Landscape transformations and the archaeology of impact: social disruption and state formation in southern Africa.
LANE, PAUL
WARREN R. PERRY. Landscape transformations and the archaeology of
impact: social disruption and state formation in southern Africa. xv+180
pages, 23 figures, 1 table. 1999. New York (NY): Kluwer Academic/Plenum;
0-306-45955-8 hardback $62.
Roughly between the 1790s and 1830s, a series of events involving
widespread social and economic upheaval, a rise in warfare and
militarism, and major population displacements and migration took place
across much of southeastern Africa and immediately adjacent areas. Known
in the Nguni languages as the Mfecane, or `crushing', and in
Sotho-Tswana vernacular as `the scattering', or Difaqane, this
period of history, the events that unfolded, and their probable causes
and subsequent consequences have long been a feature of scholarly debate
and popular discourse. Much of this has had as its central theme the
rise of the Zulu state as an expansionist and despotic regional power,
under the authority of powerful military leaders, most clearly
epitomized by the life and times of king Shaka KaShenzangakhome. By the
1960s, a degree of consensus had been reached among historians that this
episode of state formation was an entirely indigenous, African-led
process, although considerable disagreement remained over the precise
catalysts, of which population growth, environmental degradation and/ or
long-distance trade were considered the most likely. Subsequently, this
view came to be challenged by a new generation of scholars, most notably
Julian Cobbing (1988), who argued that the Mfecane/ Difaqane was little
more than a myth that had initially taken shape among the European
settlers of the region as an alibi to obscure the instigation of
commodity slavery. Cobbing's critique, in turn, has generated
renewed interest among historians in the issues and enhanced critical
assessment of the available sources, including those used by Cobbing
himself (e.g. Eldredge 1992; Hamilton 1995). Warren Perry's book is
the first attempt to identify potential archaeological indicators of
these troubled times and to assess the competing explanations of the
rise of the Zulu state. The book is essentially in three parts. An
initial review of the historical debates and identification of the
potential archaeological indicators of impact and correlates of the
pre-Cobbing theories of Zulu state formation and the Mfecane/Difaqane,
which Perry characterizes as the `Settler Model' (chapters 1-3);
testing of the Settler Model using previous published archaeological
material and the results of the author's own fieldwork in Swaziland
(chapters 4-5); and a reappraisal of the historical debates and
identification of issues requiring further archaeological research
(chapter 6).
Perry's own fieldwork and the results of his surveys and test
excavations are presented in some detail. Given that this material has
not been published previously, the importance placed in his analysis on
settlement size and layout, and more generally, the scarcity of
archaeological research on the Iron Age and later periods in Swaziland,
it is regrettable that no site plans, trench plots or section drawings
have been reproduced. Nor are any of the recovered finds illustrated.
Perry, nevertheless, derives from this and his review of related studies
a detailed settlement typology and hierarchy so ,as to provide a series
of archaeological correlates for different elements of the Settler
Model. He tests each of these correlates against his data at both a
regional scale and the various regional subdivisions for both the pre-
and post-Mfecane/Difaqane phases. In all cases, the archaeological data
diverge from the pattern predicted by the Settler Model. Moreover,
assessment at the regional level tends to obscure the widely different
trends evinced at the sub-regional (Zululand, Swaziland etc.) level.
Interestingly, the best fit is Zululand where three out of the five
tests conform to the patterns predicted by the model. In all oF the
other sub-regions, the observed changes in population size, degree of
social hierarchy, number of cattle enclosures, evidence for trade in
European goods and the rank-size relationships between settlements in
the majority of cases differ from the predicted trends. Accordingly,
Perry concludes that the archaeological data do not support the premises
of the Settler Model, and argues instead, with Cobbing, that the primary
causes of the Mfecane/Difaqane and its consequences were the impact of
colonialism and especially the rise of racial commodity slavery. Perry
concludes by pointing to the neglect of archaeological investigation of
this issue in southern Africa, as well as the lack of physical evidence
concerning settlement distributions etc. during this period in the
critical areas around Delagoa Bay.
Perry's conclusions are sufficiently provocative to stimulate
further research. However, omissions and commissions throughout the text
suggest a lack of familiarity with the regional archaeology. Various
different issues are often confused or conflated. Perhaps the most
glaring example comes on p. 22, where he cites an outdated secondary
historical source (albeit concerned with the Mfecane/Difaqane) in
support of the suggestion that scholars still conventionally assume that
large areas of Southern Africa were unoccupied until the 18th century.
This is manifestly untrue, and archaeologists and historians working in
the region have been some of the leading critics of this myth (which was
a central element of apartheid ideology) since at least the 1970s. By
the same token, when outlining the parameters of what he terms the
Settler Model Perry chooses to ignore the concern of the earlier
generation of historians to attribute the rise of the Zulu, Swazi and
Sotho states and the various Tswana kingdoms to indigenous forces and
innovation, while at the same time he overlooks the varied criticisms of
Cobbing's use of historical sources (e.g. Eldredge 1992; Hamilton
1995). In short, I would argue that Perry's Settler Model is a
parody of current, and even earlier, interpretations of the events
surrounding the Mfecane/ Difaqane and, as such, his subsequent tests of
this model are flawed. Even if one accepts the Settler Model, one can
take issue with some of the tests, such as those concerning stock
enclosure sizes. It is highly questionable that the very small (c. 1-2
sq. m) `enclosures' were used for cattle, as Perry seems to
believe, and it is more likely that these stone features represent the
remains of grain bins. Unfortunately, it is impossible to check this
against the original site reports since no list of all the sites
included in these analyses is provided. By the same token, one can
question Perry's reliance solely on published sources for compiling
his database. Many more sites associated with this period are known (and
are recorded in the local Sites & Monuments registers) than have
been published, and inclusion of these data may well have produced
markedly different results. In summary then, an imperfect assessment,
but one can concur with Perry's general conclusion that this is a
topic highly deserving of more targeted research.
References
COBBING, J. 1988. The Mfecane as alibi: thoughts on Dithakong and
Mbolompo, Journal of African History 29: 487-519.
ELDREDGE, E.A. 1992. Sources of conflict in Southern Africa, c.
1800-30: the `Mfecane' reconsidered. Journal of African History 33:
1-35.
HAMILTON, C. (ed.). 1995. The Mfecane aftermath. Johannesburg:
Witwatersrand University Press/Durban: Natal University Press.
PAUL LANE
British Institute in Eastern Africa, Nairobi
pjlane@insightkenya.com
