Cover Image
close this bookAfrican Agriculture: The Critical Choices (UNU, 1990, 227 pages)
close this folder5. Mauritania: Nomadism and peripheral capital
View the document(introductory text...)
View the documentPastoral production
View the documentIntegration of the pastoral world into the market economy
View the documentEvolution of the social and political framework of nomadism
View the documentConclusion

(introductory text...)

Abdel Wedoud Ould Cheikh

North-western Africa, known today as the Sahel, has always been an area particularly suited to animal husbandry From the 1,000 head of zebu that Askia Ishaq II used as a protective cover ahead of his troops in March 1591 (Kati 1964. 264) when the Songhai Empire collapsed under the blows of the men of Djouder, to the vast herds all over this area today, there is ample evidence of enormous wealth in livestock of the countries of the Sahel. Before 1972 this was represented by a total of 21 million head of cattle (Gallais 1977, 268).

It was not this wealth of livestock that attracted the Moroccan conquerors. Nor does it appear to have motivated more recent onslaughts, notably the French colonial occupation. But, inevitably, the organization of the pastoral societies of the Sahel, largely centred on cattle and their resources, was profoundly and permanently affected by this occupation. Other factors, such as the recent drought of the 1970s, have contributed to a dangerous acceleration of the process of disarticulation of the Sahelian pastoral systems, but the major factor in the evolution of these systems remains their marginal integration into a monetary economy centred in towns, which themselves are experiencing accelerated and disorganized growth. For Mauritania, which is examined here in order to provide an illustration of recent transformations of Sahelian pastoralism, two figures are enough to indicate the scale of changes that have occurred in recent years: whereas in 1965, nomads made up some 65% of the Mauritanian population, by 1976, they represented only 36%. Over the same period, the urban population grew threefold, rising from 90.000 in 1961 to 300,000 in 1977.

Leaving aside the terminological problems - the relationship between 'pastoralism', 'nomadism' and 'semi-nomadism' (Salzman 1980)- this chapter first presents a short outline of the factors of pastoral production, and then examines the forms and effects of the integration of pastoral society into an economy dominated by commodity relations. Finally, the social and political aspects of this integration will be considered in order to attempt clarification of the specific forms taken by the contradictions - of clans, groups, classes within a pastoral society undergoing massive upheavals

Pastoral production

The original complex constituted by the symbiosis of agriculture and animal husbandry in the prehistory of human societies (Leroi-Gourhan, I, 227-37) developed and survived at the cost of a specialization often involving violence. This complex and the more or less voluntary association between agriculturalists and herders constituted a permanent feature of the organization of Sahelian societies, and especially of Moorish society.

The environment and its resources

The key features of the Mauritanian climate are high aridity and extreme variations of temperature. It is dominated by dry winds (the maritime alize from the Azores anticyclone and the continental alize, the Irivi in the Moorish dialect) often laden with sand.

In July-August the monsoon winds from the high pressure zones in the South Atlantic provide the bulk of the rainfall, which barely exceeds 600mm in the best watered regions of the country (the far south). Further north rainfall declines to less than 100mm above a Nouakchott-Atar-Oualata line and to less than 50mm along the coast in the far north.

The influence of the maritime and continental alizes and the monsoon winds, combined with distance from the ocean, makes it possible to distinguish two broad climatic zones, each with a coastal and a continental aspect: the Sahara and the Sahel. North of Nouakchott, the coast, which has constant humidity, low rainfall and relatively low temperatures in winter, is a 'tropical coastal desert'.

The Saharan climate properly so-called, covering the vast bulk of the country, is marked by large variations in temperature, low rainfall and high evaporation.

The northern limit of Sahelian Mauritania is usually placed along the 150mm isohyet. Temperature variations in the coastal parts of this Sahelian zone are small and, on average, temperatures are lower than in the continental Sahelian climate.

Naturally, the types of vegetation and grazing land vary, depending on the climatic situation. The plant cycle also depends on the latitude and the nature of the soils. From north to south we can distinguish:

1) A vast desert zone stretching over mineral-poor soils, with sparse vegetation generally concentrated in areas with many streams (steep slopes in the mountains, oued courses and so on). Tree cover consists mainly of Acacia Raddiana and herbaceous plants, mainly of hardy graminaceae (Stipagrostis Pungens), making it possible, apart from short periods when the very irregular rainfall produces fresh pasture, to engage in animal husbandry (camels, sheep, goats) that sometimes requires movement over only short distances.

Apart from these areas of greater or lesser concentrations of plants (in the Saharan context, of course!) the sporadic grazing areas in the Saharan regions and notably the salty grasses (Cornulaca Monacantha) constitute excellent winter pastures for camels, virtually the only animals able to get any nourishment from them, given the distances to he travelled and the scarcity of watering points (in cool weather, camels feeding on these grasses can survive without drinking for over two months).1

2) A Sahelian zone with pastures renewed each year by relatively regular rainfall. The vegetation becomes denser and more varied as one moves south. Alongside a scattering of large trees (Adansonia Digitata and Combretum Glutinosum) that herald the Sudanic savanna, more or less dense copses of various varieties of acacia (Acacia Senegal; Acacia Flava: Acacia Nilotica, among others) as well as trees or shrubs (notably Commiphora Africana and Zizyphus Mauritania) of lesser forage value, dominate a bushy vegetation in which the typical Sahelian graminacea (Cenchrus Biflorus) occupies a preponderant place in the rainy season. This zone is mainly a cattle area, but there are sheep and goats too. Dromedaries, on the other hand, can stay there only in the dry season: in the wet season there is an abundance of tsetse flies carrying the very dangerous camel sleeping-sickness (tàburit).

Mauritanian livestock

The adaptation of Sahelian animal species to these harsh, natural conditions, the result of a long process of selection, has been stressed by several observers (Charles Toupet 1975. 227 and 29). The mediocre results so far obtained by attempts to acclimatize exotic species or by the few attempts at cross-breeding them with indigenous breeds, suggest that the Sahelian-Saharan species, with their qualities of resistance and moderate food and water needs, are not about to be replaced by new breeds, despite their low yields.

Goats, sheep, camels and cattle are the predominant animals reared in these areas. The so-called 'Sahelian' goats are the most widespread. These are of varying colours, high on the hoof and quite light when they reach maturity (averaging between 15-20 and 40 kg): females produce 70 litres of milk annually for a lactation period of 120 days, and adults are estimated to produce 10-15 kg of meat.

There are two kinds of Mauritanian sheep: the 'Fulani' sheep, quite large, with a short, white or black and white coat, and an adult weight of 30 to 50 kg, producing up to 30 kg of meat, excellent for human consumption: and the smaller, longer hatred 'Moorish' sheep, which provides the main raw material of nomad tents- and produces more milk than the Fulani: 1.5 to 2 litres per day in the rainy season.

All the camels raised in Mauritania are the single-humped Camelus Dromedarius. Used for transport, and for milk production, the camel is also used as a draught animal and as a source of meat for human consumption. Its hair, especially that of young animals, provides a valuable supplementary raw material for making Moorish tents. The adults weigh on average 450 to 550 kg, and can provide 150 kg of meat. Females produce an estimated average of 400 litres of milk per head per year for a lactation period of 270-360 days.

The cattle in Mauritania - Bos ludicus - are distinguished es 'Moorish' zebu and 'Fulani' zebu. The Moorish variety, with an average adult weight of between 320 and 360 kg, provides 500 litres of milk annually for a lactation period of 180-200 days. The heavier Fulani zebu (up to 400 kg on average in adulthood) produces less milk (some 300 litres for a lactation period of 180-200 days). Average meat production from adult males can reach 150-250 kg.

The low productivity of Sahelian livestock reflected in these figures is somewhat compensated for by the large size of flocks and herds that some observers consider to be excessive compared to the meagre resources available. The load capacity of Sahelian pastures will be assessed later.

There are great disparities in the quantity of livestock owned by individuals and families. These disparities confirm a hierarchical stratification marked by a longstanding system of pseudo-castes which has been considerably accentuated by recent imbalances consequent upon the urban commercial sector's domination of the pastoral economy that has developed in the wake of the colonial and post-colonial administration.

Particularly since 1968 (the beginning of the recent wave of droughts in the Sahel) as a result of the steep fall in cattle prices (in 1968 a milch cow was selling at Francs CFA I.500 at Boutilimit, whereas two years earlier it had been worth 20 or 25.000) there has been a massive transfer of cattle from traditional herders to traders and bureaucrats in the towns. This development and its economic and social significance will be discussed later.

Regarding the average size of traditional-type family herds, no established estimates exist, but from a limited empirical knowledge of Mauritanian and especially Moorish pastoralism it can be said that for large ruminants, herds bigger than 100 head are exceptional. For many families of herders, livestock resources are limited to a few dozen sheep and goats, and rarely to more than 20 head of cattle or camels.

Charles Toupet, citing pre-drought estimates, suggests for 1968. 1.9 cattle per inhabitant for a total Mauritanian population then estimated at 1.091,500 (Toupet 1975. 240). The author stresses that compared to the FAO's figure (0.31) at the same time for the whole planet, this is very high. The annual growth rate of Mauritanian livestock was recently estimated at 8% for a population growing annually by some 2.5%, which testifies to the devastating effects of the drought in the Sahelian region since the beginning of the 1970s.

Table 5.1

Livestock numbers 1969-80

('000)


1969

1970

1971

1972

1973

1974

1975

1980

Cattle

2,000

1,850

1,550

1,500

1,115

1,150

1,135

1,400

Sheep and goats

7,000

6,750

6,500

6.500

5,850

6,300

6,800

6,500

Camels

710

705

700

670

680

680

685

750

More significant for the management of family flocks and herds, particularly for a proper assessment based on what some observers consider to be an excessive accumulation of redundant and useless livestock (for example, a large number of young males and old animals) would have been a precise assessment of the average composition of flocks and herds.

