ancestors is only of the most abstract and unspecific kind. Nobody ever sponsored a famadihana in order to cure someone who was ill, to bring success to some financial project, or cause someone infertile to conceive. In any of these situations, one might make a vow at the tomb of an ancient king or Vazimba spirit, or one might consult a magical specialist of one kind or another—and I don’t think there was anyone I knew in Madagascar who hadn’t done at least one of these things at some time or another—but no one would consider appealing to their own ancestors. Unless, perhaps they thought that their ancestors had been responsible for causing the problem to begin with.

It’s true that some people would occasionally “ask for a tsodrano” from their ancestors by placing small offerings of rum, candies, or honey on the roof of their tomb, accompanying the gesture with a prayer (the offerings are mainly the same as the “tokens of request for tsodrano” given ancestors at famadihana). This appears to have been a common practice: at least, there are almost always one or two empty bottles or the remains of other offerings to be seen on the tombs of deme founders; and occasionally similar offerings on less prominent tombs as well. But it’s hard to say exactly who did this and why, since with one exception, I never found anyone willing to admit to ever having done it themselves. This was in itself unusual: I rarely ran into anyone reluctant to talk about, say, offerings she had made at the shrine of some ancient king, or the rituals performed in consulting an astrologer or spirit medium.

The one man who did admit to having made one was something of a social pariah, notorious for having offended his razambe by violating a number of ancestral taboos. He was said to have fallen into abject poverty and debt as a result. One night, while drunkenly celebrating an unexpected windfall, he declared to his neighbors that he had appealed to this same razambe for relief from his debts, and that his prayers had been answered. It was clear to everyone that his real motive was to broadcast as far as possible that the ancestor had forgiven him. They were not convinced. I strongly suspect that, in most, if not all, cases where people left offerings on the tops of tombs “to ask for their ancestor’s tsodrano,” what they were really doing was appealing to them for relief from some punishment which those ancestors had themselves inflicted. At any rate this would explain their reluctance to admit having done so.[132]

The word tsodrano literally means “to blow water.” At its simplest it refers to a domestic ritual in which a child or younger person requests his elder’s blessing, and the latter responds by sprinkling him with water, usually adding a few words of benediction, which, using a relatively conventionalized language, wish good health, prosperity, and many descendants on the person being blessed.

There are two very important points to be made here. The first is that elders never give such blessings on their own initiative. A tsodrano must always be requested. In the past, I was told, children had to “buy” their parents’ blessing by presenting a coin or small piece of money to them as a token of request. The giving of small change and other gifts to ancestors as “tokens of request” would seem to echo this same ritual logic.[133]

The second point is that the effect a blessing has on its recipient is the precise opposite of that of cursing or ozona. By cursing, parents impose taboos and restrictions on their descendants. By “blessing” they remove them. In one village, for instance, I heard that the local elders gave such a blessing after a number of teenagers who were studying in Antananarivo approached them complaining it was impossible to maintain their deme’s fady on pork while living in the city. The elders blew water over them, so freeing the whole deme from the taboo. In fact, in almost every context in which I heard of someone asking for a tsodrano, giving it could be construed as releasing the recipient from some constraint or restriction to which the giver would otherwise have had the right to hold them. The archetypical example was that of a young man leaving home, whether to pursue his education or simply “look for money.” Such a person, I was told, will always go to his parents and ask for their blessing, particularly if he is leaving the country or going very far away. In common speech, one can say that two lovers have “blown water over one another” (mifampitsodrano) if, on parting temporarily, they agree that each is free to see other people until they are reunited. Shortly before leaving Madagascar in December 1990, just as the war in Kuwait was heating up, I heard on the radio news that “the American Congress has given President Bush their tsodrano to use force in the Persian Gulf.”

More elaborate rituals were often organized in terms of requests for tsodrano. This was true especially of the rhetorical contests which surrounded the payment of the vody ondry (a kind of symbolic bridewealth—cf. Bloch 1971: 175–205, 1978; Keenan 1973). In all the examples I myself witnessed, the speechmakers treated the entire affair as a request, by the boy’s family, for the girl’s family’s tsodrano. The theme was repeated over and over in their speeches. Even the money which the suitors present was referred to as a token of respect, given in way of requesting a tsodrano. In fact, the payment itself is often divided into a large number of small payments, each named after some task that a woman would normally be expected to perform in her parents’ household—maka kitay (gathering firewood), tsaka rano (fetching water), alam-bolofotsy (plucking out her mother’s white hairs), and so on—all of which clearly imply that the money is at least symbolic compensation for the services the daughter would have provided her aged parents were she to have remained at home. After the money has finally been accepted and the woman’s parents have formally agreed to the match, the latter actually did blow water over the couple, adding some conventional words of advice and wishing them seven male and seven female children. It seems clear to me that it was this act of tsodrano which really effected the change of status of the woman: by giving it, her parents release their rights in her, or more precisely the constraints their authority as parents allows them to place on her freedom of action and of movement. In effect, it parallels the tsodrano a boy’s parents give him before he leaves home to seek his fortune: both are a release from the obligations and constraints of parental authority.[134]

