States and Social Processes in Africa: A Historiographical Survey

Author(s): John Lonsdale

Source: African Studies Review, Vol. 24, No. 2/3, Social Science and Humanistic Research on
Africa: An Assessment (Jun. - Sep., 1981), pp. 139-225
Published by: African Studies Association
Stable URL: http://www.jstor.org/stable/523904 .
Accessed: 17/06/2014 14:46

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].

African Studies Association is collaborating with JSTOR to digitize, preserve and extend access to African
Studies Review.

http://www.jstor.org

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA:A
HISTORIOGRAPHICALSURVEY

John Lonsdale

PREFACE
This survey can only be one historian's view of a large and controversial
matter.' African states are contested objects of study on two grounds, the one
particularto the changing preoccupationsof Africanists,the other common to the
humanities in general. Europeancolonialism was a living denial of the ability of
Africans to organize their own sovereignties. African studies emerged out of
colonialism's decay, in part a sort of recantation,in part a dimension of informal
empire. The analysis of past and present African states could not fail to act as a
lightning conductor of warring emotions, ideologies, and theories, categories
which have often not been very distinct. The first African histories after the
colonial era tended to be, as the first studies in politicalscience were necessarily,
studies in state-formation as achievement. In more recent years, it has been
objected that these were really chronicles of injury, not, as was thought, of pride;
for states were and are engines of oppression, not civilization (for early protests,
see Soyinka, 1967; Wrigley, 1971; Chanock, 1972b; and later, Markovitz, 1977:
25-55). Moreover, most Africans did not actually live in states until colonial rule
fastened Leviathan's yoke upon them. Indeed, the most distinctively African
contribution to human history could be said to have been precisely the civilized
art of living fairly peaceablytogether not in states.
The dispute was really over the evolutionist assumption that it is in some
sense "better" to live in states, a premise so deeply rooted in most of us that it
can seem almost an insult to explore analytically"the notion of statelessness, long
abandonedby historiansof Africa" (Uzoigwe, 1980: 115).2 This essay is based on
no such premise, however narrowlyfocused it may be upon states and upon the
social processes which are derived from states' formalities of function. Its
concerns are connections which govern processes, the connections between the
institutionalization of power and the conflicting constituencies of class,
community, and nation; between the exercise of force and the rule of law;
between modes of dominationand modes of production.
Here lies the second, more general area of controversy.A state is a presence, a
structure, which delivers power, an action (Poulantzas, 1978). And the
relationshipsbetween structureand action, between necessity and possibility,will
in the social sciences be for ever unresolved (Dunn, 1982). There is a wide

AfricanStudiesReview,vol. XXIV, nos. 2/3, June/September1981.

139

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
140 AFRICAN STUDIESREVIEW

variety of theoreticalapproachto this conundrum, whether employed consciously

by social scientists or, less frequentlynow than a few years ago, unconsciouslyby
historians.It is only fair therefore to give some sketch of my own.
It is not possible to perceive anything without the aid of some principle of
categorization. But once that is conceded, it can still seem to me that the
relationship between the social sciences and whatever reality may be held to
constitute Africa resembles nothing so much as the predicamentof the Saints of
the Atlas: too much theory chasing too little empiricaldata. For theory multiplies
by spontaneous generation, data are added to only in travail (cf. Gellner, 1969:
140). The behavioralcircularitiesof modernizationtheory, for instance, were as
difficult to use as have been, more recently, the market determinisms of
underdevelopmenttheory-which perhapsexplains their incandescentbrevities of
life. To this (doubtless very insular) outsider, structural-functionalpolitical
science of unimpeachable conservatism bears an uncanny resemblance to the
more structural brands of Marxism, and both, in their impenetrability,to the
Prince of Denmark's reading matter-save that where once the words, words,
words were in American, they are now more often French. But then, to one
rearedas I am in the positivist, empiricistEnglish historicaltradition,the clash of
theoretical combat must perhaps inevitably appear to cloak the field of African
historiographyand social science in the fog of war, leaving the footslogging
research workers uncertain not only as to the direction in which they should be
facing, but whether their sacrificesmay not be altogetherin vain.4
No matter how straightforwardlyone may wish one's history to be asked its
questions, it must nevertheless be true that greaterunderstandingcan only come
by moving to and fro-like migrantworkers, intellectual navetanes-between the
theory which suggests hypotheses and the facts uncovered by hypothesis, which
must also interrogate the theory (cf. Thompson, 1978; Law, 1981: 316). If a
hypothesis is tried and found wanting, there ought then to be severe limits to the
extent by which it can acceptablybe made more abstractor more complex so as to
accommodate the prickly particularitiesof the data; the simplest alternative
hypothesis ought to be preferred (cf. Southwold, 1968: 139; Wilks, 1978: 515;
Prins, 1980: 7; on the desirabilityof Occam's razor). If the simpler hypothesis
implies a different grand theory to that employed (however "silently") elsewhere
in one's explanation, it may well mean not that one's explanationis fundmentally
flawed but, rather, that human actions are not reducible to any single theory of
causation. There seems to me no good reason why one should not employ the
approaches of both Marx and Weber, particularly at different levels of
explanation. For empiricists to fund such eclecticism attractive is a deplorable
"law of ideological gravitation" (Therborn, 1971: 83), should one believe (with
Bernstein and Depelchin, 1978: 9f) that facts have no independence apart from
the theory or problematicwhich construes their significance.But since those who
hold this latter view accept that the facts and ideas createdby one problematiccan
also be used to answer the questions prompted by a contradictoryone (as do
Bernstein and Depelchin, 1979: 32), then perhaps there is something still to be
said in favor of epistemologicalimpurity.
At all events, it seems to me profitableto work with a set of what might be
called axiomatic uncertainties,each dependent on the initial premise that for any
reasonablysignificanthistoricaldevelopment, monocausalexplanationis ipso facto
wrong. Those questions beloved of examiners which ask "either-or" should
invariablybe answered "both-and." In other words, determinationis not a proper

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 141

term for historians, save in the weak and rather confusing sense of setting the
limits of the possible,as distinct from the necessary,in any historicalprocess. In
the study of states therefore, the most inclusive and powerfulof social structures,
the following axioms at least ought to be inscribed in whatever amulet one may
wear to ward off dogmatic error:that political structuresare not only systems of
constraint but also fields of opportunity;they work, generally speaking, but they
are also there to be worked (Morton-Williams,1968: 4; Lewis, 1968: xxii), so that
"dynamic links between ... different institutionsare best seen when one follows
the career of successful men" (Vansina, 1973: 18); that by themselves, structures
can be held to explain neither their origins nor their functions (Prins, 1980); that
"change is structured,and structureschange" (Burke, 1980: 13); and that even in
the process of supportingeach other, structuresconflict.
The relationships between structures, and between structure and action, are
perhaps even more pressing issues for African studies than for other areas of
social enquiry. For African politicalformations have been subject to what appear
to have been two radical transformationsin the past century; each entailed not
inconsiderableviolence. Precolonial polities were first incorporatedinto colonial
dependencies; colonies, most of them twenty or more years ago now, have since
become sovereign states. How far, it has often enough been asked in some
perplexity,were these breaks caused, or at least accompanied,by social rupture?
(e.g., by Wrigley, 1976: 508)? Where, to ask a more toughly systemic question,
would one look for the infrastructural quakes which might-or might
not-account for such superstructural shock? Was not the second shock,
decolonization, merely a trick of the light, by which the color of rulership (but
nothing else) was turned from pink to black? Even if it was more than that, can
the exercise of state power in Africaever more than marginallyaffect the main
determinants of its peoples' well-being? To put it more directly, "what was the
colonial contributionto economic development?", and then, "does independence
count?".
In edging toward more satisfactory answers to such questions, students of
history and society have been able to choose from among six divergent premises
on the correspondence between social, political, and ideological structures. One
premise is that social structures have a natural life much longer than (and only
precariouslyrelated to) the fragile timespan of politicalforms; they are like "coral
reefs" (Wallerstein, 1974: 3; and the Annales school generally). Conversely, it
can be argued that politicalstructures, far from being epiphenomenal,do actually
matter in socio-economic history precisely because they too are not shortlived.
They tend to be redefined and added to and patched as seems expedient, rather
than cleanly superseded, so that, over time, a "legal and institutionalchaos" is
laid down by "the geology of tradition."5 State institutions are therefore
structurallypredisposedto be the protagonistsof conflict at least as much as are
social classes, and by their own conflicts may release social forces which are not to
be contained short of revolution (Skocpol, 1979). Both these approachesassume
some considerableseparation,however difficultit may be to define, between state
and society.However valid the first propositionmay be in analyzingthe life-spans
of whole civilizations, it is the second which seems to offer more to the historian
of specificstates.
The next two premises assume that there is little if any distinction between
state and society but, again, from radicallyopposed directions.One would be what
I understand to be the behaviorist assumption-at least until the 1960s--that

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
142 AFRICAN STUDIES REVIEW

mature state institutions are those which are so consensually secure that they are
able to registerand process all the demands which surface in the free marketof a
pluralist (that is, essentially individualist) society, where state is synonym for
nation. Set against it is a Marxist premise, that the state embodies the political
and coercive conditions required to maintain any economic society (mode of
production) which is based upon a contradictoryrelationshipbetween the direct
producersand those who control their laborand appropriatetheir product.
How far, subject to this fourth premise, the state is simply the proprietary
instrument of the dominant class, or how far the organizationallycondensed but
abstractly determined social relations of production-and therefore only rather
equivocally available for the maintenance of class domination-depends largely
upon which of two premises, finally, one adopts on the potentialitiesof ideology
in social and political life. One may hold, first, that ideology is, or at least
originates as, the symbolic expression of what factually is the code with which
men signal their necessary relationships. It can be consensually agreed or
essentially contested according to the degree of social innocence in those
relationships;entirely contemporaryin its referents or subject to cognitive drag,
by which what has been "weighs like a nightmare on the brain of the living"
(Marx, 1954: 10). Against the implied Menshevism of this premise, one can set
its Bolshevik contradiction,that by "imaginative ideological labor" (Eley, 1981:
92) the conditions for effective politicalaction can be inspired where structurally
speaking they seem not to exist-whether to weave a magic web of hegemony for
the oppressors,or to rouse to revolution the heroic voluntarismof the oppressed.
Support will be found for all of these divergent presumptionsin the survey
which follows. This is in part because African studies are a broadchurch, in part
because of my own belief that each has its own illuminationto give to historical
understanding.What makes the study of states and social processes so absorbingly
controversial is that the more narrowly consistent one's choice among these
premises, the greater one's theoreticalrigor, so much the more mechanicalone's
history is liable to become;6on the other hand, always provided that one is not
self-contradictory,the less intellectuallyfastidious one is, or the more empirical,
so much the more likely one is to be able to explain what, looking about one,
seems to be the case: that if history teaches anything, "it is that the diversity of
ways of solving problems is so great that no safe predictions can be made"
(Vansina, 1968: 110). One might wish to add a parentheticalqualification,"within
limits," somewhere there, but that said, and given my intellectual upbringing,
that is in itself a predictableenough point of departure.
A PERIODIZATION OF PARADIGMS
Like any other branch of professionalized knowledge, African history and
social science are products of the middle-class mind. It matters not whether the
problematic is bourgeois or Marxist, nor if the scholar be African or Euro-
American, for the chief characteristicof the middle-class mind is its radical
ambiguity.It has been formed, whether in Africa, Europe, or North America, by
the contradictorybases of its class position. Privatelyto themselves, and publicly
to those whom they rule, middle classes generally justify their authority by an
ethic of service and by the universalisticvirtues of achievement by merit. And yet
this achievement is guardedby the very particularpower of culturaltransmission,
through the generations, of the habits and social networks of command-a
process itself assisted by class-specific inequalities of property-ownership(F.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 143

Parkin, 1979: 44-71; A. Cohen, 1981). This conflict at the heart of middle-class
culture, between professional pride and class shame, is even more painful when
projectedonto the enlargedscreen of the internationalrelations between the West
and the Third World, North and South, so that we can no longer with confidence
distinguish between aid and imperialism(cf. Leys, 1975: 23n; Warren, 1980: 1ff).7
Herein lies a particulardifficulty for Africanists. Africa occupies an unusually
tender spot on the West's cognitive map, at once guilty burden and wayward
child. It tends to divide its observers all the more sharply. Not only have the
dominant premises of our analysis shifted rapidly, they have always been
contested. Any attempt therefore to periodize our paradigmsof African states is
foredoomed to simplify what ought properlyto be complex (cf. Hopkins, 1980:
142).
Before coming to the paradigmsthemselves, one must first say that there seem
to have been two major stimuli to the changes of paradigmwe have employed
over the past twenty-five years or so to analyze African polities, whether
precolonialkingdoms, colonies, or contemporarystates. Africa itself has changed,
not apparentlyfor the better. Modes of academicanalysis have also changed. Our
conceptualization of power has become more complex. Inter-disciplinary
connections are being made in its analysis, so that social studies are beginning
again to be as historicallyaware, and historiographyas societally alert, as in the
nineteenth century both fields once used to be.
Africa's own transitions have been our first intellectual goad. In our anxious
middle-class way, Africa's students seem to have formed themselves, even
consciously, into a Committee of Concerned Scholars for a Free Africa.8 For
some of us, the study of Africa has been at least in part both a surrogatefor and
an aid to the study of our own societies, its "otherness" sharpeningperceptions
of similar but rather too familiar problems at home (e.g., Wallerstein, 1974: 3f;
Goody, 1976; ix; Greenberg, 1980: ix; A. Cohen, 1981: 14, 223). Thus events in
Africa have come to have "a moral dimension which is fundamentalto our age
and which brings the struggle for political freedom in Africa close to the social
problems which engulf our own nation" (Apter, 1961: vii). As Africa has
changed, so has the academic definition of its freedom (cf. Rodney, 1972: 66).
The first definition was simple; its moral premise seemed clear. It was based on
Africans' claims for political and racial equality. Individual self-realization,
political order, social freedom, and equity seemed destined to be joined together
under the renewed sovereignties of independent Africa. States were seen to be
beneficial, necessary for the collective good. "Nations," it was argued-certainly
by contrast with colonies-"are the most efficient and effective way to mobilize
human resources in a social unit large enough to permit the benefits deriving
from an extensive division of labor combined with a universalistic-achievement
orientation" (Wallerstein, 1964a: 127). Nation-buildingcarriedgreat moral force,
as in the present so in Africa's past (Wrigley,1971: 122; Triulzi, 1981: 289).
Freedom has become a more complicatedmatter since then, and not only in
Africa. The crises of capitalismand (writingas martiallaw is imposed in Poland)
the crises of socialism have also changed the questions with which we approach
Africa. Unsurprisingly,Africa's first definition of freedom, and ours, failed to live
up to the hopes invested in it. Committedscholarsbecame concernedthat African
states were unable to promote growth and in many cases deliberatelyabridged
civil freedoms. Academic opinion began to pull apartto its extremes. Where once
every study had concluded with pious hopes for Democracy-even if the faith of

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
144 AFRICAN STUDIES REVIEW

some was weaker than that of others-two rather strongly contrasted views
emerged in the 1970s. Some plumpedfor order, often in martialdress, as the best
means of coping with "the strains of modernization" (O'Brien, 1972, reviews the
literature). Others, the underdevelopmentalists,considered that there was little
that any peripheralstate of whatever political hue could do about either order or
freedom; the problem, and the solution, lay with the world system. The Marxisms
which have been more recently applied to the same issue are not quite so sure,
because, in contrast to most underdevelopmentalists,Marxists see classes not so
much as the transmission-beltsof power as its generators.Are Africa's dominant
classes locally resident or in the Western metropoles (the metro-bogey)?
Whichever the answer, politicalstructuresare now conceived as selfish, promoting
the freedom of the few by denying it to the majority.
Wrigley (1978) has correlatedthese shifts in the climate of academicopinion
with changes in Africa's terms of trade. Optimisticmodernizationtheory reflected
the rising terms of trade of the early 1950s. Underdevelopmenttheory followed
their fall in the 1960s. In the 1970s, intellectual uncertainty was buffeted by
contradictorymovements in the markets.
The second stimulus to re-thinking the problem of the state has come from
the cyclical movement in the academic analysis of power which has led us back
from political science to the older concept of politicaleconomy (Coleman, 1967).
If this is a general trend, however, it contains at least three different schools of
thought. The first has informed a critical revision within American political
science (Lukes, 1974). During the behaviorists' heyday, the 1950s, power was
argued to be founded on consensus and to be seen in action, in conscious
decision or visible conflict. It was wielded by individuals who enjoyed a rough
parity of power, particularly the power of expression. I shall call this the
symmetricalimage of power. More recently, it has been challenged by an
asymmetricalview, which focuses on the disparities of power. In Braudel's
mathematical analogy (1973, Vol. 2: 892), the plane geometry of analysis has
been replaced by a solid geometry. Nowadays, consensus is seen not as the
uncontested base of power but as a consequence of its successful exercise.
Furthermore,the power that matters is wielded not by individualsbut by groups
and institutions whose enjoyment of it is decidedly unequal, and all the more so
when conflict is kept inarticulate,suppressedbeneath the visible surface of social
relations.
No particularengine of change is posited in either of these views of power;but
a notion of a dynamic contradictionbetween them has long inspired the best of
Anglo-Saxon political historians, for whom the explanation of change is their
profession. This school traces the path to historicaltransitionby the tendency for
symmetrical conflict between power-holders, over the raising and enjoyment of
revenue especially, to demand more from their asymmetricalsources of power
than every they had imagined, and thus to politicize social relations which had
much better been left hidden. It is a habit of explanationwhich emphasizes the
politicalstrain of mobilizinga state's resources for war as an unintendedlever for
prizing open an existing constitution of power (e.g., Skocpol, 1979; Middlemass,
1979). It finds a ready acceptance in the analysis of royal or chiefly succession
disputes in pre-colonial or contemporaryAfrica (e.g., Southwold, 1968; Turner,
1957). It informs non-Marxist interpretationsof the decomposition of colonial
authority (Robinson, 1972; for India, Seal, 1973). Through the conflict of
successive political generations, it can be adapted to explain the dynamics of a

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 145

nationalist movement's expansion (e.g., Moore, 1965: ch. 1; 1970: ch. 2). It can,
finally, be used to predict the manner of socially progressivechange in the post-
colonial African state, whether by pressureat a criticaljuncture from below (Leys,
1975: 275) or consciously promotedfrom above by the radicalsection of a middle
class (Saul, 1974a).
Two implicationscan, I think, be drawnfrom this brief survey. One is thatjust
as there can be no very determinateconcept of determinationin history, so there
can be no historical conflict of any consequence which is either symmetricalor
asymmetrical;important conflicts are those which engage both levels of power.
The second implication amounts really to the same thing. It is that contested
affairs of state may often release the safety-catch on social processes. But can
underlying social processes decide the fate of states? That is another question
altogether.
It is a question addressedby Marxism, the second school of thought which has
contributedto the revival of politicaleconomy. To summarizeMarxistconcepts of
historical change in one paragraph is one of life's determinate impossibilities;
there are now too many Marxisms. Marx discerned not two but three levels of
historicalreality. There was a materiallevel below symmetricaland asymmetrical
relations; it supplied the historical drive for change. Save in the simplest of
societies, it also makes both relationships highly asymmetrical, political
domination supporting economic exploitation. Further, there is an upward
connection between all three levels, as opposed to the descending line of
causation in political history. It is held that development-or crisis-within the
productive forces determines, but only after a decisive conflict within the class-
divided relations of production, radicalrealignmentsin those relations and, with
them, fresh possibilities of productive development. But in Marx himself (G.
Cohen, 1978) and in Marxisms subsequently, there has always been some
ambiguityover quite how powerfullydeterminatethe materiallevel of reality is. If
it is indeed the prime mover, then one can talk of productivelogic in history; in
modern times, capital logic. But it can be vigorously argued (e.g., Brenner, 1976,
1977; Cooper, 1981) that what decides whether productivepossibilitieswill be put
into action at all, and then governs the efficiencies of their use, is the struggle
between classes over their economic benefits and social advantage, so much so
that it is class logic which ought to be given explanatoryprimacy.Debates over
the causes of decolonization and the nature of Africa's current predicamentwill
providean opportunityto consider these rival positions.
Ur-Marxism argued that material forces could be moderated or accelerated,
but scarcely deflected, by conscious political action. More recent debates within
Marxism, however, have been stimulated by the way in which this process of
radical change does indeed appear to have been deflected. At least in its
heartlands, modern capitalism is less convincingly in a pre-revolutionarycrisis
than modern socialism. So the uppermost level of reality, the politicalfact of the
state, has come under much more vigorous scrutinythan in the Marxistclassics.9
There are many unresolved dilemmas, the central issue being the degree to which
the state can distance itself from and thereby counter what history otherwise
decrees to be its fate. It can be argued in an abstractand functionalway that,
reduced to its duty to protect the reproduction of any given society and its
productive relations, a state has simply got to do what a state's got to do.
Relationships which work now are built into the conciliatory and coercive
structures of the state and can therefore be made to go on working. Change is

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
146 AFRICAN STUDIESREVIEW

bought off and locked up. If one puts people into the state, however, as historians
will keep on doing, then one can add two further possibilities. For people have
interests. Those who most intimately control the state's institutions have a self-
interest as third parties in adjudicatingsocietal conflict; yet this role is so
convenient to others (Service, 1975) that the state may be seen as more
ingeniously independent of, or autonomous from, material forces than many
Marxistswould accept as theoreticallydecent. But if the people are also divided by
class interest, then the state may well be seen as the legal instrumentof the power
and profit of one class over against another. Yet, it is possible to fault this
paradigmof the state for being too dependent upon class logic, neglecting the
deeper rhythms in the economy, which is subject to crisis and recomposition
independently of the wills of those who by conflict strive to shape it to their
purposes. One would then have to see the state as itself a complex form of social
process-derived in part from but not simply reduced to the necessities of its
mode of production-in which neither all-healingfunction, nor autonomous will,
nor canny instrumental conspiracy can be more than partial, contested, and
fleeting aspects of a system of domination which is always subject to stress and
forever incomplete. Without this insistence on the intractablematerial level of
necessity, Marxist analyses of the state must forfeit their distinctive character
(Jessop, 1977: 353). Only with this insistence does it become clear that however
much their visible forms and symbols of power may seem to resemble each other,
capitaliststates are, ineluctably,utterlydifferentfrom pre-capitalistones.
The concept of three levels of historicalreality is shared with Marxismby the
Annales school of French social history, the third intellectual contributorto the
new history and social science. Moreover, for Annalistes as much as for Marxists,
"everything is connected" (Braudel, 1974: 236). But quite how they are
connected is deliberately kept an open question (for a critique, Stone, 1979).
Braudel (1973, Vol. II: 892) refuses to pin down the correlationsbetween the
materialand other rhythmsof history, visualizing,rather, "a series of overlapping
histories developing simultaneously. It would be too simple, too perfect, if this
complex truth could be reduced to the rhythms of one dominant pattern."
Annales has come to Africa fairly late in time (Clarence-Smith,1977), imported
apparently by that transcultural man Vansina (1978: 10). The school's
complexities are illustratedby the differences between the three major works of
African history which it has so far inspired.Vansina's book on the Kuba (1978),
so rich in its detail on daily life, stresses the creative productivityof political
domination. In Iliffe's masterworkon Tanganyika (1979), the demands of the
state constitute, rather, an ordeal, arbitrarilyimpedingman's strugglewith nature.
Two characteristicallyautonomous states, one beneficial, the other cruel, emerge.
Prins (1980) is altogether more enigmaticover the relationshipbetween history's
three levels and chronologies, long-enduring material and cosmological time,
intermediateeconomic trends, and the day-to-dayismof politicalevents.
The Annales school does seem to be better attuned to the long and loosely
articulatedrhythms of pre-capitalist(or, more precisely,pre-industrial)life, just as
Marxist historiography grapples with the more insistent connectedness of
industrialsociety. And for Africanists especially(orhas this old wound ceased to
throb?), the school of Braudel and Le Roy Ladurie has that calm sense of the
evanescence of political pomp, armed with which we could at last ignore those
who scoff at the study of the "unrewarding gyrations of barbarous tribes"
(Trevor-Roper,1965: 9). The Spain of Philip II could not do much better. It is

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 147

not at that political level anywhere (indeed, more often in resistance to it) that
popular culture has been formed, nor, until our own day, the cycles of
demographyor the intensity of labor even approximatelyfixed (cf. Prins, 1980:
9).
What then can politicalaction achieve-a question of some seriousness in the
study of states (Dunn, 1982)? Marx's teasing reply, that "men make their own
history, but not under conditions of their own choosing," is probablyas near as
we shall ever get to an agreed answer. It is possible to take a very severe view of
the matter and to argue, with Perry Anderson (1980: 16-58), that anything short
of intended projects to change whole socio-politicalstructuresis disqualifiedas a
significantexercise of human agency. He excludes from the causes of meaningful
effects private or even public ambition to seize a more profitablehold on existing
productive arrangements.The effect, however, would be to banish purposeful
human action from historicalrelevance altogether. For social revolutions are not
only rare, their revolutionaries' visions are generally only marginallycausative,
and their outcomes never quite as anybody intended (Skocpol, 1979). Marx
himself (1954: Preface) was inclined to this view, conceding only-but
witheringly-that the vital uncertainties of struggle in a period of rapid social
change might make it possible for a "grotesque mediocrityto play a hero's part."
In the analysis of more normal, non-revolutionary times, a similarly strict
emphasis on structure rather than action (e.g., Poulantzas, 1973) makes it
impossible to distinguish the exercise of power from mere functional
determinationand rules out, therefore, all judgment on politicalresponsibilityand
irresponsibility(Lukes, 1974: 46-56).
A functional stress in explanation seems to imply that a given mode of
production will determine the quality of human life, that for those living at
comparableclass levels in each majorera their sense of libertyor oppression, their
enjoyment of materialgoods, will be about the same. This is manifestly not the
case. Functional history cannot comfortably accommodate differing degrees of
material success or moral acceptability in different but equivalent social
formations. Social reproductioneither succeeds or it does not; it is then replaced
by a new orderingof society. It is difficultto see what, other than some allowance
for individualagency, can account for the fact that history's course meanders for
the most part somewherejust short of these starkalternatives.Even class-struggle
seems too anonymous a version of agency to explain all the myriadsof different
textures in human life which clothe its objective necessities. Social being does not
wholly determine social consciousness. Rulers do have some discretionwhether to
use any instrumentalmeans to an end or to hold fast, as does for instance Julius
Nyerere, to the ultimately moral insistence that political means are themselves
ends (Pratt, 1976). Conversely, the oppressed do not always even see a remedy
for their misery, let alone collectively apply it; a constructive and articulate
diagnosis of injustice represents an extraordinaryeffort of will and imagination
(Moore, 1978). It would be an appallinglysterile historiographywhich denied all
agency to creative politicalgenius and purpose. Without some acceptanceof what
Anderson (1980) rejects, the "productivity of politics" (Ilchman and Uphoff,
1971: 29-33), it would be impossible to envisage any of the paradigmsof the state
which have been most commonly used in the study of African polities, the
autonomist paradigm,the instrumental,or even the process model.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
148 AFRICAN STUDIES REVIEW

The Autonomist Paradigm

The dominant assumption in the first modern studies of African polities,
whether past or contemporary, was that political power was autonomous; the
power assembled in the state was moreover symmetrical, active, creative, and
agreed. There were two further premises associated with this view, one on the
nature of human agency in history, the other on the power of ideology. As to the
first, it was taken for granted that politicalaction could be productive.It was not
requiredof men that they organize to overthrowexisting social relations in order
to be deemed to be making their own history but, rather, that they seize more
efficient hold on such social structuresas were in any case emerging. States were
often fused by "revolutions in government" engineered by existing power-
holders. These reorganized opportunity and redirected ambition by creating
"free-floating resources" for which there were no established rules of
competition. By such means, it was possible "to enlarge the volume of power" so
that all might benefit, not just the rulers (Partridge,1963: 123). Marx too had
argued that force could be an economic power, but only when acting as the
midwife of a revolutionarynew society and mode of production.The autonomist
vision of the state current in the 1960s was subject to no such restriction:rather,
organizationwas its own revolution (Sklar, 1976: 82).
The productivityof ideas was the other ruling assumption.Politicalgenius was
the motor of history, not social determinism. Any history of nationalism, the
founding social process of the modern African state, is bound to emphasize the
power of decision. Political identity is not entirely given by political situation;
there is always a variety of ways in which a situation may be interpretedand a
choice, therefore, between the teams which would be mobilized:nationalismwas
not inevitable (Iliffe, 1969: 152; cf. Adam and Giliomee, 1979: 83). Men of vision
had to construct nationalism, men prepared to invest "imaginative ideological
labor" (Eley, 1981: 92) in conjuring up a national identity; by "cultural
entrepreneurs"(Young, 1976: 46) who createda national languageby using it; or
by "political entrepreneurs" (Zolberg, 1966: 19), men who gambled on the
reliabilityof strangers.Academic analysts therefore saw as formativevirtue what
colonial governors had dismissed as self-interested fraud, "the premature
synthetic nationalisms of demagogues."'1 Incumbents also, however, need the
ideas which are providedby what Gramscicalls "organicintellectuals,"the people
who can convert the conflicting interests of varied elites into a coherent
philosophy of class domination (Bozzoli, 1981). Even if African nationalismdid
little other than take over and render fruitfully legitimatethe colonial institutions
created to serve the purposes of an alien power, then the charismaof leadership
was still required for what essentially was an act of faith in its followers (Apter,
1955).
In this nationalist era, therefore, the analysis of nationalismcould scarcelybe
anything other than voluntarist, even joyously idealist. Nationalism was seen
above all as a movement of ideas, the organizationalweapon by which the
moderns ousted the traditionalsas well as the colonial power (Smith, 1971, for a
critique). Values distinguished true nationalist parties from their opponents.
Nationalismwas a mobilization system,creatinga society as it should be; the others
were reconciliationsystems, expressing society as it was. But it was not only
African nationalism which was analyzed in these terms; given the prevailing
symmetricalanalysis of politics, with its emphasis on decision and open conflict,
the same autonomy of political power was ascribed to African states of all eras:

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 149

colonial, pre-colonial,and, soon, post-colonialalso.

In these early days, the most relevant attributeof colonial states was thus the
distinctive constitutional and cultural tradition which they severally transmitted.
The historicalexperiences of the colonial powers now became the different molds
in which were shaped their successor African elites (Hodgkin, 1956: 17, 29-59;
Young, 1970a:479-99). It was not by accident that one of the earliestand still one
of the best academichistories of colonialism was cast as a biographyof one of its
culture heroes, Lord Lugard (Perham, 2 Vols.: 1956, 1960). This academic
interest in who colonialists were and what they intended, rather than in what by
function they unavoidably did, was of course a reflection of the attitudes of
colonial rulers themselves. Since native administrationwas best carried through
"by personalityratherthan by legislation," colonial officialswere thought to need
"not so much brains as personalityand character" (Coryndon, 1908, quoted by
Kirk-Greene, 1980a: 44; Furse, 1962: 142). They were not, on this view, very
convincing idealists-though in the next section I suggest that many were-but
they were certainlymen of action.
Africa's pre-colonial past was similarly invoked with the drum-and-trumpet
history of kingdoms, autonomous actors endowed with a politicalwill, midwives
apparentlyat their own birth, whose "genius was for integration";so that Africa's
contemporarystates were not emerging "out of the void of a motionless past"
(Davidson, 1959: 267). To the contrary, "Africa [had] greatness in her past as
well as her present" (Wallerstein, 1961: opening line). History was the
legitimatingcharter for an independence renewed. There was also some truth in
the remark that the new African historiography was offered by whites in
"sentimental reparation"to Africans for the sins of empire (Owen, 1955: 28).
Our intellectual energies were "pitted against the parrot-cryof the expatriate
clubman in Africa, that 'They never invented the wheel'." " But it was not just
that we were, as always, writingcontemporaryhistory, for the emphasis in British
anthropologyat any rate upon the sharp opposition between society,the realm of
kinship, and the state, which was "never the kinship system writ large" (Fortes
and Evans-Pritchard,1940: 6), implied that an enormous creative heave was
indeed necessary to move from stateless societies to states. Moreover, states did
supply valued efficiencies in resolving social conflicts, no matter whether their
power had been assembled by selfish men. It was this external managementrole
that made them independent of society, so that in a work which epitomized a
dominant trend of thought, Service (1975: xiii) could write of governors "who
created themselves, so to speak, rather than having been the creation of others,
such as a 'propertied',or economicallyfavored, class."
It would be difficult to envisage a more autonomous power than that, but it
was certainlyequalled in the expectationsinvested in the states of the new Africa
which sought to continue nationalistmobilizationby governmentalmeans. Often
the beneficiariesof the "initial organizationalspurts"of party-formation(Zolberg,
1966: 134), political leaders were, or at their best could be, animators.They had
the potential to possess their country of "a kind of creative madness ... which
comforts the heart of man, gives him fresh confidence in the history of mankind
and disarms the most reserved of observers" (Fanon, 1965: 75, 78). The first
critics of African governments in this autonomist school, puritansthereforerather
than socialists, paid the sincerest tribute to the productivepossibilitiesof politics
when they called for a cleaner public morality. Developing nations could be
distinguished according to "the amount of public spirit and devotion to duty

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 150 AFRICAN STUDIES REVIEW

shown by their elites"; a country "in the grip of a self-seeking and corruptelite"
could not but forfeit its claim on those of its people's efforts which were "not
solely motivated by the hope of personal gain" (Leys, 1965: 229; cf. Dumont,
1966: ch. 6). Nor should the analysts of the day be convicted of a misplaced
idealism. Rather, the belief that political enthusiasm might release a vent for
surplus social energy was grounded in the dualist economics then current, under
which traditional societies could without damage, or exploitation, yield up
unlimited supplies of labour (cf. Katz, 1980: 26; Cooper, 1981).
African history and social science can nevertheless properlybe characterizedas
idealist at this time, and in two senses. They were teleological;they measuredpast
and present change against contemporary hopes for African democracy. In
addition, politicalintentions and values were seen as primecomponents of power,
especially of the power to institute change (cf. O'Brien, 1972: 353-57;
Lemarchand, 1973: 309f; Crummey and Stewart, 1981: 18). The histories were
Whiggish, Pilgrim'sProgresses, full of buildersfor the future. Coincidingwith the
"Apter to Zolberg" era of modernization theory (Coleman and Rosberg, 1964:
v), nationalisthistory was also rather short lived, little longer than the Black Star
decade of Nkrumah's lease on power, 1957 to 1966. Liberal optimism over the
renewed autonomy of African power was soon to be challenged by a "radical
pessimism" (Ranger, 1968: xxi), which in tentative form had long been present.
The radicalpessimists viewed African power as an instrumentin the hands of the
few and too often wielded, still, on behalf of external imperialism.
Autonomist and instrumental views of the state were derived from entirely
different perspectives on what divided society. The autonomists' state was an
independent composer of a concert of interests within a new political culture.
These interests were conceived in their unregenerateform as either individualor
communal. Political leadership was most fully exercised with cajoling
"transitional" individuals, men and women shaken out of their old "tribal"
communities, whether in precolonialtimes by war, or in the postcolonialera by
"social change," into the new community of the kingdom or nation (Miers and
Kopytoff, 1977; Deutsch, 1961). The argument for the necessary cultural
autonomy of the postcolonial state over against Africa's cultural pluralismswas
carriedfurthest by Young (1976), and at a time when the prevailingparadigmof
the state had become altogether more skeptical of the power of political
conviction. Young sought to explain why ethnicity, far from subsiding in a
growing market of transactionalidentities as modernization theory would have
liked, was increasinglya badge of confrontation in the contemporaryworld, and
by no means only in Africa. It seemed best explained as one of the teams most
easily mobilized either to resist, or to gain access to, the growing interventionist
power of the modern state. Within the "hardeningmatrix" of each state he could
discern core elites, a "national mandarinate," intellectuallyconvinced that the
state must create a nation, while also being themselves the self-interested
"human incarnationof the state-consolidationprocess" (1976: 82, 212). Here we
have the presentation of an autonomous state which creates not harmonies but
fresh discords, not exercising its independenceso much as coping in a process.
If human society is seen as divided by classes rather than cultural
communities, then this drive to nation-buildingought, rather, to be interpretedas
the creationof a class hegemony or legitimacy (Saul, 1974a:351; Tamarkin,1978:
312). Perhaps the most enduringweakness of African studies is that we have not
yet devised means of analyzingsocieties, doubly divided by both community and

