webAfriqa Library - Anthropology"


webAfriqa Custom Search Engine

Questia: Trusted Online Research

Find Free Money For College!

The New York Times Store

Get home phone line plus three calling features

Get books for $9.95 plus FREE shipping with BOMC2

Start Living a Life You Love with Revival Soy

Raw Organic Coconut - Live Superfoods

Cellular Country

FRANCE: 10% Remise - WS_FTP Professional 12

George Peter Murdock

Professor of Anthropology, Yale University

Africa. Its Peoples and Their Culture History

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


Part One: Orientation

1. — Geography

Among the continents, Africa stands second to Eurasia in size. It has an area of about 11,700,000 square miles, which is nearly 40 per cent greater than that of the Soviet Union and more than three times that of the United States including . Alaska. It lies astride the equator,. with about 80 per cent of its surface within the tropics. With approximately 240 million inhabitants, it ranks third in population, behind Eurasia and North America and appreciably ahead of South America, but it has more distinct peoples and cultures than any other continent.
Geologically, Africa consists largely of a single rigid block of rock of marine origin laid down perhaps 200 million years ago and later uplifted. This still lies on the surface in many places but is elsewhere covered by sedimentary rocks deposited subsequently on the floors of invading gulfs, or shallow inland basins. The processes which raised this massive block thousands of feet above sea level produced remarkably little folding.

Relief of Africa
Map I. Relief of Africa

Consequently, except for the relatively recent Atlas ranges in Morocco; Africa has no mountain chains comparable to the Alps, Andes, Himalayas, and Rockies of other continents. As shown in Map 1, it consists of a series of plateaus, generally higher in the south and east than in the north and west, falling sharply to an extraordinarily narrow coastal plain. On even the highest plateaus, with average altitudes of more than 7,000 feet, the surface is level or undulating rather than broken by ridges.
The chief exception to topographic monotony is presented by the great extinct volcanoes and spectacular rift valleys of East Africa. Among the former, Mounts Elgon, Kenya, and Ruwenzori average about 17,000 feet in elevation, and majestic Kilimanjaro lifts its glacier-tipped cone 19,400 feet above sea leveL The rift valleys, like the volcanoes, resulted from a series of stupendous north-south fractures which occurred during the uplifting of the continent and produced giant trenches often thousands of feet in depth. Their lake-studded courses can readily be followed on Map 2. The principal rift valley starts at the mouth of the Zambesi River and runs northward via Lake Nyasa and the great interior drainage basin of Tanganyika and western Kenya and thence northeastward across Ethiopia to the coast, after which it forms the floors of the Red Sea, the Gulf of Aqaba, and the Dead Sea, and terminates in Syria. A second rift valley, later in origin, is marked by a chain of lakes to the west of the first-Lakes Tanganyika, Kivu, Edward, and Albert-of which the first has a bottom several thousand feet below sea level.

Drainage systems of Africa
Map 2. Drainage Systems of Africa

The geological history of Africa accounts for some of the peculiarities of its drainage system (see Map 2). Its great rivers, like the Congo, Niger, Nile, and Zambesi, are navigable for great distances on the interior plateaus but invariably plunge over impassable rapids or cataracts as they approach the coastal plain. The most spectacular of these, Victoria Falls on the Zambesi, has a drop of 343 feet, or more than twice that of Niagara. Rivers enter the ocean, not through navigable estuaries, but through deltas, often obstructed by shifting sand bars. Because of uplift, other harbors are extraordinarily few, except along . the western Mediterranean coast. The irregular elevation of the continental mass also isolated a series of interior drainage basins which filled with shallow, brackish lakes, most of which have long since found outlets or, in arid regions, have dwindled to salt marshes. Curiously enough, the largest survivor, Lake Chad, contains fresh rather than brackish water, presumably through underground seepage into other basins.

Mean annual rainfall in Africa
Map 3. Mean Annual Rainfall in Africa

Despite the tropical location, altitude moderates the temperature, which on the plateaus of East Africa averages 20 degrees Fahrenheit lower than on the adjacent seacoast. Frosts occur only in the extreme north and south of the continent and on the higher mountains and consequently do not exert a limiting influence on either agriculture or the natural vegetation. The chief controlling factor is drought, for in Africa rainfall is vastly more critical than temperature. Not only do various parts of the continent show wide differences in mean annual precipitation (see Map 3), but these are accentuated by uneven seasonal distribution and by large variations from year to year. The inter-tropical weather front, which regulates the climate over much of the interior, fluctuates widely because of : the lack of mountain barriers, so that regions which receive plenty of rain in one year may suffer severe drought the next. Because of this extreme irregularity, the productivity of agriculture and the consequent density of population, as well as the nature of the vegetation cover, reflect not the annual means of precipitation but rather the minima of current bad years.
Africa has three areas of desert-the Sahara in the north, the margins of the Horn in the east, and the coast and some hinterland in the southwest. Despite their substantial geographic extent, these arid regions belie the popular conception of desert as barren rock and sand dunes bereft of vegetation, for such features characterize less than 8 per cent of the area of the continent. Most of the desert surface supports scattered scrub and even, especially on the margins, grass, which provides excellent pasturage. Moreover, the Sahara in particular is dotted with oases which support intensive irrigated agriculture.
Popular opinion also greatly exaggerates the extent of tropical rainforest, or “jungle,” in Africa. Actually this covers only restricted coastal strips in Madagascar, Mozambique, and the Guinea coast and a larger area in the northern Belgian Congo and adjacent Cameroon and French Equatorial Africa, and even here it is less dense than in comparable regions in Southeast Asia or South America. Small areas of temperate-zone woodland and brush occur at the northern and southern extremes of the continent. More extensive than tropical rainforest is the so-called “dry forest” of Northern Rhodesia and adjacent Angola, Belgian Congo, and Tanganyika. This consists of a relatively open stand of flat-topped; deciduous trees, which shed their leaves during the hot dry season, with a fairly substantial grass cover underneath an'd between them.

Vegetation zones of Africa
Map 4. Vegetation Zones of Africa

By far the preponderant type of vegetation cover in Africa (see Map 4) is savanna, or grassland with scattered trees. Among the latter, baobab and species of Acacia tend to prevail where rainfall is sufficient, with thorny shrubs in more arid sections. At unusually high elevations savanna is replaced by upland grasslands, e.g., the High Veld of Transvaal, or by forests, where mountain slopes are steep. When one adds desert steppe and dry forest to savanna and mountain grasslands, the proportion of the surface of Africa which can support herbivorous animals is really extraordinary. Herein lies, of course, the geographical reason for both the fabled wealth of the continent in wild game and the exceptional role that pastoral activities play in the economies of the indigenous peoples.
A comparable richness in mineral resources suggests the probable economic position of Africa in the world of the future. They lie, however, outside the scope of this book, since only one of them, namely, iron, was important in the native economy prior to the present century. Iron has been smelted and made into tools and weapons by most of the inhabitants of the continent for at least the past thousand years, and the author greatly regrets not having included metallurgical technology among the subjects of his survey.

Selected Bibliography
Bernard, A. Afrique septentrionale et occidentale. 2 vols. Paris, 1937-1939.
Fitzgerald, W. Africa. 7th edit. London, 1950.
Klute, F., L. Wittschell, and A. Kaufmann. Afrika in Natur, Kultur und Wirtschaft. Wildpark-Potsdam, 1930.
Stamp, L. D. Africa. New York, 1953.

