Feudalism in East Africa
            Buluda. Itandala
            Senior Lecturer                                       Utafiti Vol VIII No.2, 1986, Journal
            Department of History.                          of the Faculty of Arts and Social Sciences,
            University of Dar es Salaam                                     University of Oar es Salaam
        Feudalism has been a subject or'debate among non-Marxist and Marxist
  scholars for a long time. The 'debate has centred no~ only on w.hat it is sup-
  posed to be but also on whether it is a universal stage in man's historY or an
  exclusively medieval west European phenomenon.
        Among the non- Marxist scholars, there are two main interpretations on
 the issue. One school of thought regards it as a group of political and legal
 institutions which regulated the relationship between overlords and their vas-
  sals in medieval western Europe. J In this sense, feudalism is portrayed as be-
 ing mainly "the story of baronial and knightly contracts of service"                          ,z that is,
 a decentralised system of government in which the relationship l:>etween free
 men was governed by a specific set of laws and property rights. This type of
 government is said to have arisen during the second half of the ninth centurY
 as the Frankish Empire of Charlemagne disintegrated.3 The Lord-vassal rela-
 tionship or vassalage, associated with the granting of a fief in return for obe-
 dience and military service, is regarded by the advocates of this school as the
 core institution of feudal society.
        This school of thought does not, however, believe that feudalism was neces-
 sarily bound up with any specific economic system. Its proponents point out,
 for example, that feudal institutions continued to exist even during the late fhir-
 teenth century when the natural economy was being superseded by a fast ex-
 panding money economy and that, instead of being given real estates, vassals
 were then money receiving fiefs.4 This school sharply distinguishes feudalism
 from manorialism too by pointing out that the former was 3; system of political
 and legal relationship involving free men, while the latter was an agrarian sys-
 tem i~volving dependent peasants. 5•
 Furthermore, the advocates of this political-legal interpretation tend to be scep-
 tical about the use of the term feudalism with reference to non-European his-
 tory. In their opihion, as we have already pointed out, feudalism is a specific
 kind of political system which prevailed in western Europe from the mid-ninth
to the thirteenth century. Therefore they consider it in appropriate to use such
 a concept for the analysis of non-European societies ..
    The second non-Marxist interpretation of feudalism is much broader than the
 first one. Its supporters use the term in a general way to describe the dominant
 forms of social, economic, and political organisation which prevailed in medieval
western Europe. In other words, its supporters do not merely conceive of feudal-
ism as the relationship between lords l:\lldtheir vassals or the system of depen-
deii. land tenure but also a way of life centering on lordship. Feudalism in
this sense was many things: it was a political system, an economic system as
well as a system of values. One prominent scholar who has depicted feudalism
• This paper was originally presented at the 13th Annual Confe~nce of the Canadian Association  of African
    Studies held at Laval University in Quebec Citv, Canada, 15th to 19th, 1983.
                                                     29


in this way is Marc Bloch. In a book which has become a classic on the subject,
he summarised the main characteristics of feudal society as he understands it
as follows:
         A subject peasantry; widespread use of the service tenement (Le., the fief) instead
         of a salary, which was out of question; the supremacy of a class of specialized
         warriors; ties of obedience and protection which bind man to man and, within
         the warrior class, assume the distinctive form called vassalage; fragmentation of
         authority .... such then seem to be the fundamental features of European
         feudalism.6
        Those who lean towards this broad definition of feudalism tend to regard
it as a stage in human history which has existed not only in medieval wester n
Europe but also in non-European parts of the world such as Japan, China, In-
dia, Ancient Egypt, Mesopotamia, Byzantium, and Russia, at different times.7
What all of them have done is to take the institutions of medieval western Eu-
rope as their model and then compare them with those of their specific case
studies in the non-European world.
        Marxist scholars, on the other hand, use the term feudalism to identify
a particular epoch in the process of socio-economic development. In other words,
they use it to characterise a particular mode of production in human history.
In this sense, feudalism is regarded as one of the pre-capitalist stages of socio-
economic development; one which immediately preceded the rise of the capitalist
mode of production. One Marxist scholar has defined feudalism as follows:
   • It was a mode of production dominated by land and a natural economy, in which
     neither labour nor the products of labour were commodities. The immediate
     producer - the peasant - was united to the means of production-the           soil-by
     a specific social relationship. The literal formula of this relationship was provid-
     ed by the legal definition of serfdom ... serfs had jurisdically restricted mobility.
     The peasants who occupied and tilled the land were not its owners. Agrarian
     property was privately controlled by a class of feudal lords, who extracted a sur-
     plus from the peasants by politico-legal relations of.compulsion."
Although this definition may not be exhaustive or adequate, it nevertheless in-
dicates what feudalism is as a mode of production. Its essence in the Marxist
sense, as Rodney Hilton has rightly pointed out, "is the exploitative relation-
ship between landowners and subordinated peasants, in which the surplus be-
yond subsistence of the latter, whether in direct labour or in rent in kind or
in money, is transferred under coercive sanction to the former".9 Other main
characteristics of feudalism as a mode of production indicated in the above defi-
nition and whi~h must be emphasised are the fact that the goods produced in
this economic system are mainly for local consumption and not for sale, that
the producers are still attached to the means of production but do not own them,
that the means of production which may be land, a cottage industry, or some-
thing else are privately {)WIled b)'-individual members of the ruling class, and
that production in this type of economy largely depends on family labour and
not on wage labour. This then is how feudalism is generally defined by Marxist
scholars.