The only other occasion I know of, aside from famadihana and the tomb ritual, when anyone was said to request a tsodrano from the dead was a ritual said to be performed privately by a widow who wishes to remarry. She has to “ask for her husband’s tsodrano” before being free to do so. This she does by approaching his tomb carrying two stones. One, a piece of quartz, is called “stone of the living”; the other, granite, is called the “stone of the dead.” The ceremony itself is simple—she throws the dead stone at the tomb, and carries the living one home—but one can see it too as a capsule famadihana, at least in so far as it involves the same combination of violence and request for release.

A War Against Death?

What I am arguing then is that, since there is no clear line between positive benefits and the benefits of simply being left alone, the notion of tsodrano can be used as a kind of euphemism. This became particularly clear when, instead of asking what you accomplished by performing famadihana, I asked what would happen if you didn’t perform them at all. While answers to the first question was usually preceded by a good deal of reflection and casting about for the right words, the second response was instantaneous: your children will die. Or you will fall desperately ill. Or you and your family will fall deeper and deeper into poverty. The catalog of misfortunes could, admittedly, be seen as simply a negative image of the fertility, health, and prosperity tsodrano was said to bring; but since people were always much more concrete and specific in speaking of the misfortunes than they were of the benefits, it would make better sense, I think, to look at it the other way around.

The danger of ancestors coming to kill a family’s infant children was, in fact, a constant concern. Ghosts (lolo, angatra, matoatoa) were said to linger around tombs and anyone unwise enough to come in too close contact with a tomb in ordinary circumstances should light a small fire in the doorway of their house and enter by stepping over it lest a ghost follow them inside. The same thing is done after attending funerals. There are any number of customs having to do with placement and maintenance of tombs which are explicitly concerned with keeping the dead from having access to the living—and the sure way of knowing that one has failed to maintain the separation is that the young children in one’s family begin to die. Most of the people I knew could tell stories about waking in the middle of the night because they (or someone in the same room) were in the middle of being strangled by some malevolent ghost—which, when they appear in one’s sleep, are typically characterized by their naked, black forms and huge size—and any marketplace would be sure to contain two or three vendors selling charms aimed at keeping ghosts away or getting rid of them.[135] These ghosts were anonymous, generic beings and contrasted in this with individualized, “good” ancestors who, when they appeared in dreams and visions, were usually robed in white. But even such relatively benevolent ancestors were, to say the least, troublesome: one of the most frequent reasons for their appearance was to complain of being cold and demand that their descendants perform famadihana—and I’ve already mentioned what is considered likely to occur if they are not satisfied with the results. When asked about the origins of the dark, murderous specters that disturbed children’s sleep or otherwise plagued the living, most people immediately suggested they were ancestors whose descendants no longer “took care of them.”

Since some would say that it was most often the recently dead who demanded famadihana, one might be tempted to look to Hertz’s secondary burials once again for a parallel. In the societies he discussed, the vindictive ghosts of the recently dead were believed to linger near their old habitations; and the ritual served to release them into another world where they would be harmless to the living. Famadihana could be thought of as doing something similar: dissolving away the identities of the dangerous, recent dead, so they could ultimately be absorbed into that of a relatively benevolent razambe. But, as the example which began this essay makes abundantly clear, razambe are not necessarily all that benevolent.

One married couple from Betafo—who had, in fact, just earlier told me their own version of the story about the fire in 1931—mentioned that, after the most recent famadihana for Andrianambololona’s someone broke into his tomb and stole several expensive lambamena that, having been bought for the ceremony but never used, had been left behind inside it. “That’s odd,” I said. “You’d think a thief would be afraid to enter such a tomb.” “Well this one must not have been.” “But he’s supposed to be so powerful and fearsome! Isn’t this the same one who burned down the town?” “Well,” they both replied, more or less at once, “he wasn’t cold any more, was he? If he starts appearing to you, it can only be because he’s cold.” But, in this case, there had just been a famadihana. He’d just been wrapped; he wasn’t cold at all, and unless he was, the husband added, “he’s really nothing but a pile of dust.”