2. — Race

Africa was probably the cradle of mankind. For several decades hardly a year has passed without some exciting new evidence of early man or of manlike apes in East or South Africa to strengthen this conclusion.
These findings, however, lie beyond the time horizon of the present volume, which begins with the end of the Paleolithic period. From that time to the present, Africa has been inhabited by representatives of only five races: the Bushmanoid, Caucasoid, Mongoloid, Negroid, and Pygmoid.
Racial factors in themselves, of course, do not assist in explaining the development or distribution of cultures in Africa, since scientists have long since disproved popular assumptions of inherent racial differences in the capacity to create and maintain culture. We also recognize that the anthropometric and somatological criteria by which the five races have been distinguished are themselves becoming increasingly suspect as physical anthropology comes to lay its main emphasis upon genetic factors.
Unfortunately, the distribution of 0, A, B, and AB blood groups is rather similar in all five African races, and other genetically precise data are not yet available in sufficient quantity to render much assistance. Though we consequently neither regard race as relevant to culture nor consider traditional typologies as particularly respectable from the scientific point of view, we nevertheless insist that the older anthropometric and somatological criteria still serve a useful function as an aid in historical reconstruction. They show enough uniformity over limited areas, enough stability over time, and enough persistence in mixtures, to provide the archeologist and the ethnologist with a highly welcome additional tool for tracing important culture-historical movements in the past. We thus offer no apologies for the fivefold division presented herewith.
The Bushmanoid peoples, however insignificant numerically, clearly constitute a distinct race. The Xam, or Southern Bushmen, selected as the type because they are probably the least mixed, reveal the following physical characteristics: short stature, averaging no more than 4 feet 10 inches for adult males; exceedingly wrinkled skin of a light yellowish-brown color; very scanty facial and body hair; short, black hair on the head, clustering in very tight “peppercorn” spirals; medium to narrow head, with a cephalic index of 76; low forehead; flat, triangular face; high cheekbones; extremely broad nose, with a nasal index of 115; brown eyes reduced to narrow slits by a prominent but un-Mongoloid eye fold; thin but slightly protruding lips; pointed chin; slender body and limbs; and, in women, marked steatopygia (prominent buttocks) and the so-called “Hottentot apron” (elongation of the labia minora). Among more northern Bushmen the stature increases to about 5 feet 2 inches, and the prevailing skin color becomes reddish brown. The average Hottentot, with cattle and a milk diet, attains 5 feet 3 inches in stature and has a narrower head (cephalic index 73) and nose (nasal index 100).
The Pygmoid peoples, though perhaps ultimately somewhat more closely akin to the Negroid than to any other stock, display enough specialized features to warrant classification as a distinct race rather than as a sub-race. The Mbuti of the Ituri Forest region, the group least affected by Negro contacts, exhibit the following physical traits: short stature, averaging about 4 feet 9 inches in adult males; light yellowish-brown skin densely covered with downy body hair (lanugo), which is blond to reddish in infants but dark brown in adults; black, kinky head hair; high, bulging forehead; large head of medium breadth, with a cephalic index of 76.5; broad, flat nose, with a nasal index of about 100; protruding brown eyes; oval face; lips of medium thickness, but not everted; slender legs and body, with protruding abdomen; and an average body weight for males of only 88 pounds. Other Pygmy groups are usually darker in color because of Negroid admixture but range in stature from as short as 4 feet 6 inches to as tall as 5 feet 1 inch.
The Negroid race, being familiar to most readers, requires a less detailed description. The Mande peoples of the western Sudan, who may be selected as a fairly representative group, are tall in stature, averaging 5 feet 8 inches for adult males, and dark brown in skin color. Other characteristics are a narrow head (cephalic index 74); black, kinky hair; bulging forehead; broad nose (nasal index 95); prognathous face; thick, everted lips; and scanty facial and body hair. On the Guinea coast the stature is usually somewhat shorter, and the skin color frequently approaches true black. The Bantu tend to be somewhat lighter, shorter, and less prognathous than other Negroes, and in the central Sudan a round, or brachycephalic, head form occasionally appears. The Nilotes of the eastern Sudan are sufficiently divergent to warrant their classification as a distinct sub-race. The Dinka, a typical Nilotic tribe, exhibit the following characteristics: very tall stature, averaging 5 feet 10 inches for adult males; long limbs; extreme slenderness; dark-brown skin; narrow head (cephalic index 73); black, kinky hair; broad nose (nasal index 92); and medium to thick everted lips.

Table I : Descriptive Characteristics of African Races
Race and people Stature Skin color Head form Nose form Hair form Lips Special features
Pygmoid
Mbuti Very short Light yellowish brown Narrow to medium Broad Kinky Medium but not everted Protruding eyes Downy body hair Delicate frame
Bushmanoid              
Xam Short Light yellowish brown Narrow to medium Very broad Pepper corn Thin, slightly everted Wrinkled face High cheekbones Steatopygia  
Negroid
Mande Tall Dark brown Narrow Broad Kinky Thick, everted Prognathism
Dinka Very tall Dark brown Narrow Broad Kinky Medium everted Very slender Long limbs
Caucasoid
Kabyle Medium Light brown Medium Narrow to medium Straight or wavy Medium to thin Some hair and eye blondish
Galla Tall Medium brown Medium Narrow Curly to kinky Medium to thick Prominent nose
Mongoloid
Merina Short Yellowish brown Medium to broad Broad Straight Medium Mongolian spots Epicanthic fold

 

The Caucasoids of Africa belong to the Mediterranean sub-race but tend to be much taller and somewhat more dolichocephalic (narrow-headed) than their kinsmen of southern Europe and the Near East. The Berbers of North Africa and the Cushites of Ethiopia and the Horn diverge in some respects. The former reveal certain affinities with the Nordic sub-race of northwestern Europe, notably a definite, though relatively low, incidence of blandness in hair and eye color. The Cushites have long since incorporated a not insubstantial infusion of Negroid blood. This reveals itself in different ways in different tribes. Thus the lowland Somali are much darker than the peoples of highland Ethiopia but have hair that is wavy or occasionally straight and only rarely kinky, whereas this typically Negroid form prevails in 60 to 70 per cent of the plateau population. The Galla, a typical Cushitic people, exhibit the following traits: tall stature, averaging 5 feet 8 inches in adult males; medium-brown skin; curly to kinky hair on the head, with an appreciable amount also on the face and body; a head of medium breadth (cephalic index 77); narrow nose (nasal index 69); dark-brown eyes; medium to thick but not everted lips; prominent nose; and strong chin. The Kabyle of Algeria, a representative Berber people, may be characterized as follows: medium stature, averaging 5 feet 5 inches in males; light-brown skin; straight or wavy hair, which is usually black or dark brown, but occasionally red; abundant facial and body hair; head of medium breadth (cephalic index 77); moderately narrow nose (nasal index 74); dark- to light-brown eyes, with occasional lighter shades; medium to thin lips; and prominent nose.
Mongoloid peoples survive today only on the island of Madagascar, where they are usually heavily admixed with Negroids. The interior peoples of Borneo, their nearest kinsmen in Indonesia, have approximately the following characteristics: short stature, averaging 5 feet 2 inches for adult males; yellowish-brown skin; straight black hair, which is scanty on the face and body; head of medium breadth (cephalic index 77); broad and rather flat nose (nasal index 80); dark eyes; lips of medium thickness; and a high incidence of Mongolian spots and the epicanthic fold. The Merina of central Madagascar, in whom the Mongoloid element is strongest today, resemble the Borneo Dayak in most respects, although the head is somewhat broader (cephalic index 81).
For purposes of ready comparison, the distinctive descriptive characteristics of the five African races are summarized in Table 1.
At the close of the Paleolithic period the Mongoloids had not yet made their appearance in Africa, and the other four races divided the continent in not notably unequal proportions. Subsequent historical movements, to be described in detail in later chapters, have practically eliminated the Pygmoids, the Bushmanoids, and even the latecoming Mongoloids.

Distribution of races in Africa
Map 5. Distribution of Races in Africa (B-Bushmanoid, C-Caucasoid, M-Mongoloid, N-Negroid,
P-Pygmoid, --> Displaced by)

Although the Caucasoids have made modest advances in certain areas, most of the territorial gains have accrued to the Negroid race. Map 5 shows the racial distribution in Africa at the end of the Paleolithic period and today. Regions whose ethnic occupation has remained unaltered are indicated by the initial of the race. Where one race has replaced or largely absorbed another, ( show the order of succession.
Thus B-->C-->N, for example, indicates that the area in question was originally occupied by Bushmanoids, who have since been replaced by Caucasoids, and these in turn by Negroids.

Selected Bibliography
Clark, W. E. LeG. Fossil Evidence for the Evolution of Man. Chicago, 1956.
Coon, C. S. The Races of Europe. New York, 1939.
Gates, R. R. Human Ancestry from a Genetical Point of View. Cambridge, 1948.
Gusinde, M. Pygmies and Pygmoids. AQ, 3:3-61. 1955.
Haddon, A. C. The Races of Man and Their Distribution. New York, 1925.
Martin, R. Lehrbuch der Anthropologie. 2d edit. 3 vols. Jena, 1928.
Schapera, I. The Khoisan Peoples of South Africa. London, 1930.
Seligman, C. G. Races of Africa. New York, 1930.

3. — Language

In the absence of written records, linguistic relationships provide by far the most dependable evidence of historical connections. If two peoples speak related languages, however much they may differ in race or in culture and however remote their geographical location, either both have descended from a single ancestral society or the ancestors of one have at some time had such intimate contact with a group thus related to the other that they abandoned their own language and adopted that of their neighbors. Even great paucity or a complete lack of other evidence cannot invalidate this conclusion. Thus the fact that the Malagasy speak a Malaya-Polynesian language proves beyond doubt the former existence of some direct contact between the inhabitants of Madagascar and of Indonesia, though this might never have been guessed from other clues.
This volume draws heavily upon the results of linguistic research.
Inferences from this source have frequently given the first intimation of some major historical movement, whether of migration or of cultural diffusion. Moreover, these have never proved false, for a careful search has invariably turned up corroborative data from written records, archeology, botany, physical anthropology, social structural analysis, ethnographic distributions, or more commonly several of these sources.
Although generations of linguists have analyzed the languages of Africa and have added immensely to the sum of our knowledge, various defects of method have delayed until recently the formulation of a comprehensive and dependable classification of all African languages comparable to that achieved in 1891 by J.W. Powell for native North America. Far too much stress, for example, has been laid on single traits like sex gender. Even more serious has been the reliance placed upon inferences from other aspects of culture, especially the presence or absence of a pastoral mode of life. Innumerable linguists have seemed to regard the herding and milking of cattle as a linguistic trait, and an overriding one at that.
Repeatedly, for example, they have classed the Fulani, the Hottentot, and the southeastern Nilotes erroneously as “Hamitic” in language because of their dependence upon cattle, and have placed the Hausa, who are really Hamitic, in some other linguistic group because they are not pastoral.
They have resorted to inadmissible processes of language formation such as the emergence of mixed intermediate languages implied in such terms as “Semi-Bantu” and “Nilo-Hamite.” Finally, they have clung too tenaciously to traditional groupings, treating Bantu as an independent linguistic stock long after evidence was available to show that it is merely an enormously expanded division of one branch of a single subfamily of a very much larger stock.
Fortunately we need no longer depend upon this earlier work, for Greenberg (1949-1954) has now given us a complete and thoroughly satisfactory classification of the languages of Africa, superseding for our purposes everything that preceded it. Our own research has confirmed Greenberg on every point where he differs from his predecessors, and our classification of stocks and families is derived entirely from his. It deviates only on minor and inessential points, which should be noted.
First of all, we have renamed several of the stocks. Khoisan is preferred to “Click” since we accept Greenberg's own objections to classification in terms of a single linguistic trait. We propose Kanuric in place of “Central Saharan” since we prefer terms with ethnic connotations, like Algonkian and Arawakan, to geographical names. Similar considerations pertain in the cases of Greenberg's “Afroasiatic” and “Niger-Congo” stocks. Instead of the former we use Hamitic. The more generally accepted term “Hamito-Semitic” incorrectly implies that Semitic is paired with a coordinate Hamitic division, whereas it actually has four parallel subfamilies—Berber, Chadic, Cushitic, and Egyptian. Moreover, the name Hamitic is now available, having been replaced by Cushitic for the southeastern subfamily of the stock. For the awkward “Niger-Congo" we prefer the archaic name Nigritic, which seems especially fitting since its speakers comprise, and are largely confined to, the more typical representatives of the Negro race.
Two of Greenberg's stocks—Nyangiya and Temainian—are omitted from our list. Both are based on very short word lists, and their independence therefore awaits confirmation 1. On Map 6 they appear in the areas of Eastern Sudanic and Kordofanian, respectively. All other changes are minor and obvious. A complete classification of the linguistic stocks and subfamilies of Africa, as used throughout this volume, is presented in the following list.

