I. Historic, Political-Economic, and Social Levels of Experience

2. The Political Economy of the Sambirano

The Sambirano Valley is one of the few regions of Madagascar where one finds evidence of prosperity. Malagasy often refer to this region as one of the most fertile and productive areas of the island, providing a sharp contrast to the rest of the nation. The town of Ambanja lies in the heart of this river valley and is the commercial, political (including royal), religious, educational, and entertainment center for the region. Even though national trends exhibit a downward movement of the economy, the Sambirano has become well-known as a region where “there is work” and “money” (misy asa,misy vola), so that many people have come here from other regions of the island to “seek their fortunes” (hidaty harena). On the whole this nation is plagued by fierce shortages of such essentials as foodstuffs, medicines, and construction materials, but in Ambanja one is struck by the relative beauty of the area and plentitude of goods—though overpriced—that are available. In comparison to other regions of Madagascar, the development that has occurred in Ambanja and the Sambirano Valley is striking. This chapter will explore the factors that account for this over time.

Ambanja, A Plantation Community

The greatest economic force existing in the Sambirano Valley is that exerted by the enterprises, the large companies that own much of the land in the valley. Originally these were private plantations, established by foreign planters around the turn of the century. Following the Socialist Revolution of the 1970s, many foreign nationals fled the country, their lands confiscated by the state. By the late 1970s nearly all of these plantations became fully or semi-nationalized, and today they are managed by Malagasy rather than foreign staff. The shift from private holdings to state capitalism has had an effect on land tenure and work relations; it is also important for understanding Malagasy notions of historical experience. For this reason, throughout this study I will use the term plantation when referring to these farms during the colonial (1896–1960) and early postcolonial (1960–1972) eras, and enterprise when speaking of the period since the Socialist Revolution (1972 to the present).

The plantations (and, more recently, the enterprises) have transformed the geography of the region and shaped, directly and indirectly, the economic, political, social, and cultural orders of the region. Their activities have been characterized by intensive development, the landscape transformed within a few decades into large estates of manioc and sugar cane and, more recently, cocoa, coffee, and perfume plants such as ylang-ylang. During the colonial period, Sakalava living throughout the valley were relocated to make room for plantation lands. Sanctioned by French colonial policies, the activities of the plantations led to the introduction of foreign capital and the proletarianization of local and migrant labor. Colonial efforts to undermine the authority of local royalty also caused the breakdown of local, indigenous power structures and, ultimately, Sakalava cultural identity. As a result of plantation activities, by the 1920s Ambanja had also become a major religious center for the Catholic Church and the district headquarters for the colonial administration. The presence of these forces in the Sambirano has contributed, to a large extent, to the establishment and growth of Ambanja as a major northern urban center.

The Town and its Environs

Regardless of the direction that one takes out of town, the view is the same: shady, damp forests of cocoa and coffee. Almost all of these lands belong to the enterprises. Occasional breaks in the fields of cocoa and coffee reveal independent villages, a handful of small company worker settlements, and rice fields, fruit trees, or stumpy and gnarled trunks of ylang-ylang. There are also the administrative and production centers of the enterprises. Here one finds warehouses where produce is collected, dried, and sorted by hand in preparation for export and concrete colonial-style villas with wide tin roofs and large verandas, which serve as the offices and housing for the management staff. The borders of the enterprises may be drawn where the terrain changes—at the steep ridge of the Tsaratanana mountains to the east; at the semi-arid terrain to the north and south, which is more suitable for wild cashews and for grazing hardy hump-backed zebu cattle; and at the mangrove swamps that flank the seacoast. In contrast to these bordering areas, the land throughout much of the valley is verdant and lush.

The town of Ambanja lies at a major commercial crossroads for the north (see figures 1.1, 2.1; plate 1). One may approach it from the port of Antsahampano (approximately fifteen kilometers to the northwest), from the national capital, Antananarivo, which lies to the south in the central high plateaux, and from Diégo, the provincial capital, which is a day’s drive (240 kilometers) to the north. Until a decade ago the only practical way to reach Ambanja from other parts of the island was to fly or travel by boat to the neighboring island of Nosy Be, take a ferry to the small port of Antsahampano on the main island, and then catch ground transport to town.

For a long time this road between the port and the town was by far the best in the area. It was originally constructed for the early plantations and it is still maintained by the enterprises, which use the port to export their produce. The significance of the role played by the enterprises in terms of road maintenance and construction was particularly evident in 1987 when telephone poles and electric wires were installed, stretching from Ambanja all the way to the port on the coast. This ensured that electricity eventually would be available for the inhabitants of the villages along the road. Nevertheless, such progress had its price, since the poles were mounted on the south side of the road, making it necessary to cut down fruit trees in private yards, while the northern side, where there are groves belonging to the enterprises, remained untouched. This road to the port is also used by many people wishing to travel by ferry to the nearby island of Nosy Be, which is both a tourist resort and a site for other large-scale enterprises that grow sugar cane. The ferry provides inexpensive and convenient transportation for passengers wishing to visit their kin or the shops on Nosy Be, while young people may make the journey to enjoy the nightlife of discos and organized parties (bals). Since land on this smaller island is at a premium, zebu cattle, bound for slaughter, are carried from the Sambirano on the boat’s lower decks to supply meat for local inhabitants and for the lucrative tourist industry.

Today the conditions of the roads of this region are exceptionally good, since Madagascar is a nation that relies heavily on air and sea travel and invests little in road development and repair. Only in the past few years have these been all-weather roads. For a long time they were little more than footpaths, and until very recently only cars equipped with four-wheel drive were able to traverse them with ease. Within the last decade these roads have been improved considerably, so that now they are graded and large portions are paved. Local Sakalava dislike this new development, fearing that good roads will only make it easier for migrants or vahiny from other parts of an otherwise extremely economically depressed nation to come here to settle.

The south-north road that runs through Ambanja is a major national route, and it is by means of this road that the majority of vahiny arrive in the valley from points south. In the past, the trip from the south was long and difficult: many traveled more than one thousand kilometers by foot in an effort to find work. Today most migrants come to this region by bush taxi (taxi-brusse) by way of the national capital of Antananarivo in the central highlands or from the southern coastal city of Mahajanga. There are also airports in Nosy Be and Diégo with runways equipped for landings by Boeing jets, and Ambanja has a small airport where propeller-driven Twin Otters land twice each week. The northern part of this route is used frequently by those who travel to the provincial capital of Diégo either for business or pleasure. This road is paved as far as the town of Ambilobe (102 kilometers). This town, like Ambanja, is a county seat, a slightly smaller commercial center, and the residence of the king of the Antakarana who inhabit the neighboring Mahavavy region. Near Ambilobe is another large sugar cane enterprise called Sirama.

Approaching Ambanja from the south, one passes through fields of cocoa and coffee, crosses over the bridge that spans the Sambirano River, and then enters the oldest quarter of town. Here one finds the bazarbe, a wide and circular-shaped daily market. Ambanja is a major urban center of the northwest, serving as a county seat for the region, and offers many goods, services, and sources of entertainment that are unavailable in other regions of the island. This is immediately evident in the bazarbe. This market is surrounded on all sides by the concrete structures that house the shops and residences for Indian, Arab, Chinese, and Comorean merchants who first came to Ambanja after hearing of the development of the area by foreign planters. Behind the merchants’ shops, just to the west and beyond the bazarbe, the top of the Indian mosque can be seen. To the east is the petrol station, a soccer field, and the town cemetery, where a wall separates Christian and Muslim graves. On a Sunday afternoon in the dry season, the music at a morengy or boxing match can be heard playing in the distance. On most evenings a crowd will gather before City Hall to gaze at the color television suspended above the front door, while lovers and groups of teenagers stroll by or pause momentarily on their way to a disco or to the cinema. The bazarbe is also one of two stations for the bush taxis, the small Peugeot pickups that serve as the main form of ground transportation in Madagascar. Here in Ambanja they dart madly between the bazarbe and the Tsaramandroso, the site of the Thursday market at the other end of town (see plate 2).

The enterprises are a dominant force in the daily lives of all people living in the Sambirano, whether or not they are employees. The town itself is flanked on the east and west by enterprise lands that, in some places, border streets (see figure 2.2). As a result, urban growth has pushed south along the banks of the river and north following the main road out of town. There is some order to this development, for as neighborhoods expand they continue to form a rough grid of wide, unpaved, and generally very dusty streets. Nevertheless, for the most part, settlement patterns have been sporadic, haphazard, and random. Apart from the old quarter by the bazarbe, it is difficult to categorize any one neighborhood as rich or poor or populated by a majority of any one ethnic group. The structures of the town reflect this diversity. There are a few impressive and imposing two-story villas plopped down among neat two- or four-room houses—made either of corrugated tin or fiber from the traveler’s palm (Ravenala madagascariensis)—which rest on concrete foundations or on stiltlike wooden legs. Children, chickens, goats, geese, and ducks wander in and out of houses or root about in the grass beneath the shade of mango, banana, coconut, and jackfruit trees.

Since the Sambirano lies in northern Madagascar near the equator (just north of 14°S parallel), it is always hot in Ambanja. There are only two seasons here: the dry season from May to October, and the wet, hot season from November to April, which is accompanied by the threat of cyclones from December to February. Everyday greetings often include a commentary on the weather: in the dry season, “Ah, the heat, the dust!” (mafana é! misy poussière!), or, in the wet season, “Ah, the heat, the mud!” (mafana é! misy goda!). Most people prefer to walk along the main street, beneath straight, tall shade trees. These trees were planted earlier in the century, and are the same varieties which are used on enterprise lands to shade cocoa from the heat of the sun (again, see plate 1).^[1]

It takes about twenty to thirty minutes to walk, at a leisurely pace (mitsangantsangana), from one end of town to the other by way of the main street. Lining this street are small shops that sell colorful fabrics and household necessities; hotely, or the small bars and restaurants; the colonial style, wooden homes that belong to select royal families; the impressive and freshly painted structures of the Peasant’s Bank and the county seat; a variety of mosques and Protestant churches; and the courthouse, gendarmerie, and prison. All of the enterprises have an office centrally located in Ambanja, since the directors prefer to live in town. These offices are generally placed on the ground level of an imposing villa. Here on the main street there are also offices and agricultural experimental stations of a variety of government ministries, so that as you follow the main street you can see cocoa growing and can smell the pungent blossoms of ylang-ylang.^[2]

All along this main street women and young girls sell snacks to passersby, and there is usually a considerable gathering of their stands placed strategically so as to draw clients from the local schools and the municipal hospital. Near the latter is the private occupational health clinic established for the workers and their families and funded by member enterprises and local businesses. Across the street are the grounds of the Catholic Mission school and printery and the tallest building in town, the cathedral. Behind it is the Catholic hospital, which, when completed, will provide the first operating room facilities in the area.^[3] Near the other end of town is the Thursday market, (again, see plate 2) where one finds a cluster of oxcarts, yet another bar, the office of Air Madagascar, and many houses of all sizes extending to the town border and beyond. If one visits this second market early in the morning or at dusk, bright red flatbed trucks from the enterprises will have stopped there and men and women—some with small children—will be climbing in or out on their way to or from work.

An Economic and Political History of the region

Precolonial History: The Bemazava-Sakalava^[4]

When French military troops arrived in the Sambirano in 1896, this valley was indisputably the territory of the Bemazava, the northernmost dynastic branch of Sakalava speakers, who today form the fifth largest of the eighteen officially recognized ethnic groups of Madagascar (Covell 1987: 12).^[5] The Sakalava as a whole are organized as a collection of kingdoms occupying the island’s west coast, having been formed as a result of disputes over succession and the subsequent movement north by new founding dynasties.^[6] The Bemazava-Sakalava trace their origins to Boina, the royal and sacred capital that lies near Mahajanga on the central part of the western coast. The Bemazava dynasty was established by Andriantompoeniarivo (see figures 2.3, 2.4).^[7] According to oral tradition, he left Boina following a dispute over royal succession. Andriantompoeniarivo was accompanied by his followers and by a powerful moasy (HP: ombiasy) or herbalist named Andriamsara. As they traveled they carried with them royal relics, and among these was a container of sacred water. When the party reached the river valley, however, they discovered that the container had run dry. While trying to decide whether or not to turn back or to use the local river water to replenish their supply, Andriantompoeniarivo is said to have remarked, “it makes no difference, they are each/both water” (samy ny rano). Thus the Sambirano River was named, and it is here where they chose to settle.^[8]

In the nineteenth century, just prior to French conquest, the Bemazava lived in small villages that were scattered throughout the Sambirano. They farmed plots of dry rice and manioc, banana, and other fruits, but the majority of the land of this fertile alluvial plain was used to graze herds of zebu cattle (see Dury 1897). The Bemazava of the Sambirano appear to have been united into a loose confederation under a common ruler (ampanjakabe) who was the living successor of the tromba or spirits of the royal ancestral dead. An important royal duty involved the ruler serving as the representative for his or her living subjects in ritual contexts. It was the ruler, with the help of assistants (male assistant: ngahy; female assistant: marovavy or ambiman̂angy) and tromba mediums (saha), who invoked the royal spirits. The ruler also served as a mediator and judge in secular disputes. The Bemazava ruler lived in the coastal village of Ankify, while the island of Nosy Faly served as the sacred ground where royal dead were entombed and where the mediums for the greatest of the royal tromba spirits resided. On Nosy Faly the first Bemazava king, Andriantompoeniarivo, was laid to rest, along with his successors and other members of the royal family (ampanjaka). Today both Andriantompoeniarivo and Andriamsara are regarded by the Bemazava as their founding ancestors. They are the most important of the local Bemazava tromba spirits, and their mediums live on the sacred island of Nosy Faly. As will become clear in Part 2, these spirits and their mediums wield much power in the Sambirano.

Prior to 1896, the Bemazava were well aware of the existence of the French, who had been active in the north for approximately sixty years. The queen of another branch of Sakalava—the Bemihisatra of Nosy Be—had previously invited the French to her island. In so doing, she sought to gain a powerful ally against her enemy, the Merina of the central high plateaux (see Mutibwa 1974 for a discussion of Merina expansion in the eighteenth and nineteenth centuries). She extended this invitation to the French only after she had failed to acquire firearms from the Sultan of Muscat: giving way to his conditions, she and other Bemihisatra royalty agreed to convert to Islam, but the Sultan only sent velvet hats, not arms (Dalmond 1840). In 1840 the French established their first permanent settlement on Nosy Be, and the members of this party included Jesuit missionaries and planters. The latter started sugar cane plantations that still exist today (they are administered from the town of Djamanjary). Although the Sambirano was not far from Nosy Be, contact between Bemazava and Europeans remained fairly limited due to the relative difficulty of traveling to the interior, and to the reputation of the fierceness of Bemazava warriors.

The Establishment of French Control

The end of the nineteenth century marks the beginning of the colonial era in Madagascar. In 1895 Madagascar was declared a protectorate and, in the following year, a colony of France. Under the direction of the military strategist Général J. S. Galliéni, the Merina monarchy was abolished and the queen, Ranovalona III, exiled to Réunion. French troops marched from one end of Madagascar to the other as part of Galliéni’s pacification program. Galliéni sent military expeditions to comb the entire island, collecting enormous quantities of data of strategic import, making note of relevant social and cultural institutions, and eventually setting up a network of military posts. From records made by commanding officers, it appears that the Sambirano was visited twice, one battalion moving west from the Tsaratanana ridge to Nosy Be, the other moving up from the ancient northern Sakalava capital of Boina (Boucabeille 1897; Brown 1978: 236ff; Dury 1897; Galliéni 1900, 1908; Raolison 1966; see also David-Bernard 1943). Lieutenant S. V. Dury, who led the battalion from the east, described the Sambirano as follows:

Dans ces plains [Sambirano and neighboring ones] l’herbe est abondante en toute saisons, ce sont des pâturages magnifiques, qui nourrissent les troupeaux les plus beaux et les plus nombreaux qui existent de la Mahajomba [Mahajanga] au Sambirano. Les boeufs y sont superbes et donnent de 100–150 kilos de viande; on les vends sur place 20 Francs du maximum.…Cette region du Sambirano est donc aussi très riche. Le débouché vers la mer, qui donne dans la large baie de Passandava, est commode et la construction d’une route simplement. (Dury 1897: 443, 445)

It was in 1896 that the first military post, under the command of Captain Verdure, was established in the Sambirano area in the village of Ambato. Ambato lay near the coast en route to the sacred island of Nosy Faly, the location of the Bemazava-Sakalava royal tombs. Within a few years, military men—some with families—had started to farm in the area, and by the turn of the century the Sambirano had become attractive to planters from the neighboring islands of Nosy Be, Réunion, and the Seychelles. In 1903 the French moved the military post to a more advantageous location inland and upstream and situated on a high hill that provided a spectacular view not only of the river but of the entire valley. Placed strategically at a major crossroads, it is around this post that the town of Ambanja eventually grew. From the buildings constructed by the French one may now look down upon the oldest quarter of town which encircles the bazarbe. Throughout the colonial era the post’s buildings served as the residence for French colonial officers; since Independence they have been used to house local Malagasy civil authorities.

