Fanany: people who come back as snakes
In this essay I would like to talk about people who lost everything. Imerina (the traditional name for the northern half of the central plateau of Madagascar) is a place where people attach enormous importance to the memory of their ancestors and the lands on which their ancestors once lived. History, in Imerina, is largely a matter of placing the living in an historical landscape created by the dead. About a third of the Merina population, however, is made up of the descendants of slaves, and in Madagascar, slaves are by definition people without ancestors, ‘lost people’ (obna very) who have been ripped from their ancestral landscapes, left unanchored to any place. These were people who had been literally stripped of history. Even today, almost a hundred years after emancipation, most ‘black people’ (as their descendants are called) remain in a kind of historical limbo, unable to make a real claim to the territories in which they live and are buried.
The question I want to ask is: what forms does historical memory take for such people? What forms can it take?
It has become a commonplace, nowadays, to argue that historical consciousness is ultimately about identity. Memories of the past are ways of defining who one is in the present—and perhaps too, of defining what kind of action one is capable of, of enunciating collective projects (e.g., Connerton 1989; Friedman 1992). Clearly, this would leave the historical consciousness of slaves more than a little problematic. Slaves’ identities were created by events their descendants would not wish to commemorate, events which not only annihilated any link to their previous histories, but left their victims generically incapable of producing new ones. Not surprising then that most descendants tried their best to avoid having to admit to their ancestry. It was embarrassing. Almost all the stories I did manage to cull about the ‘days of slavery’ centered on the insidious means masters used to ritually pollute their slaves—rubbing excrement on their heads, making them sleep alongside pigs— and so destroy their hasina, a word whose meaning in this context falls about halfway between ‘state of grace’ and ‘power.’ The ultimate message was often quite explicit: it was only by destroying their ability to act for themselves that masters were able to keep slaves in subjugation. But it is telling that these were just about the only memories of slavery I ever heard recounted. It was as if, having explained how slaves were rendered people who did not have the right to act, or even to speak, for themselves, there was nothing left to say.
The experience of slavery could not be directly told, as history, if only because admitting to such a past deprived one of the authority with which to speak. It was inherently shameful. Of course, the feeling that one is not entided to have an opinion, or a history, is a common phenomenon among the dispossessed of any society (Bourdieu 1984). However, what I am going to argue in this paper is that Merina slaves did, in fact, develop a ritual idiom with which to reflect on their history and their condition, and even to speak to others with the voice of authority. It was, perhaps, somewhat veiled and indirect. But here too the ability to speak was inseparable from the ability to act; it was through the very process of seizing the authority to speak that the descendants of slaves, in so many cases, began to take back for themselves the capacity to act as historical agents in their own right, actors as well as narrators, and so perhaps to begin to recuperate a little bit of what they had lost.
From around 1820 to sometime around 1850, hundreds of thousands of people were taken prisoner by Merina military expeditions and carried back to the central highlands of Imerina. The Merina kingdom’s army, armed and trained by its British allies, made common practice of massacring all the adult males of ‘rebel’ villages, and carrying off everybody else to be sold as slaves. Sometime around 1855, Queen Ranavalona I’s secretary Raombana wrote of these campaigns:
As to the miseries which these continual wars brought to the provincial people, that is indescribable, for by fighting, but more deceit, that is, the offer of life and pardon if they yield and submit, thousands and thousands have thus been murdered in cold blood, and their numerous wives, children and cattle seized and reduced to slavery.
Mothers are separated from their tender offspring and other relations, as they are divided and distributed amongst different masters, and are thus taken into different parts of the country where they never discover one another again, with very few exceptions.
An officer who is a real Christian informed me that the pains in hell could not be more than the pains suffered by these unfortunate people in being separated from one another to be taken away to their different masters. Their cries, their weepings and their lamentations, said the above Christian, is such as almost sufficient to raise the dead from their graves for to take their parts....
It is important to remember that for its victims, the very first thing slavery meant was a complete rupture with everything that had made their lives meaningful: of all the ties of love, kinship, shared experience that had bound them to a home, to parents, friends, lovers, to everyone and everything they had most cared for. It was in this sense too that slaves were ‘lost people,’ alone, in an alien place among people who did not know them.
It is even more important, perhaps, to emphasize that this situation did not end once a captive had been sold and settled in Imerina. It was extremely difficult for slaves to create any kind of enduring ties, either to people or to places. Free Merina lived in permanent towns or villages on hilltops; slaves, in makeshift settlements in the valley bottoms, near the paddy fields. These hamlets were mostly structured around a handful of older men or women; most younger slaves had no fixed abode at all but circulated between several hamlets in different parts of Imerina, often between different masters, as well as between scattered family, friends, and lovers. Many slaves managed to win a remarkable degree of independence from their masters; they came to make up the majority of both Imerina’s petty merchants and its petty criminals, as well as almost all wage laborers; but the mobility which made this possible also ensured their uprooted condition remained permanent.
The most obvious symptom of slaves’ placelessness was their lack of proper tombs. For free Merina, tombs were—as they remain today— the ultimate link between people and places, the anchors of group identities (cf. Bloch, 1971). It is through collective stone tombs that each descent group is fixed to an ancestral territory; one may spend one’s entire life far from one’s ancestral lands, but one nonetheless expects to be buried on them when one dies. To be buried in a magnificent tomb, to be remembered there by one’s descendants, and to be periodically rewrapped in beautiful silk shrouds called lambammay is the ultimate aspiration of any important man. But the ideal is hard to achieve; to guarantee one will not be forgotten, one has to acquire enough land to settle a large number of descendants around one’s tomb. For slaves, all this was nearly inconceivable. Most slaves were not buried in proper stone tombs at all—many were buried in simple graves or improvised pit-tombs near their settlements, usually with people of very different ancestries buried together. It was only after emancipation (which came in 1896, a year after the French conquest of Madagascar) that most of those former slaves who chose to remain in Imerina began to create solid and substantial tombs of stone like those of free people. Usually, a number of families would have to pool their resources to do so. To acquire the land to keep one’s descendants around was even more difficult. Almost all ‘white’ Merina claim their ancestors forbid them to marry or sell their ancestral land to the descendants of slaves; when they make exceptions, it is only for vastly inflated prices. In the village of Betafo, where I did my fieldwork, for example, almost every ‘black’ (mainty) family that had managed to establish itself over time had only been able to do so because of some exceptional windfall: a gift of land from former masters (for whom they were also obliged to work as sharecroppers), a military pension, an unusually advantageous marriage.... The majority still spent their lives traveling between places which were not really theirs; for most, lives of endless striving ended only in failure and oblivion, to be buried in the tomb of a distant relative in a land where they had likely never lived.
There is an irony here, because the very difficulty of creating tombs and ancestors meant they came to take on a very different meaning for the descendants of slaves than for the other two thirds of the Merina population. As I have argued elsewhere (Graeber 1995), attitudes of ‘white’ Merina towards ancestors are profoundly ambivalent. People do wish to be remembered as ancestors when they die; but in part for that very reason, the memory of existing ancestors is seen as an imposition on the living, supported by the constant threat of punishment for lapses of memory or neglect of ancestral restrictions. Memory itself is felt as a kind of violence. As a result, the famadihana rituals in which the bodies of the dead are periodically removed from the tomb to be rewrapped in new lambamena, have a dual meaning too. While represented as ways of remembering the dead, their covert purpose is to make it easier to forget them: reducing ancestral bodies to dust so their names can be forgotten; then, locking them inside the tomb. For the descendants of slaves, on the other hand, it was not the pressure of history and memory that was felt as a kind of violence, but the very lack of it, and for that reason, ancestors took on a far more benevolent countenance. It is hard to be certain, but I did find that descendants of slaves were much more likely to insist that ancestors really did provide concrete benefits for their descendants, and I had a strong feeling that, while the form of mortuary ritual was the same, the content was slightly different; that there was an honest piety in ‘black’ attitudes towards the dead often lacking in their ‘white’ neighbors.
