Chapter 4
Frontiers, borderlands, and Saharan/world history
James McDougall
Places which yield only the bare necessities of men’s lives must be inhabited by barbarous
peoples, since no political society is possible. […] The least populous countries are thus the
most fitted to tyranny; wild beasts reign only in deserts.
Jean-Jacques Rousseau, The Social Contract III, 8.
It would not be much of an exaggeration, and may even be a commonplace, to say that the
question of how best to assess northwest Africa’s place within a wider world and its history
has engaged travellers, writers, and scholars since Herodotus. For the great Greek compiler of
eyewitness veracity and astounding tales alike, the imaginable world (centred, of course, on
the Mediterranean) was bounded to the south by what he believed to be the curve of the Nile,
cutting east and north through the desert, as it was to the north by that of the Danube,
meandering symmetrically east and south from the land of the Celts. Beyond both were
unimaginable barbarian lands without comprehensible language or civilisation.1 The Sahara’s
credentials as a limit, an edge of the world, are thus well-anchored in the European history of
ideas about the world and its inhabitants, and the extent to which they might be known;
unlike the deep forests of central Europe, of course, the ‘great desert’ maintained this ancient
mystique well into modern times. Even from the much closer perspective of medieval Arabic
writers, the bilād al-sūdān functioned in historical and geographical literature as a limit to
what (and whom) could be known and included in the recognisable world, and what could be
1
Herodotus, Histories II, 32; V, 11.
1
left to the imagination: somewhere out across the desert there was a modicum of law,
religion, manners and settled life, and connections between Sijilmassa and Awdaghust or
between Ouargla and Tadmekkat/Es-Souk were so regular that Ibadi texts could use them as
the setting for morality tales (Levtzion and Hopkins 1981: 90-91). But there were also singlebreasted women and kings riding horned beasts. From the dimly-perceived states and
potentates of the medieval era (like those depicted on the famous Catalan Atlas of 1356,
showing the fourteenth-century Malian king, Mansa Musa, holding an apple-sized nugget of
gold) through to much more recent images of the region, the Sahara’s vast attraction as an
unknowable frontier, alternately empty of anything or full of fears, has been inversely
proportional to the extent to which it has been understood. This continues to be a barrier to
broader understanding of the region in the wider world, and to some degree a potential danger
for the region’s peoples themselves, today (E. A. McDougall 2007c).
Fernand Braudel, also looking at the world from the Mediterranean, nonetheless
enabled a very different angle of view when, in the English translation of his monumental and
massively influential The Mediterranean and the Mediterranean world in the age of Philip II,
he turned the map of the world upside down to illustrate the world-historical location of the
middle sea, and the Sahara as its ‘second face’ (1972: 169, 171). Startlingly, on Braudel’s
map, the Sahara is suddenly the centre of the world, or at least, the centre of a large part of
the eastern hemisphere. For Braudel, who had taught at a secondary school in Algiers before
the Second World War and was influenced there by a very formidable tradition of French
colonial geography and Orientalism, the Sahara constituted a vast, connected zone unified by,
and enabling the unity of, the classical Islamic world, just as the Mediterranean was unified
by, and had enabled the unity of, classical Greco-Roman civilisation and the Europe thought
to have sprung from it. Still, and despite his own insistence to the contrary, Braudel’s
‘Turkish’ (i.e. Ottoman, Islamic) world lay beyond the frontier of the world he was most
2
interested in knowing. The Islamic zone, stretching from the western Sahara through
Cyrenaica and Egypt, through the successive deserts of Sinai and north Arabia across to Iraq
and the Iranian plateau, was an anoikoumené (1972: 174), an ‘un-environment’ defined by its
aridity and oases more than by the dynamic, long-term rhythms of interaction between
environment and society that he so brilliantly traced out in the temperate zone immediately
adjacent to it.2
More recent scholarship, especially from West Africanists, has, of course, greatly
improved on this older state of affairs, and as Ann McDougall’s chapter above explores,
Saharan studies properly so called can now claim several decades of extremely distinguished
existence. But, for reasons partly explored both in that chapter and in the introduction to this
book, the location of the Sahara itself in much of even the best recent work continues to pose
something of a problem. The most recent general overview (Austen 2010) calls its subject
‘trans-Saharan Africa’, which—the importance of its contribution notwithstanding—is
ultimately a way of writing the Saharan commercial economy into world history while
writing the Sahara itself out of it. Austen’s overarching narrative is one of a delimited period
(albeit a millennium long, from around the eighth century to the late nineteenth/early
twentieth) in which the Sahara is open as a corridor of the world economy, connecting the
producers of West Africa with consumers of their gold and slaves in the Mediterranean and
Middle East. When this period closes, at the advent of the colonial railway, the Sahara returns
to what it seems to have been in the Roman and Phoenician age before the advent of the
unifying bonds of Islam; an area marginal to the currents of world trade and a backwater of
2
Although Braudel professed himself convinced that ‘the Turkish Mediterranean lived and
breathed with the same rhythms as the Christian’, he also acknowledged the limits of his own
(vast enough) sources, through which he could ‘glimpse the Turkish world from the outside
only’ (1972: 13-14).
3
world history, dividing Africa ‘proper’ to its south from the Arab world to its north and east,
belonging properly to neither and unable to assert a singularity of its own.
