Finish the HW with quailty work.

Economy and Feudalism in Africa Author(s): Jack Goody Source: The Economic History Review, New Series, Vol. 22, No. 3 (Dec., 1969), pp. 393-405 Published by: Wiley on behalf of the Economic History Society Stable URL: http://www.jstor.org/stable/2594117 . Accessed: 06/08/2013 15:51

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

Wiley and Economic History Society are collaborating with JSTOR to digitize, preserve and extend access to The Economic History Review.

http://www.jstor.org

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

THE

ECONOMIC HISTORY

REVIEW

SECOND SERIES, VOLUME XXII, No. 3, December i969

Economy and Feudalism in Africa

BY JACK GOODY

IN an earlier paper in another journal I discussed the application of the term "feudal" to pre-colonial African states from the formal point of view and I concluded that "there appears little to be gained by thinking of African soci-

eties in terms of the concept of 'feudalism'."1'. The usage, and indeed debate, have continued; the number of feudal states in

Africa has proliferated in the last few years, the authors of the accounts explicitly or implicitly rejecting my caveat about the use of this term.2 I should make it clear that these doubts are not about the possibility of finding resemblances be- tween the states of medieval Europe and those of pre-colonial Africa: there are plenty of similarities in the structure of monarchical systems of government, wherever they are found.3 My queries had to do with the utility of any analysis that rested on such vague and all-embracing concepts and which approached the situation in terms of fixed categories rather than sets of variables. But there are more precise reasons why the overall comparison with medieval Europe seems in- appropriate and this has, initially, less to do with government and politics than

1 J. Goody, 'Feudalism in Africa', Journal of African History, iv ( I963), I I . I would like to thank G. C. Homans, E. Miller, and M. Postan for their comments at various times,

especially helpful to one whose knowledge of medieval Europe is as limited as mine. I gave an earlier version of this paper as the StJohn's College lecture to the University of East Anglia, which offered me the opportunity of developing certain themes in a more general context.

2 See E. M. Loeb, In Feudal Africa (Bloomington, I 962); J. Beattie, 'Bunyoro: an African feudality ?', Jnl. African Hist. v (i964), 25-35; R. Cohen, 'The Dynamics of Feudalism in Bornu', in African History, ed. J. Butler (Boston Univ. Papers on Africa, vol. i I, I 966); J. Lombard, Structures de typefdodal en Afrique noire: Jtude des dynamismes interned et des relations sociales chez les Bariba du Dahomey (The Hague, I 965). For an extreme position see Pierre Bettez Gravel, 'Life on the Manor in Gisaka (Rwanda)', jnl. African Hist. vi (i965). Gravel's article is based on field work in i960-I; in it he "purports to describe certain specific aspects of life on a small 'manor' in eastern Rwanda, and to show how remarkably similar they are to the same aspects of life on a baronial manor of medieval continental Europe" (p. 323), especially with regard to the absence of "true markets", the "economy of subsistence", and the "self-sufficiency of the com- munity". For a view much closer to my own see Edward I. Steinhart, 'Vassal and Fief in Three Lacus- trine Kingdoms', Cahiers d'Jtudes africaines, VII ( I967), 606-23 . In general these discussions place very little emphasis upon the basic technology of medieval Europe, the use of mills, animal traction, etc.

3 For a discussion of some of these similarities see my introduction to Succession to High Office (Cambridge, I 966).

393

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

394 JACK GOODY

with economics and technology. In my opinion, most writers about African social systems, particularly when they are dealing with class and government, have failed to appreciate certain basic differences between the economies of Black Africa and of the Eurasian continent, and this failure has led to superficial com- parisons not only in relation to "feudality" but also in relation to marriage, descent groups, property, and other important spheres of social action.'

Before I continue, I add a disclaimer. The identification of "feudalism" in Africa has been associated with the left rather than the right. Palaeo-Marxists accept a fixed progression, inherited from their nineteenth-century predecessors, from tribalism through feudalism to capitalism, though a greater element of flexibility is introduced with the recent publication of some of Marx's writings.2 Others regard the discovery of the same processes in Africa that earlier occurred in Europe as crucial to the recognition of African history as a proper subject for academic discussion. Such an approach seems to me understandable, often correct, but in this case misguided. We are here concerned with the utility of analytic concepts and whether "feudalism" illumines more than it obscures. I suggest that we need to take a closer look at the means and organization of pro- duction in Africa and Europe instead of tacitly assuming identity in these impor- tant respects.

There are three interrelated aspects of the society I want to discuss: the system of exchange (that is, trade and markets), the system of production (especially the ownership of the means of production), and the military organization (and especially the ownership of the means of destruction).

My thesis is that while the pattern of trade showed little difference from Eurasia, and while the military organization displayed some similarities, at least in the savannah country of West Africa, productive relationships did differ in certain major respects. And secondly that these differences are relevant not only to the understanding of the past but to the present as well, and need to be taken into active consideration when formulating development programmes.