The first story I have to recount, in fact, is largely about how difficult it is for ‘black people’ (olona mainty) to lay claims to Merina mortuary ritual.
The story of Rainitaba
The community of Betafo, the focus of my own fieldwork which I conducted between 1989 and 1991, consisted of something like 13 settlements and perhaps 400 people, occupying a stretch of rolling country about forty minutes’ walk to the north of the town of Arivonimamo. It was a community divided between the descendants of an andriana, ‘noble,’ descent group, and the descendants of their former slaves. The latter made up about a third of the total population.
The main reason they were there was because shortly after emancipation, members of the wealthiest andriana families had given their former slaves a large grant of land to encourage them to stay on. Almost all of it went to two men, who were, at the time, the chief men of a hamlet called Antandrokomby which sat, like most slave settlements, on the edge of a stretch of terraced rice fields. Both men built substantial tombs and kept large numbers of descendants around them; in fact, most ‘black people’ in Betafo were descended from one of them in some way or another. Betafo was also a somewhat anomalous community because on the whole, its ‘black people’ were doing rather well: partly because with their greater acclimatization to wage labor, they had been better able to adapt to the economic crisis that had hit rural Madagascar since the late 1970s; also, because some of the families that had not originally received grants of land had been able to capitalize on their reputations as astrologers and magicians to acquire enough money to buy land from their increasingly impoverished andriana neighbors. On the whole, in fact, Betafo’s mainty were, by the late ’80s, doing rather better than the andriana) certainly that was the andriana perception of the matter. But there were some mainty families who held on more tenuously. One of the most famous was made up of the descendants of a certain Rainitaba, a man who had, apparently, also lived in Antandrokomby, but who had already died, leaving behind only a single daughter, before the land was distributed in 1896.
It had never been a prosperous family, and most people I knew in Betafo were convinced that it had entirely died out. Nonetheless, everyone had heard Rainitaba’s story.
The story goes like this:
When Rainitaba died, his children wrapped his body and buried it in the tomb, but when they opened the tomb to rewrap him, the body had disappeared. Only an empty cloth remained. Later, Rainitaba appeared in a dream to his daughter and told her he had taken the form of a snake and abandoned the tomb, becoming a Vazimba spirit in the waters to the east of the village of Betafo.
Some would repeat the story with amusement, others with a trace of scandal. But almost always, the narrator would then go on to point out the course of Rainitaba’s meanderings since death. First he descended to the waters around the spring to the east of the village—a secluded spot, full of shade trees and quiet pools, where people used to be afraid to take children after night lest ghosts seize them and they waste away and die. Later, he followed the waters that flowed by the rice fields to the north of Betafo downstream until he reached the ruins of Antandrokomby—by then long since abandoned. Near the ruins was a reedy pool, and there he was supposed to have resided for some time, before again disappearing. No one was quite sure what happened to him since, but most were sure he was not in Betafo any more.
Now, there is every reason to believe that a man named Rainitaba did indeed live in Antandrokomby in the 1870s or ’80s; his daughter, Rabakomanga, was still living there with three sons of her own when she died (probably at the age of 45 or 50) in 1912. She was apparently the one who had the dream in which her father revealed that he had left his tomb, but, after that, almost ail of her descendants were said to have had dreams of him, and sometimes unrelated people, too. I was told Rainitaba would always appear whenever one of his descendants conceived a child, and would often give advice about how to ensure a healthy birth.
It was only after I had been working in Betafo for some time that I discovered that there were still several descendants of Rainitaba living in Betafo or, anyway, quite frequently around. The most notable was Razanamavo, an old, poor woman who spent most of her time doing odd jobs for her slightly better-off neighbors in the mainly quarter of Betafo, or seeking work in town. She didn’t really have a house in Betafo, but lived in an outbuilding—litde more than a shed really— which a man named Armand had given her as an act of charity. Most people in Betafo did not have any idea of her ancestry; ordinarily, most tended to forget she existed at all.
I first met Razanamavo in town. Armand—who was a good friend of mine—kept a room in the town of Arivonimamo, from which he conducted a small business selling bananas; people from Betafo often used to gather there, or drop by seeking news. Once I happened by while Razanamavo was visiting, and Armand’s wife, Nety—always helpful in tracking down bits of Betafo history for me—immediately seized the opportunity to see if we could get her to tell us something about her famous ancestor.
Actually, Nety had previously wondered whether Razanamavo would be willing to talk to me at all: ‘she might consider it embarrassing, having an ancestor who was a Vazimba.’ But at first she seemed quite happy to tell us. Rainitaba, she said, was originally a nobleman from Betsileo. ‘Back then, you know, people would be bought and sold’ she said. ‘And that was the origin of Rainitaba. He was a lost person.’ He had been captured and sold into slavery in Betafo. The fact that people referred to him as a Vazimba did not bother her. What bothered her, she told Nety, was that people said he had turned into an animal. ‘Rainitaba is not an animal,’ she insisted, ‘but a Vazimba—a person, a person like a Kalanoro. Haven’t you ever dreamed of him?’ Nety hadn’t. Well, Razanamavo said, many have; he used to appear to her grandmother and her father regularly. He had appeared to her, too, before the birth of her first child.
Razanamavo had an oddly distant manner of speaking, somehow absent; she crouched wrapped in a yellow cloth staring off into the courtyard as she talked, as if looking at something far away, or perhaps nothing at all, and never once gazed at the other people in the room, even when she was more or less speaking to them. In part she was probably exhausted from an afternoon at work; but her manner seemed to complement the content of her discourse, which was much more evocative and dreamy than the usual, matter-of-fact style of historical narration. After telling her story, and answering a few of Nety’s questions about her relatives, she seemed to just fade away, staring off as if so lost in thought that she didn’t even notice anyone was talking to her, until after a little while, we gave up and started talking about something else.
Razanamavo’. He, you know—our grandmother said that when he was about to die, he said: ‘I am about to die now, so take me to the north of the village, to the dam. And as for me’ he said, ‘don’t bury me in a tomb but just release me in the current of the river. And get a lamba arin-drano  he said, ‘like you would for burying a son-in-law.’
And they said: ‘maybe we won’t put you in the river, because we’d be embarrassed.’
‘No, don’t be embarrassed,’ he said, ‘because you’ll receive a great blessing if you do it that way.’
But they didn’t do it. They just buried him normally and left him there.
Someone: This was Rainitaba?
Razanamavo: A little hole like that they buried him in—there was no tomb, no entry to the west. Later they got ready to do a famadihana [to move him into a proper tomb], and looked for the cloth, and they set the cloth around him nicely.
‘If you do this thing, then a great blessing will come to you.’ Then the water flooded... he was dead. ‘I don’t like lambammd he said, ‘but lamba arin-drano, and...’ those were to be the mourning clothes. ‘So your children will never become poor, nor the generations of your descendants to come ...’ This is something they all dreamed, all of them absolutely. But we didn’t get the blessing because we didn’t do it. They put him in the center of the top shelf to the north of the tomb, and he still hasn’t been moved to this day.
Once, there were a good number of descendants, but there are few left any more.
She too told how he had moved from the pool to the east of the village to the one near Antandrokomby. She wasn’t sure where he had gone after that, but she suspected he had finally returned to his original home in Betsileo.
The ‘little hole’ she refers to was a temporary grave—it seems that when he died, his daughter simply buried him—as was often done by slaves—until such time as a group of slaves could pool enough money to create a proper tomb. The opportunity only arose around 1910, when several ‘black’ families got together to build a collective tomb on a hill overlooking Betafo to the northeast. Rabakomanga contributed some money to the effort, and when the tomb was done had her father’s body wrapped properly in cloth, and then transferred it to one of its most prominent shelves. It was after this, the first time they returned to perform a famadihana,, that they found the body gone.