       Each of these different approaches to feudalism has been used by differ-
ent writers for analysis pre-colonial African regimes. In the interlacustrine region
of East Africa, for example, the term feudalism, and the vocabulary usually
associated with it, has been used implicictly and explicitly to describe the states
which emerged between the 15th and 19th centuries in Buganda, Busoga,
Bunyoro, Nkore, Rwanda, Burundi, Buha and other areas. Among the writers
who have made indirect reference to the existence of feudalism in these states
                                               30


are Roscoe, Mair, Faller and Oberg,lO and those who have explicitly described,
them as feudal include Beattie, Maquet, Tawney, Hall and Cory, and Grav-
el.lI Beattie in one of his earlier writings on Bunyoro,for instance, described
this kingdom in this way:                                         .
    Although it is changing rapidly,.it still presents many of the characters of a ceu-'
     tralized, "feudal" state, oddly reminiscent in many ways of die feudal kingdoms
     which existed centuries ago in Europe and elsewhere. I use the word "feudal"
     here in its simplest sense, to refer to the kind of political system which is based
     on the relation between a superior and his iDferior or vassal, where the latter holds
     lands, and authority over the people living on these lands, "in feud" from the
     former. This means that the vassal must render homage and services of various
     kinds ... to the superior lord from whom he holds his lands and authori\y. Tradi-
     tional Bunyoro has many features i.n common with such a system .... 1
These references, which have been made mainly by social anthropologists, have
tried to compare the social and political institutions of these kingdoms with
those of medieval western Europe as the above quotation clearly shows. In some
of the kingdoms, it is the existence of a clientage system based on cattle owner-
ship which these different writers regarded as the basis for feudal relations, while
in others it is the existence of a patronage system based on land ownership.
        Such analogies, how~ver, have attracted a lot of criticism for those who
adhere to the political-legal interpretation of feudalism. Their contention, as
we have already explained, is that feudalism was essentially a medieval west
European phenomenon, and for that reason, it should not be applied to Africa
and other parts of the non-European world. This is the position which Jack
Goody has taken. According to him, most of the studies which have described
some pre-colonial African states as feudal have done so on the basis of superfi-
cial resemblances between their soci;ti and political institutions and those of
medieval European states.13 Hence he concludes that the term. feudalism and
the vocabulary associated with it should not be applied at all in describing them
because it has little or no meaning in their case.
        Even those who support Marc Bloch's broad definition and accept the use
of the term for comparative purposes have tended to be sceptical about its use
in the analysis of pre-colonial African societies. This has been the case because
they have tended to regard only certain aspects of the medieval west European
social formations as the determining factors of feudalis,m, Paradoxically, one
of these sceptics was Beattie who in the 1950's had explicitly described Bunyoro
as a feudal state. After the publication of Goody's ariicle on feudalism in Afri-.
ca in 1963, however, he seems to have changed his rrVnd completely on the is-
sue and decided to support the former's position. His new conclusion was based
on the fact that when Bunyoro's political iQstitutions are tested against Bloch's
essentials of fel,ldalism, they fail to match two of the most important of them,
namely the existence of a specialised military class and the existence of a decen-
tralised state system. I. He regards the last named characteristic as being the
most important apparently because European feudalism in its beginnings im-
plied the weakening of the state and the distribution of powers formerly held
by the central authority among private individuals, while in Bunyoro and other
African states there has been no such weakening or breakdown. As a matter
of fact, many, if not most, of the policies which have been described as feudal
in Africa represent, in hi~ opinion. ~ trend towards centralisation of power.
     Another important point of difference which Beattie now stresses between
the European feudal states apd those of tl:te Interlacustrine Region "relates to
                                              31


 the question of which categories of persons were linked as subordinate and su-
 perordinate. In Europe the characteristic feudal bond was between the tenant'
 or fief-holder and a nearby chief or lord, from whom he held his fief and to
 whom he looked for protection; there was no direct link between the fief-holder
 and the king" at the topY In Bunyoro and other similar states, however, vir-
 tually all political authority was vested in the omukama or king, and the peo-
 ple could appeal directly to him. Moreover ,even the chiefs who,were appointed
 by the omukama to administer regions of the kingdom were in those areas as
 admitjistrators,and not as fief-holders or landlords, and the tribute which they
 received from the people was a form of taxation and not a share from their
 estates as lords. He suggests also that the chiefs in the regions were there in
 order to strengthen the position of the central government vis-a-vis the local
 clan communities which stood in opposition to it. So, as far as Beattie is con-
 cerned, all these factors show clearly that Bunyoro was not a feudal state despite
 the fact that it had several of Bloch's essentials of feudalism in common with
 the European feudal states.
       Another scholar who has rejected the existence of feudalism in the Inter-
 lacustrine States during the pre-colonial period is E.M. Chilver. Her premise.
 is that the most important features of European feudalism, namely feudo-
 vassalage and the dispersal of political power, were absent. in the Interlacus-
 trine States.16 She elaborates this by pointing out that, unlike the medieval Eu-
 ropean states which arose from the ruins of larger highly centralised political
 anits, which were linked by a universal church, the Interlacustrine States did
 not arise out of a similar situation. In fact, the historical situation from which
 the latter, and other African states which have been described as feudal, arose
 did not lead to decentralisation of politial power as was the case in medieval
 western Europe. Therefore, unlike in medieval western Europe, where the feu-
 dal states were confederations or paramountancies,        the lnterlacustrine King-
 doms and other African states which have been referred to as feudal were highly
 centralised whether they were big or small.
       Having shown that the dispersal of political power in the medieval west
 European sense was absent in the Interlacustrine Kingdoms, Chilver, like Beat-
 tie, critici~ those who have e<luated the ownershi~ of estates or portions of ~and
 by. appointed chiefs to fief-holding. She has rejected this equation because the
 chiefs were given their estates by the kings without entering into legal bonds
 of dependence with them as lords and their vassals as in Europe. A similar view
 has been taken by Edward Steinhart who regards feudo-vassalage as the essence'
 of feudalism.17 Steinhart explains that the chiefs in states such as Buganda and
.B~yoro did not offer themselves as vassals; they were nominated officials whose
 prune function was to exercise delegated authority in their areas. Therefore there
 was no fendo-vassalage here in the medieval west European sense, which in turn
  means that feudalism did not exist either.
