Slavery Came First
by Martin Legassick
(SARoB, 43 May/June 1996)
Martin legassick reviews two substantial books on legal slavery and
coerced labour at the Cape. He disagrees with both over the role slavery
played in shaping race relations in the 20th century
Children of Bondage: a Social History of the Slave Society at the
Cape of Good Hope, 1652-1838
by Robert C-H Shell.
Wits University Press, 1996, xlii + 501pp, R79.95
Slavery in South Africa: Captive Labour on the Dutch Frontier
edited by Elizabeth A. Eldredge and Fred Morton.
University of Natal Press, 1994, xviii + 311pp, R75.95
The experience of slavery in South Africa has been largely forgotten among
its descendants and in public historical representation, as well as in historical
writing. These two books form part of a corpus of work that, over the last
decade, has begun to pierce this amnesia and to examine the nature and influence
of slavery. Shell's book - fifteen years in the making, and eagerly awaited
by specialists - concentrates on the period of legal slavery in the Cape
Colony. Eldredge and Morton's book, inter alia, argues for the persistence
of slavery on the Highveld after its formal abolition in the Cape.
Shell's book, relatively unusually, links research on the slave trade with
research on slavery itself. The core of his work has been the laborious
and diligent compilation of the most comprehensive quantitative data yet
on Cape slavery. It has been frustrating work: see his appendix on 'The
records of the slave trade'. To give only one example: slaves owned by officials,
being illegal, were systematically excluded from censuses and must be traced
from other sources.
The result has been heralded as a 'magnificent' study, as outstanding as
Van Onselen's work on the social history of Johannesburg. I cannot agree.
Analysis of his data does indeed allow Shell to make confident and sometimes
novel assertions. 'The popular notion that most Cape slaves were imported
from the East ... has to be modified. The slave trade to the Cape started
in West Africa, turned east after 1706, and finally became re-Africanized
after 1780. A slim majority of slaves imported to the Cape were African'
(p. 65). The 63,000 slaves imported to the Cape between 1652 and 1808 were
not merely diverse in origins to a degree 'unparalleled in any other recorded
slave population anywhere in the world' (p. 46): also their origin changed
significantly over time. To each other they were 'somatic and linguistic
strangers' (p. 88). This had obvious consequences for language and culture.
The ratio of male to female slaves was 'among the highest ever recorded
for a slave society' (at least until 1808), inhibiting family creation (pp.
66-77, 290, 405). For the first century, over 50% of slaves lived in urban
Cape Town (pp. 138ff, 408). In contrast to the Americas, female slaves were
not put to work in the fields, but remained in the house (p. xxxiii, 290-1,
406) - where some became incorporated as wet-nurses or concubines, and some
freed as wives of owners (see pp. 246, 397).
We knew previously that the overall ratio of slaves to freemen was of the
order of 1 to 1, in contrast to the 5 or 10 to 1 in plantation slave societies
(see also pp. 155-6). Shell shows that while by the eighteenth century most
freemen owned slaves, less than 7% owned most of the slaves (see pp. 153ff).
On the basis of distribution of slaves over different properties of an owner,
he estimates that the average slave-holding on a property was 3-5, with
few properties having more than 20 slaves (p. 407). He has also, less quantitative,
but nonetheless interesting insights into the naming of slaves (chapter
8) and the influence of slavery on Cape architecture (chapter 9).
Quantitative analysis thus generates some interesting results, though sometimes,
as in Shell's treatment of the internal slave market and its effects on
slave family formation (chapter 4), it tends towards the tedious. But what
the book centrally lacks (as has been pointed out by fellow historian of
slavery Andrew Bank) is a depiction of the experience of slavery. 'The slave
response to bondage, the leisure activities of slaves, the working lives
of slaves ... even slave interactions within the household - these are all
conspicuous silences ... The actual experience of slavery slips through
the grid of graphs and tables.' Rather than the subtle teasing out of how
slavery appeared to the slave, as in, for example, John Mason's recent work,
we experience slavery largely through the eyes of the slaveowner.
Some of Shell's more controversial arguments, in fact, appear to rest little
on his quantitative data. Recent writing (Robert Ross, Nigel Worden) challenged
the previous historical verdict that Cape slavery was 'mild', pointing to
the intensity of exploitation and the violence with which the system was
maintained. While concurring that it was not 'mild' (p. 395), Shell criticises
their idea that Cape slavery 'depended principally on whips and chains'
as a 'naive neo-abolitionist view' (p. 206). He argues (a) that laws limited
punishment of slaves by their masters to 'the same type of punishment a
husband and father could apply to his wife and children', and excluded whipping
(p. 208) and (b) (somewhat contradictorily) that violent punishment was
not restricted to slaves. However, slaves could be legally whipped by their
owners (but not wives and children, surely?). Moreover, the state reserved
a series of punishments for slaves (such as breaking on the wheel) that
were not applied to anyone else.