The scattered figures that can be found, which must be accepted warily - fear of the devil eye', and of heavy taxes, mean that statements made by herders are not entirely trustworthy - and which relate almost exclusively to cattle rearing, are, due to their very imprecision, difficult to interpret. A few are, however, given below.

In the early 1960s, the following distribution was suggested for the cattle herds in the Moudjéria region (central Sahelian Mauritania) made up on average of 120 head (Toupet, op, cit.. 249): 5 bulls; 10 bullocks: 67 cows and heifers: 38 calves of both sexes.

The livestock department's official estimates for the same period suggest a much higher proportion of males: males 4 years old and above. 8%; females 4 years old and above. 38.5%: males aged I to 3 years. 18%: male calves. 7.5%: female calves. 7.5%.

Following are more recent figures for the Mauritanian region of Tagant, the same region that was the focus of Toupet's surveys (Grosser and Ba. 1980. 30). Bearing females: sheep and goats 40%: cattle 40% camels 30%. Adult males: sheep and goats 10%: cattle 10%; camels 20%. For young growing animals, 50% in each category: and for numbers of births per female per annum: sheep and goats. 2%; cattle 0.6%; and camels 0.4%.

In order to explain the 'rationality' or 'irrationality' of the economic behaviour of Sahelian herders these figures must be put in the overall socioeconomic context. Changes in this context since colonization and the gradual integration of Sahelian pastoral societies into a dependent and dominated market economy will be discussed later.

Rather than any individual Sahelian herder's desire to accumulate cattle, the reasons for the growth in numbers of livestock supported with increasing difficulty by the poor pastures of the Sahelian region must also be examined in the socio-economic context. The pre-colonial pastoral system's relative functionality is generally recognized today, involving as it did building up a stock for food and trading purposes adapted to environmental conditions and periodically readjusted to these conditions by razzias and natural disasters. But it did not withstand the growing grip of commodity relations, which the economic crises of capitalism and local climatic crises have helped accelerate. The emergence of a market for cattle, and the enhanced dependence that the development of an essentially unequal exchange signified for herders, were processes contemporary with the transformation of political and social conditions (notably replacement of the disorganized razzia violence by the 'rational' violence of what was called 'pacification') that affected political control of the pastoral space. These factors, combined with the establishment of an embryonic modern health and water infrastructure, helped increase demands on the grazing areas of the Sudan and the Sahel.

Examination of the 1938 figures - even judged by the colonial administrator Beyriès, who reported them, to be one-third short of reality - shows, in comparison to Table 5, 1, how big this increased burden was: in that year, less than five years after the final stabilization of colonial rule, the Governor of Mauritania's annual report (the Hodh circle was not yet part) gave the following figures for the whole country: 75.871 camels: 212.175 cattle; 1,713,631 sheep and goats.2

Compared to current estimates of the forage resources of Mauritania, figures in Table 5.1 show the first phase of a saturation process which, according to some forecasts, will become total and effective by the year 2000. In fact according to a recent study for the Mauritanian Ministry of the Economy and Finance (RAMS. 'Livestock subsector study'), the load capacity of the 55 million hectares in Mauritania that can be grazed (one unit of tropical cattle requiring between 4 and 70 hectares depending on the state of the biomass) will, if current trends continue, be reached by the year 2000.

This type of assessment must be treated with caution since, while the volume of cattle/useful area ratio is convenient to give a rough assessment, it glosses over many parameters whose complex interaction can alone provide the bases for a well-founded forecast. Independent of the strictly geographical factors, social and technical constraints decisively alter this ratio, which is also heavily dependent on the forms of state intervention (forage policy, reserve policy, and so on) and its regional and sectoral economic choices (Gallais. 1979, 121).

In the space available here it is not possible to survey all of these factors. The changes in the system of pastoral production, the various techniques of acquiring and processing livestock products, all of which directly affect herders' consumption and incomes,3 are of relevance, but here only one crucial aspect for the future of pastoralism will be highlighted. This is mobility, which constitutes both the distinguishing feature of pastoral life and a permanent means of adjusting human and animal occupation to the precarious and unevenly distributed resources of the natural environment of the Sahel and the Sahara.

Mobility and herding

Aside from its cultural significance as a focal element of pastoral civilization the nomadics' life-style aims, above all, to procure access to basically rare and precarious resources. It is the decisive role of mobility, that constitutes both the form and the major means of subsistence of the pastoral community.

Nomadic movements show a degree of permanence in to where, how, and when they take place. These movements are tightly conditioned by the seasonal character of the rainfall, although the distance covered by herders annually has markedly decreased, resecting the weakening of large- in favour of small-scale nomadism, a tendency that is often the prelude to sedentarization.

Mauritanian and more generally Sahelian-Saharan nomads move back-and-forth from north to south, following the annual rainfall pattern. From the first tornadoes over the southern regions of the country (June-July) a slow movement northwards gets underway and continues until the end of the cold season. A movement in the opposite direction then begins, taking large numbers of nomads as far as the banks of the river Senegal and the Niger bend by the end of the dry season.

In contrast to this regular north-south movement, a second, much more diffuse and irregular form of nomadic movement corresponds to the development of exceptional grazing areas in the beds of oueds or in the gràyar (areas into which water runs off mountainous regions). This form, typical of the Saharan part of the country, generally involves small-scale movements.

The annual range of north-south migration, which varies for both full and semi-nomads, depending on the volume and distribution of rainfall, reaches its maximum among camel-keeping nomads who may cover over 1.000 km annually (UNESCO, 1965). For example, the Hmu dnnãt of the Dhar of Oualata spend the winter around Agweylil Nmädi, over 200 km north-east of Oualata on the saline hãt pastures, often several days from the nearest watering point; their dry season and early winter stay is generally on the Tãgurarut cliffs, near to the well of the same name some 260 km south-eat of Agweylil. In winter, the camels are sometimes left to roam freely, moving much further north than their owners who meet up with them at the beginning of the hot season at watering points where they are accustomed to drink. Occasionally some cattle thief- rustling is still quite common in this region- upsets this almost automatic mechanism.

In bad years, such as have been the case since 1968, much greater distances are covered. In March 1980, at the watering-place at 'Weynãt r-razzat, about 10 km south-east of Néma, we met a young hmu nni shepherd (from the hmu nnãt, a D-dlãkne 'fraction'), who had come down with his family from the remote regions of the dhar where they had spent the winter and who, when we met them, had already travelled 400 km southwards. Given that there had been almost no rain in the Oualata region in 1979, they had only a very vague idea of how far south they would go. Perhaps, they said, as far as Ras el Me (lake Faguibine, in Mali). Their quest for grazing lands in the south would then have taken them over 800 km, which they would have to retrace in the opposite direction as soon as the rains started in order to save the camels from contracting taburit (trypanosomiasis).

The movement of semi-nomadic cattle herders and shepherds has the same seasonal characteristics as that of camel herders (acceleration in the rainy season, slowing down in the dry season) it generally extends, barring a major climatic disaster, over much smaller areas.

For example, in January 1980, we encountered a camp of S-sxaymàt cattle herders 75 km south-east of Magta Lajar. Usually, when rainfall was more or less normal, they moved throughout the year between el Wàd Lebya in the south-east and Wàd Leyrdi in the north-west. When we met them, as in the previous year, they were just beginning a journey southwards, due to the catastrophic failure of the rains in 1979. Only two men and a paid Fulani shepherd were accompanying the herd, on a journey that would take them 300 km to the Selibaby region. These were full nomads whose sole activity was herding and who, in order to save what was almost their only resource, were capable of a great burst of large-scale nomadism. With the increase of agriculture in their activities the capacity to undertake this sort of move is becoming lost.

As some Idegg molle agriculturalists explained, since the construction of the Magta La jar dam (1,400 hectares flooded) in the late 1940s they have tended to settle alongside the fields, thus creating a large village with 3.821 inhabitants by 1977. 'In the beginning,' a notable belonging to this group told us in January 1980, 'we used to send a lot of people with the herds and very few to the fields, but today almost everybody is farming and only a few people are sent with the animals.' The correlation between nomads' practice of agriculture and the increasingly lesser distance they cover each year is obvious.

Recent data from the 'Provisional Results of the General Population Census' conducted in January 1977 by the Mauritanian Ministry of Planning show that almost 30% of nomad households practice agriculture, with a low percentage in the north and far east (regions of large-scale camel-herding nomadism) and a much higher percentage in the southern regions.

A high percentage can be observed in the Adrar (46.1 %) and Tagant (32%) despite the relative aridity of these regions. This is linked to the presence of a large number of palm groves which, without any loss of pride, the nomads can cultivate freely (palm growing is more 'noble' than working in the fields, which is generally poorly regarded and even despised).

The 1977 census figures show a relatively small proportion of large-scale nomads among the non-sedentary population. Only 17% of those counted make annual moves of more than 200 km; these are mainly in the camel-herding regions of the far north-west and south-east of the country.

The importance of semi-nomadism and transhumance compared to long-distance nomadism indicated in Table 5.2 is only the most striking manifestation of an erosion affecting every aspect of Mauritanian pastoralism, principally the conditions and forms of mobility, the key feature of pastoral life.

It has been noted that the relative abundance of cattle, periodically controlled before colonization by natural disasters and razzias, might appear to be a compensation for the mediocre productivity of these Sahelian-Saharan breeds. The rapid growth in numbers, especially from the 1950s.4 in which some observers claim to see a process of saturation beginning, and as both cause and effect of desertification, the consequences of which became particularly dramatic after 1968, resulted in fact from the complex interaction of numerous factors in which the hegemonic extension of commodity relations played a central role. Mobility itself, the key factor of pastoral production and the reproduction of pastoral society, was profoundly affected by changes that gradually increased semi-nomadism and transhumance compared to large-scale nomadism, and often led to nomads settling in rural villages or on the outskirts of the new urban agglomerations.