At the time, Rabakomanga was by no means penniless; she had apparently received a small amount of land from her former owners, which she had passed on to her sons. Her sons apparendy were not able to hold on to much of it. What land there was was sold or mortgaged off. Those who remember her sons, Ingahivelona and Rakotonanahary, remember them as landless laborers, and desperately poor. In fact, the history of Rainitaba’s family was always represented as one of loss, poverty and dispersal. They never received the blessing that they were offered. They scattered; now they’re gone. In fact, it is one of the ironies of their history that it took me a long time to realize that the lineage had been really quite prolific: most of Rainitaba’s grandchildren had numerous sons and daughters. Some died in infancy; others were fostered by relatives in other places. Almost all of them would leave Betafo before they were thirty, there being no property to speak of or reason for them to stay. When once or twice I tried to make lists of their names, I found it was impossible: people would just shrug and said something to the effect of ‘oh, there were lots of them. Who remembers? None of them live around here any more.’
Unrealized promises, currents, dispersal, disappearance ... the traditions surrounding Rainitaba seem to echo the sense of loss and displacement inherent to the experience of slavery, and to make it a figure for the lineage’s own eventual dispersal, its withering away as a presence in Betafo. In fact, it is a very complicated story, which draws together a series of very old ideas and images—some Merina, some Betsileo—into a narrative so powerful that it has gone from an obscure piece of family history to an essential part of the historical consciousness of the community, a story everyone could repeat.
Part of how it could do this was by seizing on the richness of the term ‘Vazimba,’ a word which can be used to refer to ancient aborigines, lost ancestors, or dangerous spirits of the water—categories which tend to overlap considerably. It might be helpful to explore some of the term’s meanings: not least because it has become something of a notorious issue in the scholarly literature.
The so-called ‘Vazimba problem’ has, in fact, generated a very long and (in my opinion) largely poindess intellectual history. It all started in the 19th century, when early missionaries heard stories about dark, diminutive Vazimba spirits lurking in wild places, and concluded that they must reflect the memory of an ancient ‘aboriginal race’ that had occupied the highlands of Madagascar before its present-day inhabitants (see Berg 1975, 1977). The logic seems to have been this: the people of Imerina tended to have straight hair and more Asian features than most other Malagasy. Therefore, they had to be the descendants of recent immigrants from the Malay archipelago. The Vazimba, then, would have to be the people already living in the highlands when they arrived: backward, dark-skinned savages, originally from East Africa. For English and French missionaries working in Imerina, this soon became a matter of simple common sense. There was some speculation the Vazimba might have been pygmies; others argued that the Crace of pygmies’ (called Kimosy) was an even earlier strata the (perhaps pastoral) Vazimba drove out, before they were in turn put to flight by conquering Malays.
Needless to say, no evidence was ever produced to back up any of this, and there would be litde reason to go into it were it not for the fact that this picture of Malagasy history has become entrenched in schoolbooks and, therefore, that anyone who has been to school has been exposed to it. When the descendants of free Merina call themselves ‘white people’ today, in contrast to the descendants of slaves, who like people of the coast are called Cblack’ they draw on this picture of Malagasy history. On the other hand, popular conceptions of Vazimba themselves seem to have changed little from the ones Gerald Berg (1977: 7–12) documents for the early 19th century—the stories that missionaries first seized on and misinterpreted.
First and foremost, Vazimba were ancestors whose bodies had been lost. If a man or woman drowned or died in a far-off country and their body was not recovered, they were often said to have ‘become a Vazimba.’ This could be a simple figure of speech; one did not necessarily mean anything more than that the person would never become a proper ancestor, never be wrapped and placed inside the tomb. But, more often, the term Vazimba was applied to the ghosts of such unfortunates, dangerous spirits, angry because they were cut off from proper relations with their descendants.
For all they lacked bodies, Vazimba were always identified with a specific place. Most often, their ghosts inhabited marshy places far from human habitation: little springs or pools between the rice fields, grottoes often marked by the presence of red fish or red crabs, knots of bamboo, reeds and rushes, sometimes, in certain kinds of tree. One might occasionally encounter a Vazimba by a rock or spring on an isolated hillside or even amidst the crags of a mountain, but it was unusual to find them far away from water.
I heard a lot of speculation about the origin of such ghosts. One medium from Arivonimamo told me they were usually ancestors whose descendants no longer ‘took care of them.’ If descendants stopped conducting famadihana, stopped keeping up the tomb, eventually the ancestor’s fanahy or soul would leave the crumbling tomb entirely to settle in watery places, having become a fierce creature full of resentment towards the living. Others suggested most Vazimba were the spirits of travellers from other parts of Madagascar—Bara, Sakalava, Betsileo— who happened to die while passing through, and were buried hastily on the spot by whoever found them there. Others would point to the existence of Kalanoro: small human-like creatures rumored to live in distant lakes and marshes. Vazimba, they suggested, were the ghosts of Kalanoro. Finally, some (for instance, Betafo’s schoolteacher, or one of its former pastors) did speak of Vazimba as if they were a former population, long since driven away, and therefore, whose ancestors no longer had any descendants to remember them.
Many refused to even speculate. The important thing about Vazimba, Armand’s brother Germain once told me, is that you don’t know what they are or where they come from. They are by definition mysterious, invisible, a kind of unknown power:
Germain: Vazimba are a kind of thing that isn’t seen. They don’t show their bodies like, say, people do, or the divine spirits who possess mediums and cure people. If you carry pork to a place where one is, then that night, as soon as you kill the light you look and there’s this hand moving towards you. As soon as you light the candle again, it’s gone. Or, say you’re washing your face in you don’t know what... and likely as not your face will swell up hugely like this, and it absolutely won’t go away until you bum incense over it. You take it to someone who will make offerings, and then you’re cured. But that’s all you know—you have absolutely no idea what was in the water.
As this quote makes clear, when people thought about Vazimba, it was usually not as a matter for abstract historical speculation but as one of immediate practical concern. One discovered that a place was haunted by Vazimba because someone had taken ill. A child playing in the fields had drunk some water from a reedy pool, or taken fishes that they shouldn’t have, or they had been tending pigs or taking some other polluting substance to the place where a Vazimba was. Such children would often fall into a fever, or parts of their body would swell up; usually, they would be tormented by dreams or apparitions. Vazimba were normally invisible; when they did appear, it was almost always in the nightmare visions of an adolescent or a child. Normally they appear either as horribly mutilated—fingerless, noseless—reaching out to snatch the children, or else, especially with older children as extremely attractive members of the opposite sex, trying to lure them into their watery domains. The ritual for expelling Vazimba was similar to rituals for expelling hostile ghosts: the curer would burn things, there would be incense, smoke. But one would usually also leave offerings at the place, almost exactly those one would give to ancestors at famadihana: rum or honey, candies, ginger, suet, bananas, bread. If nothing else, these rituals would ‘clean’ the place of the pollution that had offended the creature and made it ‘fierce,’ to soothe it, placate it, and at the same time, ensure it remained confined there.
Some places thus develop reputations. I knew at least seven in the eastern half of Betafo alone where there were rumored to be Vazimba; I wouldn’t be surprised if there were more. Often it was not entirely clear, because people differ on whether there is still a Vazimba in a given spot, or if there is, whether it is still masiaka, ‘fierce,’ still a force to be reckoned with.
For present purposes, what is really important is the relationship between Vazimba and slaves. This relationship appears to be longstanding. Many of the captives brought to Imerina quickly developed ritual ties with local Vazimba. It makes a certain sense that they should feel an affinity, since Vazimba were themselves figures of loss and dispersal. The one common feature in all stories about Vazimba is that they involve people being uprooted, cast out of their proper place. Vazimba are people who have been driven from their homes, ancestors whose descendants have dispersed and forgotten them, who have themselves left their solid tombs to enter confused, watery places, like slaves, then, their defining feature is that they are lost; they embody the complete negation of those ties of descent that bind the living to ancestors buried in ancestral soil. If slaves were people wrenched from their ancestors, Vazimba were ancestors lost to their descendants. It is not difficult, then, to understand why slaves might have seized on these images as a way of capturing their own experience—and in many cases at least, translating it into a source of power with which to restore some of what they’d lost.