       Professor Samwiri Karugire is another scholar who haS participated in this
 debate on feudalism in the Interlacustrine Region. Writing specifically on Nkore,
 he ~as categorically stated that pre-colonial Nkore was not a feudal society in
 which the Bahima rulers were the feudal lords and the Bairu agriculturalists
 were the serfs. "Feudalism", he says, "is not an economic system, but a form
 of government. That being so. an analysis of Nkore's traditional institutions
 of government tends to show that Nkore was not even remotely feudal, if by:
 !he expr~ssion one means, a form of government in which political authority
 IS mopohze~ by a small group of military leaders but evenly distributed       among
                                          32


the members of that group, and in which the k'ing, at best, can merely keep
peace among the lords and usually is unable even to do this. ,,18 To begin with,
he argues, there is nothing in Nkore's history to show that a group of chiefs
or individuals ever challenged the authority of the Mugabe or king. ;Secondly,
neither the mugabe nor his chiefs could stop anyone from moving in and out
of the country at will. Moreover, although he admits that Nkore was a class
society in which the Bahima pastoralists constituted the upper class while the
Bairn agriculturalists constituted the lower class, he does not think that the form-
er exploited the latter economically. Actually, he maintains that the economic
relations between the Bahima and the Bairu were based on mutual"exchange
of the goods produced by each group. Even the tribute which the mugabe and
his chiefs appropriated from their subjects is said to have been presents given
in exchange for other benefits such as assistance in times of hardship and pro-
tection against internal and external enemies.19 In view of all thIS,to speak of
feudalism in the context of Nkore, according to Karugire, seems to 1gea mis-
taken superimposition of the classical patterns of European feudalism on an
African tribal structure.
      It is evident, from their own arguments, that all these scholars view feudal-
ism as a set of certain social and political institutions and not as an economic
system in which people belonging to different classes play different roles. I think
they do this because they have no conception of feudalism as a mode of produc-
tion. That is to say, they fail to distinguish between the feudal mode of produc-
tion and the social formations which constitute it as Mamdani has correctly
observed.20This lack of understanding of the difference between the two applies
 also to'the otber non-Marxist scholars who have asserted that feudalism exist-
 ed in some of the Interlacustrine Kingdoms on the basis of similarities between
 their social and political institutions and those of medieval European states.
 Feudalism as a mode of production can exist across time and space. In other
 words, it can exist in different places or countries at different,times. The feu-
 dal social formations, on the other hand, cannot be expected to be the same
 or uniform throughout the world because each one of them is a product of specif-
 ic historical and prevailing material conditions. For this reason, they are bound
 to differ from country to country. Even in medieval western Europe they differed
 from one country to another.21 Hence to deny the existerrceof feudalism in the
 Interlacustrine States or in any other states on account of the differences in
 social and political institutions between them and those of medieval European
 states, as these scholarshav~done, is definitely wrong.
      Like their non-Marxist counterparts, Marxist scholars have differed too
intheir use of the term feudalism with regard to Africa and other non-European
areas. First there are those who believe in what has been referred to as "the
five member scheme,,22or "the traditional unilinear scheme", 23 which is a
model of development founded on the five modes of production schemesketched
out by Marx and his close associates to show the historical stages of develop-
ment through which Europe has passed up to the twentieth century. This Euro-
pean model of successive stages of development from primitive communalism
through slavery, feudalism, capitalism and socialism or communism has been
taken by orthodox Marxists sincethe 1930'sto be the general path through which
man kind everywhere must historically pass.24• It is thus taken as an obligato-
ry schedule of development for all mankind. Those who subscribed to it con-
sider feudalism to be a necessary universal stage in human historical
                                          33


 development. Such scholars have therefore been able to apply the concept of .
 a feudal mode to the analysis of some pre-colonial African state forms without
 any hesitation. A good example is Potekhin.2S
     The unilinear model, however, has been found unsuitable by some Marxists
 for the analysis of many non-European social formations apparently because
 it tends to oversimplify the Marxian theory of development and is a mechanis-
 tic way of viewing development. Adherence to it meant also a rejection or a
 total disregard of the Asiatic Mode of Production formulated by Marx after
 discovering that nineteenth c~ntury Asian social formations could not be easily
 int~ated       into ,the European model. The essence of this mode, according to
 Godelier, is "the existence of primitive communities in which ownership of land
 is communal and which are still partlyorganisedon                 the basis of kinship rela-
 tions, combined with the existence of state power, which expresses the real or
 imaginary unity of these communities, controls the use of essential economic
 resources, and directly appropriates part of the labour and production of the'
 communities which it .dominates" .26 This mode of production differs from the
 feudal mode in two main respects. One of the differences between the two is
 that the means of production in the former are owned by the state and not by
 private individuals as is the case in the latter. Secondly, in the Asiatic Mode
 of Production the state mobilises the producers to engage in major economic
 undertakings and exploits them directly by appropriating part of their labour
 and produce, while in the feudal mode a class of independent small producers
 is exploited by a class of independent owners of the means of production.