First of all I want to stress that the difference between Africa and Eurasia does not lie in the presence or absence of markets and, in some spheres of activity at least, of a market economy. Much writing about non-European economics is based on the island communities in the Pacific-of the Trobriands, Tikopia, or Rossell Island. These communities are atypical in that certain primary features of the economy arise from the fact that they are small, relatively isolated groups rather than just "simple societies". The concept of non-monetary economics is hardly applicable to pre-colonial Africa, except possibly for certain hunting groups of minimal importance. Africa was involved in a vast network of wide- ranging trade long before the Portuguese came on the scene. For East Africa we have a late first-century sailors' guide, the Periplus of the Erythrean Sea, 3 to the trade along the coast. Long before the Europeans arrived there were trade routes from Madagascar up to the East African coast, through the Red Sea and into the Mediterranean, along the Persian Gulf to India, South-east Asia, and Indonesia. By the time the Portuguese had reached East Africa, the Chinese had already

1 See 'Inheritance, Property and Marriage in Africa and Eurasia', Sociology, iII I 969), 55-76. 2 See Karl Marx, Pre-capitalist Economic Formations, ed. E. J. Hobsbawm ( I965), and the comment by

M. I. Finley, 'Slavery', International Encyclopaedia of the Social Sciences, XIV (i968), 307-I3. 3 Trans. W. H. Schoff (New York, I 9 I 2).

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

AFRICAN ECONOMY 395

been active there; before the development of the gun-carrying sailing ship on the Atlantic seaboard, the maritime commerce of the Indian ocean made western Europe seem an underdeveloped area.' Indeed the trade between Ethiopia, the Mediterranean, and the Indian Ocean had much to do with the developments in the Arabian peninsula, including the rise of Muhammad.2

In West Africa the medieval empires of the Niger bend were built up on the trade which brought salt, cloth, and beads south from the Sahara across to West Africa and took gold and ivory and slaves back to the Barbary coast and from there into medieval Europe. When the British defeated the Ashanti in I 896 they found a war shrine in Kumasi consisting of a bronze ewer and jug. The jug now stands in the medieval section of the British Museum as one of the finest examples of English craftsmanship at that time. It bears the arms of England and the badge of Richard II, and is inscribed with the following proverbs:

He that will not spare when he may he shall not spend when he would. Deem the best in every doubt till the truth be tried out.

Were the story known ofhow these pots reached the tropical forest of West Africa, it would encapsulate much of the economic history of trans-Saharan trade.3

From the point of view of mercantile economy, parts of Africa were similar to western Europe of the same period. Metal coinage was in use on the East African coast. In the west, currencies consisted of gold, brass, and salt, but more especially cowrie shells which, coming as they did from the Maldive Islands off the south of Ceylon, filled most of the necessary attributes of money. In certain respects this was a monetary economy. Trade was highly organized and in kingdoms such as Dahomey and Ashanti important sectors of the economy were under state con- trol, whereas in the savannah regions exchange was left largely in private (Muslim) hands. Most of the kinds of economic operations that were found in pre- industrial Europe were also to be found in Africa; even in the stateless societies of the interior, barter had been superseded by more complex forms of exchange and again production was rarely limited to subsistence alone; the extensive use of cowries from the Maldives and carnelians from Gujerat shows that they were all in some degree part of the economic system of the Old World. The impact of trade on social organization of course depends upon the degree to which productive activity is diverted to serve the purposes of external trade. My point is that except in the special fields of the wine and wool trade,4 the differences between the ex- ternal exchanges of Africa and early medieval Europe appear to have been rela- tively slight.

1 C. M. Cipolla, Guns and Sails in the Early Phase of European Expansion, I400-I700 (I965); R. B. Serjeant, The Portuguese off the South Arabian Coast (Oxford, I 963) .

2 The early trade in the Indian Ocean had important consequences for the population of Madagascar, the spread of Asian food crops to Africa, and for the change in the social organization of the coastal areas of eastern Africa and parts of the Arabian peninsula prior to the rise of Muhammad. Some aspects of these latter changes are discussed by E. R. Wolf, 'The Social Organisation of Mecca and the Origins of Islam', Southwestern Journal of Anthropology, vii (I 95 I), 329-56.

3 Guide to AMIediaeval Antiquities (I 924), fig. I 56. See also the fourteenth-century bronze ewer, in the same museum, which was "the great war fetish of the Ashanti Nation" (British Museum Quarterly, VIII, I 933, 52). For the European side of the trans-Saharan slave trade see Charles Verlinden, L'Esclavage dans 1'Europe mididval, I (Bruges, I955).

4 See E. M. Carus-Wilson, Medieval Merchant Venturers (I 954) .

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

396 JACK GOODY

If mercantile activity, then, was not vastly different from that of medieval Europe, what about other aspects of the productive system? It has been claimed by some writers that land tenure was feudal in kind; others have disputed this contention, denying the utility of the concept of a landed fief in Africa. Most of this discussion has taken place on a politico-legal level, but there is one crucial and obvious difference which has been largely overlooked. It is a difference which means that land tenure (and hence vassalage and landed fiefs) in Africa were basically different from Europe and indeed from the Eurasian continent gener- ally; and it has to do with the means of production rather than with productive relations, though its influence upon these relations is of some importance. Basically Africa is a land of extensive agriculture.' The population is small, the land is plentiful, and the soils are relatively poor. Moreover, one fundamental invention that spread throughout the Eurasian continent never reached Africa south of the Sahara, with the exception of Ethiopia. I am referring to that Bronze Age invention, the plough.

What effect does the plough have? In the first place it increases the area of land a man can cultivate and hence makes possible a substantial rise in productivity, at least in open country.2 This in turn means a greater surplus for the main- tenance of specialist crafts, for the growth of differences in wealth and in styles of life, for developments in urban, that is, non-agricultural, life.3 In the second place, it stimulates the move to fixed holdings and away from shifting agriculture. Thirdly (and not independently), it increases the value (and decreases the avail- ability) of arable land.