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 151

class, which do not seem crassly biased towards the explanation of one division
alone. Where class divisions are perceived, there we have a social formation
incorporatingone or more modes of production whose states cannot be the
orchestratorsof harmony. They become instead (if one is not very careful) the
guarantors of exploitation, for they are certainly parties, twice over, to the
fundamental struggle by which dominant classes derive surplus from the
dominated. Lords or proprietorsdo need the help of the state's coercive power,
however much it may be dressed up as reciprocal rights and obligations; and
without some taxable surplus, the state itself could not survive. In the
instrumental view of things, therefore, the study of states is not to do with the
spread of civilization but, at least in the modern era, useful only insofar as it
bears upon the spreadof capitalism.
The Instrumental Model
One can distinguishat least four different versions of state instrumentalismin
the literature, all as yet more symmetrical than asymmetrical in form, seeing
power in its visible exercise rather than as a many-sided social relation. The
earliest version was in fact still couched in the language of cultural pluralism.
Furnivall (1948), and Kuper and Smith (1971), argued that external force,
wielded by a dominant minority, was the only possible factor of cohesion in
culturally plural societies whose several communities had literally nothing in
common save a market place, to which in any case they had unequal access. It
was a description which most easily fitted colonies, or South Africa; but it has
been applied to postcolonial African states as well, where unity is said to rest
more on coercion than consensus. In some cases, the single partywas said to act
as "a mechanism of domination by one ethny" (Andreski, 1968: 65, 62). Such
analyses, to which Young (1976) is an extended refutation (also, Adam and
Giliomee, 1979: 42-46), suffer from a fundamental flaw. They see cultural
difference, or ethnicity, as a given unit of political action, but the creative link
between culture and politics is precisely what needs to be explained as, for
example, Young does. It is difficult to imagine an aroused ethnicity prior to the
state. it is a response to state power, even a condition of its successful exercise, in
providingthe categories between which men divide in order to rule. One also has
to ask whether a simple coercive instrument best serves the interests of the
oppressors. Restraint in the exercise of force was as necessary for precolonial
warriorelites as it is for white South Africans, lest it destroy the very resources
which it is designed to exploit (Roberts, 1980b: 407f; Adam and Giliomee, 1979:
9f). Here, then, is a version of instrumentalitywhich also needs to be looked at
again, with the state thought of not so much as an apparatusas a process.
The three other symmetrical models of instrumentalityhave been mostly,
though not exclusively, applied to the colonial states and their successors. The
first two were more or less contemporaneouswith each other, but divided on how
convincingly educated minorities could also be seen as an entrepreneurialmiddle
class. One we can call class-instrumental,the other neo-colonial;both can be seen
as contested aspects of the third model, class conflict. Politicalscientists were not
very quick to see modern nationalism as a class organization; it seemed too
composite a bandwagonfor that, although with hindsight one can see that such a
vehicle was exactly right to bear hegemony. The first observers to distinguish the
private interest which the "class of professional politicians" might possess in
nationalism from the hopes of the masses were the colonial powers, and they

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
152 AFRICAN STUDIES REVIEW

were scarcely disinterested (Creech Jones, 1947, 1951; Robinson, 1974; Cell,
1980). Journalists also asked the question, "Freedom for Whom?" (Legum,
1954: ch. 26); and Hodgkin, in the first book-length analysis of African
nationalism, hinted that those who organized politics might also be its
beneficiaries (1956: 115ff, 138, 168; also, 1976: 12). Perhapsironically,the most
enthusiastic interpreterof nationalism as the weapon of the nationalist middle-
class was Frantz Fanon. The middle class had to sweep away the barriers of
feudalism and obscurantism in order to develop as a class; in promulgating
modern ideas, they also took the bread out of the chiefs' mouths (Englishedition,
1965: 89). Among professional political scientists, Sklar seems to have been in a
small minority in also describing political parties as "veritable engines of class
formation" at this time (1965: 204), perhaps because Fanon's Wretchedof the
Earth, first publishedin Paris in 1962, does not seem to have been well known to
anglophone scholarshipuntil it appearedin translationin 1965.12For both Fanon
and Sklar, the middle classes were formed primarily (but not exclusively) in
commerce. Nevertheless, when analysis moved on from nationalism to the
postcolonial states to which it had given birth, the only class which anybody saw
as purposefullywielding power were the bureaucraticbourgeoisie, those with the
unique expertise of rule, learned at the colonialists' office desks (Meillassoux,
1970; Shivji, 1976). At this stage, however, no scholar thought at all extensively
about what difficulties other classes might put in the way of the state as an
instrumentof bourgeois hegemony.
In any case, the question always was, the instrumentof which bourgeoisie? It
was perhaps a journalist, again (John Gunther, 1953), who first published the
notion that by decolonization the imperialpowers were really "giving in order to
keep," through the neo-colonial agency of a contrived middle class. Balandierhad
also recently shown that colonial regimes were far from being the rationalizing
instances of modernizationtheory. Their social engineeringmay well have divided
natives into classes, but for the native bourgeoisie as much as for the proletariat,
the colonial situation was "a trial . . . a crude sociological experiment" (1951,
translated 1966: 38), a "blocked transition" (Meillassoux, 1970), not a field
cleared for capitalist accumulation. Fanon agreed (1965: 122). A constructive,
inventive national middle class was an impossible product of colonialism since
expropriationhad been the monopoly of the colonialists. The local bourgeoisie,
traders in particular,could aspire only "to keep in the running and be a part of
the racket." This has been an influential analysis. Those who have followed it
have not perhapsbeen sufficientlyalert to the fact that Fanon was writingwith the
effects of settler colonialism primarilyin mind-although in another specific case
Leys (1978) has shown that the accumulationof settlers may indeed clear the way
for a vigorous national bourgeoisie. Fanon's insights have promptedthe concept
of the overdeveloped postcolonialstate, which was originatedby Alavi (1972) and
debated and revised by, among others, Saul (1974a) and Leys (1976). The
argument turned on the issue of how instrumental to the metropolitan
bourgeoisie the neo-colonial state could be. If (as arguedby Fanon) the new state
was nothing more than a political manifestation of a weak commercial
bourgeoisie, then clearlyone would have a pure neo-colony. But few neo-colonies
are pure; the relationships between the West and the Third World are by no
means uncomplicated. Alavi sought to explain this by arguing that since their
commercialand other dominant classes had been suppressedin the colonial era as
obstacles to metropolitan gain, postcolonial states could scarcely be local class

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 153

instruments.They were controlled, rather, by bureaucraticand militaryelites who

exercised some discretion on what was, therefore, a distinctly sectionalized
transmission belt between the metropolitan and national bourgeoisies. The
autonomy of the state's apparatuswas coming back, to explain the vagariesof the
politicsof dependence.
For all their discussion of class, neither of these models was really a class
analysis. Middle classes tended to be seen as groups in themselves. Their
relationshipwith capital was barely touched upon, their confrontationwith other
classes in productionnot at all. The little power which they did wield was a thing,
not a social relation. Here the analyses were symptomaticof the general tendency
of underdevelopment theory, to which they were a contribution.
Underdevelopmenttheory argued that the unequal exchange of global capitalism
determined class formation; class struggle was not allowed to govern in any way
the level of surplus which might be realized by that exchange (see further,
Cooper, 1981). The colonial state and its successor were therefore imposed
according to the needs of the market rather than developing as the continuous
culmination of internal social conflicts. Neither analysis considered how the state
might be used in internal class struggle except, possibly, as a mystifyingagent in
the hands of the popularleader (Fanon, 1965: 134ff).
There was a logically prior question, however, which now began to be asked:
how were middle classes themselves formed? It was not enough to point to one
specific asset, property,or state office. No bourgeoisie can reproduceitself in its
fragmentations. Social reproduction is meaningless unless interpreted as a
sequence of compromises with equivalent groups in order successfully to pursue
confrontationswith others. One has to look at the formationof coalitions through
conflict in order to understandhow classes gather. Neither the class-instrument
nor the neo-colonial paradigmof the state went so far. If, however, coalition-
buildingis acceptedas a problem, then the state ceases to be a thing and becomes
rather a channel of action (Sklar, 1981), or a resource. It is at this point that the
model can apply not only to postcolonialbut to colonial and indeed to precolonial
states in Africa. Precolonialkingdoms began to be seen as politicalarenas for the
conflict of separatepower holders, rather than as autonomous third partiesabove
such competition (Lloyd, 1965; 1968a; 1968b; 1971). Even colonial rulers had to
seek out power by sharing it with African elites in "the politics of collaboration"
(Robinson, 1972). The national bourgeoisies of contemporaryAfrica were not
now seen as single-strangedin their interest but derived ratherfrom a "fusion of
elites" (Sklar, 1979). They may not have had any joint interest other than the
maintenance of a social order, but quite what social order, capitalistor socialist,
was itself a question. Its answer depended very much upon the way in which the
fused elites had themselves been formed and the type of power they now
exercised, whether private, party, or state (cf. Murray, 1967). At one end of the
spectrum of possibility of this class conflict model, revolutionaryelites could be
fired in the crucible of people's war (Fanon, 1965: 85-117, 150-59; Cabral, 1969;
Davidson, 1977; Chabal, f.c.). At the other extreme, a confident proprietaryclass
could tolerate some show of bourgeois democracy, as can fleetingly be seen in
Kenya. The plasticityof most African bourgeoisies, however, seems to allow for
an almost infinite variety of political practices-at least, until one adds to the
analysis some of the sterner necessities of material reproduction and the
disparitiesof asymmetricalclass conflict.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
154 AFRICAN STUDIESREVIEW

Reproductionand conflict are, or should be, historicalterms. Neither state nor

class can be understood at a moment in time. They accumulate meanings,
symbols and power, over time. In ever looking for more powerful or inclusive
explanation, historians have increasinglyhad to examine asymmetricalrelations
between unequal conditions of men. In most political history, I have suggested,
asymmetricalpower is generallyseen as being opened up for reinterpretationas an
unintended consequence of symmetricalconflict between power-holders.But it is
also possible to concentratemore directlyon the hidden underpinningsof power,
its sociology rather than its politics, upon the perpetual antagonisms stirring
between the surface of politics, as contrasted with its periodic uproars. Both
approachsto historicalchange ask more insistent questions about the nature of
the state than do either the autonomist or instrumentalperspectives.
The Process Model
Otherwise than in myth, states do not have origins; they are formed. Origins
are magical events; formation is slow, often very slow, social, and political
process. States are formed through conflict, in which people are hurt and
humiliated; by contract, in which wounds are bandaged (even in time perhaps
healed); and by successful self-defense against neighbors, by which hierarchyis
justified and, should defense be carried over into the offensive, an important
means by which supportersmay be rewarded.Each facet of state-formationhas its
peculiar face of power, so that as Miller (1976: 269f) characterizedthe Mbundu
kingdoms of Angola, states are a fabric of ordered tension between "a variety of
competing forms of authority," each with different myths of legitimacy and
principlesof allegiance.Successive pragmaticrules of powerwhich become norms,
new layers of institutions to cope with new problems, get written into the
historicalstructure, partiallyrubbed out and written over again, so that all states
are "palimpsests of contradictions" (Lonsdale and Berman, 1979: 491; cf.
Berman, 1980: 3); nothing quite "fits" (Bradbury,1969: 29; Vansina, 1973: 401);
one instrument of state clashes against another; and there are always anomalies.
As Bismarck probably and Lord Salisbury certainly never said in a common
misquotation-and both played a large part in the formation of the states of
modern Africa-"If the state is the executive committee of the bourgeoisie, then
what am I doing here?"' If then there is a crucialrelationshipbetween the power
of the state and the power of class, the historicaldimension of their interactionis
essential to its understanding.Only states without a past look like the simple
instrumentof their dominantclass; and even revolutionarystates have pasts.
There are periods when states do appear to have origins, to take on new
forms, autonomously to exert or instrumentally to transmit power more
intensively in a burst of law-making(Service, 1975: 90). Historicaleras do come
to an end; new ones do begin; even people living at the time can be awareof that.
Three broad approachesto explaining such discontinuities in or surges of state
formation seem to be particularlyapplicableto Africa. Each turns on a distinction
or dichotomy, the first between state and nation, the next between state and class,
and the third, to put it at its simplest, between coercion and consent.
Nation and state are a familiar conceptual division in unity. One is an
organism, the other an organization.The modern exaltation of the nation state
presupposes that, together, they each bring out the best qualities in the other;
power becomes accountable,and the communitycan defend itself. But the nation
state is a recent phenomenon. In Africa, Somalia is often said to be its sole

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 155

representative,although presumablyTunisia and Egypt have equal claims, as also

do Swazilandand Lesotho. Two opposed sequences are commonly traced in the
coming together of state and nation and, with them, two contrastingtendencies in
the political culture of nationhood (Young, 1976: 69f; Eley, 1981). In northwest
Europe, states created nations; in central Europe, nations created states. In the
first case, kingdoms inaugurated the modern era when they elaborated the
bureaucraticpower of absolutism against the customaryand representativestatus
of society's "Estates"; this absolutist engorgement of power was then challenged
by representativebodies, parliaments,themselves foci for the organizingenergies
of parties which derived authority from their claims to speak for the nation;
subjects had finally become citizens. Nationhoods so derived were inseparable
from liberal, bourgeoisfreedoms. By contrast, in the centralEuropeancase, where
the "creative political action" of cultural nationalism was organized against the
alien state, sovereign nationhood was not so much contractual as organic,
potentially authoritarian (Eley, 1981: 85-90). All nationalisms exhibit both
characteristicsin varying measure, and it is not certain that Africa's are so very
unusual in their combinations of institutional formation from above and
ideological imagination from below. To some extent, nationalism is the end-
product of colonialism's transformation of Africa (Coleman, 1954). Equally,
Africa's nationalistssecured state power before they had createdmuch of a nation
(Young, 1976: 70f). The uses to which they may now put the state in order to
build the nation, however, were to be found not so long ago in establishednation
states. Even in France, the revolutionary Adam of the new age of popular
sovereignty, it was state action which finally turned peasants into Frenchmen,
with road-building, military conscription, and compulsory education (Weber,
1977).
There is also a third element in the process to which proper weight has not
alwaysbeen given, and that is the degree to which the right of representation,and
therefore opposition, is constitutionallyrooted in any given historicalexperience.
In Europeanhistory, royal absolutismwould not have been so offensive had it not
shouldered aside the customaryrights of the civil Estates and, often, co-opted the
Church still more firmly as an auxiliaryarm of government. The question needs
to be faced in African history, too, and it may be here that modern Africa's
experience is most impoverished in the dialectic of state and nation. It is not
enough to hypothesize, as once was not uncommon (e.g., Fallers, 1959), that by
requiring the appointment of low-born officials in order to break aristocratic
privilege, precolonial despotisms forged something akin to a national culture in
the ensuing tumult of social mobility. For the terrifyinguncontrolabilityof the
nineteenth-century kings of Buganda might well not have been so coherently
confronted by its "Christian revolutionaries"had they not been able to assume
the discarded rights of Buganda's own tradition of a "priestly opposition"
(Wrigley, 1959). In those same years, it has been argued (and no less
controversially)that Egyptianautocracycould by contraststand anythingsave the
suspicion of its own weakness. For all their privaterights in land and exposure to
Western thought, Egyptian notables had no inherited constitutional perspective
within which they might have even begun to think of opposing the sovereigntyof
the nation to that of the dynasty (Schi1ch, 1974; but see Abu-Lughod, 1967).
Modernizationtheory once extolled the virtues of secularizationas an aspect
of political development. One might concede that point in lands where there had
been something like a national church as a countervailingpower to the state; and

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 156 AFRICAN STUDIES REVIEW

the modern welfare state is in large part simply the secular descendant of
ecclesiastical-or Islamic-provision for the destitute. Africa's own religious
traditions, however (Islam excepted), were certainly not national at the time of
national independence, and missionary Christianity was too identified with
colonial rule to be a confident keeper of the state's conscience after decolonization
(cf. Lonsdale et al., 1978). Africa's new states have simply not got a traditionof
constitutional politics (cf. Ruel, 1969), primarilybecause they have not got a
relevant religious tradition.But Tunisia, for instance, has-one in which a three-
hundred-year-olddynasty was rightly guided by Islam. In colonial times, its first
nationalist party took the name of "constitution": Destour. Since 1956, the
government of neo-Destour has in consequence been more hesitant than many in
unleashing the violence of the state (Moore, 1965).
There is therefore no simple relationshipbetween state and nation. It has been
mediated by religion, or, more broadly speaking, a cultural traditionwhich very
often stands obliquely to state power. Nor is there any simple relationship
between state and class, or coercion and consent. But it is not possible to explore
further these versions of the processual understandingof states before one has
unpackedthe concept of the state itself.
THE DIMENSIONS OF STATES
As in African historiographythe concept of the state has grown more complex,
so in analyses of contemporaryAfrica it seems generally to be becoming more
stark. Yet precolonial African kingdoms were often neither complicated nor
powerful; and modern states take on ever more varied measures of authoritative
intervention. The paradox is easily enough explained. In history, processes of
change are most densely documented in conflict between institutions, or in the
multiplicationof offices (notablyWilks, 1975). In contemporarypoliticaleconomy,
the function of the state is the gravest issue, and its unity, therefore, if with
varying degrees of internal contradiction, a facilitating assumption. The
characteristicsof states depend upon the questions one asks. To explore the
relationshipsbetween state and class, a more differentiatedconcept of the state is
required than that which tends to structure arguments about coercion and
consent. Yet both are needed.
In everyday talk, most people know what they mean by "government." But
"state," if it is not merely to be a redundantsynonym for government, implies
somewthing wider, more inclusive, more confusing. Scholars who are happy to
talk about bureaucracies,the law, armies, politicalparties, class structures,and so
on, can be reluctantto lump them together. It seems truer to historicalexperience
to use bits and pieces of a model of the state, as the Nyakyusa have come to
observe their rituals (Wilson, 1959: 211). Nevertheless, to try to see things whole
is now as much the ambition of African studies as of any other branch of the
humanities. And it appears that most of our bits and pieces can be assembled
within a four-partconceptualizationof the state-its apparatus,its representative
estates, its ideology, and its materialbase.
With increasing historical knowledge, it has become plain that in any known
state, these four structuresought to be seen as cockpitsof conflict, interconnected
but not mutually reducibleeach to the other. Their connectedness is what makes
the analysis of states so difficult. Our common terms of description betray this
tension and circularity in our understandingof state power, whether "input-
output" (Easton, 1957), "presence-action" (Poulantzas, 1978), "large-scale

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 157

feedback systems" (R. Cohen, 1978: 70), or, indeed, "nation-state." And
structure is itself an ambiguous term. Following Levine (1974: xiii), I am clearly
using it here to refer to "three kinds of social facts-institutional orders,
interactionalpatterns, . . . culturalcodes," and a fourth: modes of production.
It is an historiographicalcommonplace that between the holders of state
power, and not least in Africa, there is constant conflict over policy and profit;
kings try to subdue barons, executive governments attempt to control
administrative departments. Moreover, each of these segments of a state's
apparatuswill probably have competing forms of access to or identity with the
society's estates-in one era priests, nobles, elders; in another, intelligentsia,
owners, and workers. Each estate is in conflict with the others, not least over
whether they should be dignified with representative status at all. Further, if
human power, any power, is not in some way mystified, it is mere violence. The
ideologies of power are also its disguise. But disguises fool nobody when the
hidden face of power becomes too widely separatedfrom its public mask. Finally,
the materialbase of states will always be riven between the provisionof economic
citizenship for all and the perquisitesof control for the few. Given the complex
nature of states, all these fields of conflict overlap. On paper, they have to appear
in sequence. In reality, they are Chinese boxes.
Strugglesbetween kings and subordinatechiefs have long been a staple theme
of precolonialAfrican history. Control over succession to office has generallybeen
taken to be the key prize, but as to its significance, opinions have radically
changed. It used to be argued that conflict over royal succession was functionalto
the health of the community at large; symmetrical competiton between noble
electors, the champions of rival princes, would mobilize their asymmetricalbases
of power, their followers, so that the nation was at least temporarilyinvolved in
political competition and the relations of power thereby renewed (Gluckman,
1955: 27-53; Moore, 1970: 15f). The sphere of functional conflict was then
expanded to embrace royal efforts to control succession to subordinate office.
Weberian concepts were applied, perhaps most illuminatinglyto Buganda. The
increase of monarchicalpower was measured by the degree to which clan election
or aristocraticdescent were usurped by patrimonialappointment (Fallers, 1959;
Southwold, 1961). Such functionalism, however, was really the language of
incumbents in power, the terminologyof self-justificationrather than an analysis
of reality. It could be most misleading, even when the analysisof power remained
at the symmetrical level, as Southwold (1968) found when re-examining his
Buganda material. For the chiefs whom he had earlier seen as coming more
closely under the thumb of individual kings were acquiring,instead, a collective
interest in the power of the monarchy as an institution (contrast especially
Southwold, 1961: 4, with 1968: 149). What was once seen as an autonomous king
now looked more like an instrumentalstate.
This shift in perception turns out to be in fact a pervasive model of the
changing relationshipbetween state and class in Africanist writings. It can be as
well applied to precolonial kingdoms, colonies, or postcolonial states. Briefly, it
seems that revolutions in government, surges of politicalcompetitionwhich result
in new coalitions of power, may provide the conditions for the formation of
classes whose perceivedinterests then increasinglydiverge from the politicalorder
which gave them birth. Asante provides the classic example for eighteenth-
century Africa, with its Kwadwoan revolution in government (Wilks, 1975);
Bradbury (1969: 28) has suggested that a similar upheaval occurred under

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
158 AFRICAN STUDIES REVIEW

Ewuare, a fifteenth-centuryOba of Benin. In precolonial Africa, and there are

many such cases, it seems always to have been militaryexpansion which provided
not only the new administrativechallenges, but also the increased rewardswith
which to govern. Public rewardmight then be converted to privategain. Colonial
Africa's successive "crises of displacement" of native elites, on the other hand
(Stokes, 1970: 103), were caused at least as much by the changing purposes of
colonial government as by the tendency of its politics of collaborationto provide
Africans with ever greater resources of class formation. In other words, not only
can classes consolidate after organizationalrevolutions, but states go on changing.
This double process has been most fully worked out by Greenberg (1980), with
reference to twentieth-century South Africa. At the turn of the century, its
revolution in government was a necessary condition for the intensification of
capitalist exploitation. Harsher race relations reflected more systematic labor
control. Half-formed classes among white South Africans matured in the
coalition-building which more intensive government evoked. But for some
interests, their class reproduction has more recently appeared, so Greenberg
argues, to be not only independent of but endangered by the coercions of the
state, whose institutions have become self-interestedlygeared to racialrepression
for its own sake rather than as an adjunct to capitalisthegemony. Whereas white
manufacturers may hesitantly see some advantage in allowing otherwise
unpredictablyangry African workers to become a legitimate estate, such a canny
concession is anathema to the bureaucracieswhich administerapartheid.A similar
tension between the interests of the state as the core of a power-blocand the
classes whose dominance rests upon it has often been seen in postcolonialAfrica.
In extreme cases, states have broken up into "kleptocracies" whose rivalries
strangle any possibility of the accumulation they seek (Andreski, 1968). Some
revolutions in government do not work, as will be discussed furtherbelow.
The autonomy or instrumentalityof a state can now be seen to be a comment
upon the degree of class consciousness in society or the capacityof classes to
reproduce themselves in conflict. Dominant classes are formed, or reconstituted,
by fighting for their share in the revolution in government, whose political
dynamics have a certain autonomy; stabilityonce achieved (cf. Plumb, 1969: 13),
self-conscious classes will thereafter seek to impose their competing
instrumentalitieson the state.14 I have already argued that it is not possible for
any dominant class to succeed in this endeavor, at least in any outright sense, in
any state which has an historical constitution rather than an abstract rationale.
The ambiguities of instrumentalityare nowhere better illustrated than in the
cultural codes of states' whole histories of past conflict continuously restated in
the rule of law and the ideologies of power.
Some years ago, Jack Goody (1963: 11) complained that there was very little
African legal history. The situation does not seem to be very greatly improved
even now, and it is treacherousground for the non-specialist.One of the marks
which distinguishes states from stateless societies is the application of law as
enforceableadjudicationratherthan socially sanctionedmediationbetween parties.
One of the great state-buildersof precolonialAfrica, Dingiswayo,justified all his
bloodletting by his wish "to do away with the incessant quarrels that occurred
among the tribes, because no supreme head was over them to say who was right
or who was wrong" (Service, 1975: 108f). But it was not as simple as that. If
states are upheld by coalitions of estates, then there must always be difficultyin
reconciling impartialadjudicationwith the diversities of interest between those

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 159

who most nearly support the judge (e.g., Gluckman, 1955: 36f; Ruel, 1969). That
is one reason why a rule of law is so difficult to establish yet so necessary to
sustain, if the state is not to be wrackedby the dissensions of its power-holders.
But there is a second reason, the obverse of the first. In all societies, presumably,
a definition of a "big man" is that he can break social rules and yet be admired
(S. Roberts, 1979: 33). But where a man is "big" by virtue of his office, then to
break the rules is also to subvert a division of power. States are governed by
officials, not by big men. The rule of law is therefore also a means by which to
limit the discretion of subordinates. As the Asantehene upbraidedan unduly
enterprising traveling commissioner: "no man must dare to do good out of his
own head" (Wilks, 1967: 225). Paradoxically,it is precisely this constrainton his
arbitrarypower which robes even the most junior officialin all the majestyof the
state (Churchill, 1908: 26f). Conversely, the rule of law, if it is not to mock the
legitimacyof the state, must also act to curb, if only sometimes, the selfishness of
particularruling interests (cf. Thompson, 1977: 262-69). Where it does not,
then -under the competing rapacitiesof power-the people may well sigh "King
rules or barons rule; we have suffered various oppression" (Eliot, 1938: 12); and
the moralityof dominationhas gone.
The need for a rule of law is not the only reason why the culture, ideology,
even theology of power are always wrappedin the "wax and gold" of doubtful
ambiguity (Levine, 1965). There is also the pragmaticconsiderationfor power-
holders that the more evasive a concept may be, whether a spiritual one like
baraka (Gellner, 1973: 39-42) or a political one like "multi-racialpartnership"
(Low and Lonsdale, 1976: 52), so much the more magicallyvalid it can seem to
be through all contingencies. Further, in the chaotic geology of every state's
asymmetries of power, successive methods of ordering productive relations get
hallowed by custom; there may well be mutually incompatiblerules of access to
resources therefore, a thicket of conflicting propertyrights, and divergent moral
communities. If one takes Africa's lineages, for example, and accepts what non-
Marxist anthropologists seem increasingly to doubt, that their material
reproductionwas once in fact structured by kinship relations, then that kinship
idiom of communal reciprocitycan become "a powerful force in its own right, a
force capable of surviving to some extent the conditions that brought it into
being" (Horton, 1971: 83f). This "happily pluralist" culture of much of Africa
(Dunn and Robertson, 1973: 101) is also due in some part to the stratified
geology of colonization and invasion by successive ethnic populations and the
symbioses of their different modes of subsistence. Popular religion is also an
historicallylayeredcomposite for the same reason, with the ancestralfertilitycults
of subjects not infrequently represented in the councils of intrusive rulers (Van
Binsbergen, 1981: esp. 18ff). Religious and ideological incongruities may be
attributableto some or all of these factors in any given society, and in much of
precolonial Africa the absence of literacy will very likely have precluded that
"cumulative traditionof criticaldiscusion" which exposes inconsistenciesin belief
(Goody, 1977: 47).
The problem of ideology in a state, however, is more fundamentalthan all of
these. It has to do with rule. Rule over others must always be an essentially
contested phenomenon. This is primarilya problem for rulers rather than the
ruled, for whom deference is probablyat least as effective a mystificationof the
ruler as his hegemony is of the ruled. No ruling group can escape the politicaland
indeed moral dilemma of reconcilingprivate privilege with public function. They

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 160 AFRICAN STUDIESREVIEW

need "a recipe for domination,and at the same time an alibi" (Berque, 1967: 77).
The oldest and still the most respected African secondary school in Kenya, the
Alliance High School, is gracedwith the motto "Strong to Serve." It was founded
by British protestantmissionaries, some of them the productof English public (in
fact, private) schools, the nurseryof the elite, at which a favorite text for visiting
preachersis (or used to be) "unto whomsoever much is given, of him shall much
be required" (Luke xii:48). A private elite culture which nurtures cohesion,
reliability,and restraintin rulers is essential to the exercise of power; delegation
of authority is otherwise impossible (Berman, 1973: 32ff, 88-113; A. Cohen,
1981). Class consciousness is actuallynecessary at the top of the heap, while only
contingent at the bottom. But their culture and ideology can never be entirely
privateto a ruling class for, if only for reasons of economy, rulers like to be loved
(Dunn and Robertson, 1976: 167f). If in order to curry favor they proclaimtheir
purpose to be the service of their subjects, then they must expect to be taken at
their word. They may even believe (be interpellated by) their own alibi. The
moral queasiness of hegemony can be a serious matter. Not only does it put
doubts in the minds of rulers, it lends legitimacyto the complaintsof the ruled. It
was perhaps particularlyimportantduring the colonial period, when both rulers
and subjects actually came to believe in the myth of the civilizing mission-and
found its practice wanting (Robinson, 1979; 1980). Consensus among rulers is
more crucial than the consent of the ruled in preserving a particularpattern of
power (Therborn, 1980: 109). The perplexities of the powerful are a necessary
condition for those creative reformulations of political obligation by which so
often historicalchange has been given its shape.
The uncertainties of rule must also be the chief reason why all states are
theaters of power. They dramatize the core values of a society, whether
precapitalist(Prins, 1980) or capitalist,so that one set of possible social relations
is elevated to a universal standardof civilization (Corriganet al., 1980: 15). This
is one salutaryeffect, but the first purposeof the rituals of power is surely to give
confidence to the actors, rulers. The colonial rulers of Kenya must have been in
sore need of the reassuring orderliness of ceremonial to have left five hundred
files on the subject in the Nairobi archives (Berman, 1973: 113)! But successful
theater demands audience participation.About the most spectacularcelebrationof
King George VI's coronation in 1937, anywhere in the Empire, must have been
staged in Kampalaby the young EdwardTwining. Among its tableaux was one
entitled "From Savage to Soldier, showing the process of turning out real
soldiers.... Everythinghad gone without a hitch and without any delay and most
amazing of all no Europeanhad been in the arena the whole evening" (Twining
to his mother, in Bates, 1972: 102ff). All imperialpageantryrests uneasily on this
sort of ambiguity (cf. Ranger, 1980; Lonsdale, 1981: 113f), but even Marx would
not have expected the ruling bourgeoisie to take such pride in trainingtheir own
grave-diggers.
Students of Africa have accordinglybeen able to draw somewhat conflicting
conclusions from consideringthe dramaturgyof power. Certainly,all power save
brute violence is elaborated through symbols. Hegemonies are established;
alternativevisions of social order are difficultto formulateand still more difficult
to establish. Nevertheless, symbols are subject to conflictinginterpretations;they
can be shared very unequally; they can be channels of revision as much as
replication.The parable"From Savage to Soldier" is clear warningof the perils of
delegating power, armed might especially, to alien slaves or mercenaries whose

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 161

cultural perceptionsare literallyforeign. Rulers throughout African history might

ruefully agree, whether in precolonial Segu or Oyo (Bazin, 1975; Law, 1977a:
183-200, 250-60); or in colonial Africa, apprehensivelyawaiting the return of
demobilized soldiers after the Second World War; or even Idi Amin, when
contemplatingthe excesses of his Nubians. Yet to share out the magic of civil
power may equally be to create competitors rather than allies. Colonial Africa
provides one of the clearest examples. If the paraphernaliaof bureaucracywas
understood to be its ritualcode (Miller, 1977), then those who first decipheredit,
lawyers and other literati, had obvious claims to be colonialism's most effective
usurpers (Hodgkin, 1956: 140). But for most of history, that is too sharp a
distinction, between incumbentsand insurgents.The conceptualcodes and models
of power are more normally the domain of successive "working
misunderstandings"(Dorward, 1974); for the rulers' need to categorizepeople in
order to know how best to dominate them gives to their subjectsan inexhaustible
resource of (mis)informationto manipulate-either to protectthemselves from or
to gain a purchaseon the interventionof the state. Rulers on stage are continually
plaguedby mimics, understudies,and false cues.
All these possibilities are of vital relevance to the processes which link class
formation and state consolidation and, to begin to move the survey and the
argument on, the problems of coercion and consent. The causes of the
organizationalrevolutions which, often after long periods of prior "incubation"
(Steinhart, 1981: 128), have periodically thrown up new or caused the
reconstitution of old political orders are almost infinite in variety and endlessly
disputed (R. Cohen, 1978). The disorganizationof legitimaciesafter defeat in war
(Therborn, 1980: 106); the availabilityof new means of destruction, with new
requirementsfor supply or battlefieldcontrol (e.g., from a fast-growingliterature:
Goody, 1971: ch. 3; Inikori, 1977; Kea, 1973; Law, 1980; Low, 1975; Smaldone,
1977); the free-floating resources, as already suggested, created by military
expansion, whether in land or slaves, which allow or force a redivision of power;
the oppositions stirredup by rulers who breachan existing rule of law in order to
recoup from their subjects revenues which they had previouslyderived from war
or trade (Klein, 1972; Bonner, 1980: 88-92); examples of all these and many
more can be found in African history. At issue now, though, is the outcome of
these organizationalrevolutions, the productivityand stability of the regimes to
which they give rise. merely because they appearto have become necessary they
will not necessarily happen nor, if they happen, will they necessarilysucceed. If
politicscan be productive,its counter-productivityhas to be allowedfor, too.
In Africa's history, as in all history, productiveorganizationalrevolutions have
tended to be factional rather than subversive in their drive, revolutions in
government rather than social revolutions. To revert to and then move forward
from the class conflict paradigmof the state, they have succeeded insofar as
dominant elites, each with its separatebase of power, became fused together as a
ruling class. Revolutions in government were moments of sudden consolidation
around a composite answer to disparateneeds. It is at this point that the theater
of power differentiates dominant from dominated classes. Powerful groups are
those which are more cohesive than others, more self-conscious. Social
consciousness is shaped mostly by one's work (cf. Iliffe, 1970); and the
specificitiesof the work process, whether in field, down mine, behind desk, divide
more than they unite. So coalitions of interest have to be built by politicalaction,
they are not determined by differentiated class position, ruling is itself a