By the late 1890s, the government of France began to grant land titles to foreign-born planters, titles that were authorized by Galliéni and issued from the national capital of Antananarivo. Within a few years the majority of the land that previously had been used and occupied by Bemazava had been transformed into large private plantations. Much of this land was acquired through purchases of communal grazing lands, transactions the Bemazava misunderstood. Many Bemazava were forced out of the choicest areas of the valley and onto indigenous reserves (FR: reserves indigènes), often pushed up against steep hillsides that were difficult to farm and unsuitable for grazing cattle. This policy of designating territory as indigenous reserves was unique to the northern and western provinces of Antsiranana (in which Ambanja is located), Mahajanga, and only a few areas of the high plateaux. Furthermore, Nosy Be and the Sambirano each had over twenty reserves, while Mahajanga had only two (Service Topographique, Nosy Be, n.d.). This period is also marked by a large exodus of Bemazava to the drier Mahavavy area that lies one hundred kilometers to the north in Antakarana territory. By the 1920s the Sambirano Valley had developed into one of the most prosperous areas of Madagascar. Merchants with origins as diverse as southern China, southern India, and Yemen came and settled permanently in the town, establishing a large market (the bazarbe) beside the river and just below the military post.

The Development of the Sambirano

It is important to realize that the development of Ambanja into a town and commercial center occurred as a result of, first, French occupation of the region and, second, the subsequent activities of the early plantations. Although it is possible that there was a Bemazava village at this location prior to the arrival of the French, a surveyor’s map drawn at the turn of the century depicts simply a military post with a flagpole to mark it. As the town has grown, it has developed along a clear axis, sandwiched between the fields of the large plantation fields of the enterprises that flank it to the east and west (figure 2.2).

The activities of a number of other foreigners reveal that Ambanja was growing rapidly into an important cosmopolitan center. The Catholic Holy Ghost Fathers from France were soon active here: in 1921 they performed their first Catholic marriage, and in 1936 they completed the construction of the cathedral. By 1927, surveyors were employed to measure property systematically and lay down landmarks for both private individuals and the owners of plantations, who thus acquired official deeds to their lands. An early road map of Madagascar, published in 1938, categorizes Ambanja as a commercial center, complete with a post office and lodging for weary travelers.

The development of the Sambirano into a plantation area was shaped primarily by two men, Louis Millot and Guy de la Motte St. Pierre. By 1905, each had acquired land grants from France which together covered nearly all of the Sambirano Valley. In the next few years Millot and de la Motte St. Pierre also bought out other smaller planters when their efforts to farm failed.^[9] By the 1920s a third plantation had become active in the Sambirano, one that is now known as the Compagnie Nosybéenne d’Industries Agricoles (CNIA). This was an extension of Djamanjary Sugar in Nosy Be.

Both Millot and de la Motte St. Pierre began by planting what was locally available. Millot chose his fields carefully, and it is said that he sent soil samples back to Europe for analysis before making large investments in his lands. He began by planting coconut palms, rice, and manioc for the production of tapioca. By the the 1930s, he had imported cocoa plants from the Ivory Coast, and today cocoa remains the dominant crop on Millot lands. In more recent years, this enterprise has also planted some pepper, as well as fields of perfume plants. De la Motte St. Pierre also started with coconuts, but soon switched to sugar. With the arrival of CNIA in the 1920s, sugar remained the dominant crop of the Sambirano until the 1940s.

After World War II, when the price of sugar fell on the world market, it was decided that the sugar companies on Nosy Be, near Ambilobe, and in other regions of Madagascar were sufficient to satisfy the exports needed from Madagascar. As a result, plantations in the Sambirano cut back on sugar production, cocoa production was expanded, and coffee was introduced as a new crop. Eventually cocoa and coffee replaced sugar altogether as the major export crops of this region. Although French planters maintained a monopoly over cocoa production, by the 1950s the government was also encouraging private farmers who had small plots to grow coffee as a cash crop. Tapioca production continued to be a major industry in the Sambirano until the 1970s, when the last factory shut down in response to a diminishing market. Meanwhile, within the last two decades, cashews—which grow wild in the drier areas to the north and west—have joined coffee and cocoa as one of the region’s three major export crops. As will become clear, coffee and cashews are important in Bemazava-Sakalava constructions of their local history. The meanings associated with common historical experience are played out through tromba possession, where coffee and cashews figure as prominent symbols (see Part 2).

Following the Socialist Revolution in 1972, all of the large plantations except E. Millot were seminationalized.^[10] Millot and the now combined businesses of CNIA/SOMIA (Sociéte Malgache d’Industrie et d’Agriculture) are the largest and most impressive enterprises in the Valley.^[11] These have grown out of the original farms of the two men, Millot and de la Motte St. Pierre. The plantation E. Millot had more than seven thousand hectares at its height in the 1940s, although it is now down to two thousand. CNIA/SOMIA, which possesses much of de la Motte St. Pierre’s original lands, at one time boasted close to a million hectares.^[12] In the Sambirano there are also a number of smaller enterprises, small private farms, and companies (referred to as concessions) which buy agricultural produce and prepare them for export.

The labor requirements of the cash crops grown in this region demanded a large and reliable workforce, since each had specific needs for both seasonal and general year-round upkeep and care. The plantations in other regions nearby—such as Djamanjary Sugar of Nosy Be—recruited prison labor from the south in the early 1920s.^[13] In contrast, the Sambirano quickly became well known throughout the island for the availability of wage labor, and so the plantations of the Sambirano did not find it necessary to engage in an aggressive recruitment of workers from other areas of the island.^[14] Unlike the Bemazava, who had (and still have) a reputation for being a fixed population, peoples from the high plateaux and the arid south left their homelands and settled in the Sambirano. For example, Betsileo—who are famous in Madagascar for their skills as rice farmers—came and settled on CNIA property as land tenants, developing the area around Antsakoamanondro, just north of Ambanja, into fertile irrigated paddy land. Antandroy, Antaimoro, and other peoples of the economically depressed south and southeast also came to work as manual laborers in the fields.

Effects on Land Tenure

Although many of the Bemazava had been alienated from their original territory through relocation, this same policy enabled them to maintain access to arable land throughout the twentieth century. In addition, numerous small villages and private family plots remained scattered throughout the Sambirano, interspersed with plantation lands. A perusal of colonial property records reveals that subsequent transactions of land involving Malagasy were uncommon during the colonial period, so that migrants were unable to acquire land when sales did occur. Instead, priority was given to members of the Bemazava royal family. This was an important trend during the colonial period, whereby the French granted royalty special privileges, assuming that if they could control the Bemazava rulers they could control their subjects. For example, the father of one of my informants was a member of the royal family and was among those royalty (ampanjaka) who were in the direct line of succession; in addition, he was a favorite among a number of French colonial officers. As a result, he was able to purchase several large plots throughout the Sambirano during his lifetime, including one in the 1930s which was nearly fifty hectares in size. Throughout the colonial period, pieces of land of this magnitude generally only went to Bemazava royalty and foreign-born planters.

To some extent, French colonial law favored all local Bemazava, and this enabled commoners as well to maintain control over valuable parcels of land in the valley, albeit smaller in size than those owned by royalty. Among those Bemazava commoners who remained landed, a bilateral rule of inheritance prevailed. Land was generally divided fairly equally among the spouse and male and female offspring when the parent (biological or classificatory) was either too old to farm or had died. For example, two of my informants (one male, the other female) owned two and four hectares of land each, where they grew primarily dry rice. Each had inherited their land from their mothers. The man was an only child whose father had died when he was very young, and previously his mother had inherited land, along with her mother and siblings, when her own father had died. My female informant could trace land inheritance among her kin through six generations. Each time someone died their land was divided fairly equally among their children (sometimes including favorite classificatory offspring) and their surviving spouse. In her kin group land was also acquired in four other ways: as gifts from royalty for performing royal service (fanompoan̂a), as previous residents of indigenous reserves, by periodically purchasing plots of two to three hectares each, or as a result of Napoleonic law, which honors the land rights of squatters who make productive use of land. These methods of land acquisition helped to offset the common trend whereby the size of inherited plots grows smaller with each generation.

Today, tera-tany continue to be favored over vahiny as a result of government reforms following Independence in 1960 and, to a lesser extent, from the sales or confiscation of large private holdings through the nationalization of large private estates that followed the 1972 Socialist Revolution. In the Sambirano, those who own land are truly at an advantage, for even plots rejected by foreign planters give high yields. Land, however, continues to become increasingly scarce, and, as the story below illustrates, it is a source of much contention among kin.

Zaloky’s Homestead

Zaloky estimates her age to be about fifty years old, although she looks as though she could be another twenty. She is Sakalava tera-tany, born and raised in an area which, in her childhood, lay on the outskirts of town; within the past twenty years, however, it has become a thriving neighborhood. She lives on a small patch of land (approximately one quarter of a hectare) which she inherited from her widowed mother. In the past she had a garden on an adjacent plot of land but now she is boxed in on all sides by new houses, several of which are concrete. She now farms a small field in her mother’s native village, which lies eighteen kilometers from town. Zaloky was married at age sixteen and she had four children. At age thirty-five her husband died; five years later she remarried (by common law) a Tsimihety vahiny named Marcel who would come each year to the Sambirano for the harvest season. He now works off and on at the enterprises. Because he is old and suffers from a bad back, however, it is difficult for him to find steady work. In 1987 he also worked as a night watchman.

The homestead on which Zaloky and Marcel live is dusty and in disarray; rather than having a neatly swept yard bordered by flowers, like most Sakalava homes, it is littered with metal scrap and old papers that get swept up by the wind and caught in the fence. Marcel sells these items when he can as a way to supplement their income; Zaloky also has two scrawny chickens that she keeps so she can sell the eggs. Their income is meager at best, and nearly every day they beg for food from their neighbors and the Lutheran church that is behind their house.

Her homestead is one that is a source of much conflict: ever since her first husband’s death her children have fought to have rights over this land. Since she married Marcel, two of her children (a son and a daughter) have filed a case in court to take it away from her. As Zaloky explains, her children believe that Marcel only married her for her land, and she states flatly that she certainly would rather give it to him than to them! Within the last year a disco has also been built next door, and the proprietor threatens them several times a week, saying he wants them to die so he can expand into their yard. He has already tried to tear down the fence, and so each day Zaloky and Marcel check to make sure it is in its proper place.

Zaloky’s problems are extreme: a tale such as this, involving children seeking to evict their mother from her land, is one most Sakalava in Ambanja would listen to in disbelief. Nevertheless, it is instructive, since it reveals the severity of tensions underlying the scarcity of land in the Sambirano. The fact that her second spouse was not tera-tany, but a migrant, brought familial conflicts to the fore. Zaloky’s story is one that threads its way invisibly through the following chapters, for she had at one time been a medium for a prestigious tromba spirit, and it was through this work that she met Marcel, who had often consulted her in times of personal crisis. The continuation of her story will appear in chapter 10, for ten years ago she converted to Lutheranism, a choice that (as the next chapter will illustrate) is an unusual one for Sakalava. As tensions with her children, and now her neighbor, have worsened, Zaloky has recently decided to give her land to the church, and the pastor has solicited the national center for funds to pay the surveyor’s and lawyer’s fees that will enable them to make a formal and legitimate claim.

Local Power and Reactions to Colonialism

Local Authority and Power

Today the power structure of the town of Ambanja is multifaceted and complex, the town serving as a center for both secular and sacred power. There are a number of institutions that operate here. First, officially recognized power lies with the state, a structure that originated during the colonial era and which continues today under the independent government of Madagascar. It is composed of a hierarchy of national, provincial, county, town, and neighborhood authorities. There are also other, more informal, power structures at work in Ambanja. The state-owned enterprises wield much control in the Sambirano, since they play an extremely important role economically, one that affects the everyday lives of local people in many ways. Even though directors are state employees, they exercise great freedom in making decisions that affect the local economic and social order of the Sambirano. For example, much of the town’s funding for celebrations, transportation, roadwork, and so forth comes from the enterprises, with donations made by the directors themselves. The patron-client relationship between the enterprises and the townspeople is so strong that a common response to an important family or personal problem is to turn to one of the directors for assistance, whose responses can be either rewarding or devastating. Likewise, religious groups—be they Islamic, Protestant, or Catholic—provide support through divine injunction and financial assistance. The Catholic church is by far the wealthiest (as is reflected by the extent of its landholdings in town), and it is also the most influential. The Catholic church also provides access to rare opportunities beyond Ambanja and outside of Madagascar, particularly through education.

Prior to French occupation of Bemazava territory, political power and authority lay almost exclusively with the local royalty, that is, with the ruler and his or her advisers—be they living or dead. As will become clear in Part 2, by far the most important role was played by the tromba spirits and their mediums on Nosy Faly. With the arrival of the French, however, control shifted to colonial administrators and planters. This shift in the power base led to the relocation of the Bemazava ruler from the ancestral village of Ankify to the commercial center, Ambanja. This occurred during the reign of Tsiaraso II (Andriamandefitriarivo), who was the father of the present ruler, Tsiaraso III. Tsiaraso II reigned from 1945 to 1966, and from conversations with informants I assume that the move occurred early in his career (figure 2.4). Furthermore, French policies were designed deliberately to undermine the royal family’s power. The most effective effort involved recruiting royal children to be schooled at the local Catholic Mission. These children were trained to be civil servants for the colonial administration, whose loyalties and interests lay with the French and not their own people (cf. Feeley-Harnik 1991b: 137).

Although royal power has diminished considerably (again, cf. Feeley-Harnik, 1991b), the present ruler (ampanjakabe) remains an important figure in the lives of the Bemazava. He is often preferred over the local court as a mediator in private disputes, and as the official living guardian of the local tanindrazan̂afa, he is the first authority one must consult prior to any further development of local land (see chapter 6). Spirit mediums also seek his blessing as well as his guidance in matters involving royal ancestors. The extent to which royalty more generally may become involved in the reassertion of royal power is exemplified in the actions of an Antakarana ruler to the north, who has become heavily involved in the revival and reinvention (Hobsbaum 1983) of royal rituals, which may draw a crowd of thousands of participants.

Local government officials honor the importance of royalty at public (state) occasions. A respect for Malagasy customs (HP/SAK: fomba-gasy; also referred to as fombandrazana/fombandrazan̂a or “customs of the ancestors”) remains strong in Madagascar. As part of malagasization, the national government has encouraged and institutionalized the observance of fomba-gasy at public events. Tromba mediums in particular have become very active, since the permission of the spirits must be sought through them, for example, prior to the naming of a local school after a past ruler or at the beginning of each season before boats of a state-owned fishery enter the waters near Nosy Faly (see chapter 6). Nevertheless, one has a sense of cooptation of living royal authority by the state, ironically illustrated by the fact that the present Bemazava ruler is Catholic and works at the county office as a tax collector.

Resistance and Revolt in the Sambirano

Bemazava reaction to the French occupation and the eventual transformation of the landscape was neither accepting nor passive. As Stoler’s work in Sumatra’s plantation belt illustrates, what may be presumed to be a peaceful occupation, settlement, and development by colonial forces may in fact be marked by acts of violence by the seemingly powerless against the powerful (Stoler 1985). At first glance, for a number of reasons it does not appear that French activities in the Sambirano were greatly hampered. First, it would have been difficult to present a united front against the French, the Bemazava being relatively few in number and scattered throughout the valley. Second, although Bemazava warriors were known to be fierce, their access to firearms was limited. Third, they regarded the French as their allies, since they had conquered an old enemy, the Merina.

Nevertheless, acts of violence against the French did occur. An incident cited frequently by the Bemazava involved a private dispute that quickly became a public—and political—one. Shortly after the French military post in Ambato was established, a group of Bemazava men were captured and imprisoned for having killed a Frenchman. The Frenchman had been living in the royal village of Ankify and was married to a Bemazava woman of royal descent. Jealous of his wife’s interactions with other men, he had confined her to the house. Since she was royalty (ampanjaka), other Bemazava regarded this as offensive and killed the husband. Captain Verdure had the prisoners beheaded in order to discourage future violence against French citizens. This episode is still remembered with great bitterness by the Bemazava. Much of their anger focuses not only on the French but also on Senegalese soldiers who were brought there by the colonial government to help maintain order. These soldiers are said to have tried to eat the prisoners’ bodies until Verdure stepped in to prevent such an outrageous act.^[15]

The Sambirano became a politically charged area in 1947, a year that today is commemorated in Madagascar as a time of revolution and early nationalism. The front of resistance was located on the east coast of the island and began in March 1947. Malagasy in other areas of the island subsequently staged their own revolts against the French (see Covell 1987: 26; Rajoelina 1988; Tronchon 1974). Following the lead of revolutionaries elsewhere, inhabitants of the Sambirano behaved disrespectfully to French as they passed in the streets and the workers of some plantations went on strike. By November 1947, a small rebel group had formed in the Sambirano with a camp established in a village southwest of town, where they secretly manufactured knives and guns. Within a few months, however, the leaders were captured and imprisoned, putting an end to a potential uprising.

Resistance may also take more insidious forms, which are culturally more appropriate yet perhaps less effective against foreign invaders (Scott 1985). The Ramanenjana or “dancing mania,” which occurred in the Merina capital of Antananarivo in 1863, is perhaps the most large-scale resistance movement against foreigners involving religious forces. At this time the streets were crowded with hundreds of people possessed by the dead Queen Ranavalona. The possessed often disturbed and even attacked Europeans, many of whom were Protestant missionaries (see Davidson 1889; Sharp 1985).