  1. Furian [the Fur of Greenberg]
  2. Hamitic [the Afroasiatic of Greenberg, the Hamito-Semitic of others]
    1. Berber
    2. Chadic [the Chad of Greenberg]
    3. Cushitic Branches: Central, Eastern, Northern, Southern, Western
    4. Egyptian [the Ancient Egyptian of Greenberg]
    5. Semitic
  3. Kanuric [the Central Saharan of Greenberg]
  4. Khoisan [the Click of Greenberg]
    1. Bushman Branches: Central, Northern, Southern
    2. Kindiga [the Hatsa of Greenberg]
    3. Sandawe
  5. Koman
  6. Kordofaman
    1. Katla
    2. Koalib
    3. Tagali
    4. Talodi
    5. Tumtum
  7. Maban
  8. Malayo-Polynesian [the Austronesian of some]
  9. Nigritic [the Niger-Congo of Greenberg]
    1. Atlantic [the West Atlantic of Greenberg]
    2. Bantoid [the Central branch of Greenberg, the Semi-Bantu of others] Branches: about six in number, of which one, herewith called MacroBantu, has seven divisions: Bantu, Batu, Galim-Mambila, Jarawa-Mbula, Ndoro, Tiv [the Munshi of Greenberg], and Zuande [the Bitare of Greenberg]
    3. Eastern Nigritic [the Adamawa-Eastern branch of Greenberg]
    4. Ijaw [the Ijo of Greenberg]
    5. Kwa — Branches: Edo, lbo, Kru, Nupe, Twi, Yoruba
    6. Mande ·[the Mandingo of Greenberg]
    7. Voltaic [the Gur of Greenberg] Branches: Gur [the Mossi-Grtinshi of Greenberg], Senufo
  10. Songhaic [the Songhai of Greenberg]
  11. Sudanic [the Macro-Sudanic of Greenberg]
    1. Bertan
    2. Central Sudanic
    3. Eastern Sudanic — Branches: Barea, Beir [the Beir-Didinga of Greenberg], Dagu, lngassana [the Tabi ,bf Greenberg], Merarit, Nilotic [the Southern branch of Greenberg], Nubian
    4. Kunaman
African languages about A.D. 1500
Map 6. African Languages about A.D. 1500
(1-Furian, 2-Hamitic, 3-Khoisan, 4-Kordofanian)

Map 6 shows the location and geographic extent of these eleven stocks about A.D. 1500 and of three lesser groupings-Bantu, Central Sudanic; and Chadic: mapped separately because of their special historical significance. Several factors have prompted the selection of 1500 rather than the present as the basic date for the map. For one thing, this eliminates the necessity for indicating an extraneous linguistic stock, the Indo-European, in such regions as Algeria and South Africa. More important, it makes possible considerable simplification in areas where recent intrusive populations have shattered and dispersed the indigenous inhabitants and thereby produced complex distributions that are difficult to map clearly on a small scale. The principal deviations from a more recent situation are the following:

  1. In South Africa the distribution of Khoisan remains unaffected by European settlement and by the late encroachments of the Tswana.
  2. In the northern part of the eastern Sudan Arabic does not appear, leaving the linguistic situation approximately as it must have existed prior to the invasions of the Baggara, or Cattle Arabs.
  3. In the southern part of the eastern Sudan the situation is reconstructed to show the Central Sudanic peoples before they were shattered by the Azande spearhead of the Eastern Nigritic migration from the west, the Nilotic penetration from the northeast, and the Arab slave raids of the nineteenth century.
  4. On the middle Niger the Tuareg had not yet migrated south of the great bend of that river.
  5. In Northern Nigeria the boundary between the Chadic and Nigritic languages is readjusted to discount the displacements that resulted from Fulani penetration.
  6. In East Africa the Bantu are assigned areas which they apparently occupied prior to the last southward advance of the Masai and the expansion of the Somali and Galla toward the southwest.

The map likewise omits a few enclaves of intrusive tribes. It has the definite advantage, however, of delimiting quite clearly the major linguistic divisions of Africa. For specific tribal locations and more· recent distributions the reader may consult the tribal map in the pocket at the back of the book.

Selected Bibliography
Greenberg, J. H. Studies in African Linguistic Classification. SW]A, 5:79-100, 190- 198, 309-317; 6:47-63, 143-160, 223-237, 388-398; 10:405-415. 1949-1954. [Also published in book form, New Haven, 1955].

4. — Economy

A survey of the entire range of indigenous economic life in Africa lies outside the scope of the present volume, which attempts to cover intensively only the subsistence economy, i.e., the major types of food acquisition. These are hunting and gathering, fishing, animal husbandry, agriculture, and their various combinations. Map 7 shows the distribution of these types of subsistence activity over the continent.
Hunting and gathering, by which man gained his livelihood throughout the Paleolithic period, survive today as the mainstay of existence only among a few remnant peoples, who will be considered in Chapters 8 to 10, notably the Pygmies of the Congo Basin, the Bushmen of arid southern Africa, the Kindiga of northern Tanganyika, and the Dorobo of western Kenya. As a subsidiary activity, of course, hunting still retains some importance over most of Africa, as is understandable in view of the plentitude of game animals for which the continent is famed. Gathering, too, provides a welcome addition to the diet in some regions, e.g., in the western Sudan, where locust beans, kola nuts, and shea fruits are extensively used.
Most African peoples practice fishing as a subsidiary economic pursuit wherever geographical conditions permit. One interesting exception, however, deserves special note. The Cushites of northeastern Africa impose a taboo on the eating of fish which compares in rigor to that of the Semites against pork. Fishing attains the status of the dominant subsistence activity in a few scattered regions. In no instance, however, does this represent a survival from a preagricultural economy but occurs only as a reflection of environmental influences. Coastal Mauritania, inhabited by the lmragen, and the lagoons of the Ivory Coast exemplify one type of favorable situation—a maritime location with an immediate hinterland unsuited to agriculture. A second conducive situation is presented by the Congo and Niger Rivers and some of their principal tributaries. Here numerous tribes combine fishing with a profitable boat trade and exchange dried or salted fish for the agricultural produce of their neighbors. An occasional insular people, notably the Buduma of the small islets in Lake Chad, has arrived at a comparable economic adjustment. Animal husbandry as a subsistence activity-eliminating those domesticated species which contribute but slightly to the food supply-revolves in Africa around pigs, goats, sheep, cattle, and camels. All were introduced to the continent from Asia by way of Egypt, and all except the camel, which first appeared in Greco-Roman times, arrived along with agriculture in the early Neolithic period. The pig, once extremely important throughout North Africa, has nearly become extinct through the notion of its uncleanness, which first arose in Pharaonic Egypt and was reinforced by the much later invasion of Bedouin Arabs and Islamic culture.
Except for places where it has recently been reintroduced by Europeans, the pig has survived into modern times in only a few small peripheral areas-among the Nuba and Prenilotes of the middle Nile, the Guanche of the Canary Islands, and the Senegambian tribes of Portuguese. Guinea. The goat, except for the dog the most widespread of African domesticated animals, has penetrated even the tropical forest in many places. African sheep have fat tails and hair rather than wool, the familiar woolly sheep representing a relatively late introduction. The camel is confined to arid regions, though by no means exclusively to Arabs.
African cattle represent varieties of and crosses between two species: Mediterranean cattle (Bos taurus) and the Indian humped zebu (B. indicus)—each of which entered the continent twice in different forms. Longhorned Mediterranean cattle, presumably descended from the wild B. primigenius, appeared in Egypt at the beginning of the Neolithic period; short-horned Mediterranean cattle, probably derived from the wild B. brachyceros, arrived during the third millennium B.C. Both spread early throughout North Africa and the Sudan, where derivative and hybrid breeds survive, except where Arabs or Fulani have introduced the zebu within the past ten centuries. The zebu reached Egypt about 1600 B.c. and East Africa in another form during the first millennium after Christ.
Most, but by no means all, East and South African cattle have a strong zebu ingredient. Cattle have penetrated the tropical-forest zone to only a very limited extent, and they are excluded from many sections by the presence of the tsetse fly.