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
162 AFRICAN STUDIESREVIEW

specializedform of labor. Its working conditions are unique in that they not only
impose cohesion on its practitionersbut at the same time make any independent
cohesion more difficult for those outside the ruling circle, for the rule of law
disciplines its exponents. The dominant classes get organized by the very
institutions which their conflicts have created; and by the same means, the
illegitimacy of disobedience is inculcated (Lamb, 1975; Parkin, 1979). Ready
access to state institutions is therefore literally what makes classes dominant; it
economizes their class effort with an external constraint. Institution-buildingis
therefore more correctlyconstrued as class-formationthan politicaldevelopment.
It is an exclusionaryprocess before it is an integrativeone. Dominant classes have
internal access to the state apparatusas its operators; dominated classes have
external access only, as supplicants.If there were not this differentialin the costs
of class formation, there would be little point in isolatingthe state as an object of
study. CardinalRichelieu best summarizedthe advantageof internalover external
access in the dramatizationof social reality. "In politics," he said, "the weak are
alwayswrong."
To leave the matter with Richelieu's dictum, however, would be to forget the
workingmisunderstandingsof power. The very institutionsby which it is intended
that they should be controlled, like the trade unions or compulsory producer
cooperatives (more properly"coercitives") of colonial Africa, can providea focus
of aggregationfor the dominated, with the possibilityeven of becoming legitimate
estates. The perpetualchanges of state-formationare thereforenot to be separated
from the social processes of class-formation.They are not to be separatedfrom
tribe-formationeither, as argued in my earlier critique of the cultural pluralist
model of state instrumentalism.Most historians would probablynow accept that
Africans have rarelylived in self-conscious tribes save where they have also lived
under, or in uncomfortablyclose proximityto states (cf. Southall, 1970). But the
nature of this relationship is hotly contested, between symmetrical and
asymmetricalanalyses, as is perhaps best illustratedby reference to precolonial
Morocco. Gellner (1969: 6) saw state and tribe as a "tragic antithesis between
civilizationand social cohesion." The civilized power of the towns rested upon the
social cohesion of tribes, some of which were armed to support the dynasty
againstand extract tax from other tribes, the dissidents. Tribes were the corporate
constituencies of access to and defense from the exercise of state power. Seddon
(1981) has argued, to the contrary,that tribe was an ideologicalconstruct, not a
social organism, a mask for factionalmanipulationby big men who ruled not over
co-equal tribesmen but differentiatedpeasantries.Privateexploitationby landlords
flourishedwithin the exactions of dynastictribute. Triballeaders had much more
in common with each other than with their own tribesmen.In Gellner's view, the
state existed as a form of intervention between tribes; in Seddon's, as the power
behind differentiationwithin tribes.
This is an important difference of opinion over the question of what
distinction, if any, can usefully be drawnbetween state and society. There is more
than one answer, dependingon the configurationof modes of productionpresided
over by a state. This is a matter shortly to be discussed. The examples so far cited
for colonial Africa and precolonial Morocco suggest that for states where
capitalism is dominant, as merchant capital in these instances, the division
between state and society is to be found not between institutionalcomplexes, one
set being of the state, the other self-regulatingwithin society, but ratherwithin all
such institutions. No matterwhether they be trade unions or tribes, they are torn

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 163

between obligation to the state and responsibility to their members (compare

Gluckman, 1955: 50ff, with Middlemass, 1979). It could be said, therefore, that
successful organizationalrevolutions were carried forward on this fragile and
contested linkage, without which the working misunderstandingsof power would
have become unworkable.But there is a corollary,that dominatedclasses too are
partiallydefined by the nature of their institutionalaccess to the state (Poulantzas,
1973: ch. 2). Africa's nationalist leaders found that the minimallyrepresentative
structuresof colonial rule could economize even subversive effort (e.g., Zolberg,
1966). Thus, the process of state and class formation can become linked, through
the apparatusof control which is also a channel of representation,to the growth
of nations.
It was the radical pessimist view, epitomized by Frantz Fanon, that
decolonization was a panicky revolution in government settled round a "green
baize table," not a social revolution. "They decolonize at such a rate that they
impose independence on Houphoubt-Boigny" (1965: 52, 55). But even
constitutionalnationalists, from their position as a half-acknowledgedestate, had
first to make the quite astoundingannouncement that the drama of colonial rule
was in fact an ugly farce. Among those to whom access to state institutions is
most rigorouslydenied, it must be true that imaginativeideological labor is still
more essential to political cohesion than it was to Africa's constitutional
nationalists.Among Africa's stateless people in the past or peoples made stateless
by Europeanconquest, it has often been argued, the enormous effort needed to
combine forces against states could be mobilized only through visionaryreligious
fervor (e.g., Sheppersonand Price, 1958; Ranger, 1967; Gellner, 1979; Isaacman,
1976: 193; Iliffe, 1979: 65, 168-202; Maier, 1981). More recently, in white-
dominated southern Africa, where Africans were and still are most harshly
excluded from power, the ideology of their nationalisms has of necessity been
vastly more coherent in the radicalismof its societal premises than the ideology of
constitutional nationalism elsewhere. Three reasons are generally given. People's
war can enjoy none of the economy of effort which was, however grudgingly,
afforded to mass nationalism by the administrations of colonial states. An
infinitely greater degree of sacrifice is risked by its active supporters. Most
significantlyof all, its leadershipwas and is forced to "put itself to school with the
people" (Fanon, 1965: 122). Far from achieving a mere working
misunderstandingwith the peasantry, leaders were obliged to take the peasants'
part fully and from within rural society, to meet the concrete needs of the
liberated zones (from a huge literature:Saul, 1974b; Davidson, 1977; Chabal,
1981; and more skeptically on the radicalizationof leadership, Ranger, 1982).
This process was completely to rewrite the dramaof power:the chorus moved to
center-stage. As Cabral put it, and he ought to have known, the conditions for
national liberationbelong partlyin "the sphere of morals" (Cabral,1969: 90). It
remains to be seen whether the organizationof power in the liberatedstates will
be as productivelyopen as the leadershipof people's war was obliged to be.
Quite what is in question here becomes clearer, as always, when one considers
failure, or the counter-productivityof politics. To quote the Emperor Haile
Selassie, another man with good cause to know what he was talkingabout, "Trees
that are planted do not always bear the desired fruit" (Hess, 1970: 141).
Organizationalrevolutions can fail in any one or more of at least four different
ways. They can fail to establish a rule of law, most obviously so in colonial Africa,
where at least two legal systems co-existed, indigenous and imported,an appealto

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 164 AFRICAN STUDIESREVIEW

either being a subversive of the other (Barnes, 1969); where the rationalizations
which the British might see as forging the organiclinks of IndirectRule were seen
by the Hausa villagers who endured them as, rather, yayyaga, a "tearing
asunder" (Hill, 1977: 48); or where populations had to learn afresh each
unpredictable"white man's madness" or wazimuwa mzunguwith every change of
districtcommissioner (Liebenow, 1971: 143).
Deliberate changes in politicalrules of access may, secondly, destroy not only
the social bases of the power of rivals, but, with them, the possibility of any
socially structured transactionsat all. As one of colonial Africa's most quoted
governors, Jules Cambon, reflected in 1894 on the French determination to
destroy the great families of Algeria: "We did not realize that in suppressingthe
forces of resistance ... we were also suppressingour means of action. The result
is that we are today confronted by a sort of human dust on which we have no
influence and in which movements take place which are to us unknown. We no
longer have any authoritative intermediaries between ourselves and the
indigenous population" (in Quandt, 1969: 5). In modern Egypt, similarly,
Nasser's Free Officers found that they could replace the power of landlords,
guilds, and religious brotherhoodsonly with an expanded militarybureaucracy.In
destroying their rivals, they had destroyed all politics. Without some form of
"mitigating intercession" between the state and Egyptian society, any effort at
mobilization ran into the sand (Vatikiotis, 1968). Obote followed a strategy of
"ethnic departicipation"in politics only to find that there was no other idiom of
political participationin Uganda (Kasfir, 1976; Low, 1971: 246-50). Nkrumah
discovered that Ghana's small farmers liked his bureaucraticagents, supposedly
their allies, even less than the big farmers and creditors of the countryside. At
least the latter were subject, as bureaucratswere not, to some of rural society's
sanctions and the need to invest in local reputation(Beckman, 1976).
Freund has linked this failure with a third, in Tanzania.Villagizationthere has
in his view destroyed (as bureaucratizedmarketing did in Ghana) "a nexus of
long-standingsocial relations." More than this, however, Tanzaniacould also be
said to have failed to produce a coherent ruling-class if, as many Tanzanians
themselves feel, the main barrierto managerialefficiency is "non-delegation of
authority" (1981: 491, 493). All of these failures are generally summarized in
political science by a fourth, a failure in politicaldevelopment, so that claims on
government are not channeled by state institutions but multiply outside or in
opposition to them. The mobilization of ethnicity has been the most obvious
indicator(Yung, 1976; Kasfir, 1976)
The productivityor otherwise of the politics of organizationalrevolutions is a
key issue if states are consideredas elaboratedforms of unequal social relationsof
production.From this perspective, the foregoing discussion of state and class has
been about the more or less successful construction of a hegemonic bloc of
power. But how do dominant classes extract value from the dominated and
control their labor? What degree of coercion are they able to muster, how do they
secure consent? The answers to this question point to the differences between
states. Hitherto, the argumenthas been more about their similarities.
To descend from symmetricalpolitics to the asymmetricalbases of power and
modes of production is to enter on to a recently discovered and imperfectly
charted theoretical minefield. Any mode of productionis a theoreticalabstraction
from the processes of toil and the appropriationof its productfor the purposesof
social reproduction.Quite how deep a level of abstractionit represents and in

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 165

what ways its necessary conditions of existence are also the conditions for its
antagonistic development and change are, as noted briefly already, sources of
division between Marxists (G. Cohen, 1979; 79-84; Copans and Seddon, 1978;
Law, 1978b; Wolpe, 1980; Kahn and Llobera, 1980; Kitching, 1980: 3-6). Even
the most ambitious of theoreticianshave doubted if answers can be given at the
level of general theory (Poulantzas, 1978; Bernstein and Depelchin, 1978). It is a
gross oversimplification,but nonetheless I hope helpful, to suggest that there are
three angles of attack on the problems involved, or varying emphases on the
three connected but semi-independent logics within any mode of production:
productive logic, political, and class. Through a brief consideration of each, it
should be possible to clarify some of the major distinctions between precapitalist
and capitalistmodes of productionand, with them, the changingchallenges in the
exercise of and oppositions to state power.
It was through productivelogicthat modes of production theory first came to
African studies, what Hindess and Hirst (1975: 61) condemned as the "technicist
conception of the structureof the economic level." Meillassoux (1972) and other
French Marxist anthropologists who have followed him sought a material
explanation for the kinship structures of society which pervaded much of
precolonial Africa. Agriculture,he argued, was to be distinguishedfrom hunting
and gathering (with their very direct appropriationfrom nature) by the necessity
to reproducelaborersand the fruits of their labor through time-at least from one
harvest season to the next. Domestic households could not therefore be
reproduced in their isolations but within structured social relations which
guaranteedaccess to land, to seed grain, and to women. The elders who controled
such lineage structures were scarcely a class, since their replacement by their
dependent cadets was in the nature of things, so that the legitimacyof the social
order was not seriously contested. Goody (1976) employed a similar perspective
on a global level, to explore the social-structuralcorollaries of Africa's hoe
agriculture.The hoe, wielded largely by womanpower,led to comparativelyopen
marriagestrategies in the past, since there were no great productivedifferentials
to be protected by class-endogamy.Furthermore,it was Africa's low productivity
which, within this perspective,governed the natureof its precolonialkingdoms.
Without the competitive market pressures of capitalism, there was little
incentive in precapitalist production to intensify productive methods by
technologicalchange. More could be producedonly by workingharder.It was not
easy for would-be exploiters to achieve this end, especiallywhere productionwas
already geared to the reproduction of lineages. Yet, pre-colonial kingdoms
nonetheless existed, handling quite large flows of goods and slaves through their
palaces. Because their peasant subjects were so difficult to exploit, Coquery-
Vidrovitch (1969) argued that African kings could maintain the apparatus of
power only through tribute exacted from peripheralareas, to be exchanged for
prestige goods or the weapons of domination through long-distancetrade. Surplus
precapitalistproduction depended therefore on very direct forms of coercion, or
else it was not forthcoming at all. Kings hardly dared to test to the consent of
their peasants. Power was decentralized, personal, and difficult to exert; in
precolonial Africa, therefore, there were kingdoms, not states, except insofar as
"state" is a useful generic term distinguishedfrom "non-state."
The productive logic of capitalism, capital logic, could scarcely be more
different, and its difference as a mode of productioncan be argued to demand an
entirely different mode of domination (Anderson, 1974). The more the direct

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
166 AFRICAN STUDIES REVIEW

producers, workers, have been separated from ownership of their means of

production and their labor commoditized for sale on the market, so much the
more productiveunits in which they are combined by capitalhave been obliged to
compete in the market for finished goods. Politicallyprivilegedcontrol over either
market has become a "market imperfection" or "unfair competition." Thus,
impersonal forms of exploitation and control are called for, contractualequalities
rather than status restrictions (Greenberg, 1980: 148). States and bureaucracies
are on the cards, not the personalrelationsof kingdoms.What is generallyseen to
be a necessary characteristicof states, a monopoly in force, becomes possible for
the first time, since the militarycan now live on its pay and not off the land. Here
is firm theoreticalground for the starknessof state analysis in the modern world,
strippeddown to simplifiedrelations of coercion, however mystified they may be
through the representativeinstitutionsof consent.
At the opposite end of the modes of productionspectrumto the functionalism
of productive logic stands the autonomist insistence on politicallogic, used with
reference to precapitalisthistory by scholars who would not consider themselves
to be writing within the modes of productioncanon at all. Vansina is perhapsits
chief exponent; he stresses the political determinants of productivity,with the
tributarydemands of the equatorialforest kingdoms obliging households to labor
more intensively and to plant higher-yieldinggrains, maize rather than sorghum
(1966: 5; 1978: 172-86). This emphasis reflects a growing tendency in non-
Marxist African anthropologyto break away from ascriptivedeterminantsof local
community to the fluid, indeterminate manipulationof resource-controlby big
men with clients rather than elders with lineages (Guyer, 1981). New Guinea
models of multiple affiliationare replacingAfricanmodels of lineage solidarity(cf.
Barnes, 1962). It is a perspective which makes the emergence of precolonial
kingdoms less difficultto explain.
With capitalism so different from any preceding economic system, it is not
surprisingto find that this politicallogic can be turned upside down to become a
functional condition for the capitalistmode of production.The inner apparatusof
the capitaliststate can be argued to stand some distance apart from the needs of
capitalists, the better to serve them. This relative autonomy of the state echoes
Rousseau's thinking. The state comes to representthe General Will of Capital,as
against the competitive and mutually destructive will of all capitalists, for
capitalism has been a profoundly subversive force. Its commoditization of the
factors of production,labor, materials,even land, has freed them from the grip of
political authority, whether within states or, increasingly,internationally.Despite
this division between the political and the economic under capitalism,so unlike
the precapitalistera, state power has proved essential to capitalismas its General
Will, in at least three respects (e.g., Murray, 1971; Jessop, 1977; Holloway and
Picciotto, 1978; Berman, 1981). Capitalists needed from the state the legal
structure of economic liberalization. Aristocratic privileges and the communal
rights of the poor alike had to be outlawed. Capitalistsalso secured the legislation
which orchestratedtheir needs for labor, credit, their own particularprivileges in
tariffs, and so on. Finally, employers have often had forced upon them the
"consensual interventions" of the state (Murray,1971: 98). These have provided
for the social wages of workers in health care, education, and so on from out of
the general taxable fund of capital's profits. They have nevertheless lifted from
the backs of individualcapitaliststhe reproductivecosts of their labor and bought
off its discontents.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 167

These are all heavy duties, more difficult to carry out in Africa, whether
colonial or postcolonial, than in the capitalisthomelands of the West. In Africa,
capitalism's"drasticsimplifications"(Oliver, 1963: 456) were wroughtby external
conquest, not domestic liberalization.The orchestrationof an African workforce
needed harsh state interventions where labor had barely begun to be
commoditized (e.g., Marks and Trapido, 1979; Berman and Lonsdale, 1980;
Cooper, 1980). The orchestratedreconciliationof capitalism'scompetingfractions
in settler Africa was no less explosive a politicalprocess. It accountedin large part
for the modern revival of Afrikaner nationalism (O'Meara, 1977; 1978). The
labor requirements of white farmers, mine-owners, and traders, were very
divergent; each implied a different underlyingracialorder of control (Greenberg,
1980). In the contemporaryworld of independent Africa, there is the different
considerationthat each state is only one of many weak contendersfor capitalism's
interest, all of its fractions powerfully based overseas. Finally, consensual
intervention was not, clearly, something at which colonial regimes could be very
convincing. Their multi-cultural inheritance has not made it much easier for
independentAfrica's "managerialbourgeoisies" (Meillassoux, 1970; Sklar, 1979).
Class logic, as argued earlier, intervenes between and breaks up the
functionalisms of both approachesto capitalism and questions the autonomous
power of politics in the precapitalistera. Conflict over the labor process mainly
determines the degree and type of investment which the proprietaryclass is
obliged to undertake in order to maintain or increase its share of the social
product (Brenner, 1977). Marx was succinct: "No antagonism, no progress" (in
G. Cohen, 1978: 207). Africa's history raises the insistent question whether its
common people have not been better able than most of their kind around the
world to resist the demands of those who exploit them. African villagers have
been known to complain that they would be better off if left to their own devices,
without the interferenceof the state (e.g., Mbee, 1965: 199; Wrigley, 1976: 516).
This, and they may increasinglyrealize it, is really an expression of class opinion.
Certainly, precolonial African kingdoms were plagued by secessions beyond the
reach of tribute raids, colonial states by tax-evasion across frontiers, and
independentstates now by smuggling.
The resistance of the poor to the exactions of the rich may be greatlyassisted
by what is known as the articulation of modes of production, conveniently
summarized by Van Binsbergen (1981: 43) as an economic conjuncture far
removed from the mutuallybeneficent symbioses of, say, farmingand herding, in
which the "surpluses generated in one mode are expropriatedso as to serve the
reproductionof another, dominant mode." But that is only one side of the coin.
Equally, the resources of one mode, say domestic agriculture,can be used to
resist the expropriationsof the dominant mode -in modern times, capitalism.The
enjoyment of land rights and reciprocallabor services outside the market can be
pitted against the exactions of politicallysupervisedcommodityand labor markets.
Conversely, the ability to sell one's labor to capitalistsmay make it the easier to
flout the claims of chief or parent-not that it is common to regardmigrantlabor
as a form of emancipation (see, further, Rey, 1973; Cooper, 1982). The
opportunities for popular resistance by mobility between and withdrawalfrom
marketshave not yet been extinguishedand may in some partsof Africa even be
increasing.
Dominant classes have had much less freedom of maneuver. Warriorclasses
had little alternative but to reproducethemselves through war; traderscould not

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
168 AFRICAN STUDIES REVIEW

survive without regular supplies of commodities, nor employers without labor.

The history of states in Africa could be said to have been the history of their
collusion together the better to capturethe resources of the poor (Wrigley,1971).
It has not been easy. Societal survival has set a double curb on the self-interested
exercise of power. The symmetricalconflicts between those who have sought to
profit, no less than the asymmetricalstruggle of those who have resisted, have
both helped to shape the culturalcodes of a rule of law. States, it is true, need
satisfied ruling classes, as much as dominant classes have needed states. It is
arguable that the need of the common people is as great. States organize the
upward flow of resources (MacGaffey, 1970), but that may be preferable to
disorganizedflows. If the king rules, he may afford some protection against the
barons, as peasant legitimists have believed from Wyatt in 1381 to the Ganda
bakopi in the 1950s (Low, 1964). If therefore class antagonism does result in
progress, and in the 1980s that must increasinglybe in question, then that is not
least because in controllingthe competitions for power states have tended not to
leave their dominant classes to their own devices either, but to mediate between
them to secure a more reproductivelyproductiverelationshipwith the poor.
It is in this struggle between power and endurance that one of the stock
debates in African studies, between formalist and substantivistinterpretationsof
socio-economic behavior, ought finally to be resolved. Both positions were
idealist, the one stressing economic individualism and market rationality
unhampered by political distraction, the other the primacyof moral, social, and
political restraints upon self-interested economic activity. But the same restraints
which protect an "economy of affection" among the poor (Hyden, 1980) will, in
the hands of the powerful, foster an economy of authority (MacGaffey, 1970:
262f). Community can justify either defense against power or its totalitarian
exercise; it depends upon who is speaking.The same must be true of free market
enterprise.It is all very well for the capitalistif the law has first endowed peasants
with enough rope to hang themselves in negotiable title to land, but not so good
if peasant enterprise can still be put to freeing themselves from a particular
market. Market and community are, simply, different ideologies of social
mobilization.Each is certainlysubversive of the other, but only in their dynamics,
not because they inhabit differentmoral universes. What was the first thing that a
Lozi commoner did when civil war disruptedthe fabricof power in 1884? He sold
ivory direct to traders,not through his chief; and his chief ruled with a much less
certainhand (Prins, 1980: 134).
This brings us back to the relationshipbetween coercion and consent and two
final considerations, one within the class conflict approach to state power, the
other within the discourse which contrasts state and constituent communities, or
the state and plural society. If one asks, first, what it is that constitutes the
relative autonomy of the state in capitalist society, then the answer in mature
capitalism must stress the compromises of self-interest imposed upon any one
sector of capital by the sheer problem of coalition building within a liberal
frameworkof representation.The relative autonomy of the state is a consequence
therefore of capitalist hegemony as much as its guarantee. If, however, the
hegemony of capitalismand the power of dominant classes are weak, as in the
Africa of articulatedmodes of production,then the relative autonomy of the state
must take a more deliberate form. In Marxist analysis, this is termed
Bonapartism, after Marx's account (1954) of Louis Bonaparte's military-
bureaucraticregime, which rescued the French bourgeoisie from the popular

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 169

consequences of the revolution which they themselves had instigatedin 1848. The
political elite was a corporate grouping separate from the "actual classes of
society." In terms of its own politicallogic, it sought to constrainbourgeois power
even while the productive logic of the day was ineluctably promoting capitalist
economic relations. It is this distinction between the contingenciesof politics and
the inexorabilityof the laws of capitalistexpansion which has made Bonapartism
so attractive a model for analyses of postcolonial Africa (Leys, 1975; 207-12;
Davies et al., 1976; Langdon, 1977). It is a concept which can comprehendboth
the idea of the overdeveloped state removed above the conflicts between Africa's
weakly formed classes, the alliances between managerial bourgeoisies and the
military which have featured so largely in recent years (Markovitz, 1977: 310),
and yet also the fact that despite all the politicalturbulencea workingalliance has
been establishedbetween local and multinationalcapital.
It may, however, be a concept at once too partial and too abstract. A
Bonapartiststate elite is not after all so differentfrom the "nationalmandarinate"
of Young's (1976) analysis of the cultural autonomy of states which engineer a
national ethnicity from out of a maze of parochial ethnicities, and whose
reproductivebase lies in the very real networks of education, administration,and
inter-marriagerather than in the abstractneeds of capital.Power does have to be
organized, by a mixture of alliance and expropriation.It is not a new problem in
Africa's history. Probablya majorityof Africa's myths of dynasticorigin aver that
her kings were aliens, separatetherefore from the actual classes of society. Many
of them were uncouth, dangerousmen, drunks, Tom Paine's "crownedruffians."
They were hairy hunters, outcast magicians, sometimes incestuous lovers, men
like Rukidi, Mbegha, Cibinda Ilunga, or Shyaam (de Heusch, 1972; Wrigley,
1973; Feierman, 1974: ch. 2; Miller, 1976: 129ff; Vansina, 1978: 59-63; Neale,
1980: 413-57). They were outside the kinship relations of agriculturalproduction,
sometimes scandalouslyso, the better perhapsto arbitratebetween them. In myth
too they had various relatively autonomous bases of support for this relatively
autonomous function. In examining the claimed hunter origins of dynasties in
middle Nigeria, Boston (1964) pointed out that the successful hunter depended
on individualprowess to enlist men in wider patternsof cooperationthan allowed
by kinship. In Tio society friendships formed in hunting associations were more
reliable than the bonds of kin (Vansina, 1973: 81ff, 121-28). Bambaratraditions
disputedwhether their dynastywas based on a hunting associationor the revolt of
a young warrior group (Roberts, 1980b: 403ff). These were possible sectional
bases of support from within existing relations.The growingliteratureon African
slavery (e.g., Meillassoux, 1975; Miers and Kopytoff, 1977; esp. chs. by Miller
and MacGaffey;Cooper, 1979) stresses, rather, the utility of external resourcesof
power. Kinless outsiders, refugees from famine, pawns and debtors as well as
captives, could be crucial additional supporters, free from the constitutional
restraintsin establishedmodes of authority, "gens sans feu ni lieu" (Bazin, 1975:
1968). Slaves indeed sound rather like the transitionalsof modernizationtheory,
once seen to be so vital as the raw material of nationalist mobilization. Men
shaken out of their traditionalcommunity who had yet to find a modern role,
they have since found one among the ranks of the managerialbourgeoisies.Their
profit lies in their power to allocate permits and projects.As slave warriorsdid so
often in the past, these modern tonjonor tyeddohave become one of the actual
classes in society, indeed, the most powerful.They would be just as threatenedby
the rise of an entrepreneurialbourgeoisie (Wilks, 1979). Capitalistpenetrationin

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 170 AFRICAN STUDIESREVIEW

the Africa of the 1980s is weaker than the dynamism of capitalin the Europe of
the 1850s, and colonial rule bequeathed to independent Africa highly politicized
economies. It may be that a more straightforwardstress on the interest of those
who organize power in Africa, what it is rationalfor them as a class to tolerate in
others, would make one doubt the transitionalfunction of African Bonapartism,
without which the analyticalpower of the analogymust falter.
To doubt the transformingpotential of bourgeois capitalismis not to deny the
importanceof the modern state as a social relationof production.Its differencein
this respect from most precapitalistkingdoms-but not all, as will shortly be
suggested of the forest kingdoms of western and central Africa-is best seen with
reference to the thorny question of Africa's tribalisms.I have alreadyargued that
ethnicity is best seen as a form of constituencyof access to state power, but that
may be true only of certain kinds of states and in particulareconomic contexts.
Because their rulers were dependent on revenues derived mainly from war and
trade, on Coquery-Vidrovitch'smodel of an "African mode of production"
(1969), it has been argued that the western Sudanickingdomswere less ethnically
integratedthan, for example, Ethiopia,where the aristocracydependeddirectlyon
peasantproductionand for whom a tight politicalcontrol was necessary (Crummy,
1980: 118). Goody (1971: 57-72) has also drawnattention to the widespreadritual
opposition of Horse and Earth in the same Sudanic area, between the bearer of
domination and the shrines of cultivatingtribesmen. This was a ritual expression
of separationbetween the concentrationof power we call the state and the society
over which it periodically,but not within which it regularlyexercised power.Often
true enough of kingdoms, this is a degree of separationwhich simply cannot occur
in capitalist states. If precolonial Sudanic pluralism is attributableto the state's
disinterest in local production, modern tribalismis as plausiblyexplained by the
need of producers under capitalism for the range of services which only states
provide. Precolonialmarginaltribalismwas a defense against state power, modern
tribal sectionalism often the most economical means of sharing in it (e.g., Leys,
1967). If one finds that the boundariesbetween state and society in the modern
world are indeed between tribes and the state rather than within the hierarchical
multi-classcoalitions we call tribes (Post, 1972), we would then have to conclude
that we were dealing with a species of politicalpathology.It is only where the state
no longer trusts in the reciprocitiesof the rule of law and relies instead on force
as, say, in the recent history of Uganda, that villages do literallypeer through the
bushes and pray that their state may not come roaring down the road in four-
wheel drive (Robertson, 1978). The corollaryhas been that, oppressed by a state
with which all working misunderstandingshave been severed, peasantshave then
severed their own connections with the market.
It is time now to move on from the models of social and political science to
some of the specificitiesof history (cf. Wrigley, 1971: 121; Stokes, 1978: 19).

KINGDOMS

Formation
After the Second WorldWar, scholarswere naturallyexcited by the emergence
of new nations. What Coleman (1963: 409) called "the knotty problem" for
contemporary African leaders-"how to create nations out of heterogeneous
cultural materials"-became the problematicfor historians. We took the part of
kings and used incumbents' talk. It was difficultto account for kingdoms because

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 171

of two contradictoryassumptions about state and society in precolonialAfrica.

The first was that state and community were fundamentally opposed to each
other. And the Gemeinschaft-Gesellschaft distinction was as clear to African
informants as to Western sociologists, certainlyto the Banyoro,for whom "rule"
meant "oppression" (Beattie, 1971: 6ff). This opposition was derived from a
model of African forms of production, epitomized in non-Marxist terms by
Goody (1971: ch. 2) and Hopkins (1973: ch. 2) which modes of productiontheory
has not radicallychallenged (Coquery-Vidrovitch,1969; Seddon, 1981: 38). Save
for Egypt and Ethiopia, it was argued, Africa was a continent of colonizing
societies. Scatteredpopulationstamed innumerableinternalfrontiers. Agricultural
productivitywas low and uncertain. African societies could carryonly lightweight
political baggage, the tribal chiefdom. Kinship rather than power mapped out
social obligations. There was insufficient social surplus to support governmental
institutions. Law was a matter of mediation rather than adjudication.If power
were exerted beyond mere management, for extraction, it was defeated by
secession, popularprotest by politicalmigration.That was the first assumption.
Functional social anthropologyprovided the second assumption. Social forces
tended towardsequilibrium.In however simple a society, cross-cuttingaffiliations
preserved the Peace in the Feud (Gluckman, 1955: ch. 1). Where kingdoms
existed, there was "only one theory of government"; the constitution was a
sacred value-system. Even in kingdoms, there was little inequality.Wealth could
not be stored; it could only be used by being shared out. In Ankole, the king was
the Mugabe, "giver of gifts." Force could not therefore be centralized. Since
subjects took "for granted. . . the foundation of their social order," rebellion was
directed not to changing the system but to repairing the fragility in authority
(Fortes and Evans-Pritchard,1940; Gluckman, 1955: ch. 2).
These assumptions had three implicationsfor state formation. They led, first,
to the search for origins, events rather than revolutions in organizationmaturing
out of chieftaincies.Secondly, kingdoms looked to be imposed upon society by an
autonomous will with a politicalvision. Murdock (1959: 37) envisaged all Africa
being permeated by a "mental blueprintof a despotic culture." Oliver and Fage
(1962: 44-52) also explained state formation, in the absence of all the data which
has since been collected, mainly in terms of conquests which facilitated "the
deployment of a considerablefund of common political ideas" derived from a
common source. Any other explanation was difficult to sustain if the Sudanic
State was "a parasiticgrowth, fastening itself upon the economic base of pre-
existing agriculturalsocieties."
The third implicationwas that the superiorwill was best exercisedduringsome
"cataclysmicprocess" (Miller, 1976: 1-10; Service, 1975: 15). Conquest remained
a favorite explanation; it was not appreciatedthat military occupation does not
create a political order, but a need to create one. Law (1980: 181-84) has more
recently urged the administrativeas well as the militaryvalue of the horse in the
West African savannahs. Dense or rapidly rising population was another
candidate,with the appropriationof power growingout of conflict over resources.
Thornton (1977) gave us the first hard demographicdata in support (see also
Stevenson, 1968). The articulationof two or more forms of subsistence, typically
farming and herding (but sometimes fishing as well), also gave the opportunity
for a mangerialbasis of power (Ogot, 1964; 1967: 169-73; Roberts, 1981; more
generally, Service, 1975: 75-79). A similar opportunitycould arise at deposits of
scarce but necessary minerals, salt, say, or iron, as "disorders increased among

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
172 AFRICAN STUDIES REVIEW

strangers" who flocked there and mediation became a resource (Miller, 1976:
272). The most popularexplanationfor the rise of state power was the growth of
long-distance trade, the Pirenne thesis of medieval Africa. Trade encouragedthe
growth of distributablewealth, of skills, of towns with politics, of communications
(Grey and Birmingham, 1970; Iliffe, 1979: 52f, both skeptical accounts). The
point of all these hypotheses was that something ratherexceptionalwas needed to
explain any concentrationof power in a logically tribal Africa. Furthermore,they
were all based on pluralsociety assumptions, not on social divisions accordingto
inequalitiesof access to resources.
Precolonialhistory is now much richer in data than ever it was in the 1960s, so
much so that it can support sharply divergent logics of enquiry. Marxist history
and historical anthropology have pursued productive logic, non-Marxists seem
increasingly to stress political logic. This may to some extent reflect different
geographicalareas of enquiry, with Marxists thus far tending to focus on the West
African savannahs and non-Marxistson the West African and equatorialforests.
It is possible that productionin the resource-richforests was more amenable to
politicaldirection than in the less fertile savannahs, with their greaterdependence
on seasonal peaks of labor accordingto rainfall (Tosh, 1980). At all events, the
embedded communal modes of production remain more obdurate for Marxists
than for others, and African kingdoms therefore more clearly elevated above
them in a tributarymode of production which while supporting warriorcastes
which raided outwards was nonetheless more respectfulof the domestic lineages
(Coquery-Vidrovitch, 1969; Meillassoux, 1964, 1972; Terray, 1972). Non-
Marxistsare beginning to argue to the contrarythat the communities of tribe and
clan were either deliberatecreations to facilitatepoliticalcontrol (Moseley, 1979)
or else developing relations of labor absorption and exploitation (Wilks, 1978).
Any process model of the state would suggest that in most cases these
antagonisticconditions are likely to be found in all social formations, as forms of
class conflict.
More materialhas taught us three things. First, state-formationreallywas very
slow: it was frequentlybotched and startedagain. The slowness of the process can
clearlybe seen in Alagoa's studies (1970, 1971) of the Niger Delta, with layers of
changinglivelihoods convertingfarming to fishing villages and these to mercantile
city states over two centuries. Colson (1969) and Miller (1976) insist that the
decay and fall of kingdoms is as important a process as their rise, with the
communal modes of production beneath them continuing to produce and
exchange with a vitalityuninterruptedby the politicalgyrationsabove them.
What we have learned, secondly, is a great deal more about the politics of
state-formation. This is important. It was so often to be the politics of state
collapse as well. The notion of fusion of elite is useful; it suggests a considerable
independence of power in those who held together a kingdom, the decentralized
hallmark of a precapitalist society. Early kings were probably mediators, a
relativelyautonomous, but also relativelypowerless, convenience. In a number of
commentaries and studies, Lloyd (1965, 1968a, 1968b, 1971) used a conflict
model to examine how Yoruba politics developed from "tribal kingdoms" with
royal mediators to "centralized monarchies" under royal autocrats. There were
developing struggles over the enjoyment of new resources, booty, slaves, guns.
There was certainly no one theory of government during these organizational
revolutions. Autocracy fought to free itself from aristocraticconstitutionalism.
Centralizationwas a euphemism for the violent expropriationof local autonomies.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 173

The emerging states were held together not by consensus but by coercion. Nor did
the Kuba have only one political theory. They used to argue subversively about
their own monarchy (Vansina, 1978: 209). Was it really endowed with mystic
worth? Were kings merely rich enought to have the best charms? Worse still,
could they be arch-sorcerers?Sociologicalthinking had in any case become more
skeptical about the necessary consistency of ideology. Gellner (1973: ch. 2) even
argued that a wholly consistent ideology would make society unworkable.
This symmetricalconflict between dominant elites is perhapsmore useful than
a model of class domination in explaining state formation. The class domination
came afterwards, a silent cataclysm (Fried, 1967: 183) more plausible than the
noisy ones of an earlier historiography.Where formal class analysis of precolonial
African states has been attempted (e.g., Terray, 1975a and 1975b; Rusch, 1975),
the dominant class has looked fairly secure, mainly because its power to exploit
was so limited by the low level of the forces of production.Marxists, however,
need to take intra-classconflict, especiallyconflict at the top, just as seriously as
inter-class struggle: it is so much the more easily politicized, it is political. From
this perspective, dominant strata everywhere in Africa seem to have been very
insecure. Not only were they, as senior members of cultivating or herding
communities, subject to the same kind of naturaldisasteras everybodyelse, with
a constantly fragile balance between the manpower under their control and the
material environment, holding authoritywithin a dispersed network of legitimacy
which was always open to test by the ambitious, but in West Africa at least there
does seem to have been a persistent tension between commerce, and political
control through warfareand tribute. As in our own day, the merchantclass was
transnational.It did not need states to supervise business, however much (or
little) states may have structured production. It had its own cultural diaspora,
Islam in the interior, secret societies or missionaryChristianityalong the coat, to
create and police a commercial morality beyond the bounds of kin (Dike, 1956;
A. Cohen, 1971; Ajayi, 1965: 51f). The tradersof the Niger Delta were careful to
keep political power sufficiently small not to interfere with commercial
competition (Dike, 1956; Jones, 1963; Northrup, 1978). In Senegambia,
merchants were quick to withdrawfrom a polity when a king died; it was too
expensive to get caught up in the politics of succession (Klein, 1968: 19). In the
Kololo kingdom of Barotseland, trade was always being jeopardized by cattle-
raiding chiefs (Flint, 1970: 83ff). There was no end to the oppositions between
tradersand states in precolonialas in postcolonialAfrica, and no end, therefore,
to symmetricalconflict over the inhibitionsimposed by the rule of law.
Emphasis on the local politics of state-formation has, thirdly, largely
supplanted the concept of the diffusion of state forms. No longer so concerned
with the idea of the state, historians now seek to explain the rise of particular
kingdoms. The image now is of separatelittle local internationalsystems, clusters
of kingdoms reacting against their neighbors by competition in war and trade. If
African kingdoms did share a common fund of political ideas, that was because
there is a limited stock of symbols and justificationsof human authorityanyway.
If there was a movement of ideas, it was by exchange, by "saturation" rather
than diffusion (Curtinet al., 1978: 31f). The SudanicState has come to look more
like European enlightened despotism or even liberal democracy, powerful
principlesadopted from elsewhere but by alreadyestablished politicalsystems-or
like the constitution of the British Labor Party, which although used by the CPP
in Ghana and TANU in Tanzania, was scarcely responsible for aggregatingtheir