In Africa, then, there was little use of machines, even elementary ones; agri- culture meant hoe farming, which was carried out by men and women or both, depending upon the particular society. Indeed animal power, that drew the Eurasian plough, was not used for any form of traction. One immediate reason was that the wheel,4 though it crossed the Sahara, both in the West (as evidenced in the two-wheeled chariots liberally engraved upon Saharan boulders) and in

1 The difference between extensive and intensive modes of agricultural production is clearly relative, and one which has to be related to the nature of the soils, the labour force, and the terrain; shifting culti- vation continued in less fertile and accessible parts of medieval Europe long after the plough dominated the agricultural scene, and the same is of course true of Ethiopia today.

2 In forests it clearly has many limitations. Nor does the plough improve vegeculture to the same extent as it does the cultivation of cereals. The main point, however, was effectively made by V. G. Childe. See also W. M. McNeill, The Rise of the West (Chicago, I 963): "The harnessing of animal power for the labor of tillage was a step of obvious significance. Human resources were substantially increased thereby, since for the first time men tapped a source of mechanical energy greater than that which their own muscles could supply. The use of animal power also established a much more integral relation between stock- breeding and agriculture. Mixed farming, uniting animal husbandry with crop cultivation, was to be- come the distinguishing characteristic of agriculture in western Eurasia. It made possible a higher standard of living or of leisure than was attainable by peoples relying mainly or entirely upon the strength of merely human muscles" (pp. 25-6).

3 Urban centres of course existed in pre-colonial Africa; they ranged from the agro-cities of the Yoruba to the trading and administrative towns of the Saharan fringes and the coastal regions.

4 Animal disease was another factor limiting the use of the plough. Of Ethiopia, which may have obtained the plough from South Arabia or Egypt even in pre-Semitic times (about IOOO B.C.), F. J. Simoons has written: "Where there are animals suitable for ploughing, both Cushites and Semites use the plough; but where, as along the Sudan border, these animals are excluded by disease, even Semites turn to the hoe or digging stick for preparing their fields."-'Some Questions on the Economic Prehistory of Ethiopia', YnI. African Hist. vi (i965), i I. Iron-working, however, appears to have arrived from South Arabia at about the same time as writing, that is, in the fifth century B.C.-Anfray, 'Aspects de l'archeo- logie ethiopienne', JnI. African Hist. IX (i968), 352.

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

AFRICAN ECONOMY 397

the East (in Ethiopia and in the early Sudan), never penetrated pre-colonial Africa (or rather was never adopted there). Nor was this because of the lack of a metal technology. While Black Africa escaped the civilizing influence of the Bronze and Copper ages, the smelting of iron diffused from the Mediterranean down both sides of the continent.

In the East the technique of iron-smelting travelled to the Sudan, where Mero6 has been described, with some exaggeration, as the "Birmingham of Africa"; this movement occurred in the sixth century B.C., roughly the same time as iron was found in Ethiopia. From there it spread to Chad in the first century A.D., together with horsemen using a long lance. In the West, iron-working was transmitted from Carthage and the Barbary Coast to the Niger towns in the third century B.C. The technique spread down eastern Africa, being introduced to Zambia in "the first few centuries A.D." by a number of small related groups of immigrants who brought not only metallurgy but also food production and pot- making from the area west of Lake Tanganyika. Copper and bronze were em- ployed in many parts of pre-colonial Africa, but before the coming of iron the extent of this use was negligible.

The absence of the wheel meant that man was unable to use not only animal power but hydraulic power and wind power as well. This is why the recent intro- duction of the lorry, the bicycle, and the engine-driven mill has been so import- ant for the rural economy of Africa. But the lack of the wheel had another con- sequence for agriculture, since it limited the possibilities of water control. In the drier regions of the Eurasian continent the wheel has played a dominant part in raising water from wells to irrigate the land. Simple irrigation there is in Africa, as in almost every agricultural area. Some of the inhabitants of Birifu (Lo Wiili) in northern Ghana channel a permanent water supply to run among their fields, and thus get two crops a year in place of one. The Sonjo of Tanzania practise more developed water control. Rice-growing in the Western Sudan (and it should be remembered that Oryza glaberima was domesticated independently of Asian rice in the Senegal-Mali region) demands yet more positive measures.

There are other means of water control that do not involve the wheel, using various techniques of temporary storage. Methods of this kind did of course exist. Everywhere there was some improvement of natural pools. In Gonja and in neighbouring areas of northern Ghana there are many ancient cisterns hollowed out of the laterite; in the famous market town of Salaga, the city of i,ooo wells, these are cylindrical in form and do not seem wholly dependent upon surface water. But these storage systems are very different from the village tank of South- east Asia; while there is no lack ofwater in Africa, the problem ofits distribution is enormous. And in terms of agriculture what is lacking is any mechanical device for drawing water, such as is used in the Middle East and even in the Saharan oases.

One example of the pragmatic effects of the technological gap between Africa and Eurasia lies in the military field. When the Portuguese spearheaded Euro- pean expansion into other continents, they succeeded largely because of their use of gun-bearing sailing ships.' Through these they could dominate their African

1 At first they depended upon the cannon on their floating castles, later upon hand-guns. See Cipolla's useful discussion where he quotes Pannikar as saying that by I498 "the armament of the Portuguese ships was something totally unexpected and new in the Indian (and China) seas and gave an immediate advantage to the Portuguese."-Op. cit. p. 107.