Heat did play an important role in the symbolism of famadihana. Honey, rum, cow fat, ginger, and even candies, all of which are prominent among the “tokens of requests for tsodrano” given to the ancestors, are also things one eats when one has a cold—precisely because they are considered food with heating properties, that can relieve the coldness in one’s head or chest responsible for coughing or congestion. In fact, these gifts were supposed to be placed roughly where the ancestor’s head and chest ought to have been.

Fire too had a complexly ambiguous relation with the dead. Ghosts were frightened by it. Everyone knew that, if in danger of being accosted by a ghost, the best thing to do was to light a match: a flashlight I was told wouldn’t do, because it isn’t light ghosts fear, but actual flames. I’ve already mentioned that stepping over a candle or other flame when entering a house prevents ghosts from following one in. Charms to drive away ghosts almost always involve heat and flames: most involved incense. But, at the same time, I heard people insist that one had to carry a candle or lantern—again, a flashlight would not do—when descending into a tomb to fetch the dead during a famadihana; and it was common practice to burn candles at the tombs of ancient kings or other benevolent spirits, or while invoking them elsewhere.

A friend of mine called Ramose Parson, a biology teacher at the Catholic secondary school in Arivonimamo, told me that he always thought of the practice of famadihana as being basically the same as cremation, except carried out over a much longer period of time. Cremated bodies are reduced to dust through the application of heat; afterwards, the dust is encased in an urn which ensures it never mixes with the surrounding earth. All of which, he pointed out, is also the case in Malagasy mortuary ritual—the place of the urn here being taken by the lambamena, which is valued for its hardness and durability, and by the care people take to ensure the ancestral bundles never come in contact with the earth. This is of course one man’s theory, and a rather eccentric one at that, but, if nothing else, it would make the story with which I began this essay all the more poetically appropriate: by forgetting to carry out the famadihana in its entirety, the hapless descendants of Andrianambololona ended up bringing the destructive fire on themselves instead.

Some Conclusions

It can also be interpreted to mean that remembering and forgetting are equally matters of violence; that it is only the direction of the violence that varies between the two. This is, of course, the argument I have tried to develop over the course of this essay. The ancestors whose enduring memories give shape to social groups—whether these be recent Ray aman-dReny or ancient razambe like Andrianambololona—do so in practical terms mainly by their power to constrain and punish their descendants, by ancestral violence; while famadihana, seen as the highest expression of group unity, were occasions on which descendants could turn a form of violence precisely modeled on that of their ancestors against them, and by doing so gradually blot those memories away.

In Imerina, the rather commonplace dynamics by which genealogies are made and transformed—ones which, it has been clear since Evans-Pritchard’s work on the Nuer, require a continual process of forgetting people’s names—are transformed into a veritable struggle for existence between the living and the dead. This was true in the most literal sense. The dead, as a Malagasy proverb puts it “wish to become more numerous”—by murdering the living; the living respond by crushing and consolidating the bodies of the dead so as to keep their numbers low.

This is not to say that famadihana were not also memorials to the dead, just as participants said they were. The memory of ancestors was in its essence double edged: particularly so from the point of view of the most important men in rural society, who wield an authority and fame largely borrowed from ancestors who are ultimately their rivals—as well as being people they knew and cared for while they were alive. As I’ve said, the contradictions of their position often seem to put such men in a position of wanting to deny the existence of such violence altogether. For all that, in describing the moral unity of the community that ancestors create they are in effect speaking of the effects of that violence itself. Women, whose position in relation to ancestors is very different, though equally complex, feel much more comfortable talking about such matters, but even they did not really know how to reconcile the “cruelty” of which ancestors were capable when enforcing moral principles and the sheer egotistical violence of ancestors who simply wanted to be remembered for their own sake. It was presumably this latter dilemma—itself a transformation of the same central contradiction—which caused images of the dead, in effect, to split in two: between the one idea of benevolent elders who bring their descendants together in a moral community, and the other of rapacious ghosts who carry their descendants’ children off to join them in the tomb.


Astuti, Rita

1995 People of the Sea: Identity and Descent Among the Vezo of Madagascar. Cambridge: Cambridge University Press.

Bloch, Maurice

1971 Placing the Dead: Tombs, Ancestral Villages, and Kinship Organization in Madagascar. London: Seminar Press.

1982 “Death, Women and Power.” In Death and the Regeneration of Life (M. Bloch and J. Parry eds.). Cambridge: Cambridge University Press, 211–230.