Integration of the pastoral world into the market economy

It scarcely needs stressing that there are internal sociological, cultural and economic reasons for these changes and the vulnerability of Sahelian pastoral societies, enclosed as they had been for several centuries in the poverty and monotony of a particularly Spartan way of life. It is equally obvious that these changes and their recent manifestations were related to the effects of colonization and the unification of a 'world-economy', to use the terms of Wallerstein and Braudel, centred on the capitalist West.

Table 5.2

Distances travelled by region (%) and camp sizes (numbers)

Region

Households present

Households in permanent camps

Households moving over 200 km

Average camp size (numbers of persons)




Total

Those moving


Eastern Hodh

54.6

0.6

27.2

27.4

25.6

Western Hodh

51.6

6.4

31.3

33.4

21.8

Assaba

33.5

15.8

6

7.1

25

Gorgol & Guidimakha

11.1

11.3

4.3

4.9

30.2

Brakna

33.7

4.5

7.4

7.8

26.5

Trarza

49.3

3.2

6.4

6.6

12.8

Adrar

32.1

0.6

14.2

14.3

14.1

Tagant

57.6

2.9

23.2

23.9

25.3

Nouadhibou & Inchiri and the two Tirises El Gharbia

12.4

1.8

70

71.3

11.9

National total

33

29.3

16.7

17.6

21.4

Note: There seems also to be some correlation between the scale of annual migration and the size of the nomad units even if the gaps shown by this latter variable in all regions remain rather small: in the north-west where the proportion of nomads annually moving over 200 km is 71.3% the camps have an average of 11.9 persons, or more or less two-tent households (average size of nomad households: 4.84), whereas in Gorgol-Guidimakha where the lowest regional proportion of large-scale nomads is recorded (4.9%) the size of camps rises to an average of 30.2 persons. Or six-to-seven-tent households.

The major consequence of colonization - and possibly its essential vocation - was the creation and development of a dominant commodity sector in the pastoral societies of the Sahel and the Sahara whose production, organization and values had been hitherto based upon pre- or non-capitalist structures even though commodity exchanges, including with European traders, frequently occupied a not insignificant place within these societies.

The various historical stages of the establishment of the hegemony of commodity and monetary economy in Mauritania will not be detailed here, but simply a brief overview to clarify the present situation to be followed by an examination of that situation.

The establishment of a market in cattle

Commercial activities have always represented a significant proportion of the economic life of the western Sahara regions which, since the Middle Ages, have been crossed by caravans taking the products of the Sudan and the Sahel (gold, slaves, ivory, even cereals) northwards in exchange for the products of the Sahara and the Maghreb (salt, metals, weapons, fabrics and so forth). The role of gum Arabic in the Moorish world's external trade from the eighteenth century onwards is also well known, as are the effects of this 'long-distance trade' on the economic, social and political structures of the groups most immediately in contact with it, the emirates of Trarza and Brakna, 'controlling' the 'trading posts' of the river Senegal in particular (Hamès 1977).

But neither trans-Saharan nor river trade (on the Senegal) or the Atlantic trade led directly to the widespread establishment of a dominant commodity and monetary sphere within Moorish pastoral society. Not until the completion of colonization (1902-34) was there substantial progress in the incorporation of the pastoral economy into the market economy. This progress can be seen particularly in the establishment and development of a market in cattle.

Here some results of Pierre Bonte's research on this topic are summarized (Bonte 1981). He writes: 'The formation of a market in cattle was the essential and almost immediate - consequence of colonization.'

Leaving aside France's political and military interest in securing control of Mauritania, that territory was, at the beginning of the twentieth century, seen as a food crop and commercial appendage to the economic development represented for the regional colonial authorities by the groundnut zone of neighbouring Senegal.

Without neglecting the development - or rather the exploitation - of Mauritania. [they] gave it as a result two main functions: 1) to supply the labour force for the rural and urban sectors of groundnut production (that involved essentially the Black Africans along the Senegal valley who migrated southwards in large numbers): 2) to supply cheap food and especially meat to the groundnut producers and to the wage-earners in Senegalese towns. (Bonte, 1981)

It was this second function that affected the evolution of pastoralism more directly since it contributed to the rapid emergence of a market in cattle that more specifically concerns us here. Two main reasons explain the rapidity with which this market was established.

The first relates to the central role of animal ownership among the nomads' resources. They turned increasingly to livestock husbandry to meet the needs for cash that colonization helped shape and extend (role of colonial 'security', markets, improved communications, consumption patterns diffused, however modestly, by the colonial schools and those close to the 'commandant', in short what Hamid El Mauritanyi (1975) called the 'colonial boyarchy'). The second reason seems to be directly linked to the accelerated growth of urban demand, arising particularly from the large Senegalese agglomerations, which experienced continuous demographic growth. By the beginning of the 1920s, the market at Louga was channelling the hulk of Mauritanian cattle exports on the hoof far ahead of the market at Goulimine (southern Morocco) where camels were the main, if not only, item of trade, cattle being unable to go into the desert regions of northern Mauritania.

But 'economic' reasons alone are not responsible for the increase in Moorish cattle sales during the first 30 years of the twentieth century: Pierre Bonte rightly stresses the role of pressures from the colonial administration. These pressures might take the form of commandeering animals for human consumption, paid for (when they were) at the official rate. As early as 1909, the 1.000-man-strong 'Guiraud column', about to conquer the Adrar, collected a heavy levy in cattle from those regions where an economic and military lesson was seen as necessary to help staunch any spirit of resistance. Colonel Guiraud was simply extending, in very disturbed circumstances, an administrative razzia that had been institutionalized in the conquered area of southern Mauritania after 1902. Despite the heavy burdens for the nomads, especially in bad years, such requisitions, open or covert, continued virtually until independence in 1960, and even beyond - as a tax.

In 1926, in the Adrar, the requisitions were for 1.500 camels (out of 9,000). The administrator himself noted that this burden was too heavy and was contributing to the famine then raging in the Adrar. Each fraction in turn had to provide transport animals (one-fifth of animals levied), at least 55 camels per month for administrative transport; these animals had to be maintained near the posts, even if there were no grazing areas! In addition, they had to provide animals for the annual transport of supplies to the Adrar from Rosso and Podor. (Bonte 1981. 8)

As the free market developed, the requisitions, whose unpopularity is easy to imagine, tended to diminish.

The small-scale development of motorized transport saw the first regular commercial links between Atar and Rosso made by lorries belonging to the Lacombe company in 1935. On this route. Lacombe's 35 lorries had a virtual monopoly and transported 2,770 metric tons of freight and 4.782 passengers in 1950, as against 2.500 metric tons of freight and 3.990 passengers in 1947 (Brechignac. 1952).

During the Second World War, however, there was a temporary return to camel transport, both for the needs of internal Mauritanian traffic for which requisitions reached 8.000 camels per year, and for transport to Senegal which required 11,000 camels. In addition to the administration's requisitions there was the cattle tax, opportunistically named zakat (Muslim religious tithe) by the colonial authorities.

This tax was levied on a lump sum basis for each tribe or fraction of a tribe. Tribal chiefs, sometimes accompanied by guards, collected zakat and received a percentage of what they collected as payment; an arrangement that gave rise to countless abuses. For the most prosperous herders payment was relatively easy (helped by a great deal of under-declaration), but it was a heavy burden on the livestock resources of smallholders, especially in had years. Nevertheless, the combination of administrative 'incentives' and economic motivation resulted in bringing growing quantities of stock on to local and foreign markets.

The figures available, quoted below, cannot claim to be other than notional. In 1940, when the effects of the long crisis of 1933-36, which had contributed to a sharp fall in cattle prices, were beginning to diminish and as an era of lasting market disturbances associated with the War was about to begin, the archives provide the following figures for exports of Mauritanian livestock to the Senegal markets (Saint-Louis, Louga and Dakar): 9,723 camels, 9,853 cattle. 126,765 goats and sheep.

It is equally difficult to estimate the recent development of these exports. The lack of strict border control, of controllable marketing structures and administrative means of control, and the fluctuations of official policies for livestock exports, are factors that explain the inevitable imprecision marking any effort to quantify recent exports.

As regards official policies, a belated realization of the danger to national animal resources in the continuation of massive exports of cattle on the hoof began to emerge in the late 1960s. Thus, in 1969, a Mauritanian meat marketing company (COVIMA) was set up, with the Mauritanian state taking a major stake in it. The COVIMA, thanks to the cold stores at Kaedi (3.000 metric tons of frozen meat per annum), was the first step towards replacing exports of cattle on the hoof by marketing meat, at a time when annual sales outside Mauritania were estimated (for 1968) at 52.000 cattle and 330,000 sheep and goats. Financial difficulties, partly associated with transport problems and - as this was the beginning of the drought that has been affecting the Sahel for over ten years - the poor quality of the meat it had been hoped to sell on the open markets in the Canaries and Libya, rapidly immobilized the company. In 1975 it was completely nationalized, becoming the société Nationale pour l'Industrialisation et la Commercialisation du Bétail, SONICOB, but this did little to improve its situation, and today, it works- at a loss- simply to meet the local needs of Kaedi.