Types of Subsistence Economy
Map. 7. Distribution of Types of Subsistence Economy

Cows, goats, and, where they occur, camels are milked throughout North, East, and South Africa, and over most of this area butter is made at least for cosmetic purposes if not for food. The inhabitants of the western Sudan, of the Guinea coast, and of northern Angola, however, do not milk their animals, except on the very fringe of the Sahara and in a few other places under recent Fulani influence. It would thus appear that the Negroes originally borrowed animal husbandry from the North African Caucasoids without its associated dairy complex and that the latter diffused southward only at a later date and over a smaller area. Map 8 shows the distribution of cattle and of milking. Animal husbandry plays varying roles in the economies of different parts of Africa. It may make a negligible contribution to subsistence, as in the tropical-forest zone. It may provide a significant, though subsidiary, supplement to the products of tillage, as in many agricultural areas. It may combine with cultivation in a balanced economy with approximately equal dependence upon both activities, as in many societies that are commonly regarded as “pastoral,” e.g., among the majority of the Berbers, Galla, Nilotes, and Cattle Bantu. Or it may be in large measure detached from agriculture, becoming the basis of an independent nomadic mode of pastoral life in which subsistence depends primarily upon milk and other animal products.
Independent pastoralism of this last type has existed from the dawn of recorded history among the Beja peoples between the Nile Valley and the Red Sea, deriving presumably from the neighboring Bedouin Arabs to the east. Elsewhere in Africa, however, it did not develop until a surprisingly late date, probably nowhere much earlier than A.D. 1000. It has nevertheless spread very widely during the past thousand years and is the prevailing pattern of life today among the Afar and Somali of the Eastern Horn, most North African Arabs, the Tuareg of the Sahara, the Fulani of the western and central Sudan, and the Herero and Hottentot of southwestern Africa. Pastoralism may center on sheep and goats, on camels, or on cattle. Primary dependence upon small livestock characterizes the Arab nomads in the hinterland of the Mediterranean coast in North Africa. The camel assumes first place among the Beja, the Afar and Somali, and the Arab and Berber tribes of the Sahara. Cattle play the dominant role among the Galla, the Baggara Arabs of the eastern Sudan, the Fulani, the Hottentot, and all Negro pastoralists.
Agriculture is practiced throughout the continent except among the surviving hunters and a few of the most exclusive fishing and independent pastoral peoples. Africans grow approximately nine-tenths of all the cultivated plant varieties known to man and have assembled them from every originating center in the world, though borrowings from China and highland America have been relatively few. Since the various plants first appeared in Africa at widely different places and times and have commonly been controlled by climatic and other geographical factors in their spread, the analysis of their distribution yields clues of the utmost importance in unraveling the paths of past migrations of peoples and diffusions of culture. The historical reconstructions in this volume consequently depend heavily upon this type of evidence.

Distribution of Cattle and of Milking
Map 8. Distribution of Cattle and of Milking

Botanists can frequently identify the approximate place of origin of an indigenous cultigen by locating the native habitat of the wild form from which it was ennobled. Cultivated cotton (Gossypium herbaceum), for instance, can be traced to the Sahara-Sudan borderland because of the occurrence there of the wild G. anomalum, its presumptive ancestor.
Botanists can also demonstrate the alien provenience of particular plants on the basis of genetic analysis. All African bananas (Musa paradisiaca and M. sapientum) are hybrids of two wild forms, M. acuminata and M. balbisiana, natives respectively of Malaya and India, and can have come originally only from Southeast Asia since neither wild form occurs anywhere in Africa. Cultivated plants have an immense advantage over domesticated animals as aids in historical reconstruction. One imported bull can alter appreciably the breed of a herd, but indigenous and introduced plants either retain their separate identities or, if they cross, readily yield evidence of their origins to genetic research.
Introduced plants rarely replace completely the species previously cultivated. The latter nearly always survive in some subsidiary status, as revealed in the full lists of cultivated plants compiled by conscientious ethnographers. In Africa, ethnographies written in the French and German languages are much the most useful from this point of view. Over wide regions of Africa today plants introduced in recent centuries from the New World, notably maize and manioc, have attained a dominant role as the staple crops, but analysis of ethnographic lists of subsidiary cultigens invariably reveals those which provided the basis of subsistence prior to A.D. 1500.
Each of the major complexes of food plants which have played a role in African culture history will receive full consideration in the regional chapter dealing with its introduction. At this point only a brief orientation is needed. Two of the complexes originated on the continent itself, one in West Africa and one in Ethiopia, considered, respectively, in Chapters 11 and 22. Two derive from Southwest Asia by way of Egypt and are treated in Chapters 14 and 18. They are differentiated on the basis of dates of introduction, i.e., prior or subsequent to the close of the Pharaonic, or Dynastic, period marked by the conquest of Egypt by Alexander the Great in 3 3 2 B.c. Two complexes, discussed in Chapter 26, derive from Southeast Asia by way of East Africa and are differentiated on the basis of whether their component plants were first domesticated in India or farther to the east. The last complex to reach Africa is the American (see Chapter 29), whose elements were introduced to the coasts of West and East Africa by Europeans during the course of the slave trade. A partial list of plants from these various sources, classified according to type of cultigen, is presented in Table 2.
It may also prove helpful to single out for brief mention all the plants which have attained the status of an outstanding staple in at least several of the societies of the continent. These are, roughly in order of their importance:

Table 2: Classification of African Cultivated Plants by Type and Origin

Type Place of origin
West Africa Ethiopia Southwest Asia Southeast Asia America
Cereal grains Fonio, Pearl millet, Sorghum Eleusin,e Teff Barley Wheat Rice Maize
Legumes   Cow pea Broad bean Chick pea Lentil Pea Gram bean Hyacinth bean Pigeon pea Sword bean Haricot bean Lima bean
Tubers and root crops Coleus, Earth pea, Geocarpa bean ,Guinea yam Ensete Beet Chufa Onion Radish Taro Yam Malanga Manioc Peanut Sweet potato
Leaf and stalk vegetables Okra Cress Cabbage Lettuce Jew's mallow    
Vine and ground fruits Fluted pumpkin Gourd Watermelon   Grape Melon Cucumber Eggplant Pineapple Pumpkin Squash Tomato
Tree fruits Akee Tamarind Date palm Fig Pomegranate Banana Coconut palm Mango Avocado Papaya
Condiments and indulgents Kola Roselle Coffee Fenugreek Kat Coriander Garlic Opium poppy Ginger Hemp (hashish) Sugarcane Cacao Red pepper Tobacco
Textile plants Ambary Cotton   Flax    
Oil plants Oil palm Sesame Castor Remtil Olive, Rape      

Among economic plants and cash crops grown for export the most important are cacao, coffee, cotton, flax, kola, the oil palm, peanuts, and sesame.

Selected Bibliography
[For sources on African agriculture see the bibliographies appended to Chapters 11, 14, 18, 22, 26, and 29.]
Bisschop, J. H. R. Parent Stock and Derived Types of African Cattle. SAIS, 33:852-870. 1936.
Dyson, R. H. Archeology and the Domestication of Animals in the Old World. AA, 55:661-673. 1953.
Phillips, R. W. Cattle. Scientific American, 198(6):51-59. 1958.
Simmonds, N. W., and K. Shepherd. The Taxonomy and Origins of the Cultivated Bananas. Journal of the Linnean Society, Botany, 55:302-312. 1955.
Simoons, F. The Non-milking Area of Africa. Anthropos, 49:58-66. 1954.