         ~hen some Marxists started questioning the European-centred                  develop-
 ment model, particularly since the 1960's, they reasserted also the Asiatic Mode
,of Production, a concept which had originally been devised for the analysis of
 Middle Eastern and South Asian societies. But its use was now. extended to
 alltypes of early class societies, including those of Africa, mainly by French
'Marxist scholars working at the Centre d'Etudes et de Recherches Marxistes
 in Paris such as Maurice Godelier, Jean Suret-Canale, and Jean Chesneaux.27 •
.suret-Canale, for example, is reported to have stated that "A.fter having our-
  s"elvesspOken of African 'feudal societies' we feel compelled to abandon the
  term. The mode of production in traditional Africa is never 'feudal' and'to
  use the word in a purely political sense may cause confusion". 28 Instead, he
  advocated that "the most common mode of production in the more advanced
  areas of traditional Black Africa can be compared to the Asiatic mode". 29
  Godelier was even more eXDlicit on this issue:                                        ..
     In our view, this hypothesis allows us to understand 'Yhy t~e concept o~ ASIatic
      mode of production is referred to more and .more to Illurmnate the ,Penods a;nd
      societies of ancient Europe ... for Black Afnca ... and "r~-ColUJI.l1?Ian Amenca
     ... A common. element appears throughout these many speCIfic realIties - a com-
      mon structure which combines community relations an~ the embryo of cl~s and
      is related to the same situation of transition to class socIety. Becuase of this rela-
       tion between the situation it is possible to explain the geographical and historical
     universality of the form of social organisation which emerges when the conditions
     for the transition to class society develop; maybe at the end of the fourth milleni-
     urn B.C. in the case of Egypt .... or in the nineteenth century with the birth of
     the     Bamoun      Kingdom       in the     Cameroons.       By providing      many
     examples      of societies   in the process _of transition          to class oq~ani-
     sation, the archaeological and ethnological knowledge accumulated since the
     nineteenth century provides the concept with a field of application which Marx
     or Engels could not have envisaged. In becoming more and more widely applica-
     ble both in time and space, the concept n~ longer applies exclusively to ,.,.
                                                 34


'Thus the Asiatic Mode ofJSroduction was elevated to a Universal mode' of de-
 velopment covering not only class societies in the Middle East and Southeast
 Asia but also in pre-Columbian American and pre-colonial Africa.
        But it is now evident that some of the scholars who originally advocated
 the universality and necessity of the Asiatic Mode have discovered that they
 were making the same mistake as the supporters of the Euro-centric unilinear
 model. Melotti intimates that the main problem with them is the fact that "they
 ignore the historical and geographical peculiarities of the Asiatic mode of produc-
 tion ... and arbitrarily elevate it to a universal 'stage' in a modelof historical
 development still seen as a single path" .310ne of the first to realise this mis-
 take was Chesneaux who is reported to have, retorted that "there is no question
 of replacing the dogma of the universality of slavery and feudalism with a new
 dogma of the universality of the Asiatic mode of production"32. This implies
 that pre-colonial African states and those of pre-Columbian America were based
 on a structure which was entirely different from oriental despotism and there-
 fore cannot be included in the Asiatic mode. A similar position has been taken
 by Catherine Coquery-Vidrovitch who, along with others at the Centre d'E-
 tudes et de Recherches Marxistes in Paris, has introduced the concept of an
  'African Mode of Production' .33 This African Mode of Production is supposed
 to deal with the specific state forms of this continent which, according to her,
 cannot be ,adequately analysed under the Asiatic Mode of Production. She ar-
 gues against including the African state forms under the Asiatic Mode of Produc-
,tion because the only thing common between them and Asian state forms was
 the existence of self-sustaining village communities. Otherwise the two state sys-
 tems differed fundamentally from each other because there existed a bureaucratic
 regime with absolute power in tQe latter which directly exploited. the subject
 communities under it through major production works, while in the former
 abureaucracy without despotic power exploited the subject peasantry under it
 only indirectly by appropriating'surplus     labour or produce.34 In other words,
 she disagrees with Godelier's proposition that there could be an Asiatic Mode
 of Production "with major works" and despotism or true AMP, which existed
 in Asia itself, and another one 'without major works" and despotism, which
 existed in Africa and elsewhere.3s. Such ll. proposition, in her opinion, not only
 deprives the Asiatic Mode of Production of its dynamic element by eliminating
 its,economic basis. of major works but also denies the African states their speci-
 ficity and dynamic elements ..                                               -
         The implications of these arguments are very progressive in themselves
 because; as Rodney says', they are concerned with the concrete conditions of
 Africa rather than with pre-conceptions of development brought over from Eu-
 rope.36 But the scholars concerned seem to be bent on trying to find a single
 concept to cover the analysis of a variety of social formations which existed
 in Africa from the beginning of the iron age to the eve of European colonial
 rule. This may lead to difficulties should some of the social formations fail to
 fit into the proposed ,African Mode.
       Rather than be rigid on this issue of modes of production in the African
 situation, flexibility seems to be the way out. It is now evident that the modes
 of production outlined for Europe were not reproduced in Africa and other
 hon-European areas in the same sequence. This does not, however, mean that
 they were not reproduced at all in Africa, for there are probably cases where
 either s~ave or feudal relations of production or a combination of both foIIo'Yed
                                           35


    the disintegration of communalism. It is also possible that there were some so-
    cial formQ.tions which could be fitted either in the Asiatic Mode of Production
    (e.g., Ancient Egypt) or in the proposed African Mode of Production (e.g.,
    the Sudanic Empires of West Africa). We are, in other words, not arguing
    against the possibility of having an African Mode of Production. Weare sim-
    ply suggesting that, given the existence of a large variety of social formations
   .following the collapse of communalism in Mrica, there might have been some
    that could be adequately included in each of the modes mentioned.