Placing a quota on exports of cattle on the hoof, instituted as a counterpart to the measures aiming to gradually replace them with the export of meat, has since been abandoned, in the framework of ECOWAS agreements. Quite clearly, exporters had not waited for these measures to end before resuming and intensifying a trade encouraged by price differentials with neighbouring countries and doubtless, indirectly, by the creation, in June 1973, of an inconvertible Mauritanian national currency (the ouguiya) that further accentuated and bureaucratized the monopoly of import-export activities in the hands of a handful of Nouackchott middlemen. The Mauritanian cattle export networks that were the source of leading Moorish fortunes thus continue, on the back of a movement largely begun in the colonial period, to fuel the ever-expanding regional markets.

The creation of a meat marketing company was also designed to act as a means to regulate a domestic market rendered particularly unstable by regional geographical disparities (variations in rainfall), considerable distances between the producing areas (especially the south-east) and the importing towns along the Atlantic coast (Nouakchott, Nouadhibou, the mining region) and finally the inadequacy of means of transport. Before dealing with meat consumption and how it has evolved, it is necessary to look at commodity circulation for cattle on the hoof within the Mauritanian market.

As already stressed above, the distribution of Mauritanian livestock resources, and the circulation of these resources before colonization was carried on mainly in a non-monetary framework. The global significance of livestock and its products and circulation will be dealt with later, as part of social (family, hierarchical, clientage) relations within the Moorish pastoral community.

The role of razzias in the circulation of Mauritanian cattle in the pre-colonial period is well known. Usually, much less stress is put on the role of other forms of more 'voluntary' distribution - which indeed were more or less associated with the permanent insecurity engendered by the institutionalization of the razzia - such as the mniha. This loan on the usufruct- manaha means supplying the products of an animal lent - offered the double advantage of consolidating kinship and clientage relations while ensuring a prudent dispersal of the herds which were never safe from some unforeseeable occurrence or some enzootic disease.

Alongside these forms of circulation, plus gifts associated with matrimony (before colonization dowries were predominantly settled in cattle) or status (payment of tribute, for example), or particular ceremonial circumstances (births, deaths, festivals, marriages and so on) commodity exchanges in the internal circulation of Mauritanian livestock in the pre-colonial period were marginal.

The role of economic incentives (the rising monetary needs of herders) and administrative coercion in the establishment of a 'free' domestic market in cattle in Mauritania, has already been noted. While this market's needs, notably for meat for human consumption, remained relatively modest throughout the colonial period (urbanization did not take off until after Mauritania's independence in the early 1960s) the continuous growth of supply combined with a relative fall in the cattle prices, compared to imported products (sugar. Guinea fabrics, among others), gradually put traditional herders at the mercy of administrative or commercial middlemen, who themselves had relatively stable monetary incomes. There were sharp upsurges of this process during local climatic crises (1917,1942-43 and so on) and during economic or political crises in the dominant capitalist system, and well as when there was a catastrophic conjunction of the two es in 1940-43 or since 1970. On the basis of an index of 100 in 1940. Pierre Bonte (1981,20) calculated that the price of one kilo of millet officially reached 433 in 1942 (1,333 on the black market), on kilo of sugar from an index of 100 in 1939 increased to I,500 in 1942, the price of one metre of Guinea cloth (the Moorish people's traditional clothing) rose from 100 in 1939 to 737 in 1946, and a kilo of imported rice rose from an index of 100 in 1939 to 2.533 in 1942. Over that period livestock prices tended to stagnate and not until 1948 did these price curves begin a significant recovery.

Recently, in the areas most affected as a consequence of the cycle of drought that began in the late 1960s, livestock prices have fallen sharply: in some localities of Trarza, in 1968, impoverished nomads who had been forced to settle were selling milch cows for 1.500 Francs CFA; two years earlier the selling price would have been 20.000.

But conditions in periods of crisis cannot form a basis for deriving conclusions about some unilinear or continual deterioration of the trade terms between the Mauritanian pastoral economy and the commodity system of world capitalism. A comparison of prices of local products (notably of livestock) and imported goods, from the beginning of colonization to the present day, shows a succession of movements that are far from uniformly unfavourable to herders. But the weakening of the herders' position as they became integrated into the commodity economy cannot be measured solely by the relative market prices between their products and imported ones. This is only one element in an overall situation marked by a global erosion that affects both the system of production and exchange and the cultural models and social relations that govern reproduction in nomad society.

The move of more and more livestock from traditional herders to bureaucrats and traders in the towns in the wake of crises - especially that which has continued since 1969 - is not due simply to a difference in incomes between the two groups but also to a whole range of different attitudes towards livestock and its upkeep.

An essential condition of this transfer lies in the 'Liberation' of the labour force (servile or tributary) from traditional leaders. This is a vital prerequisite to the extension of wage-earning that has become the dominant relation of production in the sector we have called elsewhere (Bonte and Ould Cheikh. 1981)'second herds'. After the damage to Mauritanian livestock caused by the drought' and given the apparent recovery of rainfall over the last three years, it is clear that the increase in stock maintained in this ('second herd') way, sold at the right time, can bring in considerable profits. Unfortunately, there are no statistics to facilitate an appreciation of the scale of this livestock transfer into the hands of the new (and not so new) rich in the towns which, in our opinion, is destined to play a key role in the future of Mauritanian pastoralism.

Another aspect of internal commodity circulation is linked to the local market's need for meat. Here, too, precise data are lacking, largely because most slaughtering is done privately, even in the large towns where wealthy families prefer to buy a whole sheep to be consumed as required, rather than to buy meat from butchers. It is also well known that traditional herders rarely eat animals they themselves have reared.

Compared to other Sahel countries. Mauritania's high level of meat consumption (according to some estimates) is probably a recent phenomenon. Estimates for annual per capita meat consumption according to the FAO's Annual Report on Production and Trade for 19776 are: Mauritania, 26.8 kg; Mali, 13.8 kg: Senegal. 14 kg. The reason for this increased meat consumption since the early 1960s is, above all, the development of urban areas, for example, in 1961 the inhabitants of Nouakchott consumed 70 kg of meat per head (Lacrouts et al.. 1962).

At this time the building of Nouakchott and other mining towns in the north, such as Zouérate and its port of Nouadhibou, were underway, and together (with Akjoujt after 1967) accounted for virtually all the modern sector jobs in the Mauritanian economy. Consequently they experienced extremely rapid population growth, sometimes to the detriment of older rural centres.

The big (in the Mauritanian context) towns in the west and north attract large quantities of cattle from the livestock regions of south-east Mauritania. Facilitated by the tarred road, which will soon link Néma and Nouakchott, this movement, especially for sheep and goats, will doubtless grow in size. Already, on the Nouakchott-Kiffa stretch, articulated lorries regularly carry cargoes of livestock to the capital for slaughter; some of them are then carried northward.

It is not possible to provide figures on the full extent of this trade or to accurately quantify Mauritanian domestic meat consumption, but, as an example, Table 5.3 shows the number of controlled slaughterings in Nouakchott in 1980.

Whatever the precise volume of meat consumption, the estimates, rough though they are, suggest that, given the rate of population growth in Mauritania and the limits on the load-bearing capacity of the grazing areas, it cannot be maintained at its present level over the next 20 years. A recent study Ministère de l'Economie et des Finances. RAMS Project) suggests that the Mauritanian authorities should immediately reduce the population's consumption of red meat in order to preserve the number of livestock still recovering from the depletion caused by the drought in the early 1970s. Moreover, the author of this advice assumed that the traditional system of livestock organization and management, extensive nomadic herding, would be maintained, it being considered that this was the form of land-use best adapted to the scarcity and fragility of natural resources in the Sahelian-Saharan environment.

Table 5.3

Livestock slaughtered in Nouakchott, by month (1980)

Month

Cattle

Sheep

Goats

Camels

Total

January

1,286

457

493

171

2,407

February

1,226

489

480

212

2,407

March

1,252

376

383

308

2,319

April

1,184

286

298

326

2,094

May

1,128

216

207

418

1,969

June

604

199

236

599

1.638

July

486

253

145

894

1,778

August

540

385

393

983

2,301

September

1.140

869

577

527

3.113

October

1,219

990

597

242

3,048

November

1,558

1,186

624

293

3,661

December

1.650

1,532

387

244

3.813

Total

13 273

7,238

4 820

5,217

30,548

Average

1,106

603

402

434

2,545

But this system, as has been stressed several times, was centred on spatial mobility and dispersion, and involved a set of technical, economic and ideological behaviour patterns: in short, it derived from a culture which, since the advent of colonial occupation and particularly since the upheavals engendered by the latest cycle of drought, has shown signs of ever-more widespread erosion. The transformations suffered by the Mauritanian nomadic way of life in relation to the advances of the dominant commodity sector can be seen particularly in an examination of the incomes and expenditure of nomad households.

Recent changes of nomads' incomes and expenditure

Surveys in which we took part were carried out in 1979 and 1980 in the framework of a project aimed at providing basic data and defining possible scenarios for the principal projects of the Fourth Mauritanian Economic and Social Development Plan, especially in the rural sector. Part of these surveys concerned the structures of the income and expenditure of rural people, including the nomads with whom we are particularly concerned here. For many reasons, both material and methodological considerations (for example, the small sample surveyed, shortness of time, lack of training of those involved) and the obstacles inherent in any survey among nomads (dispersion and mobility of the target populations, mistrust of the 'administration', fear of the 'evil eye', among others), the figures emerging from these surveys cannot be regarded as a faithful reflection of the socio-economic reality of nomad life. Even if they were shown to be absolutely accurate, they would still be a partial image of a nomad milieu heavily marked by community and hierarchical values that further limit the scope of the key concept -'family budget' or 'budgetary unit'- of the survey a few results of which follow.

For the sedentary rural population, monetary income was derived from: wages 26%: gifts 17%; trade 17%; herding 14%; agriculture 6%; pensions and family allowances 5%; fishing 2%; loans 4%; handicrafts 1%; other 8% Nomads' monetary income was derived from: herding 60%: handicrafts 7%; agriculture 2%; trade 7%; wages 4%; gifts 16%: loans 4%.