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

398 JACK GOODY

opponents who were armed only with sword, spear, bow and arrows. But, by the end of the fifteenth century, when the expansion of Europe began, their guns were also far in advance of Asia as well. These technological innovations soon spread from Europe, just as simpler forms of gunpowder and "cannon" had earlier diffused there from China and the Middle East. But the way in which they did so is of interest. Beachy notes that "the Africans never seem to have learnt to make fire-arms as good as those of the Europeans, unlike the sixteenth and seventeenth-century Japanese and Sinhalese, who soon achieved virtual parity with the Portuguese in this respect."' Already in the mid-sixteenth century the Japanese were producing matchlocks and in the I 590's they were followed by the Koreans; Ceylon had become a centre of production by the end of the sixteenth century, when muskets and cannon were made at many different points in the Indian sub-continent. By the seventeenth century the inhabitants of the Malabar coast were exporting muskets to Arabia.2 The reason for the failure of Africans successfully to take up the manufacture of this powerful new weapon is a simple one. They did not possess the requisite level of craft skill in iron-working. As a result, Africans were at an enormous disadvantage when the scramble for their continent began, since they had to fight against the very people who were supply- ing them with arms.3

What does all this add up to in socio-economic terms? Firstly, rights in land were less highly individualized in most of Africa, partly

because land was not a scarce commodity. 4 Among the Bemba of Zambia rights 1 Review of B. Davidson, jnl. African Hist. III (I 962), 5 1 0. 2 See Cipolla, op. cit. pp. I 27-8, and Al-Djamfizi's history in Serjeant, op. cit. p. I I 7: "These Malabaris

are Muslims: the Munaibdri muskets are called after them." 3 The position was not quite as desperate as at first appears, since one European power was quite

willing to benefit at the expense of another; the sources never entirely dried up. And even within the same political unit Ce.g. the British-administered Gold Coast), the interests of merchants and administrators were often at odds. It should be added that, though sulphur had to be imported, there was certainly some manufacture of local gunpowder (e.g. in northern Ghana) but this was recognized to be of inferior quality (because it was not "corned"). Equally guns could be repaired. But there is no evidence of any extensive manufacture of components. Even today it is imported bicycle tubes that are often used to replace barrels.

The one exception may have been the late nineteenth-century Mandingo warrior Samory. In 1898 Nebout was reported to have been shown a small-arms factory "which was capable of turning out three repeaters a week". The gun is described "as a wonderfully close copy of the French article, but not bearing near inspection as to its details, more particularly in the matter of rifling".-Dir. Mil. Intell. to C.O. I4Jan. i898, C.O. Confid. Print, African (West), no. 549. Legasick quotes a similar report ('Fire- arms, Horses and Samorian Army Organisation, I 870-I898', Ynl. African Hist. VII (I 966), I o4). Peroz and Binger, who separately visited Samory in I 897, make no mention of locally made guns; Henderson, who was captured in i897, writes only of a cartridge factory. Local numu blacksmiths may have learnt new techniques from the invaders, but it is difficult to see how the low-temperature forge could be used for the manufacture of guns from scratch, as distinct from the kind of repair work that Binger describes (Du Niger au Golfe de Guind (Paris, I 892), I, I 9 I -2). In Europe the temperatures needed to emerge from the wrought- iron phase were first produced in the fourteenth century by using water-powered bellows for blast pur- poses, and this in turn depended upon rotary motion. But whether or not Samory succeeded in manu- facturing firearms, his continued attempts to secure imported guns by exchanging captives or by more direct methods indicate that his efforts in this direction met with little success. Indeed right to the end the military economy of his empire-building was based upon the necessity of acquiring sufficient income from booty to import guns, powder, and percussion caps.-J. Goody, Introduction, Ashanti and the North- west, Institute of African Studies (Legon, I 965), pp. 75-8.

4 The tendency for rights in land to be more highly individualized the greater the population density has been noted by C. K. Meek, Land Law and Custom in the Colonies (I 946), pp. I 49-50. See also G. I. Jones, 'Ibo Land Tenure', Africa, XIX (I 949), 3 I 3, and J. Goody, The Social Organisation of the Lo Wiili (I 956), p. 37. Of course, land shortage is always relative to the available resources of labour etc. But the tractor can create a shortage more easily than the plough, the plough more readily than the hoe. In my view

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

AFRICAN ECONOMY 399

in the productive use of land, other than the small proportion under cultivation at any one time, hardly existed; the same is still true among the Gonj a in northern Ghana. Under such conditions neither individuals nor kin groups bother to lay specific claims to large tracts of territory, since land is virtually a free good. Else- where, among the Lo Wiili or Tallensi of northern Ghana, the population densi- ties are aroundI 00-200 per square mile and rights in land more highly developed; productive resources tend to get tied down to small kin groups, minimal lineage segments, although residual rights are vested in larger descent groups which often see themselves as property-holding corporations. But in fact rights to the assets are divided up among the smaller units, which are themselves constantly splitting and reuniting, depending upon the distribution of the births of male children. But, overall, people were thin on the ground. Even today, the total figure for Black Africa is not much more than the population of the United States, although the surface area is four times as great.

Not only was land plentiful; it was also less productive than on the Eurasian continent, partly because of the technological limitations and partly because tropical soils are often of poor quality. Moreover, since processes of soil regenera- tion (either by manure or by special cropping) were limited in nature, the fer- tility of land soon fell off. Under these conditions, the answer usually lay in mov- ing one's farm to a new site (though not necessarily one's residence), that is, in shifting cultivation.