          The Interlacustrine Kingdoms of East Africa, for example, do not fit into
    the Asiatic Mode of Production because they were not based on major produc-
   'tion works. Nor do they fit into the proposed African Mode of Production be-
   'cause they were not based on long-distance trade as the ideal type state forms
    of this mode were - the Sudan Empires of West Africa. What seems to have
    happened is that the social formations which emerged in this region following
    the collapse of communal relations of production acquired feudal relations of
    production. This means that the state which emerged in the region between the
     15th and 19th centuries did not go through a slave moCle of production before'
    acquiring feudal characteristics as was the case in medieval western Europe.
          In order to understand how these Ind:rlacustrine Kingdoms became feudal,
     we must first examine how some of their social formation developed after
    the 15th century. And in order to do this, it is useful to divide the region into
    two economic zones, namely the banana culture zone and the cattle culture zone.
    The former included. among others, the kingdoms of Buganda, Busoga and
  :Buhaya, and the latter included t~e ~gdoms           pf BUJltIOro, Nlcore, Karagwe,
-,Mpororo, Buhweju, Rwanda, Burundi and Buha. In the banana culture zone, the
    basis of prOduction relations was land on which the plantain banana, which
    was the main food crop, could thrive. This banana economy forced people to
    live in permanent settlements because the banana was a perennial crop and
    thrived only in certain areas of the zone. In other words, the plantain banana
    economy prevented people from changing their settlements from time to time
    as was the case in pastoral and grain-producing areas. As a result, this type
    of ec.onomy led to the development of relations of production in which one
    group or part of the population became dominant because it seized political
    power and control of most of the productive land and the other became subor-
    dinate because it lost control of both.
          One area which shows clearly how these relations of production      came   into
    being in the banana zone is Buganda. Buganda emerged as a state in about the
    15th century consisting of a number of clans which were still very powerful and
    owned the land in their respective areas. In subsequent centuries, however, a
    vigorous struggle for ;power ensued: between the kahoka (king) and the various
    clan leaders or ahotaka (chiefs). Eventually the kahoka was able to reduce the
   power of the hereditary ahotaka chiefs in the regions of his kingdom by ap-
   pointing people of his own choice as regional. heads who became known as
   abakungu chiefs. This means that, instead of appointing clan leaders as regional
   heads, he now appointed any individuals whom he thought would be loyal to
   him. These appointed ahokungu chiefs were rewarded with estates of produc-
  tive land by the kahoka as he gained control of it from the clan leaders. The
   second step which the kahoka took in order to consolidate his position was to
  .eliminate competition from princes of the blood known as aho/ang;ra.37 In-
   ~       of giving them political' of~ces, h.e gave them estates of prfX.Iuctive land.
                                             36


 Thirdly .. the kiI~      &ave out 1arid to his favowites and supporters in exchatJ.Bc
  fC?rraising ~,        supplying tw:k-cloth and pqf9I'}lling ritual se!"'1ces'. Estates
  gIVento IQyal S9,pporters of the ~      as rewards for service were known as ebitOn-
 gole and their owners as abato~o/e. 31 All these land owners, including cl.(li1
 leaders who retained control of portions of land in their areas, had full contt'ol
 of their estates and the peasants who were living in them. What this means is
 that, as the kabaka consolidated his political position after the 15th centurY,
 a land-owning class "of chiefs or abami Consisting of the kabaka, the abakun ....
 ~, aba/angira, abatongo/e and abataka ;elQjergedwhich depended for its sub-
 siStence on the subject peasantry known as abakopi.39 This land-owning ruliJl8
 class lived partly on what was produced by the abakopi in their estates.It alSO
 received tribute annually from peasants living in Clan-owned hind in the for:t1l
 of agriculturalWoducts,        crafts and services ..
       Similar developments took place in the Buhaya states in which the abakO-
 ma seized control of m~st of the banana-producing land apd divided it up into
 large estates known as nyarubanja. The nyarubanja estates were then given
 by the abakama to people whQm they appointed as abakungu chiefs and to -their
 relatives and friends as rewards for their service and loyalty ..co The landlords
.of the nyarubanja estates became known as abatwazi and the tenants in thetIl
 as abatwarwa. When clan lands were seized by the abakama and given to the
 abakungu and other supporters, their former owners became tenants andclient5
 of the new owners - the abatwazi. Through the creation of these estates, the
 abakama of Buhaya were ultimately able to break up clan-held lands so that
'a fragmentary land tenure pattern emerged on which the new.relations of produc-
 tion were based. The development of these new relations of production has been
 described by Schmidt:
        The development of nyarubanja was a device employed by the &kama           to frag-
        ment clan land and to undermine the social unity of the clans. The creation of
        a nyarubanja estate out of clan-held land inevitably led to the creation of a neW
        loyalty. No longer was a clansman directly answerable to the elders (Bogurusi).
        but his primary obligation as a tenant lay to his immediate superior, his Mutwa-
        zi.,. Furthermore, the landlord as superior and as patron had judicial rights over
        disputes within his limited domain .•. 41
 As tenant and clients, the people living in nyarubanja estates stopped paying
 tribute directly to the abakama. They now I?rovided goods and semces to their
 landlords, the abatwazi. In this way, the vanous officers of each omukama were
 maintained in office. The abakama themselves continued to receive tribute from
 peas@ts who were still living on clan-owned land.