Shell insists that 'Physical coercion alone ... can never explain why slavery
worked in South Africa ... The chains of slavery were, in the main, psychological
- a more subtle but no less certain form of cruelty' (pp. 206, 212). The
main vehicle for their domination, he claims, was the 'household', which
slides over into the 'family': 'The metaphor of family control was projected
over all members of the Cape household' (pp. xxvi, 214ff, 395). Apart from
in the Company Lodge (on which he has a good chapter) he argues slaves were
distributed on properties in relatively small numbers and that this is the
basis 'for all subsequent sociological and psychological assumptions I have
made about the family and the incorporation of slaves into the settler household'
(pp. 155). Slaves were controlled as 'permanent children' (pp. 222ff, 410).
He lays out three categories of 'household': patrician, patriarchal, and
paternal, and tells us that 'patriarchal' constitutes a more severe, less
sentimental, and more chauvinistic form than 'paternal' (pp. xxvii). 'Patrician'
in this respect, hangs in the air. He variously (and confusingly) ascribes
patriarchy to the colonial family in general (p. 403), or the 'frontier'
(p. 407) and paternalism to early settlers (p. 403), urban slaveowners (p.
407) or British rule (p. 399, 403).
I can only agree with Robert Ross in questioning this emphasis on patriarchy
or paternalism. The 'family', for post-Engels, post-Freud, post-feminist
analysis, is of course a site of conflict and exploitation rather than the
affective care which 'paternalism' at least suggests. Yet the evidence for
slave-masters regarding and treating slaves as 'quasi-kin' is meagre - let
alone the acceptance of this status by slaves. 'Master', not 'father', was
the common form of deferential address by slaves to their owners. The 'paternalism'
of the early nineteenth-century period of British-inspired 'amelioration'
of slavery is, moreover, a much looser use of the term, descriptive, surely,
of the state rather than the household.
Shell correspondingly challenges the concept of the slave as a social 'outsider'.
Locally-born slaves, he argues, who generally began to outnumber imported
slaves by the mid-eighteenth century (and not just after the end of the
slave trade) were consistently preferred by slave-owners (pp. xxxviii, 46-8,
55-8). However - somewhat confusingly - he accepts in a qualified way the
related idea of the slave as 'marginal', including to the family (pp. xxxviii,
Where Shell has made significant innovation is in his insistence that the
condition of Khoisan labourers, from early on, began to approximate to that
of slaves. In particular, he draws attention to the onset of the system
of 'apprenticeship' from the 1720s and not from its formal beginning in
the 1770s (pp. 29-30). This was, initially, a form of bonded labour by the
children of Khoisan women and slave men, but was extended in the Cape to
captured 'orphan' Khoisan children - particularly after the ending of the
slave trade. Though Shell describes this system as one of 'serfdom', it
is this idea of the capture and enslavement of the indigenous population
- plagium in Shell's terms - which is developed in the collection of essays
edited by Elizabeth Eldredge and Fred Morton.
Eldredge and Morton trace the capture, trading, and holding as 'chattel'
of indigenous people from the time of the early Cape through the nineteenth
century along the expanding 'Dutch frontier' in the interior - a term used
not merely for white Trekkers but including Griqua, Baster and other frontiersmen
(see pp. 8, 268). Chapters by Eldredge, Morton and Jan Boeyens provide extensive
evidence of raiding for children, who were bound to service by white farmers
in the Cape, Transvaal, and Orange Free State. Chapters by Shell (a repeat
of sections of his book), Nigel Penn and John Mason are not so relevant
to this theme. A chapter by Barry Morton on slavery among the Tswana - the
so-called bothlanka system - is apparently intended to draw the analogies
of 'Dutch frontier slavery' to 'African' rather than New World plantation
Eldredge and Morton's book is in part stimulated by the historiographical
revision spurred by Julian Cobbing, who made the argument that the events
hitherto known as the mfecane and attributed to the rise of the Zulu state
were in fact the product of colonial slave-raiding from Delagoa Bay and
the Cape. Another chapter in this volume by Eldredge casts substantial doubt
on the Delagoa Bay end of Cobbing's thesis - but her chapter on the early
Cape reinforces the idea of an expanding market for captives, and the effects
of raiding for them on the indigenous peoples of the Highveld.
Fred Morton takes to task previous historians for failing to recognise that
indigenous South Africans were enslaved, and that slavery extended beyond
the Cape. Here it seems to me he lumps together too readily earlier historians
with the 'revisionist' writing of the 1970s. It was in fact (as recognised
by Boeyens, p. 192) pioneering pieces by Stan Trapido, and Peter Delius
and Trapido, that first drew attention to the inboekselinge of the
Transvaal as 'part of the evolving social process of the society'. If they
hesitated to describe this system as 'slavery' - preferring the idea of
a 'dependent servile class' - it was because of the grey areas and special
features of this system (brought out, incidentally, in Mason's chapter).