This research revealed how small is the annual monetary income of nomads compared to that of sedentary rural dwellers: rising to 9.280 MU (US$ 206) per capita for nomads, and to 13,494 MU (US$ 300) for sedentary rural dwellers. A nomad's annual monetary income thus represents only 65% of that of a settled rural dweller.

Herding continues to be the essential means of meeting nomads' domestic needs and the principal source (60%) of monetary income. Wage-earning (4% of monetary income), as indicated here, is less than had been previously supposed given the development of wage-earning in herding, nevertheless, it reflects a significant change in the traditional relations of production. While the stress given here to the transfer of livestock into the hands of pseudo-managers from the bureaucratic and commercial sector has some foundation, if this transfer is accompanied by an extension of wage-earning, the fact that the percentage of wage-earnings in herders' monetary income is so small might be due to a low monetary portion of the wage paid by owners to shepherds; the shepherds themselves are often small herders, preferring payment in kind in the framework of traditional-type contracts such as: usufruct of herd, right to a new-born calf when an animal gives birth to twins, right to one animal of a given age and sex each year fixed on the basis of the number of animals herded, for example.

Gifts in kind from relatives or clients resident in or working in towns (16% of herders' income in kind) testify to nomads' growing dependence on temporarily migrant or permanently settled kin. But it only partly reflects the scale of the rural-urban drift and of sedentarization, which over the years 1964-76 was responsible for a decrease in the proportion of nomads in the population from 64% to 36%.

The insignificant monetary profit accruing to herders from agriculture (2%) does not fully reflect the nomads' role in agricultural labour; a relatively high proportion (29.8% overall) of nomads engage in some agricultural activity.

This figure is the outcome of a change that became apparent from especially the late 1940s. The vertiginous rise in cereal prices (30 times higher in 1949 than in 1940!), the support of the colonial administration, which saw the development of agriculture as a means to make good the food deficit and facilitate the settlement of tax payments, and the gradual emancipation of the former slaves of Moorish herders, large numbers of whom settled in independent villages (àdwàba) and devoted themselves to agriculture, are all factors that explain the development of nomads' agricultural activities. The repercussions of these events on herders' mobility has already been underlined. Combined with other factors transforming the economic and social environment of nomadism they led to large-scale and disorderly sedentarization of nomads.

It is, however, hardly surprising that the current monetary contribution from this agricultural activity appears marginal in the herders' annual accounts. The best lands (courses of oueds, in particular) are still sometimes the object of violent disputes, associated with the demands of newly sedentarized groups, but they have long since been divided up. Nomads anxious to diversify their food resources are left with only the very unreliable product of rain-fed cropping, yielding barely 300 kg of millet per annum, thus to secure a marketable surplus requires exceptional circumstances. Harvests can, therefore, for the most part provide no more than a little extra food for subsistence. This is why the bulk of the expenditure of nomad budgetary units is devoted to food.

The 1980 RAMS survey puts the annual individual consumption level of nomads at 13,778 MU, distributed in the following proportions: cereals 115%; fruit and vegetables 6%; meat 11%: milk products 57%; tea and sugar 9%; other 2%.

The movement towards monetarization of nomad consumption continues to be very uneven (only 49% passes through the monetary circuit), in so far as regional disparities reflect the uneven penetration of commodity relations. This is itself a function of how close an area is to the centres from which these relations are diffused (the strip along the river Senegal and the Atlantic coast since the 18th century, the capital and the mining towns since 1960) or of the particular role of a regional community: in the Eastern Hodh, the most useful inland region of the country, and where pastoralism has remained most vigorous, only 17% of the nomad consumption involves monetary exchanges: while in Tagant, the homeland of the most active trading tribe in Mauritania (the Idawa'li) 78% involves money.

Among the non-food expenditure of nomads, clothes come first for the modest sum of 245 MU (barely 25 French Francs) per capita per annum. In this area, old habits- since the eighteenth century, the so-called 'Guinea' cloth and percale have remained the undisputed leading choice for Moorish veils and boubous - limit the needs of nomad households to one or two purchases a year.

Some expenditure on leisure or toiletries, which is sometimes an extension of old practices - the purchase of glassware and other 'trade' goods in the slave trade period - completes this picture of the consumption of nomad households whose energy needs (estimated at 445 kg of wood per capita per annum) are met in the framework of traditional gathering methods.

According to the most recent data, these are some essential features of consumption and incomes of nomad households in Mauritania. We have no illusions about the real significance of the figures provided by the survey on highly mobile populations, traditionally suspicious of anything that they see as deriving from central government, and living in a system still heavily marked by communal values that are hardly compatible with the individualization of budgetary units practicing rigorously autonomous accounting. But despite these reservations, the estimates obtained do throw a significant light on the socio-economic evolution of the Mauritanian nomad milieu. The major feature of this evolution is identified as a growing monetarization associated with colonization and the postcolonial heritage. The gradual emergence of a market in cattle, sustained by the growth of nomads' monetary needs, administrative requisitions and, more recently, the development of urban centres and the beginnings of a modern transport infrastructure, has been the driving force in a profound change in Mauritanian pastoral society which climatic and/or political and economic crises have helped to speed up. The social and political implications of this change demonstrate the contradictory character of a transformation which has all the appearances of a crisis.

Evolution of the social and political framework of nomadism

Claude Levi-Strauss wrote of functionalism: 'to say that something functions in a society is a truism, to say that everything is functional is an absurdity'. A statement with which I agree, and grant the structures of nomad society only a relative functionality, expressing in various domains (economic, technological, institutional, ideological and so on) the adjustment of a necessarily discontinuous human and animal occupation to the scattered and often ephemeral resources of a poor natural environment. I have earlier stressed the role and influence of the natural and economic constraints in the evolution of Mauritanian pastoralism. In the following pages the evolution of the political organization of pastoralism will be examined, and an attempt made to determine the role of the social and political structures of nomadism in the contradictions of present-day Mauritanian society which, in turn, has diverse effects on the structures of nomadism.

Women' herds and capital

Numerous observers have stressed the uniqueness of the social and political structures of pre-colonial Moorish society in which the apparent rigidity of status and hierarchical positions both confirmed and cut across the political control of a territorial space with ill-defined boundaries. What was the nature of this hierarchy and political control? How has the evolution described above of the economic context, and particularly the extension of commodity relations, affected this organization? What do the changes in the institutional, juridical and ideological framework of nomadization that have occurred since colonization mean for the future of pastoralism?

One of the most prominent characteristics of African pastoral societies in general and those of the Sahel in particular, and one that has engaged the interest of numerous ethnologists,7 is the close and many-sided relationship between pastoralists and their herds. These bonds largely conditioned the social and political structures of Moorish pastoralists, in which the necessary mobility of livestock and the hierarchical circulation of animals and their products (gifts, loans, tributes) played a central role. But other considerations must be taken into account, notably those concerning meeting nomads' cereal and agricultural needs and the specialization that they required, or specific features peculiar to the constitution of the ideological and political arena of this pastoral society.

The works of many anthropologists (Bonte, Stenning. Hopen. Asad. Galaty and others) who have observed nomad societies stress, over and beyond the immediate functions of livestock as a means of meeting elementary needs, the place occupied by livestock in the reproduction of the nomad community itself. The privileged role of livestock in the social relations and community life of nomads that ethnologists designate as a 'cattle complex', even 'boolatrie' (cattle worship) is thought by some observers to be essentially the obstacle to a 'rationalization' of the exploitation of the herds.

We have already mentioned the role played by mniha, the loan on the use of animals, especially milch cows, in the creation and consolidation of bonds of reciprocity and clientage in a society in which agnatic solidarity and its clientelist extensions represented the principal recourse against the constant threat of a razzia. The forms of mniha, the nature and number of animals it involved, the period of loans and the relations between givers and receivers naturally varied a great deal. A hartàni (former slave) with a flock of ewes may loan a former master, or some needy marabout or notable, one or two milkers for as long as they are giving milk, although he may well be concerned for their kids, which generally benefit very little from their mother's milk and may even simply have their throats cut before the mothers are returned. The religious benefit (receiving the marabout's baraka) or the political one, in the widest sense, or sometimes simply the inequality in power would usually be enough for such an 'irregularity' to be disregarded. A wealthier dependent or client would grant the loan of a herd for a long or even indeterminate period, contenting himself with a periodic inspection or occasionally collecting dues. Between kin too, great disparities in the distribution of animal wealth may be the source of mnayah (pl, of mniha) helping both to affirm kin solidarity and establish a hierarchy between givers and receivers.

In addition to the practice of loaning, tributes, ranging from a few kilograms of wool or a goatskin of rancid butter, to one or more head of cattle per adult male, contributed to ensuring that the circulation of livestock and its products had permanent role in maintaining and reproducing the social relations peculiar to the nomad group. Such tributes continued to be collected until 1951, when, under the aegis of the colonial administration, tribute-payers settled their last dues, partly in livestock, partly in cash.8 Finally, while in Moorish society, Islamized from a very early date, gifts, and ceremonial and ritual sacrifices were on a smaller scale than among other groups of pastoralists, they still made a significant contribution to the reproduction of the social order.

In this connection, the fundamental role of livestock in the payment of bridewealth (the dowry in the pre-colonial period was almost wholly settled in livestock) points to a very clear conjunction between the renewal of the pastoral society and the growth of herds. Summarizing generally the meaning of the reproduction and circulation of livestock as a privileged vehicle of social relations within Sahelian pastoral communities, Pierre Bonte writes:

It is in so far as the reproduction of domestic groups appears simultaneously as the reproduction of the communal conditions of pastoral production within domestic units that it requires the existence of a large surplus over and beyond the immediate needs of reproduction of labour and is integrated into a wider cycle of livestock circulation, resting on successive changes in the value of livestock. It is the moment when this livestock circulates as a social value that appears as determining of other moments of production because these are moments when it makes possible the simultaneous reproduction of the domestic groups as a whole and the community as such.
(Bonte 1977, p. 49)

The reproduction of pastoral society and its chief means of subsistence was effected in an institutional and territorial framework that has been largely undermined by the changes that began with the colonial occupation.