5 — Society

Anthropologists have devoted special attention to the field of social organization for nearly a century and have made more striking scientific progress here than in any other branch of their subject. As a consequence, they have developed a somewhat complex technical vocabulary which it is necessary to employ in any discussion of African society. Since the concepts employed by specialists, though seldom difficult to grasp, are not always familiar to the general reader, and the terms are not always fully standardized, each subject considered below will be introduced by a set of definitions.
In some parts of the world, though rarely in Africa, marriages may be concluded relatively informally, involving only the consent of the parties and their kinsmen, a minimum of ceremonial, and no property transfer or other material consideration, except perhaps for gift giving or an exchange of presents. Most African societies, however, require some material consideration to legitimize and stabilize a marriage, to recompense the bride's parents for the loss of their daughter, and to serve as a guaranty that the husband will fulfill his obligations. Such a consideration may be tendered in goods, in services, or in kind. If paid in goods, e.g., in livestock or in currency, it is technically called a bride-price. If rendered in services, e.g., in agricultural labor for the wife's parents either before or after marriage, it is called bride-service. A payment in kind involves giving a woman of one's own group, e.g., a sister or a daughter, to a kinsman of the bride in exchange for her-a transaction known as an exchange marriage.
Payment of a bride-price strongly predominates in Africa as a whole and in the great majority of its individual culture provinces. Exchange marriage prevails in a number of societies on the Nigerian plateau, in a smaller cluster in western Ethiopia, and in sporadic instances elsewhere. Bride-service has a higher incidence. It often occurs as a supplement to, or a substitute for, a bride-price, and stands alone, as one might expect, chiefly among peoples who are poor in livestock or other valuable possessions, e.g., the Bushmen and the eastern Central Bantu.
Marriage can assume any one of three basic forms. Monogamy unites one man with one woman only, but a society cannot be characterized as monogamous unless plural marriages are either forbidden or disapproved, since the normal sex ratio assures a preponderance of monogamous unions in any society except under highly exceptional, and usually temporary, circumstances. Polyandry unites one woman with two or more husbands; polygyny, one man with two or more wives. The more familiar term "polygamy," which embraces both forms of plural marriage, will not be used in this book. Polygyny may be either sororal or nonsororal depending upon the presence or absence of a preference for marrying sisters rather than co-wives who are unrelated to each other. Polygyny often results from preferential rules governing the remarriage of widows, e.g., the so-called "inheritance" of widows by a brother, nephew, or son of the deceased husband. Actually these rules represent, not a form of property transmission, but a social security device assuring the support and maintenance of women in their old age. The rule prescribing secondary marriage with a brother of the deceased husband is known as the levirate, or if only the husband's younger brother is eligible, as the junior levirate. A parallel rule, called the sororate, gives a widower a preferential right to marry a sister of his deceased wife.
Polyandry is virtually nonexistent in Africa, and monogamy, except for intrusive Europeans and missionized natives, is confined almost exclusively to the Berbers of North Africa, the Monophysitic Christians of Ethiopia, and the remnant hunting peoples. Polygyny preponderates to an overwhelming extent, prevailing in 88 per cent of a representative sample of 154 societies drawn from the continent at large (Murdock, 1957). Africans have discovered means of making the institution work to the satisfaction of both sexes. No woman lacks a male provider. No polygynous wife has trouble finding a helper or baby sitter in time of need. Since the first wife normally enjoys for her lifetime a position of superior authority and prestige, every woman knows in advance of her marriage what her future status will be and has no fear of being superseded. Since men almost universally establish each of their vvives in a separate dwelling and endow them individually with land and livestock, sources of friction are reduced to a minimum. Custom normally requires the husband to treat each wife with equal consideration, to eat and sleep with each in regular rotation, so that no married woman suffers public humiliation through any overt manifestation of favoritism. In consequence ofthese cultural adjustments, missionaries seeking to institute monogamy in African societies frequently encounter their strongest opposition from the women.
A household may consist either of an independent polygynous or monogamous family or of an extended family. In the latter type of organization two or more families of different generations are united by consanguineal kinship ties and common residence under a single head. The precise composition differs with the prevailing rule of residence (see below), but one common type, the patrilocal extended family, comprises a patriarchal head, his wife or wives, his unmarried children, his married sons with their wives and children, and not infrequently also his younger brothers or other collateral relatives with their wives and offspring. The distribution of the various types of family and household organization in Africa defies easy summarization and must therefore await consideration in the regional chapters.
Besides families and households, all societies possess kin groups of some kind whose members are bound together by ties of consanguineal kinship, or blood relationship, but not necessarily by common residence. Affiliation with such groups is accomplished by a rule of descent, of which there are two main types. Bilateral descent, which affiliates individuals with their close kinsmen on the basis of relationships traced equally through males and through females, need not concern us here since it prevails in only an insignificant handful of African societies. In contrast to most of the peoples of Europe and the New World, and many of those in Asia and Oceania, those of Africa adhere almost unanimously to one of the two alternative modes of unilinear descent, i.e., the patrilineal rule, which affiliates individuals with kin groups exclusively on the basis of relationships through males, or the matrilineal rule, in which it is only relationships through females that count. A not inconsiderable number even practice double descent, which affiliates individuals with some kin groups through the patrilineal rule and with others through the matrilineal rule, thus differing fundamentally from bilateral descent, which affiliates people with the same group regardless of the sex of the parent or other kinsman through whom they are related.
Bilateral descent doubtless once prevailed among the Bushmanoid hunters of South and East Africa since it still survives today among the less acculturated of their survivors. The Caucasoid peoples of northern Africa seem to have been characterized by patrilineal descent from time immemorial, since both matrilineal and bilateral descent occur among them very rarely and only in situations suggesting derivative developments. On the other hand, the Negroes, though preponderantly patrilineal today, may very probably once have been largely matrilineal. In nearly every subfamily of the major linguistic stocks in Negro Africa, e.g., the Nigritic, Sudanic, and Kordofanian, there are at least some tribes who adhere to the matrilineal rule even today, and techniques developed elsewhere by the author (Murdock, 1949) for ascertaining earlier forms of social organization demonstrate rather conclusively that in at least these linguistic stocks matrilineal descent prevailed generally in the not very distant past. Map 9 shows the distribution of the various patterns of descent in Africa today.
Both rules of unilinear descent produce kin groups of parallel type. These are called lineages, especially when relatively small and characterized by unmistakable corporate functions; sibs, especially when the members are numerous and geographically dispersed so that they cannot maintain a genuine corporate identity; phratries, when they are still larger and comprise a number of sibs; and moieties, when an entire society comprises only two very large groups, so that every individual must necessarily belong to either one or the other. To indicate succinctly whether the organizing rule of descent is patrilineal or matrilineal, these kin groups are commonly called patrilineages, matrisibs, etc. Some societies exhibit a tendency to constitute as separate groups the descendants of ancestors at each ascending generation, producing a hierarchy of lineages within lineages in what is called a segmentary lineage system. The Arab peoples provide particularly striking examples. In such systems it has been found helpful to employ a standard terminology for segments of increasing size and generation depth: minimal lineage, minor lineage, major lineage, maximal lineage, subsib, sib, subphratry, and phratry, or as many such terms as the number of recognized levels in the particular society requires.

Distribution of Rules of Descent
Map 9. Distribution of Rules of Descent

Kin groups may or may not regulate marriage, and the same is true of local groups, social classes, age-grades, etc. Any social group which does not regulate marriage, i.e., which prescribes neither that its members marry amongst themselves nor that they marry outsiders, is called agamous. A group which does regulate marriage is called endogamous if in-marriage is favored, exogamous if out-marriage is strongly preferential. In general, kin groups tend to be exogamous, especially at the lineage level. Phratries, however, are often agamous or even endogamous, and sibs occasionally so. Endogamy at the lowest lineage level is confined almost exclusively to the Arabs of North Africa and to indigenous peoples who have been subject to strong Islamic indoctrination.
Kinship systems, i.e., the classification of relatives into categories with reference to the kinship terms applied to them, reveal unmistakable correlations with particular configurations of kin groups. The complexities of kinship lie for the most part beyond the scope of the present volume, which will report on only one phase of the subject, namely, typology in the terminological classification of first cousins. Before presenting this, it will help to make one important distinction. A cross-cousin is the child of a father's sister or of a mother's brother; a parallel cousin is the child of a father's brother or of a mother's sister. In other words, cousins of the former type are the children of a brother and sister, whereas those of the latter type are the offspring of two brothers or of two sisters. The distinction has implications for subjects other than kinship. Cross-cousins, for example, are commonly permitted or even expected to marry, whereas unions between parallel cousins are usually forbidden as incestuous, except in the Islamic world.
First cousins can be designated in the six basically different ways discussed below.

Eskimo pattern
Cross- and parallel cousins are called by the same terms but are terminologically distinguished from siblings. The English term "cousin" illustrates this pattern, which is most typically associated with bilateral descent, independent nuclear families, and kin groups of the type known as personal kindreds.
Hawaiian pattern
Cross- and parallel cousins are both called by the same terms as brothers and sisters. This pattern occurs most commonly in bilateral societies with extended families or corporate kin groups but can also result from an incompletely assimilated shift from one rule of descent to another.
Descriptive pattern
Cousins of every category are terminologically distinguished not only from siblings but also from each other, and are called by compound terms which identify precisely their relationship to Ego, e.g., "father's brother's son" or "maternal uncle's daughter." In Africa this pattern occurs commonly in the Islamic area and on the Guinea coast and occasionally elsewhere, and seems to be correlated with patrilineal descent, a segmentary lineage organization, and the absence or weak development of exogamy.
Iroquois pattern
Cross-cousins are terminologically equated with each other and differentiated from both siblings and parallel cousins. This pattern normally coexists with a matrilineal or patrilineal rule of descent, under which cross-cousins cannot be members of Ego's kin group whereas parallel cousins either necessarily are or may be.
Omaha pattern
Cross-cousins are terminologically differentiated alike from siblings, parallel cousins, and each other, the children of a mother's brother being equated with kinsmen of a higher generation, e.g., being called "uncle" or "mother," whereas a father's sister's children are equated with kinsmen of a lower generation. This pattern occurs almost exclusively in patrilineal societies with strongly functional lineages and reflects a feeling that membership in one's mother's patrilineage, to which of course one's mother, maternal uncle, and the latter's children all belong, is more important than a difference in generation.
Crow pattern
This type is similar to the Omaha pattern. It reflects the lineage principle in the same way except that it occurs in matrilineal societies. Here it is the father's sisters' children who are classed with a higher generation, e.g., as "father" and "aunt," and the mother's brothers' children with a lower generation. When found in a patrilineal society, the Crow pattern provides an infallible indication of former matrilineal descent.