Without doubt, during the colonial period Malagasy throughout the island appealed to the ancestors and other spiritual forces for assistance. In the Sambirano, it is certain that Sakalava sought to cause harm to or drive these foreigners from their homeland. They appealed to specialists such as herbalists (moasy), diviners (mpisikidy ), and tromba spirit mediums, making use of Malagasy medicine, or what Europeans referred to as magic (fanafody, fanafody-gasy). According to one informant, at the time of the 1947 revolt in the Sambirano, dead and living forces united against the colonial government; Malagasy who sympathized with the French were victims of “occult forces” (forces occultes) including tromba and other spirits, and fanafody. As will become clear in subsequent chapters, the use of fanafody continues to be a powerful way to control events in everyday life. Tromba mediums in particular play a special role in this manipulation of the spiritual realm. In addition, tromba as an institution has been a source of Sakalava pride and identity, as well as a means of Sakalava resistance to foreigners. During this century it has operated as a reminder of things past. It has also provided an idiom through which to critique contemporary experiences.

The Social Construction of Work

Throughout the twentieth century, by far the most significant form of resistance, in terms of its impact on European attitudes and policies, has been the Sakalava’s refusal to work as wage laborers. Although a requirement under the colonial regime was ten days per year of enforced (corvée) labor, ownership of land enabled many local Sakalava to avoid having to work full-time for the plantations (for a discussion of similar policies throughout the island see Thompson and Adloff 1965, chap. 23). Furthermore, many Bemazava were able to pay the mandatory head tax through cash-cropping or by selling their cattle. As a result, early in the century the plantations began to hire laborers who came from other areas of the island. To understand the logic behind Sakalava actions, it is necessary to analyze the meaning of work in their society.

Feeley-Harnik, in her discussion of slavery and royal work among the Sakalava in Analalava, states that Sakalava distinguish between two types of work: asa, or work owed to kin, and fanompoan̂a or “royal work or service that is part of their politico-religious and economic obligation to the monarchy” (Feeley-Harnik 1984: 3). In Ambanja (and elsewhere in Madagascar), asa also has other meanings: as a noun, it can mean “task” or “project,” and the verb miasa, which is often translated “to work,” is used more generally to mean “to be occupied” or “active” or “to be working on a task, a project.” For example, if a woman is doing the laundry, one says “miasaizy” (“she is working”). Beginning in the colonial era, asa took on additional meanings, as a result of the penetration of a capitalist economy and its associated labor relations. The French used asa to mean wage labor, as well as enforced, mandatory work that involved clearing roads, for example (cf. Feeley-Harnik 1991b, especially p. 349).

French concepts of work ran contrary to Sakalava ones. Since, for Sakalava, work was something that one did out of loyalty to kin and local rulers (Feeley-Harnik 1986, 1991b), requirements imposed by the French government and plantations negated local custom. From a Sakalava point of view (and certainly from a French one as well) fulfilling these work requirements was a sign of loyalty to the colonial administration. Feeley-Harnik (1984), in her discussion of slavery and royal work, argues that this action by the French was deliberately designed to undermine local royal authority and to disrupt the local sociocultural order. (For other discussions of Malagasy conceptions of work see Decary 1956; on the implementation of French colonial policy elsewhere in Africa see Crowder 1964 and Gifford and Weiskel 1974.)

In the Sambirano these efforts were only partially successful. Although royal authority has certainly diminished, many people still look to the local Bemazava king as an adviser and mediator in local disputes. It is also acknowledged that his authority is legitimated by the tromba spirits. Those who today turn to royalty (living or dead) for guidance are not only local Bemazava-Sakalava, but also many others who have become involved in tromba possession and serve as mediums for these spirits. The local government often must rely on the support of the king if it wishes to encourage participation by community members in local activities. On International Workers’ Day (May 1), for example, only the blessing of the king ensures that townspeople will help to repair the roads in town—this time as a sign of support for the present socialist state.

Sakalava resistance to work—in the form of wage labor—is often cited today by employers, who speak of them as “lazy” (kamo) or as too independent and proud to be willing to work for others. Sakalava youth express a preference for working in fields owned by kin to wage labor, be it manual or more highly trained supervisory positions. As the chapters in Part 2 will show, tromba possession also provides mediums with a means through which they may resist wage labor, particularly at the enterprises. A sentiment that many Sakalava express is that in such work there is no freedom: the hours are long, the pay is too low, and the fruits of their labor belong to someone else (see plate 3). The refusal to work was and still is a strong form of resistance to capitalist discipline. Such relations involve, first, a readjustment to clock time (Thompson 1967),^[16] and, second, the displacement of reciprocal relationships by having services paid for with cash, both of which Sakalava despise. Throughout this century Sakalava have resisted engaging in such relations with non-Sakalava. In the past, these were foreigners. More recently they are Merina, who now form the majority of managers at the enterprises. This action, however, has its price: after Sakalava refused to work for foreign planters they witnessed the inundation of their territory by a variety of non-Sakalava migrants.

Notes

3. National and Local Factions: The Nature of Polyculturalism in Ambanja

As a booming migrant town, Ambanja is a mélange, a mixture of peoples of diverse origins, not only from within the boundaries of this large island, but from abroad. This town is a microcosm of factions that operate on a national scale. Of these factions, ethnicity is the primary category, one that overlaps with a variety of other group orientations and allegiances. These are shaped by historically based geographical, economic, and religious differences. Such groupings or identities, when taken in total, reveal the complexity of the tensions operating in Madagascar and, in turn, in Ambanja. An individual’sability to make sense of and become established as a member of one or more groups determines his or her well-being in this community.

The complexity of categories that exist in Ambanja is extreme when compared to other areas where the population is more homogeneous. As a result, in Ambanja and the surrounding Sambirano Valley, defining what it means to be an insider or an outsider is complex. For this reason, the discussion on migration is divided into two chapters. This chapter will address the nature of social and cultural divisions that exist, first on a national scale and then in Ambanja. Chapter 4 provides case studies drawn from migrants’ lives and analyzes those factors that facilitate or inhibit their integration into local Sakalava tera-tany society.

National Factions: Regionalism and Cultural Stereotypes

Ethnic Categories

The Malagasy

Today in Madagascar there are eighteen^[1] (Covell 1987: 12) officially recognized ethnic groups (FR: ethnie,tribu; HP: foko,karazana; SAK: karazan̂a) which are relevant for census purposes (figure 3.1).

The use of the term ethnic group is problematic from an anthropological point of view, since all of these groups are actually subgroups of the general category Malagasy, whose members share common cultural elements such as language and religious beliefs. In other words, the concept of ethnicity is one of perspective and scale. Outside Madagascar, Malagasy are viewed as the dominant ethnic group of the country, and Sakalava, Merina, and other peoples are considered subgroups of this larger category. From a Malagasy point of view, however, Merina, for example, are viewed as the dominant ethnic group, and other non-Malagasy peoples (Arabs, Indo-Pakistanis, Chinese, Comoreans, and Europeans) are grouped separately as “strangers” or “foreigners” (étrangers). Since this study is concerned primarily with Malagasy peoples, I will use ethnic group as the Malagasy themselves do. These ethnic divisions are significant in everyday discourse, as Malagasy use them to define themselves in relation to each other. Ethnic groups also overlap with other categories based on geographical, economic, and religious differences.

Ethnic categories have changed over time. As Covell points out, they are flexible constructs, and although they are in part a reflection of changes in the political climate of Madagascar, it would be false to conceive of them solely in these terms:

This form of identification hardly constitutes a key to Malagasy politics. The groups themselves are riddled with internal subdivisions and several are, in fact, political constructions created from small groups in the eighteenth and nineteenth centuries: the Merina, Sakalava, and Betsimisaraka are the most important of these. Others, such as the Betsileo and Bara were first grouped together as administrative subdivisions of the nineteenth-century Merina empire. (1987: 12)

The French also made use of these categories. They did not begin systematically to take official censuses, with Malagasy broken down into different ethnic groups, until 1949, following the 1947 revolt. This practice was continued by the government of the Malagasy Republic after Independence (1960) up until the time of the Socialist Revolution in 1972. For approximately a decade afterward, no census information was collected (publications in general came to a halt in Madagascar at this time). In the Sambirano Valley it is only in 1985 that new efforts were made to gather census data in preparation for national elections that occurred in early 1989. On these recent censuses, the ethnic categories no longer appear, although logbooks kept of Ambanja’s neighborhood residents, for example, still make note of ethnic affiliation. In everyday discourse these categories are used by Malagasy to label one another.^[2]

Today certain factors unite members of each ethnic group: a shared dialect of Malagasy; similar religious customs, most notably in regard to mortuary rituals and a strong regard for local ancestors; observation of fady or taboos; characteristic regional dress; economic activities; and affiliation with a specific geographical territory (figure 3.2). These categories and their associated characteristics are used by Sakalava as they define themselves in opposition to others. Ultimately, these differences define boundaries between insider and outsider. Among some peoples the boundaries are fluid, while among others they are very rigid.

Non-Malagasy Strangers (Etrangers)

In addition to Malagasy speakers, there are also a number of minority groups (see Vincent 1974: 377) that consist of non-Malagasy peoples who have settled on the island. The largest of these groups are Arabs, Chinese, Comoreans, Europeans (especially French and some Greek), Indo-Pakistanis, and peoples of mixed heritage from the neighboring islands of Réunion and Mauritius. No recent statistics are available for the size of these populations, and so it is very difficult to estimate their numbers. This is in part due to recent political events. The majority of Europeans fled the island following the Socialist Revolution in 1972. Comoreans and Indo-Pakistanis have also fled periodically, since they have been the targets of violence in the last decade or so. Out of a total estimated population of 9.9 million for the entire island (figure for 1984; Covell 1987: xiii), population estimates for each group of étrangers are: 10,000 each of the Indo-Pakistanis and Chinese; 15,000 to 20,000 Comoreans (Covell 1987; 84–85); and 12,000 French (Bunge 1983: 49; for more details see Covell 1987: 84–85; on the Chinese see also Slawecki 1971 and Tche-Hao 1967).

Malagasy Ethnic Groups: How Difference Is Perceived

A variety of factors delineates boundaries between ethnic groups. To illustrate how these operate, I will briefly discuss two such factors: physical differences and the fady or taboos. Sections that follow provide discussions on other distinguishing characteristics, such as territorial affiliation, economic specialization, and religion.

Highland and coastal peoples (and, in turn, specific ethnic groups) are distinguished from one another by dialect and phenotype, reflecting the diverse historical and cultural origins of the Malagasy. As Bloch explains:

Madagascar has always been considered an anthropological oddity, due to the fact that although geographically it is close to Africa the language spoken throughout the island clearly belongs to the Austronesian group spoken in Southest Asia; more particularly Malagasy is linked to the languages spoken in western Indonesia. These surprising facts are also reflected in the biological and cultural affinity of the people. Although there is much controversy over the relative importance of the African and Indonesian element in the population, there is general agreement that we find the two merged together throughout the island. In some parts one side of this dual inheritance is more important; in other regions it is the other side that seems to dominate. For example, all commentators agree that among the Merina…the Southeast Asian element is particularly strongly marked. (1986: 12)

In addition, style of dress and specific customs serve as markers of difference. Clothing styles vary as one moves from one area of the island to another. In the central and southern highlands, which have a cool, temperate climate, most Merina and Betsileo wear western-style clothes. A style that is considered to be more traditional among these people is the akanjo, a knee-length shirt made of plaid flannel which is worn by peasant men throughout the highlands. Merina women (regardless of class) are also easily recognized since they often wear a white shawl (lamba) draped over their shoulders for formal occasions. The coastal areas are humid and tropical, and all around the rim of the island men and women wear body wraps made of brightly printed cloth (called a lambahoany). Among the Sakalava this consists of the kitamby for men, which is a waist wrap worn like a sarong, and for women, a salova (salovan̂a), which is wrapped around the waist or chest, and kisaly, draped over the shoulders or head. This style of dress is very similar to that worn by Swahili of the East African coast. The far south, which is occupied by such people as the Bara, Mahafaly, and Antandroy, is an arid region. Here men often dress in shorts, a short-sleeved shirt, plastic sandals, and a straw hat, while women wear western-style dresses or lambahoany. Throughout the island different hats as well as hairstyles also serve as markers for ethnicity. For example, the name Sakalava means “[People of] the long valleys,”^[3] while the Tsimihety (“Those who do not cut their hair”) were so named by the Sakalava after they refused to cut their hair in mourning as an assertion of their independence following the death of a Sakalava ruler (Société Malgache 1973: 46, 47).

Fady (taboos) are another important aspect of Malagasy identity. They are widespread yet specific to particular peoples and regions and have been carefully catalogued in books on Malagasy folklore (see Ruud 1960 and Van Gennep 1904; see also Lambek 1992). Fady work at many different levels: all ethnic groups have their own particular ones and smaller groups, such as villages or kin groups, may observe specific fady. Individuals may have personal fady that are determined at birth by the vintana or the Malagasy cosmological zodiac system (see Huntington 1981), or which they have collected over time as a result of sickness or other experiences. Fady may consist of dietary restrictions or clothing requirements, and locations or particular days of the week may have fady associated with them. Complex constellations of fady may surround particular ritual settings or events, such as burials for commoners or royalty. They are also a key aspect of tromba possession.

Among the Sakalava, the following are important examples. Pork is fady for many, requiring that they avoid contact with pigs and their products. Nosy Faly (“Taboo Island,” faly being an older form of the word fady), is Sakalava sacred space that requires respectful behavior from all visitors. One can not enter the village of the royal tombs (mahabo/zomba) on a Tuesday, Thursday, or Saturday, and so visitors who arrive on these days camp out on the beach, waiting for a more auspicious time. In the tomb village (Mahabo) one must go barefoot, and women are prohibited from wearing a kisaly on their heads. Dogs are prohibited and must be killed if they enter. Tortoises, on the other hand, are sacred, and it is fady to harm them. Because of these and other fady associated with the village, visitors must carefully monitor their actions so as not to incur the wrath of the local royal ancestors. Non-Sakalava visitors to Nosy Faly must also conform to local, Sakalava fady. Thus, on this sacred island and elsewhere, when one is among strangers, fady serve as markers of difference. They also operate to control the actions of outsiders. Visitors are expected to respect local fady or risk harming themselves or others. For this reason, Malagasy, when they travel, generally inquire about local fady, and they are usually listed in tourist guidebooks (see, for example, Société Malgache 1973).

Geographical Territory and Ethnicity

The Highlands versus the Coast

Among Malagasy, distinctions are made in reference to geographical areas and their corresponding ethnic groups, which in turn are relevant to the development of political power in Madagascar. There are, first, the peoples of the high plateaux. The most important group here is the Merina. The Betsileo, who live to the south of the Merina, occupy an ambiguous position in the minds of other Malagasy, since they are of the plateaux but are non-Merina people. I know of no general term in common usage that is applied collectively to Merina and Betsileo. Instead, more specific ethnic labels are used. Contrasted to these two highland groups are the côtiers (“peoples of the coast”), a term coined during the colonial era by the French to label all other Malagasy groups, many of whom Covell (1987: 13) points out live nowhere near the coast. The term côtiers is used most frequently by highland peoples and carries somewhat derogatory connotations.

These two general geographical categories have evolved out of politically defined divisions as a result of Merina expansion and subsequent French occupation of the island. During the eighteenth and nineteenth centuries, the Merina conquered much of the island and established a powerful kingdom. By the early 1900s Merina royalty in Antananarivo established an alliance with the British against the French, but this alliance ended when the French conquered the island. Throughout the twentieth century the distinction between Merina and all other peoples has remained significant. With Independence, President Tsiranana (who himself was from the coast) and his party members “tried to mobilize support by portraying themselves as defenders against ‘Merina domination’ ” (Covell 1987: 13). Today, this tension persists. As Covell states, “The conflict cannot be reduced either to ethnic or class competition, but has elements of both” (1987: 13; she in turn cites Tronchon 1975).

Relations between members of these two major geographical groups are characterized by mistrust and racist attitudes. African versus Austronesian origins are a subject of important debate in Madagascar. Merina and Betsileo claim to be the most “Polynesian” of the Malagasy (see Bloch 1971: 1–5) and look disdainfully on coastal peoples, whom they often refer to as being “more African.” These comments are made most often in reference to skin color and hair texture. For example, the term Makoa (which often appears on censuses as an ethnic category, see figures 3.1, 3.2, 3.3) is used as a descriptive (and somewhat derogatory) term for people who have dark skin and kinky hair. Makoa were originally brought to Madagascar as slaves from an area in the interior of Africa that is now part of Mozambique (see Lombard 1988: 88 and Smith 1963 on the “Makua,” especially pp. 257 and 273). Highlanders in general view côtiers as backward and uneducated. Coastal peoples, in turn, express resentment of Merina who were once favored by the French and who form the majority of the population in the capital city. Merina continue to dominate the national political arena and maintain access to the best education, health care, and other services and facilities, and they fill many of the country’s civil service jobs. (For a discussion of these trends after Independence, especially in reference to education, see LaPierre 1966.) Ideas shared by members of each of these groups toward one another also include notions of uncleanliness; concepts of physical beauty, especially in regard to skin color (“white/black”: HP: fotsy/mainty; SAK: fotsy/joby); and a reluctance to intermarry (cf. Bloch 1971: 1–5, 198–201; J. Ramamonjisoa 1984).