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
174 AFRICAN STUDIES REVIEW

power.
A brief glance at the changing treatmentof two cases of state formation in the
early nineteenth century serves to illustratethe argument.The stimuli to jihad in
the Western Sudan have always been recognized to be complex. No scholar has
ever proposed a single-strandedexplanation.Nevertheless, it is not stretching the
spectrum of hypothesis too much to argue that whereas in earlier studies (e.g.,
Smith, 1961; Waldman, 1965; Last, 1967; Levtzion, 1971), the jihads were placed
within what one might call Islamic, that is, ideological, time, more recent ones
(e.g., Adeleye, 1971a; Klein, 1972; Johnson, 1976; Usman, n.d.) have situated
them in local, that is, socio-political, time. Previously, the frictions between the
normative Islam of the clerical and commercialdiasporaand the popularIslam of
local authority provided the spark of action. Islam is now being treated more as
Thompson (1965) treated Methodism in relation to the English working-class,
available to give further shape and direction to a consciousness already half
formed in local conflicts. In the Western Sudan, it was of course a ruling class
ideology which was formed. Willis (1979) seems to have finally discredited any
Fulani tribal explanation for revolt. Rather (and cf. Usman, n.d.), occupational
categories of people, whether scholar-officialsor herders, can be seen coming
together in a cultural ethnicity which had almost nothing to do with affinity and
descent. Islam was a ready-made rule of law within which to reach the
compromises of ruling-classformation from out of the previous conflicts between
war, production,and trade. Of the three majorjihads, the first, dan Fodio's, now
begins to look more like an organizationalrevolution and consolidationof state-
formation, driven by dissident members of a previouslydivided ruling class. The
second, Ahmadu's, looks like an independence movement by one occupational
category, the herders, in a social formation whose modes of production were
closely articulated. But the third, Umar's (Oloruntimehin, 1972), seems likely
always to remain essentially idealist, the wilful construction of a man with a
mission.
The three theocracies which eventuated were markedly different in their
formation and productivityof their politics. Only the smallest of them, Massina,
can be said to have had an origin, in sudden revolutionarywar. Umar's kingdom
was never fully formed. The Rulers of the Sokoto Caliphatespent a century in
never quite establishing central control over the several emirates which claimed
their authority for rule. Their organizationalrevolutions variously illustratedthe
point that not only does trade encourage the growth of authority but power
promotes productionand trade. War paid for itself in slaves; a coherent political
order enabled slaves to be controlled at work; taxes and efficient administration
stimulated the flow of commodities; traders could share the same religious and
social assumptionsas rulers (Johnson, 1976; Lovejoy, 1978; Roberts, 1980a: 172;
Azarya, 1980). The concept of social reproduction,as seen through the filter of
the rule of law, can be further put to good use in relatingthe formationof classes
to the actions of states. Roberts (1980b) shows how alien warriorclasses can be
distinguished from those which emerged internally by reference to the laws on
private raiding. Before Umar's conquest, Segu's slaves, although politically
localized warriors, were legally, if not always effectively, forbidden to engage in
raiding on their own account. Umar's invading army, for whom the local
productiveeconomy meant not a sustaining politicalorder but a source of booty,
was encouraged to plunder at will. The Tukulor evidently could not face the
political costs involved in subduing their occupying force to a locally productive

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 175

rule of law. In Sokoto, the community of Islam for long kept the emirates more
or less at peace with each other, while they were also "predatorystates organized
for war" againstoutsiders (Smaldone, 1977).
New interpretations of the post-mfecanekingdoms of south-eastern Africa
have owed much to raids on the static concepts of modes of productiontheory for
the purposes of understandingan extraordinarilydynamic process. When Omer-
Cooper (1966) opened the modern enquiry into the problem, the formal influence
of structural-functionalanthropology, it is now clear, caused him to over-
emphasize the revolutionarychanges involved in institution-buildingamong the
Nguni and Sotho peoples, however much he also stressed the adaptation of
existing structures.Creative leadershipwas his major explanatorydevice, and the
differences between the successor kingdoms were determined largely by their
varying social and historical distance from the blueprint of the Zulu military
system.
There have been three majorchanges in more recent historiography.First, the
response to populationpressurethesis has been put on firmerfooting with the aid
of dendrochronology(Hall, 1976-77) and detailed knowledge of the mosaic of
seasonal grasslandswhich successful herd management required (Daniel, 1973).
Competition for the more humid coastal grazing during the period of failing
rainfall in the later eighteenth century now reinforces the earlier trade and state
emphasis on competition for the trade of Delagoa Bay (Smith, 1969). Here,
however, one must insert a general caution. Drought seems to be emerging to
challenge trade as a major explanation of state-formation (Weoster, 1979;
Steinhart, 1981). In any mode of production with direct appropriationfrom
nature, drought is clearlyas dreadfula crisis as marketfailure is in capitalism.But
does reaction to drought build states? East Africa's history furnishes many
examples of individualsand families from neighboringpeoples deliberatelyforging
relationships of marriage, fostering, or gift, across the ethnic and ecological
boundaryprecisely to insure against local climatologicalor other naturaldisasters.
At times of drought, the unlucky borroweda subsistence from neighbors whose
friendship they had cultivated; they did not try to conquer them (e.g., Spencer,
1973; Muriuki, 1974; Waller, 1978; Sobania, 1980; Anderson, 1982). Two
questions arise. Should one most expect conquest across boundarieswhere such
ties have not previously been created, as perhapsin the hundredyears of warfare
between Africans and Europeanson the bordersof the EasternCape? Or, rather,
are the kingdoms which are most convincingly attributable to an aggressive
response to ecological crisis based upon these cross-ethnic ties and, therefore,
conquests from within existing social relations rather than from without? The
latter would bring state-buildingmore comfortablywithin the ambit of modes of
productiontheory.
This points to the second change in the historiographyof the mfecane. Hedges
(1978) and Bonner (1980 and f.c.) have used the concepts of the lineage mode of
production to argue that in many respects the state-building of the early
nineteenth-centurywas less revolutionarythan was earlier thought. They show
that inter-clanrelations of marriage,cooperationin hunting, cattle-loaning,and so
on, all of which supportedthe parochialauthorityof elders and chiefs each within
their own lineage segment, constitutedwhat, in retrospect,can be seen as a latent
state, whose complex affiliationscontinued to play an importantpartin the politics
of the post-revolutionarystates. Ballard(1980) has given a detailed pictureof the
marriage-politicsof a post-mfecane chief (albeit English) which belies the old

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
176 AFRICAN STUDIES REVIEW

picture of a great divide between the Zulu state and Nguni society. The
continuities of pre- and post-mfecane politics are now beginningto appearat least
as strikingas the discontinuities (e.g., Morrow, 1972-73).
Finally, whereas virtually all the post-mfecane kingdoms once looked like
political answers to economic problems, it is now being argued that their
revolutions in government were at the same time transformations in the
techniques and relations of production.This had alwaysbeen known of the Sotho,
with their social differentiatedaccess to the horses, wheeled transport, merino
sheep, and wheat of the South African white economy (e.g., L. Thompson, 1975:
190-201), but Guy (1979: ch. 1; 1980) has now analyzed the Zulu regimental
system not only, as was previouslyaccepted, as a brutalweapon of appropriation,
although it was certainly that, but also as the social embodiment of a larger
system of pastoralproductionin the royal interest. Bonner (1980) too has traced
how the institutional changes of the post-mfecane Swazi state were means to
extract more surplus production internally for the enjoyment of the royal
establishment. Just in case, however, one should be tempted to postulate as
revolutionary an economic change as was once claimed in the political realm,
Cobbing (1974) has demonstrated how the reproduction of formal military
structuresthrough time in precapitalisteconomy can be achieved only through the
informal-and therefore real-decomposition of their structuresinto a network of
productivelycooperatinghouseholds, more tribe than regiment.
Change
The same emphasis on the continuities within change is emerging in the
African history more generally. Previously they were too much opposed. There
could be only two interpretationsof change. Either the blueprint of African
despotism faded away into political decay (Miller, 1976: 1-11); or else the only
politicalchanges worthy the name were violent, revolutionary,or unconstitutional.
In the latter case, the cause was a foreign intrusion into a domestic equilibrium
(Lloyd, 1968b: 27), whether conquest, the opening of export markets for slaves
or vegetable oils, the import of firearmsor missionaryideas, Islamic or Christian.
Externalchallenges were important,but not all African kingdomsreactedto them
in the same way. Their foreign policy was closely related to domestic affairs, as in
any other state. Two dimensions of change are now becoming increasingly
apparent, and their dynamics were created by a fundamental continuity. Some
changes were symmetrical,in the constitutionaldispositionsof power-holders.The
other changes were asymmetrical, between the power-holdersand those from
whom they derived their materialsupports, clansmen, the city poor, slaves. The
continuity lay in the means of livelihood for most people: "the lineages were
constantly in the background"(Miller, 1976: 282). In was on this continuity that
change was founded.
Symmetricalchange focused on the dissension between king and barons over
the distribution of enlarged state powers where success in war forced tribal
kingdoms, to use Lloyd's terminology (1971) to become centralized monarchies
through organizationalrevolution or else fall apart.Their governing capacitieshad
to be developed in range, depth, and efficiency (Wilks, 1967: 212). But to whose
gain? King's or staff's? The process of decision throughconflict has perhapsbeen
most systematically laid out by Southwold (1961, 1968), with reference to
Buganda. The Bemba experience has been discussed rather similarly (Werbner,
1967; Roberts, 1974). The nineteenth century saw the working out of the two

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 177

points at issue, introduced earlier, between king and barons. The kings of
Bugandarepeatedlyinvested the resources created by fresh conquests in funding
dependent new orders of chiefs as older authorities converted power from "a
borrowedgarment" delegated by the king into their private, hereditable,property
(Fallers, 1964: ch. 2). The power of monarchical government undoubtedly
increased, but that was not the same as the power of the king. For as government
enlarged, so too did the destructive effects of succession war and fraternal
usurpation.The rules of succession were narrowedto limit eligibility to younger
sons, striplingswho were no threat to their reigningfather, in order to protect the
unity of the kingdom. The powers of its chief officerswere enlargedby the same
measures, since a young king was necessarily dependent upon his dead father's
ministers. A further consequence, not explored by Southwold, was to give each
reign a political history, that is, to separate reigns from governments, as every
successive ruler sought by force and intrigue to replace yesterday's men with an
executive which owed everything to him. Buganda's Christian Revolution was
what might occur when the executive cadets were also a fifth column for
ideologicalsubversion from outside.
Party politics animated other African kingdoms as well, even the outwardly
stern military despotism of the Ndebele (Brown, 1969). It seems to have been
institutionalizedin ancient times in the Kuba kingdom (Vansina, 1978: 155-60).
In Dahomey, the administrative face of regulated exchange and the rigid
separation of different levels of society on Polanyi's (1966) celebrated
substantivistmodel, has now been modulated into conflict over divergentpolitical
orders, between kings and generals who tried to monopolize externalcontacts and
noble traderswho, if they could not break into the circle of monopoly, sought to
do away with it altogether (Law, 1977b; Moseley, 1979; Johnson, 1980).
Underneath, the asymmetries of power were also shifting, at least in West
Africa, although its savannahand forest kingdomsagain need to be distinguished.
In the savannahs, the alliances of dynasty and warriorcaste were increasingly
threatened as the nineteenth century progressed.Warfareturned in on itself in
the far interior (Adeleye, 1971b: 109-13; Smaldone, 1977: 138-62). On the
Senegambian coasts, resistance to warfare and tributaryexaction rose with the
opportunities for trade in vegetable products (Klein, 1968: 44f; 1972; Curtin,
1975: 187-96). Cultivating communities, whether ex-slave or free, turned to
Muslim traders for protection, "in reality an opposition of the proletariatto the
aristocracy, a class struggle" (R. Arnaud, quoted in O'Brien, 1971: 29). This
sharp change in the control of rural production, however, was not paralleledin
the forest areas. Politicalchange was more subtle, because their economies were
more diverse. "Trade or taxable peasants?" is no longer an illuminatingquestion
to ask of their kingdoms' sources of revenue. One of the consequences of more
effective government was the ramifyingabilityto bring ever more activities within
the tributary, fee-paying or directly profitable net (Newbury, 1969; Hopkins,
1973; Vansina, 1978, ch. 10; Law, 1978a; Dumett, 1979). Kings partly lived of
their own from the slave productionof foodstuffs and export crops; partlyfrom
tolls and licenses levied on trade; partly from tribute; and in some cases from
direct taxes on private plantations. The social relations were those of a mixed
economy, not merely articulated modes of production. Lineages and clans,
however fictitious, provided the cooperativegrids of farmingwhile clientage gave
access to trading privilege or the fortunes of war. It was on this foundation that
political fortunes turned, as power-holdersvariously led their followers to war or

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 178 AFRICAN STUDIES REVIEW

sought to profit by their agriculturalsales. A number of examples suggest that

while such communal ties were apparently submerged under the symmetrical
politics of patrimonialregime-building,they in fact endured through the changes
of the nineteenth century and became once more the bedrock of politicalloyalty
during the time of troubleswith which the century ended.
The experience of Asante has been the best documented, a paradigmof the
successful revolution in government followed by the diverging hegemonies of
class formation. Asante fought its way out from tribal kingdom to centralizing
monarchy during the eighteenth century, making its provinces pay for their
conquest with what the British in India in the same years called "subsidiary
alliances" (Arhin, 1967). Arhin, and Wilks (1975), have shown how Asante's
rulers were then forced to civilianizecontrol to cut the costs of empire. Wilks has
shown how they had the power to do so (1975: 445-76). In brief, the dislocations
of conquest made Kumasi the focus of ambition for young Akan everywhere,
from Kumasi itself, from the surroundingHome Counties of the original tribal
kingdom, from the conquered provinces. The kings were also able to assemble
their own following of hit-men, the ankobea, from these transitionals,although
centralizationseems to have relied more upon ambition to enjoy the free-floating
resources of the new asomfo, or service, stools than on intimidation by
revolutionary guards. The new administrative institutions and relationships
evoked new sources and solidaritiesof conflict within the enlarged politicalclass.
The segmentary politics of provincial rebellion, focused on the geographical
allocation of power, became overshadowed by party politics within the same
patrimonialofficialdomwhich had been designed to govern the provinces. Party
politics disputed not so much the allocation of power as its use in policy,
especiallyin matters of war and peace with the Britishat the coast (see also Wilks,
1971). As in Buganda, one must begin to distinguish between reigns and
governments. Asante notables increasinglymobilized their asymmetricalbases of
power in their conflicts with each other, the contradictoryresourcesof tributeand
trade, with profoundlydifferent implicationsfor the ideology of the politicalorder
(Wilks, 1975: ch. 15). Intra-classpolitics at the top then opened up the possibility
of inter-class politics from below. The poor of Kumasi, the rural poor too, slave
and free, entered history for the first time on their own account, in protest
especially at the rising burdens of war upon those who stood least to gain and
most to suffer from it (Wilks, 1975: 504-9). Possibly the best analogy would be
Haile Selassie's Ethiopia.
When Asante was reconstructedafter the civil wars of the mid-1880s (Wilks,
1975: 549-88), however, it was essentially on the basis of the old tribalkingdom,
the Home Counties of the Confederacy, from which the outer provinces
increasingly defected, or from which they were enticed by the British (also
Tordoff, 1965: ch. 1). Recent writing on Buganda too (e.g., Kiwanuka, 1971;
Wright, 1971; Twaddle, 1972; 1974) has stressed the durability of old, pre-
patrimonialpatterns of affiliation, which in the civil wars, again, of the 1880s,
protected fellow clansmen of opposing patrimonial parties from each other's
wrath. For the Tanganyikaarea, Iliffe (1979: 60-66, 98-122) has shown how old
civic loyalties and spears outlasted the new patrimonialcreations of the firearms
trade in the ordeal of resistanceto German conquest.
The collapse of the Oyo Empire (Law, 1977a) and the long wars between its
successor states (Ajayi and Smith, 1964) perhaps illustrate most starkly the
counter-productivity of politics, the fragility of new-style centralized

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 179

administrationwhen superimposedover the old constitutionalismsof lineage and

arbitration. The chief weakness of Oyo's centralization, when compared with
Asante's, was that its new staff was composed of kinless outsiders, the palace
slaves, the ilari. Asante's new officials were very often the sons of the old.
Asante's political class was linked by innumerableties of marriage,descent, and
culture (McCaskie, 1980). As in Benin, it competed in "a single promotional
field" (Bradbury, 1969: 31). Oyo's politics was unbridgeablydivided between
palace and Oyo town. When Oyo too had its civil war and under-class revolt, it
lacked the political means to reconstitute a central authority. The dangers of
political centralization based on slaves were nowhere more clearly shown, the
limiting case on Cooper's (1979: 116f) argument for the utility of slaves as a
free-floating political resource. And when Oyo collapsed, constitutional
relationshipsbetween the Yoruba collapsed, too. Many reasons can be given for
the debilitating length of the Yoruba wars. The reason which the Yoruba
themselves gave (Ajayi and Smith, 1964: 63-75) was that there was no longer any
authoritystrong enough to arbitratea peace. The rule of force had supplantedthe
rule of law. Any revolution in government was blocked, for Ibadan could never
establish a hegemony. It was a parvenu power, "a people without a king or even a
constitution" as the Ekiti rebels complained (Akintoye, 1971: 37, 156). Ibadan
was just as illegitimate, or could usefully be made to seem so, as the Portuguese
prazeros of Mozambique, who had all the attributes of chiefly power save,
crucially, the legitimacy which came from ritual connections with the land,
reinforcedby dynasticpolitics (Isaacman,1972).
Any concept of an African mode of productionwhich insists on the durability
of community and the fragilityof rule, yet which also maintainsa rigid separation
between the communalities of production and the administrativelevel of the
state, cannot begin to explain these interconnectionsbetween politics and societal
change or dislocation.
Fall
Not enough work has yet been done on the European conquest of Africa
which asks the question: How did the forms of opposition within African polities
affect their external opposition to the European? (but see Kenya-Forstner,1969).
Until recently, societies have been taken too much in the round. Even when one
does ask the question, it is not clear whether the breakdownin African authority
under increasing European penetration released forms and ideas of opposition
which would have been literally unthinkable under the old order. If historians,
rightly, insist on process as the moulder of subsequent possibilities, how good is
the evidence for previous slave unrest to be found in the great flights of slaves
from their owners after European-ledarmies marchedacross the western Sudan?
(e.g., Roberts and Klein, 1980). The earlierhistory of Calabargives firm evidence
of slave discontent certainly (Latham, 1973), and the examples of Asante and
Oyo also suggest that in their discontents the African poor were no differentfrom
those elsewhere in the world. They took their chance to express it only when and
where they could.
But what about the discontents of the wealthy and the powerful? Christianity
was not attractive just to the poor and outcast (Ajayi, 1965), refugees and
dissidents (McCracken, 1977: 47-61), "the unstable, the rebellious or the
rejected" (Etherington, 1978: 67). Bugandais the classic (in many ways the only)
case of the cadets of a political class being drawn to Christianityas a dissident

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
180 AFRICAN STUDIES REVIEW

ideology through the strains of political competition, in much the same way as it
has been suggested of West African chiefs that the burdens of office made them
susceptible to Islam (Levtzion, 1971: 34). But if Buganda can nowhere be
paralleled,not even among the Lozi and the Tswana, there are other indications
of how insecure or dissatisfiedmembers of the African politicalclass could be-in
the way in which the Pedi royal women were attracted to Christianity (Delius,
1980: 158f), or the interest which Owusu Ansa was able to arouse in Kumasi
(Wilks, 1975: ch. 14). More concretely, there is all the evidence from West Africa
of what one might call a crisis of monarchy.There were the Asante officialtraders
who made privatedeals with Europeans,who resisted the paymentof death duties
(Wilks, 1979); the tradingchiefs of Yorubaland,whose armed conflicts filled the
vacuum left by the fall of Oyo (Awe, 1973); the merchants of Dahomey, who
outmaneuvered the king in selling palm productsoverseas (Law, 1977b); and, all
over Senegambia, the muslim traders who took the part of peasants against the
chiefs and their slave armies (Klein, 1972). This period has been called a "crisis
of aristocracy" (Hopkins, 1973: ch. 4), on the hypothesis that the big
entrepreneurs,kings and other members of the politicalclass, could be outsold by
peasants once the export staple turned from slaves to vegetables. On the evidence
referred to here, however, this seems too undifferentiateda picture of state
structures in crisis. It appears, to the contrary,that African power-holdersbegan
to shift for themselves, increasinglyin competition with the central powers of
their kingdoms, for the continued control of marketable surpluses. To some
extent, therefore, the intensificationof Europeantrade and the onset of colonial
rule promotedand benefited from the rise of a newly conscious class, men whose
career opportunities had been created first by centralization, then by the
increasing power of chiefs over kings, and now, with the growing commodity
trade, by the opportunityto put public power to privateprofit.The colonial period
would not by any means everywherecut the ground from under their feet.
COLONIES
It is beginning to be possible to discern the outlines of a historiographical
revolution in our treatment of colonial Africa. Symmetrical and idealist
explanationsare giving way to asymmetricalones. The best indicatorof change is
the shifting periodizationof colonial rule, the break between its formation and
fall.
Two decades ago, colonial rule could be analyzedonly with the techniques and
assumptions of political science. It was one among several contemporary
revolutionary systems of "social change in the context of preliterate or pre-
industrial societies," not merely "a dominating social and political system"
(Kilson, 1966: 37). The most influentialapproachto understandingnew states at
the time had virtually excluded domination from its political vocabularly
altogether; power was used for "integration," for "adaptation" (Almond and
Coleman, 1960). Capitalismwas understood as a normativeratherthan economic
system; it was called modernization.Its values were in the mind, not the market.
Too many scholars had forgotten the caveat with which Roland Oliver (1952)
opened one of the first professionalworks of African history, in his quotation of
Archbishop William Temple to the effect that "the only purely spiritual
phenomena are good intentions, and we all know what portion of the universe is
paved with them." Too few scholars had read Balandier (1951), with his bleak
awareness that colonialism was indeed a system of domination and the social

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 181

change which it evoked scarcely a liberation. As BarringtonMoore (1978: 156)

has more recently said, what for social scientists are the "strains of incipient
modernization" are really "causes of acute human suffering." Studies of the
colonial period do now focus on domination and subordination,their purposes,
their pathologies, and the changes to which they were subjectwhen they could no
longer be endured.
It would be too simple to infer that an autonomist paradigmof the colonial
state must be supplanted by an instrumental one. After all, colonies were
increasinglyjustified to metropolitan and African publics in terms of the good
they did to Africans, and on those terms colonies themselves came increasinglyto
be judged. They were certainly external apparatusesof control "created for the
express purpose of dominatingthe local society" (Dunn, 1978: 5), but they were
also far more than that. Colonial revenues required the intensification of
production and trade, the deepening of capitalist relations. Colonial peace
therefore depended on coping with a variety of contradictions (Lonsdale and
Berman, 1979). Capitalists had to be satisfied; their competing claims,
metropolitan and local, had to be reconciled. Their rapacities often had to be
restrained lest they destroy the African production on which they fed. As an
organizational revolution, colonial rule sometimes enlisted African political
authorities in the supervision of production, as landlords (Cooper, 1981); more
often, they were looked to simply to ensure some continuity in social order lest,
when "the natives [were brought] from the first Chapterof Genesis to the Book
of Judges," the colonial state itself should "be plunged into Revelations"
(Governor of Upper Nile, 1928, quoted in Tosh, 1981). A properunderstanding
of colonial states requires, therefore, an awarenessof a more than usual number
of historical dimensions (cf. Low, 1973). Colonies suffered an extraordinaryfall
from grace in the twentieth century history of ideas; their rulers ran out of self-
confidence as autonomous agents; they looked increasingly dubious as
instruments of metropolitanprofit the more they stirredup local and international
antagonisms. As social processes, they could engender nations or classes only by
opposition, they rarely disposed of large forces of coercion and had no means of
mobilizing active consent. It is probablyas social relationsof productionthat they
are best looked at in the round and most to be found wanting. They were not
especiallyproductiverevolutions in government.
Colonial rulers prided themselves in opening up Africa, an accurate enough
description of the articulation of its indigenous economies with external
capitalism.A concept introducedearlier, articulationcan now be further explored
as a means of analyzingcolonial rule (Rey, 1971; 1973). Like Marxistconcepts of
history generally, articulationhas three layers of meaning. At the bottom, at the
level of productive logic, the oppositions between precapitalistand capitalist
modes of productionseem overwhelming.It is difficultto see how dependent and
wage labor can co-exist in one system, or productionfor use and productionfor
exchange, politically allocated productive resources and a market in them.
Theories of dual economy were founded in this common-sense dichotomy. When
one considers class logic, then it is these very oppositions which, I have
suggested, make class conflict so endlessly intricate,so full of tacticalresource.To
take just one example, if migrantwage labor providedan opportunityfor youthful
independence, the elders' control over women and the cattle of bridewealthoften
but not always served to claw it back again (e.g., Wilson, 1977; Beinart, 1979);
employers and traders could profit in either event. The political logic for states

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
182 AFRICAN STUDIESREVIEW

was just as contradictory,and specific to each colony. The economic liberalization

demanded by capitalism meant the overturning of African propertyrights, the
subversion of family controls on labor. Orchestrationof supply entailed taxation;
it could mean the enforcement of wage labor or production quotas of export
crops. When so little of Africa's land and labor was commoditized, however,
consensual intervention could only mean droppingsuch pressures in the face of
African resistance, not softening the conditions of employment. Articulation,
therefore, embracedmany conflictingsocial processes, dependent upon the nature
of the local food productionsystem, the most readilyexploitableexport resources,
and most of all upon the coalition of dominant interests, capitalist,precapitalist,
and combinations of the two to which governors had to defer if they were to
secure profits, peace, and promotion.
Four processes of articulation can be readily distinguished: intransitive,
transitive, tributary,and syncretic.The first was impossible in colonial Africa; the
second was often an economic ideal but for long a political anathema; the third
the settler's dream;and the fourth a general condition.
Intransitive articulation implies the strengthening rather than weakening of
existing relations of productionthrough the stimulus of exchange relations with
capitalism. If they could get away with it internally, slave-owners might impose
harsher work disciplines, elders might convert ceremonial or legal funds into
sources of investment for tighter control of the lineage, each untroubled by
external intervention. Both occurred in precolonial Africa. Neither was fully
possible once a colonial apparatuswas inserted into local productiverelations, not
so much because they became illegal but because the whole economic and social
context of conflict was so greatly widened. Authority faced new competition,
resistance found new escape routes. Nevertheless, the old relations of power
remainedas resourceswith which to resist a capitalisttransition.
Transitivearticulation,the transformationof all existing relationsinto capitalist
relations, was therefore a deeply contested process and has rarely occurred. In
Africa, as elsewhere, it requiredprimitive accumulation,an exclusionaryprocess
whereby productiveresources to which there had been many rightsof access were
appropriatedby the few, to be reconstituted as the capital without whose
employment laborers could not reproduce themselves. It has been virtually
unknown for private capital to appropriateAfrican land and labor on its own
account. The earliest white settlements in South Africa spread under the
protectionof the Dutch East India Company, in Algeria as auxiliariesto military
conquest. It took British imperialtroops finally to protect the Transvaalfrom the
Pedi, Natal from the Zulu, and the Rhodesian pioneers from their folly. Where
capital directly employed its own coercive force, as in the Congo of the
concessionarycompanies, it broughtforth death and destructionmore than profits
(Slade, 1962: ch. 8; Anstey, 1966: ch. 1); it was "a system with limits"
(Jewsiewicki, 1979: 36). Where colonial states abstracted these relations of
violence from the private into the public domain, they brought forth
protests-and the beginningsof the historiographyof colonial rule.
Colonialism's internal critics were believers in capitalism'sfree-marketvalues,
morally preferableto, economically more efficient than its most obvious African
predecessor, slavery (Cooper, 1980: ch. 2). They had been brought up to believe
in the autonomous role of the state as a referee, paid to see that nobody broke
the rules by the exercise of political privilege or restrictivepractice.Whether, to
take a few at random, they were humanitarianofficial, metropolitancommunist,

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 183

or liberal economist, all deplored the instrumentalinequalitiesby which the white

settlers in East Africa or white traders in the West sacrificed long term
development for short term advantage by measures to limit African productivity
or bargaining-power(N. Leys, 1924; Aaronovitch, 1947; Bauer, 1954; Colonial
Office, 1955). Colin Leys's analysis of white politics in Rhodesia (1959)
epitomized this liberal critique: it could not possibly be in the settlers' true
interest to deny full market access to Africans. The theorists of political and
economic modernization who followed them could apply their metropolitan
models of possessory individualism to Africa only by taking as environmental
givens the politicalconstraintswhich their spiritualpredecessorshad condemned.
Dual economy analysis agreed that the economicallyprivilegedmight misguidedly
hinder African emergence from the subsistence to the money economy, but
scarcelythat the state had by this exclusion actuallycreated the conditionsof rural
poverty (Barber, 1961). Neo-classical economic historians since then have also
played down the instrumental role of colonies, either by stressing, for West
Africa, the productivealliance between public infrastructuresand private African
entrepreneurship,both governed by the terms of trade (Hopkins, 1973: ch. 5) or,
in eastern, settler Africa, by showing how little could be done to subsidize white
agriculturewhen the tributaryresources available from Africans were anyway so
slender (Mosley, 1980).
Marxists, on the other hand, had no problem in seeing colonial states as the
instruments of primitiveaccumulation.Settler capitalismclearly had no chance of
prospering according to the market's hidden hand until enough resources had
been appropriatedfrom Africans to force their "free" entry to the labor market
(Arrighi, 1970; Brett, 1973; Van Zwanenberg, 1975). Laissez-faire colonialism
was a contradiction in terms, no more than a myth of legitimacy. To put the
argumentquite as starklyas that was, however, to miss three importantstimuli to
further change in the politicaleconomy of colonialism, to be elaboratedbelow. It
was to forget the importanceof myths of legitimacyin the minds of colonialism's
critics, some of whom were also colonial rulers. It was to understate the
continuing complications of articulated modes of production, which the 1930s
depression, war and post-war reconstruction,would demand to be simplified. It
was also to ignore the difficulties which otherwise diverse colonial regimes all
faced when dealing with capitalist class formation. It was not until after the
Second World War that some colonies were obliged to come to grips with the
implicationsof a capitalisttransition,in a second painful organizationalrevolution.
The drive to this further change in the 1940s was in part due to growingcrisis
in the rural Africa of the labor reserves (Amin, 1972) impoverishedby what can
be called tributary articulation. This concept seems to be central to
underdevelopmenttheory-the partialdestruction of peasant economies, so that
they were capable only of reproducing migrant male labor-powerby women's
work within their households. They would supply laborerscost-free to employers
and care for them in old age without charge to the state. They could not compete
on the produce market, they could not engender the social differentiationswhich
might push out full proletarians,a standing challenge to capital. However much
such subsidizationof labor may have been a universal settler dream (cf. Wrigley,
1965: 246), tributarywas as rare as transitive articulation;it can only be said to
have occurred in South Africa, and in its pure form only between the wars; or,
more precisely, between the 1913 Natives Land Act, which rang down the curtain
on the possibility of surplus production by black South African peasantries

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
184 AFRICAN STUDIES REVIEW

(Bundy, 1979) and 1948, when the new coercions of apartheidrecognized that
only the law could keep Africans out of towns: the Homelands were now so
exhausted and overcrowded that they could no longer sustain even paupers
(Wolpe, 1972). But the South African state was unique in three respects; it does
not seem wise to use its experience as an export model for understandingthe
relations between states and peasantries further north (Palmer and Parsons,
1977). Its coercive power was unequivocally derived from the local whites'
representative politics; white politics was divided by unusually bitter ethnic and
class conflicts which were best composed at the expense of blacks; and the
profitabilityof gold, from the 1930s, meant that mining has been able to subsidize
other sectors of productionmore convincingly than cheap African food supplies
did in white Africa north of the Limpopo. South Africa had the power and the
political reasons to constrict the African Homelands to the supply of tributary
labor and, critically, the financial resources with which to ignore the economic
consequences.
No colonial state was so constituted, not even Rhodesia or Algeria. Tributary
articulationwas never so unequivocalelsewhere; and if transitivearticulationwas
also rare before the Second World War, then syncretic articulation,the uneven
and combined development (cf. Post, 1972) of both capitalistand precapitalist
economy, remains as a possibility. Its probabilitywas shaped in the politics of
colonial conquest.