In its broad traits, of course, this long-term history of northwest Africa is perfectly
lucid, especially as a general account within a concise approach to major geographical
regions of world history. But there is, inevitably, a good deal that is thereby left out of the
picture, and in some respects it may be that the overall picture itself might be considerably
altered. As archaeological research proceeds apace in the Fezzan and around the Niger bend,
we may now need to revise the long-held, cautious view of the absence or extreme
tenuousness of pre-Islamic commerce across the Sahara (Wilson 2009, Fentress 2009,
Schörle, this volume). And just as the caravel is no longer assumed to have displaced the
caravan when European Atlantic shipping began to tap Saharan trade from West African
seaports, it may be that patterns of intra-Saharan and cross-Saharan exchange have in fact
rebounded, since the calamities of the colonial period, such that the ‘closure’ of the desert
corridor as a route of world trade in the early twentieth century might now look more like a
brief parenthesis in a longer, continuous history than a final and defining death-knell.3 Of
course, the mobilities and connections of the later twentieth and early twenty-first centuries
obey rather different logics, and play out over very different patterns of possibility and
constraint, than did those of the preceding two millennia. And as we shall see, having long
3
For the endurance and expansion of Saharan trade after the sixteenth century, Newbury
(1966), E. A. McDougall (1990, 2007a), J. L. A. Webb (1995). Austen (2010: xi) notes that
global economic integration from the fifteenth century onwards ‘marginalised trans-Saharan
commerce in global terms but stimulated its accelerated local growth.’ For a similar argument
in relation to the continent’s economic history as a whole, see also Austen (1987). For the
persistence or resumption of exchange (and migration) in the Sahara, Choplin, Brachet,
Scheele, Oudada, this volume.
4
maintained a certain ‘internal sovereignty’ over the spaces they inhabited, and over the
internally-managed connections and exchanges that made them habitable, the region’s
peoples have certainly found themselves more recently facing a world turned increasingly
outward. But looking at the region in the very long run, there seem to be good reasons for
directing attention not simply at the routes across the Sahara that have made it a global
corridor, but at the ‘spaces in between’ that have made it, and make it today, a ‘world
crossroads’ (Choplin 2009). More broadly, this chapter will argue, a recalibration not so
much of timeframe, as of our level of analysis, helps to put the Sahara back into focus as an
intelligible region in its own right, one that certainly connects its contiguous—and very
different—worlds of Mediterranean Africa to the north and highland-and-forest West Africa
to the south, the Atlantic coast to the west with the Nile valley to the east, but which also has
a distinct significance of its own. There is, in other words, a properly Saharan history within
world history, one that might repay comparative attention as giving visibility to a global
region whose interest (as other chapters in this volume argue in more detail) lies at least as
much in what constitutes the region from within as what moves across it from outside. One
way of thinking usefully about this from a historical perspective is to return to the question of
northwest Africa’s location in a wider world, as a set of overlapping, interdependent places in
between, and thus to resituate the Sahara within the recent comparative history of the frontier,
and the closely associated study of borderlands.
Saharan, African, and world history: frontiers, borderlands, and spaces in between
Beyond the ancient notion of the ‘edge of the unknown’ with which we began, the concept of
the frontier and models of historical development articulated around it have been very
productive for scholarship in a variety of world-historical regions over the past half-century.
While they have been applied to cases as far removed from each other as Latin America and
5
China, the problematic of the frontier has remained for many tied to the ‘Turner thesis’ and
its successors in North American history.4 This in itself has been a fruitful comparative angle:
the inversions and inflections experienced by Turner’s progressive, pioneering, expanding US
frontier when transplanted to Africa or Latin America have enabled new perspectives on the
social, cultural and political histories of a variety of distinct but comparable world regions.
From the original, triumphalist frontier, ‘the hither edge of free land’ as Turner defined it (no
doubt as much a poetic expression of a particularly late-nineteenth century American
sensibility as an analytical construct) to more recent concerns with the ‘edgy’ intercultural
margins, the shared-and-contested ‘middle grounds’ of imperial expansion and indigenous
survival struggle (White 1991, Silver 2008), the frontier concept within North American
history has continued to stimulate debate and illuminate the complexities of early American
history well beyond the significance of the east-west driving force of nineteenth century
‘progress’. On a larger scale, one of Turner’s students, Walter Prescott Webb (1952), already
anticipated more recent trends in the ‘Anglo-world’ literature that has emerged from the ‘new
(British) imperial history’ when he wrote of a ‘world frontier’, comprising Australia, New
Zealand and southern Africa as well as South and North America in a vast horizon of
European expansion, settlement, and capitalist enterprise. Elsewhere, other definitions of the
frontier have brought other historical patterns to light. In the 1980s and early 1990s, the focus
in frontier history shifted from the purely territorial line of demarcation at the edge of
4
If in 1960 it was possible to pinpoint ‘a neglect of comparative research’ (Mikesell 1960:
64) as ‘the principal failing’ of Frederick Jackson Turner and those who followed or critiqued
him within US history, the subsequent fifty years have seen considerable attempts to make
good. From a vast literature, see especially: Turner (2008 [1894]), Lattimore (1940), W. P.
Webb (1952), Lamar and Thompson (1981), Kopytoff (1987), Weber and Rausch (1994),
Faragher (1998), Adelman and Aron (1999).