The social consequences were twofold. Politically, chiefship tended to be over people rather than land, and these a leader tried to attract as well as restrain. The conditions for the forms of domination that obtained in the European Middle Ages hardly existed, with the exception of slavery.' It is highly significant that only in Ethiopia, which had the plough, was there any landlordism in Africa; here, in true medieval fashion, estates in land supported a nobility that filled the important offices of state in both the staff and line organization, a nobility that was at the same time a leisure class in Veblen's sense. Besides the nobility one also found ecclesiastical landlordism-functionaries whose time was devoted to the glory of God (though individual commitment to the monastic life was often tem- porary rather than permanent in character) derived their "living" from the church with which God had been endowed. 2

there is no doubt that the "communal" nature of African land tenure must break down rapidly with the introduction of plough or tractor, just as the economy of large households must tend to fragment when individual members acquire their own pay-packets, by selling their labour outside the domestic group.

1 Prof. Edward Miller has pointed out to me that there is another situation which leads either to serfdom or to plantation slavery. Both institutions can emerge under conditions where land is relatively plentiful and where it is necessary to prevent the escape of rent producers or labour producers which a chief or land requires. Perhaps an extreme form of this is the slavery found in pioneer settlements in certain parts of early medieval Europe. However, in Africa labour requirements led to slavery but not serfdom; trading towns like Kano and Bida in northern Nigeria, or Salaga and Bole in northern Ghana, were sur- rounded by villages of slaves which supplied the ruling and commercial groups. Domestic slaves, depen- dent kinsfolk, and clients filled other servile roles, but the supply of land and the degree of control made it difficult to exploit labour by anything other than slavery.

In Middle America, despite ample land, the Spanish conquerors imposed on the population a system of peonage, which was something less than slavery. But the conditions of productivity (the plough, soil, crops) and control (guns and horses) were very different.

2 In Ethiopia the church was the largest landowner after the Emperor. Some of these lands were worked by the monks; others were farmed by peasants. The landlord's rent, lay or ecclesiastical, could be enor-

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

400 JACK GOODY

If there are landlords, there can also be tenants and serfs; unfree tenancies mean little unless land is highly valued and the peasantry has nowhere else to go.1 Under conditions of shifting cultivation, it means little. Slavery was important throughout most of Africa: war captives were given household or agricultural work to perform for their victors or their purchasers. But ties of subordination arose not out of shortage of land but as the result of purchase or conquest, giving rise to slavery rather than to serfdom.

Though there were no landlords, there were of course lords of the land-the local chiefs of centralized states who, from the standpoint of food production, were in a sense carried by the rest of the population; we may look at this either as a return for services rendered or as the exploitation of the weak, for there is, I think, no real test. Chiefs could not be expected to hear complaints on an empty stomach. Their families did better than those of commoners; they rode where others walked. But in general, owing to the limited nature of the technology and to the relatively low differentiation in terms of levels of consumption, standards of living, as measured by the usual tallies, were not markedly different. As Gluck- man has remarked of Zulu chiefship, one man can eat only a limited amount of porridge; the rest has to be distributed.2 Probably the maximum differentiation in terms of what Weber called "styles of life" was to be found in the trading states of the coast, in the empires of the Niger bend, and in the emirates of northern Nigeria, where there were certain modes of behaviour that distinguished the urbanites of the capital (the dynasty, the merchants, the learned men, the specialist workers and their hangers-on) from the rural population. But in most cases such differentiation was confined to chiefs themselves rather than to the dynasty as such, being a function of roles rather than social strata; and the most noticeable aspect of the difference lay in control over women and slaves (war booty), and guns and horses, rather than goods and land. The exceptions were to

mous. At the end of the eighteenth century, Bruce stated that the tenants of Tigre usually surrendered at least half their crop, the landlord supplying the seed. But it was "a very indulgent master that does not take another quarter for the risk he has run".-See Richard Pankhurst, An Introduction to the Economic History of Ethiopia (I 96 I1), pp. I 93-5. A similar situation existed in Egypt with regard to the waof lands. See, for example, Ibn Khaldun's reference to chieftains who "are wont to build Mosque schools, shrines and almshouses and to endow them with Waqf [Mortmain] lands. . ."-Charles Issawi, An Arab Philo- sophy of History (I 950), p. I 44.

Of course, monasticism could also be supported by means other than land-e.g. by "taxes" on trade or by providing services for travellers. But this was not, I think, possible to any great extent under pre- colonial conditions in Africa.

1 The reduction of population in relation to resources can have the opposite effect on serfdom. When in the fourteenth century land in Europe became relatively plentiful (through either the Black Death or other causes) rents were halved in a decade, wages doubled, and the institution of serfdom was greatly weakened.

2 Gluckman suggests that the failure of the south-east Bantu to develop more extensive political units before the time of Dingiswayo at the beginning of the nineteenthcenturymayhave been due to thelimited technology and the availability of land, so that there was possibly little point in building up power. "The tribal economy was simple and undifferentiated; even in a good year the available technology did not allow a man to produce much beyond his own needs. There was little trade and luxury, so even a con- queror could not make himself more comfortable than he had been before. One cannot build a palace with grass and mud, and if the only foods are grains, milk and meat, one cannot live much above the standard of ordinary men."-M. Gluckman, 'The Rise of a Zulu Empire', Scientific American, ccii (i960), I57-68. In West Africa the economy was more developed in these respects, particularly the military economy insofar as it depended upon cavalry and the importation of firearms and gunpowder. But the general point still holds.