Nymphs and Mediums
The only well documented example of how such ritual ties first developed is a story preserved in the Tantara ny Andriana, a collection of 19th century Malagasy texts (Callet 1908: 240–243; Dahle and Sims 1984: 197): about a woman originally from Betsileo, who became the medium for a Vazimba spirit named Ranoro.
First, a word about Ranoro. Ranoro remains, even today, probably the most famous Vazimba in Madagascar (Domenichini 1985: 416–445; Rajaofera 1912; Aujas 1927: 16–17; Peetz 1951; Haring 1982: 358–359; see Bloch 1991). She is considered one of the most ancient ancestors of a large and historically significant descent group called the Antehi-roka, whose territory is just to the north of the capital. The Antehiroka are sometimes described as ‘Vazimbas themselves—if only because they were the original inhabitants of the hill on which Antananarivo, the Merina capital, was later built, displaced when it was taken over by an early king.
According to the story the Antehiroka ancestor Andriambodilova was strolling by the banks of the river Mamba one day when he chanced on Ranoro, a beautiful water nymph (zazavavindrano), sunning herself on a rock. He proposed marriage. Ranoro was not necessarily disinclined, but she warned him that marriage with supernatural beings was difficult; there were always all sorts of taboos. If he wished to marry her, he would have to agree, among other things, never to use salt or even to pronounce the word.
Some versions explain the reason for this unusual demand. Ranoro knew that if she abandoned the waters to marry a mortal man, it would mean never again seeing her father, whose name was Andriantsira, ‘Lord Salt.’ Therefore, she made him promise never to say or do anything that would remind her of him. He agreed, the two married and had children. But one day many years later, during a domestic argument, he spat it out in anger, calling her ‘daughter of salt.’ No sooner had she heard the word than she turned her back on him and walked to the banks of the river, dived in, and was never seen again. The place where she disappeared, a rocky grotto by the river, has been a place of worship ever since, and her present-day descendants continue to maintain a taboo on salting food—in fact, many versions add that any salt that comes near the grotto immediately dissolves.
The taboo on salt is the main claim to fame the Antehiroka had among people I knew in Arivonimamo: it was considered the most difficult taboo anyone had ever heard of. And salt does seem the key to the story. In fact I suspect the whole story is a kind of play on a Malagasy proverb: sira latsaka an-drano, tsy himpody intsony, ‘like salt fallen into the water, it will never again return to its previous form.’ As with Rainitaba, a broken trust leads to a very literal dissolution: what was once a single object becomes an infinity of tiny things which flow away in all directions.
Already in the 19th century, Ranoro’s sanctuary, like her husband’s tomb, had become a place where people came from far and wide to make vows, and ask for favors; Ranoro is still famous for helping infertile women to conceive. But in the 19th century, these were places which slaves were not allowed to enter. The presence of slaves was considered to be polluting, in much the same way as pork. This makes it all the more surprising that the most famous disciple of Ranoro of that century was, in fact, a slave—she was a woman originally from the Betsileo country in the southern highlands of Madagascar. During the reign of Ranavalona I (1828–1861), this woman—always herself referred to as Ranoro—began periodically to fall into trance and be possessed (tsindriana) by the Vazimba’s spirit. Her fame began to spread after she cured a woman who had been struck blind for having sullied Ranoro’s grotto; soon, even free people were beginning to frequent her, seeking advice and cures. According to the story preserved in Callet (1908), the news eventually reached the Queen, who ordered the woman to be put to the poison ordeal. When she survived, Ranavalona recognized her claims to be legitimate and granted her an honorary guard of thirty Merina soldiers.
The mortal Ranoro appears to have become a figure of some fame and influence—despite her continued status as a slave. She is said to have slept on an elevated bed suspended from the rafters, to have walked across the room on a tightrope when possessed, and performed other remarkable feats. For her last miracle, she went to Ranoro’s grotto and dived into the water; it was only three days later, according to the story, that she emerged. In the interim, she told her followers, she had lived with Vazimba in the bottom of the cave, who fed on raw fish and raw crabs. They tried to make her join them in their meals, but the food repelled her, and having refused them, she was returned to the surface. The spirits had rejected her. Claiming her contact had thus been broken, the woman left and went back to find her father, who she believed was still alive somewhere in Betsileo.
In her case at least—and hers was clearly very unusual—it was possible to use access to Vazimba as a way of restoring the severed bonds of descent.
Fanany: people who come back as snakes
It is worth exploring the connection to Betsileo in more detail, since most ‘black people’ in Imerina today claim Betsileo origins.
Betsileo is the name given the country directly to the south of Imerina, including most of the southern part of the vast plateau that forms the center of Madagascar, as well as to the people who live there. In the 18th century, they were divided into a number of independent kingdoms; in the beginning of the 19th, Betsileo was conquered by the
Merina king Andrianampoinimerina. From the point of view of present-day Merina, the Betsileo are a bit of an anomaly. On the one hand, like all other Malagasy they are considered ‘black people’—if only because they are much less likely to have straight hair. However, in almost every other way, they are indistinguishable from Merina. Their way of speaking is similar; so are their houses, clothes, and ritual practices. If any differences are widely remarked upon, it is that Betsileo tend to be more open and easy-going than Merina (those from Amba-lavao are widely held to be the most talkative people in Madagascar), and are much more sophisticated farmers: their skill at irrigation and terracing, for example, is famous throughout Madagascar.
Nowadays, descendants of slaves in Imerina almost always claim to be Betsileo (I met dozens, in fact, who insisted they were descendants of the famous Betsileo king Andriamanalina). There are any number of reasons why the identity might seem appealing. Not only were Betsileo also ‘black people’ who were otherwise indistinguishable from Merina, they also had a renown as migrant laborers. During the 550s and ’60s, thousands used to cross Imerina every year, following the rice replanting and the harvest; many ended up marrying local people and stayed on. Since local attitudes towards Betsileo migrants were so strikingly more accepting than they were of former slaves (whom most claimed their ancestors forbid them to marry) it is easy to see how mainty laborers already living in Imerina would wish to blur any distinction between them.
To understand the story of Rainitaba, though, one has to understand something about Betsileo mortuary custom. While ordinary Betsileo are buried in much the same way as Merina, the souls of royalty were believed to transform themselves into snakes; specifically, a species of water snake called fanany or fangany (Shaw 1878: 411; Abinal 1885: 242–246; Sibree 1880: 170; 1896: 198; Dubois 1938: 716–18; Delord 1958; Razafintsalama 1983). Betsileo royal funerals involved a variety of sometimes elaborate processes aimed at separating the wet portions of the body from the dry. Sometimes the entrails were removed and thrown into lakes, where they are thought to transform into fanany\ in other cases fluids were drained from the corpse, and the snake was believed to develop from a worm that fed from those fluids; the worm having been placed inside the tomb at the same time as the dried cadaver, along with a bamboo tube by which it can escape.
Fanany are striking-looking snakes, easily identifiable because of their coloration, which is said to resemble that of a lamba arin-drano, with bands of white, orange, and black. But they are rarely sighted on dry land.
When one of these is found the people assemble and ask it if it is the fangany of So-and-so, mentioning in succession the names of the various chiefs who are dead; and the animal is asserted to nod its head when the right name is mentioned. The relatives of the man at whose name the beast moves its head then take possession of it by inducing it to wriggle on to a clean lamba, by which it is carried to the former residence of the dead man. Oxen are killed, feasting commences, and a scene similar to that at a funeral ensues. A little of the blood is presented to the fangany, after which it is set free in the neighborhood of the chief’s grave (Shaw 1878: 411).
While this was a 19th century Betsileo custom restricted to royalty, just about everyone I spoke to in 20th century Imerina had heard of it. Certainly, all olona mainty. Almost all of them though were under the impression that this was something that can happen spontaneously to anyone from Betsileo.
