webAfriqa / Library / Anthropology

webAfriqa Custom Search Engine

George Peter Murdock
1897 — 1985

Professor of Anthropology, Yale University

Africa. Its Peoples and Their Culture History

New York. McGraw-Hill. 1959. 456 p.

Previous Home Next

Explore also (a) the SemanticAfrica Peoples Vocabulary
(b) the Malagasy mind-mapping diagram

Part Seven
Cultural Impact of Indonesia
— 27 —

Madagascar, the fourth largest island in the world, is about 1,000 miles in length and some 300 miles in width and has a total land surface of nearly 230,000 square miles and a population of more than 4 million. Dspite its size and topographical diversity, its inhabitants reveal a considerable degree of homogeneity in language and culture. In ethnic composition, on the other hand, they represent a highly complex mixture of phyical types—Negroid, Mongoloid, and Caucasoid. The Negroid element predominates, especially in the coastal regions, whereas the Mongoloid element is strongest on the interior plateau. The Caucasoid element, much the least significant, is most noticeable in precisely the regions where there is definite historical evidence of Arab or European settlement. Malagasy culture, however, is basically Malaysian (Mongolian) with a strong Arabic (Caucasian) overlay and a relatively weak Bantu (Negro) component. The language belongs to the Malayo-Polynesian stock but has a modest increment of loan words from both Arabic and Bantu.
Recorded history gives little help in explaining or reconciling the above facts and may therefore be summarized very briefly. Dependable records do not begin until about A.D. 900, i.e., at about the time when the Arabs replaced the Persians as the dominant maritime power in the Indian Ocean and began the establishment of a series of strong mercantile colonies along the East African coast, first at Mogadisho and Brava but ultimately extending to Sofala. These Arabs were orthodox Sunnites, and as a consequence of their expansion various heterodox Arabs from Oman, Yemen, and Azania sought refuge in eastern Madagascar, eventually establishing themselves in the Antaimoro country in the southeast and introducing certain Arabian and Islamic cultural elements to the population already established on the island. The expanding Sunnites soon followed them to Madagascar for purposes of trade and established colonies, especially in the northwest. In contrast to their predecessors they were illiterate, and the culture they brought with them was basically that of the mixed Swahili type which developed in former Azania after the advent of the Bantu. They imported large numbers of Negro slaves to perform their own agricultural labor and for sale to the Malagasy. The arrival of the Portuguese in 1500 put an end to Arab domination. As Portuguese power waned, the British and French competed for control, with the latter achieving political occupation in 1896.

Merina Dancers
Merina Dancers.
(Courtesy of the Peabody Museum of Archaeology and Ethnology, Harvard University.)