1985 “Almost Eating the Ancestors.” Man (n.s.) 20: 631–646.

1986 From Blessing to Violence: History and Ideology in the Circumcision Ritual of the Merina of Madagascar. Cambridge: Cambridge University Press.

Callet, R. P.

1908 Tantara ny Andriana eto Madagascar. Tananarive: Académie Malgache.

Condominas, Gerard

1960 Fokonolona et Collectivités Rurales en Imerina. Paris: Berger-Levrault.

Cousins, William E.

1963 Fomba Gasy. (H. Randzavola, ed.) Tananarive: Imarivolanitra.

Dahle, Lars

1984 Anganon’ny Ntaolo, Tantara Mampiseho ny Fombandrazana sy ny Finoana Sasany Nanganany. (L. Sims, ed., original edition 1878). Antananarivo: Trano Printy Loterana.

Evans-Pritchard, Edward E.

1940 The Nuer: A Description of the Modes of Livelihood and Political Institutions of a Nilotic People. Oxford: Oxford University Press.

Feeley-Harnik, Gillian

1989 “Cloth and the Creation of Ancestors in Madagascar.” In Cloth and Human Experience (J. Schneider and A. B.Weiner, eds). Washington, DC: Smithsonian Institution Press.

1991 A Green Estate: Restoring Independence in Madagascar. Washington, DC: Smithsonian Institution Press.

Gueunier, Noël Jacques

1982 “Review of Le Tsimahafotsy d’Ambohimanga, by Adolphe Razafintsalama.” Asie du Sud Est et Monde Insulindien 13: 235–241.

Haile, John

1891 “Famadihana, a Malagasy burial custom.” Antananarivo Annual and Malagasy Magazine 16: 406–416.

Hertz, Robert

1907 “Contribution à une étude sur la représentation de la mort.” L’Année Sociologique 10: 48–137.

Jully, Antoine

1894 “Funerailles, tombeaux et honneurs rendus aux morts à Madagascar.” Anthropologie 5: 385–401.

Kottak, Conrad Phillip

1986 “Kinship Modeling: Adaptation, Fosterage, and Fictive Kinship among the Betsileo.” In Madagascar: Society and History (C. P. Kottack, J.-A. Rakotoarisoa, A. Southall and P. Vérin, eds). Durham: Carolina Academic Press.

Lambek, Michael

1992 “Taboo as Cultural Practice among Malagasy Speakers.” Man 27: 245–266.

Raison-Jourde, Françoise

1991 Bible et pouvoir à Madagascar au XIXe siècle. Paris: Karthala.

Rasamimanana, Joseph and Louis Razafindrazaka

1957 Contribution a l’histoire des Malgaches: Ny Andriantompokoindrindra. Tananarive: Volamahitsy.

Razafindratovo, Janine

1980 “Noms Passés, Noms Presents chez les Merina.” Asie du Sud Est et Monde Insulindien 11: 169–182.

Razafintsalama, Adolphe

1981 Les Tsimahafotsy d’Ambohimanga. Organization familiale et sociale en Imerina (Madagascar). Paris: SELAF.

Ruud, Jørgen

1960 Taboo: A Study of Malagasy Beliefs and Customs. New York: Humanities Press.

Southall, Aidan

1971 “Ideology and Group Composition in Madagascar.” American Anthropologist 73: 144–164.

1986 “Common Themes in Malagasy Culture. In Madagascar: Society and History (C. P. Kottak, J.-A. Rakotoarisoa, A. Southall and P. Vérin, eds.) Durham: Carolina Academic Press.

Standing, Henry L.

1883 “Malagasy ‘Fady.’” Antananarivo Annual and Malagasy Magazine 8: 253–265.

Van Gennep, Arnold

1904 Tabou et Totemism à Madagascar. Paris: Ernest Leroux.

Vogel, Claude

1982 Les quaters-méres d’Ambohibao: étude d’une population régionale d’Imerina (Madagascar). Paris: SELAF.

Wilson, Peter

1977 “The Problem with Primitive Folk.” Natural History 81: 26–35.

1991 Freedom by a Hair’s Breadth: Tsimihety in Madagascar. Ann Arbor: University of Michigan Press.

From :

Chronology :

January 07, 2021 : Part 2, Chapter 6 -- Added.
January 17, 2022 : Part 2, Chapter 6 -- Updated.

HTML file generated from :