The way I usually heard the story was this: should a fanany happen to appear in a Betsileo’s house, the family immediately assembles and lists the names of their ancestors, until it nods its head to tell them which it is. After that, they feed it rice with milk and honey, sometimes play music or otherwise celebrate its presence, but afterwards, the head of the family makes a speech, thanking it for coming, but asking it to leave now ‘because’ (as Nety put it) ‘you’re frightening the children.’
Here are two different stories I heard from old men from Betafo, stories about Betsileo wage laborers which both seem to date back to around the 1920s—well before the massive migrations of the 1950s and ’60s. One was told me by the catechist from the village of Amba-ribe, in the far west of Betafo’s territory. His uncle, he explained to me, had once employed an ancient Betsileo cattleherd, who was also a medium, and his great-uncle had also once employed him to catch a Vazimba and move it into a nearby moat—which he did, trapping the spirit in a giant sobika basket. Later the medium himself grew sick and was about to die.
Ramena: However, when he was about to die—and he was very old—he wouldn’t let them wrap him in a lambamena, or bind him. They put him in the bottom of a sobika, and covered it up with earth. Three days later, when they looked again, sure enough he wasn’t in there any more. The hole was wide open.
The refusal of proper shrouds of course echoes the story of Rainitaba, as does the disappearing body. But Ramena leaves the rest ambiguous. An old man from Andrianony told me a much more explicit story about a Betsileo laborer who turned into a fanany:
Ramwao: It was a long time ago, back when they used to drive cattle over the rice fields instead of harrowing them. And there was a hireling of Rakotomanga’s, who lived in Antanety. He was Betsileo. The man had been hired to harvest rice, but he got sick and died while he was working in the fields. So they buried the Betsileo in the ground above the rice field. After he’d been in there for a year, he came out as a fanany. Came out as an animal.
And having come out as an animal, he went up to Rakotomanga’s house. And they said ‘What?’ And they said: ‘Are you what’s left of So-and-so?’ The creature nodded his head. They gave it rice. The creature ate the rice (they say it was really huge, but I didn’t see it with my own eyes) and after a while, they told it ‘if this is truly you who are here before us, then leave us and go home. Go back to your ancestral lands.’ When the creature had its fill of rice, it headed off. And when it had arrived there, then the man had a dream where it told him ‘I have arrived at my ancestral lands.’
Such stories—about isolated Betsileo workers—already overlap with ideas about lost travelers who die away from home and become Vazimba. But in this case, transformation contains the possibility of resolution; a number of people told me that Betsileo who happen to die and are buried far from home will almost always turn into serpents and go home.
At this point, it is clear how the story of Rainitaba—which probably also took form in the 1920s—patches together elements taken from Vazimba stories with those about fanany. According to his descendants,
Rainitaba was an andriana from Betsileo who had been kidnapped and taken to Imerina. When he died, he refused to be tied down and contained like a Merina ancestor; instead, he wanted his children to release him in the currents of the river, wrapped in a Betsileo cloth that was colored like the coat of a fanany. When they couldn’t bring themselves to so completely offend the sensibilities of their Merina neighbors, he made the transformation on his own; but as a result, they lost the blessing he would otherwise have given them. His subsequent history had him following the course of the waters downstream, but also in a sense, moving backwards in time: from Betafo to his old settlement of Antan-drokomby, and finally, back to Betsileo once again. But as with Ranoro (who was also sometimes said to have moved steadily downstream, ever further from her former home), returning home meant orphaning one’s own descendants.
Of course, it is only now that his descendants have almost entirely dispersed, that people say he’s abandoned Imerina entirely. For decades, he lingered in the waters around Betafo, visiting his grandchildren in their dreams, suspended halfway in a movement of escape he could not yet bear to bring to its conclusion.
One reason the story seemed to strike such a chord with people was that it encapsulated something fundamental about the experience of slaves and their descendants. Perhaps, one can even say, the experience of slavery itself. There are those that have proposed that slaves are by definition human beings who have been wrenched from the society which formed them, the web of social ties which has made them what they are; plunged into a kind of ‘social death’ (Meillassoux 1991; Patterson 1982). From this perspective, slavery as an institution is founded on the destruction of social worlds, and it is in fact the moment described by Raombana, when children are tom from their mother’s arms and families broken apart, which makes a slave a slave.
It is difficult to assess the full implications of such a moment for the historical consciousness of those who passed through it. Rarely do large numbers of people go through a rupture so utter and extreme. It brings to mind Elaine Scarry’s (1985) observation that physical pain empties worlds of their meaning. In normal life, one is invested in a thousand ways in one’s surroundings, in people, places, projects, things one cares about; so that one’s sense of self expands outwards to imbue and become entangled with a much larger social world. One effect of extreme physical pain, she says, is to empty these investments of all meaning; one’s sense of self collapses into the narrow confines of the hurting body. For that moment nothing and no one else is real. The scene described by Raombana in a way reverses this: the victims are, for the most part, physically unscathed, but as they are lead off from burning villages, most of the men they have ever known lying dead in bloody pools, women and children dragged from each others’ arms; in a matter of hours, the entire universe of social relations in which they have come into being was utterly annihilated. The result, as Raombana’s friend himself observed, was a trauma so intense that no mere physical pain could possibly surpass it.
All this does not mean that memories of such a moment are likely to become a part of historical consciousness. In fact they are just the sort of events that would not; that one would normally suspect survivors would prefer never to have to talk about. Certainly the memory of them is not preserved in oral histories of the present day. But if for slaves and their descendants, that one moment, when worlds dissolved away, seems to have reverberated endlessly, it is because such experiences did not stop. Dispersal, families drifting apart, people uprooted from their memories: for most, it was repeated with every generation.
It also took place within a cultural milieu which placed an extraordinary emphasis on the politics of memory. The manipulation and transformation of such memories—particularly, women’s memories of their parents and ancestral homes—was a constant theme of Merina ritual. In marriage negotiations, for instance, the suitors’ family offered a series of cash payments which compensated either for the nurturance and care the girl’s parents had provided her—such as the valim-babena, the ‘answer for having carried an infant on one’s back’—or services the daughter herself would no longer be able to provide—the akana kitay, ‘gathering firewood,’ or alana volo fotsy, ‘pulling out white hairs.’ In the latter case especially, an image so intimate of a daughter poring over her father or her mother’s head, searching for white hairs to pluck out, evokes a whole world of domestic sentiments: protective affection, the fear of aging and resultant loss, the pain of ruptured domesticity when the woman moves away. The money, officially meant to ‘ask for the parent’s blessing’ for the marriage, can equally be seen, I think, as compensation for that pain. In famadihana the evocation of emotionally-charged memories becomes even more explicit. When one places the corpses of women’s relatives on their laps, the effect is to break the power of women’s most vivid, intimate memories of people that they loved. It evokes that entire world in order to efface it, to free the living from their attachments to the dead.
For women who had been carried into slavery, evocation of such memories could only serve as a reminder of acts of such irreparable violence, that the entire world of those memories had been brutally destroyed. It is hardly surprising that the Betsileo woman should have felt such an affinity with the figure of Ranoro. Ranoro was a woman who could not bear to hear her father’s name; the memories it evoked for her would be too painful. It was a story about salt dropped in water, things that could never be brought back together or attain their previous form. In this one case, of course, the story may have taken on a different level of poignancy because the woman possessed by Ranoro believed her father was still alive—at least, in the end she managed to win her freedom so as to try to find him. Though one cannot help but wonder whether the dream of finding her father was really as much a projection of her imagination as Ranoro herself had been.
Containment and Redemption
At this point, let me return briefly to the question with which I began: about stories that can, and can’t, be told.