These historical facts probably account sufficiently for the Arabian and Islamic elements in modern Malagasy culture and in part for the Negroid and Caucasoid elements in the population of the island, the one introduced through sla,·es and the other through traders and colonists. They leave unexplained, however, most of the problems raised by the complex ethnic, linguistic, and cultural composition of the Malagasy. Theories attempting to answer these problems are legion. They have, however, been built largely out of conjecture and inference, and they conflict with one another on practically every point. Their authors, moreover, with the notable exception of Linton, have lacked either experience in other areas or competence in general ethnology. Instead of discussing the theories, consequently, we shall first list a serie of facts which seem indisputable, indicating the prevalent theories which each renders untenable, and then formulate a historical reconstruction that is consistent with all the facts and accounts for them with the highest degree of probability.
First, no archeological site on the island has revealed either skeletal remains of any racial type other than the three already mentioned or evidence of any culture notably different from, and possibly antecedent to, that of the present Malagasy. We therefore reject the almost unilerally accepted theory of an earlier Pygmoid occupation, whether of Oceanic Negritos or of African Twides, and see no reason for assuming that the island was other than completely uninhabited when the first Malagasy arrived on its shores.
Second, competent physical anthropological studies of the Malagasy, when compared with similar studies of both African and Oceanic Negroids, have demonstrated an incidence of blood types and somatic traits consistent with the derivation of the Negroid element in Madagascar from a Bantu source but inconsistent with the Papuo-Melanesian origin postulated by Grandidier and others. Third, archeology on the adjacent continent has established conclusively that the Bantu did not reach East Africa until after the middle of the first millennium after Christ and that there were no other Negrors on the coast or its immediate hinterland prior to that time. This renders untenable the assumption of several leading theorists that a Bantu migration preceded that of the Malagasy to the island. The present egroid element must therefore have entered the island after it had been populated by other elements or else, in part, contemporaneously with the latter.
Fourth, the Malagasy language not only belongs to the Malayo-Polynesian linguistic stock but is especially closely related to the Maanyan language of south central Borneo, as was pointed out in the preceding chapter. We must consequently reject the common assumption that the Mongoloid element in Madagascar is derived primarily from Sumatra or Java, together with the alternative conjecture of a Polynesian origin.
Fifth, the fact that the dialects of Malagasy are similar, all of them being reported as mutually intelligible, suggests immigration by a relatively unified linguistic and cultural group. This contradict the various theories of the peopling of the island by diverse Indonesian elements arriving at different times and places from different points of origin, for in that case Madagascar should still preserve some measure of linguistic diversity.
Sixth, none of the remote islands of the Indian Ocean, such as Mauritius or Reunion, was inhabited at the time of their discovery by Europeans, nor has any of them revealed a trace of any earlier occupation, either permanent or transient. Coupled with the indiputable evidence of Indonesians on the Azanian coast, this fact argues strongly against the various theories which would bring one or more of the elements in the population of Madagascar to that island by direct voyages from Indonesia or the Pacific islands, and supports the alternative hypothesis of a coastwise passage via India, South Arabia, and East Africa.
With the underbrush of untenable theories cleared away, the landmarks of attested fact stand out in sufficient clarity and number to enable one to map the actual topography of events with only a minimum of inference. The story, as reconstructed, may be summarized as follows.
At some time prior to the beginning of the Christian era the coastal Maanyan of southeastern Borneo, close kinsmen of the people bearing this name who still survive in the adjacent interior of that island, had developed skills in navigating their outrigger canoes in trading with the inhabitants of nearby islands. Their territory lay along a trade route no less famous than the Sabaean Lane—one connecting Malaya, Sumatra, and Java with the southeast coast of China through the Philippine Islands and Formosa. The first ambitious overseas ventures of the coastal Maanyan doubtless occurred along this route. That it was being plied at the time, and had already been connected with the Sabaean Lane, is indicated by information received from Professor H. Otley Beyer of Manila, in a personal communication, that coins of Alexander the Great have been unearthed in the Philippine Islands in sites along the route.