6
‘progress’ to the often-fraught and always fuzzy lines of contact between cultures: the
frontier became ‘a territory or zone of interpenetration between two previously distinct
societies’ (Lamar and Thompson 1981:7), ‘geographic zones of interaction between two or
more distinctive cultures […], places where cultures contend with one another and with their
physical environment to produce a dynamic that is unique to time and place’—in this latter
sense, the frontier became ‘both place and process, linked inextricably’ (Weber and Rausch
1994: xiv). The places and processes that emerged from these analyses were often
considerably darker than the (deceptively) bright open sky of Turner’s West; in Latin
America the frontier seemed ‘a brutal place where the weak are devoured by the strong’,
characterised, when written about by city-bound nineteenth century Latin American
intellectuals, by ‘ignorance and primitivism’ (Weber and Rausch 1994: xvii). In southern
Africa, the ‘opening frontier’ of white settlement and black servitude in the eastern and
northern Cape, Griqualand and Transvaal, produced unstable, gun-and-liquor-running bandit
economies, endemic regional warfare and, in the 1857 Xhosa cattle-killing, catastrophic and
suicidal millennialism.5
In African history, the major point of reference in this literature remains Igor
Kopytoff’s (1987) edited volume The African Frontier: The Reproduction of Traditional
African Societies, which combined the frontier as space at the leading edge of settlement with
the frontier as cultural process. In his long introductory essay to the volume, the ‘internal
African frontier’, a perennially renewed space of social reproduction ‘lying at the fringes of
the numerous established African societies’, provides a continuous, cyclical incubator for the
5
In 1857, in a deep societal crisis provoked by continuous and brutal frontier war, the Xhosa
of the eastern Cape responded to a prophetic vision of salvation that demanded they slaughter
their cattle and destroy their grain. The ensuing mass starvation may have killed up to two
thirds of the population.
7
‘perpetuation of a pan-African political culture’ (1987: 3, 25-26). The socio-cultural patterns
that Kopytoff saw as being generated on the frontiers of existing societies, as new groups
constantly broke away from established ones, ‘stood Turner’s thesis on its head’ in that his
African frontier, instead of being a force for dynamic, progressive cultural transformation,
instead serves as ‘a force for culture-historical [sic] continuity and conservatism’ (1987:3).
This vision was important (and positively reviewed) in particular for its critique of what
Kopytoff called the ‘tribal model’ of progressively aggregated African polities emerging in
‘evolutionary’ fashion from wandering bands of proto-historical hunters or herdsmen, placing
the emphasis instead on the dynamics of settlement and the ‘spinning-off’ of new waves of
population movement at the edges of societies as they ‘mature’. For historians of central and
southern Africa, in particular, this was an attractive approach to ethnogenesis and polityformation, and in a sense it chimes in well with John Iliffe’s later characterisation, within a
largely demographically-oriented history of the continent, of Africans as ‘the frontiersmen of
mankind’ whose history is structured by the progressive settlement and colonisation (in the
general sense) of an especially rich but also especially challenging set of productive
environments (Iliffe 1995; see also, for example, Schoenbrun 1998).
None of the studies in Kopytoff’s volume concerned northwest Africa, which figured
in his deep history of ethnogenesis only as an almost mythical, ‘ancestral “hearth”’ where,
sometime before 2500 BCE, ‘the “incubation” of the ancestral pan-African culture patterns
took place, often under frontier conditions and in contact with the kindred patterns of the preIslamic Near-East’ (1987: 9). The recourse to a posited ‘cradle of ancestral culture’,
especially one so vaguely formulated, was perhaps unnecessary to Kopytoff’s argument. But
in this, as in the overarching conception of a single, ancestrally-defined ‘pan-African political
culture’, despite both its conceptual originality and its detailed empirical observation,
Kopytoff’s account simply displaced the old essentialism of ‘tribal’ origins onto a more
8
sophisticated but still essentialist theory of African history as ‘closed’, cyclical, and selfreplicating . In this sense, Kopytoff’s frontier shared with other frontier histories a
paradoxical tendency to reproduce the boundedness of the histories of the areas thus studied,
reproducing their particularism, exceptionalism, and enclavement rather than opening them
up beyond their various essentialisms to comparison and connection. The search for a more
properly comparative frontier history that might enable the histories of different world
regions to have something to say to each other remains here, as in much of the older
American frontier literature, unresolved.
One possible way around this difficulty might be sought in a field that has developed
in tandem with the frontier literature since the 1970s, and which again derives mainly from
work on America, this time the southwestern United States, namely the study of borderlands.
Together with the tradition of writing on the Hispanic frontier in the Americas (a rather
different, northward rather than westward and ‘inclusive’, mestizo rather than exclusive,
binary-racialised, boundary), and the newer history of early settlement and coexistence in
areas such as the American Great Lakes (White 1991), the focus on borderlands has stressed
the importance of more fluid, open, transitional zones than is often envisaged by the frontier
as a demarcation of settlement. Borderlands can usefully be seen, for example, as spaces in
which ‘autonomous peoples of different cultures are bound together by a greater, multiimperial context’ (Haefeli 1999: 1224). At the same time, such zones are set necessarily
within the sphere of inter-state competition for territorial sovereignty, access to resources,
and jurisdiction over people, i.e., within struggles to create, maintain and enforce the
recognition of borders.6 Here again, the possibility of engaging a more properly comparative
history on the basis of the relationship between frontier/boundary and borderland, however
6
See Baud and Van Schendel (1997), for a European perspective; Adelman and Aron (1999),
for an American one.