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

AFRICAN ECONOMY 401

be found in the coastal strip, where ruling and merchant groups financed them- selves out of the European and Asian trade: there are certainly some sumptuary distinctions reported by travellers from Dahomey and Ashanti. But these differ- ences do not come near to those of medieval Europe, or even of Ethiopia, where a series of sumptuary rules confined to the nobility the playing of certain musical instruments, the brewing of honey wine, and similar forms of behaviour that were by definition of high prestige.1

I have been arguing that although there existed local chiefships (a line organ- ization) supported partly out of agriculture, partly from trade, there was nothing equivalent to estates in land of the European kind. This should perhaps be quali- fied by "only in some places and under limited conditions". For, turning from the line to the staff organization, the suggestion is found that the large-scale develop- ment of central government, which as Weber had argued involves the creation of appointive (as distinct from hereditary) office, was possible in only a few African states because of the land situation that I have described.2 The southern Bantu failed to develop in this way. But some states succeeded. Among those that did were the highly centralized kingdoms of Buganda,3 of Barotseland (Zambia), and of Dahomey. In each of these special conditions existed. In Buganda, the banana provided the basis for continuous cultivation; in Barotseland, there were the fertile mounds of the Zambesi flood plain; in Dahomey there was the develop- ment of some plantation agriculture.4 In all these states there was a system of "office estates" (rather than fiefs proper), i.e. estates attached to the staff rather than the line organization.

But if limited estates in land, sufficient to support appointive offices, were only rarely built up under conditions of this kind, was it possible to use benefices of a different kind in a similar way? In Ruanda, estates in cattle were employed to create relationships of subordination that resembled the cattle clientage of early Ireland.5 Such estates are less easy to assign for the support of appointees to offices. They easily become fused with the personal property of the incumbent; and in any case, support by livestock is the formula for a very much looser polity than the predominance ofappointed office suggests; it is difficult to centralize cows.

A more likely form of benefice to substitute for the estate in land is one based upon the allocation of income from tax or trade, which also provided one of the bases of support in western Europe. Few states of Africa were not involved in

1 Levine notes that the upper stratum of Abyssinian society "created and shared what might be called a 'gentry sub-culture', even though it was not so differentiated from the general culture as was the case in Europe and China."-D. H. Levine, Wax and Gold (Chicago, I 965), p. I 56. It should be added that sumptuary laws in the strict sense often arise when certain individuals or groups are challenging the prestige behaviour of old-established classes; they are a defensive mechanism to help maintain the existing social hierarchy. See, for example, Edward II's Ordinance against extravagant housekeeping, 6 Aug. I 3 I 6.-W. Stubbs, Chronicles of the Reigns of Edward I and Edward II, I (I 882), 238-9.

2 E. Colson speaks of the variable as the "attitude to land", but the context of her remarks allows us to give this a more concrete reference. See 'The Role of Bureaucratic Norms in African Political Structure', in Systems of Political Control and Bureaucracy in Human Societies, ed. V. F. Ray, Proceedings of the American Ethnological Society (Seattle, I958).

3 See C. C. Wrigley, 'Buganda: an Outline Economic History', Economic History Review, 2nd ser. x (I957), 69-80.

4 And more importantly, perhaps, the European trade. 5J.J. Maquet, The Premise of Inequality in Ruanda (1 96 I), pp. 129 ff; E. M. Chilver, 'Feudalism in the

Interlacustrine Kingdoms', in East African Chiefs, ed. A. I. Richards ( I960), p. 390.

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

402 JACK GOODY

long-distance exchange. In the case of direct coastal trade with Europeans (as in Dahomey and Ashanti) goods could be channelled through the machinery of government. Trading privileges could be granted directly by the king and taxa- tion held few administrative problems.1 But even in the savannah kingdoms where state import-export systems did not exist, the trading communities could be forced to contribute to the support of office-holders in a variety of ways-by market taxes, by customs dues, and by transport charges.

Other more problematic methods of endowment did exist. These lay in the field of military organization, which is the third aspect of the social system I am considering. Here, at least in the savannah country of West Africa, the military technology resembled that of feudal Europe in that war was dominated by the use of horsed cavalry. At this point I refer to the thesis elaborated by Heinrich Brunner concerning the relationship of feudalism and cavalry in Europe. He maintained that the great change from infantry to cavalry fighting came in the eighth century, some time after the Frankish king, Charles Martel, faced the Saracens and their horsemen near Poitiers in 733. In this battle Charles was vic- torious, but because of an absence of cavalry could not follow up his success, and Brunner held that the change to cavalry occurred shortly afterwards and marked the beginning of feudalism. For this "was essentially military, a type of social organisation designed to produce and support cavalry". 2

Discussing this thesis, Lynn White doubts whether the Saracen invasions them- selves produced the change, for the redistribution of lands had already begun by 732, and moreover the Muslims made little use of horses. It was rather the advent from the East of a technological device that made mounted warfare possible, namely the foot-stirrup. The traditional Frankish arm had been infantry. The stirrup, however, enabled the horseman to gain greater control so that he could wield a sword or charge his enemy with lance at rest. With the coming of the stirrup, thefrancisca, the distinctly Frankish battle-axe, and the ango, or barbed javelin, disappear. The long sword comes into use and so too does the spear with a heavy stock and a spur below the blade to prevent too deep a penetration. And this kind of horsed warfare led in turn to developments in the sphere of defence, especially the use of increasingly heavy armour. This new form of combat required a considerable capital expenditure; the military equipment of one warrior cost about 20 oxen, or the plough-teams of at least ten peasant families.3 To maintain cavalry of this kind there had to be some pay-off for those making the investment. In France the redistribution of church lands appears to have met this need, providing an income, a pasture, and a training ground for "the feudal class" which, as White remarks, "existed to be armed horsemen, cavaliers fighting in a particular manner which was made possible by the stirrup".4 The original and basic knight's service was mounted shock combat; it was around the cavalry that notions of chivalry arose, that there developed a knightly ethic based upon loiautee to the leader and proesce in arms. "The feudal aristocrat might, indeed, be a ruler, but this was incidental to his being a warrior."5