Like kin groups, local groups-or cmnmunities-represent a universal form of social organization. Under nomadic or seminomadic conditions the prevailing type of local group, comprising a number of families which wander and camp together, is called a band. With sedentary life the families which compose a local community may live in a village, or aggregate of clustered households; in a neighborhood, or aggregate of dispersed households; or in a settlement pattern of some intermediate or more complex form. Many societies recognize local groups intermediate in size between the household and the community. These are most widely known as wards, but in the African literature are more commonly called "quarters." Wards may adjoin one another closely or be separated by an intervening space, which, if great enough, may even convert them into dispersed hamlets.
The inhabitants of a local community, or of a ward within it, may be associated on the basis of common religious beliefs or ethnic ties or social status or economic specialization, or as a by-product of individual enterprise and self-interest; in such circumstances their integration rests on propinquity and interdependence alone and any kinship bonds among them are purely incidental. Members of bilateral societies of European origin tend to assume that this kind of community structure is universal. In unilinear societies, however, and perhaps especially in patrilineal cultures, this assumption commonly proves false, and the rule of descent often provides the primary bond of association in local groups as well as in kin groups, at least in the ward and the village, though admittedly only rarely in large urban centers. In Africa, particularly, no adequate conception of community life can be achieved without due recognition of the integrating role of unilinear kinship ties.
Very widely in Africa the village, or at least the ward, is itself a genuine kin group, united by unilinear descent. It cannot, of course, be a sib or a lineage according to the definitions given above, since such groups are normally exogamous. If husband and wife always and necessarily belong to different lineages or sibs, it is obvious that no local group, whether a household, a ward, or a village, can be composed entirely of members of a single unilinear consanguineal kin group. How, then, can we call any local group a kin group? The answer lies in the fact that only the core of a local group is affiliated through unilinear descent, their spouses being attached to the group through affinalc rather than consanguineal ties. A local kin group is thus a compromise kin group, or, technically, a clan (Murdock, 1949).
A clan invariably owes its origin to. a compromise between a rule of descent and a consistent rule of residence. Whenever a man and a woman marry, since unions between members of the same family are universally prohibited as incestuous, one or the other or both must change residence if they are to establish a common household. The alternative possibilities are few, and each has a scientific name. In the first possible solution to this problem, known as neolocal residence, both spouses leave their parental homes and establish a new household whose location is not determined primarily by the kinship ties of either. Though normal in many European societies, neolocal residence is all but unknown in Africa. In an alternative solution, known as duolocal residence, neither spouse leaves his natal home; instead of founding a common household they maintain separate establishments, each with his own kinsmen. This occurs nowhere in Africa as a permanent arrangement, although a few Twi tribes in Ghana practice duolocal residence for the first few years of married life.
With these insignificant exceptions, all African peoples adhere to a rule of unilocal residence, in which one spouse continues to reside with or near his or her kinsmen and is joined there by the other. Unilocal residence may assume any of three variant forms. In the first of these, called matrilocal residence, it is the wife who remains at home, the husband who leaves his relatives to reside with her. This rule, though common in matrilineal societies elsewhere in the world, occurs with extraordinary infrequency in Africa, being confined to a minority of the Central Bantu and not more than two or three scattered tribes elsewhere. In the second variant, called patrilocal residence, the wife leaves home and joins her husband, either in the household of his parents or in a new one in the vicinity. This rule has by far the widest distribution of all on the African continent. Ranking second in incidence, far behind patrilocal residence but several times as common as all other rules put together, is a third unilocal variant, known as avunculocal residence. Under this rule it is again the woman who leaves home, but the couple set up housekeeping with or near, not the parents of the husband, but his maternal uncle. The man has also, of course, left his parental home, but in most cases he has done so long before marriage, in adolescence or even in boyhood. The term "virilocal," which has recently been gaining currency, is to be abjured as ambiguous since it needlessly confuses patrilocal with avunculocal and even with neolocal residence.
Clans can develop only in the presence of consistent rules of descent and residence. The conjunction of patrilineal descent and patrilocal residence produces a patrician, a compromise kin group with the male members of a patrilineage as its core, to which are added their wives from other lineages who have joined them in patrilocal residence, and from which are subtracted their adult sisters who have left to reside in the clans of their husbands. The combination of matrilineal descent and matrilocal residence similarly gives rise to matriclans around a core of matrilineally related females plus their in-marrying husbands and minus their out-marrying brothers. When combined with avunculocal residence, matrilineal descent yields avuncuclans, in which the core consists of the males rather than the females of a matrilineage, and the fringe, as in the case of a patrician, comprises the in-marrying wives of the adult men. Clans cannot arise or exist under other combinations, e.g., of matrilineal descent and patrilocal residence, and any clan organization dissolves as soon as social change produces any serious inconsistency between the rules of residence and descent.
Like other kin groups, clans vary in size. One which is coextensive with an entire village, neighborhood, or band is called a clan-community. One which is confined to a ward or comparable segment of a community is called a clan-barrio. On a still more reduced level, extended families can be regarded as minimal clans since they possess an identical structural composition. Whatever their scale, clans and lineages in the same society always share a common core, consisting of the members of one sex in the prevailing line of descent, and always differ in the relationship of the other sex to them, i.e., as spouses in the one case and siblings in the other. Where both types of group coexist in the same society, they typically reveal very different functions. In general it is the lineage which regulates marriage and inheritance, acts as a unit in life-crisis situations, and is associated with totemism and ancestor worship, whereas the clan functions primarily in the economic, recreational, political, and military spheres of life. These differences are unfortunately often obscured in the descriptive literature through the indiscriminate use of the term "clan" for both the sib or maximal lineage and the compromise kin group, to which the name is here restricted.

Selected Bibliography
Baumann, H. Vaterrecht und Mutterrecht in Afrika. ZE, 58:62-161. 1926.
Murdock, G. P. Social Structure. New York, 1949.
Murdock, G. P. World Ethnographic Sample. AA, 59:664-689. 1957.