Conquest
Colonies were undeniably among Africa's conquest states; their formation
illustrates the point that even conquest was part of a politicalprocess ratherthan
an event. As an organizationalrevolution, conquest did often enlarge not only
European but also African power, if in both cases at the expense of other
Africans (cf. Lonsdale, 1977). Here too the historiographyis in flux. When
autonomist paradigms of power held sway, Imperial history was quite distinct
from and opposed by Africanist history. As explanation becomes more
asymmetrical,so the two traditionsconverge.
Theories of imperialismhad little to say about African history until Robinson
and Gallagher (1953; 1961) pointed out that it was a formal empire, colonial rule,
which needed explaining, not just economic penetration; it was imposed only
where precapitalistpolities collapsed under, or were able to resist, the subversions
of free trade. For all the anti-Marxist approachof this "peripheraltheory" of
empire (cf. Fieldhouse, 1973: 76-84, 460-63), it does in fact fit remarkablywell as
a concrete, political corollaryto the abstractionsof articulation,giving them life
and movement. Hopkins (1973: ch. 4; 1980b) has done most to apply this
approachin detail, so that conquest does not so much bring African history to an
end but respond to and provide new structuresfor the expression and control of
its continuingconflicts.
The underside of conquest and colonial state-formationwas one of the first
magnets for the new African historiography of the 1960s, of which Ranger's
(1967) study of the Rhodesian rebellions of the 1890s remains the most
memorable.Its historiographicalfate is, however, a cautionarytale. Palmer (1972:
293f) was moved to protest that the creative autonomy of political action had
been so completely transferredin the literaturefrom colonial to African initiative
that it was no longer clear what Africans were reacting against, or within.
Ranger's own study had made that clear enough, but there was considerablepoint

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 185

in Palmer's complaint, which was that the structuresof Europeanpower and their
rules of access were just as formative of the African reaction (and especiallyof its
evidential legacy) as any imaginativeAfrican ideologicallabor-the importanceof
which in the Rhodesian case has been called into doubt by Cobbing (1977) and
Beach (1979). Power, we had to re-learn, is a reciprocal process; it can be
misleadinglysimple to separateit into initiative and response.
The new growth point in the historiographyof conquest is the question of
production, in the guise of the spread of disease and its consequences for
demography, or in competitions for labor. Both enquiries link imperial and
African history, in the bush. It has recently been suggested (Patterson and
Hartwig, 1978: 4) that the first four decades of colonial rule were "the
unhealthiest period in all Africanhistory" as the uneasy equilibriumbetween man
and disease was disruptedby trade, conquest, and migrantlabor.If Ford (1971) is
right, rinderpestin cattle, attributableto the imported meat supply of conquest,
was responsible for the expansion of tsetse-infested bush and the terriblescourge
of sleeping-sickness. Migrant labor could also mean a losing battle between
African cultivatorsand their hostile environment (Iliffe, 1979: ch. 5). Openingup
Africa could entail the literal destruction of modes of production. And military
conflict between whites and Africans was often occasioned by demands for labor,
demands resisted not just by peasants but by the chiefs who were determined to
hang on to their control (e.g., Weiskel, 1980). Ecologicaldisasterand conflictover
labor supplies had no one effect on the processes of articulation.Rinderpestcould
force out to work those who had lost their cattle (e.g., Van Onselen, 1972;
Beinart, 1979), but the spread of tsetse reduced the scope of any productive
activity in Tanganyika.The opposition by chiefs to government demands for labor
or controls over land could make colonial governments stop in their tracks,
abdicating direct supervision of the export trade to the chief-and-merchant
coalitions whose competitive drive for profit had called in the consuls in the first
place. In coastal British West Africa, it is just plausible to see imperialismled
captive in this way by its collaborators,so that colonialism "only scratched the
surface of Africa's historic continuities" (Robinson, 1977: 162), a good enough
metaphor for syncretic articulation.The same image will not do at all for South
Africa, where the better organizationof African labor supply was early seen by
the British as a necessary condition for resolving the area's white competitions
into a federatedunity loyal to Empire (Etherington,1979).
State Formation: The Case of South Africa
The historiographyof South Africa is a huge and fascinatingsubject in itself;
but perhaps an outsider can offer one suggestion, emboldened by the recent
publication of four helpful, generalizing works (Adam and Giliomee, 1979;
Greenberg, 1980; Marks and Atmore, 1980; Fredrickson, 1981). It is that, as in
the study of African kingdoms and colonies more generally, so too for South
Africa, historicalunderstandingis best advancedby turningfrom originsand ideal
types-Calvinism, capitalism, class, for instance-in favor of complex and
interlocking process. The politics of South Africa's formative revolution in
government, from 1899 to 1910, was not only complex but well documented;
until its symmetrical uncertainties are fused with the rather functionalist,
Althusserian, asymmetries which have so far been developed, the South African
state will appear to be too abstractlyautonomous, too manipulativelyintelligent,
dealing with each separate aspect of political economy with too untrammeled a

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
186 AFRICAN STUDIESREVIEW

will.
South Africa's modern historiographywrestles with the causative connections
between industrial development and African unfreedom. Liberal, idealist
historiography-but it could as well be Menshevik, legal Marxist, as in Arrighi's
(1967) analysis of Rhodesia-has argued that the apartheidstate represents an
extreme example of how previouslyexisting social relations may act as a fetter on
the development of the forces of production, although scarcely in those precise
terms. The more rigorous and brutal entrenchment of white supremacyafter the
Second World War was on this view of "counter-revolution,"to reverse "those
processes which were drawing the peoples of South Africa away from plural
division towarda common society" (Kuper, 1971: 459). Marxistrevisionists have
replied that South Africa represents, rather, an extreme illustrationof the need
for state power to assist late-start industrializing capitalisms to a parity of
competitiveness with more mature ones, by suppressingthe bargainingpower of
its labor force and by a comprehensiverange of subsidy and protection (Trapido,
1971).
Liberal interpretations assumed a dual economy. Industry represented a
dynamic capitalist enclave struggling against the irrationalitiesand inertia of
traditionalsociety, both Afrikanerand African (Houghton, 1964; Horwitz, 1967).
All state intervention in protection of white privilege was an institutional
expression of the old frontier mentality, by commandos in white collars;it sought
to arrest the full logic of a transitive articulationwhich, however much it might
break down racial exclusiveness in transactionalcooperation, would have more
than compensated for the loss of ascribed racial status by greater rewards for
meritocraticachievement.
Marxist revisionists argued that liberal historiographyhad in mind the wrong
counter-factual. South Africa's main industry was mining long before it was
manufacture, and without mining's export earnings and accumulated profits,
import substitution industry would never have prospered. It was an innocent
delusion to argue as if South African industrialpotentialwas blightedfor lack of a
skilled and mobile black labor force or a large black domestic market. Mining's
market was overseas, its laboring skills minimal. Mining prospered best with
tributaryrather than transitive articulation.This conclusion was based on one
main propositionand carriedone majorimplication.The propositionwas that the
peculiarly varied geological structure of the Witwatersrand,with so much low-
grade ore, together with the fixed price of gold on the market, made mine-owners
more than usually attentive to reducing such costs as they could. The price of
unskilled labor was the most easily slashed, by theirjoint operationof a recruiting
ring within all the customary and statutory restraints which already militated
against free African labor. The implicationwas that it was not white society as a
whole which set its face against the color-blind rationality of the market, but
specifically white workers (increasingly Afrikaaners) who struggled against the
efforts of white capitalists (generallyEnglish) to value their labor at kaffirlevels.
Nor was the differential price of white and black labor necessarily culturally
determined; the supply price of black labor had been previously suppressed by
conquest. The liberals'free marketnever existed, except as an ideologicalweapon
againstwhite workers.South Africa was a raciallydivided class society, not merely
a culturally plural one (Legassick, 1974; Johnstone, 1976). But what did this
imply for the workingsof the state?

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 187

For Johnstone, the South African state was more or less simply the executive
committee of the white propertyowners. Davies (1979) has added some of the
complexities of representation. In his view, class antagonism between white
owners and white workers gradually changed from direct conflict to the
suppression of white working class politics by their co-optation and protection
from above. An internally divided white bourgeoisie as a whole needed their
assistance in controlling black workers. A more stable capitalist order was well
worth the marginalcost of class bribery.Capitalistdominance created a cohesive
white supremacyin order to make South Africa safe for capitalism.Bozzoli (1981)
has tracedthe culturalcorollaryto these arguments, the ideologicalformationof a
consolidatedwhite ruling bourgeoisie, a fusion of elites with diverse origins and in
perplexingdifficultiesof accumulation.
All this is a salutaryantidote to the previous race-fixationof much of South
African historiography.Yet, from the literature so far cited, there emerges a
narrowlypurposive yet rather abstractpicture of the South African state; it has a
mind of its own and is seen to do things virtuallyon its own, despite (or rather
because) of the great press of conflictingcapitalistinterests upon it (Davies et al.,
1976). From out of a useful figure-of-speechthe state has emerged as a sentient
being, organizing capital's hegemony. If the state is indeed reducible to the
institutionalized social relations of production, with, in capitalism, their
characteristic tension between the competitive will of all capitalists and the
General Will of capital in conflict with labor, then it is difficult to see how that
particularreificationis to be avoided. But it comes close to Bonapartismwithout
visible means of support. A positivist reaction is beginning to set in. Capital
cannot actually do anything, any more than the state. It gives more power to
groups which are sufficientlywell organizedto seize and protectit, it sets limits on
the purposes to which power can be put. By the late nineteenth century, however,
capitalwas also so diffused a relation in South Africa, so variouslyconstituted in
private and public forms, that a historiographywhich concentrates narrowlyon
one strand only of its social formation through time will produce a functionalist
rather than historical, processual analysis. The social relations of production in
late nineteenth-centurySouth Africa, as expressed at the level of the state, were
in fact divided into at least three competitive units. Each contributed to the
subsequent formation of the unified South African state. Moreover, the
oppositions of British imperialism and Afrikaner nationalism can neither be
excluded from the analysis nor reduced to-nor of course entirely divorced
from-the needs of capital.
Cape Colony had witnessed the nearest to a capitalist transition, as strong
exchange relations with the overseas market combined with the prolonged
disruptionsof conquest to separatemost non-whites, as well as whites, into either
laborers or property-owners.The dominance of merchant capital encouraged the
adoptionof typicallymid-nineteenthcentury liberalsolutions to social and political
order, with the non-racial if highly unequal mediation of master-and-servant
relations in the courts, and with a franchise restrictedby propertyand education
rather than race (Trapido,1980). Natal, more recently colonized, was also a more
syncretic formation. Its control evoked other techniques, other ideologies, which
emphasized the separation of black and white rather than their economic
commingling (Welsh, 1971). In the interior, the several Boer republicswere until
the 1870s rather tenuously connected to the wider market, they were barelystates
at all except as coalitions of big men; they therefore permittedgreat freedom to

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
188 AFRICAN STUDIES REVIEW

individual propertyowners in the control of their dependent workers (Legassick,

1977; 1980).
It was not just the further penetrationof capitalistrelations but more precisely
their sectional embodiment in British imperialismwhich forced the amalgamation
of this incongruous inheritance.British colonial state power was the one constant
in the baffling array of contradictions which marked the process. British state
institutions outraged the "personal pluralism" (Legassick, 1977) of the Boers by
making labor relations the business of the courts; British liberalism had also
abolished slavery. Conversely, the British state institution of electoral
representation provided a means for converting outrage into an organized
opposition within the state far more easily than the ad hoc petitioningpracticedin
the Republics;and Britishcolonial conquests capturedmany more Africans in the
toils of taxation and vagrancylaws. Moreover, for colonial officials, the control of
natives was as important as the recruitment of labor was to the mine-owners,
whose unprecedentedlyvast investments demanded regularityof supply (Jeeves,
1975). By the 1890s, colonial rule was an essential element in South Africa's
relations of production,in Mozambiquewhere so many workerswere recruitedas
much as in the Britishcolonies. Indeed, if social relations structurea given mode
of production, then it may be as suggested by Frederickson(1981: 220) that the
revisionist Marxist proposition that low grade ores and fixed gold prices
necessitated a tributary-labor solution is as "technicist" as Meillassoux's
agriculture-logicof social reproduction or Goody's guns and despotism. The
ultra-exploitabilityof labor under its strict colonial control may, rather,have made
the extractionof low-gradeores into a business proposition,setting the patternfor
what came after. When the second mfecane of rinderpest, the British war, and
drought disrupted pastoral Boer society, it was the politics of British
decolonizationwhich regroupedit in the Union. And in facilitatingthis process, it
may be that the specificallycolonial white perception of the collective need for
white prestige was as important an ideological code as the more fiercely tribal
sense of vocation which the Century of Wrong had imprintedin most Afrikaners
(for which, see Moodie, 1975).
To attributeall South Africa's ills to Britishcolonialismwould be as stultifying
ahistoricalas to relate them all to inherent racism or to capital.Nevertheless, the
existence of its state power did provide particularinstitutional economies; and
these compensated for the costs of the collusions (Bates, 1979: 241f) which were
requiredto pacify the disordersamong the white strangers (cf. Miller, 1976: 272)
during the competitive process of South Africa's state formation. Nor does one
have to postulate an abstractGeneral Will to discover the intelligenceof the state.
One need search no further than the British colonial officials, with their specific
yet all-embracinginterest in creatinga stable white society in South Africa which
would also be a loyal link in the strategicchain of Empire.These were the sort of
men who with the sound of battle in the distance looked down on a Boer town,
"built castles on the air and on the Kopje and laid out terraces and avenues in
imagination," hoping for "some good young architects" to come out from
England (Curtis, 1951: 135f). Marks and Trapido (1979) have stated this
argument well but perhaps too narrowly, with Milner's administrationas the
orchestrator of the mine-owners' needs for efficient government, economic
infrastructures,and disciplinedcheap labor. There were in fact violent clashes of
interest between the architects of capitalist society and the needs of capital,
illustratedat their simplest in the dispute between the mines and the government

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 189

over Johannesburg'smunicipalboundaries.The mine-ownerswanted none of the

costs of inclusion in municipalresponsibility,but to the British, it was vital that
Johannesburg "should become a place in which people will desire to remain,
rather than a mine of wealth whence men look to rapidlyaccumulatefortunes for
export and enjoyment elsewhere." They had so to engineer political structures
"that each class shall bear the political and social burdens which should fall to
their lot as members of an economic whole, and that one class should not be
allowed to separate its life from another class with which it is bound up by an
inseparableeconomic tie" (Curtis, 1951: 258, 260). It would be difficultto find a
more concrete statement of what is said to be the abstract need of capital-in-
general. And it was a statement of a distinctlypoliticalopinion concernedwith the
creation of a civil order in which white capital and white labor should both be
represented.
Moreover, for this colonial bureaucracy,the first-formedamong South Africa's
managerial bourgeoisies, racial segregation was by 1900 a customary British
defense of the state against the belligerentracism of Britons overseas. It may be
as "subjectivist" to make these points as it is to stress the cumulative racism of
Afrikaners. To this there are two replies, both recapitulations of earlier
observations. First, that no group can operate at all in human society without a
shared ideologicalcode, however much that may have been sociallyconstructedin
the past. And secondly, that while in so extraordinarilycomplex and contradictory
a society as South Africa, some degree of Bonapartismdoes seem to have been a
necessary attribute of the state until the consolidation of a national bourgeoisie
after the Second World War, Bonapartismcannot rest upon functional needs
alone. It requiresan institutionalframeworkstaffed by a class of officers,cohesive
and habituated to power. In Greenberg's view (1980), the Bonapartistelite has
now acquired a self-interest in the administrationof racial domination quite
distinct from and increasinglyat odds with the more subtle hegemonies available
to manufacturingand even agriculturalcapitalists,but without the priorautonomy
of the state these more recent possibilities of bourgeois hegemony would never
have existed.

Change and Continuity in Colonial Africa

In South Africa, a capitalist transition has been imposed upon a base of
tributary articulation; the present debate within white politics centers on the
question how these two processes may best be combined consistent with the
preservationof white supremacy (Adam and Giliomee, 1979; Greenberg, 1980).
The regimes of colonial Africa faced a similar question in the 1930s and 1940s,
but in an entirely differentcontext. A transitivearticulationhad proved impossible
where conquests had merely scratched the surface; tributary articulation in
support of migrant labor was scarcely thinkable where revenues were derived
from the export of penny-packetsof diverse argiculturalcommodities rather than
gold or other minerals. White supremacy,outside the scattered patches of white
settlement, was more a matter of commercial profit than the control of labor.
Household production continued to be the material base of the politics of
collaboration.There had thus been a syncreticarticulationof modes of production
which cannot even from the standpoint of Africa in the 1980s be analyzed as a
necessarily prolonged and contradictoryphase in an essentially transitiveprocess.
It was not an economic order which served the needs of capital;it was not a tidy
solution to the problemsof social control. It was an ever-changingkaleidoscopeof

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
190 AFRICAN STUDIES REVIEW

struggle within the endless range of social relations of productionwhich uneven

and combined development has thrown up. Its conflicts can be traced in four
spheres which often overlappedin specificcolonies.
Merchant houses strove constantly and often not very successfully to control
the quality and quantity of agriculturalexport production by limiting African
choices of crop and buyer (e.g., Brett, 1973: ch. 8; Miles, 1978). White farmers
increasinglydisputed with their African labor whether their terms of employment
were those of a tenantryor a proletariat;and when the Kenya settlers forced the
issue in the 1940s, they stirred up their tenantryto revolt in Mau Mau (for East
Africa, e.g., Furedi, 1975; Wambaaand King, 1975; Cooper, 1980; Kanogo, 1980;
for Rhodesia, Rennie, 1978; for South Africa, Morris, 1976; Trapido, 1978). It is
a struggle which has continued in Kenya since independence, between big farmers
and peasant corporations(Leys, 1971), and it will be the majorform of ruralclass
conflict in Zimbabwe,whether mediated through the state or not. Africanfarmers
too have encountered constant conflict with their own migrant labor, always
changing in its source of supply and challenging the nature of the labor relation,
pressing for wages, tenancy, or share-cropping according to their changing
comparativeadvantage (Richards, n.d.). Finally, syncretic articulationis seen in
the continual oscillations of public policy in southern Africa over migrant
industrial labor and its stabilization, an issue bedeviled by incompatible
calculations over labor costs, production efficiencies, forms of political control,
and the social composition of the domestic market.
Underdevelopmenttheory did not take these conflicts seriously enough. They
were not rationalproductiveresponses to world marketopportunity;they were the
best that both parties to the social relations of productioncould get out of a not
too unequal bargain.They followed no capital logic. If a class-logic interpretation
is preferred, then clearly it cannot focus simply on the purposes of Africa's
capitalists,because what Africa's capitalistscould do was decided as much by what
Africa's workers allowed or forced them to contrive as by what the world market
required (cf. Denoon, 1978). How much bargaining power was available to
Africa's workers in their conflicts depended upon how much of Africa's resources
in productive land colonial states allowed them to retain, for allocation within
social relations which, however much they may have been distorted by partial
incorporationin the apparatusesof the state, nevertheless protectedthe most vital
factor of production from the inequalities of the market. It was a mark of
syncreticclass conflict that it was not bounded by any one work place but ramified
uncontrollablybetween town and country (Cooper, 1981).
If syncreticarticulationdoes typify most of colonial and contemporaryAfrica's
productiverelations, then colonial states were scarcelyconvincing instruments of
capital logic; they neither wholeheartedly expropriated the direct producers
themselves nor allowed unrestrictedaction by would-be expropriators(for India,
cf. Washbrook, 1981). By the same token, colonial states were scarcely efficient
agents of modernization. No colonial state bore a single character. They were
intended to be instruments;otherwise, the recruitmentof their officialswould not
have been so jealously guarded by the metropolitanpowers. Their mode of rule
was Bonapartist.Capitalismwas weak in Africa, its productiveefficiencycontested
by pre-capitalistforms whose reproductivecosts were not a liabilityin the market.
The power of colonial states did bringcapitalismcloser to the point of production.
Colonial officials were an external elite, not locally connected. They were more
like the palace slaves, the ilari, of Oyo, than the upwardlymobile adventurersin

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 191

the asomfo stools of Asante. There is, therefore, room for debate within the
bonapartist paradigm. Kay (1972) and Wayne (1981) argue that the alien
bureaucracieswere so distanced from local politics that they suppressedcapitalist
endeavor as a form of political subversion, or at least tried by all means to
enmesh it in official regulation. Colonial rule was therefore self-defeating in the
very real sense that what was instrumentally optimal to the metropoles was
directly antagonistic to their bureaucratic agents, since increasing production
necessarily fostered the boisterous politics of local class-formation.As in Oyo, it
was impossible to reconcile proconsularpalaceand native town. Colonialismcould
not cope with a capitalisttransition.Conversely, it can be maintainedthat, for all
their lack of formal representativeinstitutions, colonial administrationswere so
feeble in coercive potential, so dependent on the politics of collaborationwith
African notables (Robinson, 1972), that the syncretic articulations of colonial
Africa's economies were in fact faithful registers of local class conflict, and
expressed within the changing rules of access to colonial government ratherthan
ragingoutside and against it (Lonsdaleand Berman, 1979; Berman, 1980).
At some point in time, colonial rule began to lose such initiative as it had
possessed. Its administrativeinstitutions lost the power to co-opt African energies
and instead served as the platform for ever-increasing African demands. Its
symbols lost their masteryfor rulers and ruled alike; it began to belie its myths of
legitimacy.Its economic potential began to look like frustration.The responsibility
for mapping the future began to shift from rulers to ruled (Berque, 1967; Iliffe,
1979: ch. 11). All of these uncertaintieshad dogged various colonies at various
times since their foundation, but graduallythey came together in a momentum
for change which was not to be gainsaid. How one interprets this change and
where one places it in time depends on one's general interpretationof colonial
rule; and on this interpretationhangs one's understandingof Africannationalism,
the social process which gathered the unease of transitioninto a positive drive for
transformation.
In yet another vast simplification,one can say that in the literaturethere have
been four contending stimuli to the second revolution in government which
seized hold of most colonies some three or four decades after that which had
brought them into being. The earliest to be recognized was taken from the
autonomist paradigm of colonial states, and concentrated on their changing
ideologies; 1945 was the conventional turning-point.More recently, we have been
made aware of the claims of the 1930s, when largely open economies began to
close against the stringencies of the depression; colonial regimes became more
protectively instrumentaltowards their major capitalistinterests. Building on this
institutionalbase of public managementof the market, colonies began to nurture
local manufacturingafter the Second World War, but whether following the logic
of capital or intervening in a new arena of class conflict is in question. Finally,
syncretic articulationin the countrysidealso began to run up against the limits of
expansion. All these shifts of public policy contained the contradictoryelements
of closer state control and encouragementto further capitalizationof production.
All contrived to find new structures of African collaboration, new
misunderstandingswhich Africans strove to work to their own advantage. The
economic and institutionalbases on which the elites of African nationalismfused
together as legitimate estates were as various as the successive crises with which
colonialism tried to cope.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
192 AFRICAN STUDIES REVIEW

The obvious break in colonial rule for those who wrote in an idealist idiom and
with a symmetricalanalysis of power came in 1945. Before the war, colonial rule
(Britishperhapsmore than French) suffered from what Coleman (1963: 160f) has
called a "fatal dualism" in its alibi for domination, torn between a politic respect
for Africa's natural rulers, the chiefs, and an inability to conceive of any future
other than a nation state ruled by its unnaturalleaders, the educated elite. The
middle class official mind was hopelessly divided. Administrators' "normative
reference group" (Lee, 1967: 2) included both metropolitan colleagues
accustomed to the parliamentaryexpression of class politics and fellow prefects
alarmed lest capitalist relations rot the communal estates of colonial rule. The
backboneof social relationsat home, capitalismwas their usurperoverseas.
The war did not resolve these doubts, but a job had to be done. Mobilization
of men and commodities put unprecedented power into the hands of
businessmen, brokers, and clerks. The expansion of state interventions at home
infused with social welfare the purposes of colonialismabroad;and there were the
American allies to appease (Louis, 1977). The war opened Africa's eyes, too. It
was certainlya time of economic and social stress (Coleman, 1963: 251-55); more
importantly,militaryservice overseas, official anti-fascistpropaganda,accelerated
urbanization, all constituted a normative revelation. After the war, French and
British attempts to co-opt all their necessary new African collaboratorsin systems
of electoral representation provided the vital stimulus for them to mobilize
support, from the new flood of transitionalswho had been drawn out from the
traditional communities on whose parochialismscolonial stability had hitherto
depended (Coleman, 1954; Austin, 1964). With this periodization, African
nationalismwas an idealist phenomenon, the politics of mobilizationby educated
elites. It was The productof a colony's educationalhistory. the further dynamism
of post-war change was charged by the dialectical relationship between
metropolitan self-doubts and African certainties within the ideology of the
civilizing mission. The matrix of their accommodation was provided by the
condensed constitutional myths of the metropoles, local devolution in Britain,
centralizationfor the other powers (Young, 1970a; Smith, 1978); it was Portugal's
domestic dictatorship which outlawed any such movement in her colonies
(Anderson, 1962). More recent work (Robinson, 1979; Kirk-Greene, 1979; Cell,
1980) has revealed how the exciting prospects of innovating with the grain of
African nationalism rather than against the grain of African conservatism
reinvigoratedthe moral purpose of the men who ruled the British Empire, of all
rulers perhaps the queasiest, certainly the mandarins in Whitehall and the
administrativecadets, if not the governorswho had grown up in quieter times.
Nobody would want to deny the importance of the war, but there was an
earlierand different crisis in the 1930s. The collapse of marketsand revenues, so
soon after all colonies had borrowedto the hilt on metropolitanmoney marketsto
improve their exporting infrastructures, posed a real threat to the fiscal
reproductionof colonial states and, with it, to the continued flow of merchant
capital. So the global crisis of capitalism pushed tottering states and traders
together to displace the costs of private export losses and soaring public debt on
to the productive poor. Open economies began to close; markets began to be
administered(Bauer, 1954) for the same reason that they had been periodicallyin
the kingdom of Dahomey: to protect the powerful (Law, 1977b). Just as states
needed to sustain their rural African tax base, so they had to give to merchant
capital the licensed monopsonies which met its long-standing desire for

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 193

predictabilityin African supplies. The productive choices available to household

producers within syncretic articulation were closed down under the weight of
departmentalextension campaigns, inspected markets, compulsoryco-operatives,
production quotas, and buying rings. The inefficiencies of syncretic articulation
could no longer be tolerated. In default of a capitalisttransition, the state had to
assume the supervisory role of landlord, banker, and employer. It was in the
1930s that one can see the origins of the over-developedstate, but it was scarcely
over-developed;rather, it was the class-biased solution to crisis (from a growing
literature,e.g., Chanock, 1972a; Coquery-Vidrovitch,1977; Jewsiewicki,1977).
As its interventions increased, so the power of the state increasinglybecame a
bone of contention. Nationalism becomes a product of economic history, the
politics of appropriation.Faced with the twin disaster of collapsing markets and
the state-sponsored onslaught on the redundant middleman (for India, cf.
Washbrook, 1981), Africa's farmers and businessmen also organized, classically
in the Gold Coast (Kilson, 1970; Rathbone, 1973), if only to become the licensed
dependents of the state. From licensed dependents, they fought to become a
legitimate estate within the revolution in government which was now under way.
Whether conceived as the politics of mobilization or of appropriation -and the
two paradigms, as every, must be combined-nationalism was shaped by the
changing institutions of colonial control. As with the decomposition of ruling
solidaritiesin precolonialkingdomsor republicanSouth Africa, so also in colonies
(and in postcolonial states) one can trade a sequence of alliance and defection.
Colonies, like any state, needed the support of those their rulers feared most
(Wallerstein,1964b: 333). Earlycolonial regimes were ratherlike tribalkingdoms;
their politicaladministrationsfostered decentralized,clientelist channels of access
by chiefs to state power. The centralizingregimes of the 1930s and 1940s did not
abandon their earlier allies, but they sought to add to them the wealthy and the
educated.Like the new orders of chiefs in nineteenth century Bugandaor Asante,
it was intended that they should be obedient. The representativeinstitutions by
which they were to be disciplinedenabled them, instead, to convert the garment
of power loaned out by the state into privateproperty,and a class was born.
After the war, the momentum of change took hold of Africa's towns as well.
Colonial Africa was rent by industrialdisputes (Cooper, 1981) as soon as industry
put in an appearance.Moreover, it is often argued that manufacturingcapital's
need for prosperouspeasantries,disciplinedlabor, and bourgeois friends, required
the overthrowof all the politicalconventions by which colonies had hitherto been
governed (notably by Arrighi, 1967; Leys, 1975). Further, local manufacturecan
be seen as a consequence of global capitallogic. The Second World War hastened
the "recomposition of social capital" (Bernstein and Depelchin, 1979: 18f). It
became internationalizedunder the new hegemony of the United States. The Old
World's colonial powers scurried to prop up their archaic privileges against this
threat (and with memories too of the Japanesepenetrationof colonial markets in
the 1930s) by encouraging their own capitalists to invest in production in the
colonies (Cowen, 1980). Here was a reversal of policy indeed. Two qualifications
have to be made, however. First, there is little evidence yet for official
consultation with business over the strategies of decolonization (Hopkins, 1976;
Milburn, 1977). Secondly, colonial states had their own reasons for encouraging
manufacture, in an attempt to discipline by work the "guerilla army of the
underemployed" (Cooper, 1981) who were flocking from the countryside to the
towns. The new institutions of Trade Unionism were as much a device for

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
194 AFRICAN STUDIESREVIEW

politicalcooptation in the purposes of colonial rule, like legislative assemblies, as

an answer to the class interests of manufacturingcapital.Trade Unions, however,
proved to be rather unreliable nurseries of labor aristocrats;their elites were not
so much fused with as suppressed by independent Africa's manageriaal
bourgeoisies (e.g., Bates, 1971; R. Cohen, 1974; Sandbrook,1975; Jeffries, 1978).
This class conflict in the towns took its origins from the troubles of the
countryside. Here lay the final, most formidable stimulus to the colonial
revolution in government, a veritable "second occupation" (Low and Lonsdale,
1976: 12). Syncreticarticulationcould no longer bear the demands placed upon it
by the state, by rising African aspirations,by rising African population-the last
being the single most important social process in modern Africa and a direct
result of the powers of colonial states in ending the periodicdemographicdisasters
of epidemic and famine. Soil erosion, declining crop yields, the closing of the land
frontiers of extensive cultivation, conflicts over land rights, and challenges to
recognized tribal authorities-all these forced decisions which governments had
previously preferred not to face. The thicket of conflicting propertylaws which
characterizedsyncretic articulationhad to be cleared away if the intensificationof
agricultural productivity within finite resources were to be achieved. One is
accustomed to seeing agrarian crisis as characteristicof the eastern African
savannahs, but it ought not to be forgotten that the classic study of successful
West African cash-cropproduction,Ghana's cocoa, was made possible only by its
collapse in the post-wardisasterof swollen shoot (Hill, 1963).
Colonial governments had an alternative, either to dragoon battalions of
peasant laborers in soil conservation works, or, like pre-revolutionaryRussia, to
wager on the strong, encouraging rural class formation to take its course in a
capitalist transition rather than, as previously, blocking it with a reconstituted
traditionalism.Characteristically,they did both, and in the contradictionsof policy
lay the frustrationsof Africans. Officialsdreamt of creatinga stable rural middle
class, but it is not clear how seriously this should be taken in historicalanalysis.
The classic agriculturalrevolution, in Kenya, no longer looks as startlingas once
it did in its statements of policy. The grant of negotiabletitle to land was probably
less important than the simple removal of the restrictionswhich had prevented
Africans from rowing the more profitable cash crops (Smith, 1976; Kitching,
1980: ch. 9). Cowen, whose research is the most impressivelybased on the rural
data (1972; 1974; 1975; 1976; 1981), has argued that, far from accelerating
agrariancapital accumulation,the state's generalizationof extension services and
the linking of merchantcapitalto producercooperativesmay have retardedit. The
rise in household productivitywas spread too widely, allowing the independent
survival of cultivatingfamilies who would otherwise have been forced to sell out.
One of the men in charge of the process also doubted the capacityof state action
to create a middle class, although it might ease the transition;policy "was actually
based upon what a few more advanced African farmers were doing" (Brown,
1968: 79).
There are perhapsthree lessons to be drawnfrom this survey of colonialism's
second revolution in government. The first is that, like other revolutions, it was a
long and complex process, not an event. Periodizationhas to take account of the
habit of states to be in more or less continualformation.But secondly, it did mark
a profound change, a silent cataclysm in the deepening of state power over the
lives and choices of Africans. It was in part what it was intended to be: an
officially sponsored capitalist transition. Previous government interventions had

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 195

helped specific capitalistinterests, mining, farming, and commercial;they had not

been intended to foster capitalist relations in general. After the Second World
War, nation-building became official policy, if possible of course without the
nuisance of nationalism.Widerfranchises, modern forms of laborcontrol through
the unionization of stabilized work forces, credit provision for farmers and
traders, land consolidation under private title of ownership, all were necessary
elements in liberal capitalist society. Nevertheless, finally, the revolution was a
turning point which did not quite turn; as in the centralizationsof precolonial
kingdoms, underlyingcontinuitiesremained.
The shift from syncretic to transitive articulationwas blocked. The terms of
trade were against it, with the downturn in the commodity markets in the later
1950s. More fundamentally, official controls, designed first to grapple with
uncaptured peasantries and migrant workers, became the networks of African
opportunityas power devolved. Class formation did occur, but more within the
institutions of the state than by exclusionaryaccumulationon the ground. Many
successful careers involved straddling,as it has been called, between a variety of
occupations, farming, trading, government employment, all linked by the use of
public wages as private investment funds (Kitching, 1980). The best investment
was education. It was a period of acute social disorder, with many competing
forms of status and power, the mark of any period of rapidchange. There were
still many different moral communities and no one rule of law. It was a time of
vulgarization;old authorities waned, new freedoms were tested. "The dusty and
disintegratingharness that tells of former glories" hung useless in the huts of
ancient savannah towns (Goody, 1971: 36); elders in Uganda blamed the
uncontrolledrise of sorcery and spirit-possessionamong the young on bad habits
picked up in migrantlabor (Southall, 1956: 279); Ethiopiannobility lamented that
where once songs were sung only for them, anybody now could sing them
"cheaplyin drinkinghouses" (White, 1980: 537).
Modernization theory took an optimistic view of all this. Nativistic religious
enthuasiams were a passing phase. Towns were melting pots. As the politically
relevant strata expanded so tribes would fade away, residual categories in the
brave new world of voluntaryassociations (Deutsch, 1961). Africa's transitionals,
like those in the middle east, were "nomads of the spirit in search of a new
identity" (Riseman, 1958: 5). Scholars did indeed capture the precious
atmosphere of excitement in this time of hope, this "stumbling from night into
day" (Davidson, 1975: 44). However much nationalism may now be argued to
have been a delusion, it did for a few years appearto make all things new. Then
we did at least have human actors on our pages, even if too many of them were
rootless individuals. Now, unless we are careful, we have the bearers of class
forces. That is the language of cardboardcaricature.Modernizationtheory and
rigid class analysis are equally empty in this respect. As Balandier(1951) was well
aware, men cannot so easily detach themselves from old identities and
associations in so uncertaina world, nor cease to be troubled by the problem of
evil. Individuals they were, members of a class perhaps, but also human beings
tugged at every step by all the culturalsymbols with which their elders had taken
such pains to endow them.
As a result, the multi-class alliances of tribe (Post, 1972) and myriads of
clientelisms (Berman, 1974) survived as the effective social relations of power.
They were still the estates through which to gain access to institutions which
colonial regimes had created for controlling what they had anticipated would

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
196 AFRICAN STUDIESREVIEW

become quite a different world. Colonial Africa remainedunpredictableto its end.