First, at the level of the social hierarchy, was the apparent rigidity of a structure that fixed the heredity status of each Moor, thus assigning him by virtue of his descent to one of the following groups: warriors, marabouts, tributaries (servant group), artisans, griots, former slaves and slaves. While this distribution of the Moorish population does not directly reflect a division of pastoral labour itself9 (only those tributaries specialized in animal husbandry appear to have been directly affected), the links between this structure and livestock, the focus and principal weapon of all social competition, may be clearly distinguished in almost every aspect of Moorish nomad life and activity.

This summary of the hierarchical and ceremonial circulation of livestock will not here be dealt with in detail nor will the role of the razzia, the activity that formed the base of the warrior aristocracy's power. It should be stressed that the work of artisans, paid for by the products of herding was, obviously, aimed at meeting needs associated with pastoral life (saddlery, tent equipment, milking equipment, shearing equipment, veterinary equipment etc).

The status of former slaves, many of whom were agricultural workers in the oases and in the more watered regions in southern Mauritania, was largely instrumental to maintaining the necessary complementarily between pastoral production and agricultural production referred to at the beginning of this chapter. This specialization of former slaves was, of course, maintained and reproduced to the advantage of the dominant groups who could thereby ensure for themselves a monopoly of political control of the land. Here, it is necessary to digress briefly in order to comment on the political control of the land, in which the changes over the last 50 years have had a crucial effect on pastoral mobility.

The apparent simplicity of what survives today of pre-colonial Moorish land tenure institutions- a tribal-type collective appropriation of the land- should not permit the earlier complex control of the area of pastoralism, in which use-rights and ownership rights were associated and interwoven with status and personal ties, to be forgotten.

First, it must be emphasized that appropriation of land, claimed to be operative over a territorial area with more or less recognized and acknowledged boundaries, actually concerned only the most useful parts of the territory appropriated: watering places and growing areas. Control of watering places, which largely commanded control of transhumance circuits, was itself subject to varied and contradictory norms.

Apart from during exceptional circumstances and notably in the event of armed conflict, permanent water sources (lined wells, natural springs, large winter ponds and so on), while deemed to belong to this or that 'tribe' or tribal 'fraction', were freely accessible to all nomads Cases of enclosing a winter pond in order to restrict its use to a particular group were rare and often hotly disputed The realization of seasonal draining wells, where adequate surface water made this possible, was not a problem, since this did not justify any permanent appropriation of the land.

The relationship between watering places and grazing areas explains the nomad tribes' resolute opposition should elements external to the tribe propose to dig a permanent well on 'their' land Beyond an expression of tribal nationalism (the tribe being, above all, a political unit) this opposition obviously reflects the nomads' fear of restrictions that might affect the free availability of grazing areas.

In pre-colonial Mauritania, possession of watering places formed part of the complex game of status relations and marked out the institutional framework of control of the pastoral space rather than did the fluid territorial boundaries.

Wells, for example, were mostly in the hands of marabout groups, the leading organizers of the pastoral economy, but they generally had to take account of the political hegemony of the warriors who were prepared to use their weapons to ensure permanent access to forage and water resources for themselves, unless established privileges enshrining their hegemony in this area were institutionalized.10

Control of the use of grazing areas often assumed its most stable form in the payment of individual or collective dues that, in principle, were part of a contract for protection in a society which was constantly prey to razzias.

Despite the precarious nature of their authority, only the emirates (Trarza. Brakna, Adrar, Tagant), which emerged in the seventeenth and eighteenth centuries, tried, within often indeterminate boundaries, to provide a precise territorial framework to the political and hierarchical system whose unity they both expressed and guaranteed. There too, imposing a tax on nomads from outside the emirate, a sort of entry tax in exchange for their (relative) security, was a means of regulating the use of grazing areas as well as an instrument to affirm the hierarchical power of warriors within the emirate since only they were able to impose and collect this sort of due. The instability of the political power of the emirates, not only in the sense of a chronic precariousness of the leaders' personal positions but also in the sense of uncertainty as to the legitimacy and effectiveness of its ideological and hierarchical foundations (the real or supposed power of some great marabout, reputed to act on occult forces might, in some cases, effectively oppose that of the emir) in addition to the absence of fixed and recognized borders in fact left herders a freedom of movement limited mainly by the extent of their alliance network or their capacities to defend themselves against possible aggressors.

Such are the broad outlines of the situation before the French colonization of Mauritania. The upheavals that have since occurred, affecting the territorial and political organization, and hierarchical structures of Moorish pastoralism, have profoundly transformed a nomad social milieu made increasingly vulnerable to the vicissitudes occasioned by an unpredictable climate.

Pastoralism and the Mauritanian state

With colonization, the web of relations that was the key characteristic of any group or category: caste, tribe, religious brotherhood, rank, status, and so on, and the power relations it underpinned which, as we have noted, were closely linked to the political organization of the pastoral space, came under the direct control of the French administration. The overall movement of nomadism in a now unified territorial area - with the disorderly violence of the razzia supplanted by the violence of colonial rule - as well as local population movements, were considerably changed by this new situation. Nomad society itself was integrated into a larger territorial and political grouping embracing communities of sedentary black agriculturalists and suffered the levelling effects of the market economy. Its reproduction became increasingly difficult. Since independence (1960) pastoralism was generally only indirectly affected by the official options of the Mauritanian authorities but these, nevertheless, helped accelerate what has in recent years become a pronounced trend towards the sedentarization of nomads.

The 'security' inaugurated by colonial peace and the 'freedom of grazing' decreed by the colonizer created major obstacles to the free movement of men and livestock. Through those close to the 'commandant' (petty administrative staff, guards, servants, and so on) the creation of administrative centres and the opening of a few schools favoured the diffusion of modes and models that underpinned the extension of the commodity relations.

This extension itself precipitated and required, if not a dissolution, at least a profound weakening of traditional social relations that was essential to the 'freeing' of labour mentioned above. This meant at least a partial 'Loosening' of the hierarchical relations that provided the framework for the production and reproduction of pastoral society. Some progress, though ambiguous, has been made in this direction in recent years.

In the past, as today, the principal issue in 'freeing' labour is the status of dominated groups in Moorish society (probably over half the total Moorish population) and the nature of their relations with the groups that dominate them.

With colonization, as already remarked, the warriors" autonomous military power ceased. But by adapting to the new situation, the power of the marabouts, whose essential vocation lies in the production of what Weber called 'the goods of salvation', tended to be maintained and even to grow, in order to handle the diffusion of the ideological negotiations of a society in crisis.

The colonial administration claimed that one motive for its action was to eliminate marabouts' and warriors' exploitation of groups whose support it hoped to win over. But it soon adopted an attitude of compromise with the ruling classes of Moorish society, once their military resistance had ended (Coppolani and Gouraud columns in 1902-5 and 1908-9). The 'Patey instructions' (from the name of the then Governor of Mauritania) of February 1910 set out the broad outlines of the policy to be followed in reestablishing a now loyal tribal chieftaincy. By taking on tribal notables, or recruiting them as functionaries after a brief spell in the a 'School for the Sons of Chiefs', a system of Collaboration', which offered the advantage of maintaining the traditional social hierarchy apparently intact, was rapidly established. These were the first steps in the formation of the post-colonial bureaucratic class in which elements from the traditionally dominant groups or families have retained much of their hegemony.11

The same 'prudence' was observed by the colonial administration in its attitude towards the dominated groups in Moorish society. Under colonial rule, tribute-payers continued for a long time to pay dues to their lords and, only belatedly, after 1946, was the redemption of tributes by the payers themselves speeded up under the auspices of the administration; the last transactions took place in 1951.

Slavery, although officially abolished, was- and still is- widely tolerated by the authorities.12 Contrary to widespread belief. Moorish slavery, an institution probably as old as the Moors themselves, does not, for the most part, derive from the isolated kidnapping of blacks by raiders. Doubtless there were many such cases and the insecurity maintained by Moorish razzias among the sedentary black peasants in neighbouring areas left memories that are still very much alive. But our few brief surveys show that over 80% of Moorish slaves who still remember where they or their ancestors were enslaved (only a minority of those we questioned) were from Bambara country where they were bought in the nineteenth century. Slaves were used as herders and shepherds, and as well-diggers in agriculture, or generally as share-croppers - in which case they were usually former slaves paying dues that might range from a symbolic quantity of seeds to virtually the whole harvest. At the time of the slave trade, they were also used to collect gum. Moorish slaves were not a servile mass employed in heavy communal labour, as in Graeco-Roman antiquity or on the American plantations, but 'house' (or rasher 'tent') slaves whose status and personal living conditions varied, depending on the master's status, from beast of burden to confidant and personal adviser.

While officially abolishing slavery, the colonial administration did not embark on any systematic action to make such a measure a reality. There were even frequent cases of complicity between some local representatives of the authorities and owners who had come to claim slaves who had run away. It should be noted, however, that fear of a possible repression and the support that maroons could sometimes count upon, particularly among the black auxiliaries of the colonial administration, tended to make administrative centres places of refuge. Since the end of European slave trading, and given how little traffic in slaves there was across the western Sahara, sales of slaves in Mauritania which had only ever involved small quantitites of human merchandise, or even just individuals, became rarer. Isolated cases of sales (some are still reported today) thus increasingly took on an exceptional and more or less clandestine nature.