       From these two examples it can be seen clearly that relations of produc-
 tion based on control of land developed in the banana zone afer the formation
.of states. The J?C&l!ants,be they abOkopi, abatwarwa or those living in clan-
 held land,remamed attached to the lana as before but they w~re now subject
 to the domination of landlords wh~ extracted goods and seI'V!cel!from t-!tem
 in return for some form of protection. In other wordsz explOlt~tive relatIons
 developed between the land-controlling ruling class ana the subject peasantry
 in which the surplus beyond the subsistence of the latter was transferred to the
.former under coercive sanctions in return for some form of security. Goods in
 this economy were produced by individual families mainly. for th~ satisfaction
 of local consumption needs and not for sale or I>rofit-m~king: This m~s that
 even the tribute paid to the ruling class was in kind and It~ prunary object was
 consumption not accumulation. The existence of an a~cUltural surpll}s and,
 the expansion of regional trade in the eighteenth and nmet~nth centunes led
 to the formation of a ~oup of artisans patronised by the kIOgs. But even the
                                            37


.development of trade with the east coast during the nineteenth century did not.
 change the basic relations of production in the zone because all major business
 transactions were monopolised by the kings. It only provided more surplus and
 luxury goods for the ruling class. In Buganda, for instance the central authori-
 ty was' enhanced by the imJ?C?rtationof guns from the coast. Since only the kaba-
  ka's army was equipped with firearms, he was able to use them to exercise more
  control over the population, which in turn led to a more efficient collection
'of surplus from within and from without. Therefore, whichever way one looks
  at it, the dominant mode of production which prevailed in this zone during the
,eighteenth and nine-teenth centuries was without doubt, ~eudal.
     Unlike in the banana zone. where land was used as the chief means of estab-
  lishing control over the population by the ruling class, it was cattle which was
 used for the purpose iri the western part of the Interlacustrine Region. Here,
  unlike in the-former, where agriculture predominated, there was a dual econo-
'my of pastoralism and a~riculture. Each of these sectors of the economy was
 ;the monopoly of one SOCialgroup - pastoralism was carried out by the Bahi-
 ,ma in Bunyoro, Nkore, Mpororo, Buhweju and Karagwe and by the Batutsi
  in Rwanda, Burundi and Buha; agriculture was the monopoly of the Bairu in
  the former group of states and of the Bahutu in the latter. But besides being
  cattle-owners the Bahima/Batutsi, formed the ruling class and the Bairu/Bahutu
  the subject peasantry in their respective areas.
         Like their counterparts in the banana zone, the kings in the cattle zone
  appointed their favourites to senior administrative and military positions in order
  to consolidate their own political power. All senior chiefs and military com-
  manders were recruited exclusively from the Bahima/Batutsi socio-economic
  $roup. In Nkore, for example, the Muhima king or mugabe consolidated his
  political position by appointing some of his supporters, friends and relatives
  as regional military leaners known as emitwe and as civilian administrative offi-
  cials and tribute coUectors. all of whom were ~enerallv referred to as abakun-
  gu. In order to make them entirely dependent on him. he gave them cattle. He
  was able to do this bacause he was theoretically the owner of all cattle and land
  in his kingdom. Having given them political positions and cattle, these appointed
  officials were ex{)CCtedto serve the mugabe faithfully and had to demonstrate'
  their loyalty to him by attending his court as often as possible and by occasion-
  ally givmg him "presents" of cattle, lion or leopard skins, hone~ or beer. This
  giving of 'presents' to one's superiors was known as okutoija. It applied to
  ordinary people as well as officials of the mugabe and it was not a voluntary
  act as the word 'presents' seems to suggest because as Karugire explains. "It
  was essential for every head of family to keep himself'visible' to the authori-
  ties" in order to secure more reliable protection against sudden hardships and
  internal and external enemies. This means that it was an obligation and not
  something which people could choose to do or not to do without facing serious
  consequences. Tms applied to the officials of the mugabe too, for it anyone
 of them 'stopped- giving these gifts or attending atcourt for a lengthy perio<~,
  it was usually assumed that he had 'rebelled' and, in such a case.l he lost hiS
  position and, often, his property at the orders of the Mugabe." 3
         Thus the king in Nkore, and elsewhere in this _zone, was a giver and a
  receiver of ~oods. He distributed benefices and received more goods and serv-,
  ices from hiS subordinate officials and subjects. This means that he practised
  what is known as patronage, the giving of favours and privileges to one's subor-
  dinates in exchange for service and loyalty. This system is sometimes referred
  to as clientage. But it was not only the king who used cattle as a means of es-
  tablishing patronage in Nkore and in the other kingdoms of the cattle zone.
  All people who owned cattle used it for acquiring client. Officials of the mugabe
  at different levels ~ave cattle to their juniors as rewards in exchange for loyalty
 just as he did at the senior level.. They also loaned cattle to poor people, both
                                             38


  BaJ:.1im<;t and Bairu,. in exchan~e for service. Ordinary cattle-owners acquired
  then chents by l~anmg. and givmg cattle tq poor people in exchange for service
  and other benef!ts. ThIs means that the clIentage system was used by all cattle
  owners to expl9It n0!l-catt~e-owners, especially the agriculturalists. This clien-
  tage. system eXIsted I!l vanous forms throughout the western Interlacustrine
  RegIOn and reached Its most elaborate form in Rwanda where it was known
  as buhake# Apparently, the buhakewas so wide spread and so elaborate that
  th~ onl~ person 10 Rwanda who ha.d no patron or lord above him was the mwa-
  ml or kmg, and th~ only people WItJ:.10Ut      clients were the poor or cattleless Ba-
  hutu or Batwa. GIven the ~ay socIety was organised, It was imperative for
  everyone to have someone nch and powerful as a protector and provider of
  cattle. "To live without a lord", as Maquet has pointed out "was to invite
 trouble" .45 because when faced with a lawsuit, a famine or so~e othcr misfor-
,tune 0!le could b~ quite helpless. But to have one was equally cumbersome be-
  cause It meant being subservient to him throughout one's life. It also meant
,p'rovidin~ labour,military.service,crafts,      agricultural products and so on from
 tIme to tIme.