Students of working-class history have noted that it is relatively easy to cull oral histories of periods of successful strikes, political advances, in which workers had some control of their destiny; much harder for periods of massive retrenchment or defeat. When Italian workers told the stories of their lives, for instance, the two decades of fascist rule often seemed to drop out entirely (Passerini 1987). In a fascinating essay, Michel Bozon and Anne-Marie Thiesse (1986) asked: what happens, then, to workers who have never known anything but defeat? Their research focused on farm laborers from the countryside near Paris, people who had begun their lives at the bottom of the social heap and then moved down steadily: made redundant by mechanization, set to scrabbling for endless miserable jobs wherever they could get them. They discovered that, indeed, most found it impossible to give any account of their lives since childhood; many found it painful to even try. Instead, they tended to fall back on quasi-ethnographic descriptions—£how we used to do things in the old days’—and anecdotes about their own experience of famous historical events—mainly, of France’s wars. These anecdotes, however, were in almost every case and themselves little images of loss and dispersal: peasants fleeing before unearthly German horsemen, the government fleeing Paris by balloon.... It was as if, having been told all their lives they had no right to speak of or for themselves, they could only do so through the borrowed authority of ‘national’ history.
Malagasy understandings of what history is, and what gives one the authority to tell it, are rather different. Histories are, indeed, matters of privilege, but they are also intrinsically tied to places—where ancestors lived and are buried, where famous events took place. And there is a very deep-seated feeling that only those who live near a place can really know its history. Even the wealthiest and most powerful descendants of Betafo’s noble families would look mildly irritated when I asked them about the histories of their illustrious forbears, unable to speak about the place because their families had long since relocated to the city. Several ended up referring me to the descendants of their former slaves, who still lived there. In this sense what Vazimba pools provide is not just a way to conceptualize a history of pain and dispersal, but the right to speak of it: after all, most of these pools were in wild places in the valley bottoms, the very places to which slaves too were once exiled.
Even if what they spoke of was, ultimately, their own sense of loss, their own disempowerment, the ability to speak about such things itself opened up possibilities of taking action and beginning to reverse the situation. In the case of Ranoro this was fairly obvious; less so, perhaps, in the case of Rainitaba. But histories keep changing, and Vazimba provide endless possibilities of moving from speech to action.
One of the last descendants of Rainitaba everyone remembered, for example, was an old man named Pascal, a landless laborer who had died several years before I came, when, while working a neighbor’s field, he had an epileptic fit and fell on his own shovel. Pascal, I was told, was haunted by a Vazimba, who would periodically possess him. Some said it was Rainitaba himself; others insisted it was a different, nameless spirit, that he had first encountered while swimming in a pool to the north of the village. All though remembered how Pascal would periodically announce he felt the Vazimba beginning to move in him, and how practically the entire population of the mainty quarter of Betafo would set out across the fields to the north of the village, and then gather to sing and clap, encouraging the spirit to emerge. The sessions would always end the same way. After some time, the Vazimba would come to him, and Pascal would bolt off randomly into the surrounding woods or waters, whereon everyone would chase after him, to bring him back again—in an endless drama of dispersal and retrieval.
These gatherings were organized with the help of another man— one I will call Rainibe. Unlike Pascal, Rainibe was an experienced medium. This is not how people around Betafo put it, though. Rainibe, people would tell me, ‘had’ a Vazimba. Many insisted that he had moved out from his old home in Betafo to found a new hamlet at the end of a long valley to the northeast just to be closer to the field where it was. Every night when there was a new moon his whole family could be heard out there, clapping and singing to bring it out. Rainibe himself never admitted any of this to me—in fact, he denied that he even worked as a curer. But this sort of coyness was typical of people who had Vazimba, because being too close to a Vazimba is a morally dubious thing: such spirits can not only help one in curing, they can also take vengeance on one’s enemies, making one little better than a witch. Most people in Betafo were very careful not to pick a quarrel with Rainibe, and from the money he got from curing and the fear he inspired he had managed to acquire quite a bit of land, and to keep a very large number of his children and grandchildren around him in his little hamlet by the fields. This was the main thing that struck me when I would visit him in his hamlet, occupied entirely by his own descendants. He was always surrounded by children.
Rainibe did not entirely deny his links with Vazimba. His grandfather, he said, had many years before been mysteriously pulled into the waters underneath the dam to the northeast of Betafo, and wasn’t seen for days. When he finally re-emerged, he told little, except that the spirit was an old man with a long beard, dressed all in red, surrounded by endless numbers of tiny children. He had stayed there for three days in all, and all he had to eat was crabs.
‘Raw crabs?’ asked my companion, an andriana from Betafo.
‘I have no idea if they were raw or they were cooked.’ Then, deciding this might be a bit too coy: ‘well, I guess there wouldn’t have been cooked ones.’
‘Because you know what they say’ (everyone, apparently, knew this story): ‘if you eat those, then they become your friend for life. But if you refuse, they might even kill you.’
‘Yes,’ he smiled. So they say.
Just so as to show that anything is possible, let me end by noting that towards the end of the time I was in Betafo I discovered there was another of Rainitaba’s descendants living there. He was one of Razanamavo’s sons, a man in his thirties named Tratra. Tratra too claimed to be a medium. Nobody I talked to took these claims particularly seriously: most considered him a drunken blowhard, and a bit of a buffoon (‘if you really have spirits,’ Armand told me, ‘you don’t go around telling everyone.’) But a few years before he had built a little house near the ruins of Antandrokomby, just a few meters away from the reedy pool Rainitaba is said to have inhabited before he disappeared. He couldn’t afford to be around very often; most of the time, like his mother, he was off looking for work. But it at least suggests the possibility that, were one to come back in twenty years, Rainitaba might have acquired a new and entirely different history.
I should thank Bruce Applebaum, Jennifer Cole, Jean Comaroff, Nhu Thi Le, Stuart Rockefeller, and Hylton White for their many helpful comments and suggestions.
1885 Vingt ans a Madagascar: colonisation, traditions historiques, moeurs et croyances. Paris. Aujas, Luis
1927 Les rites du sacrifice a Madagascar. Tananarive: Memoire de l’Academie Malgache, fasc. II.
1975 Historical Traditions and the Foundations of Monarchy in Imerina. Doctoral Dissertation, University of California, Berkeley.
1977 ‘The myth of racial strife and Merina kinglists: the transformation of texts/ History in Africa vol. 4, 1–30.
1980 ‘Some words about Merina Historical Texts.’ In The African Past Speaks: Essays on Oral Tradition and History (J.G. Miller, ed.). London: Dawson, pp. 221–239.
1971 Placing the dead: tombs, ancestral villages, and kinship organization in Madagascar. London: Seminar Press.
1978 ‘Marriage amongst Equals: An Analysis of Merina Marriage Rituals.’ Man (n.s.) 13: 21–33.
1979 ‘Slavery and Mode of Production in Madagascar: Two Case Studies.’ In Asian and African Systems of Slavery (L.J. Watson, ed.). Berkeley: University of California Press, pp. 123–45.
1982 ‘Death, women and power.’ In Death and the Regeneration of Life (M. Bloch andj. Parry eds.). Cambridge: Cambridge University Press, pp. 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.
1994 ‘The Slaves, the King, and Mary in the Slums of Antananarivo’ in Shamanism, History and the State (N. Thomas and Caroline Humphrey eds.), Ann Arbor: University of Michigan Press.
1984 Distinction: A Social Critique of the Judgement of Taste. Cambridge: Harvard University Press.
Bozon, Michel and Anne-Marie Thiesse
1986 ‘The Collapse of Memory: the Case of Farm Workers (French Vexin, pays de France).’ History and Anthropology 2: 237–259.
1972 ‘Cultes de possession dans la plaine de Tananarive.’ Cahiers du Centre d’fitudes des Coutumes 9: 33–66.
1908 Tantara ny Andriana eto Madagascar, Antananarivo: Imprimerie officielle.
1989 How Societies Remember. Cambridge: Cambridge University Press.
1896 ‘The Abolition of Slavery in Madagascar: with Some Remarks on Malagasy Slavery Generally’ Antananarivo Annual and Malagasy Magazine 21: 446–50.
Dahle, Lars and Louis Sims
1984 Anganon’ny Ntaolo, Tantara Mampiseho ny Fombandrazana sy ny Finoana Sasany Nanganany. Antananarivo: Trano Printy Loterana.