6. — Government

In the survey of African ethnography conducted for the present volume, political systems received appreciably less attention than social organization. They were examined less from the point of view of function, or even of structure, than from that of form, which it was hoped might shed light on historical movements and the diffusion of culture in the past. This hope has been realized only in small part. A comprehensive survey of African political organization is still urgently needed, although Fortes and Evans-Pritchard (1940) have made a promising beginning. The present writer can do little more than offer a tentative classification and characterization of the major types of indigenous structures and propose a few suggestions for future research.
The first and simplest, as well as the most widespread, type of political system is one which may be called a primitive democracy. Leadership and a measure of prestige, but not authority, are vested typically in a headman and a council of elders or family heads with perhaps a few other semispecialized functionaries to direct hunting or conduct particular rituals. The headman, though often hereditary, is merely primus inter pares. Neither he nor any other leader has the power or the right to compel compliance. He can only advise or persuade. Decisions are reached through discussion and informal consensus, and sanctions are applied exclusively through the operation of informal mechanisms of social control. With rare exceptions political integration does not transcend the bounds of an autonomous local community, and nothing remotely approximating the structure of a state is detectable. Political relations tend to be indistinguishable from kinship relations. Under unilinear descent the structure of the clan is the structure of government. If a community comprises several clan-barrios, a very modest hierarchical order may emerge, with the senior or founding clan providing the headman and the others relating themselves to it in some real or fictive junior kinship capacity, e.g., as younger brothers, sisters' sons, or affines.
A second distinguishable type, particularly characteristic of the Bedouin Arabs and the Tuareg, may be named the gentile aristocracy . The political structure still rests primarily on a kinship foundation, specifically a segmentary lineage system. Coercive power, however, makes its appearance, based on superiority of weapons, mobility through the camel or the horse, lineage esprit de corps, and the fortunes of war. Through such advantages one powerful kin group or a confederation of such conquers weaker groups or defenseless alien peoples and reduces them to tributary status, serfdom, or slavery, exploiting them remorselessly. Extremely sharp class distinctions develop as a consequence. Stratification is not ordinarily achieved, however, through the subordination of individuals to individuals but of groups to groups, with the dominant kin group forming a ruling aristocracy at the apex of the hierarchy. Political hegemony tends, nevertheless, to be transitory since no adequate devices for achieving administrative stability have been achieved. Except for Islamic religious and judicial functionaries, leadership rests exclusively in the heads (sheikhs) of the hierarchical lineage segments, who ordinarily lead the same mode of life as their followers and must continually validate their authority through valor, generosity, sagacity, or piety. Since rivalries are rampant, strong fissive tendencies prevail. Power alignments are readily shattered, with a constant reshuffling of lineage segments. Even when dominant gentile aristocracies acknowledge the theoretical sovereignty of neighboring states or of colonial governments, they commonly remain essentially autonomous and undergo little significant structural change.
A highly distinctive type of political system, confined today to the less strongly Arabized parts of North Africa but formerly much more widespread, is the Berber republic. Its characteristic structure, unveiled by the researches of Montagne (1930), will receive detailed consideration in Chapter 15, and the briefest of summaries must suffice at this point. Except for emergency war leaders there are no chiefs or other authoritative functionaries. Government at the local level is conducted exclusively by a democratic assembly of all adult males, headed by an elected president who serves for a single year and functions only as the chairman or presiding officer at assembly meetings. At the district level authority is vested in a council or senate, with a similar elective presiding officer, which is composed of a representative of each lineage of the district. Since these tend to be the older and wealthier men, the body has an oligarchical rather than a democratic character. It is, however, in no sense anstocratic, for the independent Berbers are notably egalitarian and have developed no significant class or caste distinctions. Above the district level political integration is achieved solely through traditional alliances between groups of districts. Although its lineage substructure is still clearly apparent, this system has evolved some specifically political features of rather exceptional interest.
A fourth and genuinely unique political system, the Gada republic, has a fairly extensive distribution in East Africa. Especially characteristic of the Galla and neighboring Eastern Cushitic tribes of southern Ethiopia, it has spread to a series of adjacent Bantu and Nilotic peoples, though often in an attenuated form. It rests, not on kinship, but on a structure of formal age-grades. An age-grade may be defined as an organized social group in a hierarchical system of such groups stratified according to either the actual or the socially defined ages of their members. An age-set comprises the individuals who occupy a particular grade at a particular time. They become constituted as a group through initiation into the lowest grade of the system. After a specific period of time-eight years among the Galla—they are promoted collectively to the next higher grade, being succeeded by another set and themselves replacing the set ahead of them, which advances into the next grade. After a set has filled successively all the grades of the system-five in number among the Galla—its surviving members pass out of the system entirely. Specific communal and political functions are associated with each grade. The set or sets occupying a junior grade or grades serve as warriors. The senior set serve as respected elders and advisers but exercise no political authority, which is invariably reserved for the members of the set occupying the grade behind them, i.e., the second from the top in the hierarchy. The Gada republic is a genuine state, integrating not a single community but entire tribes. Even among peripheral peoples like the Masai, among whom the formal political structure has become greatly simplified, the entire society constitutes a peace-group unified by a tribe-wide agegrade organization. Where the system is highly developed, political authority is vested in the holders of a diverse series of offices, including a tribal chief, a speaker of the tribal assembly, headmen of local communities, judges, legal advisers, and religious functionaries, all chosen from the age-set occupying the semifinal age-grade, all holding office only for the duration of their occupancy of that grade, and all retiring together at the termination of this period to become honored but politically powerless elders. Officeholders are elected on the basis of merit and past accomplishment, but eligible candidates for the highest positions are restricted to the sons of previous incumbents. Like the Berber system, from which it differs in so many other respects, the Gada political structure does not rest on kinship, aristocratic status, or despotic power. Since, moreover, neither system is truly democratic, though in both instances authority derives more from popular choice than from hereditary right, it seems appropriate to characterize both of them as republican.
A fifth type of political system may be called the Oriental despotism, after Wittfogel (1957), who has delineated its characteristic features and associated institutions with acumen and painstaking thoroughness. Pharaonic Egypt, with its agrarian economy based on irrigation, its absolutistic monarchy, its hierarchical administrative bureaucracy, its massive monuments and public works created by corvee labor, the conspicuous consumption of its rulers, and the complete domination of the state over property, religion, and other potential sources of autonomous power, represents the veritable prototype of the Oriental despotism. Under the same category clearly fall the political systems established in maritime North Africa by Rome, Byzantium, the Ottoman Turks, and the Arabs, as well as the derivative states known historically from Nubia and highland Ethiopia.
What concerns us more is the African despotism, a parallel form of political structure found widely in Negro Africa and extending to Madagascar and to the Cushitic-speaking Sidamo peoples of southwestern Ethiopia. Its striking resemblances to the Oriental despotisms analyzed by Wittfogel raise two extremely important questions. The first is historical. Do the despotic states of Africa represent an independent but parallel development, like those of Mexico and Peru in the New World, or did the peoples of the Sudan borrow the fundamental pattern of organization at some early date from Pharaonic Egypt, adapt it to their own needs, and transmit it to the rest of Negro Africa? The writer feels unable to give a categoric answer to this question and recommends it as a significant research problem for some historically oriented anthropologist. He can, however, record his own very tentative conclusions. These are, first, that African despotisms all derive from a single common- source; second, that the balance of evidence favors an origin in the western Sudan rather than on the lower Nile; third, that significant increments to the basic pattern were in fact borrowed from Egypt at some subsequent period or periods.
The second important question is sociological or comparative. Is the African despotism merely a subtype of the Oriental despotism, or does it represent a separate major type of political system like the maritime state, the feudal state, the modern industrial democracy, and-we might add from the previous discussion-the primitive democracy, the gentile aristocracy, and the archaic republic? Again the writer cannot give a definite answer, and can only recommend the problem as one well worth independent investigation and record his own impressions. On the whole, the African despotism, even when all accretions reasonably attributable to diffusion from Pharaonic Egypt are stripped away, strikes him as much too similar to the Oriental despotism to warrant its establishment as a distinct major type of political system. On the other hand, he finds it impossible to ascribe the system to an economic basis in irrigated agriculture — the hydraulic foundation on which Wittfogel believes all except marginal and derivative Oriental despotisms rest. Irrigation is, to be sure, by no means unknown in Negro Africa, but the regions of greatest political complexity are almost invariably characterized by shifting swidden, or slash-and-burn, agriculture. What, then, provides the economic basis for the African despotic state? To the scholar who may wish to pursue the subject, the writer can offer only one suggestion: Is it, perhaps, the exploitation of slave labor, which Wittfogel finds surprisingly minimized in his Oriental despotisms but which is enormously developed in many of the most typical African states?
Specialists steeped in the African literature commonly discern wide differences among the complex political systems of different areas. To the present writer, coming to the subject after a survey of another continent, such differences appear superficial in comparison with the extraordinary resemblances in fundamental features and even in external forms. As contrasted with the diversity of complex political structures in aboriginal North America, exemplified by those of the Aztec, the Creek, the Iroquois, and the Natchez, the states of Negro Africa appear essentially as similar as the peas in a single pod. However geographically remote and however dissimilar the other aspects of culture, political forms seem everywhere to conform to a single fundamental pattern. Nor does the size of a state seem to make any essential difference. Even a petty paramount chief who has subjugated a few neighboring communities and destroyed the preexisting primitive democracy seems invariably to institute, in so far as he can on a small scale, the forms prevailing in larger states in the vicinity or even at some distance. It is almost as though all of Africa south of the Sahara were permeated, as it were, by a mental blueprint of a despotic political structure, transmitted from generation to generation as a part of traditional verbal culture, and always available to be transmuted into reality whenever some individual arises with the imagination, enterprise, strength, and luck to establish, with the aid of his kinsmen, an authoritarian regime over people residing beyond the limits of his .local community.
Some of the widespread similarities in basic pattern and in details of form which have led the author to his conclusion of the essential uniformity and single origin of despotic states in Negro Africa may now be listed.

Monarchical absolutism
Each king or independent paramount chief enjoys absolute power, at least in theory.
Eminent domain
All land, livestock, and wild game in the state belong in theory to the monarch, providing a basis for his right to derive an income from them.
Divine kingship
Either the ruler himself is divine or he has unique personal access to the dominant divine powers.
Ritual isolation
The king is isolated from physical contact with all except a few attendants and intimates. Often he eats in private or must be fed by others, or his feet may not touch the ground, or he is concealed by curtains because his glance is considered dangerous.
Insignia of office
Royal status is symbolized by the possession of distinctive regalia, among which stools, drums, and animal tails are especially common.
Capital towns
The ruler resides in a capital town along with his attendants and ministers. Typically each new monarch founds a new capital or at least establishes a new royal residence.
Royal courts
The monarch maintains an elaborate court with pages, guards, entertainers, personal attendants, treasurers, and a variety of chamberlains with specialized functions.
Protocol
Behavior at court follows detailed rules of protocol, of which abject prostration in the presence of the monarch is a nearly universal ingredient.
Harems
The ruler is invariably surrounded by a large number of wives and concubines.
Queens
At most royal courts a Queen-Mother, a Queen-Consort, and a Queen-Sister, or at least two of the three, enjoy extraordinary prestige, even sometimes technically outranking the king himself. Queens are commonly endowed with independent estates and often exercise restricted political authority.
Territorial bureaucracy
For administrative purposes each state is divided into a territorial hierarchy of provinces, districts, and local communities with bureaucratic officials at each level responsible for maintaining order, collecting and transmitting taxes, and levying troops and corvee labor. Even where bureaucratic posts are hereditary rather than appointive, their occupants are firmly subordinated to the central authority.
Ministers
Resident at the capital as assistants to the ruler in the exercise of centralized authority are always a number of ministers of state, the most important of whom form a supreme advisory council. They are distinguished by specialized functions, e.g., a vizier or prime minister, a military commander in chief, a chief justice, a royal executioner, a custodian of the royal tombs, a supervisor of royal princes and princesses.
Duality of ministerial roles
Almost universally, the ministers combine their specialized functions at the capital with offices as provincial governors in the territorial organization.
Titles
Characteristic of African states is a great proliferation of titles. Although a few or many may be hereditary, there are always a large number bestowable by the monarch in return for loyal services, and competition for these is often keen.
Security provisions
To prevent palace revolutions a king's brothers, as the most likely usurpers, may be killed, blinded, incarcerated, or banished from the capital. To prevent revolts in the provinces, positions as governors are commonly filled, not by members of the royal lineage, but by persons of categories ineligible to succeed to the throne, e.g., commoners, elevated slaves, eunuchs, or, where succession is patrilineal, sisters' sons.
Electoral succession
Although the ruler often designates an heir presumptive, and may even invest him with ministerial authority, succession to the throne is almost never automatic. The decision usually rests in the hands of a committee of ministers with constitutional electoral powers, who are free to follow or ignore the late king's wishes. Not infrequently the succession shifts regularly from one to another branch of the royal lineage.
Anarchic interregnums
Since there is always a plurality of candidates with strong supporters, and considerable political maneuvering may be necessary before the electors can agree upon a successor, a period of several days or even weeks usually intervenes between the death of one king and the selection of the next. During this interregnum laws are relaxed and social disorder prevails, often accentuated by a resort to arms by the partisans of rival claimants.
Human sacrifice
In many Negro states the funeral of a king is accompanied by human sacrifices, sometimes on an extravagant scale.

Naturally not all the above features occur in every despotic state south of the Sahara, but the great majority usually do, and the degree of association among them is far too strong to be attributable to chance .

Selected Bibliography
Fortes, M., and E. E. Evans-Pritchard, eds. African Political Systems. London, 1940.
Irstam, T. The King of Ganda. PEMS, n.s., 8:1-203. 1944.
Montagne, R. Les Berbères et le Makhzen dans le sud du Maroc. Paris, 1930.
Onneken, D. A. Die Konigskultur Kaffas und der verwandten Konigreiche. Frankfurt am Main, 1956.
Schapera, I. Government and Politics in Tribal Societies. London, 1956.
Wittfogel, K. A. Oriental Despotism. New Haven, 1957.