1 K. A. Busia, The Position of the Chief in the Modern Political System of Ashanti (I 95I), pp. 79 ff; I. Wilks, 'Aspects of Bureaucratization in Ashanti in the Nineteenth Century', Jnl. African Hist. VII (1 966), 2 I 5-32; M. J. Herskovits, Dahomey, an Ancient West African Kingdom (New York, I 938).

2 L. White, Medieval Technology and Social Change (Oxford, I 962), p. 3. 3 Ibid. p. 29. 4 Ibid. p. 28. Parts of the White thesis seem too rigidly tied to the stirrup. 5 Ibid. p. 30.

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

AFRICAN ECONOMY 403

The horse penetrated into Africa down the Nile valley into the Sudan, and across the Sahara, from north to east, to the region of the Niger and Lake Chad. So later did the stirrup-the toe-stirrup in Ethiopia, but elsewhere the foot- stirrup. However, there was no changeover to heavy cavalry equipment; the heavy lance never replaced the javelin, and the coats of kapok, mail, and Koranic leather charms worn by the horsemen were mainly a protection against infantry rather than against opposing cavalry.

The failure to develop heavy cavalry was no doubt partly due to the weaker economic base, which inhibited the accumulation of the required capital. But in any case lighter cavalry was characteristic of the Berbers of North Africa, who were not directly faced with Christian arms; ' and also the lighter cavalry of the Turks and other Muslim powers had the advantage of speed, in attack, in pur- suit, and in escape.2 The added mobility placed a premium on raiding rather than warfare. In Europe, the cavalry partly reimbursed themselves by loot, by slaving, and by the ransom for their more valuable captives; at other times they had their estates to sustain them. In Africa, the cavalry were maintained partly by tribute and by protection payments, but largely from booty. This was not taken primarily from material goods for there was little to seize from the average farmer except his family. The pay-off was in human booty, captives to be sold as slaves. Booty was indeed part of the productive system of the ruling class. A measure of this close interdependence of cavalry and raiding is the situation that obtains in the eastern Gonja capital of Kpembe, the twin city of Salaga, at the present day; a cavalry headquarters at the end of the last century, it now boasts only one horse; yet on the wall of every entrance-hut hangs the dusty and disintegrating harness that tells of former glories.3 The establishment of colonial overrule spelt the end of the raiding and control, thus undermining the economic position of the ruling estate.

Although there was no development of heavy cavalry, the military system had nevertheless some basic similarities with that of medieval Europe. Here I refer not so much to the trappings of war, some of which were indirect borrowings, but rather to associated features of the political system. For the very nature of cavalry warfare imposes certain patterns upon the social organization.

In the first place the cavalry usually provide the ruling strata, since they also provide the means of destruction, the means of conquest, the means of booty. Unlike military technologies based upon bows and arrows or upon iron infantry weapons, there is a built-in stratification between horse-soldier and foot-soldier. 4 Secondly, horsed warfare demands a considerable investment of skill as well as of money-a long training in horsemanship which again tends to make for differ- ences in "class consciousness", for ideas of nobility and the knightly ethic. Thirdly, the investment of skill and capital was not a free gift to the nation. Not

1 Ibid. p. 36. 2 Cipolla, op. cit. p. 9i. 3 Seej. A. Braimah andJack Goody, Salaga: the Strugglefor Power (i967). 4 The horsemen are particularly useful for fighting peasants and collecting captives, but the archers

still have an advantage at close quarters, behind defensive positions. The layout of buildings and settle- ments in the western Sudan was profoundly influenced by cavalry (for examples of fortifications see Binger, op. cit. pp. 93, I 7). With the advent of the horse, the cavalry became the offensive or predatory arm of states, whereas the bow was employed mainly for defence, though from the later sixteenth century it was supplemented by imported firearms, initially in very short supply.

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

404 JACK GOODY

only was there a pay-off in booty but there was also a political increment. Since military power was centred upon the cavalry, they were well set to demand a share in political power. This they might achieve by acting as kingmakers, or else as strong divisional chiefs. But in West Africa the diffuse government characteristic of European feudalism often took the form of rotational succession by which the paramountcy passed between the various segments of the ruling estate. 1

Cavalry states have certain common features both in feudal Europe and in circum-Saharan Africa. But most of Africa was unsuitable for the horse, which is more sensitive to tsetse fly than other livestock. Not only is the animal absent from all forest areas; it never penetrated south of the Sud marshes of the Sudan. Con- sequently the structures of East African kingdoms display some major differences from those found in the west. Even in Buganda, where war was an economic activity, it required little investment of skill and goods;2 no large-scale endow- ment of land and peasant services was needed to support a permanent body of knights. Nor was there the sharp dichotomy between horse-owners and agri- culturalists that existed in the savannahs of West Africa, though the rigid strati- fication that marked Ruanda and other interlacustrine states of central Africa had a similar root in the division between a ruling estate whose economy was based on livestock and a conquered group who worked the land.