It is true that inboekselinge, supposed to be freed at age 25, were
often retained in service. But, as even Morton points out, 'Upon reaching
adulthood, inboekselinge were known to marry, raise families, and
even to begin farming on their own. But they settled close to their former
owners and remained ever liable for service, as did their children' (p.
174) - and he refers several times to 'ex-slaves' (pp. 178-80). This was
not slavery as a uniform inheritable status. At best these were, in Mason's
ironic phrase, 'fortunate slaves' (see chapter 4).
In fact the book as a whole, while providing useful additional evidence
on the question (particularly on the Western Transvaal and Zoutpansberg),
sensationalises and moralises the issue of 'slavery' somewhat in the manner
of nineteenth-century missionaries like Livingstone, rather than examining
the actualities of social relationships. Trapido and Delius's account of
the inboekselinge is a much more subtle one of the system as the
outcome of a balance of forces between relatively weak Boer societies and
stronger African societies.
Both Shell and the collection edited by Eldredge and Morton attribute a
dominant shaping role for slavery to the twentieth-century system of South
African race relations. 'Slavery, not the frontier and certainly not the
process of industrialisation', Shell tells us, 'shaped South Africa' (p.
395). Fred Morton writes of the inboekstelsel as 'the "missing
link" in the purview of South African labor relations. As a form of
indigenous slavery that originated in the Cape and was transferred inland
in the nineteenth century, inboekstelsel appears to serve for analytical
purposes as a transitional form of coerced labour the history of which begins
with the plantations of Stellenbosch and ends with the mining compounds
of Witwatersrand' (p. 266).
Certainly slavery had a general formative effect on shaping the master class's
attitudes to labour and labourers. In the Western Cape it also, arguably,
moulded the consciousness of post-slave labour. Those who argue that the
present-day coloured vote for the National Party is a consequence of 'slave-mentality'
have a part of the truth. But whether the institutional effect of slavery
was transmitted into the twentieth century primarily through the inboekstelsel
(a view which is also implicit in Shell: see for example pp. xx, 171, 403)
is doubtful. This looks like another attempt by historians to ascribe twentieth
century racial institutions solely to the 'frontier tradition' and to Afrikaners.
Slavery meant close physical proximity between slave and master in a living
and working relationship of a permanent kind - regulated by mechanisms establishing
social distance. Whether through long-distance transport or through capture
of children, enslavement had the consequence of the 'natal alienation' of
the slave: extreme cultural alienation from his or her origins. Migrant
labour and segregation, on the other hand, the products of the mineral revolution,
were based on maximum separation of living relationships consistent with
a working relationship of black labour for white masters. They also precisely
aimed to avoid 'natal alienation' by retaining the link of the (male) labourers
with their rural place of origin. Moreover, all the evidence goes to show
that the migrant labour system and segregation were constructed out of a
dialectic of relationships between agents of the (English-speaking) mineowners
and 'tribal' African societies - by-passing the labour relations established
by Boer farmers.
Even in agriculture on the Highveld, the inboekstelsel eroded. Share-cropping,
labour tenancy, and other forms of social relationship on white farms were
the predominant link into the twentieth century.
If there is a 'missing link' between slavery and the labour relations of
the mineral revolution it is more fruitfully to be sought, in my view, in
the nineteenth-century Cape than in the Boer Republics. Clifton Crais and
others have pioneered the historical examination of post-formal-slavery
labour relations in the Cape: the reconstruction of the ideology and institutions
of coerced labour under conditions of liberal post-Enlightenment British-dominated
capitalising society - and their 'export' to other parts of British Africa.
Shell's book, while tracing the evolution of stereotypes based on origins
and descent, and pointing to the 'preconceived, very generalized, pre-Darwinian
ideas about inferiority of people of non-European descent' (p. 409), implicitly
supports the conclusion drawn by others that during the period of DEIC slavery
no explicitly racial ideology emerged in the Cape. Instead, as Crais and
Worden put it in their introduction to Breaking the Chains, a set
of essays on the transition from slavery in the Cape, it was 'The very era
that ended bonded labour and, in 1853, established a non-racial franchise,
[which] also saw the emergence of new forms of unfree labour and, perhaps
most tragically, the emergence of modern racist ideologies in South Africa'.
This period, the first half of the nineteenth century, is also the subject
of a forthcoming book by Tim Keegan, who lays emphasis on the role of English
merchants, land speculators, and missionaries in the creation of the new
institutions and ideologies.
The books under review are diligently researched, and - despite flaws -
shed some light on their areas of focus. As guides to the more general contours
of South African history, however, they should be treated with circumspection.
Martin Legassick is Professor of History at the University of the Western
Up One Level
Back to document index