The Tanindrazana or Ancestral Land

The most important concept used to define personal and group identity is that of the ancestral land (HP: tanindrazana; SAK: tanindrazan̂a). This may not necessarily be the locality where one lives or even grew up; it is where one’s ancestors are located and, ultimately, where one will be entombed. For all Malagasy, identity is intrinsically linked to the ancestral land. Although it is considered rude to ask what one’s ethnicity is, asking the question “where is your ancestral land” (HP: “aiza ny tanindrazanao?”) will generally provide the same information (see Bloch 1968 and 1971, especially chapter 4).

Even when Malagasy migrate to other parts of Madagascar, the notion of the ancestral land continues to tie them to a particular locale. It is not simply the geographical space, but the ancestors themselves that serve as the locus of identity and that define an individual’s point of origin. As a result, as one moves about the island, one continues to have a strong sense of ethnic identity that is geographically defined. As Keenan states of the Vakinankaratra (a group that is culturally considered to be Merina), “The worst fear of a villager is to travel far from the tanindrazana and fall sick and, perhaps, die alone” (Keenan 1974). The same may be said for the majority of Malagasy, regardless of origin.^[4] Migrants and their children, who are born far from their ancestral land, may continue to invest money in a family tomb that is hundreds of miles away, so that they can be placed there when they are dead.

Economic Specialization

Another system of ranking and categorization is based on forms of economic specialization that, in turn, correspond to ethnic and geographical categories. In a country where, for the majority of peoples, the staple is rice (see Linton 1927), highland peoples pride themselves on their talents as paddy rice farmers. The abilities of Betsileo farmers are a source of great national pride: their paddies are located in the valleys and tiered on the steep hillsides of the temperate regions of the southern plateaux. Coastal peoples of the east, west, and north practice swidden agriculture, where rice is, once again, a major crop (see Le Bourdiec et al. 1969: map no. 51).

The peoples of the arid south (for example the Antandroy, Bara, and Mahafaly) are pastoralists who raise cattle and goats. Although some grow rice, manioc is an important staple. Among many Malagasy—coastal and highland alike—pastoralists are regarded as being “simple,” “primitive,” and “African,” and are said to speak dialects that are unintelligible by the vast majority of Malagasy. Pastoralists are feared by others who say that they are thieves (mpangalatra). This comment is made in reference to (and is a result of a misunderstanding of) the dahalo, or cattle raiders, since among these groups cattle raiding is an important social institution for young men. Antandroy and other pastoralists are taller than many Malagasy and the men often carry long staffs with large blades mounted on one end. As a result, other Malagasy are wary of them.

Conditions are severe in the south, made worse by a drought that has extended throughout the past decade. This has forced a large proportion of men to spend their lives as migrants, working in different parts of the island and sending remittances home (usually by registered mail) to their spouses and other kin, where much of the cash is invested in animals.^[5] These migrants are drawn to the major urban centers of other regions where they are hired as night watchmen and as cowherds. They are preferred by employers because they are willing to brave the elements and sleep outdoors, even during the cold, wet winters so characteristic of the highlands (for detailed discussions of southern pastoralists see Decary 1933; Faublée 1954; Frère 1958; Huntington 1973).

Religious Affiliation

Religions of foreign origin—Islam, Catholicism, and Protestantism—provide another means for distinguishing Malagasy from one another. They define divisions that are both ethnic and geographical. Estimates for religious affliation for the entire island in 1982 are as follows: 57 percent adhere to traditional (or what I will refer to as “indigenous”) beliefs (fomba-gasy) and 40 percent are Christian, with equal representation of Roman Catholics and Protestants (Bunge 1983: 62). I assume that the remaining 3 percent is mostly Muslim, but also includes Indian Hindus.

Islam is strongest in the north and west. In contrast to other Malagasy, many Sakalava and Antakarana are faithful to Islam, their conversion having occurred in response to their efforts to find allies against Europeans in the eighteenth and nineteenth centuries. In addition, Arabs, the majority of Comoreans, and many Indo-Pakistanis are Muslim (see Delval 1967, 1987; Dez 1967; Gueunier and Fanony 1980; Vérin 1967).

Christianity has played an important role in shaping factions in modern Madagascar (Gow 1979; Mutibwa 1974; Southall 1979). Attempts were made by Portuguese and French Catholics to Christianize coastal peoples in the fifteenth and sixteenth centuries (Brown 1978: 30ff; Bunge 1983: 14), but these missionaries were killed by indigenous peoples shortly after they arrived. The first successful missionaries were Protestants from the London Missionary Society (LMS), who were welcomed into the Merina kingdom in 1818 by royalty who sought to have the British join them as allies against the French. Other sects soon followed and rivalries developed as they sought to stake out territory and win converts. While Protestants won the favor of Merina royalty (andriana) and elites (hova), Catholics worked outside the capital. Among the highland Merina they were most successful with the slave population (andevo). They also worked among the Betsileo and throughout the coastal areas of the island. In 1863, Protestants were expelled from the island by the Merina Queen Ranavalona I, but they were eventually allowed to return in the 1880s following her death (Brown 1978: 152ff; Gow 1979; Mutibwa 1974).

Rivalries between different sects have been fierce up to the present. According to Covell, the number of Christians has declined since Independence (1987: 95; she in turn quotes Raison 1970). As she states, “The Christian Churches of Madagascar claim an estimated four million ‘adherents’; a term that obviously covers a wide range of commitment. Both Catholic and Protestant churches have their largest number of adherents in central Madagascar, and one of the common physical features of the plateaux villages is the presence of two spired churches, usually confronting each other from the opposite ends of the village” (1987: 95).

Today, certain faiths are associated with specific ethnic groups. The Catholic Church is strongest in the southern plateaux among the Betsileo and in coastal areas. Protestantism is split between those churches that have a tradition of strong political ties with the Merina in Antananarivo and others in the south; the latter rely on regional missionary activities (for example, American and Scandinavian Lutherans have been active in southern Madagascar since 1866). In recent years, attempts have been made to forge new unions, alleviating the competition between Protestants. The FJKM church (Fiongonana Jesosy Kristiany Malagasy or the Malagasy Church of Jesus Christ) was formed in 1970 and draws adherents from such Protestant groups as the original LMS, Quaker, and French Protestant Mission churches (Covell 1987: 95). Subsequent ecumenical agreements among Protestant sects have led to their working in what was predominantly viewed as each other’s territories. (For more detailed discussions on the Catholics see Chandon-Moët 1957; Decary 1966; Hübsch 1987; Judic 1987; L’Hermite 1968. On Protestants see Belrose-Huyghues 1979; Birkeli 1957; Gontard 1971; Gow 1979; London Missionary Society 1881–1900; Vérin et al. 1970.)

The majority of Malagasy remain faithful to indigenous religion or fomba-gasy, in which ancestors are central. In a sense, fomba-gasy touches the lives of all Malagasy, regardless of faith, since all Malagasy express an interest in their origins and a concern for their respective ancestors. The contrasts and contradictions between fomba-gasy and other religions have served as a means to divide peoples throughout the recent history of the island. This is certainly true for the inhabitants of Ambanja.

Social and Cultural Divisions in Ambanja

Malagasy are, historically, migratory peoples, first in terms of the original settlements of the island and, more recently, in terms of subsequent movements within the boundaries of this island (see Deschamps 1959). The movements of peoples have increased dramatically throughout this century (figure 3.3). Individuals and entire groups have moved for a variety of reasons: to find new grazing lands in response to population pressures (the most important case being that of the Tsimihety, see Molet 1959 and Wilson 1971); accompanying the expansion of Sakalava and Merina kingdoms (see Brown 1978; see also assorted essays in Kent, ed. 1979 and Kottak et al., eds. 1986); and in order to flee from more powerful enemies, as was true with the Tanala (Linton 1933: 24ff). Throughout the latter part of this century, Malagasy have also migrated in search of wage labor, and the Sambirano is an important destination.

The Social Categories of Ambanja

As a result of the in-migration of non-Sakalava to Ambanja, the classification of its inhabitants is extremely complex. The most important distinction is between insiders and outsiders, or tera-tany and vahiny. When loosely defined, tera-tany may refer to anyone who owns land, but in the strictest sense, this term refers to the indigenous Bemazava, who form the largest single group represented in the area.^[6]Vahiny is used to refer to other migrant Malagasy. Much tension exists between members of these two social categories. In addition to Malagasy-speakers, French, Comoreans, Chinese, Indians, Arab-Yemenis, and Greeks have also been drawn to the Sambirano. As mentioned above, these peoples are “foreigners” (étrangers; Western Europeans are also called vazaha), and this label applies even if they were born in Madagascar. According to the 1986 census, foreigners collectively numbered 493, or 0.5 percent of the total population of the Sambirano Valley.

Contrary to customs of ethnic endogamy observed in other regions of Madagascar, in Ambanja intermarriage is common, not only between different groups of Malagasy-speakers, but also between Malagasy and peoples of foreign origins. Métisization—that is, intermarriage between members of diverse groups and their subsequent offspring—involving all groups is extremely high. Special labels are used when speaking of the offspring of these unions, and they reflect that an individual’s parentage is a mixture of Malagasy and foreign heritage. For example, a child of a Sakalava mother and an Arab father is called “Arab-métis(se).” Ambanja has developed into a polycultural community, marked by the sharing and overlapping of cultural and social norms of diverse peoples. As a result of this blurring of ethnic boundaries, what it means to be tera-tany or, more specifically, Sakalava, has become increasingly problematic, so the terms insider and outsider require constant redefinition.

Malagasy Ethnic Groups and Their Points of Origin

Ambanja, as a migrant town, may be viewed in certain ways as a microcosm of Madagascar as a whole, since all groups—regardless of whether they are defined by ethnic, geographical, or religious categories—are represented. According to the 1986 census, the town had a population of 26,288 (25,945 nationals and 343 foreigners)^[7] (Madagascar, Service du Planification, 1986). Over the latter half of this century the population of the town has doubled every ten years, whereas the average national growth rate in Madagascar is 3.1 percent per annum (Covell 1987: xiii), the national doubling time being approximately twenty to twenty-five years. This population increase in the Sambirano has occurred as a result of the immigration of non-Sakalava. The rate of immigration has increased dramatically since 1975, following the Socialist Revolution and the subsequent expansion of state-run agricultural enterprises (Andriamihamina et al. 1987: 25). Estimates for the population of the town of Ambanja since 1950 reveal this doubling of the population (Andriamihamina et al. 1987: 21) (figure 3.4).

Statistics on the ethnic breakdown of the population of the district of Ambanja (the entire Sambirano Valley) are scanty, since they were collected regularly only from 1950 to 1971. Further problems arise since changes were made in the size of the district over time, as well as in the procedures used to collect census information. It is also unclear whether or not these figures include temporary residents of the valley. Since 1985 the timing of the census has been designed to avoid the coffee season (June and July), when there are many temporary laborers in town. The figures are now collected by local representatives for each neighborhood, drawn from registration books where they keep a log of all residents living there during the course of that year. Despite these disadvantages, I have reproduced these figures since they provide a rough picture of the ethnic composition of the Sambirano Valley (see figure 3.5).

These figures show that, first, the Sakalava remain the largest single group, comprising 37 to 54 percent of the population. Second, all officially recognized categories are well represented in each census, reflecting the presence of large numbers of non-Sakalava. The most significant groups, in terms of number (listed in descending order) are as follows:

1. Makoa are the descendants of African slaves; Makoa was an official ethnic category as late as the 1970s. Makoa are now considered to be indigenous to Sakalava territory in western Madagascar and are sometimes subsumed under the category designated for Sakalava. From the point of view of local Bemazava, they occupy an ambiguous category vis-à-vis tera-tany status: although they are perceived as a distinct group, they have special rights and privileges in reference to this status because they served the Sakalava in the past.
2. Tsimehety are pastoralists who originally came from the south and who have since migrated northward throughout this century. Today they occupy the territory to the south of the Sambirano; many also live along the northeast coast in an area referred to as the “Vanilla Triangle.”
3. Antaimoro are from the southeast, near Farafangana.
4. Antandroy are pastoral peoples from the far south.^[8]
5. Betsileo and Merina are the two major groups that occupy the high plateaux.

Dating the arrival of these different migrant groups to the Sambirano is difficult. By consulting the records of land titles held by one of the first plantation owners, de la Motte St. Pierre, it is clear that by 1907 he had worker villages situated on his lands (Title 140, Le Gabés) and by 1908 the Sakalava were living on indigenous reserves (Title 130 B. P.) (Service Topographique). Planters on Nosy Be used slave labor until slavery was abolished by the French parliament in 1896. After that they relied on land tenants and corvée labor. Attempts were made throughout Madagascar to recruit foreign laborers from China, Yemen, and the Comoro Islands, but these efforts were relatively unsuccessful (see Brown 1978: 249–250; Stratton 1964: 95–96; and Thompson and Adloff 1965: 442ff). The first Antandroy laborers were brought to Nosy Be in 1922 by Djamanjary Sugar (established 1918). These workers were recruited from the prison in Tulear and brought to Nosy Be by boat. Many of these Antandroy settled permanently in the north. It is possible that a few planters in the Sambirano continued this recruitment practice until migrants began to arrive on their own.

Betsileo are also among the early migrants to the Sambirano; evidence of their work can be seen on the road to Diégo, just north of Ambanja, where there are large rice paddies. As one Betsileo informant explained:

My parents decided to come north in the 1920s to search for work, and they came directly to the Sambirano, having heard that there were colonists here who needed laborers. My father came to work in the sugar cane and manioc fields, and he became a land tenant of one of the plantation owners. Under a system called miasa-talata [lit. “to work on Tuesday”; the French called these workers talatiers, cf. Feeley-Harnik 1984: 11] he was given land to farm where he grew rice; in exchange he worked every Tuesday for the landowner. When French and other large landowners left Madagascar in the 1970s, my father bought the land, so that now he owns approximately 110 hectares.

Today my informant’s father farms his land successfully with the assistance of a tractor and the labor of his wife and fifteen descendants, including children and grandchildren. Clearly, by the second decade of this century, the Sambirano was well-known throughout Madagascar as a place where one could find work, and it attracted people from all areas of the island.

In the 1959 census, Jonoro Houlder, the Chef du District, devoted special attention to a discussion of immigration and emigration of peoples to and from the Sambirano. He noted that a number of non-Sakalava were well established in the area, contracted by European planters to work as manual laborers. These included the Tsimehety (from Analalava, Antsohihy, and Bealanana), and peoples from the south and southeast (Antandroy, Antanosy, Bara, Antaimoro, Antaisaka). Others from the east coast (Betsimisaraka) and high plateaux (Betsileo and Merina) had also arrived in the area and had put down roots:

Le district d’Ambanja est un district plutôt d’immigration et d’emmigration. Il forme une plaque tournante où transitent obligatoirement tous ceux qui, venant d’Analalava, d’Antsohihy ou de Bealanana par voie de terre, se dirigent vers Ambilobe et Diégo-Suarez.…Dans cette région particulièrementfavorisée par une terre riche, la colonisation européene a introduit des travailleurs salairiés ordinaires du Sud, engagés par contrat, et qui forment l’essentiel de sa main-d’oeuvre: Antaimoro, Antaisaka, Antandroy, Antanosy, Bara.…Profitant de la décadence et de l’apathie des Sakalava qui peuple-rent initialement ce pays, d’outres éléments allongènes plus laborieux, se sont infiltrés et y ont fait solidement souche: Betsileo, Tsimehety, Hova [Merina], Betsimisaraka, etc. (Madagascar 1950–1971, report from 1959: 2)

Houlder then went on to discuss the presence of non-Malagasy, including Comoreans and Indians, and their activities as local merchants.

Similar comments are made about the immigration of Malagasy in the report from 1965, where the author spoke again of the importance of contract workers, especially Antandroy and Antaimoro; he also commented on Tsimehety migration, which had continued to grow. He stressed that most Tsimehety came from Bealalana (which is about 350 kilometers south of Ambanja) and that Tsimehety villages had been established in the Sambirano (Madagascar 1950–1971, report from 1965: 2).

A recurring theme in subsequent reports is the reluctance of Sakalava to emigrate to other areas of Madagascar. Such a statement is made by Houlder, above, and is reiterated by the author of the 1968 report: “les elements aborigènes sont ordinairement sedentaires et quittent difficilement le village natal” (Madagascar 1950–1971, report from 1968: 1). The authors of these reports comment on Sakalava resistance to work as wage laborers (cf. Feeley-Harnik 1984: 6, 1991b, especially p. 191ff and chapter 5).^[9] As discussed in the previous chapter, today these are themes which are voiced by Sakalava about themselves as well as by frustrated employers. The complexities that have arisen as a result of this constant immigration of non-Sakalava will be discussed below in this chapter, as well as in the chapter that follows.

Local Territory and the Ancestral Land

As described above, the tanindrazana or ancestral land is central to Malagasy notions of identity, both personal and collective. In the strictest sense, the region of the Sambirano is the exclusive ancestral land of the Bemazava-Sakalava. Since Malagasy identity is tied to the ancestors and, thus, to the land, all other peoples, regardless of how long they have lived here, are outsiders by virtue of this rule.

Sakalava maintain tombs that are located throughout the valley, often in small remote villages that may be a long drive or walk from where they reside.^[10] In some cases these tombs are fairly new. When the French relocated Sakalava, they obliterated all traces of old cemeteries, so that the Sakalava had to create new tombs for their dead (cf. Colson 1971 for similarities with British colonial policy).