       Along side the buhake, okutoi}a or ubugabire in Buha and Burundi,46 was
 the state'administrative     machinery which was also used for the appropriation
 of goods and services from the dominated class. In the case of Rwanda, every
,di~trict had a land-chief and a cattle-chief for administrative purposes. The land-
'chlef controlled the agricultural population and was responsible for the collec-
 tion of tribute .in agricultural products and r~quired labo.u~ serv~ce from every
 household, while the cattle-chIef was responSIble for admmistratIOn among the
 Batutsi pastoralists and for collecting jars of milk from them for the state. All
 administrative officials involved in tnbute collection at hill and district levels
,retained a portion of the goods and services for their own use and forwarded
 the rest to court or wherever the king wanted them to be sent. Thus the ..king
 and his officials exacted tribute in the form of goods and services from both
 peasants and common cattle-owners.
        As a result of the development of political institutions and a clientage sys-
 tem based on the control of cattle by one social group, two classes emerged
 in this zone, a cattle-owning Bahima/Batutsi aristocracy and a Bairu/Bahutu
 subject peasantry. This was not simply a division of labour between pastoralists
 and agriculturalIsts with status distmction between them; it was a class division
 in which the pastoral aristocracy appropriated a .large share of the surplus
 produced by the agricultualists in the form of tribute. 'Ethnic' distinction be-
 tween the two groups led to the development of a more rigid class system here
 than that which emer~ed in the banana zone. In fact, it almost became a cas
 te system, a factor WhIChreduced the possibility of assimilating members from
the subejct ~asantry into the ruling class.
      The origins of the system of social inequality as well as the political imbalance
     must have had their roots in economics, and cattle caused it. The sense of ine-
     quality was s6mething that developed gradually from social contacts. It was
     promoted by the distribution of goods highly prized in society. The distribution
     differences were themselves the result of an earlier acquaintance with cattle by
    .one group. The prestige that goes with property ~ily spr~d to the physical f~a-
     tures which were then held to be proof of supenor qualItIes of anatomy. With
     the use of hindsight the theory and practice of inequality one sees as a clever
     manipulation of a ruling class to perpetu!1te their hol~on political powt:r. 47, '
Thus the ownership Ol cattle by the BahIma/Batu~sI playe? a very sIgmflc~nl
role in determining their relationship with the agncultura~Ists. Cattle was 1m.
portant because its products such as milk, ghee,. meat,. skmsand manure werl'
required by the agriculturalists too. It is, therefore, eVIdent. that by t~<: ~ve 01
colonial rule there existed in the cattle zone a Soclo-economlc and political ,\-
tem in which one class controllIed both politics and the major meam of produ,
                                             39


 tion, land and cattle, and in which the dominated class, which was still attached
 to the means of production, laboured for itself and for the subsistence of the
 former .
     . It would th~refore appear th8;t con~rary to t~e ~enial by Beattie, Karugire,
 Chtlver and SteInhart that feudalIsm dId not eXIstIn the precolonial states of
 the Interlacustrine Region of East Africa on the basis of lack of similarities
 between their social and political institutions and those of medieval western Eu-
 rope, the relations of production and the nature of exploitation which prevailed
 indicate clearly that they were feudal. In other words, when one examines the
 way the means of production were controlled, how the goods were produced
 and the purpose for their production, how the producers were related to the
means of production, and how the surplus was appropriated in the different
 examples discussed, one is bound to conclude that these societies were feudal
in the Marxist senseof the term. It is my considered opinion that the concept
of feudalism is more appropriate for the analysis of these states than either the
concept of the Asiatic or the African Mode of Production.
                                          NOTES
 1. F. Pollock and F. W. Maitland, The History of English Law, University Press, Cam-
     bridge, 1898, VoL n, p. 66-67; F.L. Ganshot (trans. P. Grierson), Feudalism, Long-
     man, London, 1952, passim; J. Strayer, "Feudalism in Western Europe", in R.
     Coulbom (ed.), Feudalism in History, Archon Books, Hamden, Connecti~ut, 1965,
     pp. 15-26; P. Vinogradoff, English Society in the Eleventh Century, Clarendon
     Press, Oxford, 1908, pp. 39-89, 208-304.
2 . M.M. Postan, foreword, in M. Bloch (trans. L.A. Manyon), Feudal Society, Rout-
     ledge and K. Paul Ltd. London, 1962, p. xiii.
3. G. Fourquin (trans. Iris and A.L. Lytton Sells), Lordship and Feudalism in the Middle
     Ages, Allen and Unwin Ltd., London, 1970, pp" 19-29, 39-60.
4. Ibid. p. 70.
5. The proposition that manorialism was an agrarian system involving dependent
     peasants is supported by many writers. See, for example, Bloch, op. cit., pp.
     242-248; Vinogradodd, opp. cit., pp. 305.369.
6. Bloch.....op. cit., p. 446 ..
7. One classic example is the Coulborn symposium, which includes a comparison of
     feudalism in different parts of the world .. See Coulborn, op. cit. passim: see alsoJ.
     S. Critchley, Feudalism, Allen and UnwIn,win, London, 1978, passim.
8. P. Anderson, Passages from Antiquity to Feudalism, NBL, London, 1974, p. 147.
9. R. Hilton (ed.), The Transition from Feudalism to Capitalism, Verso, London, 1978,
    p.30 ..
10. J. Roscoe, The Banyankole, University Press, Cambridge, 1923, pp. 14-18; L.P.
    Mair, "Clientship in East Africa". Cahier d'Etude Africaine (hereinafter, CEA),
    II, 6 (1961) pp. 315-25, L.A. Fallers, Bantu Bureaucracy, CUP, Cambridge, 1956,
    p. 173, K. Oberg, "The Kingdom of Ankole in Uganda", in Fortes and Evans -
    Pritchard (eds.), African Political Systems, London, 1966, pp. 126-34 ..