With experience gained in this eastern maritime trade the coastal Maanyan ventured ever farther to the west—to the Nicobar Islands and Ceylon, to the Laccadives and the west coast of India, to southern Arabia, and ultimately to Azania, where we have already found evidences of them in the first century. Such dauntless mariners would certainly not have hesitated to venture farther south along the East African coast in the quest for profit, and in doing so they would inevitably have discovered the Comoro Islands and Madagascar, at that time completely uninhabited.
If they found products of commercial value in Madagascar, some would doubtless have settled there to exploit them. They would have found the virgin forests of the island ideal for the swidden (slash-and-burn) cultivation of the dry rice and the root and tree crops which they were wont to grow for subsistence on their home island. It is significant that the cultivation of wet, or paddy, rice, already established at that time in Java, Sumatra, and Malaya, had not yet been introduced to Borneo, where most inhabitants subsist by swidden cultivation even today.
In Azania the Maanyan probably functioned mainly as merchants and mariners. Those who settled there, temporarily or permanently, probably constituted an urban group, leaving agriculture to the local Cushites. The latter, who may be assumed to have perpetuated the crops (including perhaps ensete, or the Abyssinian banana) and the irrigation techniques of their forefathers, would have found the Malaysian banana and root crop highly acceptable but may well have rejected rice because they were not accustomed to swidden cultivation. This would account for the failure of this crop to become established on the East African coast even though it doubtless became the staple on Madagascar from earliest settlement.
Through intermarriage with the local Cushites the Indonesians may have incorporated a certain amount of Caucasian blood and even a trickle of Negro blood, derived ultimately from the Cushitic population of southern Ethiopia. With the arrival of the Bantu in East Africa this trickle must have increased to a steady stream. The spread of the Bantu will receive full consideration later. For present purposes it is sufficient to note the slave-trading record of the Arabs, the testimony of Zimbabwe, and the evidence of race mixture in the present-day Swahili, and to draw the conclusion that in Azania the Arabs, Cushites, and Indonesians alike acquired numbers of Bantu slaves, mixed with them, and produced hybrid offspring.
One gains an impression from the historical record that the India-Africa trade was open to the ships and merchants of all nations during the Yemenite and Persian periods but that after the Moslem Arabs achieved dominance—especially after their establishment, beginning in the early tenth century, of a series of fortified trading posts from Mogadisho to Sofala—they strove for a mercantile monopoly and sought to discourage their rivals of other nationalities. This was clearly the situation when the Portuguese arrived and shattered the monopoly by armed force. We may assume that the Indonesians and Chinese were among the unwelcome visitors who were thus gradually squeezed out. At any rate, as previously noted, all historical mention of Indonesians on the Azanian coast ceases completely by the end of the twelfth century.
Traditions and genealogies of the Merina tribe of the Madagascar plateau ascribe their ruling caste, the Andriana, to a migration occurring around the beginning of the second millennium, and practically all authorities infer from this that the lighter-skinned and more Mongoloid inhabitants of the interior of the island owe their origin to a separate and late movement from Indonesia. The traditions conceivably have a basis in the evacuation of the Malayan Azanians to Madagascar, but the assumption of a special migration has none. On the contrary, by this time the Indonesians of Africa had probably already been permanently cut off from their homeland, or at least would be before long.
The Indonesian and Bantu components in the Malagasy race and culture can now be simply accounted for. The original culture was exclusively Indonesian, and all immigrants prior to the middle of the first millennium were overwhelmingly Mongoloid. Later voluntary immigrants from Azania were at first predominantly Indonesian in culture but with considerable absorbed elements of Negro, Arab, and Cushitic blood, varying probably with the place and date of their departure, and later increasingly Negroid in physique with a culture of the emerging mixed Swahili type. During this period and after the final exodus from Azania the involuntary immigrants, or slaves, an element probably appreciably larger than either of the others, consisted exclusively of Negroes. As in the tropical New World, their effect upon the ethnic composition of the population was large but their contribution to the total culture was relatively small—and for identical reasons. The variable distribution of physical types by regions in Madagascar doubtless reflects specific historical influences which can no longer be reconstructed, although Linton's hypothesis of natural selection on the basis of the differential susceptibility of the two races to malaria may offer a partial explanation. The numerous Malagasy tribes can be conveniently reduced to eleven major groups, as follows:

  1. Anraisaka (Taisaka, Tesaki), with the neighboring Antaimoro (Anrimorona, Taimoro, Tenuro), Anrambahoaka (Tambahuaca), Antifasy (Antifasina, Taifasy, Tefasi) , and Sahafatra (Antimanambondro). They number about 420,000 and reveal a noticeable infusion of early Arabic physical and cultural traits.
  2. Amandroy (Tandruy), with the neighboring Anranosy (Tanosy, Tanusi). They number about 390,000.
  3. Bara, with the kindred Barobe, Imamono, Sautsaura, Timonjy, and Vinda. Their population totals about 180,000.
  4. Betsileo, embracing the Arindrano, Ilalangina, Isandra, and Manadriana. They number about 500,000.
  5. Betsimisaraka, with the Betanimena. They number about 630,000.
  6. Mahafaly (Mahafali). They have a population of about 80,000.
  7. Merina (Antimerina, Hova, Imerina, Ovah). These plateau people, markedly Malayan in physical type, have exerted political hegemony over their neighbors since about 1600. Their present population is about a million.
  8. Sakalava, embracing the Antankarana, Antiboina, Antifiherena, Anrimailaka, Antimaraka, Antimena, Antimilanja (Amambongo), and Vezu (Veso). Though probably the most Negroid of all in physique, they are the only group in Madagascar that still preserves the Indonesian outrigger canoe. They number about 300,000.
  9. Sihanaka (Antisihanaka). They have a population of about 85,000.
  10. Tanala (Antanala), embracing the Ikongo and Menabe and including the neighboring Bezanozano (Antaiva, Anrankay, Tankay). They number about 200,000.
  11. Tsimihety. Their population totals about 300,000.