Some migrants have established their own tombs or cemeteries on land they or older kin have bought. This is especially true for the Antandroy, who bury their dead locally, since for them the costs for transporting a body as far as one thousand kilometers to their ancestral land in the south is prohibitively high. In the town of Ambanja there is also a cemetery where Christians and Muslims are buried, and it is used by both Sakalava and non-Sakalava. Migrants who choose to be buried locally may not be accepted as full-fledged tera-tany by Sakalava. Nevertheless, by Malagasy rules of affiliation, such a choice is an indicator that one’s sentimental ties to the region are strong, since where one is buried carries the sense that one is from—or belongs—there.

Ethnicity and Economic Specialization

In Ambanja, different economic activities are associated with different ethnic groups. In reference to the four largest migrant groups represented in the Sambirano (see above), the following generalizations can be made.

Tsimihety have an established history as labor migrants throughout this century (Deschamps 1959). Many come to the Sambirano as temporary migrants during coffee harvest season in June and July. Others have settled permanently in the Sambirano.

Others from the south and southeast, such as the Antaimoro and Antandroy, were among the earliest migrants to the region. Today they form the bulk of manual laborers at the enterprises and at businesses in town. Antandroy men also work as night watchmen (see Frère 1958: 117–140 for information about the Antandroy).

It is the Betsileo and Merina who hold university degrees and who fill the majority of civil service and other government sponsored positions. As state employees, they are schoolteachers, the managers and engineers of the enterprises, the president of the local bank, the doctors at the town’s hospital and worker’s clinic, and the directional staff of government agricultural and communication agencies. There are also a number of rural peoples from the high plateaux who have set up small businesses or who work as “traveling merchants” (mpivarotra mandeha). The best known of these are the Vakinankaratra, a group occupying a southern region of Merina territory around the town of Ambatolampy (which lies to the south of Antananarivo). Many of these traveling merchants have established Ambanja as their home base, and they travel periodically to the national capital where they buy goods, return to Ambanja, and sell them in the local market at a significant markup.

Religion

Ambanja is the religious center for the Sambirano and adjoining regions. Reflecting national trends, Muslims, Catholics, and Protestants are divided along ethnic lines. Estimates of religious affiliation in the Sambirano are 4–6 percent Muslim, 6 percent Catholic, up to 6 percent Protestant, and approximately 80 percent fomba-gasy (Jaovelo-Dzao 1983, 1987, and personal communication). The faiths of foreign origin which boast Sakalava membership are Islam and Catholicism. Protestants, on the other hand, usually are non-Sakalava from the high plateaux and the south.

Islam

As mentioned in chapter 2, in the nineteenth century, the Bemihisatra-Sakalava royalty of Nosy Be converted to Islam. Pierre Dalmond, a Catholic priest who arrived in Nosy Be in 1840, reported in his journal that when he was first received by Tsimiaro I, the king of the Antakarana, the ruler was dressed as a Muslim, wearing a white robe and a hat of red velour (Dalmond 1840). A photo taken perhaps half a century later of “Tisaraso [Tsiaraso I] (1871–1919), souverain des Sakalava Bemazava” (see Raison-Jourde, ed. 1983, between pages 128 and 129) shows the ruler clothed in similar fashion. It is this same style of dress that mediums wear today when they are possessed by the greatest of the Bemazava, Bemihisatra, and Antakarana tromba spirits.

Today, among the Bemazava, the majority of Muslims are royalty. Most are “sympathetic Muslims” (SAK: muselmans/silamo sympatiques) who do not go to mosque regularly and only observe the fast of Ramadan (SAK: Ramzan). A Sakalava expression that reflects their lax attitude is tsopa tsy haramo, which literally means “[drinking] tsopa [a small bottle of very strong alcohol] does not violate the Koran” (Jaovelo-Dzao, personal communication; compare the kiSwahili expression chupa tsi haramu, J. Bergman and M. Porter, personal communication).

There are five mosques in Ambanja, each identified according to the ethnic makeup of the majority of its members. These are the Comorean (the majority of members are originally from Anjouan); Comorean and métis (Comoreans and Malagasy of mixed origins); Arab-métis (offspring of mixed marriages between Arabs and Sakalava); the Indian mosque; and a splinter group that I will refer to as “Modern Islam” (see chapter 10), whose members are primarily Sakalava, Antakarana, and some Comorean. These labels aside, membership in all mosques (except that of the Indians) is fairly heterogeneous, reflecting the town’s high rate of intermarriage between peoples of diverse backgrounds. The number of regular members (that is, those who attend Friday prayer) of each mosque varies between fifty to one hundred, with perhaps half as many attending daily prayer. The largest is the Comorean mosque.

The Catholic Church

When the northern Sakalava failed to acquire firearms from the Sultan of Muscat, they turned to the French for assistance against the Merina. Captain Passot led the French military into Nosy Be in 1840, and he brought with him Catholic missionaries. Among these was the Jesuit priest Dalmond, who had already been active on the small islands off the east coast of Madagascar (see Dalmond, n.d.). Nosy Be became an important northern post for Catholic activities, while additional evangelical work was conducted on Nosy Faly, Nosy Mitsiosioko, and in the Sambirano Valley. In the early part of this century, however, Ambanja replaced Nosy Be as an important colonial and commercial center. By 1921 the Catholics were active in Ambanja, having followed the military there, and in 1936 they built a cathedral, which continues to be one of the largest buildings in town. The Catholic church has by far the largest congregation in Ambanja, the cathedral seating approximately two thousand (it is packed during Easter). Catholic adherents include Sakalava as well as well as Tsimihety, Antandroy, Antaimoro, and some Betsileo and Merina.

Catholicism is far more popular than Protestantism among the Sakalava, and the style of Catholic evangelizing has much to do with this. Although Sakalava royalty have remained Muslim, Sakalava commoners were encouraged to attend the mission school, and those who did so often converted to Catholicism. The present king, Ampanjakabe Tsiaraso III, is an anomaly: tera-tany insist he is Muslim, but he himself professes to be Catholic, since he converted to the latter faith when, as a child he attended the mission school. The Catholic church has been successful in winning Sakalava converts because it is especially sensitive to Sakalava culture. In recent years this approach has been formalized under a policy known as enculturation which evolved following Vatican II. Enculturation encourages the tolerance of local beliefs and customs and emphasizes syncretism over orthodoxy. This policy plays an important role in shaping Catholic practice in Madagascar, as in many other parts of Africa (cf. Aubert 1987).

Although Ambanja’s Catholic clergy includes nuns and priests from France, Italy, and Germany, the majority are Malagasy, who are primarily Sakalava. The leaders of the church in Ambanja and nearby towns are all locally born Sakalava, while Europeans occupy the backstage in terms of policy matters. In addition, whereas the Protestant Bible is written in official Malagasy (and thus is based on the Merina dialect), the Catholic Bible used in Ambanja has been translated into Sakalava. Catholic services are also performed in the local dialect. They are lively and innovative and incorporate elements that characterize local town life. For example, the church hosts discos as fundraisers and has an electric guitarist who accompanies the singing of hymns at Sunday services. Rumor also has it that a number of priests have taken mistresses and have had children by them, and that several of the nuns delight in dancing with their students at the church discos.

The Catholic church has made attempts to incorporate local customs into the services. It is also tolerant of fomba-gasy, including tromba possession. In contrast, the Protestant groups regard fomba-gasy as sinful and the work of Satan (the full ramifications of this stance will be discussed in chapter 10). Furthermore, while Protestants complain about the Sakalava’s lack of respect toward such social institutions as marriage, the Catholics perform several marriage ceremonies each year (this rarely occurs at the Protestant churches).

Protestantism

Thus, Sakalava, if they “pray” (mivavaka), are Muslim or Catholic; they are rarely Protestant. Sakalava and their Antakarana neighbors have resisted Protestantism since they associate it with their enemies of the past century, the Merina, among whom Protestants have been active since the nineteenth century. Today it is rare to find Protestant Sakalava in the Sambirano. The only northern people who historically have been associated with Protestantism are the Tsimihety.

Ambanja is, nevertheless, home to a wide variety of Protestant churches. The oldest of these is the FJKM, which has been active in the area since the 1930s. Other groups include Lutherans, Adventists, Anglicans and a handful of Pentecostal groups that split from the FJKM more than a decade ago. The FJKM is the largest Protestant church in town, with approximately 250 to 300 people who regularly attend Sunday services. The Anglicans, Adventists, and Lutherans all have small churches that can hold up to about two hundred people, but on a regular basis there are only, perhaps, ten to thirty in attendance. Each of these churches is relatively new, having been established within the last ten to twenty years, following an ecumenical agreement among different sects to allow one another to expand into each other’s territories. In contrast to the Sakalava, the majority of migrants are Protestants. The Lutherans, for example, who are associated with southern Madagascar, established a church in Ambanja in 1975. This congregation is composed primarily of peoples from the south and high plateaux, although they hope to gain converts among the Sakalava. More generally, the parishoners of all Protestant churches are Betsileo, Merina, Tsimehety, and Antandroy and Antaimoro. In Ambanja, Protestantism is an institution that embodies that which is not Sakalava.

Fomba-gasy, or Malagasy Religion

As the figures for religious affiliation show, fomba-gasy remains the dominant religion in Madagascar; in Ambanja, the percentage (80 percent) is much higher than the national average (59 percent), because Sakalava of Ambanja have been reluctant to convert to faiths of foreign origin. Jaovelo-Dzao attributes their resistance to Christianity to the power and influence of the Bemazava royalty.^[11] Since royalty had already converted to Islam prior to the arrival of Catholic priests, they dissuaded commoners from converting to this new religion. Following the dictum of royalty, Sakalava often state that it is fady for them to embrace other faiths (fady mivavaka, lit. “praying is taboo”). It is also easier for a Muslim to marry a Sakalava than it is for a Christian to do so, since Sakalava marriage and other institutions have incorporated elements derived from Islam (Jaovelo-Dzao 1983; personal communication).

Sakalava hostility toward non-Sakalava has played a role in their resistance to conversion. For the Sakalava, the fomba-gasy define what it means to be Sakalava, and so to give up tromba and other local practices would, from a Sakalava point of view, require denouncing one’s ethnic identity. Sakalava express their reluctance to attend Protestant services by saying that many of the pastors, and other officíants, are from the highlands. Several Lutheran evangelists spoke to me of the Sakalava with bitterness; since they themselves are Merina, it is impossible to gain converts. During the course of my fieldwork two young Sakalava men joined the church, and the pastor and evangelists hoped that at least one would choose to be trained as a pastor and return to Ambanja to evangelize among his own people. Catholicism and Islam reveal a tolerance for fomba-gasy, but Protestants are strongly opposed to these traditions. Tromba possession, which is central to Sakalava religion, provides an appropriate example. Protestants say they are fady tromba (“[have a] tromba taboo”) and so only those mediums who choose to opt out of possession by having their spirits exorcized by the Protestants convert (this will be discussed in chapter 10).

Studies of Zionist churches in Africa provide examples of how faiths of foreign origin may answer questions that arise as a consequence of massive social change (Comaroff 1985; Jules-Rosette 1975; see also Colson 1970). Yet in Ambanja, a town so rife with conflicts, the general population shows little or no inclination to convert to Christianity or Islam. Instead, what has occurred is a virtual explosion in the incidence of tromba possession, involving Sakalava and vahiny.

The Effects of Polyculturalism

Although there are no statistics available for the valley’s present ethnic composition, it seems clear that Sakalava remain the single largest group represented. Sakalava own the majority of small, privately owned plots of land which are scattered throughout the Sambirano. Many are peasants who raise both subsistence (especially rice) and cash crops (especially coffee and cocoa). Those who live in town usually own the land they live on, supporting themselves by running small family-owned groceries on the premises. Others sell produce, charcoal, and other essential items in front of their homes. Townspeople may also own land in the countryside. It is the Sakalava who, more generally, dominate formal power structures. Since they comprise the largest voting block, Sakalava dominate elected positions in the city and county governments. The only other group that has managed to have a few members elected are the Tsimihety. Sakalava royalty—both living and dead—also wield power in the Sambirano. Because Sakalava exercise the greatest influence of any ethnic group in the Sambirano, it is important to understand how they perceive the social world of Ambanja.

Sakalava Perspectives

Sakalava characterize other Malagasy groups as being closer to or farther from themselves. These conceptualizations in turn reflect notions of similarity and compatibility on the one hand and dissimilarity and distrust on the other (figure 3.6).

These categories have profound implications for patterns of informal association. They may affect more formal decisions as well, as in choosing an appropriate spouse or electing an official to the local government. On one end are groups thought to be the most similar to the Sakalava: these are their neighbors the Antakarana and Tsimihety. Makoa occupy a slightly more ambiguous position since they are the descendants of slaves who served local Sakalava in the past. The other extreme includes pastoralists from the south, like the Antandroy and the Merina of the high plateaux. Sakalava, like many other non-southern peoples, rank themselves above pastoralists, regarding themselves as more sophisticated. Like other coastal peoples, they express resentment and distrust of the Merina (whom Sakalava refer to derogatorily as borzany). There is a body of fady which places Merina in an extreme position outside the realm of association for Sakalava. Some Sakalava tromba are said to be fady Merina (“[having a] Merina taboo”), so that Merina may not attend tromba ceremonies, nor may they approach Sakalava tombs or other sacred locations. Many Sakalava extend this fady to everyday interactions as well and deny Merina access to their homes. More generally, there are fady and other customs that set various groups outside the Sakalava realm. Among the most common is the ingestion of pork, for the majority of Sakalava have a pork taboo (SAK: fady komankory).

What I wish to stress here is that it is difficult to pass as Sakalava: dialect, name, dress, and physical appearance give away a person’s ethnic background.^[12] Overlapping and confusion generally only occur among peoples from the same general regions of Madagascar. For example, Antandroy is a label often applied to Mahafaly and Bara. Also, a number of my informants were Merina but in public they claimed to be Betsileo to facilitate interactions with Sakalava. In general, peoples such as the Merina, Sakalava, and Antandroy have distinct physical characteristics that make it easy for the trained observer to distinguish one from the other. Malagasy are accustomed to thinking in this way about each other. Maps of Madagascar, for example, often include photos of people who are considered to be archetypal of their given ethnic group (see for example Société Malgache 1973; map opposite p. 40).

The logic of the Sakalava framework of similar and dissimilar peoples is based on several general principles. The first involves a sense of shared cultural features and historical origins. For example, in the past, Sakalava and Antakarana were members of one group. As a result of a dispute over royal succession, the Antakarana split off and moved north, so that now Sakalava and Antakarana are members of related yet separate descent groups. These descent groups of the northern Sakalava and Antakarana are called the Zafin’i’mena and Zafin’i’fotsy, or “Grandchildren of Gold” and “Silver,” respectively. Relations today between the two are cordial. The mother of the present king of the Antakarana, for example, is of the Bemazava-Sakalava royal lineage. Sakalava and Antakarana share many fady and honor many of the same ancestors. Ceremonies are similar and at times indistinguishable: tromba ceremonies may include the same spirits, and at royal occasions for either one may see a dance called the rebiky performed (see Feeley-Harnik 1988).

The second principle is one of proximity. The Sakalava’s closest neighbors are the Tsimihety (to the east and south) and Antakarana (to the north). This places all other peoples geographically at a distance from them. As one moves farther away to the east (towards the Betsimisaraka) and to the south (into the high plateaux and farther south and southeast), social distance increases. With the Merina, however, historical factors take precedence, for they are still seen as enemies while the Betsileo, who live farther to the south, are tolerated to a greater extent.

Defining Tera-Tany and Vahiny

The increased immigration of non-Sakalava to the Sambirano Valley creates a dilemma for the Sakalava. Since they would rather not work for the enterprises, they are faced with the constant influx of unwanted strangers who are willing to do so. Sakalava express anxiety about migrants, and this anxiety has increased since the recent improvement of the north-south highway, which connects Ambanja with Mahajanga and, ultimately, Antananarivo. In 1987, it was being turned into an all-weather road, and the Sakalava feared it would bring still more people, leading to an increase in crime rates and other problems characteristic of other urban centers in Madagascar.

The lack of social cohesiveness among members of different groups is graphically illustrated in the haphazard layout of the town. Malagasy exhibit a preference for socializing with individuals who share similar regional or ethnic origins. Nevertheless, in Ambanja, like peoples are rarely clustered together in neighborhoods. Instead, each area of town presents a wild array of people of diverse origins and class backgrounds. The effects that accompany the ever-increasing influx of migrants have become a major concern in the town. A recent study conducted by the Urban and Housing Service of the Ministry of Public Works (Andriamihamina et al. 1987: 7–9) echoed these concerns, particularly the potential dangers associated with water shortages and insufficient sewer facilities. They stressed the need for more careful city planning if these problems were to be avoided. An area of special concern was the neighborhood of Ambaibo, near the bazarbe, where houses are now crammed against the banks of the Sambirano River (see figure 2.2).

Since Sakalava are the dominant group of the Sambirano, it is they who determine who is tera-tany or vahiny (insider versus outsider). The factors that define these social categories should be clear by now. Yet the dynamic between Malagasy kinship and ethnic identity often blurs these categories. In Madagascar, offspring are said to share the ethnic identity of their biological father. In Ambanja, individuals born into families that have lived in the region for generations, who own land, and who identify culturally with Sakalava generally consider themselves to be tera-tany. Nevertheless, others may insist that they are, and always will be, vahiny because they ultimately share the identity and thus origin of their father (or father’s father, and so on).