11. J. Beattie, "A Further Note on the Kibanja System of Land Tenure in Bunyoro,
    Uganda", Journal of AjricanAdministration,        VI, 4 (1954), pp. 178-179; J.J. Ma-
    quet, The Premise of Inequality in Ruanda, IAI, London, 1961, pp. 129-142; J.J.
    Tawney, "Ugabire: a Feudal Custom among the Waha", Tanganyika (Tangania)
    Notes and Records, (hereinafter, TNR), No. 17 (1944), pp. 6-9, R. de Z. Hall and
    Cory, "A Study of Land Tenure in Bugufi, 1925-1944", TNR, No. 24 (1947), pp.
    35-36 and 39-43; P .B. Grvel, "Life on the Manor in Gisaka (Rwanda)", Journal
    of African History (hereinafter, JAH), VI, 3 (1965), pp. 323-331.                    .
12. J. Beattie, Bunyoro; an African Kingdom, Holt and Co., New York, 1960, p.5.
                                           40.


  13. J. Goody, :'Fe?dalis~ in Africa", JAHIV, 1 (1963), pp. 9-13; idem: "Economy
       and Feudalism In Afnca", The Economic History Review, 2nd series, XXII, 3 {1969),
       pp. 393-405.
  14. J. ,Beattie, "Bunyoro: an African Feudality?" JAH, V, I (1964), p. 26.
  15. Ibid., pp. 26-27.
  16. E.M. Chilver, "Feudalism in the Interlacustrine Kingdoms", in A.I. Richards (ed.),
 ,     East African Chiefs, Faber, London, 1959, pp. 378-81.
  17. E.I. Steinhard, "Vassal and Fiefin Three Lacustrine Kingdoms" CEA VII 28(1967)
       pp. 606-617.                                                  '      "               ,
  18. S.R. Karugire, A History of the Kingdom of Nkore in Western Uganda, Clarendon
       Press, Oxford, 1971, p. 67.
  19. Ibid., p. 66
  20. M. Mamdani, Politics and ClassFormation in Uganda, Monthly Review Press, New
      York, 1976, footnote on pp. 23-25.
  21. Anderson, op. cit., pp. 154-172; Strayer in Coulborn, op. cit., pp. 20-25; Bloch,
      op. cit., pp. 176-189.
 22. A. Agh, "Development Paths in Africa," Southern African Universities Confer-
      ences, Dar es Salaam, 1979, Vol. I, pp. I, 50-55.
 23. U. Melotti (trans. P. Ransford), Marxand the Third World, Macmillan, London,
       1977, pp. 8-12.
 24. J.V. Stalin, Dialectical Materialsm Internal Publishers, New York, 1975, pp. 34-46;
      Anonymous, ABC of Dialectical and Historical Materialism, Progressive Publish-
      ers, Moscow, 1976, pp. 282-296 and 301-320, and M. Cornforth, DialecticalMateri-
      alsm, Lawrence and Wishard, London, 1962, Vol. II, pp. 47-54.
 25. 1.1. Potekhin, 'On Feudalism of Ashanti", A paper read to the Twenty-Fifth Inter-
      national Congress of Orientalists, Moscow, 1980.
 26. M. Godelier, "The Concept of the 'Asiatic Mode of Production' and Marxist Models
      of Social Evolution" in D. Seddon (ed.), Relations of Proquction, F. Cass, Lon-
      don, 1978, p. 22.
 27. Melotti, op. cit., pp. 14-17; Godelier in Seddon, op. cit., pp. 209-251.
 28. Surer-Canale, quoted in Melotti, opp. cit., p. 163, note 8.
 29. Ibid. p. 202, note 32.
 30. Godelier, in Seddon, op. cit., p. 241.
'31. Melotti, op. cU., p. 15.
'32. Ibid., p. 165, note 25.
 33. C. Coquery- Vidrovitch, "Research on an African Mode of Production", in Seddon,
      Op. cit., pp. 261-285.
 34. Ibid., pp. 268-270.
 35. Godelier in Seddon, op. cit., p. 242 .
.36. W. Rodney, How Europe Underdeveloped Africa, Tanzania Publishing House, Dar
      es Salaam, 1973, p. 46.
 37. Richard, op. cit., pp. 44-47, L.P. Mair, "BugandaLand    Tenure", Africa, VI (1933),
      pp. 189-190. ,                    .'
 38. Richards op. Clt., pp. 50-51; MalT, op. Clt., pp. 190-198.
 39. L.P. Mair, African Societies, CUP, Cambridge, 1974, pp. 183-198 ..
 40. P.R. Schmidt, Historical Archaeology, Greenwood Press, Westport, Connecticut,
      1978, pp. 28-33.
 41.,Ibid.., pp. 32-33.
 42. Karugire,pp. cU., p. 64, Oberg in Fortes and Eva~s-Pritchard,   l}p.,~it.!pp. 128-130.
      See also S. R. Baitwababo, 'Foundations of RUJumbura 5<?cIety , In D. Denoon
      (ed.), A HistoT." of Kigezi in South-West Uganda, The NatIonal Trust, Kampala,
      n.d. (19705), pp. 87-88.
 43. Karugire. op. cit., p. 65 ..                                              42 160 172
 44. For'll detailed discussion on buhake, see Maquet, op. 'Clt., pp. 129-1,          -    .
 45. Ibid,. p. 13.1.
                                              41


46. For a discussion on ubugabire, see Tawnye, op. cU., pp. 6-9: G.C.K. Gwassa ad
    J. F. MbwiIiza, 'Social Production, Symbolism and Ritual in Ruha, 1850-1900",
    TNR, No. 79 (1976), pp. 20-21;
47. Baitwababo in Denoon, p. cU., p. 89.
                                          42