9
productive the analysis in each particular case, has been questioned. North America’s
frontiers, for example, are said to have been ‘remarkable for their instability and fluidity’
relative to the frontiers of other regions and societies, in ‘deserts, deep forests and vast steppe
lands’, that ‘tended to form along ecological boundaries and last for centuries’ (Haefeli 1999:
1224).
This, though, is perhaps to overstate the case for the stability of frontiers in the ‘old’
world. In northwest Africa, ecological boundaries themselves have moved, sometimes quite
rapidly; they have been less fixed features of the landscape than they have been expressions
of the frequently-shifting balance between the environment, its resources, and human
populations. (And as stated elsewhere in this volume, it is worth remembering that resources
are not always simply given by environmental conditions—like salt—but, like date palms and
irrigated gardens, are themselves produced by human action in the environment. Even salt
must be exposed, dug out or evaporated, water tables tapped and channelled.) At the shortest
end of the scale, the availability of pasture and water in parts of the Sahara and Sahel can be
vanishingly brief, exploited at the moment of their seasonal appearance and almost
immediately exhausted. Changing practices of settlement and land- or animal-husbandry can
effect dramatic shifts in ecology over a few decades, as in the case of the water-meadows
(bougoutières) of the Niger bend (Grémont, this volume). At the other extreme, the very
longue durée environmental history of ‘desertification’ since the seventeenth century,
punctuated (notably in the twentieth century, in the 1910s and again recurrently since the
1970s) by dramatic shorter-term episodes of drought, might be seen as an inexorable, secular
march of the arid frontier into the West African savannah; in between, the shorter-term
spread of bayoud, the ‘whitening’ disease of date palms caused by the fungus Fusarium
oxysporum albidensis from Morocco eastward through the Algerian oases over the course of
the past century can be regarded as a pathogenic advance forcing (or at least, threatening) the
10
retreat of cultivation. Such shifts in the short-, medium- and long-term suggest that caution be
exercised before we pronounce on the plasticity of new world political frontiers relative to the
permanence of old world ecological ones. Perhaps, again, the level of analysis as well as the
timeframe needs to be adjusted if we are to gain a more satisfactory grasp on the uses of
frontiers and their borderlands for a comparative world history.
Edges, empires, and borderlands: a politico-ecological history of the Sahara?
Comparative history necessarily seeks to determine both what is distinctive and particular to
a specific case through examining how the distinctiveness and particularity of one time and
place can be related to others; how setting one time and space in relief against other times,
across wider or different spaces, enables us to trace the lineaments of commonalities and
differences, and (if possible) identify or at least posit their causal factors. Some of the
distinguishing features of Saharan history are clear enough in themselves, and doubly so
when compared to other regions that have provided the focus for frontier or borderland
history. Its vast geographical size and minuscule, exceedingly sparsely distributed (but also
locally concentrated) population; the cultural centrality over most of the recorded past of a
single religio-legal system, Islam, to both social practice and the inhabited landscape
(Moussaoui, this volume); long-established patterns of settlement but the extreme precarity as
well as the dynamism of ‘towns’ (Scheele, this volume); the key importance of a very few
commodities to local and long-distance production and exchange, all set Saharan Africa off
against other parts of the continent (central or southern Africa) as well as from the
Mediterranean, let alone western China, southeast Asia, the American West or the interior of
Canada or Latin America. What does Saharan Africa, understood through the lens of
frontier/borderland dynamics, have to say to the very different histories of such regions?
11
What does the understanding of such a Saharan history have to contribute to the broader
comparative world history within which I have suggested it should be located?
If the problem of location has been particularly acute for the Sahara, perhaps the most
‘peripheral’ region of Africa and of African studies, African history more generally has also
long been preoccupied with the location of the continent relative to the wider world; Africa’s
long isolation from Eurasian connections, then its incorporated-and-marginalised position as
a periphery of world systems centred elsewhere, the internal dynamics of the continent that
have turned on dispersed and separate centres with only occasionally contiguous edges and
frequently conflictual contact zones, are recurrent themes in the literature. (They are also, to a
degree, mirrored in the conditions of the production of Africanist scholarship itself; think of
the extent to which, in academic departments, teaching programs, literatures and fields of
research, the various subregions of the continent—West, East, Central, and South—remain
relatively distinct from each other, a fact surely not sufficiently accounted for by the
undoubted difficulty of any single scholar working with any great expertise on more than a
very small part of the continent. The centres of interest and lines of demarcation also, of
course, vary whether one is looking at, say, American, British, or French fields of African
studies, let alone Africana or African-American studies in the US academy.)