Liberal assumptions about the development of individual citizenship were no
further off the mark than conventional Marxist class analysis (Kitching, 1980).
One presumed too much on the prospectsfor bourgeoisdemocracyin postcolonial
Africa, the other on the chances of socialist revolution. Equally, to see half-
formed, dependent bourgeoisies as the dupes of neo-colonialism (Leys, 1975:
25ff) was to adopt a conspiracytheory of decolonization consistent perhaps with
capital logic but plausible only for colonies through which the fires of mass
nationalism had never run. Except possibly in French West Africa, no colonial
power was able to actually to choose its successors. Independence, the power of
decision, did count, even if one must also ask, How much?
STATES
No paradigmof state power has long been able to structureour perceptionsof
contemporary Africa, since Africa has continued its habit of confounding
predictions.Social and politicalscience, unlike history, proceed by model-building,
so one can trade for post-colonialAfrica rathermore clearlythan for its kingdoms
and colonies the rise and fall of two paradigmsand the emergence of a third. A
clear picture is, however, a misleadingone, since perspectivestend to overlap and
cross-fertilizewith each other, and the problem on which they may have focused
has remained the same: how to explain what is generallyrecognizedto be a crisis
of authorityin the new states. But what is generallyrecognizedis not universally
agreed. There have been sharp disagreementsover what constitute the important
relations of power and, therefore, over the sense in which they can be said to be
in crisis. Africa's troubles have been attributedto the fact that it is a continent,
successively, of states without nations, of neo-colonies, and of syncretic social
processes.
States Without Nations
The autonomist and class-instrumentalmodes of postcolonial state power,
already outlined, coexisted together in time. The second was a worried
interrogationof the first, preoccupied with the elite-mass gap rather than with
class relations in productionand exchange. Centralto each was the "criticalphase
hypothesis" (Jamal, 1965; Pratt, 1976), that Africa's future was being decided in
a race between social and politicalprocesses. Consistent with the pictureof Africa
as a continent of dual economies, modern enclaves in traditionalsocieties, it was
argued that the social mobilization of men (much more than women) by
nationalist ideas, the Africanizationof jobs, urbanization,and so on, was bound
to create an increasing torrent of demands for public provision of goods and
opportunities. Two questions followed. The first was whether this largely
involuntary process of social mobilization could be outrun and contained by
conscious programsof politicaldevelopment, revolutions in government by which
state elites created nations. Under universal symbols, they had to engineer a
universal consensus, one rule of law; otherwise, social mobilizationcould result
only in increasingly conflict. The unregulated arena of competition had to be
demarcatedwith legitimatelyauthoritativeinstitutions.
The second question was whether this critical phase would be bridged by
consensual authority or coercive force. The answers lay in the nature of African
political parties and their connections with their social environment. Zolberg
(1966: 3ff), in an analysiswhich in many ways summed up an academicera which

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 197

was even then fading behind the gunshots of Africa's first militarycoups, divided
these answers into the hopeful and the gloomy. Optimistic answers, of which
Apter's (1955) was the archetype, assumed that Africa's leaders wanted political
mobilization, that they were able to evoke it with political imagination, that they
would be able to enlist private ambition for the public good in the new free-
floating resources of the party and its affiliates: trade unions, farmers'
associations, women's leagues, youth brigades, and so on. New transactional
patternswould mould new, national communities-as liberal historians, following
Marx himself, believed was capitalism'spromise for a multi-racialSouth Africa.
There were two varieties of pessimist. There was a minority, of whom Kilson
(1963) was one of the first, who doubted whether African elites actuallywanted
to pursue mobilizational policies once anti-colonial agitation had brought them
power. Mass demands were highly inconvenient once elite demands were satisfied.
This was to take an instrumental view of nationalism as the politics of class-
formation. There nevertheless seemed to be good reason for hesitating to apply
any recognizableclass analysis to the problem, as can be seen when one considers
Africa's first post-independence revolts, in the Congo (Zaire). Young peasant
rebels certainly saw their ruling elites as a new class, monopolists of freedom's
fruits (Fox et al., 1965). It is also true that their ideology was "eerily Fanonist"
(Jeffries, 1978: 339); it could as well be said to have been eerily Catholic, in its
adaptation of missionary ideals of community development, just as African
nationalism adapted the claims of the civilizing mission. Probably the most
pervading break in the perceived legitimacies of social reproduction occurred
between the generations as, with the state in pieces, rapidAfricanizationfroze to
a halt in government and business. Among the youth, educated but not yet
employed, "boundless hope was replacedby unrelieved despair" (Young, 1970b:
981). The perceptions of both African participantsand academic observers alike
were premised on the inevitabilityof upwardsocial mobility and the arbitrariness
therefore of any closure in it. There was as yet no conviction that any social
structures stood between individuals and full participantcitizenship, other than
residual communalisms below and the selfish abuse of disproportionatestate
power above.
Zolberg himself was representativeof the second brandof pessimist, believing
that colonialism and nationalism together had created so little by way of nations
that those who ruled Africa had insufficientpower to bridgethe criticalphase, not
so much that they could block it. Without the authoritywhich, say, Neo-Destour
had inherited from three hundred years of plausibly national history in Tunisia
(Moore, 1965), African rulers were bound to turn to force. No sub-Saharan
African state could be reasonablyexpected to have the common politicalculture,
the rule of law, which allowed for easy understandingsand compromise; and
where there are no rules, "clubs are trumps" (O'Connell, 1967: 183). Faced with
the unceasing demands of the politics of scarcity (for India, cf. Weiner, 1962),
politicians were increasinglyscared of their own incapacities.Frightened also of
each other, competitive ruling elites fell back on the instrumentswhich they felt
they could trust: their tribe and (increasingly) the gun (Zolberg, 1968; First,
1970).
This trend of explanation for Africa's civil instabilities suggested a ready
analysis of the militarycoups of the later 1960s. Nothing better illustratesthe way
in which colonialism had been dismissed as a system of domination than the way
in which political scientists played down the possibility of military, coercive,

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
198 AFRICAN STUDIESREVIEW

intervention in subsequent political development (Sklar, 1963: 497; Janowitz,

1964: vi). When the soldiers did move, they were seen initially not so much as
the self-interested monopolists of force as members of the one nationalinstitution
that did conform to formal organizaationtheory, which could therefore act as an
autonomous referee in managing the competitions and uncertainties of
modernization (Huntington, 1968). This image faded, too. African armies were
often the most ethnically divided of institutions, as the lower ranks of "martial
races" (Kirk-Greene,1980b) from the most backwardregions became officeredat
independence by educated young men from the richer and more powerful ones
(e.g., Luckham, 1971). Most coups started with some soldiers shooting other
soldiers; they were not very reassuring examples of Praetorianism.Clientelism
and competitive patrimonialism,organizationalineptitudes and personal greed,
political contingencies which political science was not well equipped to analyze,
seemed nonetheless to be more plausible candidates for explanation (Decalo,
1976). Few African armies in power attempted the restructuringof institutional
supportwhich was the only alternativethat Janowitz (1964) could see to alliances
with the civil service or the discredited civil politicians if armies were to rule
through power rather than force. Nasser tried most conspicuouslyin Egypt; one
can see other such attempts in Somalia, Mali, and Zaire, none of them convincing
candidatesfor productiverevolutions in government. Nigeria's army has been the
most successful because its task has been the easiest. Oil revenues have allowed
the continuous expansion of the politics of accommodationto all existing interests
which emptied most African treasuries soon after independence. Nobody now
looks to Africa's armies for its salvation.
Neo-Colonies
For underdevelopmentalists,or for anybodywho knew a little African history,
the main defect of modernizationtheory was that it was too domestic in focus. It
assumed that if only Africans could break through the critical phase to
consensually competitive politics, if states could create nations, then rational
development and equitabledistributionwould prevail.This was a redundant,even
utopian, hypothesis if Africa's crisis stemmed not from the failures of nationalism
and its organizationalrevolutions but from the limitationsof peripheralcapitalism.
Underdevelopment theory has by now spawned as many internal debates as the
modernization theory to which it was a reply. One can say, however, that while
modernization theory had an optimistic bias, seeking causes of failure in local
pathologies, underdevelopment or dependency writings have been
characteristicallygloomy, brightening only at the prospect of revolution. Their
main weakness has been an insufficient attention to the nature of the states.
Modernization theory was over-preoccupied with political logic; dependency
theory has tended to ascribe all Africa's problems to an inexorable capital logic,
driven more by the demands of the global market than by the conflicts of local
production and social order. It has been in seeking explanations for the
particularitiesof the post-colonial African experience within this encirclinggloom
that dependencytheory has begun to be transcended.
The first proposition of dependency theory was that class relations in
Africa-as elsewhere in the Third World-were derivative of its weak world
market position (Brenner, 1977). Africa's dominant classes remained the
metropolitanbourgeoisies, although, as I have tried to explain, their dominance
was scarcely unequivocal even in the colonial period when they purportedly

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 199

exercised direct control. Peripheraldependent states were not the locus of an

independent class power reproducingitself in the compromises of local conflict,
but of political strata whose power was derived from their intermediaryposition
between internationalcapitaland the peasant producersor workerswho were both
more or less subjected to capital's requirements (Leys, 1976). The second
propositionfollowed. The perquisitesof Africa's rulers were dependentupon their
allocationof state licenses to manufacturetrade, and so upon the redistributionof
rather static resources within tariff and other protections inimical to local
consumers; they were not themselves direct owners of productive assets and
unable therefore to procure the intensification of their exploitation. Africa's
instabilities reflected not so much the absence of a common political culture as
squabbles between internationalcapital and the bureaucraticclass for increasing
shares of the social product, which could be gained only at the expense of the
agrariansector (Brett, 1975). Worse still, dependence upon the imports of foreign
capital tended to favor multinational capital in manufacture, with its high
technology, limited employment potential, and feeble backwardlinkages to local
suppliers (Langdon, 1977; 1978).
It was difficult to see any way out of this vicious circle, any room for
mitigationof fate, save in the more or less idealistdeterminationof African ruling
classes to take the part of peasantsagainstcapital.It was, suggested Mazrui (1967:
26), "a profound ethical longing" which first made Nyerere so attractive to
anxious Western intellectuals;or Tanzaphiliastemmed from relief at finding an
apolitical politicianof a congeniallyFabian cast of mind who was preparedto act
upon the proposition that institutionaldemocratizationwas a necessarycondition
for economic growth-his belief, in other words, in the productivityof politics
(see also Pratt, 1976; Samoff, 1979). Tanzaphilia set a precedent for seeing
African socialism as the product of political genius, a choice rather than the
outcome of social conflict, because there was still so much uncertaintyabout how
Africa's conflicts could be seen to be structured (Phillips, 1977: 20). The
autonomist paradigmof the state lingeredon.
The difficultiesreally startedwhen the choice was made not by one man but by
ruling classes. What made one class radical while others were conservative was
hotly disputed. Shivji (1973, 1976) rather neatly solved the problem by arguing
that both were conservative in intent and equally compradorin their auxiliary
relationship to internationalcapital; it was only the means of conservation that
differed. Conservative rulers had some base in the ownership of productionand
exchange. Radicalshad no control other than what their ownershipof state power
gave them; the only thing radicalabout them was the populistrhetoricwith which
they masked in the public good their essentially private appropriation.This has
been an influential critique of voluntarist socialism, but the concept of the
bureaucraticbourgeoisie is also the logical conclusion of the dependency belief
that classes are formed accordingto the needs of the market rather than in local
conflict over productionand distribution.If the state is the intermediarybetween
internationalcapital and African peasants, and if indigenous capitalaccumulation
has been blocked by alien bourgeoisies, then, if one can talk in terms of a locally
dominant class at all, its formation could only be in and through the exercise of
state power. A double instrumentalismresulted. Internationalcapital needed the
state for the local orchestrationof supplies and the preservationof order, local
elites needed the state in order to become a class; in this mutual need, neo-
colonialism was made secure, and a relativelyautonomous Bonapartismfounds its

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
200 AFRICAN STUDIESREVIEW

intermediaryfunction (Leys, 1975: 207-12).

To conceive of elites becoming a class is, strictly speaking, to confuse one
discourse with another (Lloyd, 1966: 49-62). For scholars who believe that all
men can usefully be assigned to a class position, confusion can become worse
confounded by describingthose of uncertainclass standing-and that means most
Africans-as members of the petty bourgeoisie; it is a categoryabout as wide as
some of its equally bewildered predecessors: "natives," "the African,"
"peasants," "transitionals." The plastic political consciousness of the petty
bourgeoisie has, however, provideda catch-allexplanationfor the observablefact
that Africa's rulingclasses are themselves the site of conflict over the purposesof
state power (Saul, 1974a). Kinless outsiders of the capitalistmode of production,
their multiple ties of ideology, self-interest, and identity, can conveniently make
them capable of almost anything (Mamdani, 1976)-as, for instance, Williams
(1976) has robustly objected. This is one reason why there has been so much
scholarly interest in the People's Wars fought against the Portuguese and
Rhodesians. If the petty bourgeoisie were to win any power against these ultra-
colonialisms, they did not have any choice at all but to "commit suicide as a
class," in Cabral'sfamous phrase (1969: 89); it was a path forced upon them by
their need to take the part of peasants so as to create a base for guerrillawar. In
other words, the revolutionarypetty bourgeoisie may have started a process by
imaginativeideologicallabor; but it was also in the process of war that they were
themselves created. People's War has given class formation and the social
relations of subsequent state power an observable history, a history of class
conflict. It was also a history which dialectically linked colonialism with
independence. This turns out to be the overriding weakness of dependency
explanations for the traumas of contemporary Africa; however contradictory
capitallogic and idealism may be in appearance,they share a common inattention
to history (in this case, colonial history) and therefore to the social processes
which have gone into the makingof states.
Social Processes
Modernizationtheory was premised on symmetricalpower, on the assumption
that those who had it could use it to good or ill. Underdevelopmenttheory was
uncomfortably bestraddled between a functional capital logic and symmetrical
class decision, without the support of intervening structures of conflict. To see
modern African states as condensed social processes ought to enable one to
overcome the difficultiesof both approaches,without ignoringtheir insights. It is
not necessary to be a Marxistin order to explore this perspective;Marxismhas, I
have suggested, as many internal contradictions as any other school of
explanation. Indeed, I would go further: it is essential for scholars, whether
Marxist or not, to rediscover older truths expounded by scholars of all
persuasions and none, if the discussion of African states is not to be as narrow
and sterile as much underdevelopmenttheory has left it. A great deal has been
said in the past about Africa's social differentiationsand about the organizationof
power through access to state institutions. We need to remember this, to avoid
both too blithe an autonomy of political choice and too functional an insistence
that Africa's ruling classes have got to do what in capitallogic they simply gotta
do. As always, historicaltruth lies somewhere in between.
Four points are now, I think, being fairly commonly made in the attempt to
rethink the characterof tropicalAfrica's dominant classes. None of them is really

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 201

new. The first concerns the relationshipbetween tribe and state. The dominance
of colonial and postcolonialstates in administeringcommodity exports, allocating
producercredit, building infrastructuresand so on, all part of their orchestrating
role, has, to repeat, made politicalaccess to their investment decisions and spoils
system essential to all commodity producers. Tribe has been the most
economicallyavailableconstituencyfor mobilizingpressure for access. Indeed, the
growth and crystallizationof tribes has been recognized as one of the most
vigorous and creative social processes in modern Africa (e.g., Low, 1971: 3-7;
Young, 1976; Iliffe, 1979: ch. 10). Moreover, tribalism, far from being (as was
once supposed) an index of African states' fragility,may well be an indicationof
their strength, precisely because what states have to offer in smoothing access to
world marketsis so important.Stronger,richerareas, the ones that count, will use
assertive tribalism to gain state access; it is the weaker ones which tend to
mobilize retreatistethnicity to resist state power in its alliance with capital (Post,
1972). Tribalismmay as much strengthenstates as weaken them.
The second point relates tribe and class. Tribes are like states or, indeed,
capital. They themselves do nothing. They are mobilized by groups of like-
interested people who can persuadea latent or potential community to think and
act like one in their support. Those who most need access to the state and can
profitby it were and are Africa's emergent bourgeoisies (Sklar, 1965; 1967). Even
before access to the state was so important to them, they had learned that the
social relation of merchant capital was most easily deployed among those who
shared their cultural symbols and spoke the same language (Ehrensaft, 1977).
Laborerscould just as well use tribe as a means of workingthe misunderstandings
of employers' stereotypes. How many Luo first thought of themselves as such
when they discovered that this identity would win them a wage-premiumon
Kenya's sisal estates? Tribe is as much a weapon of struggle from below as of
control from above. It is not simply a benighted form of false consciousness
(Mafeje, 1971).
The third point connects state and classes in formation, through history. It was
the colonial response to the crises in syncretic articulation,beginning the 1930s,
which hardenedthe links between power in society and access to state institutions.
It was then that the managerialbase for contemporaryruling classes' power was
laid. Who were the grandfathersand fathers of today's ruling bourgeoisies?I have
argued, following Robinson (1972), that in their second revolution in
government, colonial states were obliged to rely increasinglyupon new African
collaborators,but often the new men were the old chiefly collaborators'educated
sons, a local illustration of the truth seen long ago by Schumpeter (1955: 151;
first publishedin German, 1927), that investment in new sources of powercan be
(but is not necessarilyalways) easier for those who possess the resourcesof older
power. There is a new interest in colonial chiefs as elements of a continually
changing ruling class, rather than the distractedbearers of intercalaryroles and
values that they were once seen to be (contrast Tignor, 1971; Tosh, 1973;
Vincent, 1977; Markovitz, 1977: ch. 4; Guyer, 1978; with Fallers, 1955). They
may have propped up colonial rule; their sons very often inherited its mantle of
power (Kipkorir, 1969; 1974). If we wish to understand Africa's managerial
bourgeoisies, it is not only the contemporarylogic of capitalthat counts but also
the past history of colonial education, native treasuries,and native tribunals.
Finally, one needs to relate dominant classes to the problem of their social
reproduction. As remarked in an earlier section of this essay, it is becoming

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
202 AFRICAN STUDIESREVIEW

recognized that there is little value in defining a class like the bourgeoisiein terms
of its fractionalinterests (Sklar, 1976; 1979). What makes a class "classy" is its
reproductionthrough coalition-buildingover time in order to deny resources to
others. How Africa's ruling classes exercise power, how they assemble it and
sustain its potential through rewardsand punishments, is as importantto know as
in what way power is functional to multinational capital. Large interests,
strategicallyplaced in terms of state access, can expect to recoup the costs of
political activity in the enjoyment of subsequent privilege. Conversely, ruling
classes can use the division of such rewardsas the resources by which to govern
(Bates, f.c.: ch. 5). There is a poor outlook for peasants on either count, but
equally a poor prospect for independent businessmen of any sort. Managerial
bourgeoisies reproducethemselves by managing,but that is incumbents'talk, like
kingdoms centralizing.From others' perspective, it means regulating,interfering,
commanding, in often arbitraryways dependent upon the internal politics of the
ruling classes. As in nineteenth century Asante, many African rulers seek now to
suppress all forms of wealth which are not derived from state access (Wilks,
1979). Such socially divisive reproductionin the present is always grateful for a
myth of legitimacyfrom the past, so that one needs to understandcontemporary
power not only in its relationsto the inheritedinstitutionsof the colonial state but
in its claimed connections with the risks of anti-colonial nationalism (Atieno-
Odhiambo, 1981). Nationalism does not cease to be an engine of class formation
when it is no longer in movement.
All these observations suggest points of comparison.The modernizationand
underdevelopment paradigmswere models of similarities, too; social processes,
being molded by history, ought to reveal differences. Development or
underdevelopmentare two perspectives on the process of Africa's incorporation
into some variouslydefined global society, but there was no one such process and
therefore no one necessary outcome. I can conclude by suggesting three sources
of difference in Africa's contemporaryexperience. There are many more. The
internal dynamicsof incorporationvaried from colony to colony, and often within
any one colony, accordingto the vigor with which the regime was pressed to assist
capitalists; according to the timing of the major expansions of commodity
production; and according to the degree to which very different precapitalist
relations of productionbecame embedded within the administrativeapparatusof
the state. On such divergencies depend the productivepotential of nationalism's
own revolutions in government.
The first of these variables is best illustrated by Colin Leys's changing
evaluation of the contemporarysignificanceof Kenya's white settlers. In his book
(1975: 39), he dismissed them as a mere epiphenomenon of the incorporative
process; they were a historicalcuriosity. In his second thoughts, however (1978:
259f), their commandeering of land and government services was seen to be
crucial to clearing the way for exclusionary capital accumulation and surplus
appropriationfrom labor by the contemporarybourgeoisie. If Kenya is more of a
bourgeois state than most in Africa, that is because it is also more fully
proletarianizedthan most (see also Njonjo, 1977); it had a more vigorous colonial
administration.Most other African bourgeoisies have still to cope with all the
intricaciesof syncretic articulation.So the shortcomingsof colonial states can be
said to be visited on the next generation.
Colonial shortcomingscan nonetheless prove to be priceless assets for today's
accumulators.This is nowhere better seen than in the current contrast between

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 203

the Ivory Coast and Ghana, superficiallysuch similarneighbors.It was admittedly

not much thanks to the colonial state that Ghana's ruralcapitalistsgot off to such
an early start in cocoa production (Kay, 1972). But it was due to the ultimately
fruitless French support for white planters in the Ivory Coast that the expansion
of cash crop productionamong the colony's peasantrieswas delayeduntil after the
Second World War. By independence, therefore, much of Ghana's cocoa lands
were exhausted, while the Ivory Coast enjoyed untappedreserves of fresh forest
soils. Whateverelse may be said about the economic rationalityof their divergent
development strategies,it must be true that primaryproductionin the Ivory Coast
has been far better placed to underwrite economic expansion than in Ghana
(Amin, 1973: ch. 2; Campbell, 1978). Africa's natural and man-made ecologies
have to be taken in deadly earnest if they are not to take man in deadlyearnest, a
point to which I will returnin my conclusion.
To come to the final area of divergence is to take up rather more distant
historical threads. Precolonial Africa's modes of production ranged from large
slave plantations under royal control to the self-sufficiencies of hunting and
gathering bands. This contrast exaggerates the difference between Northern
Nigeria and mainland Tanzania but will serve to emphasize my point. There is
probablyno more firmly consolidated dominant class in sub-SaharanAfrica than
Northern Nigeria's, for three reasons. It enjoys the ideologicalsupport of Islam,
with which its forebears forged their own revolution nearly two hundred years
ago; it is a rule of law which links trade and authority. Its emirate political
machines were preservedand strengthenedas the local levers of the colonial state.
And large concentrationsof rural population, often debtors where they are not
also the descendants of slaves, are readily available for myriad forms of
dependent production(Whitaker,1970; Hill, 1977).
The historiographyof contemporaryTanzania deserves a full-length study in
itself; it would perfectly illustrate the paradigmshift between the autonomous
state elite striving for democraticmobilization (Bienen, 1967; Hyden, 1968; Ingle,
1972), to the instrumentalities of the state for class-consolidation, first
emphasized by Shivji (1973, 1976) and taken up by most who have followed
since, although Saul (1974a) and Pratt (1976) continue to claim, if from sharply
contrastedperspectives, the primacyof politicalideology and morality.But that is
not the present issue, which is the comparativeweakness of the Tanzanianruling
class. Precolonially, what is now Tanzania had had only few and scattered
intensive production systems; in colonial times, its cash-cropping peasants,
likewise, were only rarely the base for impressive local patronagenetworks; and
TANU was a more than usually bureaucratizedparty, tending to keep a distance
from local bosses (Iliffe, 1979). This historical background, its intricacies not
always sufficiently appreciated, has informed a widely accepted critique of
Tanzanian socialism which springs from the assumption that its ruling groups'
first concern is their own social reproductionby the maintenanceof control (e.g.,
Raikes, 1975; Von Freyhold, 1979; Mueller, 1980; Freund, 1981). Hostile to the
emergence of self-regulating and self-interested estates outside the immediate
patronageof the state apparatus,party and government officials have, it is held,
preserved their own indispensabilityby policies which concentrateresources and
restrictions upon the middle peasantry, the same smallholding yeoman of late-
colonial longings for ruralcalm. On this view, ruralsocialism is the locus of class
struggle between the rationalityof public servants and the stupidity of peasants,
incumbents' talk again for the only forms of resistance to state extraction of

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
204 AFRICAN STUDIESREVIEW

which households are capable. The managerialbourgeoisie, five hundred percent

more numerous now than the administrative machine at independence, can
expand only by bearing down upon producer prices in the countryside. It is a
struggle which appears to be ultimately self-destructive of the fiscal reproduction
of the state, as peasants have both switched to crops less vulnerable to the
clawbacksof administeredmarketingand taken to smugglingacross frontiers.The
rationality of management has at the same time been fragmented by the
clientelisms of articulatedmodes of production, so that the state has become a
contradictory social relation, fatally divided between overall planning and
particularisticpatronage (Hyden, 1980). Tanzania, in other words, does not look
so differentfrom many another Africanstate.
The general hypothesis can be tested only with data which we do not appearto
possess on the size distributionof farm holdings, crop by crop (Sender, 1982).
However much such data might reveal that there is more surreptitious rural
capitalismin Tanzania than is commonly believed, it would nonetheless remain
true that there is a significant contrast between Tanzania and its northern
neighbor, Kenya, over the level of taxation imposed upon producersthrough the
official marketing boards. Tanzania and Kenya had very similar precolonial
societies. African countries are to be distinguished therefore not only by the
different indigenous modes of productionwhich became articulatedto merchant
capitalbut by the manner of their articulation,which was as much a politicalas an
economic process (Foster-Carter,1978). Colonial history matters. In the colonial
period, Kenya's class formation was doubly distinguished from Tanzania's. It
experiencedmuch sharpersymmetricalconflict between capitalistinterests, so that
its marketing boards came to represent settler producer interests against the
trading houses. Its asymmetrical conflicts welled up in the Mau Mau revolt.
Tanzania had no real equivalent to either. Its planters were a political minority;
and Maji-Majiwas in a different age. So Kenya went through the two most
formative processes in the establishment of bourgeois freedoms: competition
between capitalists,and the anger of mass despair.Its rule of law has accordingly
been obliged to accommodateboth forms of countervailingpower: the legitimate
existence of society's estates. There is a limit, much less than is endured in
Tanzania, to what producerslarge and small will accept by way of the exactions of
the state. It used to be argued that Kenya's political stabilitywas conditional on
the lack of class formation (Bienen, 1974: 22f). It now looks more plausible to
attributeit to the very process of class consolidation (Tamarkin,1978). As it has
been said, "no antagonism, no progress."
Through examination of the social processes of class formation, one may
return to look afresh at the normative criteriaof nation-buildingwhich informed
our first, autonomist, model for understanding modern African states. How
dominant classes behave does actually matter (Dunn, 1978). The autonomist
model of states without nations accorded them too much freedom, the
underdevelopment model of neo-colonies not enough; it was too functional to
allow any room for the exercise of political responsibility (cf. Lukes, 1974: 54ff;
Sklar, 1979). To situate Africa's present rulers in their specific histories is to
explain the limits on what it is possible and rationalfor them as one class to do in
relation to others, but it does not ultimately explain what they do nor, more
importantly,how they set about doing it; if it did, then it would be impossible to
account for the moral uncertainties of power through which its forms are
continually in flux. To insist on the primacyof local history is also to cast doubt

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 205

on the distinction, so central to underdevelopment theories, between resident

ruling classes and the absentee dominant classes overseas. The compromises of
local control may oblige even the multinationalsto adopt a corporatedoctrine of
local domicile if they are to bank on the longterm profits of sharing in a
productiverevolution in government (Sklar, 1976). However much there may be
a pull in the other direction, towards the transnationalization(as it has been
called) of African rulers, it is still the growth of a local class cohesion, of restraint
in the exercise of force within a rule of law, which builds nations. Furthermore,
nations can actually call their rulers to account in a way impossible to access-
seeking tribalisms.
It is not very fashionable to make these claims; they are not often justified;
and if Kenya is the latent paradigmit is clearly not untarnished.Equally, if it is
held against the experience of another neighbor, Uganda, it is not at all without
virtue. There the consequences of internal class rupture, or its imperfect
consolidation in the first place, have been truly terrible.Moreover, once they are
broken, it is difficultto reconstructthe ties between state and class which sustain
and yet also restrain each other. Where there is no rule of law, and no nation
either, there is nothing to restrain the growth of state banditry (cf. Thompson,
1977: 294); nothing to gainsay the doctrinethat government be in the interests of
the governors, by which "to be an outsider was to be a loser" (Glentworthand
Hancock, 1973: 240); nothing to prevent the slide to the point where, on the
other side of the continent, the dictatorMacias could truthfullyproclaimthat "In
politics, the victor wins and the loser dies" (Artucio, 1979: 11); and nothing,
ultimately, to prevent that doctrine expanding from individualsto the destruction
of scapegoat communities who, since the cohesion of the state has gone, become
foreigners rather than members of a competitive tribal constituency (Kuper,
1981). There is rather a lot to be said for the state's legal organizationof the
upward flow of resources. Uganda in the wake of Amin resembles nothing so
much as an institutionless arena, a pathological form in which the would-be
bourgeoisie compete for resources in unmediated, (that is, violent) conflict
precisely because there is no constitutionallyaccepted executive committee over
them (Gertzel, 1980). As in nineteenth-centuryIbadan, there is no anointed king
to restrain the barons. The oppressions suffered by the people are the more
dreadfuland various in consequence.
CONCLUSION
There is no doubt that the consolidationof class and the growthof states have
been a joint process, in Africa as elsewhere. In Africa, as elsewhere, they have
entailed a contradictoryincrease in the power to oppress and the restraint of
power. They have not ended the struggles of the oppressed, but they have
changed their arenas. They have also failed to change a great deal. Africa's modes
of production have not been so transformed by incorporation into global
capitalism that all its people have been captured.There is still a great deal that
states simply cannot control; they have proved, throughout history, to be very
imperfect relations of production.Africa's major productmust be the staples with
which to feed its populationsand, as Ranger (1976: 26) has rightlysaid, "no one
knows how to grow large quantities of food in tropical soils without ecological
damage." For all the pressure of their kingdom's fiscal demands, it still took the
Kuba about two hundred years to fully incorporate the new higher-yielding
American foodcrops into their agriculturalcalendar (Vansina, 1978: 172-86). The

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
206 AFRICAN STUDIES REVIEW

agrarian history of Tanzania since the Second World War, in colonial and
postcolonial times, gives sharp reminders of the damage which states may do if
they try to push things faster (Wood, 1950; Kjekshus, 1977). Conversely, many
of Africa's more spectacular improvements in agricultural productivity have
occurred without any state intervention, if not indeed in despite of it. The
"enrichment of food production" (Oliver, 1977: 628) in the early East African
interior owed everything to colonization, at the time a distinctlystateless process,
by differing peoples (see also Stemler et al., 1975). There have been continual
experimentationswith different seed varieties and croppingpatterns,not so much
by agriculturaldepartments of state which so often got things wrong as by
individualwomen swappingseeds and ideas along the networks of neighborhood
and affinity (Hay, 1976: 96); one wonders nowadays how much that process is
seen as deliberatelydefiant of the urgings of extension officers to plant taxable
export crops. In its defenses against state power, marginal tribalismin the past
often promoted the most intensive agricultureof all, the terracedirrigationof
hillside refuges. Finally and famously, there was the case of cocoa, another
product of family enterprise, which was variously ignored and frowned upon by
the colonial state, while all the time it was the chief providerof the Gold Coast's
revenue (Hill, 1963).
The difficulties of food production are as nothing to the subversions of
capitalism.When they are articulatedtogether in the evasiveness of peasants, the
guerrilla mobility of workers (R. Cohen, 1980) and the imperiousness of
multinationals, then one may well question the sovereignties of states and
understandthe despairingkleptocraciesof their rulers. All these continuities will
continue to demand ever changing revolutions in government in search of their
control. Africa's observers equally will doubtless be condemned for ever to
remain nomads of the intellect in search of a new paradigm.
NOTES
1. I have incurreda substantialnumber of debts in the writingand rewritingof this essay.
I shall always be grateful for the original commission from the Joint Committee on
African Studies of the Social Science Research Council and the AmericanCouncil of
Learned Societies, and to Marcia Wright, who finally pushed me into acceptingit. It
obliged me to do a great deal of readingwhich should have been done years ago and
much thinkingwhich had been similarlyneglected;I had to become much more alert to
theoreticalassumptionswhich one had employed in teaching and writing but had not
fully examined. So the commission was to me an exciting processof learning.That the
process remains incompletewill be apparentto my readers-but that, if I understand
the term aright,is a determinatefeature of the academiccondition.
This revised version of the paper has benefited enormously from the clarifications
and objections offered by my fellow panelists in their written commentaries.Professor
Jan Vansina showed that my dichotomy between state and community was highly
misleadingfor much of precolonialAfrica;and subsequentreadingalmost convinces me
that it should indeed be droppedthrough "the trapdoorof historiography"(Seal, 1973:
6) -at least if the communityis seen as pristineor pre-state,as the antiqueconstitution
opposed to new-fangled methods of organizing power or oppression. Roy Willis
reinforcedsome of my thoughts about the competitive complexitieswhich infused the
formationof the politicalclasses and kingdomsof precolonialAfrica;and he confirmed
some of my doubts about the explanatoryvalue of modes of productionas a total
concept. With Kitching (1980: 4f), I believe that the concept needs unpackinginto its
constituent elements-as, emphatically,does the concept of state-if it is to illuminate
empiricalenquiry.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIAL PROCESSES IN AFRICA 207

Professor RichardSklar thought my originalpapertoo state-centricto explain many

social processes, which I would agree, and yet also too economistic, which I would not.
To give due weight to both economic necessity and to politicalcreativityis to my mind
the central conundrumof historicalanalysis, and I have laid considerableemphasis on
the relative autonomy of the politicalrealm in shapinghuman experience.But Sklardid
greatly clarify my mind on the question of the autonomist and instrumentalist
paradigmsof the state and started off new trains of thought on the interconnectionsof
state formation and class consolidation. Finally, Ivor Wilks upbraidedme for being
much too deferentialto the social scientists and their model-building.I had not, in his
view, insisted sufficientlyupon "the peculiarautonomyof the historicaldiscipline,"with
its patient exploration of the evidence following not abstraction'sfickle fashion but,
rather, the "developing logic of the inquiry." Wilks feared that in bending over
backwardsto be fair to non-historians,I might also have been caughtwith my trousers
down. In this revised version, I think I have straightenedmy historian'sbacka bit; it is
for others to say whetherI also buckledmy belt.
I also benefited from the suggestions made in discussion at the ASA Annual
Meeting in Bloomington, Indiana. Friends and colleagues in England, at seminar
gatherings in Cambridgeand at the Institute of CommonwealthStudies in London,
have also been generous with their advice. They will know who they are. Others were
kind enough to send their comments in writing:Bruce Berman, Basil Davidson, John
Dunn, Dick Cashmore, Fred Cooper, Cherry Gertzel, John Iliffe, Allen Isaacman,
David Miller, Gwyn Prins, Terence Ranger, RichardRathbone, Sandy Robertson, Bill
Tordoff, RichardWaller,Luise White, MarciaWright,and ChristopherWrigley.Patrick
Chabalinstructedme in politicalscience while also giving me lunch. With so much wise
counsel, my head is still in a whirl. Finally, I must thank MarthaGephartat the SSRC,
who kept calm when she had very cause to despairat the recalcitranceof my authorship.
There is a great deal of overlap between this essay and Fred Cooper's on "Africa
and the World Economy." This complementarityexisted independentlyin the original
conferencepapers.I for my parthave been very glad to incorporatesome of his insights
still more directlyin this revised version.
2. "Statelessness" is a standing invitation to conquest by an external power; my
suggestion in discussion at the ASA Conference that some doctrine of revocable
sovereignty could well have been applied to, say, Bokassa's Central African Empire,
Amin's Uganda, and Macias'sEquatorialGuinea on the grounds that they were killing
their citizens, was rejectedfor this reason.
3. Richard Waller spotted this analogy. See also Campbell (1978: 66), Crummey (1980:
116), and McCaskie(1980).
4. If, that is, it should ever be more generallyacceptedthat historyis not "a coherent and
worthwhileobject of study" (Hindessand Hirst, 1975:321).
5. Colin Lucas, review of William Doyle, Origins of the French Revolution(Oxford
UniversityPress, 1980), in TimesLiterarySupplement (London, 8 May 1981).
6. This is in any case an occupationalrisk in much of precolonialAfricanhistory, with the
sources so often being silent on matters of personality,motive, ideas, religion, and
aesthetics (Vansina, 1978: 11).
7. Warren attributes this uncertaintyto the influence of Marxism, but patrioticdoubt
about the progressive justification of Empire long predated any widespread
understandingof Marxism (Hodgkin, 1976: 12), and Marx himself was preparedto
concede a progressiverole to Britishimperialismin India.
8. See, for example, the dedicationof at least two books to other scholars"because he is
committed," "because they cared" (Wallerstein, 1966; Isaacman, 1976). I would not
agree with Waterman (1977: 1) that Africanistsare revealing an increasingmoral or
politicalcommitment.It has alwaysbeen there, and often radicalby the standardsof the
time.
9. The Marxist literatureon the state is vast and growing daily. My chief debts are to
Berman (1980; 1981); Corriganet al. (1980); Gough (1979); Harris (1980); Holloway

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
208 AFRICAN STUDIESREVIEW

and Picciotto (1978); Jessop (1977); Lamb (1975); Poulantzas(1978).

10. Governor Sir PhilipMitchellto Secretaryof State CreechJones, KenyaConfidentialNo.
16, Appendix.Nairobi, 30 May 1947.
11. R. Oliver, review of Goody (1971), in Journalof AfricanHistory,XII, 3 (1971): 492.
Mitchell (1955) is a good example of clubmen's attitudes,perhapsnow as much as then.
12. To judge by the fact that Leys (1965: 229) quoted Fanon secondhand from Dumont
(1966, originally1962).
13. I owe to Norman Stone this remark, which he is almost preparedto swear is not
apocryphal.
14. Greenberg (1980: 387-90) reverses the autonomy-instrumentalitysequence, seeing
embryo classes seeking their instrumentalends in the revolution in governmentand the
state apparatusthereaftergaining its own autonomous self-interest.I would not dispute
this formulationdid it not tend to play down the importanceof the state apparatusitself
and its internal conflicts in stimulatingand guiding revolutions in government. This
question is discussed at greater length in the section on South Africa's own state
formationbelow. Autonomy and instrumentalitydepend largelyon the viewpointof the
observer.