The concept of ancestral land and the way in which it is interpreted in Ambanja is central to these definitions. Tera-tany and vahiny have conflicting views on how burial choices define one’s identity and proper place in the Sambirano. As described above, the place where one will be buried is of central concern for Malagasy, as it is tied to concepts of kinship identity and home. The kin of Malagasy who die far from home will go to extreme lengths to make sure that the body will make its way back to the family tomb. In Ambanja there are families that, even after living in the town for a number of generations, are still tied to tombs elsewhere. They are, however, a minority. Migrants who settle permanently in Ambanja generally choose to be buried in this region. Nevertheless, although migrants may think of Ambanja as home, psychological alienation may persist because of Sakalava notions that define what it means to be tera-tany or vahiny. In other words, even for those who choose to buried in the Sambirano and who consider it their (new) ancestral land, Sakalava may still consider them to be vahiny.^[13] As will be illustrated in the following chapter, it is very difficult to shake the migrant or outsider status, regardless of how long an individual or cluster of kin has been in the area and regardless of their burial choices.

Notes

4. Tera-Tany and Vahiny: Insiders and Outsiders

The complexity of factionalism makes migrant status a perplexing one in Ambanja. Clearly there are advantages to being an insider: one has first rights to the land, a greater access to and an understanding of local power structures (be they formal or informal), an understanding of Sakalava culture, and, most important, social acceptance among Sakalava. Since the essential characteristic of insider status is Sakalava ethnicity, the boundary is a fairly rigid one. Migrants face the contradictions between how they perceive their own identities, based on their sentimental ties, and how others perceive and label them.

For newly arrived migrants, having contacts in the community is essential if they are to succeed in finding lodging and employment. They must establish new networks, seeking out kin and, sometimes, friends, who come from the same region of the county and who can help ease their transition into town life. While tromba enables migrant women to be symbolically transformed into Sakalava, there are other institutions that facilitate the partial integration of migrants more generally, offering avenues that are open to both men and women. Some migrants join ethnic associations, which help to maintain ethnic distinctions, or they choose to be buried locally, instead of in their original ancestral land. Others attempt to become integrated into the Sakalava community by marrying tera-tany, or by changing their behavior, dress, and dialect. But even though these actions bring a migrant closer to Sakalava in everyday interactions, from a Sakalava point of view they do not transform the individual’s ethnic identity. The purpose of this chapter is to explore the nature of migrants’ experiences with these issues in mind. The first section illustrates this through case studies drawn from informants’ life histories. The second is an analysis of the important social institutions that facilitate or inhibit the incorporation of outsiders into the community of Ambanja.

Migrant Stories

The examples that appear below have been selected in order to illustrate the relevance of the factors mentioned in the previous chapter: ethnicity and associated characteristics such as fady or taboos, religious affiliation, areas of economic specialization, and so forth. They also reveal why migrants come to Ambanja and how they cope with life in this town. The examples presented here are for the most part success stories; the chapters in Part 3 will focus on the more extreme problems that migrants face in Ambanja.

In such a diverse community, distinctions must be made between different categories of migrants. In his study in northern Nigeria, Cohen draws on distinctions made by urban Hausa, who divide migrants into “settlers” (Dan Kasa) and “strangers” (Baki). Length of stay is significant, but it alone is insufficient to determine settler status. Instead, these two categories are defined primarily by social roles, particularly through participation in political, religious (in this case Islam), and economic arenas. More specific criteria include rights to housing within the Hausa quarter, occupation, residence patterns of wives and children, and primary relationships, including one’s relationship with the local chief (1969: 32ff and 222, nn. 9 and 10).

Cohen’s (or the Hausa’s) categories prove helpful in sorting out the dynamics of the social world of Ambanja. Cohen’s use of the term stranger is, however, problematic, since the equivalent of this term in Madagascar (étranger) is only applied to non-Malagasy. Thus, for the purposes of this study, I will divide the vahiny of Ambanja into two major categories: short-term or temporary migrants; and long-term or permanent migrants, whom I will also refer to as settlers. As with the Hausa, length of stay is, in part, a determining factor of status. In addition, as was illustrated in the preceding chapter, participation in particular types of economic, religious, and political activities are important aspects of identity. There are certain differences, however, between these two communities. Although social roles are important in Ambanja, temporary migrant and settler statuses have additional dimensions. First, whereas in Cohen’s study all migrants share a similar cultural background (they are all Hausa), in Ambanja one finds a multitude of groups, in which ethnicity is of primary importance. Thus, for members of certain groups, the desire to become a settler is already, in many cases, predetermined by point of origin. Tsimihety, for example, are more likely to settle permanently than are Antandroy. Second, Sakalava attitudes toward migrants differ according to their ethnic origin. Sakalava feel most comfortable with Antakarana, but they avoid contact with Merina. As will be made clear below, adherence to particular fady is often a clear marker of difference. A third key factor is what I will refer to as migrants’ sentimental ties to life in this town. As Wilson (1971) has illustrated, the idea of “sentimental structure” is important for Tsimihety pastoralists, who, often with each generation, move away from the place where they were born. They must, therefore, eventually choose their burial location, and this is determined by, essentially, their emotional ties to a particular place or to certain kin. As will be clear throughout this study, this idea of sentiment is a crucial one for migrants in Ambanja, evident in their daily activities, personal ties, marriage patterns, burial plans, and participation in tromba or other religious and cultural institutions.

Thus, in Ambanja, temporary migrants are those who have come to the Sambirano searching for work and who only intend to stay long enough to make their fortunes. Although some may live in Ambanja for more than a decade, they do not consider it to be their home. For these people the ancestral land lies at their point of origin, often where they were raised, and their strongest sentimental ties are with kin who remain there. Temporary migrants try to send remittances home regularly, money that is used to help support kin and is invested locally in land and animals. Temporary migrants also attempt to visit their ancestral land as often as possible (generally once a year) and they intend to return there to live permanently and retire. Although they may take lovers in Ambanja, they often have spouses and children back home. Temporary migrants also show a stronger tendency to socialize with people of similar origins; since they do not intend to stay in Ambanja, they are not particularly interested in becoming incorporated into the Sakalava community. Examples of temporary migrants, whose stories are detailed below, include the men of the handcart (posy-posy) team and Mme Razafy.

Permanent migrants or settlers, on the other hand, are those individuals who come to Ambanja and stay. For some, this was their original intention (as with Botabe’s father, see below). Others were temporary migrants whose sentimental ties to the region changed over time as a result of economic or other forms of success. In time, they have chosen to settle permanently in Ambanja. Of those who were settlers from the start, the majority come to Ambanja alone, and they may permanently sever ties with their kin in their original ancestral land. Another common pattern among settlers is to marry locally, often with Sakalava tera-tany rather than with other vahiny. Over time, their manner of speaking and style of dress also reflect their adoption of Sakalava customs. There are still others who come to the Sambirano, either accompanied by their spouses or who have their spouses (often with children) join them there later. Some of these people still feel tied to the original ancestral land and may prefer to be buried there. Their children, however, will generally choose to be buried locally in the Sambirano, considering it to be their new ancestral home.

Children of migrant parents (temporary or permament) characteristically have strong sentimental ties to the Sambirano, especially if they have never visited kin living in their parents’ original ancestral land. One day a Betsileo friend of mine, for example, looked at her four-year-old with a puzzled expression when he used an unfamiliar Sakalava term that he had learned from his playmates. She later said, laughing, “ah-ah, he is a child of the Sambirano!” playfully using the Sakalava expression for this (SAK: tsaiky [ny] Sambirano, HP: zanaka [ny] Sambirano). Ultimately, what determines where children’s sentiments will lie once they are adults depends somewhat upon their parents’ patterns of association and, to a greater extent, upon their own experiences, especially in reference to the peers with whom they associate.

Finally, the notion of what it means to be a vahiny varies. From a Sakalava point of view, all people regardless of their length of stay in or their sentimental ties to the Sambirano—are vahiny if they are not Sakalava by birth. They are forever, as the word vahiny implies, “guests” to the region. This occurs regardless of most attempts they might make to become assimilated. In general, the only exceptions to this rule are post-partum practices (which will be described later in this chapter) and, as will become clear in Part 2, tromba possession, both of which affects primarily women. From the perspectives of different categories of migrants, however, the appropriateness of the term vahiny varies. Temporary migrants generally consider themselves to be vahiny, since they have no intention of staying in the Sambirano. As the story of L. Botabe below illustrates, settlers may face a dilemma that arises from conflicting perspectives on their status: while they may consider themselves to be tera-tany, Sakalava continue to label them as vahiny.

Antandroy Migrants: The Story of Guardian

Guardian is approximately twenty-six years old and is an Antandroy man from the south. He prefers not to use his real name but instead asks that people use a term of address which is a reference to his line of work: night watchman. His employer, a lawyer in town, does not even know what his real name is.^[1] When I first met Guardian he had been living and working in the Sambirano for six years; the lawyer had been his employer for the last eight months.

When I asked Guardian why he had come to Ambanja, his reply was a typical one: “hitady harena”—“to seek [my] fortune”—in the Sambirano, where it is known that there is money (misy vola). As he explained, “It took me several months to make it here, traveling much of the way on foot.…it was a long trip.…my village is near Tulear [in the south and more than one thousand kilometers away]. In Tulear it is very beautiful, but there is no rain and no food.^[2] I stopped and worked for a while in Antananarivo to earn the bush taxi fare to the north, but life was so difficult there—it was too cold, and there were thieves everwhere! One night when I was sleeping in a doorway someone stole the sandals off my feet!”

“...When I first arrived in Ambanja I sought out my brother [FaBrSo]; he worked at a local enterprise and he helped me get a job as a laborer.…There I cut branches from cocoa trees and cleared the underbrush. I also worked at the loading station, piling sacks of prepared cocoa on transports bound for the nearby port.…Eventually I saved enough money to buy several head of zebu cattle from people living in the countryside. These I fattened up and then sold to buyers in Nosy Be, where there is no local beef. Sometimes I also sold milk and beef to neighbors.” (He then imitated the sound of a conch shell being blown, which I sometimes hear in my neighborhood at dawn to notify people that there is fresh beef for sale.) “After three years I was able to make a trip home; there I bought several goats which my father now tends for me.…I also brought my wife back to Ambanja with me.”

Guardian and his wife have since had a son, who is now two years old. Although well settled, he plans only to stay as long as necessary to make his fortune, so he can then return to his ancestral land and live there when he is an old man. He now has a herd of ten zebu in Ambanja and many goats at home.

Guardian’s story is typical for pastoralists from the far south, whose kin, for several generations, have come to the Sambirano to work as temporary migrants and who later return home.

The Traveling Merchants from the High Plateau

Population pressures, land shortages, or natural disasters (especially from floods and cyclones) have led many from the high plateaux to migrate to the Sambirano. The people best represented in Ambanja are the Vakinankaratra, who culturally are identified by other Malagasy as Merina. They come from the region of Ambatolampy, which lies to the south of Antananarivo (for a detailed study see Keenan 1974). Many of these people are traveling merchants (mpivarotra mandeha): they buy goods in the great Zoma market, which is held on Friday in Antananarivo. (Zoma is said to be among the largest markets in the world; it is here that people come from all areas of the island to sell goods.) The traveling merchants take their newly purchased goods and go to other, less central areas to sell them, generally at significant markups. A pastor I knew in Ambanja started to do this in order to supplement the meager income from his church. After an initial loan of 100,000 fmg (approximately U.S. $75 at the time) from a kinsman of his wife, he made one trip by bush taxi to Antananarivo, bought goods there, and then returned to Ambanja to sell what he had purchased, making a profit of approximately 100 percent. After this he would send money in installments to his mother, who would buy additional goods for him and then ship them by mail to Ambanja. Within a period of only four months, he managed to raise his standard of living considerably, buying luxury food items and new furniture, changes that greatly upset his parishioners, who criticized him behind his back for being a man who “loved money” (HP: tia vola izy).

Many of the traveling merchants live clustered together in the heart of the old section of town, next to the bazarbe, where the majority of them sell goods on a daily basis. It is here one can always find hats, sewing and electronic supplies, watches, ready-made clothing, and cooking utensils. One kingroup I knew had started out as traveling merchants and later set up a small factory making aluminum pots, pans, and spoons by melting down scrap aluminum (figure 4.1).

Roland’s Cookware Factory

Roland first came to Ambanja in 1981 as a traveling merchant, selling aluminum cookware to people throughout the Sambirano. When he arrived here he had no kin (havana) but made friends locally among Betsileo, Merina, and Vakinankaratra from his region. He has been married since 1979 to Nadine, who came to Ambanja to join him with three of their four children in 1987. Their oldest child lives with Nadine’s mother’s sister so that she can go to school in Ambatolampy.

Today Roland runs a thriving business in Ambanja with the help of eight adult kin. When I first visited their settlement, I was awestruck by the size of the operation. This group rents two buildings made of corrugated tin which rest on concrete foundations. The first has two rooms (one bedroom and one workroom) which they rent for 6,000 fmg per month. The second has three bedrooms, costing 5,000 fmg per month. Although the rents they pay are high by Ambanja standards, the group turns a good profit. They buy their aluminum by the kilo from a nearby village that specializes in collecting salvage metals. On average they make about thirty pots and three hundred spoons a day, although the rate at which they sell them varies greatly according to the season. The leanest months are December through February, when they sell perhaps thirty pots a week. The peak time is the coffee season in June and July, when people have more money. At this time they make as many as fifty pots a day (pots cost 1,500 to 6,000 fmg, spoons go for 300 fmg). Roland pays each worker 40,000 fmg per month, which is more than the salary paid to a junior high school teacher (who has a high school education) or a full-time manual laborer at one of the enterprises.^[3] Recently they have also built a pig shed, so that they can eat pork. (Highland peoples love to eat pork but cannot buy it in the Sambirano because pigs are fady for Sakalava and Muslims. As a result, Betsileo and Merina must go elsewhere to buy it; many make a point of buying pork sausages when they are in Diégo.) Their pig shed is a source of tension because many of their neighbors are Sakalava and Muslim. Their landlord, who is Sakalava, has permitted them to build the shed there.

I asked Roland how he had started his factory. “In 1985 I went back to Ambatolampy and persuaded several of my [male cognatic and agnatic] kin to join me here in Ambanja. I wanted them to come with me to help set up a shop for making aluminum cookware.… You see, at the time, there was no one producing these items locally. The merchants who sold these goods were buying them elsewhere and then charging a lot of money for them.…Eventually François and Clarence joined me [in 1985]: They were followed by Georges and Louis Sr. [1986], and Louis Jr., Paul, and Narda, along with Nadine and three of our children [1987].”

Although a number of these men are married, Nadine remains the only adult female of the group; all other married men have left their wives back in Ambatolampy. Since the majority are men, and a number of them are single, I asked if they go out with local women and, if so, what type (HP: foko/karazana). To this Clarence replied, “We usually go out with women from Antananarivo.” When I asked why not Sakalava, he said, “Oh, they do not speak the same dialect as we do” and then added, with a smile, “and they don’t eat pork” (fady kisoa; SAK: fady komankory).

Roland and his kin today run one of the most profitable small businesses in Ambanja. They may stay a decade or more, yet their actions reveal a decision to remain distinct from Sakalava. Although their clientele are of all backgrounds and they live in a Sakalava neighborhood, their closest social ties are with other highland people. Nearly all members of this group pray, for example, at a local Protestant church where the majority of parishoners are Merina and Betsileo. Since they are Merina, it is difficult, if not impossible, for them to become integrated into the local Sakalava community. Their choice to raise pigs in town is an obvious transgression of local Sakalava fady, one that asserts their separateness. Their marriage (and dating) patterns also reveal that they are ethnically isolated from other groups.

The Antaisaka Handcart Team

As stated before, in Madagascar, ethnicity is often paired with economic specialization, providing stereotypes for different groups. The Antaimoro, Antaisaka, and others from the southeast, for example, are well-known for their strength. Many of them work as pullers of handcarts called posy-posy (FR: pousse-pousse, lit. “push-push” or rickshaw). These handcarts have a seat that holds two people, shaded by an awning that unfolds over the seat when needed. This is an inexpensive form of transportation used in all major coastal cities, and the majority of pullers are from the southeast coast.

Ambanja has a team of five men who started a handcart business here a year ago. They range from twenty to twenty-six years in age and are originally from a village near Farafangana. Doda and Aimé had first worked in Diégo for fifteen months, going there by bush taxi directly from their ancestral land. There they worked for a man who owned several handcarts and who hired them as pullers and provided them with a place to sleep. Aimé had heard from customers that there were no posy-posy in the Sambirano, and so he came here to investigate. He returned to Diégo and arranged to have his own cart built, and then encouraged Doda to join him as a coworker in Ambanja (figure 4.2).

Within a few months, these two men had set up a business in Ambanja and were then joined by Franz, Tiana, and Piso (“The Cat”), who are from their home village in the southeast. The five men share the small cramped space of a two room house made from traveler’s palm. Here there is barely enough room for the five men to lie down at night. Each man has his own handcart, and although they do not take passengers(the town is too small for pedestrians to want to hire them) they do a lucrative business by carrying goods from one part of town to another. The largest share of activity involves delivering bottled drinks to local bars and groceries. They make anywhere from 2,500 to 10,000 fmg each day. Together they pay 1,500 fmg per month in rent for their house, which they found out about through a Tsimihety friend who works at one of the beverage distributors in town.