To a degree, the regionally-oriented world history that has aimed to escape from the
straitjacket of continental units of analysis and has given us instead studies of the Atlantic
and Indian Ocean worlds has—despite the major role reserved for Africa and Africans in
both—tended to reproduce the same patterns. The African shores of the Indian Ocean or the
Arabian seas have become arenas of greater India or greater Arabia (or greater Hadramawt),
albeit sometimes veneered with a poetically polished sense of cosmopolitanism smoother
12
than is entirely tenable.7 The Africa of the black Atlantic is above all one side of the slave
triangle, and its Africans are primarily visible—through the records of slave shipping—once
they enter the diaspora. Of course, one major stimulus for the internal history of northwest
Africa, and a real strength of the literature in Saharan studies, has been precisely the concern
to investigate the obverse of the Atlantic trade; the expanding internal slave markets of the
nineteenth century and the longer-term desert-side and trans-Saharan economy in which they
played so important a part.8
But the larger problematic of Africa’s location in global histories, and the problem of
the Sahara as a void within Africa, remain, and are perhaps also linked. Indeed, many of the
classic controversies in African history, especially (but not only) in the history of West
Africa, arise as problems because of the way that Mediterranean Africa (as part of the
Mediterranean, or Arab, and hence non-African world) has often been excised from
conceptions of the continent ‘proper’, and the way that the Sahara has correspondingly been
seen—or rather, unseen—as an empty space in between. The major concerns of establishing a
‘properly’ autonomous African history, reliant on Africa’s own internal historical dynamics
rather than on putative external stimuli, were entirely comprehensible and even necessary in
7
Barendse (2002), Bose (2006). Hadramawt is the region of southern Arabia, contiguous
with the Indian Ocean coastline, from which generations of traders and Islamic scholars,
often claiming prophetic descent as sayyids, engaged in trade and the spread of learning and
religion throughout South and Southeast Asia. On cosmopolitanism, see Scheele and J.
McDougall, this volume.
8
Again, the literature is now very extensive, though John Hunwick (in Savage 1992a: 5)
could still refer to the Mediterranean and Islamic slave trades as having ‘to date […]
stimulated little interest’ relative to the larger literatures on the African diaspora in the
Americas. See Manning (1990), Lovejoy (2000, 2004), Wright (2007), Rossi (2009).
13
the wake of colonialism, the conceits of a supposedly ‘civilising mission’, and the racial
fabrications of the Hamitic hypothesis.9 In West Africa, the historiographical struggle to
locate the beginnings of trade and state-building in the internal dynamics of the Sudanic zone
independently of what was once assumed to be the necessary ‘stimulus’ of cross-desert
contact from the Islamic Maghrib happily revealed important local and regional patterns of
change (Levtzion 1973). But the concern to locate the sources of historical ‘advancement’,
whether in ironworking, commerce, or state-formation, axiomatically within ‘Africa proper’,
i.e. to the south of the Sahara, has perhaps been founded on a category error, one that sets up
a false opposition only because the Sahara is posited as a limit. Similarly, writing, religion
and legal practice transmitted from the north have been seen, like Arab influence on the
Swahili coast, as stemming from outside Africa, rather than being themselves formatively
African.10 In this sense, Kopytoff’s proto-Sahara as an ‘ancestral hearth’ of African traditions
at least had the merit of bringing the region back into the longue durée history of Africa
‘proper’, if only as a symbolic point of origin—and only up until the third millennium BCE…
9
On the Hamitic hypothesis (attributing the origin of ‘advanced’ peoples in eastern and
central Africa to in-migration from the north), see Collins (1968).
10
It is striking, in particular, that the Africanisation of religion (Islam and later Christianity)
remains so central a term of analysis to scholarship, as if it were possible that such massive
sociocultural change could happen without codes originating elsewhere becoming
constitutively part of what it has meant to be ‘African’, and being themselves adapted in the
process. In the case of Islam, the explanation can only lie in the persistent discomfort with
‘externality’ combined with a persistent tendency to assume ‘Islam’ as transhistorically and
cross-culturally monolithic, whence the need to demonstrate how different it is in Africa.
(But of course, ‘it’ is both different everywhere, and everywhere defined by recognisable
commonality.)
14
The otherness of the desert (or the sea) and the lands beyond is, of course, a local perception
as well as a scholarly one—in the Sahel as in the Maghreb and in Zanzibar, local claims to
external genealogical origins have played a role in hierarchies of status for as long as there
has been a perceived ‘outside’ from which one’s ancestors might credibly have come. But the
fact that such perceptions of distinction are taken seriously within Africa does not oblige us
to tailor our own analytical conception of Africa to their model. Relocating the Sahara as an
intelligible, rather than as an empty, space means first of all reintegrating it into Africa; we
need no longer fear that evidence for much older regional and cross-regional connections
between Mediterranean, Saharan and Sahelian Africa (Schörle this volume, MacDonald et al
2009, Mattingly 2003, 2007, 2010), regions whose distinctiveness and diversity are simply
part of the greater diversity of the continent, somehow denies ‘Africa’ its proper history.
As suggested in the Introduction above, Saharan Africa in this sense is not simply
coterminous with the desert; it is not defined as a bioclimatic zone whose borders are drawn
along lines of minimal rainfall and humidity, or even as loosely as the space between the
Atlantic, the Senegal and Niger rivers, Lake Chad and Darfur, the Atlas, the Gulf of Syrte,
and the Nile. It is rather a set of shifting and interdependent ecologies—relationships between
people and environment—held together by a shared (though not necessarily unifying) set of
characteristics and enduring over time. Geographically, the ‘great desert’ connects these
patterns of habitation and defines the marked challenges their people face. To make sense of
the Sahara as an intelligible region within continental Africa, and within a wider global
history, this book therefore turns to histories and contemporary analyses of the environment
and its inhabitants; to demography and ecology, the changing relation of people to space, and
of both to political societies, states, and the borders they impose. It is in the relationship of
ecological to political history that the concepts of frontier and borderland can be productively
15
applied to this part of the world, so as to enable a distinctively Saharan African pattern of
history to come into view.