REFERENCES
Aaronovitch, S., and K. Aaronovitch. (1947) Crisis in Kenya. London: Lawrence and
Wishart.
Abu-Lughod, Ibrahim. (1967) "The Transformationof the EgyptianElite: Prelude to the
Urahi Revolt," MiddleEasternJournal21,3: 325-44.
Adam, Heribert, and Hermann Giliomee. (1979) EthnicPowerMobilized:Can SouthAfrica
Change?New Haven: Yale UniversityPress.
Adeleye, R. A. (1971a) "Hausalandand Bornu, 1600-1800," pp. 485-530 in J. F. A. Ajayi
and MichaelCrowder(eds.) Historyof WestAfrica,I. London:Longman.
(1971b) Powerand Diplomacyin NorthernNigeria,1804-1906:theSokotoCaliphateand
its Enemies.London:Longman.
Ajayi, J. F. A., and Robert S. Smith. (1964) YorubaWarfarein the NineteenthCentury.
CambridgeUniversityPress.
. (1965) ChristianMissionsin Nigeria,1841-1891:TheMakingof a New Elite. London:
Longman.
Akintoye, S. A. (1971) Revolutionand Power Politics in Yorubaland,1840-1893. London:
Longman.
Alagoa, E. J. (1970) "Long-DistanceTradeand States in the Niger Delta," Journalof African
HistoryXI, 3: 319-29.
(1971) "The Development of Institutionsin the States of the EasternNiger Delta,"
Journalof AfricanHistoryXII, 2: 269-78.
Alavi, Hamza. (1972) "The State in Post-ColonialSocieties:Pakistanand Bangladesh,"New
LeftReview74: 59-81.
Almond, Gabriel A., and James S. Coleman (eds.). (1960) The Politicsof the Developing
Areas.PrincetonUniversityPress.
Amin, Samir. (1972) "Underdevelopmentand Dependence in Black Africa-Origins and
ContemporaryForms," Journalof ModernAfricanStudies10, 4: 503-24.
(1973) Neo-Colonialism in WestAfrica.Harmondsworth:Penguin.
Anderson, David M. (1982) "A History of the Peoples of the BaringoPlains, 1890-1940."
CambridgeUniversityPh.D. thesis.
Anderson, Perry. (1962) "Portugaland the End of Ultra-Colonialism,"New LeftReview15:
83-102; 16: 88-123; 17: 85-114.
. (1974) Lineagesof theAbsoluteState. London:New Left Books.
. (1980) ArgumentswithinEnglishMarxism.London:New Left Books.
Andreski,Stanislav. (1968) TheAfricanPredicament: A Studyin thePathologyof Modernisation.
London:MichaelJoseph.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 209

Anstey, Roger. (1966) King Leopold'sLegacy: The Congo underBelgianRule, 1908-1960.

Oxford UniversityPress.
Apter, David E. (1955) TheGoldCoastin Transition.PrincetonUniversityPress.
_ (1961) The Political Kingdom in Uganda. Princeton University Press.
Arhin, Kwame. (1967) "The Financing of the Ashanti Expansion, 1700-1820," Africa
XXXVII, 3 (1967): 283-91.
Arrighi,G. (1967) ThePoliticalEconomyof Rhodesia.The Hague:Mouton.
. (1970) "Labour Supplies in Historical Perspective: A Study of the Proletarianization
of the AfricanPeasantryin Rhodesia," Journalof Development Studies6, 3: 197-234.
Artucio, Alejandro. (1979) The Trial of Macias in EquatorialGuinea: The Story of a
Dictatorship. Geneva: InternationalCommissionof Jurists.
Atieno-Odhiambo,E. S. (1981) "Seven Theses on Nationalismin Kenya," Nairobi Branch
of KenyaHistoricalAssociation,publiclecture:16 October.
Austin, Dennis. (1964) Politicsin Ghana,1946-1960.OxfordUniversityPress.
Awe, Bolanle. (1973) "Militarism and Economic Development in Nineteenth Century
YorubaCountry:The IbadanExample," Journalof AfricanHistoryXIV, 1: 65-78.
Azarya, Victor. (1980) "Tradersand the Center in Massina, Kong, and Samori's State,"
International Journalof AfricanHistoricalStudies13, 3: 420-56.
Balandier, Georges. (1951) "La Situation Coloniale: Approche Theorique," Cahiers
internationaux de sociologieXI: 44-79; reprinted,pp. 34-61 in Wallerstein(1966).
Ballard,Charles. (1980) "John Dunn and Cetshwayo:The MaterialFoundationsof Political
Power in the Zulu Kingdom, 1857-1878,"Journalof AfricanHistory21, 1: 75-91.
Barber, William J. (1961) The Economyof BritishCentralAfrica:A Case Studyof Economic
Development in a DualisticSociety.Oxford UniversityPress.
Barnes,J. A. (1962) "AfricanModels in the New Guinea Highlands,"Man LXII:5-9.
. (1969) "The Politics of Law," pp. 99-118 in Mary Douglas and Phyllis M. Kaberry
(eds.) Man in Africa.London:TavistockPress.
Bates, Darrell. (1972) A Gust of Plumes:A Biographyof Lord Twiningof Godalmingand
Tanganyika.London:Hodderand Stoughton.
Bates, R. H. (1971) Unions,Parties, and PoliticalDevelopment: A Study of Mineworkersin
Zambia.New Haven:Yale UniversityPress.
(1979) review of Palmerand Parsons (1977), AfricanEconomicHistory8: 238-42.
.
. (forthcoming)Essaysin RuralAfrica.CambridgeUniversityPress.
Bauer, P. T. (1954) WestAfricanTrade:A Studyof Competition, Oligopoly,and Monopolyin a
ChangingEconomy.CambridgeUniversityPress.
Bazin, Jean. (1975) "Guerre et servitude a S6gou," pp. 135-81 in Claude Meillassoux
(1975).
Beach, D. N. (1979) "'Chimurenga': The Shona Rising of 1896-97," Journalof African
History20, 3: 633-42.
Beattie, John. (1971) TheNyoroState.Oxford:ClarendonPress.
Beckman, BjOrn.(1976) Organizingthe Farmers:Cocoa Politicsand NationalDevelopmentin
Ghana.Uppsala:ScandinavianInstituteof AfricanStudies.
Beinart,William. (1979) "JoyiniInkomo:Cattle Advancesand the Originsof Migrancyfrom
Pondoland," Journalof SouthernAfricanStudies5, 2: 199-219.
Berman, Bruce J. (1973) "Administrationand Politics in Colonial Kenya." Yale University
Ph.D. thesis.
(1974) "Clientelism and Neocolonialism:Center-PeripheryRelations and Political
Development in AfricanStates," Studiesin Comparative International
Development 9, 2: 3-
25.
and J. M. Lonsdale. (1980) "Crisis of Accumulation,Coercion, and the Colonial
State: The Development of the Labor Control System in Kenya, 1919-1929," Canadian
Journalof AfricanStudies14, 1: 55-81.
Berman, Bruce J. (1980) "Articulation,Class Struggle, and the Specificityof the Colonial
State." Paper for 23rd Annual Meeting of the AfricanStudies Association, Philadelphia
(October).

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
210 AFRICAN STUDIESREVIEW

. (1981) "Class Struggle and the Origins of the Relative Autonomy of the Capitalist
State." Paperfor Annual Meeting, AmericanPoliticalScience Association.
Bernstein, Henry, and JacquesDepelchin. (1978, 1979) "The Objectof AfricanHistory: A
MaterialistPerspective," Historyin Africa5: 1-19; 6: 17-43.
Berque,Jacques. (1967) FrenchNorthAfrica:TheMaghribbetweenTwo WorldWars.London:
Faber (originallyParis:1962).
Bienen, Henry. (1967) Tanzania:Party Transformation and EconomicDevelopment.Princeton
UniversityPress.
. (1974) Kenya,thePoliticsof Participation
and Control.PrincetonUniversityPress.
Bonner, Philip. (1980) "Classes, the Mode of Production and the State in Pre-colonial
Swaziland,"pp. 80-101 in Marksand Atmore (1980).
. (forthcoming)Kings,Commoners, and Concessionnaires. CambridgeUniversityPress.
Boston, J. S. (1964) "The Hunterin IgalaLegendsof Origin," AfricaXXXIV, 2: 116-26.
Bozzoli, Belinda. (1981) The PoliticalNatureof a RulingClass: Capitaland Ideologyin South
Africa,1890-1933.London:RoutledgeKeganand Paul.
Bradbury,R. E. (1969) "Patrimonialismand Gerontocracyin Benin PoliticalCulture," pp.
17-36 in Mary Douglas and Phyllis M. Kaberry(eds.) Man in Africa.London:Tavistock
Press.
Braudel,Fernand. (1972, 1973) TheMediterranean and the Mediterranean Worldin the Age of
PhilipII. London:Collins, 2 vols.
. (1974) Capitalismand MaterialLife, 1400-1800.London:Fontana.
Brenner, Robert. (1976) "Agrarian Class Structure and Economic Development in Pre-
IndustrialEurope," Past & Present70: 30-75.
. (1977) "The Origins of Capitalist Development: A Critique of Neo-Smithian
Marxism," New LeftReview104: 25-92.
Brett, E. A. (1973) Colonialismand Underdevelopment in East Africa:ThePoliticsof Economic
change,1919-1939.London:Heinemann.
. (1975) "The PoliticalEconomy of General Amin," I.D.S. Bulletin(Brighton)7, 1:
15-22.
Brown, L. H. (1968) "AgriculturalChange in Kenya, 1945-1960," Food ResearchInstitute
(Stanford)Studiesin Agricultural
Economics,Trade,andDevelopment VIII:33-90.
Brown, Richard.(1969) "The external relationsof the Ndebele kingdomin the pre-partition
era," pp. 259-81 in LeonardThompson (ed.) AfricanStudiesin SouthernAfrica.London:
Heinemann.
Burke, Peter. (1980) Sociologyand History.London:Allen and Unwin.
Bundy, Colin. (1979) TheRise and Fall of theSouthAfricanPeasantry.London:Heinemann.
Cabral,Amilcar.(1969) Revolutionin Guinea.London:Stage 1.
Campbell,Bonnie. (1978) "The Ivory Coast," pp. 66-116 in Dunn (1978).
Cell, John W. (1980) "On the Eve of Decolonization:the Colonial Office's Plans for the
Transferof Power in Africa, 1947," Journalof Imperialand Commonwealth HistoryVIII, 3:
235-57.
Chabal, Patrick. (1981) "National Liberation in Portuguese Guinea, 1956-1974," African
Affairs80, 318: 75-100.
. (forthcoming)AmilcarCabral:Revolutionary Leadership andPeople'sWar.Cambridge
UniversityPress.
Chanock, Martin. (1972a) "The PoliticalEconomy of IndependentAgriculturein Colonial
Malawi:the Great Warto the Great Depression," Journalof SocialScience1: 1-17.
. (1972b) "Development and Change in the History of Malawi," pp. 429-41 in B.
Pachai(ed.) TheEarlyHistoryof Malawi.London:Longman.
Churchill,Winston. (1908) My AfricanJourney.London:Hodderand Stoughton.
Clarence-Smith,W. G. (1977) "For Braudel:A Note on the 'Ecole des Annales'and the
Historiographyof Africa," Historyin Africa4: 275-81.
Cobbing, Julian. (1974) "The evolution of the Ndebele Amabutho," Journalof African
HistoryXV, 4: 607-31.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIAL PROCESSES IN AFRICA 211

. (1977) "The Absent Priesthood:Another Look at the Rhodesian Risings of 1896-

1897," Journalof AfricanHistoryXVIII, 1: 61-84.
Cohen, Abner. (1971) "Cultural Strategies in the Organizationof Trading Diaspora," pp.
266-78 in Claude Meillassoux (ed.) The Developmentof IndigenousTradeand Marketsin
WestAfrica.Oxford UniversityPress.
- . (1981) The Politics of Elite Culture:Explorationsin the Dramaturgyof Power in a
ModernAfricanSociety.Berkeleyand Los Angeles:Universityof CaliforniaPress.
Cohen, G. A. (1978) KarlMarx'sTheoryof History:A defence.OxfordUniversityPress.
Cohen, Robin. (1974) Labourand Politicsin Nigeria,1945-1971.London:Heinemann.
. (1980) "Resistance and Hidden Forms of Consciousness amongst African
Workers," Reviewof AfricanPoliticalEconomy19: 8-22.
Cohen, Ronald. (1978) "State Origins:A Reappraisal,"pp. 31-70 in Henri J. M. Claessen
and P. Skalnik (eds.) TheEarlyState.The Hague:Mouton.
Coleman, James S. (1954) "Nationalism in Tropical Africa," AmericanPoliticalScience
ReviewXLVIII:404-26.
_. *(1963) Nigeria:Backgroundto Nationalism.Berkeleyand Los Angeles: Universityof
CaliforniaPress.
and Carl G. Rosberg, Jr. (eds.). (1964) PoliticalPartiesand NationalIntegrationin
TropicalAfrica.Berkeleyand Los Angeles: Universityof CaliforniaPress.
(1967) "The Resurrectionof PoliticalEconomy," Mawazo1 (June): 31-40.
Colonial Office. (1955) East AfricaRoyal Commission1953-1955Report:Cmd. 9475. London:
Her Majesty'sStationeryOffice.
Colson, Elizabeth. (1969) "AfricanSociety at the Time of the Scramble,"pp. 27-65 in L. H.
Gann and Peter Duignan (eds.) Colonialismin Africa,1870-1960,L.CambridgeUniversity
Press.
Cooper, Frederick. (1979) "The Problem of Slavery in AfricanStudies," Journalof African
History20, 1: 103-25.
. (1980) From Slaves to Squatters:PlantationLabor and Agriculturein Zanzibarand
CoastalKenya,1890-1925.New Haven: Yale UniversityPress.
. (1981) "Africa and the World Economy" AfricanStudiesReviewXXIV, 2/3: 000-
000.
Copans, Jean, and David Seddon. (1978) "Marxism and Anthropology:a preliminary
survey," pp. 1-46 in David Seddon (ed.) Relationsof Production:MarxistApproachesto
EconomicAnthropology. London:Cass.
Coquery-Vidrovitch,Catherine. (1969) "Recherches sur un mode de productionAfricain,"
La Pensbe144: 61-78; translated,pp. 77-92 in Peter C. W. Gutkind and Peter Waterman
(eds.) AfricanSocialStudies:a radicalreader.London:Heinemann (1977).
. (1977) "Mutation de l'imperialismecoloniale franCaisdans les ann6es 30," African
EconomicHistory4: 103-52.
Corrigan, Philip, Harvey Ramsay, and Derek Sayer. (1981) "The State as a Relation of
Production," pp. 1-25 in Philip Corrigan (ed.) Capitalism,State Formation,and Marxist
Theory.London:Quartet.
Cowen, Michael. (1972) "Differentiationin a KenyaLocation." East AfricanSocial Science
Conferencepaper,Nairobi.
. (1974) "Patterns of Cattle Ownershipand Dairy Production, 1900-1965."Nairobi
University:mimeo.
. (1975) "Wattle Production in the Central Province: Capital and Household
CommodityProduction,1903-1964."NairobiUniversity:mimeo.
(1976) "Capitaland PeasantHouseholds." NairobiUniversity:mimeo.
(1980) "The BritishState, State Enterprise,and an IndigenousBourgeoisiein Kenya
. 1945."
after Centre for DevelopmentStudies, Swansea:mimeo.
(1981) "CommodityProductionin Kenya'sCentralProvince," pp. 121-42in Judith
Heyer, Pepe Roberts, and Gavin Williams (eds.) Rural Developmentin TropicalAfrica.
London:Macmillan.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
212 AFRICAN STUDIES REVIEW

Creech Jones, A. (1947) Despatchfrom the Secretaryof Statefor the Coloniesto the Governors
of theAfricanColonies.ColonialOffice (25 February).
(1951) "British colonial policy, with particularreference to Africa," International
Affairs27, 2: 176-183.
Crummey, Donald. (1980) "AbyssinianFeudalism," Past andPresent89: 115-38.
and C. C. Stewart. (1981) "The Poverty of PrecolonialAfricanHistoriography,"pp.
11-34 in Crummeyand Stewart (eds.) Modesof Productionin Africa:ThePrecolonialEra.
Beverly Hills and London:Sage.
Curtin, Philip D. (1975) EconomicChangein PrecolonialAfrica:Senegambiain the Era of the
Slave Trade.Madison:Universityof WisconsinPress.
Steven Feierman, Leonard Thompson, and Jan Vansina. (1978) AfricanHistory.
London: Longman.
Curtis, Lionel. (1951) WithMilnerin SouthAfrica.Oxford:BasilBlackwell.
Daniel, J. B. (1973) "A Geography of pre-ShakanZululand," South AfricanGeographical
Journal55,1: 23-31.
Davidson, Basil. (1959) OldAfricaRediscovered. London:Gollancz.
S(1975)Can AfricaSurvive?London:Heinemann.
S
(1977) "Questionsabout Nationalism," AfricanAffairs302: 39-46.
Davies, Robert, David Kaplan,Mike Morris, and Dan O'Meara.(1976) "Class Struggleand
the Periodisationof the State in South Africa," Reviewof AfricanPoliticalEconomy7: 4-
30.
(1979) Capital, State, and White Labour in South Africa, 1900-1970. Brighton:
Harvester.
Decalo, Samuel. (1976) Coupsand ArmyRule in Africa.New Haven:Yale UniversityPress.
de Heusch, Luc. (1972) Le roiivre:ou l'originede I'btat.Paris:Gallimard.
Delius, Peter N. (1980) "The Pedi Polity under Sekwati and Sekhukhune, 1828-1880."
London UniversityPh.D. thesis.
Denoon, Donald. (1978) "The Dynamics of Dependent Development." Cambridge
UniversityCommonwealthand OverseasHistorySeminarpaper.
Deutsch, KarlW. (1961) "Social Mobilizationand PoliticalDevelopment," AmericanPolitical
ScienceReviewIv, 3: 492-514.
Dike, K. Onwuka. (1956) Tradeand Politicsin theNigerDelta, 1830-1885.Oxford:Clarendon
Press.
Dorward, D. C. (1974) "Ethnographyand Administration:A Study of Anglo-Tiv 'Working
Misunderstanding',"Journalof AfricanHistoryXV, 3: 457-77.
Dumett, Raymond E. (1979) "PrecolonialGold Mining and the State in the Akan Region:
with a Critiqueof the TerrayHypothesis," Researchin Economicanthropology 2: 37-68.
Dumont, Ren6. (1966) FalseStartin Africa.London:Deutsch. (OriginallyParis:1962.)
Dunn, John, and A. F. Robertson. (1973) Dependenceand Opportunity: PoliticalChangein
Ahafo.CambridgeUniversityPress.
(1978) "ComparingWest African States," pp. 1-21 in Dunn (ed.) WestAfrican
States:Failureand Promise.CambridgeUniversityPress.
(1982) "UnderstandingRevolutions," EthicsXCII, 2 (January).
Easton, David. (1957) "An Approachto the analysisof PoliticalSystems," WorldPoliticsIX,
3: 383-400.
Ehrensaft, Philip. (1977) "The Rise of a Proto-bourgeoisiein Yorubaland,"pp. 116-24 in
Peter C. W. Gutkind and Peter Waterman(eds.) AfricanSocial Studies:a radicalreader.
London:Heinemann.
Eley, Geoff. (1981) "Nationalismand Social History," SocialHistory6, 1: 83-107.
Eliot, T. S. (1938) Murderin theCathedral.London:Faber, 4th edition.
Etherington, Norman A. (1978) Preachers,Peasants, and Politicsin SoutheastAfrica, 1835-
1880. AfricanChristianCommunitiesin Natal, Pondoland,and Zululand.London: Royal
HistoricalSociety.
(1979) "Labour Supply and the Genesis of South African Confederationin the
1870s," Journalof AfricanHistory20, 2: 235-53.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 213

Fallers, L. A. (1955) "The Predicament of the Modern African Chief," American

57, 2: 290-305.
Anthropologist
. (1959) "Despotism, Status Culture, and Social Mobility in an African Kingdom,"
Comparative Studiesin Societyand HistoryII, 1: 11-32.
. (1964) TheKing'sMen:Leadershipand Statusin Bugandaon theEve of Independence.
Oxford UniversityPress.
Fanon, Frantz. (1965) The Wretched of the Earth.London: MacGibbonand Kee (originally
Paris:1961).
Feierman, Steven. (1974) The Shambaa Kingdom:A History. Madison: University of
WisconsinPress.
Fieldhouse, D. K. (1973) EconomicsandEmpire,1830-1914.London:Weidenfeld.
First, Ruth. (1970) TheBarrelof a Gun:PoliticalPowerin Africaand the Coupd'Etat.London:
Allen Lane.
Flint, Eric. (1970) "Trade and Politics in Barotselandduring the Kololo period," Journalof
AfricanHistoryXI, 1: 71-86.
Ford, John. (1971) TheRole of the Trypanosomiases in AfricanEcology:A Studyof the Tsetse-fly
problem.Oxford:ClarendonPress.
Fortes, M., and E. E. Evans-Pritchard.(1940) "Introduction," pp. 1-23 in Fortes and
Evans-Pritchard(eds.) AfricanPoliticalSystems.Oxford UniversityPress.
Foster-Carter,Aidan. (1978) "The Modes of Production Controversy," New Left Review
107:47-77.
Fox, Renee C., Willy de Craemer, and Jean-MarieRibeaucourt. (1965) "'The Second
Independence':A Case Study of the KwiluRebellion in the Congo," Comparative Studies
in SocietyandHistoryVIII, 1: 78-109.
Frederickson,George M. (1981) WhiteSupremacy.a comparative studyin Americanand South
AfricanHistory.Oxford UniversityPress.
Freund, W. M. (1981) "Class Conflict, PoliticalEconomy, and the Strugglefor Socialismin
Tanzania," AfricanAffairs80, 321: 483-99.
Fried, Morton H. (1967) TheEvolutionof PoliticalSociety.New York:RandomHouse.
Furedi, Frank. (1975) "The KikuyuSquattersin the Rift Valley, 1918-1929,"pp. 177-94 in
B. A. Ogot (ed.) Hadith5: Economicand SocialHistoryof EastAfrica.Nairobi:East Africa
LiteratureBureau.
Furnivall, J. S. (1948) Colonial Policy and Practice:A ComparativeStudy of Burma and
Netherlands India.CambridgeUniversityPress.
Furse, Sir Ralph. (1962) Aucuparius: Recollectionsof a RecruitingOfficer.Oxford University
Press.
Gallagher,John, and Ronald Robinson. (1953) "The Imperialismof Free Trade," Economic
HistoryReview2nd series, VI, 1: 1-15.
Gellner, Ernest. (1969) Saintsof theAtlas.London:Weidenfeld.
(1973) CauseandMeaningin theSocialSciences.London:Routledge.
Gertzel, Cherry. (1980) "Uganda after Amin: the Continuing Search for Leadershipand
Control," AfricanAffairs79, 317: 461-89.
Glentworth, Garth, and Ian Hancock. (1973) "Obote and Amin: change and Continuityin
Modern UgandaPolitics." AfricanAffairs72, 288: 237-55.
Gluckman,Max. (1955) Customand Corflictin Africa.Oxford:Blackwell.
Goody, Jack. (1963) "Feudalismin Africa?" Journalof AfricanHistoryIV, 1: 1-18.
(1971) Technology,Tradition, and theStatein Africa.OxfordUniversityPress.
S(1976)Productionand Reproduction: A Comparative Study of the DomesticDomain.
CambridgeUniversityPress.
(1977) TheDomestication of theSavageMind.CambridgeUniversityPress.
Gough, Ian. (1979) ThePoliticalEconomyof the WelfareState.London:Macmillan.
Gray, Richard,and David Birmingham.(1970) "Some Economicand PoliticalConsequences
of Trade in Centraland EasternAfricain the Pre-ColonialPeriod," pp. 1-23 in Gray and
Birmingham(eds.) Pre-ColonialAfricanTrade.OxfordUniversityPress.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
214 AFRICAN STUDIESREVIEW

Greenberg, Stanley B. (1980) Race and State in CapitalistDevelopment:Comparative

Perspectives. New Haven: Yale UniversityPress.
Gunther, John. (1953) InsideAfrica.New York:Harper.
Guy, Jeff. (1979) TheDestructionof the ZuluKingdom:TheCivil Warin Zululand,1879-1884.
London:Longman.
. (1980) "Ecologicalfactors in the rise of Shaka and the Zulu kingdom," pp. 102-19
•_ in Marksand Atmore (1980).
Guyer, Jane I. (1978) "The Food Economy and French Colonial Rule in Central
Cameroun," Journalof AfricanHistoryXIX, 4: 577-97.
. (1981) "Household and Community," AfricanStudiesReviewXXIV, 2/3: 000-000.
Hall, Martin. (1976-77) "Ethnography,Environment,and the History of the Nguni in the
Eighteenth and Nineteenth Centuries," pp. 11-20 in London University, Institute of
CommonwealthStudies, TheSocietiesof SouthernAfricain the 19thand 20th Centuries 8.
Harris, Laurence. (1980) "The State and the Economy:Some TheoreticalProblems," pp.
243-62 in Ralph Milibandand John Saville (eds.) The SocialistRegister1980. London:
MerlinPress.
Hartwig, Gerald W., and K. David Patterson (eds.). (1978) Disease in AfricanHistory.
Durham, North Carolina:Duke UniversityPress.
Hay, Margaret Jean. (1976) "Luo Women and Economic change during the Colonial
Period," pp. 87-109 in Nancy J. Hafkinand Edna G. Bay (eds.) Womenin Africa:Studies
in SocialandEconomicchange.StanfordUniversityPress.
Hedges, David W. (1978) "Tradeand Politicsin SouthernMozambiqueand Zululandin the
eighteenthand early nineteenthcenturies."London UniversityPh.D. thesis.
Hess, Robert L. (1970) Ethiopia,the Modernization of Autocracy.Ithaca:Cornell University
Press.
Hill, Polly. (1963) TheMigrantCocoa-Farmers of SouthernGhana:a studyin ruralcapitalism.
CambridgeUniversityPress.
_ . (1977) Population, Prosperity, and Poverty: Rural Kano, 1900 and 1970. Cambridge
UniversityPress.
Hindess, Barry, and Paul Q. Hirst. (1975) Pre-CapitalistModes of Production.London:
Routledge.
Hodgkin,Thomas. (1956) Nationalismin ColonialAfrica.London:Muller.
(1976) "Where the Paths Began," pp. 6-16 in C. Fyfe (ed.) AfricanStudiessince
1945:A Tributeto BasilDavidson.London:Longman.
Holloway,John, and Sol Picciotto (eds.) (1978) Stateand Capital:A MarxistDebate.London:
Arnold.
Hopkins, A. G. (1973) An EconomicHistoryof WestAfrica.London:Longman.
(1976) "ImperialBusiness in Africa," Journalof AfricanHistoryXVII, 1 and 2: 29-
48, 267-90.
(1980a) "Africa's Age of Improvement,"Historyin Africa7: 141-60.
(1980b) "Property Rights and Empire-building:Britain's annexation of Lagos,
1861," Journalof EconomicHistoryXL, 4: 777-98.
Horton, Robin. (1971) "Statelesssocieties in the historyof West Africa," pp. 78-119 in J. F.
A. Ajayiand MichaelCrowder(eds.) Historyof WestAfricaI. London:Longman.
Horwitz,Ralph. (1967) ThePoliticalEconomyof SouthAfrica.London:Weidenfeld.
Houghton, D. Hobart. (1964) The South AfricanEconomy.Cape Town: Oxford University
Press.
Huntington, Samuel P. (1968) Political Order in ChangingSocieties. New Haven: Yale
UniversityPress.
Hyden, Goran. (1968) Tanu YajengaNchi: PoliticalDevelopmentin Rural Tanzania.Lund:
ScandinavianUniversityBooks.
(1980) Beyond Ujamaain Tanzania:Underdevelopment and an Uncaptured Peasantry.
London:Heinemann.
Ilchman, WarrenF., and Norman Thomas Uphoff. (1971) ThePoliticalEconomyof Change.
Berkeleyand Los Angeles:Universityof CaliforniaPress.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIAL PROCESSES IN AFRICA 215

Iliffe, John. (1969) "The Age of Improvementand Differentiation,1907-1945,"pp. 123-60

in I. N. Kimambo and A. J. Temu (eds.) A Historyof Tanzania.Nairobi: East Africa
PublishingHouse.
. (1970) "A History of the Dockworkers of Dar es Salaam," Tanzania Notes and
Records 71: 119-48.
1 (1979) A Modern History of Tanganyika. Cambridge University Press.
Ingle, Clyde R. (1972) From Villageto State in Tanzania:ThePoliticsof RuralDevelopment.
Ithaca:Cornell UniversityPress.
Inikori, J. E. (1977) "The import of firearms into West Africa, 1750-1807," Journalof
AfricanHistoryXVIII, 3: 339-68.
Isaacman, Allen F. (1972) Mozambique:The Africanizationof a EuropeanInstitution-The
ZambeziPrazos,1750-1902.Madison:Universityof WisconsinPress.
. (1976) The Traditionof Resistance in Mozambique:Anti-Colonial Activity in the Zambezi
Valley.London:Heinemann.
Jamal, A. H. (1965) The CriticalPhase of EmergentAfricanStates. Nairobi: East African
PublishingHouse.
Janowitz,Morris. (1964) TheMilitaryin the PoliticalDevelopment of New Nations:An Essay in
Comparative analysis.Universityof ChicagoPress.
Jeeves, Alan. (1975) "The Control of MigratoryLabouron the South AfricanGold Mines
in the Era of Krugerand Milner," Journalof SouthernAfricanStudies2, 1: 3-29.
Jeffries, Richard. (1978a) Class, Power,and Ideologyin Ghana:The Railwaymenof Sekondi.
CambridgeUniversityPress.
(1978b) "PoliticalRadicalismin Africa:The Second Independence,"AfricanAffairs
308: 335-46.
Jessop, Bob. (1977) "Recent Theories of the Capitalist State," CambridgeJournal of
Economics1: 353-73.
Jewsiewicki, Bogumil. (1977) "The Great Depression and the making of the Colonial
Economicsystem in the BelgianCongo," AfricanEconomicHistory4: 153-76.
(1979) "Zaire Enters the World System: its colonial incorporationas the Belgian
Congo, 1885-1960," pp. 29-53 in G. Gran (ed.) Zaire: The Political Economy of
Underdevelopment. New York:Praeger.
Johnson, Marion. (1976) "The Economic Foundationsof an IslamicTheocracy:the case of
Massina," Journalof AfricanHistoryXVII, 4: 481-95.
(1980) "Polanyi, Peukert, and the political economy of Dahomey," Journalof
AfricanHistory21, 3: 395-98.
Johnstone, Frederick. (1976) Class, Race, and Gold: A studyof class relationsand racial
discriminationin SouthAfrica.London:Routledge.
Jones, G. I. (1963) The TradingStatesof the OilRivers.OxfordUniversityPress.
Kahn, J. S., and J. R. Llobera. (1980) "French MarxistAnthropology:twenty years after,"
Journal of Peasant Studies 8: 81-100.
Kanogo, Tabitha. (1980) "The PoliticalEconomyof KikuyuMovement into the Rift Valley,
1900-1963:the case study of NakuruDistrict."NairobiPh.D. thesis.
Kanya-Forstner,A. S. (1969) The Conquestof the WesternSudan. CambridgeUniversity
Press.
Kasfir, Nelson. (1976) The Shrinking Political Arena: Participation and Ethnicity in African
Politics,witha Case Studyof Uganda.Berkeleyand Los Angeles: Universityof California
Press.
Katz, Stephen. (1980) Marxism,Africa, and Social Class: A Critiqueof Relevant Theories.
Montreal:McGillUniversity,Center for Developing-AreaStudies.
Kay, G. B. (1972) The PoliticalEconomyof Colonialismin Ghana. CambridgeUniversity
Press.
Kea, R. A. (1973) "Trade, State Formation, and Warfareon the Gold Cost." London
UniversityPh.D. thesis.
Kilson, Martin. (1963) "AfricanPoliticalchange and the ModernisationProcess," Journalof
Modern African Studies 1, 4: 425-40.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 216 AFRICAN STUDIESREVIEW

. (1966) Political Change in a West African State: a study of the modernisationprocess in

SierraLeone. HarvardUniversityPress.
. (1970) "The emergent elites of black Africa, 1900-1960," pp. 351-98 in L. H. Gann
and Peter Duignan (eds.) Colonialismin AfricaII. CambridgeUniversityPress.
Kipkorir,B. E. (1969) "The Alliance High School and the Originsof the Kenya African
Elite, 1926-1962."CambridgeUniversityPh.D. thesis.
. (1974) "The Inheritorsand Successors:The TraditionalBackgroundto the Modern
KenyanAfricanElite, c. 1890-1930," KenyaHistoricalReview2, 2: 143-61.
Kirk-Greene, A. H. M. (ed.). (1979) Africa in the ColonialPeriod, III: The Transferof
Power- The ColonialAdministrator in the Age of Decolonisation.Oxford University:Inter-
FacultyCommitteefor AfricanStudies.
. (1980a) "The Thin White Line: The Size of the ColonialService in Africa," African
Affairs79, 314: 25-44.
. (1980b) "'Damnosa Hereditas': Ethnic Ranking and the Martial Races Imperative
in Africa," Ethnicand RacialStudies3, 4: 393-414.
Kitching,Gavin. (1980) Class and EconomicChangein Kenya:the makingof an AfricanPetite-
Bourgeoisie,1905-1970. New Haven:Yale UniversityPress.
Kiwanuka,Semakula.(1971) A Historyof Buganda. London:Longman.
Kjekshus,Helge. (1977) "The TanzanianVillagizationPolicy:ImplementationalLessons and
EcologicalDimensions," CanadianJournalof AfricanStudies11, 2: 269-82.
Klein, MartinA. (1968) Islamand Imperialism in Senegal:Sine-Saloum,1847-1914.Edinburgh
UniversityPress.
(1972) "Social and Economic Factors in the Muslim Revolution in Senegambia,"
Journalof AfricanHistoryXIII, 3: 419-41.
Kuper, Leo, and M. G. Smith (eds.). (1969) Pluralismin Africa.Berkeleyand Los Angeles:
Universityof CaliforniaPress.
(1971) "African Nationalism in South Africa, 1910-1964," pp. 424-76 in Monica
Wilson and Leonard Thompson (eds.) The OxfordHistoryof South AfricaII. Oxford:
ClarendonPress.
(1981) Genocide:itspoliticaluse in the twentieth
century.Harmondsworth:Penguin.
Lamb, Geoff. (1975) "Marxism, Access, and the State," Development and change6, 2: 119-
35.
Langdon, Steven. (1977) "The State and Capitalismin Kenya," Reviewof AfricanPolitical
Economy8: 90-98.
(1978) "The MultinationalCorporationin the Kenya PoliticalEconomy," pp. 134-
200. in Raphael Kaplinsky (ed.) Readings on the MultinationalCorporationin Kenya.
Nairobi:OxfordUniversityPress.
Last, Murray.(1967) TheSokotoCaliphate.London:Longman.
Latham,A. J. H. (1973) Old Calabar,1600-1891:theImpactof theInternational Economyupon
a Traditional Society.Oxford:ClarendonPress.
Law, Robin. (1977a) The Oyo Empirec.1600-c.1836: A WestAfricanImperialism in the Era of
theAtlanticSlave Trade.Oxford:ClarendonPress.
. (1977b) "Royal Monopoly and Private Enterprisein the Atlantic Slave Trade:the
Case of Dahomey," Journalof AfricanHistoryXVIII,4: 555-77.
(1978a) "Slaves, Trade, and Taxes: the Material Base of Political Power in
PrecolonialWest Africa," Researchin EconomicAnthropology 1: 37-52.
. (1978b) "In searchof a Marxistperspectiveon pre-colonialtropicalAfrica," Journal
of AfricanHistoryXIX, 3: 441-52.
(1980) TheHorsein WestAfricanHistory.Oxford UniversityPress.
(1981) "How not to be a Marxist historian:the Althusserianthreat to African
history," pp. 313-19 in Raphael Samuel (ed.) People's History and Socialist Theory.
London:Routledge.
Lee, J. M. (1967) ColonialDevelopment and GoodGovernment. Oxford:ClarendonPress.
Legassick, Martin. (1974) "South Africa:capitalaccumulationand violence," Economyand
SocietyIII, 3: 253-91.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 217

. (1977) "The concept of pluralism: a critique," pp. 44-50 in P. C. W. Gutkind and P.