7. — History

The fact that this volume has a primarily historical objective makes it necessary to face squarely the issue of the validity of historical anthropology and of its various methods and techniques. This is the more imperative since a substantial number of outstanding anthropologists, especially in Great Britain, exhibit a profound distrust of the historical approach and consistently refrain from using it even in an auxiliary capacity. This prejudice, for it cannot be characterized otherwise, stems directly from Radcliffe-Brown (1923 and 1950), who throughout his career habitually referred contemptuously to the work of historical anthropologists as "pseudo history" or "conjectural history." This canard cannot be allowed to stand unchallenged.
The writer agrees with the great majority of his American colleagues that history and science are equally legitimate objectives in anthropological work and that neither approach is inherently more likely than the other to produce valid results. To be sure, only some of the historical research in anthropology can be characterized as sound; much is admittedly bad, and some, though laudable in scope and method, is unfortunately simply wrong. But scientific research in anthropology can be similarly categorized as sound, bad, and wrong, and in approximately the same proportions.
Work in the African field, whether historical or scientific, which the author considers definitely bad he simply ignores in this book, whatever the reputation of the persons responsible. He also abjures the methods which he considers fundamentally unsound, but these must be specified. One is the use of similarities in tribal or place names as a basis for drawing historical inferences. Such resemblances have absolutely no evidential value of themselves, and conclusions drawn from them cannot be accepted unless solidly buttressed by other and more dependable evidence. A second unsound method consists of postulating migration or diffusion from general similarities in complex social phenomena, e.g., from the occurrence of totemism, or mother-right, or potlatches and feasts of merit in geographically remote areas in the absence of any probability of a historical connection on other and more substantial grounds.
Equally unsound, but unfortunately extremely prevalent among both ethnologists and archeologists, is the technique which the writer dubs "trait chasing." This consists in drawing conclusions as to historical connections from specific and detailed cultural resemblances occurring in different regions, e.g., the presence of a similar type of adze blade in Polynesia and South China or of a double spiral motif in both the Caucasus and northwestern Argentina. Even the compilation of a number of supportive similarities between the same two regions has no evidential value if the probability of contact or migration cannot be established on other grounds. Linguists have estimated that, for any two unrelated languages selected at random from anywhere in the world, as much as 6 per cent of their total vocabularies may reveal such similarities in both form and meaning that competent specialists would unhesitatingly accept them as cognates if the languages were known to be related. Similarly, in ethnology, if one starts with a single resemblance between any two remote and unrelated cultures, a search for others can easily reveal scores or even hundreds.
More difficult to assess are those cases, in both scientific and historical anthropology, where wrong conclusions have been reached despite essential soundness in method. The great evolutionists like Morgan and Tylor provide, of course, the classic illustration. In scientific anthropology, or "comparative sociology," Radcliffe-Brown must be placed unequivocally in this category, for his interpretations of particular ethnographic phenomena have proved erroneous in every instance where they have been brought into question, e.g., by Deacon, Junod, Kroeber, and Lawrence. His strictures against historical anthropology can rherefore probably be interpreted as a typical Freudian reaction fonn51tion against an equally just (or unjust) characterization of his own work as "pseudo science" or "conjectural sociology." This does not detract from the value of his contribution in stimulating others, for Morgan, whose interpretations were at least as erroneous, still enjoys a solid reputation as the grandfather of kinship studies.
A strictly parallel case in historical anthropology is the Kulturkreislehre of the German-Austrian "culture-historical" school. Despite a genuinely admirable methodology, e.g., in the criteria of form and quantity, this group reached erroneous conclusions in their postulation of a small series of culture complexes which allegedly spread as units over enormous areas of the world's surface. The culture history of the present volume has nothing in common with the approach of this group. A criticism, however, is not in order since the leading exponents of the theory publicly renounced it at a world anthropological congress in 1956, in what was perhaps the most laudable demonstration of scientific integrity in the history of our subject. In tribute we can at least cite Ankermanu (1905), who investigated the culture history of Africa from this point of view, even though we cannot follow him.
Thoroughly sound work in historical and in scientific anthropology, leading to essentially dependable conclusions, is well exemplified on the one hand by Spier's North American researches and on the other by the recent British social anthropologists whose reports on African societies have proved indispensable in the compilation of this book. Greater sophistication in psychology and sociology may doubtless ultimately revise some of Spier's specific conclusions, and in occasional instances a greater awareness of history and process can demonstrate errors in the interpretations advanced by British Africanists; but incidental mistakes of this kind are thoroughly excusable and in no way detract from the essential merit of the work On the whole, the present writer agrees heartily with Eggan (1950 and 1954) that the most productive procedure in both scientific and historical anthropology is one which combines the best in the two approaches and constantly tests conclusions in either by evidence from the other.
The techniques of historical reconstruction employed in this volume are those discovered to be reliable by a very simple empirical test, namely, whether their use leads invariably to conclusions consistent with those derived from other independent and dependable methods. Several sources of evidence fall unmistakably into this category. First, of course, is the direct testimony of contemporary written records. These are invaluable for North African history, but unfortunately rarely yield information of any considerable time depth elsewhere on the continent. The second is the equally direct testimony of materials excavated in archeological sites, from which, however, we must emphatically except the inferences that archeologists have drawn from these findings, unless their interpretations are buttressed by independent evidence. A third and equally dependable source is linguistic relationships, which have already been discussed in Chapter 3. A fourth is the evidence from botany and from the distribution of cultivated plants, as presented and analyzed in Chapter 4. A fifth is the methodology for inferring earlier forms of social organization in a particular society from internal structural inconsistencies reflecting the conservatism of certain features, as presented elsewhere by the author (Murdock, 1949). A sixth is ethnographic distributional analysis as applied by the more cautious American historical anthropologists, which is characterized particularly by its restriction to areas of continuous distribution of cultural forms and by its refusal to jump geographic gaps without exceedingly compelling reasons. This technique may be combined with a conservative utilization of the age-area hypothesis, although the author refrains from doing this except where the conclusions receive confirmation from at least one of the other five basic techniques.
As intimated above, these six independent methods of reconstructing the past support one another. The author has not encountered a single instance in Africa where suggestive leads from two or more of these sources have indicated inconsistent conclusions. One typical example of their mutual confirmation must suffice. The author was first led to suspect that the Barea and Kunama tribes of Eritrea might have come from the Nile Valley 300 miles to the west by Greenberg's demonstration that their languages belong to the Sudanic stock. Subsequent analysis of their cultivated plants revealed that these, unlike the crops of their present immediate neighbors, belong mainly to the Sudanic complex like those of Nubia. Their matrilineal social organization, contrasting sharply with the patrilineate of their neighbors, proved thoroughly consistent, according to the author's methods of social structural analysis, with what ethnographic distributions and age-area analysis indicated as probable in the Nile Valley before the Moslem conquest of the kingdom of Alwa in A.D. 1504. The author subsequently discovered a clinching confirmation of the validity of these various techniques in the chronicle of an Arab traveler who visited Alwa in 872 and referred to the "Barya" and "Cunama" as tribes then living on the borders of that kingdom.
The discussion thus far has ignored one source of historical information of which the African literature contains a vast quantity, namely, the oral historical traditions of the African peoples themselves. No dependence whatsoever is placed upon evidence of this type in the present volume, since repeated comparisons of local traditions regarding places or directions of origin with inferences from the six techniques of proven reliability have indicated, for time depths of over a century, an agreement of not greater than 20 or 25 per cent, or approximately what one could expect on the basis of chance alone. The author is consequently forced to conclude that indigenous oral traditions are completely undependable much beyond the personal recollections of living informants, unless they happen to be of the very unusual type, characteristic of Polynesia, where automatic self-correction results from their use for validating social status or land claims. It is ironical that British social anthropologists, as in the volumes of the Ethnographic Survey of Africa, invariably give serious consideration to native historical traditions, the one type of historical information that is virtually valueless, and completely ignore the evidence from every genuinely reliable source except written records.
The details of the culture history of Africa must, of course, await development in the regional chapters that follow. Since, however, the rationale of the particular sequence followed may not be immediately apparent, and since readers interested only in certain regions may require a summary orientation as a background, it seems desirable to insert at this point an extremely condensed synopsis of the major historical movements on the continent as a whole over the past 7,000 years.

5000 To 4000 B.C.

4000 To 3000 B.C.

3000 To 2000 B.C.

2000 To 1000 B.C.

1000 To 1 B.C.

A.D. 1 To 500

A.D. 500 To 1 000

A.D. 1000 To 1500

A.D. 1500 To 1900

Selected Bibliography
Ankermann, B. Kulturkreise und Kulturschichten in Afrika. ZE, 37:54-90. 1905.
Baumann, H., R. Thurnwald, and D. Westermann. Volkerkunde von Afrika. Essen, 1940.
Eggan, F. Social Organization of the Western Pueblos. Chicago, 1950.
Eggan, F. The Method of Controlled Comparisons. AA, 56:743-763. 1954.
Murdock, G. P. Social Structure. New York, 1949.
Radcliffe-Brown, A. R. “Methods of Ethnology and Social Anthropology.” SAJS, 20:124- 147. 1923.
Radcliffe-Brown, A. R. Introduction. ASKM, pp. 1- 85. 1950.

Note
1. Greenberg, in a personal communication while this book was in press, reports that new information definitely indicates the affiliation of both Nyangiya and Temainian with Sudanic. He likewise suggests a distant relationship between the Furian, Kanuric, Mahan, Sudanic, and probably also the Koman and Songhaic stocks. On a still more tentative basis he considers a connection between Kordofanian and Nigritic to be probable.