The Introduction to this volume also suggested that the notion of connectivity,
borrowed from the new ecological history of the Mediterranean (Horden and Purcell 2000)
might, with suitable adjustments, have considerable explanatory force in capturing the
ecological dynamics of the Sahara as a region. The Mediterranean, for Horden and Purcell,
with its intensely fragmented landscape across short, easily navigable (or indeed audible or
visible) distances, was characterised by an especially intense connectivity producing ‘net
introversion’, at least until the twentieth century (Horden, this volume). The Sahara’s
connectivity, in its very different configuration of topography and distance, might be seen to
be determined by especially intense resource scarcity; the precariousness of ecological niches
even more vulnerable (especially to annual variations in rainfall or the level of the water
table) than those of the Mediterranean; and hyper-specialisation in a small number of
products (salt, dates…) producing absolute reliance on outside inputs and connection to
markets for exchange and the acquisition of labour. In these circumstances, inhabited areas
have stimulated connectivity with other inhabited areas simply by the fact of existing, since
existence in isolation has by definition been impossible. This has been true whether the
inhabitants of such spaces were ‘nomadic’ or ‘sedentary’; and of course many groups have
always moved between both lifestyles at different times of the year, or at least have made
certain—like the Tuareg nobles (imajeghen) for whom mobility is bound up with status and
labouring the land defines its loss—of access to the products of both animal husbandry and
cereal culture, metalwork and textiles, whenever necessary. Such connectivity is not the
almost frictionless movement across easily navigable, short distances seen in the
Mediterranean, but rather a combination of relatively intense patterns of interdependence (on
kin, religious specialists, vassals or slaves) within clusters short distances apart—the chains
16
of oases that lie strung out along irrigation channels—with the arcing connections between
such clusters and others across very long distances, distances which are often difficult and
time-consuming to traverse.11 And unlike the net introversion that defined the Mediterranean,
it seems clear that Saharan ecologies, though fundamentally created in the dynamics of
production and exchange within the region’s highly interdependent networks of habitation
and market, produced a connectivity of net extraversion, both over the long run and
increasingly in the present. This can be seen in every respect, from its human inhabitants
constantly renewed from beyond the fringes of the desert to its resources in grain and
livestock that depended on relations with contiguous zones of savannah (to the south) or
mountain pasture and tell (to the north). But crucially, while the Saharan economy—unlike
the rest of Africa before the colonial period—might always have been thus turned ‘out from
the inside’, making the Sahara always a borderland in the sense of a zone constituted by its
multiple interactions with neighbouring worlds, without which it would be unable to survive,
Saharans themselves until relatively recently retained control over the terms of access, the
terms of trade, and the terms of alliance that regulated their precarious relationships across
and beyond the spaces they inhabited. Until, that is, imperial competition and state-formation
undermined them.
A defining aspect of the very recent history of the Sahara, as of Africa and the global
South more generally, has been the stress laid by regional governments, under pressure and
with (albeit financially modest) technical assistance from Europe and the United States, to
engage in ‘border strengthening’. Once looked to hopefully for the betterment of local
livelihoods and opportunities, ‘trans-border cooperation as a cornerstone for effective
subregional and regional integration’ (Asiwaju 1996: 253) is nowadays more likely to mean
11
On the relationship between locality, intra-regional interdependence, and the ‘outside’, see
Scheele 2011:2-3.
17
securitisation and anti-terror, anti-migration control (see also Choplin, this volume). But
while the more recent history of ‘bordering’—the imposition of political boundaries as
policed, bureaucratic barriers—in the region has undoubtedly had considerable and damaging
effects, it also remains the case, as it long has been, that boundary-marking ‘sets up a zone of
interaction’ (a borderland) ‘rather than representing a genuine partition’ (Nugent and Asiwaju
1996: 2). And in this sense, again as has often been the case in Africa and elsewhere, borders
have not simply been imposed out of nowhere by entirely arbitrary functionaries on high (and
equally arbitrary soldiers and policemen on the ground), but rather, as Paul Nugent has
argued for the Ghana-Togo frontier, are ‘reinforced on a daily basis by the peoples who live
along [them]’ (Nugent 1996: 36).
If we extend our concept of the frontier (as border, boundary) in this sense beyond the
geopolitical and into the ecological field, it is easy enough to see the extent to which frontiers
in the Sahara have always existed as lines of demarcation across access to resources, from
water wells, salt deposits, and pasture to uranium mines, gas fields, and oil exploration
blocks. From the very fluid and consensual—access to salt mines ‘owned by no-one’ and
worked by all comers—to the heavily encoded and stratified—access to animal husbandry,
from low-status goats to high-status camels, to rights to mobility, water or pasture, to slaveownership, to marriageability—such demarcations long existed both as impositions from
those with the power to impose on others and as systems of reciprocal rights and observances
reproducible in daily life and livelihood, well before they began to be dictated by any form of
state. If the absence of the state is a distinguishing feature for much of Saharan history,
however, political society in the sense of established systems for the exercise, and the
recognition, of coercive power must be seen to have been a mainstay of existence in the
desert over the very longue durée. Certainly the long-distance slave trade could not have
existed without it, but nor could the much more intensively intra-regional forced mobility of
18
unfree labour required to irrigate and cultivate oases. Again, unlike the Mediterranean,
movement in the Sahara, whether physical or social, is anything but friction free; the ‘friction
of distance’ and of social hierarchy is, on the contrary, extreme. As lines of demarcation
across the distribution of resources, therefore, Saharan frontiers are drawn across both
geographical and social space; and as the availability of resources (whether the means of
subsistence or those of the mobility needed to procure them) expands or contracts in time and
space, so the frontier moves, relaxes, or tightens, such that the internal ‘borderland’ zones
around or astride such frontiers similarly expand or contract, intensifying or restricting the
degree of connectivity—the intensity or absence of exchange—between them, and altering
the overall shape of the region, the location, extent, and prosperity of its inhabited spaces and
the acuity of its dependence on the outside. The emergence or disappearance of access to
resources, of course, is again not simply an external stimulus to which people merely
respond, but is also influenced by regional people’s own agency, as they move into new
employment niches—Sahelian migrants heading to booming Mauritanian towns or young
Tuareg ishumar to the labour markets of Libya (Choplin, Marfaing, Brachet, this volume),
abandon old ties of servility or accept (however reluctantly) new relationships with the state
or international development and aid organisations. The Saharan frontier, then, never simply
opens or closes, and does not move in a straight line; it expands or contacts, advances or
retreats relative to the numbers and movements of population and their access to exploitable
resources. This flexible, internally regulated frontier, however, has also been overlaid in
modern times – broadly speaking, since the sixteenth century—with contests over the control
of ecological boundaries which empires, and national states after them, have endeavoured to
turn into, or subordinate to, political borders.