All of these men intend to settle in their ancestral land after working in the Sambirano for several more years. Both Piso and Tiana have managed to go home since settling in Ambanja (Piso has done so once, Tiana twice). Only Doda is married. These men say they often seek out girlfriends (sipa) in town, but find this to be difficult, their experiences echoing those of the Vakinankaratra men. As Franz put it, “we prefer the Sakalava women, but they don’t like us! Really! They don’t like men who pull handcarts, they would be ashamed to be seen with men like us.” Piso then added, “I love to eat pork, and so no Sakalava woman will have me.” They all hope to get married, but prefer to settle with women back home, finding them to be more serious about their commitments to marriage.

These men are not unusual for many young temporary migrants. They have come to the Sambirano to make a living and earn savings and then return home, where their sentimental ties lie. Although they may attempt to socialize with local Sakalava—especially in reference to sexual liaisons with young women—they remain outsiders and look forward to later establishing more long-lasting relationships back home with women of the same ethnic background.

The Botabes

Laurence Botabe is fifty years old and is the patriarch of one of the largest corporate kin groups in Ambanja. Twenty-three kin members live in one compound, and another nineteen live close by in other parts of Ambanja or the Sambirano (figure 4.3). Laurence has been married twice (his first wife died ten years ago), both times to Sakalava women. He and his siblings conjointly own land in town which includes five houses where they live with their respective wives, children, and grandchildren. The land they own was originally purchased by their father, a Tanala man from the southeast who came to the Sambirano more than half a century ago to work for what were then sugarcane plantations. Their father married a local Tsimihety woman who had come to Ambanja with kin from Maromandia (ninety-six kilometers south of Ambanja). When their father died, Laurence and his siblings decided to occupy the homestead together rather than split up the land, and it is Laurence and his brother who manage it now. Originally there were three houses, but they have since built a fourth structure that serves as a kitchen and dining room. Together they also purchased a two-room building across the street where their widowed sister now lives. Several of the teenage boys sleep in the room next door.

All older kin group members living in the compound work for wages, unless they are students. Laurence holds the salaries of his children in a safe place (and thus maintains control over how they spend their cash). The economic activities of the household’s members are diverse and generally quite profitable. This kin group also owns collectively several hectares of land in the countryside, where they grow enough rice to feed the entire homestead for eight months out of the year. The Botabes live simply yet comfortably, and they have enough capital to enable them to send the majority of children to the private Catholic school in town (they are Catholic and all children have been baptized). Laurence, who is in his late fifties, recently retired as a truck driver for a local enterprise. He now helps his wife run several businesses out of their home. Both travel into the countryside to purchase goods that they sell out of their compound, Laurence buying supplies for building houses of ravinala palm, and his wife charcoal from nearby foresters. One daughter is a seamstress and another is a typist for a local lawyer. One son, in addition to being a high school student, works for the local radio station during his vacations, playing songs that are popular in the discos and at morengy boxing matches.

The Botabes are well-rooted in the community, especially in terms of their ties to those active in the local political machinery. Laurence knows many powerful people at the enterprises through the job he has held for much of his life. Having married Sakalava women, Laurence also has ties to the Bemazava royalty, since his second wife’s sister married a member of the royal family. His brother, who was recently elected as the neighborhood president, is responsible for taking census information and serving as a local mediator for disputes.

Unlike Guardian, Roland and his kin, and the members of the handcart team, Laurence considers himself to be tera-tany, since he was born and raised in the Sambirano, owns land, and has married Sakalava women. Nevertheless, townspeople (especially Sakalava) insist that the Botabes are vahiny, because the older male members of this kin group are considered to be Tanala by birth, and this affects subsequent off-spring as well. (Many townspeople are also under the mistaken impression that he is Antaimoro because his father was from the southeast.) This categorization as vahiny becomes blurred, however, as one moves down through the generations. Laurence’s children, having been raised by Sakalava mothers, consider themselves to be Sakalava and are so regarded by many other Sakalava, rather than being Tanala like their father. For example, the son who works for the radio station is becoming famous throughout the Sambirano. He is an object of great local pride since he prefers to play salegy music with lyrics that refer to local events and items that are typically Sakalava. Postpartum practices, which will be discussed below, also help to endow the offspring of unions between Sakalava and non-Sakalava with Sakalava qualities.

Mme Razafy, Merina Schoolteacher

Mme Razafy is twenty-eight. She was born of Merina parents and was raised in Morondava on the southwest coast. After completing high school she went to Antananarivo to study at the university, where she received her degree in education. It is here where she met her husband, who is Merina and from Antananarivo.

In Madagascar, schoolteachers are government employees and must apply to the state for job placement. Jobs are very scarce, and so when she and her husband received assignments in towns that were more than three hundred kilometers apart, they felt compelled to accept them, living separately and visiting each other only during school vacations (this was in 1982). She described the experience as follows: “The second year was especially difficult. I became pregnant and had a child. It was then that I decided to take a leave of absence for a year and join my husband…and I worked at home as a seamstress to earn more money.”

“Three years ago we finally managed to acquire jobs together in Ambanja where I was originally stationed.…Once we were both working again I tried to hire a maid to help me, but I had little success.…I went through a series of local women, none of whom worked out: several did not show up for work, one left town suddenly with a lover bound for Nosy Be.…Another made me so angry! She pinched the baby all the time.…Then my [mother’s sister], who lives in Antananarivo, eventually sent me a maid…This was a young Merina woman named Fanja [age twenty-one].…Fanja has been living with us for two years now .” I later asked Fanja how she felt about living in Ambanja, to which she replied, “tamana, tamana be aho! [I am very content here].” When I asked why, she replied, “At home there are too many people, all of them hungry. Here in Ambanja there is work and food, and I have enough money to buy some nice clothes, too.”

In contrast to Fanja, Mme Razafy and her husband hope to be able to leave Ambanja in the next few years. “It is difficult to teach in the school here, because the students are hostile toward us because we are Merina.^[4]. . . The students are not very motivated, either; they are more concerned with matters of love and money, not schoolwork.…We would like to resettle in Antananarivo, but it is very difficult to find a teaching post there.” Mme Razafy, unlike Fanja, is not content: “I am homesick [ngaoma] for my mother and sisters because they live far away in Antananarivo.”

“Te Ho Mody”: I Want to Go Home.

There are two concepts that are especially important for migrants when they are expressing their sentimental ties: these are tamana (contentment) and ngaoma (homesickness); among people who express the latter, they also use the expression te ho mody (the desire to return/go home). It is through these constructs that inhabitants of Ambanja articulate their experiences in this town. They are used deliberately either as statements of one’s personal feelings, or as questions put to others. As a foreigner, I was asked almost daily whether I was tamana or ngoama.

Because of the availability of work and the quality of the soil in the Sambirano, physical survival is rarely a problem. It is often said that in this region of Madagascar no one is really poor and rarely do people go hungry. Although I have seen sick and emaciated individuals—especially children of migrants with large families—this is a rare sight. Ambanja and other northern towns are distinct in this way from much of the rest of Madagascar, where many people are starving and where the streets are full of beggars, especially children and women. Antananarivo has a large population of people living on the street who have lost their houses, crops, or cattle to droughts and cyclones. These are the katramy (from the French quatre-murs, called this because they are people who are without “four walls”) . In Ambanja, this term is not known. Throughout 1987 there were only three people who begged on a regular basis, two of whom were considered to be mad by townspeople. My point here is that it is not simply economic problems that make life difficult for Ambanja’s migrants, but social and cultural ones, all of which are tied to the tensions and factionalism that operate locally and on a daily basis. Ngaoma (homesickness) and tamana (contentment) are sentiments that help differentiate between the experiences of temporary migrants and settlers.

Tamana is a term that is used by those who feel they belong in Ambanja. These in general are the tera-tany and permanent migrants or settlers. For example, people who would use these terms are Laurence Botabe and his kin and Mme Razafy’s maid, Fanja, who has no desire to return home. For vahiny, however, tamana is not devoid of tension, for in order for a migrant to feel tamana in the new place of residence, it means that he or she has left behind (and forgotten) other kin who remain in the original ancestral land. To neglect kin, ancestors, and ancestral land is a trying experience for any Malagasy. It is rare that a migrant feels no uneasiness about this. One must wait for the offspring who, born locally, are zanaka Sambirano, or “children of the Sambirano,” before sentimental ties to the original ancestral land are weakened or altogether broken. Temporary migrants are more likely to say they are ngaoma. It is they who feel they do not belong and who stay in the Sambirano Valley for economic—and not sentimental—reasons. Examples include Mme Razafy and the Antaisaka handcart team.

There are also other individuals who fall somewhere in between these two extremes. Examples include Guardian, especially after he was joined by his wife and after he had accumulated enough capital to invest in several heads of cattle, and Roland, who has been joined by kin and who runs a successful business in Ambanja. Neither, however, intends to settle permanently. On a daily basis they say they are content (as Guardian says with a Sakalava inflection: “Oh, the cattle are so big and fat here!” [bodabe ny omby an-Ambanja é!]), yet they still miss things and people more familiar to them and they look forward to the day they can return to their respective ancestral lands. Both Guardian and Roland have married women from their own regions and tend to associate with peoples of similar ethnic backgrounds.

Another expression that distinguishes the content from the malcontent is tsy malala fomba (lit. “don’t know how to do the customs”). This expression implies that an individual has coarse manners and was ill-raised. It is generally only used to describe children when they commit faux pas; when applied to adults it is an extremely rude thing to say. In the context of interactions between vahiny and tera-tany, it takes on a special character, and it is used by people of one background to describe the customs of others. Sakalava sometimes use it when they joke about vahiny ways of doing things. More often, it is used by migrants (especially those from the high plateaux) when they reflect on Sakalava customs. In reflecting back to the categories of temporary migrants and settlers, it is more common to hear this expression voiced by temporary migrants to express their disgust for Sakalava customs. Settlers are less critical of Sakalava, since typically they choose to be assimilated as or at least be accepted by Sakalava.

The question of where the ancestral land lies for migrants is a complicated one. Even settlers sometimes (especially among those of the first generation) prefer to be buried in the area where they were born, and they or their kin will spend enormous amounts of money so that the body of a deceased member may be transported to a family tomb located in the original ancestral land. For example, the parents of the Betsileo described earlier came to the Sambirano more than fifty years ago, yet they plan to be placed in a tomb in the southern plateau region near Fianarantsoa. Their son, however, is married to a local woman who is Indian métisse. Since it is fady for a non-Betsileo to be placed in his parents’ tomb, he plans to be entombed locally: in this way his body can lie with his wife’s. He has already built a tomb in the Sambirano for this purpose. Among the temporary migrants, only Antandroy as a group exhibit a pattern of being buried locally, because, as they explain, the cost for transporting a body more than one thousand kilometers to the southern end of this island is prohibitively high.

Patterns of Association and Means for Incorporation

The above case studies reveal how migrants who have recently arrived in Ambanja seek the assistance of kin and, sometimes, friends from their ancestral lands. There are also a number of institutions that help them to meet each other and to become involved with settlers and tera-tany. As the African literature illustrates, voluntary organizations and other forms of association are important for helping migrants to become integrated (and to be successful) in urban communties (see Cohen 1969; Epstein 1958; Hellmann 1948; Hopkins 1972; Little 1957, 1962; Mayer 1971; Meillassoux 1968; Powdermaker 1962). In Ambanja such institutions exist and they may be formal or informal. Some enable migrants to maintain ties with people of similar ethnic backgrounds, and others facilitate their assimilation into the local Sakalava community. Although tromba possession primarily enables women to become active in local Sakalava power structures, there are other local institutions that are more readily available to men and women of a variety of ethnic backgrounds. These include churches, mutual aid societies, political organizations, and sports. The following examples briefly illustrate the most common experiences.^[5]

Preserving Ethnic Distinctions: Mutual Aid Associations

At one extreme are ethnically based, mutual aid societies that are designed to help individual members in times of need. Because of the composition of their membership, they also help to maintain ethnic factionalism in the town, or, as one Betsileo friend put it, “they only promote regionalism and racism, problems which are already too much a part of life in Madagascar.” A number of such organizations exist in Ambanja, and most of these are registered with the city government. These include an association for people of the Mahavavy (an area to the north of Ambanja, near Ambilobe, which is inhabited by Sakalava and Antakarana); another for the Sakalava of the Sambirano; one for migrants from Vohemar and other nearby towns on the northeast coast (inhabited by Antakarana and Betsimisaraka); a neighborhood association where the majority of the inhabitants of that section of town are Antandroy; and three Tsimihety associations. During my stay in 1987, a Betsileo association was also being formed. There are certainly other, nonregistered groups. The members of these associations come from a wide variety of economic and educational backgrounds. Each pays dues regularly which are used by the association to assist members when they are sick or, should they die, to help transport their bodies back to their ancestral land.

Polycultural Groups

The middle ground is occupied by an assortment of formal groups and organizations as well as informal associations that tend to be polycultural in their makeup. These include groups that are based on ties developed at the workplace, local churches and mosques, different subgroups of the national party which exist for youth, workers, and specific neighborhoods, and, among men especially, soccer and other athletic teams. Studies from the colonial period throughout Africa reveal that such organizations may operate not only as social clubs but as forces for political organizing (see, for example, Epstein 1958 and Powdermaker 1962). In Ambanja, church organizations provide potentially powerful lobbying groups for local social causes, such as water, sewage and garbage control. The Alliance Française is not only a popular gathering place in the evenings for the local educated elite, but it also exerts pressure from time to time on educational policies in local schools. Athletic clubs are rallying groups for men during their off hours. One form of athletic activity which has become highly suspect in Madagascar is kung fu and other forms of martial arts, since the government fears their potential military strength.

Because of the great tension between tera-tany and vahiny, it is extremely difficult for outsiders to become insiders and, in turn, to become involved in Sakalava power structures. Polycultural groups provide a few ways in which migrants may gain access to politically powerful positions. This is difficult, however, and possibilities differ for men and women. Men, for example, try to become involved in local politics and seek to be elected to the town or county government, but so far only Tsimihety have managed to do this. Civil service appointments, which are made through Antananarivo, present another possibility, and in this context many vahiny (especially Betsileo and Merina) have been appointed to work in the Sambirano as teachers, agricultural extension agents, and directors of the local enterprises. The managers of the enterprises occupy very powerful positions, especially in relation to the national and local economy. Since they tend to be wealthy they are also able to buy favors and manipulate the outcome of decisions in the local court. Nevertheless, they are very unpopular because of their ethnic backgrounds. As a result, recently the staff in the central offices in Antananarivo have attempted to replace them with others whose backgrounds are more acceptable to Sakalava. For example, one director, who has become very popular in the area, is a métis whose background includes French, Sakalava, and Betsileo heritage. He has married a local Sakalava woman and their children were born and raised in the Sambirano. In addition, they live in the heart of town, rather than on company grounds fifteen kilometers away, as is the case for several other local directors. Because he maintains a favorable reputation among vahiny and tera-tany alike, each morning people confronted with financial or other problems line up outside his door. Other vahiny who fail to acquire such positions of power may have access to these individuals through their membership in polycultural associations.

Changes to Sakalava Social Structure

To understand how Sakalava identity is defined and how their authority is asserted, it is important to outline briefly recent structural shifts that have occurred. The most important distinction that must be made is that between commoner and royalty. Sakalava royal descent is based on truncated patrilineages that are traced primarily through past rulers. These lineages are still preserved and recorded with care by living royalty. Nevertheless, the undermining of royal authority by the French has led to a shift in focus among commoners from living to dead rulers.^[6] Today the names and deeds of dead rulers are preserved and celebrated primarily within the context of tromba lineages (see Part 2).

Prior to French conquest, Sakalava commoners defined and organized themselves in reference to their royalty. Commoners were grouped into small village-based clans (firazan̂a).^[7] These political entities were organized hierarchically in reference to specific duties or work that each was expected to perform periodically for royalty (fanompoan̂a; asam-panjakan̂a) (see the discussion on work in chapter 2). Clan membership was defined in reference to shared territory, ancestors, and tombs. In addition, there were the Sambarivo, royal slaves of diverse (often non-Malagasy) origins. Although they occupied the lowest rank, the Sambarivo were considered to be closest to royalty, since they were (and still are) responsible for tending to living royalty and the royal tombs. Sambarivo, however, as slaves, had no ancestors (see Feeley-Harnik 1991b). To quote Feeley-Harnik, “ ‘clans’ seem to be seen as having a lineage-like cohesiveness that is not born[e] out by actual relations on the ground…firazan̂a [are] very much a product of local political relations” (Feeley-Harnik, written personal communication). In addition, Baré, in his study of the Bemihisatra-Sakalava of Nosy Be, also stresses that clan membership was sentimentally determined by the “cycle de développement” (1980: 193):^[8] clan identity often hinged on where one had spent much of his or her life. In general these clans appear to have been bilaterally conceived and exogamous (although endogamous unions did occur following the performance of specific rituals that lifted marriage bans or fady). Since clans were ranked, marriages between particular clans were deemed to be the most appropriate unions.