I have argued that the social systems of Africa and Eurasia displayed some broad differences which were closely linked to the productive technologies, more especially in the rural sector. In Africa, there was no use of rotary motion or of animal traction; there was no application of machinery to cultivation, to water control, or to transport. Virtually the only machine was the loom.

A recognition of this fact is important not only for historians and for a whole range of cross-cultural studies; it also bears upon contemporary problems. In planning the rate and progress of development we have to take into account the base line from which we start. Too many projections are less than realistic because they assume a similar level of "peasant agriculture" in Europe of i850 and Africa of i96o.3 This is certainly not the case. In Africa the small-scale tech- nology of Eurasia is lacking; at the village level (and 8o per cent of the population still live in rural conditions) there are wood-carvers but few carpenters, iron- workers but no mechanics, potters but no wheels or kilns. The basic craftsmen have often still to be trained.

As a consequence the introduction of farm machinery is necessarily more expensive by way of overheads; inevitably, poorer use is made of tractors and similar tools. But the corollary of these technological limitations is that a great deal can be achieved with much less. Where farming has been done with the hoe, the introduction of the plough can achieve a notable leap forward. Where water has been scooped from pools and corn ground by hand, the use of small diesel pumps and grinding machines can do much to increase pro-

1 See R. Coulbourn, ed. Feudalism in History (Princeton, I 956). For an analysis of rotational systems see my paper 'Circulating Succession among the Gonja' in Svccession to High Office (Cambridge, i966).

2 War brought "women, slaves, cattle and ivory into the elaborate system of gift-exchange centred on the Court".-Chilver, in Richards, ed. op. cit. pp. 386-7.

3 See, for example, A. F. Ewing and S. J. Patel, 'Perspectives for Industrialization in Africa', in Man and Africa, ed. G. Wolstenholme and M. O'Connor (i965).

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions

AFRICAN ECONOMY 405

ductivity. Where loads are carried on the head, the wheel brings a radical change.1 The implications of these facts are not new; they are known to many agricul-

tural officers of the various countries; they have formed the basis of Schumacher's group for the promotion of intermediate (i.e. small-scale) technology. But the implications are too often avoided by modernizing politicians, who look first to tractors rather than to ploughs, to industry rather than to agriculture, as a means of advancing their country's economy. And in this they are encouraged by the American and Soviet models, as well as by the waves of experts from East and West that flood into every corner of the continent, though rarely for long enough to appreciate the differences in the technological base, let alone the full impli- cations of these material differences for the social system as a whole.

In placing this emphasis on differences in technology between Africa and Eurasia, differences that relate to the means of destruction as well as to the means of production, I may seem to be putting forward a crude, materialistic determin- ism. This is not my intention; man's machinery is the product of his own inven- tiveness; improvements in the technology of the intellect (i.e. in communications, such as writing) are an important factor in the economic take-off; and changes of technology are only one factor in the process of social change. A new device, as Lynn White says, merely opens the door. But that this opening up ofpossibilities is the most important part of the process, few can have serious doubt.

The plough, in particular, has had a major effect in Europe where Bloch related differences in land tenure and communal obligations to the distribution of the wheelless Mediterranean plough (aratrum) and the wheeled Germanic plough (charrue) .2 In the absence of wheel, plough, and all the concomitant aspects of the "intermediate technology", Africa was unable to match the deve- lopments in productivity and skill, stratification and specialization, that marked the agrarian societies of early medieval Europe.

St John's College, Cambridge

1 See the letter from the Rev. L. Robertson, Adviser to the Ulongwe Co-operative Society, Mlanje Mission, Malawi, in Bulletin No. i, Intermediate Technology Development Group Ltd, p. 13:

". . . These locally-made carts symbolise our problems in many ways. They represent a revolution- the introduction of the wheel as a means of transporting loads-but it is not seen as significant because it is not modern; it doesn't come into thinking or planning of the leaders of the economy (which is basically western). Further, the cart is too bulky per value to be worth transporting from Europe. It virtually has to be made locally.

"The man who sees its value is the new farmer who is part of the real revolution which is, or should be, taking place. The young leader has no experience of carts, precisely because he has missed out on the 'Intermediate Stage'."

However, this point is now being appreciated by some modernizing politicians, as shown in a recent speech made byJ. K. Nyerere, of Tanzania: "In many parts of the country we are beginning to follow the advice of our agricultural experts. But our major tool, the jembe, is too primitive for our present-day needs. We must now abandon it and replace it with the oxen-plough. We cannot make progress by waiting until every peasant is able to possess his own tractor which he can drive and maintain. Indeed, if we wait for that we shall never leave the hoe behind us, for our present methods are too inefficient ever to produce the wealth which would enable us to buy tractors for all parts of the country, or to train the people to drive and maintain them. We are not ready for the tractor, either financially or technically; but we are ready for the oxen-plough."-'After the Arusha Declaration', speech made on I6 Oct. i967, Dar es Salaam, p. 4. For this reference I am indebted to Basil Davidson.

2 M. Bloch, French Rural History (i966; first French edn I931); L. White, 'Technology and Invention in the Middle Ages', Speculum, xv (I940).

This content downloaded from 140.254.87.103 on Tue, 6 Aug 2013 15:51:49 PM All use subject to JSTOR Terms and Conditions