The crucial historical shift, then, is perhaps to be located not so much in the
appearance of borders tout court with the colonial and then the national state, but in the
19
gradual loss by the region’s peoples themselves of the ability to define, defend, and reproduce
the borders that have been most salient to the management of their ecological livelihoods and
the organisation of their social life. In their reassessment of the frontier and borderland
history of North America, Jeremy Adelman and Stephen Aron (1999) discussed the various
ways in which, caught between competing (British, French, Spanish) imperial powers in what
were initially open, borderless lands, the native American and inventively coexisting mestizo
populations of ‘the spaces in between’ were initially able to exercise control over their
territories and exchanges, retaining a certain sovereignty over the terms of trade and of
alliance with rival outsiders. Gradually, though, as the relatively open frontier gave way to
borderlands as areas of contested boundaries, and then to increasingly fixed, hierarchical and
non-porous borders (crystallised in the shift from imperial competition to the emergence of
nation states, especially the United States itself), the ‘middle ground’ vanished, and
indigenous control not only over the terms of trade with the outside, but even over the terms
of their own habitation of space, was abolished. Similarly, if over a rather longer period of
time, imperial competition succeeded by national boundary-marking might be seen to have
radically shifted the level of autonomy enjoyed by Saharans over the regulation of their own
frontiers and the shape of the borderland they inhabit.
In Saharan Africa, imperial competition first became really significant in the 1500s,
with the emergence of rivalry over Saharan influence between the Ottoman regencies of
Tripoli, Tunis and Algiers, on one hand, and Morocco on the other, just as both Moroccans
and Ottomans were themselves facing increased imperial competition from the Portuguese
and Spanish; the injection of imperial rivalry into the Saharan borderlands came with the
Moroccan expedition in 1591 to seize Timbuktu and the salt and gold trade of the Niger bend;
in the relations between the Ottomans and the state of Bornu, west of Lake Chad, and those
of Morocco with the Moorish hassan (military) emirates of the western Sahara north of the
20
Senegal. At first, and in some respects against the older narrative of regional collapse (Kaba
1981), the penetration of the region by outside imperial polities and the disintegration of the
Songhay state increased the margin of manoeuvre for regional groups.12 The late sixteenth
through the late eighteenth centuries in the southern Sahara and Sahel can be seen as a
dynamic period of both fragmentation and recomposition, the emergence of new power
centres controlled by emerging social groups and a proliferation of new states: Bagirmi and
Wadai to the east of Lake Chad, Segu upstream of the Niger bend, the Songhay successor
states downstream of Gao, as well as the—perhaps over-emphasised—arma viceroyalty of
Timbuktu and Jenné and the emergent Tuareg confederations that pressed upon it from the
north. In the central and northern desert regions, recognition and correspondence as well as
trade with one or another of the northern powers—association of ultimate, however
theoretical, authority in Shinqit or Tidikelt with ‘the sultan of Morocco’ or recourse in the
Gourara to the protection of the regency of Algiers—marked the presence but also the limits
of imperial penetration in the region. The intensity of imperial rivalry in the Sahara was
certainly far less than that between the French and British in the American Great Lakes
(especially once Morocco succumbed to civil war at the death of Ahmad al-Mansur in 1603),
12
The ‘catastrophe thesis’ of the Moroccan invasion in Songhay history begins with, and is
usually adduced from, the Timbuktu chronicles of al-Sa‘di (Ta’rīkh al-sūdān) and Ka‘ti/Ibn
al-Mukhtār (Ta’rīkh al-fattāsh) (Houdas and Delafosse,1964 [1913]; Houdas 1898, 1900,
Hunwick 1999). Paolo de Moraes Farias’ compelling argument for these texts as ‘an exercise
in catastrophe management’ by careful and accomplished ‘text craftsmen and ideological
agents’ (De Moraes Farias 2003: lxx-lxxii) might be applied, at least to some extent, to their
narrative of catastrophe itself as well as to the means they employ to cope with it; such
visions of disaster are after all not uncommon elsewhere too, and frequently overlook other
dimensions of the changes taking place.
21
but the distribution of opportunities for regional initiative, and the level of regional autonomy
after the initial effect of regional disintegration, was thereby correspondingly greater.