The Malagasy peoples fall into four subareas, differentiated mainly in basic economy: the interior Plateau, the East Coast, the Escarpment between them, and the Plains of the west coast, which also includes the extreme north and south of the island. The tribes of the Plateau (Betsileo and Merina) depend primarily on irrigated rice cultivation, with auxiliary animal husbandry. The East Coast peoples (Antaisaka and Betsimisaraka) subsist by the swidden cultivation of dry rice, supplemented by fishing and animal husbandry. The Escarpment tribes (Sihanaka, Tanala, and Tsimihery) fall into an intermediate position, characterized in many instances by an incipient transition from dry- to wet-rice cultivation. The Plains peoples (Antandroy, Bara, Mahafaly, and Sakalava) are pastoral cattle herders, who also depend heavily on fishing near the coast and to a considerably lesser extent upon agriculture.
Among the major food complexes, the Malaysian easily ranks first. All its component plants are cultivated in Madagascar. Among them, rice is the staple everywhere except in the Plains; taro and bananas are both widely cultivated; and Polynesian arrowroot is important in the Plains. Nearly all the crops of the Indian complex are likewise cultivated, but they are relatively insignificant. Of the Sudanic complex, sorghum assumes first place among the Mahafaly, earth peas are widely cultivated, and smaller quantities of gourds, watermelons, sesame, and cotton are also grown. Among plants of merican origin, sweet potatoes assume first place among the Antandroy and Sakalava; manioc and maize are important everywhere; and considerable amounts of peanuts, haricot and lima beans, pumpkins, and tobacco are also raised. Many additional crops have been introduced by Europeans within the past century. The Malagasy keep relatively small numbers of goat , fat-tailed sheep, pigs (introduced by the Portuguese), dogs, cats, chickens, and other animals of more recent introduction. Cattle, however, are exceedingly numerous. Milk, which is used everywhere, constitute a major item in the diet of the pastoral tribes. Elsewhere, however, cattle are kept primarily as economic capital and for prestige. Their manure is utilized for fertilizer, but beef is rarely eaten except on sacrificial occasions. Fishing assume considerable importance among many tribes, but hunting and gathering are inconsequential. In the division of labor by sex, the men ordinarily hunt, prepare land for cultivation, and herd and milk cattle, whereas both sexes engage in fishing and agriculture. In the Plains, however, the men do mot of the fishing and the women most of the cultivation.
All groups occupy compact villages, commonly fortified with palisades. One house type prevails throughout the island-a rectangular structure with a thatched gable roof and walls of wood, bamboo, reeds, or mats. On the East Coast, dwellings are commonly elevated on piles. Boys are everywhere circumcised in infancy or childhood, but clitoridectomy is not reported. Warfare is endemic. This usually waged on a petty scale for booty in cattle and slaves, but among the Merina and some other tribes it has served as an instrument of state building. Headhunting and cannibalism are not practiced.
All Malagasy condone sexual freedom in the unmarried, and a period of trial cohabitation is often customary before a marital union is considered definitive. A bride-price of some kind appears to be universal though its amount is frequently nominal. General polygyny prevails sometimes, but never preferentially, in the sororal form. The first wife enjoys a favored status except among the Antandroy. A man always establishes his wives in separate dwellings and distributes his time equally among them. Residence is regularly patrilocal. Among the Tanala married sons attach themselves to their paternal households, forming patrilocal extended families. Among the Antaisaka and Bara this is true only of the eldest son, who will ultimately succeed as the household head; younger sons set up independent establishments. Extended families are also reported for several other tribes, but in most cases the reference seems to be to group socially united by ancestor worship rather than to common residential units, which are normally independent nuclear or polygynous families.
Sons and daughters share equally in inheritance except among the Antandroy, where sons alone inherit, and among the Bara and Tanala, where daughters are restricted to movable property. All Malagasy tribes are patri lineally organized into sibs and lineages, which are typically localized either as clan-communities or as clan-barrios. Sib exogamy, however, prevails only among the Antaisaka, Bara, and Betsileo. Elsewhere sibs, and usually lineages as well, are agamous or even show a tendency to endogamy, as among the Merina, the Sihanaka, and the Menabe subtribe of the Tanala. Curiously enough, taboos against sex relations and marriage are usually extended farther in the female than in the male line, though corporate matrilineal kin groups do not occur. Matrilineal exogamy is specifically attested for the Bara, Merina, and Tanala, and it may possibly be inferred for the southwestern Plains tribes from the fact that they forbid marriage with a mother's sister's daughter, though permitting, or even preferring, a union with any other type of first cousin. Kinship terminology of the Hawaiian type is attested for the Antaisaka and Merina. The Tanala, however, employ Iroquois terms for cross-cousins of opposite sex.
The Malagasy do not treat smiths or other artisan groups as pariah castes, but they recognize a sharp division into three endogamous classes: (1) hereditary nobles, often with royalty as a special subclass; (2) free commoners; and (3) slaves, who are recruited through capture in war or through enslavement for debt or crime and whose status, is transmitted to their children. The Bara and the Plateau tribes also distinguish a subclass of royal slaves, who possess some of the privileges of commoners. Political organization everywhere transcends the level of the local community with its headman and council of lineage heads. Even the peripheral tribes possess despotic paramount chiefs, and the more central groups all have well-developed states with monarchical institutions. Political integration reaches its apex in the conquest state of the Merina, with its royal capital and elaborate court, its private council of specialized officials, and its territorial administrative organization. According to Grandidier, the ruler, being regarded as divine, is exempted from the usual incest taboos and may marry a sister or other close relative. Reminiscent of many continental A frican despotisms is the high rank held at court by the king's mother and his chief wife.
The social organization of the Malagasy, like their economy, clearly reflects the Indonesian antecedents of their culture. In particular, the generational character of their kinship terminology contrasts sharply with all Arabic, Bantu, and Cushitic systems. The whole social structure, indeed, corresponds almost exactly to what one would expect in a Bornean bilateral society that was subjected to patrilocal influences through Arab and Cushitic contacts and that had to adapt to the increasing importance of cattle. Matrilineal exogamy, which constitutes the sole discrepancy, possibly reflects the mass impact of Bantu slaves, most of whom probably came from societies with matrilineal descent.
Specific features of Arabic derivation in Malagasy culture are numerous, but most of them fall into categories that lie beyond the scope of our survey, such as magic, divination, and value systems. Wet-rice cultivation seems also to have been introduced by the Arabs at a relatively late period, since this has been its origin wherever else it occurs in Africa.
Influences from the African continent obtrude most prominently in the realms of animal husbandry and government. Cattle may derive, of course, from either a Bantu or a Cushitic origin, but the latter seems more probable from the presence of such diagnostic traits as the keeping of livestock in pens to conserve their manure but especially from the fact that the Northeast Coastal Bantu themselves obtained cattle from the Cushites. Dynastic incest reminds us, naturally, of ancient Egypt, but it is probably safer in this case to invoke cultural parallelism. The despotic monarchical government of the Merina, however, with its territorial administrative organization, its specialized ministers, and especially the prominence of a Queen-Mother and a Queen-Consort, unmistakably reflects a widespread pattern in Africa, perhaps derived ultimately from Pharaonic Egypt. Possible immediate sources in East Africa include the kingdoms of southwestern Ethiopia and Uganda, and, as we have seen, the form of the latter is perhaps traceable to that of the former. If the political institutions of the Merina are basically African, then this people must have entered Madagascar from Azania, as we have inferred from other evidence, not by a special late migration from Indonesia.

Selected Bibliography