In addition to these factors, the advance of agriculture, encouraged by the colonial administration which supervised the building of a few small dams such as that at Magta Lajar in the late 1940s, accompanied and reflected advances in the sedentarization of slaves and former slaves that signified not only an abandonment of the pastoral way of life but also the beginning of an emancipation movement still continuing today.

The recent stepping-up of this movement towards sedentarization, which of course does not involve only former dependents, has helped accentuate the imbalances and contradictions of a pastoral society no longer able to control its mechanisms of reproduction.

The drought affecting the whole of the Sahel since the late 1960s; the development of an urban wage-earning sector associated in particular with mining activities and currently totalling some 25,000 jobs; insecurity in the countryside associated with the war in the Sahara, which began in 1975: the development of road transport, notably the building of the Nouakchott-Néma road (begun in 1974), which more or less coincides with the track now used by nomads, who previously frequented the now decertified areas further north; are all factors that explain the scale of sedentarization and migratory movements directly affecting the organization, values, and the very existence of pastoral society.

Resort to agriculture as a substitute for or supplement to the severely reduced income from herding (herd losses due to drought have been very high, over 80% for many families) is not problem-free. The scale of the population pressure engendered by sedentarization on the few cultivable areas is leading to an exacerbation of land disputes. In some places, given the overall contradictions of Mauritanian society and the clientage and 'racial' nature of the available ideological models, these disputes are taking on more a tribal and ethnic, rather than a strictly class character.

Moorish slaves' and former slaves' challenge, encouraged in the towns by their growing importance in some sectors of the Mauritanian state apparatus, especially in the other ranks of the army, and helped or supported by the few of their number who occupy a position of some prominence in this apparatus, is thus itself set in the tribal and ethnic framework that gives it its specific features and limits. A number of contradictory aspects of a development, highly significant not only for the future of pastoralism but also for the whole of Mauritanian society, need to be stressed here.

Dominant groups in nomad society, ruined by the drought and hardly enthusiastic about agricultural work, settle close to land, often for long cultivated by their former dependents, and attempt to impose a share-cropping system upon them on lands that both parties agree belong to the tribe collectively. Previously, the nomad groups would have been satisfied, with more irregular and tolerable contributions in exchange for a few products of herding. Like their former masters, the former slaves, faced with the claims of other former slaves, justify their possession of the land on the basis that they belong to the tribe that owns the land. But the tribal juridical framework, covering the hierarchical stratification already detailed (warriors, marabouts and so on) with the former slaves and slaves at the bottom, implies the exclusion of this group from ownership. The former slaves' class aspirations and demands, and their desire to appropriate the land, thus conflicts with the tribal framework that integrates them into pastoral society as agricultural labourers but excludes them from ownership of the land, exclusive enjoyment of which is based on the political power of the dominant groups that, in turn, it contributes to establish and reproduce.

The hierarchical structures of pre-colonial pastoral society have thus been largely transferred into the present Mauritanian state order (marabouts and warriors occupy bureaucratic and commercial jobs, former slaves and slaves the lowliest and worst paid - labourers, domestics and so on). This helps to maintain, even perpetuate, a brutally hierarchical social order, involving collection of tribute, unpaid labour and corporal punishment. This order is gradually giving way to a clientelist and pseudo-philanthropic practice in which yesterday's masters who have become today's bureaucratic intermediaries, pose as protectors - or even as victims of clients who sometimes have to be lodged and fed in the urban areas.

The tribal political model's perpetuation through its clientelist extensions also partly explains the ambiguity of the political expression of the movement to emancipate former slaves. This ambiguity reflects both the continuing significance of the model according to the logic of the pre-eminence of the dominant class' ideology and the sociological heterogeneity of a rural- and 'tribal'-based movement under an urban leadership from the middle bureaucracy aspiring to convert clientelism in such a way as to legitimize what otherwise would not be fully operative. Social identity, measured in terms of the tribal system that based prestige and legitimacy on a genealogical (re)construction that excluded slaves and former slaves was not practicable for the latter, but within the range of available communal identifications there remained that of ethnicity. This was all the more tempting because it enabled their 'representatives' to play on the existing ethnic rivalries (a fundamental feature of the Mauritanian political scene) by integrating and (re)emphasizing an ethnic origin (former slaves and slaves are black Africans) that was a permanent mark of their inferiority within the tribal order. Thus, following the coup d'état of July 1978 there were tracts calling for a division of power on an ethnic base.

These observations upon former slaves and their efforts to secure emancipation (the Moorish slaves have just been solemnly 'freed' for the third time in half a century by government decree), and the integration of this phenomenon into a wider field of clientelist-type (tribal or ethnic) political contradictions and rivalries that run through present-day Mauritanian society, have only apparently taken us away from pastoralism and its future. These rivalries and contradictions are, it is true, over-determined by a regional and international context marked by competition between local micro-hegemonisms and between the commercial and strategic interests of the great powers which, inevitably, are concerned about the Sahara war and its possible 'tribal-ethnic' fall-out. But what, in reality, is at stake in this interweaving of alliances and oppositions between various social strata of pastoral society, between 'tribes', 'ethnic groups', and between sedentary people and nomads, is the future of pastoralism. Here, a particular and decisive aspect of the network of contradictions that encloses and moves Mauritanian society - the condition of servitude and quasi-servitude - has been stressed. This is because within it are articulated and expressed, in the dominant political idiom, the language of a clientelism with 'tribal' and 'ethnic' overtones, the problems and contradictions of a pastoral society profoundly transformed by colonial and commercial domination.

Conclusion

First, we must correct the impression of a possibly over-inflated contrast in presentation between the 'functionality' of the pre-colonial order and the dysfunctionality of the order that has replaced it, in short the opposition between the 'good poverty' of the past and the 'bad poverty' of the present. To exalt either the razzia or slavery, or the epidemics that periodically scarred a nomad society, which always led an extremely precarious existence, was not intended. It was, however, necessary to stress the striking resistance of a way of life resting on a fragile balance between a big-climatic threshold marked by scarcity of water and grazing land and a set of technical, economic, institutional, ideological and other behaviours.

The devastating role of the last 15 years of drought in unleashing the exodus from rural areas, and the massive sedentarization of nomads has, justifiably, been stressed. But droughts and famines are not new in the countries of the Sahara and the Sahel. The Oualata Chronicle, for example, mentions no fewer than five great famines between 1249 AH (1830-31) and 1311 AH (1893-94) plus half a dozen epidemics (especially the serious one in 1286 AH (1869-70) that carried off over 300 people), not to mention countless razzias.13 The Tichitt Chronicle14 records the same recurrence of the cycle of drought-razzia-famine-epidemic that seems to have marked the whole history of this locality. But these disturbances, serious as they may have been, never occasioned a real break-up in the pastoral way of life comparable to that we have been witnessing for the last 15 years. Before colonization, the vagaries of the climate never led to the formation of a town; nor did they contribute significantly to a rise in the population of the caravan camps which the increasing number of tributes and razzias made particularly inhospitable in times of crisis. But once the economic and ideological wellsprings of pastoral societies had been broken in a process in which violence and coercion combined with the effects of commodity relations, conditions were ripe for such a climatic catastrophe as that the Sahel has recently been experiencing, to have mortal consequences for pastoralism.

The role of the state- a mechanism to administer legitimate violence, as Weber defined it - in this process cannot go unmentioned.

The role of political control of the pastoral space in the mobility that is a key part of the nomad way of life and mode of production has been demonstrated. Some argue that this control was exercised solely through segmentary tribal structures governed by kinship and excluding the appearance of any centralized political authority, or a state. This is notably the opinion of C. C. Stewart15 who takes up the functionalist theory of segmentarism developed by Evans-Pritchard and his followers, and considers pre-colonial Moorish society to have been ianarchic', 'acephalous', 'stateiess', and in which the emir, like any other tribal chief, was simply a 'primus inter pares' with no real authority. According to Evans-Pritchard, the opposition (both antagonistic and complementary)16 between equivalent lineage segments, conceived as a correlate of the genealogical structure of unilinear descent groups ('Arab' marriage with the patrilateral parallel cousin) which leads to strong agnatic solidarity17 and to an equally strong tendency to fission, obstructs the emergence of autonomous political structures within segmentary tribal societies that are perpetually condemned to 'anarchy' end war. While this representation undeniably contains features that portray part of Moorish social and historical reality, it has the disadvantage of being unable to explain the emergence, even in an embryonic form, of political structures that were becoming autonomous: those of the emirates. For even though these structures suffered the effects of the segmentary system and kinship they nevertheless represent the beginnings of a political power, a state in the process of being constituted. Neither does it seem that the emergence of those proto-state emirate forms associated with social stratification, which cannot be explained solely in the framework of the functionalist problematic of the theory of segmentarity, can be explained within the framework of social contract theorists (Locke, Hobbes. Rousseau) and their modern descendants, solely by the desire to legitimize a domination derived from a differentiation and polarization between wealth and poverty, exploited and exploiters of which the infant state is seen as simply the scarcely veiled instrument.

Beyond this alternative, which contrasts the impossibility of the state emerging in segmentary tribal societies governed by kinship with an exclusively instrumentalist theory (the state as the tool of class domination) there is the outline of an intermediate reality: a transitional area where kinship and politics combine in a complex articulation in which the ideology and language of kinship continue to provide a body of representation to political structures that express as much domination by a class (the warrior and marabout aristocrats) as perpetuation of the genealogical principle that supposedly governs tribal unity. In the situation inherited from the slave trade period and colonization, and whatever misunderstandings may have surrounded the real nature of tribal and emirate power at that time, the effects of kinship structures on the distribution and exercise of political power have continued to operate through new forms of clientelism.