Waterman(eds.) AfricanSocialStudies.,a radicalreader.London:Heinemann.
. (1980) "The frontiertraditionin South Africanhistoriography,"pp. 44-79 in Marks
and Atmore (1980).
Legum, Colin. (1954) Must we lose Africa?London:Allen.
Lemarchand,Ren6. (1973) "African Power through the Looking-Glass,"Journalof Modern
AfricanStudies11, 2: 305-14.
Levine, Donald N. (1965) Wax and Gold: Traditionand Innovationin EthiopianCulture.
ChicagoUniversityPress.
(1974) Greater Ethiopia: The Evolution of a Multiethnic Society. Chicago University
Press.
Levtzion, Nehemia. (1971) "Patternsof Islamizationin West Africa," pp. 31-39 in D. F.
McCalland N. R. Bennett (eds.) Aspectsof WestAfricanIslam. Boston UniversityPapers
on Africa5.
Lewis, I. M. (1968) "Introduction," pp. ix-xxxviii in Lewis (ed.) History and Social
Anthropology. London:TavistockPress.
Leys, Colin. (1959) EuropeanPoliticsin SouthernRhodesia.Oxford UniversityPress.
(1965) "What is the Problemabout Corruption?"Journalof ModernAfricanStudies
3, 2: 215-30.
. (1967) Politiciansand Policies:an essay on politicsin Acholi,Uganda,1962-65.Nairobi:
East AfricanPublishingHouse.
. (1971) "Politics in Kenya: The Development of Peasant Society," British Journal of
PoliticalScienceI, 3: 307-37.
(1975) Underdevelopment in Kenya:ThePoliticalEconomyof Neo-Colonialism.
London:
Heinemann.
(1976) "The 'Overdeveloped' Post-Colonial State: a Re-evaluation," Review of
AfricanPoliticalEconomy5: 39-48.
. (1978) "Capital Accumulation, Class Formation, and Dependency-the Significance
of the KenyanCase," pp. 241-66 in R. Milibandand J. Saville (eds.) TheSocialistRegister
1978. London:MerlinPress.
Leys, Norman. (1924) Kenya.London:HogarthPress.
Liebenow, J. Gus. (1971) ColonialRule and PoliticalDevelopment in Tanzania:the case of the
Makonde.Evanston:NorthwesternUniversityPress.
Lloyd, Peter C. (1965) "The Political Structure of African Kingdoms: An Exploratory
Model," pp. 63-109 in M. Banton (ed.) PoliticalSystemsand the Distributionof Power.
London:TavistockPress.
(ed.). (1966) TheNew Elitesof TropicalAfrica.OxfordUniversityPress.
S(1968a)"The PoliticalDevelopment of West AfricanKingdoms," Journalof African
HistoryIX, 2: 319-29.
(1968b) "ConflictTheoryand YorubaKingdoms,"pp. 25-58 in Lewis.
S(1971)ThePoliticalDevelopment of YorubaKingdomsin the Eighteenthand Nineteenth
Centuries.London:Royal AnthropologicalInstitute.
Lonsdale, John. (1977) "The Politics of Conquest: The British in Western Kenya, 1894-
1908," HistoricalJournal20, 4: 841-70.
with Stanley Booth-Clibbornand Andrew Hake. (1978) "The emerging pattern of
Church and State co-operationin Kenya," pp. 267-84 in EdwardFashole-Luke,Richard
Gray, Adrian Hastings, and Godwin Tasie (eds.) Christianityin IndependentAfrica.
London:Rex Collings.
and Bruce Berman. (1979) "Coping with the Contradictions:the developmentof the
ColonialState in Kenya, 1895-1914,"Journalof AfricanHistory20, 4: 487-505.
(1981) "State and peasantryin colonial Africa," pp. 106-17in RaphaelSamuel (ed.)
People'sHistoryandSocialistTheory.London:Routledge.
Louis, Wm. Roger. (1977) Imperialism at Bay: the UnitedStatesand the Decolonization
of the
BritishEmpire,1941-1945.Oxford:ClarendonPress.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
218 AFRICAN STUDIES REVIEW

Lovejoy, Paul E. (1978) "Plantationsin the Economy of the Sokoto Caliphate,"Journalof

AfricanHistoryXIX, 3: 341-68.
Low, D. A. (1964) "The Advent of Populismin Buganda," Comparative Studiesin Societyand
HistoryVI, 4: reprintedin Low (1971).
. (1971) Buganda in Modern History. London: Weidenfeld.
. (1973) Lion Rampant. London: Cass.
. (1975) "Warbands and Ground-level Imperialism in Uganda, 1870-1900," Historical
Studies16, 65: 584-597.
and J. M. Lonsdale. (1976) "Introduction: Towards the New Order, 1945-1963," pp.
1-64 in D. A. Low and Alison Smith (eds.) Historyof East AfricaIII. Oxford:Clarendon
Press.
Luckham,Robin. (1971) TheNigerianMilitary:A SociologicalAnalysisof Authority andRevolt,
1960-1967.CambridgeUniversityPress.
Lukes, Steven. (1974) Power:A RadicalView.London:Macmillan.
MacGaffey, Wyatt. (1970) Customand Government in the Lower Congo. Berkeley and Los
Angeles: Universityof CaliforniaPress.
Mafeje, Archie. (1971) "The Ideologyof 'Tribalism',"Journalof ModernAfricanStudies9, 2:
253-61.
Maier, D. J. E. (1981) "The Dente Oracle, the Bron Confederation,and Asante: Religion
and the Politicsof Secession," Journalof AfricanHistory22, 2: 229-43.
Mamdani,Mahmood. (1976) Politicsand ClassFormationin Uganda.London:Heinemann.
Markovitz,I. Leonard. (1977) Powerand Class in Africa.EnglewoodCliffs:Prentice-Hall.
Marks, Shula, and Stanley Trapido. (1979) "Lord Milner and the South African State,"
HistoryWorkshop 8: 50-80.
and Anthony Atmore (eds.). (1980) Economyand Society in Pre-Industrial South
Africa.London:Longman.
Marx, Karl. (1954) TheEighteenthBrumaireof LouisBonaparte.Moscow:ProgressPublishers.
Mazrui, Ali. (1967) "Tanzaphilia,"Transition 31: 20-26.
Mbee, Gicha. (1965) "Letter from Mbugwe, Tanganyika,"edited and translatedby Robert
F. Gray, Africa35, 2: 198-209.
McCaskie, T. C. (1980) "Office, land, and subjects in the history of Manwere fekuo of
Kumase:an essay in the politicaleconomy of the Asante state," Journalof AfricanHistory
21, 2: 189-208.
McCracken, John. (1977) Politics and Christianityin Malawi, 1875-1940. Cambridge
UniversityPress.
Meillassoux, Claude. (1964) Anthropologieeconomiquedes Gouro de Cite d'Ivoire. Paris:
Mouton.
. (1970) "A Class analysis of the BureaucraticProcess in Mali," Journal of
Development Studies6, 2: 97-110.
(1972) "From reproduction to production: a Marxist approach to economic
anthropology,"EconomyandSocietyI, 1: 93-105.
(.ed.). (1975) L'esclavageen Afriqueprbcoloniale.
Paris:Maspero.
Middlemass,Keith. (1979) Politicsin IndustrialSociety:Theexperienceof theBritishsystemsince
1911. London:Deutsch.
Miers, Suzanne, and Igor Kopytoff. (eds.). (1977) Slavery in Africa: Historical and
Anthropological Perspectives.Madison:Universityof WisconsinPress.
Milburn,JosephineF. (1977) BritishBusinessand GhanaianIndependence. London:Hurst.
Miles, John. (1978) "Rural Protest in the Gold Coast:The Cocoa Hold-ups, 1908-1938,"pp.
152-70 in Clive Dewey and A. G. Hopkins (eds.) The ImperialImpact:Studiesin the
EconomicHistoryof AfricaandIndia.London:Athlone Press.
Miller, Joseph C. (1976) Kings and Kinsmen:Early MbunduStates in Angola. Oxford
UniversityPress.
. (1977) Review of John M. Jantzen and Wyatt MacGaffey (eds.) An Anthologyof
Kongo Religion(Kansas University Press: 1974), in Journalof AfricanHistoryXVIII, 1:
139-40.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 219

Mitchell, Sir Philip. (1955) "Africa and the West in HistoricalPerspective,"pp. 3-24 in C.
Grove Haines (ed.) AfricaToday.Baltimore:Johns Hopkins.
Moodie, T. Dunbar. (1975) TheRise of Afrikanerdom: Power,Apartheid,and theAfrikanerCivil
Religion.Berkeleyand Los Angeles: Universityof CaliforniaPress.
Moore, Barrington. (1978) Injustice:The Social Bases of Obedienceand Revolt. Stamford,
Conn.: Freimann (and London, Macmillan:1979).
Moore, Clement Henry. (1965) Tunisia since Independence:The Dynamics of One-Party
Government. Berkeleyand Los Angeles: Universityof CaliforniaPress.
(1970) Politicsin NorthAfrica,Algeria,Morocco,and Tunisia.Boston:Little, Brown.
Morris, M. L. (1976) "The development of capitalismin South African agriculture:class
struggle in the countryside," Economyand Society5; reprinted, pp. 202-253 in Wolpe
(1980).
Morrow, Lance F. (1971-72) "Pre-Mfecanecarry-oversin subsequentNdebele politics," pp.
85-95 in London University,Instituteof CommonwealthStudies, TheSocietiesof Southern
Africain the 19thand 20th Centuries3.
Morton-Williams,Peter. (1968) "The Fulani Penetration into Nupe and Yoruba in the
Nineteenth Century," pp. 1-24 in Lewis (1968).
Moseley, K. P. (1979) "The political economy of Dahomey," Research in Economic
Anthropology 2: 69-90.
Mosely, Paul. (1980) "The Settler Economies:studies in the economic historyof Kenyaand
SouthernRhodesia, 1900-1963."CambridgeUniversityPh.D. thesis.
Mueller, Susanne D. (1980) "Retarded Capitalism in Tanzania," pp. 203-26 in Ralph
Milibandand John Saville (eds.) TheSocialistRegister,1980. London:MerlinPress.
Murdock, George Peter. (1959) Africa, its Peoples and their CultureHistory.New York:
McGraw-Hill.
Muriuki, Godfrey. (1974) A Historyof the Kikuyu,1500-1900. Nairobi: Oxford University
Press.
Murray,Robin. (1971) "The Internationalization of Capitaland the Nation State," New Left
Review67: 84-109.
Murray,Roger. (1967) "Second Thoughtson Ghana," New LeftReview42: 25-39.
Neale, Caroline. (1980) "Pride and Prejudicein African History:The Idea of Progress in
AfricanHistory, 1960-1970."Sussex UniversityD.Phil thesis.
Newbury, Colin. (1969) "Tradeand Authorityin West Africafrom 1850 to 1880," pp. 66-99
in L. H. Gann and Peter Duignan (eds.) Colonialismin Africa, 1870-1960I. Cambridge
UniversityPress.
Njonjo, Apollo L. (1977) "The Africanizationof the 'White Highlands':A Studyin Agrarian
Class Strugglesin Kenya, 1950-1974."PrincetonUniversityPh.D. thesis.
Northrup, David. (1978) TradewithoutRulers:Pre-ColonialEconomicDevelopmentin South-
EasternNigeria.Oxford:ClarendonPress.
O'Brien, Donal C. (1971) TheMouridesof Senegal:ThePoliticaland EconomicOrganization of
an IslamicBrotherhood. Oxford:ClarendonPress.
(1972) "Modernization,Order, and the Erosion of a DemocraticIdeal:American
PoliticalScience, 1960-70," Journalof Development Studies8: 351-78.
O'Connell, James. (1967) "The Inevitabilityof Instability,"Journalof ModernAfricanStudies
5, 2: 181-91.
Ogot, B. A. (1964) "Kingship and statelessness among the Nilotes," pp. 284-304 in J.
Vansina, R. Mauny, and L. V. Thomas (eds.) The Historianin TropicalAfrica.Oxford
UniversityPress.
(1967) Historyof the SouthernLuo, I: migrationand settlement.Nairobi:East African
PublishingHouse.
Oliver, Roland. (1952) TheMissionaryFactorin East Africa.London:Longman.
and J. D. Fage. (1962) A ShortHistoryof Africa.Harmondsworth:Penguin.
(1963) "Epilogue," pp. 454-56 in Roland Oliver and Gervase Mathew (eds.) History
of East AfricaI. Oxford:ClarendonPress.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
220 AFRICAN STUDIESREVIEW

. (1977) "The East African Interior," pp. 621-69 in Oliver (ed.) The Cambridge
Historyof Africa3:from c. 1050 to c. 1600. CambridgeuniversityPress.
Oloruntimehin,B. 0. (1972) TheSegu TukulorEmpire.London:Longman.
O'Meara,Dan. (1977) "The AfrikanerBroderbond,1927-1948:class vanguardof Afrikaner
nationalism,"Journalof SouthernAfricanStudies3, 2: 156-86.
. (1978) "AnalysingAfrikanerNationalism:The 'Christian-National'assaulton white
tradeunionism in South Africa, 1934-1948,"AfricanAffairs306: 45-72.
Omer-Cooper,J. D. (1966) The Zulu Aftermath:A Nineteenth-Century Revolutionin Bantu
Africa.London:Longman.
Owen, David E. (1955) Commentaryon Mitchell (1955), pp. 25-29 in C. Grove Haines (ed.)
AfricaToday.Baltimore:Johns Hopkins.
Palmer, Robin. (1972) "Johnston and Jameson: a comparativestudy of the imposition of
colonial rule," pp. 293-322 in B. Pachai (ed.) The Early Historyof Malawi. London:
Longman.
and Neil Parsons (eds.). (1977) The Roots of RuralPovertyin Centraland Southern
Africa.London:Heinemann.
Parkin, Frank. (1979) Marxismand Class Theory:A BourgeoisCritique.London: Tavistock
Press.
Partridge,P. H. (1963) "Some Notes on the Conceptof Power," PoliticalStudiesXI, 2: 107-
25.
Perham, Margery.(1956, 1960) Lugard,I: The Yearsof Adventure,1858-1898;II: the Yearsof
Authority,1898-1945.London:Collins.
Phillips,Anne. (1977) "The Conceptof 'Development'," Reviewof AfricanPoliticalEconomy
8: 7-20.
Plumb, J. H. (1969) TheGrowthof PoliticalStabilityin England,1675-1725.Harmondsworth:
Penguin.
Polanyi, Karl. (1966) Dahomeyand the Slave Trade:An Analysisof an ArchaicEconomy.
WashingtonUniversityPress.
Post, Ken. (1972) "'Peasantization' and Rural Political Movements in Western Africa,"
Archiveseuropeennes de SociologieXIII, 2: 223-54.
Poulantzas, Nicos. (1973) PoliticalPower and Social Classes. London: New Left Books
(originallyParis:1968).
. (1978) State,Power,Socialism.London:New Left Books.
Prins, Gwyn. (1980) TheHiddenHippopotamus. CambridgeUniversityPress.
Pratt, Cranford.(1976) The CriticalPhase in Tanzania,1945-1968:Nyerereand the emergence
of a socialiststrategy.CambridgeUniversityPress.
Quandt,WilliamB. (1969) RevolutionandPoliticalLeadership: Algeria,1954-1968.Cambridge,
Mass.:M.I.T.Press.
Raikes, P. L. (1975) "Ujamaaand Rural Socialism," Reviewof AfricanPoliticalEconomy3:
33-52.
Ranger,T. 0. (1967) Revoltin SouthernRhodesia,1896-7. London:Heinemann.
(1968) "Introduction," pp. ix-xxii in Ranger (ed.) EmergingThemesof African
History.Nairobi:East AfricaPublishingHouse.
(1976) "Towards a Usable African Past," pp. 17-30 in ChristopherFyfe (ed.)
AfricanStudiessince 1945:a tributeto Basil Davidson.London:Longman.
(1980) "MakingNorthernRhodesiaImperial-variationson a royal theme," African
Affairs316: 349-73.
. (1982) "GuerillaWar and PeasantViolence: Makoni District, Zimbabwe."Institute
of CommonwealthStudies, London, seminarpaper (February).
Rathbone, Richard. (1973) "Businessmen in Politics: Party Struggle in Ghana, 1949-57,"
Journalof Development Studies9, 3: 391-401.
Rennie, J. K. (1978) "White Farmers, Black Tenants, and LandlordLegislation:Southern
Rhodesia, 1890-1930,"Journalof SouthernAfricanStudies5, 1: 86-98.
Rey, Pierre-Philippe.(1971) Colonialisme,neo-colonialisme, et transitionau capitalisme.Paris:
Maspero.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 221

(1973) Les alliancesdes classes.Paris:Maspero.

Richards,Audrey (ed.). (n.d.) EconomicDevelopment and TribalChange:A Studyof Immigrant
Labourin Buganda.Cambridge:Heffer.
Riesman, David. (1958) "Introduction,"pp. 1-15 in Daniel Lerner, ThePassingof Traditional
Society:Modernizing theMiddleEast. New York:Free Press.
Roberts, Andrew D. (1974) A Historyof theBemba.London:Longman.
Roberts, Richard (1980a) "Long Distance Trade and Production:Sinsani in the Nineteenth
Century," Journalof AfricanHistory21, 2: 169-88.
. (1980b) "Productionand Reproductionin WarriorStates:Segu Bambaraand Segu
Tokolorc. 1712-1890," International Journalof AfricanHistoricalStudies13, 3: 389-419.
and Martin Klein. (1980) "The Banamba slave exodus of 1905 and the decline of
slavery in the WesternSudan," Journalof AfricanHistory21, 3: 375-94.
. (1981) "Fishing for the State: The PoliticalEconomy of the Middle Niger Valley,"
pp. 175-203in Crummeyand Stewart(1981).
Roberts, Simon. (1979) Order and Dispute: An Introductionto Legal Anthropology.
Harmondsworth:Penguin.
Robertson, A. F. (1978) Community of Strangers:A Journalof Discoveryin Uganda.London:
ScolarPress.
Robinson, Ronald, and John Gallagher,with Alice Denny. (1961) Africaand the Victorians.
London:Macmillan.
(1972) "Non-Europeanfoundationsof Europeanimperialism:sketch for a theory of
collaboration,"pp. 117-40 in Roger Owen and Bob Sutcliffe (eds.) Studiesin the Theoryof
Imperialism. London:Longman.
. (1974) "The Journal and the Transfer of Power," Journal of AdministrationOverseas
XIII, 1: 255-58.
. (1977) "European Imperialism and Indigenous Reactions in British West Africa,
1880-1914," pp. 141-63 in H. L. Wesseling (ed.) Expansionand Reaction:Essays on
EuropeanExpansionandReactionsin Asia andAfrica.Leiden UniversityPress.
. (1979) "The Moral Disarmament of African Empire, 1919-1947," Journal of Imperial
and Commonwealth HistoryVIII, 1: 86-103.
. (1980) "Andrew Cohen and the Transferof Power in TropicalAfrica, 1940-1951,"
pp. 50-72 in W. H. Morris-Jonesand G. Fischer (eds.) Decolonisation andAfter:TheBritish
and FrenchExperience.London:Cass.
Rodney, Walter. (1972) How Europe Underdeveloped Africa. London and Dar es Salaam:
Bogle l'Ouvertureand TanzaniaPublishingHouse.
Ruel, Malcolm. (1969) Leopardsand Leaders:ConstitutionalPoliticsamong a Cross River
People.London:TavistockPress.
Rusch, Walter. (1975) KlassenundStaat in Bugandavorderkolonialzeit.Berlin:Verdffnungen
des Museums fur V61kerkundezu Leipzig.
Samoff, Joel. (1979) "The Bureaucracyand the Bourgeoisie:Decentralizationand Class
Structurein Tanzania," Comparative Studiesin SocietyandHistory21, 1: 30-62.
Sandbrook,Richard.(1975) Proletarians and AfricanCapitalism:TheKenyanCase, 1960-1972.
CambridgeUniversityPress.
Saul, John. (1974a) "The State in Post-ColonialSocieties: Tanzania," pp. 349-72 in R.
Milibandand J. Saville (eds.) SocialistRegister1974. London:MerlinPress.
(1974b) "African Peasantsand Revolution," Reviewof AfricanPoliticalEconomy1:
41-68.
SchOlch, A. (1974) "Constitutional Development in Nineteenth-Century Egypt: a
Reconsideration,"MiddleEasternStudies10, 1: 3-14.
Schumpeter,Joseph. (1955) Imperialism and Social Classes:TwoEssays.New York:Meridian
Books.
Seal, Anil. (1973) "Imperialismand Nationalism in India," pp. 1-27 in John Gallagher,
Gordon Johnson, and A. Seal (eds.) Locality,Province,and Nation: essays on Indian
Politics,1870 to 1940. CambridgeUniversityPress.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
222 AFRICANSTUDIESREVIEW

Seddon, David. (1981) MoroccanPeasants.London:Dawson.

Sender, John. (1982) "Tanzania: Alternative Views on Development." Cambridge
University,Centre of AfricanStudies, seminarpaper (February).
Service, Elman R. (1975) Originsof theStateand Civilization.
New York:Norton.
Shepperson,George, and Thomas Price. (1958) Independent African:John Chilembweand the
Origins, Setting, and Significanceof the NyasalandNative Rising of 1915. Edinburgh
UniversityPress.
Shivji, Issa G. (1973) TheSilentClassStruggle.Dar es Salaam:TanzaniaPublishingHouse.
. (1976) ClassStrugglesin Tanzania.London:Heinemann.
Sklar, RichardL. (1963) NigerianPoliticalParties.PrincetonUniversityPress.
. (1965) "Contradictions in the Nigerian Political System," Journal of Modern African
Studies3, 2: 201-13.
. (1967) "Political Science and National Integration," Journal of Modern African Studies
5, 1: 1-11.
_. (1976) "Postimperialism: a class analysis of multinational corporate expansion,"
PoliticsIX, 1: 75-92.
Comparative
. (1979) "The Nature of Class Domination in Africa," Journal of Modern African
Studies17, 4: 531-52.
. (1981) "Response to John Lonsdale's 'The State and Social Processes in Africa'."
Paperfor 24th Annual Meeting, AfricanStudies Association.
Skocpol, Theda. (1979) StatesandSocialRevolutions.CambridgeUniversityPress.
Slade, Ruth. (1962) KingLeopold'sCongo.OxfordUniversityPress.
Smaldone,Joseph P. (1977) Warfarein theSokotoCaliphate.CambridgeUniversityPress.
Smith, Alan. (1969) "The tradeof Delagoa Bayas a factorin Nguni politics, 1750-1835,"pp.
171-89 in Leonard Thompson (ed.) African,Societies in SouthernAfrica. London:
Heinemann.
Smith, Anthony D. (1971) Theoriesof Nationalism.London:Duckworth.
Smith, H. F. C. (1961) "A neglectedtheme of West Africanhistory:the Islamicrevolutions
of the nineteenthcentury," Journalof theHistoricalSocietyof NigeriaII, 1: 169-85.
Smith, L. D. (1976) "An Overview of AgriculturalDevelopment Policy," pp. 111-51 in J.
Heyer, J. K. Maitha, and W. M. Senga (eds.) Agricultural Developmentin Kenya: An
EconomicAssessment.Nairobi:Oxford UniversityPress.
Smith, Tony. (1978) "A Comparative Study of French and British Decolonization,"
Comparative Studiesin Societyand History20, 1: 70-102.
Sobania, Neal W. (1980) "The Historical Tradition of the Peoples of the Eastern Lake
TurkanaBasin, c. 1840-1925."London UniversityPh.D. thesis.
Southall, Aidan. (1956) AlurSociety.Cambridge:Heffer.
(1970) "The Illusionof Tribe," Journalof AsianandAfricanStudies5, 1-2: 28-50.
Southwold, Martin. (1961) Bureaucracyand Chiefshipin Buganda. Kampala:East African
Instituteof Social Research,East AfricanStudies No. 14.
(1968) "The History of a History: Royal Succession in Buganda," pp. 127-52 in
Lewis (1968).
-•.
Soyinka,Wole. (1967) "The Writerin an AfricanState," Transition 31: 11-13.
Spencer, Paul. (1973) Nomads in Alliance:symbiosisand growth among the Rendille and
Samburuof Kenya.OxfordUniversityPress.
Steinhart,EdwardI. (1978) "Ankole: PastoralHegemony," pp. 131-50 in H. J. M. Claessen
and P. Skalnik (eds.) TheEarlyState.The Hague:Mouton.
(1981) "Herders and Farmers: the Tributary Mode of Production in Western
Uganda," pp. 115-55in Crummeyand Stewart(1981).
Stemler, A. B. L., J. R. Harlan, and J. M. J. DeWet. (1975) "CaudatumSorghums and
Speakersof Chari-Nilelanguagesin Africa," Journalof AfricanHistoryXV, 2: 161-83.
Stevenson, Robert F. (1968) Populationand PoliticalSystemsin TropicalAfrica.New York:
ColumbiaUniversityPress.
Stokes, Eric. (1970) "TraditionalResistance Movements and Afro-Asian Nationalism:The
Context of the 1857 MutinyRebellionin India," Past and Present48:100-18.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATES AND SOCIALPROCESSESIN AFRICA 223

. (1978) The Peasant and the Raj: studies in agrarian society and peasant rebellion in
colonialIndia.CambridgeUniversityPress.
Stone, Lawrence. (1979) "The Revival of Narrative:Reflections on a New Old History,"
Past and Present85: 3-24.
Tamarkin,M. (1978) "The Roots of PoliticalStabilityin Kenya," AfricanAffairs308: 297-
320.
Terray, Emmanuel. (1972) Marxismand 'Primitive'Societies.New York: Monthly Review
Press (originallyParis, 1969).
. (1975a) "La captivite dans le royaume abron du Gyaman," pp. 389-453 in Claude
Paris:Maspero.
Meillassoux(ed.) L'esclavageen Afriqueprbcoloniale.
. (1975b) "Classes and Class Consciousness in the Abron Kingdom of Gyaman," pp.
London:Malaby
85-135 in MauriceBloch (ed.) MarxistAnalysesand SocialAnthropology.
Press.
Therborn,G6ran. (1971) "Jiirgen Habermas:a New Eclecticism," New Left Review67: 69-
83.
(1980) The Ideology of Power and the Power of Ideology. London: New Left Books.
.
Thompson, E. P. (1965) TheMakingof theEnglishWorkingClass.London:Gollancz.
. (1977) Whigs and Hunters: The Origin of the Black Act. Harmondsworth: Penguin
Books.
. (1978) The Poverty of Theory. London: Merlin Press.
Thompson, Leonard. (1975) Survivalin Two Worlds:Moeshoeshoeof Lesotho, 1786-1870.
Oxford:ClarendonPress.
Thornton, John. (1977) "Demographyand History in the Kingdom of the Kongo, 1550-
1750," Journalof AfricanHistoryXVIII,4: 507-30.
Tignor, Robert L. (1971) "ColonialChiefs in ChieflessSocieties," Journalof ModernAfrican
Studies9, 3: 339-59.
Tordoff, William.(1965) AshantiunderthePrempehs,1885-1935.Oxford UniversityPress.
. (1973) "Colonial Chiefs in a Chiefless Society: a case-study from Northern
Uganda," Journalof AfricanHistoryXIV, 3: 473-90.
(1978) ClanLeadersand ColonialChiefsin Lango.Oxford:ClarendonPress.
Tosh, John. (1980) "The Cash Crop Revolution in Tropical Africa: an agricultural
reappraisal,"AfricanAffairs79, 314: 79-94.
(1981) "The Economy of the Southern Sudan under the British, 1898-1955,"
Journalof Imperialand Commonwealth HistoryIX, 3: 275-88.
Trapido,Stanley. (971) "South Africa in a ComparativeStudy of Industrialisation,"Journal
of Development Studies7, 3: 309-20.
(1978) "Landlordand Tenant in a Colonial Economy:the Transvaal,1880-1910,"
Journalof SouthernAfricanStudies5, 1: 26-58.
(1980) "'The friends of the natives': merchants, peasants, and the political and
ideological structure of liberalism in the Cape, 1854-1910," pp. 247-74 in Marks and
Atmore (1980).
Trevor-Roper,Hugh. (1965) TheRise of ChristianEurope.London:Thames and Hudson.
Triulzi, Alessandro. (1981) "Decolonising African History," pp. 286-96 in RaphaelSamuel
(ed.) People'sHistoryandSocialistTheory.London:Routledge.
Turner, Victor. (1957) Schismand Continuityin an AfricanSociety.ManchesterUniversity
Press.
Twaddle,Michael. (1972) "The MuslimRevolution in Buganda,"AfricanAffairs282: 54-72.
. "Gandareceptivityto change," Journalof AfricanHistoryXV, 2: 303-15.
Usman, Yusufu Bala. (n.d.) "The Transformationof PoliticalCommunities:Some Notes on
the Perceptionof a SignificantAspect of the Sokoto Jihad." HistoryDepartmentseminar
paper,Ahmadu Bello University.
Uzoigwe, Godfrey N. (1980) Review of Tosh (1978) International Journalof AfricanHistory
Studies13, 1.
van Binsbergen,Wim M. J. (1981) ReligiousChangein Zambia:Exploratory Studies.London
and Boston:KeganPaul.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
224 AFRICAN STUDIES REVIEW

van Onselen, C. (1972) "Reactions to Rinderpest in Southern Africa," Journalof African

HistoryXIII, 3: 473-88.
Vansina,Jan. (1966) Kingdomsof theSavanna.Madison:Universityof WisconsinPress.
(1968) "The Use of EthnographicData as Sources for History," pp. 96-124 in T. O.
Ranger (ed.) EmergingThemesin AfricanHistory.Nairobi:East AfricaPublishingHouse.
(1973) The TioKingdomof theMiddleCongo,1880-1892.OxfordUniversityPress.
.
. (1978) The Childrenof Woot:A Historyof the KubaPeoples.Madison:Universityof
WisconsinPress.
van Zwanenberg, R. M. A. (1975) Colonial Capitalismand Labour in Kenya, 1919-1939.
Nairobi:East AfricanPublishingHouse.
Vatikiotis,P. J. (1968) "Some PoliticalConsequencesof the 1952 Revolution in Egypt," pp.
362-87 in P. M. Holt (ed.) Politicaland Social Changein ModernEgypt.Oxford University
Press.
Vincent, J. (1977) "ColonialChiefs and the Makingof a Class," Africa47, 2.
von Freyhold. (1979) Ujamaa Villagesin Tanzania:analysisof a social experiment.London:
Heinemann.
Waldman,MarilynR. (1965) "The Fulanijihad:a Reassessment," Journalof AfricanHistory
VI, 3: 333-55.
Waller, R. D. (1978) "'The Lords of East Africa': The Maasai in the mid-nineteenth
century, c. 1840-c. 1885." CambridgeUniversityPh.D. thesis.
Wallerstein,Immanuel. (1961) Africa,thePoliticsof Independence. New York:Vintage.
(1964a) TheRoad to Independence: Ghanaand theIvoryCoast.Paris:Mouton.
.
(1964b) "VoluntaryAssociations,"pp. 318-39 in Colemanand Rosberg (1964).
(ed.). (1966) SocialChange:TheColonialSituation.New York:Wiley.
(1974) TheModernWorldSystem:CapitalistAgriculture and the Originsof theEuropean
World-Economy in theSixteenthCentury.New York:AcademicPress.
Wambaa,R. M., and Kenneth King. (1975) "The PoliticalEconomy of the Rift Valley: A
Squatter Perspective," pp. 195-217 in B. A. Ogot (ed.) Hadith 5: Economicand Social
Historyof East Africa.Nairobi:East AfricaLiteratureBureau.
Warren,Bill. (1980) Imperialism: Pioneerof Capitalism.London:New Left Books.
Washbrook, D. A. (1981) "Law, State, and AgrarianSociety in Colonial India," Modern
AsianStudies15, 33: 649-721.
Waterman,Peter. (1977) "On Radicalismin AfricanStudies," pp. 1-17 in P. C. W. Gutkind
and P. Waterman(eds.) AfricanSocialStudies:A RadicalReader.London:Heinemann.
Wayne, Jack. (1981) StructuralContradictions of the ColonialState: the Case of Tanganyikain
theEarly Years.Toronto University:SociologyDepartment.
Weber, Eugen. (1977) PeasantsintoFrenchmen: TheModernization of RuralFrance,1870-1914.
London:Chattoand Windus.
Webster, J. B. (ed.). (1979) Chronology,Migration,and Droughtin Interlacustrine Africa.
London:Longman.
Weiner, Myron. (1962) ThePoliticsof Scarcity:PublicPressureandPoliticalResponsein India.
ChicagoUniversityPress.
Weiskel, Timothy C. (1980) French ColonialRule and the Baule Peoples:Resistanceand
Collaboration, 1889-1911.Oxford:ClarendonPress.
Welsh, David. (1971) The Roots of Segregation:NativePolicy in ColonialNatal, 1845-1910.
CapeTown: Oxford UniversityPress.
Werbner, Richard. (1967) "Federal Administration,Rank, and Civil Strife among Bemba
Royals and Nobles," AfricaXXXVII, 1: 22-49.
Whitaker,C. S., Jr. (1970) ThePoliticsof Tradition: and Changein NorthernNigeria,
Continuity
1946-1966.PrincetonUniversityPress.
White, Landeg. (1980) "Literatureand History in Africa," Journalof AfricanHistory21, 4:
537-46.
Wilks, Ivor. (1967) "Ashanti Government," pp. 206-38 in DaryllForde and P. M. Kaberry
(eds.) WestAfricanKingdomsin theNineteenthCentury.OxfordUniversityPress.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions
 STATESAND SOCIALPROCESSESIN AFRICA 225

. (1971) Political Bi-Polarity in Nineteenth CenturyAsante. Edinburgh University: Centre

of AfricanStudies.
. (1975) Asante in the Nineteenth Century: The structureand evolution of a political order.
CambridgeUniversityPress.
. (1978) "Land, Labour, Capital, and the Forest Kingdom of Asante: A Model of
EarlyChange," pp. 487-534 in J. Friedmannand M. J. Rowlands (eds.) TheEvolutionof
SocialSystems.PittsburghUniversityPress.
. (1979) "The Golden Stool and the Elephant Tail: an Essay on Wealth in Asante,"
Researchin EconomicAnthropology II: 1-36.
Williams, Gavin. (1976) "There is no theory of petit-bourgeoispolitics," Reviewof African
PoliticalEconomy6: 84-89.
Willis, John Ralph. (1979) "The Torodbe Clerisy:A Social View," Journalof AfricanHistory
XIX, 2: 195-212.
Wilson, Monica. (1959) CommunalRitualsof theNyakyusa.OxfordUniversityPress.
. (1977) For Men and Elders: Change in the Relations of Generations and of Men and
Womenamong the Nyakyusa-Ngonde People, 1875-1971. London: InternationalAfrican
institute.
Wolpe, Harold. (1972) "Capitalism and cheap labour-power in South Africa: from
segregationto apartheid,"EconomyandSociety1, 4: 425-56.
. (1980) "Introduction," pp. 1-43 in Wolpe (ed.) The Articulationof Modes of
Production.London:Routledge.
Wood, Alan. (1950) TheGroundnut Affair.London:The Bodley Head.
Wright,Michael (1971) Bugandain theHeroicAge. Nairobi:OxfordUniversityPress.
Wrigley,C. C. (1959) "The ChristianRevolution in Buganda," Comparative Studiesin Society
andHistoryII, 1: 33-48.
. (1965) "Kenya:The Patternsof Economic Life, 1902-1945,"pp. 209-64 in Vincent
Harlow and E. M. Chilver, with Alison Smith (eds.) Historyof East AfricaII. Oxford:
ClarendonPress.
(1971) "Historicismin Africa:Slaveryand State Formation," AfricanAffairs70, 279:
113-24.
(1973) "The Storyof Rukidi," AfricaXLIII,3: 219-34.
(1976) "Changes in East African Society," pp. 508-43 in D. A. Low and Alison
Smith (eds.) Historyof EastAfricaIII. Oxford:ClarendonPress.
(1978) "Sketch Maps of African Economic History." Nairobi University, History
DepartmentSeminarpaper (14 December).
Young, Crawford.(1970a) "Decolonizationin Africa," pp. 450-502 in L. H. Gann and Peter
Duignan (eds.) Colonialismin Africa,1870-1960II. CambridgeUniversityPress.
(1970b) "Rebellion in the Congo," pp. 969-1011 in Robert I. Rotberg and Ali A.
Mazrui (eds.) ProtestandPowerin BlackAfrica.New York:OxfordUniversityPress.
(1976) ThePoliticsof CulturalPluralism.Madison:WisconsinUniversityPress.
Zolberg, Aristide R. (1966) CreatingPoliticalOrder:TheParty-Statesof WestAfrica.Chicago:
Rand McNally.
(1968) "The Structureof PoliticalConflict in the New States of TropicalAfrica,"
AmericanPoliticalScienceReviewLXII, 1: 70-87.

This content downloaded from 194.29.185.25 on Tue, 17 Jun 2014 14:46:06 PM

All use subject to JSTOR Terms and Conditions