Undoubtedly, the emergence of locally competing states out of the areas and populations
previously federated by Songhay implied the generalisation of warfare and the accompanying
region-wide increase in the importance and volume of enslavement and the slave economy.
Climatic change made its impact felt as well (J. L. A. Webb 1995). But the dynamics were
internally-generated, and when Europeans in turn began to arrive on the scene, they were
more ‘drawn in’ to these African dynamics than they were simply acting upon them from on
high and afar; indeed this pattern would continue well into the more conventionally colonial
period in the nineteenth century, in the northern Sahara as in West Africa (Clancy-Smith
1994, Brower 2009; Searing 2002).
At some point, however—or rather, at a variety of points along the relatively long and
very uneven chronology of colonial state-formation, from the 1880s through the 1930s—the
ability of Saharans themselves to exercise such autonomy over the regulation of their political
and ecological frontiers, the social and economic borderlands drawn around them, and the
larger zone of connectivity encompassing them, was rapidly decreased. Intensified regional
inter-state competition, from the series of Sahelian jihads of the 1670s through the 1860s,
then the expansion of more assertive and coercive French control into and beyond the Niger
bend and southward across the desert from Algeria and Morocco, culminating in the
establishment of the region’s independent nation states, segmented and bordered the
intensively interdependent spaces of the Sahara as never before. It was not simply that, on a
global scale, the railway turned trade and people away from the inward-facing shores of the
desert to the outward-oriented world market and the ‘gatekeeper’ states that controlled access
to it (Austen 2010, Cooper 2002). Habitation, movement and exchange within and across the
desert hardly ceased; indeed, with the advent of motorised transport it began to be able to
22
expand in ways previously unimagined. But while the Sahara had always been extraverted
within its own relations of connectivity, now, instead of being turned out from the inside, it
was newly subordinated from the outside, its people relegated to frontier outbacks and no
longer able to dictate the terms of exchange, of mobility, or of alliance with—now, recast as
incorporation into—polities centred elsewhere. The Sahara would remain a global
borderland, a crossroads region of transcontinental traffic and transition (Marfaing, Brachet,
this volume). But, with their own internal frontiers now overlaid by and (however unevenly)
subject to limitation by the political boundary-making of external agencies, its people would
not recover the autonomy of control they had long held over access, movement, and exchange
within their own distinctively delimited world (Oussedik, this volume).
Towards an opening: African frontiers, and the frontiers of Africa
This loss of autonomy, of course, has played out differently for different people within the
intensely hierarchical and often literally confining social relations governed by the older
forms of politico-ecological regulation that we have described. Haratin who can leave the
oasis on a truck and go find work in the expanding city cannot be aligned with imajeghen
who abandon their unsustainable herds and settle by the river. But we ought not, perhaps, to
idealise the ‘freeing up’ of individuals from older social systems any more than we should
romanticise the ‘authentically’ indigenous ‘freedom of the nomad’ (and his relative
benevolence to his slaves); the point is rather that the outcomes of the process described
above have been intensely polarising and frequently paradoxical. The normative status of the
political frontier carries aspirations to freedom and self-determination for Sahrawis just as it
presently confines them in the refugee camps of Tindouf. Tuareg may find themselves
stripped of much that once denoted their status and autonomy, indeed their very culture may
be becoming something alienated from them as it is commodified for the world market and
23
the tourist trade, but in those processes young ishumar and impoverished imajeghen, artisans
and world music performers can make their own niches in the transregional and global
division of labour (Davis 1999).
In place of Kopytoff’s frontier, replicating coherent cultural wholes, then, the history
of Africa’s Saharan borderlands, since the sixteenth century at least, is one of constant
fragmentation and the reinvention of new norms—from jihad states through national selfdetermination to indigenous rights—through which, under pressure from the outside on
which the region has always depended, Saharans have tried to retain control over the terms of
their own existence. Conquest and incorporation here has usually meant subordination and
exclusion, and the intensification of an extraversion that was always unavoidable, but could
once be less unequally managed. Attempts to reassert a degree of autonomy now, of course,
must rely on the normative languages of a wider world, whether lobbying for the Western
Sahara or asserting Tuareg cultural autonomy and land rights over mineral resources. The
‘unbroken’ history of the Sahara is visible in the degree to which, rather than being simply a
corridor opened to the outside and then closed again, the region remains today a turntable as
well as a contested border of exchange and migration. As elsewhere, though, such continuity
is traced over a pattern of significant change and dislocation. The Sahara of the early twentyfirst century is part of wider connected networks of commodities—cigarettes, guns, labour,
livestock, narcotics and hydrocarbons—and of culture, both produced and consumed: Arabic
satellite TV beamed into the region and the poetry, music, and artisanal manufactures coming
out of it. The region’s longstanding net extraversion is radically increased, concomitant with
an internal loss of control over the terms of exchange and even of mobility and access
(Grémont, Oussedik, this volume). But there also remain, again as elsewhere in the global
South, avenues of ingenuity and opportunity, through which people retain their ability to
manipulate and exploit the frontiers imposed upon them (Cordell 1985, Scheele, this
24
volume). While new boundaries as well as new solidarities are created to preserve or limit
access to new opportunities in the state or the city (Marfaing, Choplin, this volume), as well
as to enhance regional states’ and outside agencies’ capacity for control, Saharans are no
strangers to inventing the means of mobility to circumvent, or at least to optimise, the
limitations of their environment.
25