Whether under the regime of freedom of political competition (in fact a very tightly controlled freedom) in the late 1950s (the period of the 'lot-cadre'), or under the post-colonial regime that today, has culminated in a militarization hardly concerned with political representation associated with elections (National Assembly, regional councils, local government and so on). Mauritanian politics has continued to draw on kinship networks for the establishment of a legitimacy that has been unable to kind sufficiently credible roots elsewhere.

Clientelism here, indicates the process of converting bureaucratically accumulated economic capital - payments made to local political middlemen or 'economically'- through the mechanism ('sole agency', trading commission, for example) of local distribution of central capitalism's industrial or agricultural products - into political capital ('representativeness'), without which the expansion of wealth and prestige would rapidly cease.

The close relationship of private economic prosperity and the bureaucratic, state management of dependency, even leaving aside what Hamid El Mauritanyi calls bureaucracy's 'illegal tranche of income'18 (the product of corruption, diversion of public funds, and so forth) appears in the reciprocal circulation of money and men in both directions. Every concessionaire19 must, in fact, have connections in the state marketing commission and parastatals, the sole national clients of any significance, and every politician who 'retires' almost inevitably ends up with an 'agency'.

What is thus emerging is a necessary reciprocity between 'representation' in the commercial sense of the word and 'representativeness' in the political and social sense. In a society still strongly marked by its tribal structures and in which the bureaucratic-capitalist sector employs barely 2% of the labour force, they inevitably meet in the domain of kinship.

To ensure the fruitfulness and survival of their capital stock of 'representativeness', a man who becomes rich, a professional politician or aspirant to the position, must keep the broadest possible clientele happy, first, his close kin, then dependent groups and people of the same tribe. Their solidarity, generated as much by the redistribution of economic benefits as by kinship bonds, will in turn form the basis of the representatives' (politicians or concessionaires)' representativeness' in the state and capitalist sector of the national society and economy.

Does this imply, to use Samir Amin's expression, that in common with other African states dominated by central capitalism the Mauritanian state is becoming 'transnationalized'? If this expression means the massively dominant role of managing dependency, then it cannot be denied.

Such a viewpoint can, however, be criticized as excessively reductionist if it leads to a perception of the state in dependent African countries as simply an instrument of imperialist domination. This view is strictly in line with that which Lenin, quoting Marx and Engels, helped popularize: the state as the tool of a class dictatorship culminating, according to the authors of State and Revolution, in the 'withering away' of the state in the communist society of the future. The Mauritanian situation, marked by a profound interpenetration of kinship bonds, will in turn form the basis of the representatives' (politicians or religious factors exercise a decisive influence,20 does not accord with this theory.

The Mauritanian states expresses both the present dualism of the economy and the relevance of the ancient economic, ideological and symbolic structures of kinship. It is thus a hybrid, both an instrument of centralization and a central stake in centrifugal strategies. It cannot be reduced to the role of administrator or manager of dependency.

It is tempting to see in the Mauritanian state only an instrument in the service of multinational domination combined with a sort of cannibalism in which the administration of poverty-stricken indigenous populations and organization of the aid intended for them enrich the most unscrupulous sectors of the bureaucracy. In Haiti, some tontons macoutes allegedly sell the blood of their compatriots to the USA - Africa has not descended to this level. But we should beware of a cannibalism that might be an essential part of any state, for is not governing necessarily 'devouring the substance of others'?21

Notes

1. On all aspects of grazing and herding among the Moors, see the following works: Francart. 'Le pâturage en Haut Adrar'. Bull. IFAN. 1940. II. 3-4, 285-78 end 'Nose sur le vocabulare camelin en Haute Mauritanie', Bull. IFAN. 1941,III 45-52; V. Monteil, Essai sur le chameau au Sahara Occidental. Bull. IFAN. Saint-Louis-du-Senegal. 1952: and Contribution a l'etude de la flore au Sahara Occidentale, Paris, Larose 1949; A. Leriche. 'Vocabulaire du chameau en Mauritanie'. Bull. IFAN. 1952. XIV. 3. 985-95: end 'Costumes maures relatives a l'élevage'. Bull. IFAN 1953, XV.3, 1316-20.; Leborgne. 'Vocabulaire technique du chameau en Mauritanie' Bull. IFAN. 1953, XV, 1.292-380; Charles Toupet (thesis) 'Le sédentarisation des pomades dans la Mauritanie Centrale Sahélienne'. Paris. 1975.

2. 'Rapport annuel sur le commerce et l'industrie pendant l'année 1938', Colonie de la Mauritanie. Archives de la RIM, série Q No 411.

3. See also: Ould Cheikh, Les Maures, RAMS, 1980:: P. Bonte and Ould Cheikh, Nomadisme, sédentarisation, migrations dans la société maure, Unesco. Population Division, 1980; J. P. Hervouet. 'Types d'adaptation sahéliens', Thèse de IIIe Cycle, University of Rouen. 1975.

4. Concerning water for grazing purposes, one of the main areas of herding where the administration has taken action, the second Mauritanian economic and social development plan (1970-73) estimated that there were 3.000 cement wells in Mauritania; 750 built by the administration between 1950 and 1968, including 600 in the 1950-60 period alone.

5. Urban wages (Nouakchott, Nouadhibou, Zouérate) of employees in the same category (domestics. Iabourers etc.) reach barely 3.000 ouguiya per month. It is quite common to see small, underpaid (a few hundred ouguiya) children working as domestics.

6. This is still far from the consumption levels in industrialized countries. The annual average per person in France, for example, is 94 kg (see Jean Ziégler, 'Le scandale de la surconsommation de viande dans les pays riches', Le Monde Diplomatique, November 1981, p. 10.

7. See the paradigmatic role attributed by some authors (Luc de Heusch, René Girart et al) to the ritual behaviour of East African herders, and those 'sacred' kingships of the Great Lakes region that involve cattle. See also Bonte and Becquemont, 'Travail, valeur, besoins et conscience aliénée: le cas des éleveurs de l'Afrique de l'Est'. La Pensée, 1980, pp. 90-121.

8. Of 51 agreements to redeem the tribute (noted in the Mederdra archives for 1946-47) between dominant groups and tributaries,36 involved a settlement in cash (120.475 Francs): five a 'mixed' settlement part cash (11.000 Francs), part cattle (29 sheep and two she-camels); eleven in cattle (174 sheep, seven cows, two steers, ten she-camels (including four with young), eleven male-camels, one she-ass, one vliz (strip woven from sheepswool for tent-making)).

9. The dominant ideology, of the marabout class in particular, claims to justify the Moorish social hierarchy by a decision of the Almoravid leader Abù Bakr Ben 'Umar (d.1087 in Tagant). On his death-bed, he is said to have decided to distribute the men making up his army as follows: warriors, responsible for propagating Islam by force of arms; marabouts, responsible for religious teaching and education: the tributaries, responsible for maintaining the first two groups. See in particular 'Al wasit...'. Cairo and Casablanca (2nd ed.). 1958, by Ahmad teen Al Amin Al-Sinqiti, p. 475 (in Arabic).

10. For example, the al-me clause at the end of the Sârr Bebbe war(second part of seventeenth century) stipulated that the defeated group would, when necessary, offer the victors and their descendants one-third of the water they drew from their own wells.

11. 'Virtually all the leaders of the Islamic Republic of Mauritania come from the traditional aristocratic orders: 162 out of 175. Men from tributary or artisan groups (ten) and servile categories (six) can almost be counted on the fingers of one hand... 90 out of 175 are chiefs or sons of chiefs, and 66 are from the families of notables'. F, de Chassey, in Mauritanie, 1900-1975,. Paris, Anthropos 1978, p. 286.

12. On this we have collected numerous testimonies from the masters themselves. See also one by a Dahomeyan exile: Louis Hunkarin, Un forfait colonial: l'esclavage en Mauritanie. Imprimerie moderne. Privas 1931.

13. Paul Marty (trans.). 'Chronique de Oualata et de Néma'. Revue des Etudes Islamiques. 1927. III, pp. 355-426.

14. Vincent Monteil (trans.) 'Chronique de Tichitt', Bull. IFAN. I, 1939, pp. 283-312.

15. C. C. Stewart. 'Political authority and social stratification in Mauritania', in E. Gellner and A. Micaud (eds). Arabs and Berbers. London, Duckworth 1972.

16. Evans-Pritchard wrote of the 'tribe without rulers' model of the Nuer 'Each segment is itself segmented and there is opposition between its parts. The members of any segment unite for war against adjacent segments of the same order and unite with those adjacent segments against larger sections.' The Nuer. Oxford, Clarendon Press 1940, p. 142.

17. One has only to think of the 'asabiyya of Ibn Khaldun the wellspring of the solidarity and unity of action of groups of nomad conquerors and the key concept in the cyclical conception of history developed by the great Magrhibi writer. See Al-Muqaddima.

18. Hamid El Mauritanyi. L'indépendance néo-coloniale, op, cit.

19. Perhaps there is need to distinguish 'real concession' end 'fictional consession', the representatives who actually sell something (mass-consumed foodstuffs, textiles, cars, and so on) market purveyors and other front names for licences (fishing etc.), and those solely concessionaires, the middlemen who some call ironically 'Messrs Tenpercenters'. Obviously, there are complex links between the two categories.

20. For more detail see: A. W. Ould Cheikh. 'Comment prêcher dans le desert. Fonction cléricale, fonction guerrière et émergence de l'Etat dans la société maure', in P. Bonte and J. Galaty (eds), African Pastoralism and the State, London. Sage (forthcoming).

21. A Tiv (Nigerian tribal group) saying, reported by Paul Bohannan and quoted by G. Balandier. Political Anthropology. London, Allen Lane The Penguin Press 1970, p. 60.