In addition to clans, a second form of affiliation was the kindred (tariky), which was composed of matrikin and patrikin. Residence usually determined how one conceived of one’s tariky. Different clans showed different biases toward patrikin vs. matrikin, with sentimental ties affected by residence patterns determined by the political order. Virilocality was the norm, but Sambarivo women, for example, could not leave their native villages. Similarly, members of high status groups tended to be uxorilocal. Often choice of residence after marriage was determined by the advantages and disadvantages perceived by cognates and their offspring. Baré cites an example of a man whose father settled uxorilocally, and so Baré’s informant was therefore raised among his mother’s kin. As an adult he had rights in his father’s village, but he preferred to remain where he had been raised. In order to be accepted as a full member of his mother’s kin group, he had undergone a series of ceremonies. As he told Baré: “je ne suis pas en bonne position ici (tsy tamana)” (Baré 1980: 92; note that this is the same expression used by migrants in Ambanja to mean they are “not content”).

The third organizing principle was the distinction made between the “children of men” (zanakan’lahy) and the “children of women” (zanakan’vavy), which distinguished paternal and maternal kin. In conjunction with this was the concept of uterine kin, or “children of the same belly” (kibo araiky) who shared their mother’s ancestors. These principles set restrictions on marriages between the children of sisters and excluded foster or adopted children from full participation in their adopting clan’s royal work.

In the Sambirano, clan membership is no longer a significant organizing principle. Clans have virtually vanished, save for the Sambarivo, who still live separately in special villages in the Sambirano Valley and on Nosy Faly, where they guard the royal tombs. Today few Sakalava are aware of these categories and, if they are, rarely do they know from what group their own ancestors came. Although two young adult informants stated with pride that their grandmothers had been “servants to the royalty” (mpiasam-panjakan̂a), they did not know what their duties had been. A major change that has occurred is the use of broader terms to define identity. Whereas two or three generations ago clan names served as markers for personal identity, today people of the Sambirano resort to using broader categories to define themselves as tera-tany, stating that they are Bemazava or, more generally, Sakalava. This practice stands out in contrast to the Sakalava of the Analalava region, for example, where so broad a term as “Sakalava” is rarely used by the indigenous people to label themselves (Feeley-Harnik, personal communication). Today it is the category of kindred or tariky that is important. As mentioned above, the kinship system of commoners is bilateral.

Sakalava kinship terminology is fairly complex (figure 4.4). Kin terms vary according to the sex and sometimes with the relative age of the speaker. Labels for siblings are as follows: for a male ego, brother is rahalahy, sister is anabavy; for female ego, brother is anadahy, and sister is rahavavy. The term used to address older siblings (male or female ego) is zoky and for younger siblings it is zandry. Structurally, parallel and cross-cousins are regarded as being the equivalent of siblings. The siblings of ego’s parents are differentiated from one another, again according to the sex of ego. Thus, for either a male or female ego, on the father’s side: baba (father), bababe (“big father,” or father’s older brother), babakely (“little father,” or father’s younger brother), and angovavy (FaSi). On the mother’s side: nindry/Mama (mother), nindrihely/mamahely (“little mother,” mother’s younger sister), nindrihely/mamabe (“big mother,” mother’s older sister), and zama (MoBr). Spouse’s siblings are ran̂ao, and in turn their spouses are structurally regarded as siblings. In the past, there was a joking relationship between agnates of opposite sex (for female ego: rokilahy, for male ego: rokivavy). A special term, asidy, is used instead of zanaka for the children of ego’s opposite sex sibling.

As figures 4.4 and 4.5 show, changes have occurred in kinship terminology and they are significant for several reasons. First, they reflect the effects of Sakalava contact with French and other peoples. Various indigenous terms (and, thus, structural concepts) have been replaced with French ones. Sakalava now use the terms mama(n),papa, and tonton. Sakalava younger than thirty or so are unfamiliar with the terms that designate joking relationships with agnates (rokilahy and rokivavy). Also the term asidy, for sibling’s children, is used less frequently. Finally, Sakalava kin terms (and the shifts that have occurred in the movement from village to town) are important in the context of tromba possession (this will be discussed in chapter 7). To illustrate these changes, figure 4.4 provides the older (“village”) kinship terms, while the second set in figure 4.5 corresponds to those used in town.

These structural changes—where clan membership and older (village) kinship terms have been forgotten—illustrate an important shift in how Sakalava conceive of their collective identity. Whereas in the past clan membership served to distinguish different groups of Sakalava from each other, today other categories operate to distinguish Sakalava from non-Sakalava. The broadest of these are tera-tany and vahiny. The use of Sakalava kin terms also serves as a marker for Sakalava identity or affiliation: terms such as angovavy and zama are distinctly Sakalava. Ironically, so are maman, papa, and tonton, since Sakalava are unusual among Malagasy speakers in reference to the number of French terms they have incorporated into their vocabularies (Dalmond, n.d.).^[9] When vahiny use French-derived Sakalava kinship terms they do so to signal their affiliation with the tera-tany, since kinship terms vary somewhat as one moves from one region of Madagascar to another. Sakalava kinship terms are also very important in the context of tromba, where they are operational in establishing links of fictive kinship among the living and between the living and the dead (chapter 7).

Marriage Across Ethnic Lines

At first glance marriage would seem to offer vahiny an easy means for their integration and acceptance by tera-tany, but Sakalava ideas about paternity (which will be described in more detail below) and ethnicity make this difficult. Sakalava attitudes towards non-Sakalava vary, and so it is much easier for Antakarana and Tsimihety to form unions with Sakalava than it is for groups from the south or high plateaux (it is almost impossible for Merina). The experiences of the men of Roland’s factory and those of the Antaisaka handcart team provide examples of this.

Nevertheless, the rate of ethnic exogamy is high in Ambanja when compared to other regions of Madagascar. Unions between Sakalava and non-Sakalava have led to a greater blurring of ethnic boundaries, so that defining what it means to be Sakalava is becoming increasingly problematic. As a result, tracing lines of descent and reckoning ethnic affiliation in such unions, local informants now resort to rules that are legal or what I will refer to as bureaucratic in nature. Today, by national, semilegal ways of reckoning, the father’s ethnic identity determines the “official” ethnic label to be used for offspring. This breaks down with each subsequent generation. The case of the Botabes provides an example of this: Laurence is considered to be a vahiny even though he was born and raised in the Sambirano, owns land, and has consecutively been married to two Sakalava women. Officially his children are Tanala like him and his father. By less formal rules, however, his children are generally accepted as tera-tany (they are referred to as being Sakalava), since they have grown up among Sakalava and they were raised by Sakalava mothers. Culturally (and physically) they dress, act, and speak like Sakalava, and so their status as tera-tany is rarely questioned. In other words, it is the children—or, in the case of the Botabes, the grandchildren—of migrants who may be considered to be tera-tany. Thus, marriage unions between Sakalava and non-Sakalava do not affect the ethnic identities of spouses but, rather, only potentially those of their offspring.

Some settler families have developed creative measures for determining an individual’s kin group in cases where relatives are of different ethnic backgrounds. Again, the Botabes provide an apt example. In this kin group, all female children are given a surname derived from the names of their mother’s female kin, and male children carry a different surname that they share with a male member on their father’s side. This is a creative embellishment on a bureaucratically defined assumption, since many Malagasy people do not have surnames. As a result of this system of naming, female children tend to be sentimentally tied to their mother’s Sakalava kin, but male children are closer to their father’s Tanala kin. This system also determines patterns of inheritance among the Botabes, so that female offspring inherit from tera-tany, and male from vahiny.

Sakalava marriage as an institution has also been affected by the changes that have occurred in response to colonial presence. These changes in turn further complicate the manner in which the affiliation of offspring is determined in any generation. In precolonial and early colonial times, Sakalava households were defined by monogamous or polygynous unions, their offspring, and extended kin. A marriage ceremony (fehim-badian̂a, Baré 1980: 212) was performed to legitimate the union, but now this ceremony is rarely undertaken. Laurence Botabe, for example, performed such a ceremony with his first wife, but not with his second. Among his children, only one daughter has done so. Her other married siblings went either to City Hall or the Catholic church to confirm their unions. In the urban setting in particular, indigenous marriage ceremonies have been uncommon for the past two generations. Although marriages that are sanctioned either by institutionalized religion (Catholic, Islamic, Protestant) or by City Hall do occur, they are infrequent. Unions are extremely fragile. By the time they have reached their forties, most adults in Ambanja have been involved in a series of unions, each of which has lasted only a few years.

Linguistic analysis reveals that this abandonment of marriage ceremonies has become institutionalized. The term vady is used in Ambanja (and neighboring areas of northern Madagascar) to refer to an individual’s primary partner, be they a spouse (through a union confirmed through a ceremony) or a lover (common-law spouse).^[10] When Bemazava-Sakalava of Ambanja are compared to Malagasy in other parts of Madagascar, what is striking is that elsewhere the term vady is usually used only to refer to unions confirmed through a ceremony. Baré also notes this, saying the term has become more popular among the young generations of the Bemihisatra of Nosy Be, but he stresses that “adultery” or relationships outside of marriage (vamba) were common in earlier generations as well. Bemazava informants report that unions of their grandparents’ and older generations lasted much longer: although several unions may have occurred in a lifetime, the turnover was much less frequent. Interviews with elders on their marriage histories confirm this. Feeley-Harnik (1991b: 200ff) also describes how, for the Bemihisatra of the Analalava region, the search for a compatible partner (vady; namana: “friend”) is an important yet difficult quest; the same is true in Ambanja. As will become clear in chapter 8, tromba mediums specialize in helping clients solve problems associated with romance and marriage.

Today, many children of the Sambirano are the offspring of such tenuous unions. Some remain with their mothers, although the majority are raised by their mothers’ kin (typically by their maternal grandmother) while their mother lives and works elsewhere, sending remittances back to her mother. In cases where the father of the children is absent (and, perhaps, his identity not known by others), children are considered to be culturally the same as their mother and structurally they are considered members of their mother’s kin group. Thus children of a Sakalava mother and an absent non-Sakalava father will be considered to be tera-tany. There is also a postpartum practice that is distinctly Sakalava, one that helps to incorporate non-Sakalava women and their offspring into the Sakalava community.

Post-Partum Practices as a Form of Ritual Incorporation

Determining the ethnic affiliation of a child who is born of a mixed union between a Sakalava and a non-Sakalava parent can be problematic. By the bureaucratic rule of reckoning, as mentioned above, the child is Sakalava only if his or her father is. In the past, the identity of the mother often was important for determining clan affiliation among the Sakalava. Also, uterine siblings or “the children of women” (zanakan’ vavy) had special privileges in traditonal Sakalava society that were not extended to adopted children. These categories still operate in ritual contexts, where zanakan’ vavy and zanakan’ lahy (“children of men”) must both be present and participate if a cermony is to be performed properly. Although clan affilation is no longer important, uterine ties are operative. For example, today the responsibility for raising children of temporary mixed unions usually falls with the mother’s kin, as mentioned above. Also, postpartum practices that are distinctly Sakalava help to determine the ethnic identity of offspring of mixed unions. Through such practices non-Sakalava women who bear children by Sakalava men may temporarily and symbolically become Sakalava.

All Malagasy but northern Sakalava are said to be “hot” or mifana (SAK; HP: mafana) in reference to their postpartum practices.^[11] During a specified period (typically six weeks or forty days), a woman with a newborn child is expected to take special precautions to keep herself and her infant warm, so that she will heal quickly and so that her infant will grow strong and remain healthy. Following labor, the umbilical cord and afterbirth are put in a warm place to ensure the protection of the child. The mother must also bathe several times a day with hot water and eat heated food. Some Malagasy place hot coals under the mother’s bed to keep her warm (see Hart et al. 1965 for descriptions of similar practices of “mother roasting” in southeast Asia). Among northern Sakalava, postpartum practices are very much the same as that of other Malagasy, with one important difference: they alone are ranginalo, observing a postpartum categorization that is “cold” rather than “hot.”^[12] Following ranginalo involves placing the umbilical cord in a cool place, often inside the house, and burying the placenta in a cool, shaded spot in the courtyard. The mother is expected to bathe in cool water several times a day and eat only cold or cooled food.

Marriage among peoples of similar ethnic backgrounds presents no problems as to which practice should be followed, since both parents share the same tradition. When Malagasy marry non-Malagasy, the mother must decide between doing what her own mother did and what is common among her non-Malagasy affines. In choosing one practice over the other, the mother is making a statement about where her sentiments lie, as well as about the identity of her child. I met several Chinese women married to Malagasy men who each chose to follow their husband’s mother’s orders. Malagasy women married to non-Malagasy men, however, tended to do whatever their mothers had done. In other words, women in mixed marriages that produced métis offspring generally follow the Malagasy custom deemed appropriate by their Malagasy kin. Most informants agree that Chinese, Comoreans, and Arabs are hot. As for Europeans, no one seemed sure, although since most Europeans prefer warm bath water to cold, they tend to be categorized as hot as well. No problems arise between most Malagasy and non-Malagasy, since the practice is basically the same.

It is in marriages between northern Sakalava and other groups where the decision regarding postpartum practices is complex. Sakalava insist that raginalo or “cold” is stronger (mahery) than “hot,” and so in cases where one parent (mother or father) is cold, it outweighs hot. If the mother does not practice ranginalo, she runs the risk of having great harm befall her child. In general, for Sakalava, coldness (manintsy) is associated with health. Coolness promotes healing: this is why water and kaolin (a type of clay; SAK: tany malandy or “white earth,” HP: tany fotsy), which are attributed with cool properties, are used by tromba mediums to heal patients. A Sakalava word for “sick” is mamay, which also means “hot.” Heat (mafana) is associated with danger and death, principles that are the opposite to those of other Malagasy groups.^[13] For the Sakalava, for example, June and July are said to have “cool” qualities, and it is during these months that royal work is performed at the tombs on Nosy Faly. When a royal person has recently died, it is said to be a “hot” time, and so no one may enter the tombs.

In interviews with thirty women of assorted backgrounds who had borne children from mixed unions, all but two said they had chosen to practice cold over hot. In both cases these women insisted that they felt more comfortable in doing what their mothers had done. One was living with her mother at the time, and they both thought it riskier to try to follow a practice with which neither one was familiar. The other woman was Sakalava and married to a man of Arab and Comorean heritage (hot). At first she practiced hot, but she fell ill. She was then instructed by her husband’s mother to switch to cold, after which her health improved. As one informant put it, “If you have to make a choice, go with cold, because it is the stronger of the two.” The diagrams provided in figure 4.6 illustrate how this system works.

Structurally, the choice to observe the cold over the hot practice has a number of effects. From the point of view of the male Sakalava parent, it is straightforward. Having the child’s mother follow the cold practice provides him with a means to assert his paternity, since it symbolically illustrates that the child is endowed with Sakalava qualities. (The child is already recognized as such by the bureaucratic rule that stipulates that a child shares the ethnicity of its biological father.) Since the mother must use the cold system, her offspring—as well as all future generations of children—will always be considered to be cold. If the mother is Sakalava and the father is not, however, the effect is different. It serves to label the child as being partly Sakalava and endows it and its progeny with coldness, a Sakalava trait. As a result, here it works to assert the maternal tie to the child, regardless of the bureaucratic rule. In this way it also distinguishes the child from its father, since the child will be cold like its mother while the father is hot. In cases where the father has left or his identity is unknown by others, the decision made by the mother following labor will serve to identify the child as hot or cold (and ultimately as Sakalava or something else). The offspring of a woman who has performed ranginalo (cold) are viewed as having Sakalava characteristics and they, in turn, will consider themselves to be cold when they are adults and have their own children.

Ranginalo, as a Sakalava postpartum practice, affects mothers in ways that it does not affect fathers. In unions between Sakalava men and non-Sakalava women, this cold system serves to incorporate women into the community of Sakalava. Each time that a non-Sakalava woman has a child by a Sakalava man, she must perform ranginalo. In so doing, she participates in a Sakalava institution and generally does so in the company and under the direct supervision of those who know how to do it properly: other Sakalava women. Since men are not involved in postpartum practices, it does not operate in this way for them. Instead, it serves to exclude or distinguish non-Sakalava men from their Sakalava spouses (and children). In essence, maternal ties to the child take precedence or they override paternity.^[14]

To summarize, although the first generation of male settlers will never be considered to be Sakalava, non-Sakalava women become incorporated as insiders following the births of any children whose fathers are Sakalava. All offspring of mixed unions—regardless of whether it is the mother or father who is Sakalava—will be endowed with a quality that is distinctly Sakalava. In this way, through ranginalo, offspring are retained as Sakalava, so that Sakalava-ness is not altogether lost through the effects of polyculturalism.

As this chapter has shown, significant elements of Sakalava social structure have changed dramatically within this century, but some continue to work to define what it means to be Sakalava. In the past, various categories distinguished Sakalava royalty from commoner, and, in turn, different clans of commoners from one another. Today broader categories work to distinguish Sakalava as a whole from all other groups: Sakalava conceptions of their own ethnicity, on the one hand, coupled with migrants’ participation in forms of economic specialization and voluntary organizations on the other, foster the continuation of ethnic isolation. These factors maintain the distance between adult tera-tany and vahiny. Even though marriage fails as a means for vahiny to become integrated as tera-tany, kinship is, nevertheless, the organizing principle that allows for the integration of outsiders. Postpartum practices, for example, operate in such a way as to temporarily incorporate non-Sakalava women and to permanently incorporate their offspring. As will become clear in Part 2, these same structural principles are key to understanding the significance of tromba for tera-tany and vahiny.

Notes