1974 UHistmre des Rois d’Imerina: interpretation d3une tradition orale. Paris: Klincksieck.
1958 ‘Note sur la croyance en la reincarnation des princes betsileo sous la forme d’un serpent.’ Bulletin de l3Academie Malgache 47: 155–170.
1985 Les Dieux au Service des Rois: Histoire orale des Sampin3 andriana ou palladiums roy-aux de Madagascar, Paris, CNRS, 716p.
1938 Ethnographic des Betsileo. Paris: Institut d’Ethnographie.
1838 History of Madagascar. 2 vols. London: Fisher & Son.
1954 Les esprits de la vie a Madagascar. Paris: Presses Universitaires de France.
1982 ‘The King’s men in Madagascar: Slavery, citizenship and Sakalava monarchy.’ Africa 52: 31–50.
1986 ‘Ritual and Work in Madagascar.’ In Madagascar: Society and History (C.P. Kottak, J.-A. Rakotoarisoa, A. Southall and P. Verin, eds.) Durham: Carolina Academic Press, pp. 157–174.
1991 A Green Estate.: restoring independmce in Madagascar. Washington D.C.: Smithsonian Institution Press.
1992 ‘The past in the future.’ American Anthropologist 94: 837–59.
1995 ‘Dancing with Corpses Reconsidered: an Interpretation of Famadihana (in Arivonimamo, Madagascar).’ American Ethnologist 22: 258–278.
1996a The Disastrous Ordeal of 1987: Memory and Violence in Rural Madagascar. Ph.D. dissertation, University of Chicago.
1996b ‘Beads and Money: Notes Toward a Theory of Wealth and Power.’ American Ethnologist 23: 1–32.
1996c ‘Love Magic and Colonial Labor Policy in Central Madagascar.’ Gender and History 8.
1893 Malagasy Village Life: Pen and Ink Sketches of the People of Western Imerina. Antananarivo Annual and Malagasy Magazine 18: 1–20.
1982 Malagasy Tale Index. Folklore Fellows Communications no. 231. Helsinki: Suomalainen Tiedeakatemia.
1894 ‘Funerailles, tombeaux et honneurs rendus aux morts a Madagascar.’ Anthro-pologie 5: 385–401.
1992 Making Ethnic Tradition in a Pre-Colonial Society: Culture, Gender, and Protest in the Early Merina Kingdom, 1750–1822. Doctoral dissertation, University of Wisconsin-Madison.
1991 The Anthropology of Slavery: The Womb of Iron and Gold. Chicago: University
of Chicago Press.
1987 Fascism in popular memory: the cultural experience of the Turin Working class. Cambridge: Cambridge University Press.
1982 Slxwery and Social Death: a Comparative Study. Cambridge: Harvard University
1896 ‘De l’Esclavage a Madagascar.’ Le Correspondant [Paris), 10 February 1896,
1991 Bible et pouvoir a Madagascar au XIX* siecle. Paris: Karthala.
1912 ‘La culte d’Andriambodilova.’ Tananarive: Bulletin de VAcademie Malgache 10:
1975 ‘Etude du village d’llafy.’ Annales de I’Universite de Madagascar, Serie Lettres et Sciences humaines, no. 8–10. Paris: Cujas.
1984 ‘Blancs et Noirs,’ les dimensions de l’inegalite sociale. Documents socio-linguistiques.’ Cahiers des sciences sociales (Antananarivo) 1: 39–75.
1986 ‘Hafy: terre et parente.’ In Madagascar: Society and History (G.P. Kottak, J.-A. Rakotoarisoa, A. Southall and P. Venn, eds.) Durham: Carolina Academic Press, pp. 229–244.
N.d. Annales. Manuscript preserved in the archives of the Academie Malgache, Antananarivo.
1983 ‘Les funerailles royales en Isandra d’apres les sources du XIXe siecle.’ In Les Souverains de Madagascar (F. Raison-Jourde, ed.) Paris: Karthala, pp. 193–210.
1910 Contes de Madagascar. Paris: E. Leroux.
1920 ‘Ancetres et dieux.’ Bulletin de VAcademie Malgache (n.s.) 5: 1–261. Richardson, John
1885 A New Malagasy-English Dictionary. Antananarivo: London Missionary Society. Ruud, Jorgen
1960 Taboo: A Study of Malagasy Beliefs and Customs. New York: Humanities Press. Scarry, Elaine
1985 The Body in Pain: The Making and Un-Making of the World. New York: Oxford University Press.
1876 Remarks on Slavery in Madagascar. London: Elliot Stock.
1878 ‘The Betsileo: Religious and Social Customs.’ Antananarivo Annual and Malagasy Magazine 4: 312–334.
1880 Madagascar: the Great African Island. London: Triibner & Co.
1896 Madagascar Before the Conquest’ The Islands the Country, and the People. London: T. Fisher Unwin.
Van Gennep, Arnold
1904 Tabou et Totemisme a Madagascar. Paris: Ernest Leroux.
1977 Croyances et Moeurs des Malgaches (2 volumes). Antananarivo: TPFLM.
 I would like to thank Nhu Le for first suggesting this connection.
 Raombana’s ‘Annals,’ A2, no. 5: p. 74 (693–696). Raombana wrote his history in English; in part to ensure that no one else at court could read it.
 My generalizations about slavery are based partly on European sources (e.g., Sewell 1876, Cousins 1896, Piolet 1896) but even more on my readings of 19th century Malagasy documents preserved in the National Archives, notably court cases and the AKTA series (etat civil).
 I never found anyone in the present who recalled how slaves were buried, though some pointed out that there were a few, exceptionally generous masters who set aside a shelf within their own family tombs for slaves; others built modest slave-tombs near their own.
 Most of the slaves emancipated in 1895 abandoned Imerina entirely. Some returned to their former homes—if those places existed and they still had any memories of them. Many others moved to lands newly opened up for farming to the west. Even among those that stayed, only a handful remained in their former villages. Some moved to towns like Arivonimamo, working as porters, merchants, laborers; others moved off to found their own settlements in depopulated comers of Imerina; many sought work in the capital.
 When several black families did manage to get a foothold in some village, their kin would usually follow, and often this would lead to a kind of ‘white’ flight: the more the children of the mainty prospered, the more the children of the fotsy would move out, so that after a generation or two none of the villages’ former inhabitants were left.
 French documents from the second decade of the twentieth century say that he was Rabakomanga’s father, and that she was bom in Antandrokomby in the 1860s.
 Kalanoro are diminutive, human-looking creatures said to live in watery places; one often hears that they are living versions of Vazimba.
 Armand’s brother Germain had gone much further: he had told me she hadn’t ever had a child and then happened to pass by the stand of reeds where Rainitaba was, saw the animal, and immediately afterwards found she had conceived a male child.
 A kind of silk cloth from Betsileo, marked by bands of bright color set between black and white stripes, worn during Betsileo funerals, but also on festive occasions.
 MV: Izy manko efa ho faty izy, hoy ig>izany Hay renibenay, iizaho izao efa ho faty3 hoy izy, tsy maintsy tonga aty auaratanana io, an-baragp io. Dia izaho hoy izy aza alevina ampasana fa alefasao hanarafy an3io renirano io. Dia hovidiana lamba arin-drano hoy izy, ohatra an’ilay alevina vinanto-lahy izany, dia hoe ‘angamba tsy rumdefany an’iny renirano iny, fa menatr’olona izahay. ‘A-an, tsy menatra ianareo hoy izy io, fa hahita fahasoavina be ianareo raha vitanareo izay.’ Dia tsy nanao izy’, fa nalevina ihany dia izoy lasa izoy.
?: Rainitaba io?
MV: I-e, lavaka kely ohatra an’izao no nilevenany, tsisy fasana miditra eo, amin3ny atsi-nanana. Dia nanao an’anona hamadika an’ilay olona dia nitady lamba dia nijanona tsara Hay lamba. Izy naka.
Raha vitanareo iny zovatra iny dia ho any aminareo fahasoavina be. Dia nitondraka ny rano—maty izy. ‘Izaho tsy tia lambamena, 3 hoy izy, fa lamba arindrano, dia.... Ireny ilay lamba jisaona ireny izany.
Fa tsy manjaiy mahantra ny zanakareo, olona farana mandimby sy ny olona any ivelany izy. Zavatra efa vao tsiryony daholo daholo mihitsy. Fa tsy nahazo fahasoavina izahay fa tsy nanao. Apetraka amin3ny avaratra indrindra amin3ny afovoany sy mbola tsy qftndra hatra-min’izao. Taloha, taranany betsaka ihany fa efa vitsy izy izao.
 When she died in 1913 this land was estimated to be worth about 152 francs; actually somewhat above the average legacy for Betafo as a whole. (Though this might simply be because none of the land had been parceled out in advance among her sons.) Her husband, who had died in 1912, appears to have left no property at all to his descendants, which is apparently the reason he has been entirely forgotten.
 Ingahivelona stayed in Andrianony, the mainty quarter of Betafo, most of his life, as did his daughter Rabakolava, now remembered mainly because of her unusual height. She had a number of children who either married away or, reportedly, moved to town— though I never managed to track any of them down. In her old age she lived by herself and like most solitary old women was widely rumored to be a witch. Her brother Pascal only died in 1985, but his children too have also since disappeared.
 The first settlers of Madagascar undoubtedly came from somewhere in this area: this is why Malagasy is an Austronesian language. But no one has ever managed to come up with any evidence (linguistic, archaeological or otherwise) for such a second migration. Despite this it appears to remain unchallenged in the literature, it apparently never occurring to anyone that, if the first inhabitants of Madagascar came from Indonesia, and people had been coming from Africa ever since, the inhabitants of the most isolated central highlands would be likely to look the most like the original inhabitants.
 I am not suggesting that terms like fotsy and mainty, ‘white’ and ‘black,’ sure simply the products of missionary influence: they go back to 19th century social classifications which, however, originally had a very different meaning. For further exploration of this point, see Graeber, 1996a.
 Occasionally, people I asked about Vazimba would reply with something to the effect of ‘well, you have to understand that the Vazimba were really an entirely separate race’ (always using the French word, race); and then go on to cite things they had read in books about Malagasy history, or heard professors discussing on the radio. Even Armand (who had been to college) did this the first time I asked him about Vazimba; but it was a one-time thing, a kind of bow to the authority of scholarship which seemed to have nothing to do with anything else he had to say on the subject thereafter.
 The word razana, normally translated ‘ancestor,’ actually means both ‘ancestor’ and ‘corpse.’ Vazimba were most definitively not razana. When people chanced upon forgotten skeletons in their fields or wild places, for example, they often speculated that they were the remains of witches or lost travelers; but even if they ended up propitiating the spirit of the deceased, they never referred to them as Vazimba; Vazimba lacked bodies by definition.
 GR: Ny Vazimba aloha dia karazana zavatra tsy hita. Izy tsy miseho vatana ohatry ny hoe olona sa £anaharin’ilay mitsabo olona ^anadrano ireo. Raha toerana misy azy, dia ohatra hoe mitondra hena kisoa io, izany hoe dia nentina izany tsy maintsy atao na. . .vao maty ny jiro dia hitanao misy tanana manatona anao; efa vao mirehitra nyjiro dia tsy hita. Izany hoe, ianao misasa tarehy} tsy fantatra na inona. Dia mety lasa vonto be ohatra an’izany koa, dia tsy qfaka mihitsy hono raha tsy evoahany—misy fanevokan’ny olona azy. AnateranHlay olona fanasina, izay vao qfaka. Dia izay no tena hoe misy... tsy fantatra mihitsy na inona na inona no ao anatin’ny rano fa izay fotsiny.
 In this, the basic underlying logic was not all that different than that of famadi-hana, which as I’ve argued elsewhere (Graeber 1995) was largely about the containment of ancestors.
 The connection probably would have made perfect sense to Europeans who assumed that Vazimba were themselves the remnants of an African population; slaves, after all, were mainly drawn from the coastal populations of Madagascar, who were assumed to be equally African. But clearly, this had nothing to do with Malagasy attitudes. I certainly never heard anyone refer to Vazimba as ‘black.’
 This indeed is the gist of most previous analyses. Gerald Berg (1977) for instance notes that in Merina king-lists, the earliest rulers are referred to as ‘Vazimba’ because unlike later kings they were not buried on solid ground; their bodies were thrown into lakes. Likewise Bloch (1982, 1985, 1986) draws an opposition between ancestors identified with ancestral land and Vazimba identified with water.
 Her fame—or at least her documentation—is in part due to the fact that her grotto is located less than an hour’s drive from the center of the capital.
 An interesting parallel with Rainitaba.
 Pier Larson (personal communication) has pointed out to me that in the 18th century, most Merina slaves were, in fact, Betsileo. Later this was not the case, but since no single group ever gained the same numerical dominance Bestileo once had, the identity might well have lingered.
 Betsileo in fact are the only other people in Madagascar who regularly perform famadihana.
 That is to say, just before the narrator’s living memory.
 An olon-javatra was the term he used, which means someone constantly accompanied by an invisible spirit.
 NR: Fa ny zavatra hita raha tahaka ny army dia mpiandry ombin’ny anadahin-drminay kosa atao hoe: Betsileo. Olon-javatra hono izy io. Iray soa vahiny [Ohatry ny hoe, petrahan-java-tra izany]. Izy no niandry ny ombin’ny anadahin-dreninay, tery an-tananay ambony andrefana tery, tamin’ny and.ro taloha. Dia. .. nefa efa ho faty izy, efa antitra be izy, dia tsy mba navelany ho fonosana lambamena na nofatorana. Fa dia ambody sobika no nentina. Dia totorana. Dia rehefa auy eo izany a, dia nojerena afaka telo andro, dia tsy too inisony tokoa. Ny lavaka dia nisokatra be izao. Fa izy tsy Vazimba mihitsy fa hoe: olon-javatra fotsiny. Izy efa antitra be Hay izy.
 V: Efa taloha ela be, efa naharoaka omby tamm’izany nandrangaranga. Dia nisy mpikarama tamin-dry Rakotomanga no nipetraka tery Antanety teny—Betsileo io. Ary izy nikarama nijinga vary. Ary rehefa nikarama nijinja vary by, dia marary, dia maty teny ampanoaoan-draharaha teny Hay Rangahy. Dia nileuina teo ambon’ny tanimbarin’ny teo ilqy Betsileo. Dia rehefa nipetraka teo herintaona izy dia navoaka fanany. Navoaka biby izy. Ary rehefa navoaka biby, mala too izy, dia niakatra too an-tranony. Dia hoe: ahoana? Dia hoe: Ianona teo aloha avela? Dia nanantoka ny hhan’ilay biby. Dia omena vary. Dia nihinana vary eo ny biby—ngeza be hono izy (izaho tsy nahita maso) dia rehefa cay eo dia manao hoe: ary rehefa izao ianao a, dia ianao marina no too, dia mendehana mandeha mody. Mivoaka amin’ny tanindrazanao. Dia rehefa voiky teo iky biby, dia lasa nandeha. Ary efa tonga tany izy, dia nanofy ilay Rangahy hoe Honga aty amin’ny tanindrazako. Ary ry zareo izany, rehefa lasa izy lasa nandeha teo iky fatiny. Ilqy olona tsy misy hita intsony.
 In fact, the word for ‘history,’ tantara, could be used to mean ‘privilege’ as well.
 Though almost all these ancient hamlets had long since been abandoned, the slaves on liberation having moved further up the hills.
The almost ritualized invocation of Betsileo origins might also be interpreted as serving as a kind of authorization to speak. In many rural communities I found that even old Merina men would push the descendants of Betsileo migrants forward to tell me local histories, despite the fact they had been bom elsewhere, just on the basis of their greater ability to talk.