What is the use of my describing to you a piece of ground which you ought to see, in order to judge of its nature?
(Friedrich Engels, The Moorish War)1
The heartland of Greater Kabylia is the Jebel Jurjura and the series of parallel ridges which extend northwards from it. The inhabitants of these ridges and of the central Jurjura proper are the Igawawen and they are the core of the Kabyle population. It was the Igawawen whose defeat at the battle of Icherriden in 1857 is generally taken to have marked the French conquest of Greater Kabylia as a whole and it was they who provided the majority of the Kabyle element in the leadership of the nationalist movement from 1926 onwards. In so far as the revolutionary war of 1954–62 had a theorist before the fact, it was Hocine Aït Ahmed, from the Igawawen tribe of the Ath Yahia, and it was the central Jurjura which furnished the nucleus of Aït Ahmed’s rebellion against Ben Bella’s government in 1963–5.
The political salience of the Igawawen was evident to the colonial authorities by 1857 if not earlier and ensured that the attention of the colonial ethnographers and other observers thereafter was concentrated upon them. The works of Carette and Devaux,2 Hanoteau and Letourneux and Masqueray abound with references to the villages and tribes of the Jurjura but display only a superficial interest in or acquaintance with the inhabitants of western and maritime Kabylia (and virtually none at all with those of Lesser Kabylia). This tendency to concentrate upon the Jurjura has also characterised the twentieth-century literature on the region.3
Figure 2.1: The south-west of the Igawawen district, with the peaks of the Jurjura behind. In the middle distance, the ridge bearing the Ath Boudrar villages of Bou Adnane, Ighil n’Tsedda and Ath Ali ou Harzoun and the Ath Wasif villages of Tassaft Waguemoun and Ath Eurbah.
One effect of this bias has been to encourage the assumption that the society of the Jurjura is representative of Kabyle society in general, thus obscuring the important dichotomy between the Igawawen and the rest of the Kabyle population. Because of this, the remarkable political salience of the Igawawen has never been adequately explained. For it is because the society of the central Jurjura differed in significant ways from that of the rest of Greater Kabylia that the Igawawen were able eventually to function as the core of the Kabyle population.
The same tendency to overlook the main difference within the society of Greater Kabylia has also characterised the perceptions of Arabophone Algerians. In Algerian Arabic, Zwawa (singular: Zwawi, usually written Zouaoua, Zouaoui by the French) is the term used to denote the people of Greater Kabylia in general as distinct from those of Lesser Kabylia. It is, of course, a derivative of Igawawen. This usage has arisen from the fact that the inhabitants of Greater Kabylia with whom the Arabic speakers of Algeria were acquainted were very predominantly from the Jurjura, where the migratory tradition was most highly developed.
This usage has not been confined to Arabophone Algerians. The Berbers of Lesser Kabylia have also been accustomed to refer to Greater Kabylia in general as le pays zouaoua, bled Zwawa or, in Berber, thamurth Igawawen. Jean Amrouche, the distinguished Kabyle poet, was born in Ighil Ali in Lesser Kabylia, but his mother, Fadhma Aïth Mansour Amrouche, came from Tizi Hibel, of the Ath Mahmoud tribe in Greater Kabylia, which Jean referred to in his memoir of his mother as ‘the Zouaoua region’.4 Strictly speaking, however, Fadhma Aïth Mansour Amrouche was not from the Igawawen district. The Ath Mahmoud tribe did not regard itself as an Igawawen tribe, nor was it so regarded by its neighbours or by the Igawawen tribes themselves.5 For, within the society of Greater Kabylia, the dichotomy between Igawawen and the rest of the Kabyles (Leqbayel) has long been established. In the Kabyle dialect, Igawawen (singular: Agawa) denotes the people of the central Jurjura as opposed to those of the rest of the region. This is putting matters a little vaguely, however, and there has been a degree of uncertainty about the precise limits of the range of reference of the term in Kabylia itself.
According to Hanoteau and Letourneux, the term referred properly only to the four tribes of the Ath Bethroun confederation (the Ath Yenni, Ath Wasif, Ath Bou Akkach and Ath Boudrar) and the four tribes of the Ath Menguellat confederation immediately to their east (the Ath Menguellat proper, the Ath Attaf, Aqbil and Ath Bou Youcef). This is the most restricted use of the term, however. Devaux reported that two tribes to the east of the Ath Bou Youcef – the Ath Itsouragh and the Illilten – were known locally as ‘the eastern Igawawen’,6 and Bourdieu states that the Ath Yahia are also considered to be an Igawawen tribe.8 Morizot, on the other hand, uses the term ‘Zouaoua’ to refer, in addition to the above, to all the tribes to the north of the Ath Menguellat and the Ath Yahia up to the Sebaou valley.9 In this usage, the five tribes of the Ath Irathen confederation and their eastern neighbours – the Ath Fraoucen, Ath Khelili and Ath Bou Chaïb – are also considered to be Igawawen.
Map 2.1: Kabylia: the Igawawen and the other main confederations and tribes.7
There is a sociological warrant for this extended usage, for the tribes in question resemble their southern neighbours in most essential respects and differ in the same respects from the rest of the Kabyle population. There is also a historical warrant for Morizot’s usage, for the Ath Irathen and their neighbours were referred to as ‘Igawawen’ by the tribes on the north side of the Sebaou valley, notably the Ath Jennad confederation.10 Thus Morizot’s extended usage is soundly based and I shall follow it here, although it should be borne in mind that some Igawawen were definitely more Igawawen than others, and that the Ath Irathen in particular have been inclined to distinguish themselves from their fellow Igawawen to the south, a fact which has been of political significance in certain contexts, as we shall see.11
The fact that Kabyles from the non-Igawawen tribes of Greater Kabylia should, when abroad, refer to their native region as thamurth Igawawen is evidence of how far the Igawawen had succeeded in imposing their hegemony on the region as a whole. And the way in which the ordinary Kabyles of Greater Kabylia employed the terms Agawa and Igawawen within the region is in itself evidence of the unsegmentary character of Kabyle politics in the pre-colonial period.
It is an implication of the segmentarist thesis that proper names, other than those of individuals, can identify only segments of socio-political organisation – families, clans, tribes and so forth. Only one exception to this rule is allowed for, at any rate in the case of the Berbers of North Africa – that is, linguistic groupings, such as the Chleuh, Imazighen and Rifians in Morocco and the Chaouia, Mzabis and, for that matter, of course, the Kabyles themselves in Algeria. But, within the society of Greater Kabylia, ‘Igawawen’ is neither the name of a tribe or a confederation of tribes nor that of a distinct linguistic grouping. It denotes neither a level of segmentary organisation nor the sole admissible alternative to this. There is no comparatively uniform and distinct Igawawen variant or sub-dialect of Thaqbaylith12 and the Igawawen were (and still are) divided into a number of tribes and confederations, as we have seen.
What united the Igawawen and endowed them with a particular and enduring identity which has no counterpart elsewhere in Kabylia was not kinship solidarity nor organisation nor language, but a common and distinctive culture. It is the particular culture of the Igawawen which has enabled them to function as the core of the Kabyle people. They have done so, not because they are the most ancient and long-established element of the Kabyle population, but despite the fact that they are the youngest or newest element of this population. As Jean Morizot has explained, contrary to a widespread misconception of the matter, the Jurjura was the last part of Greater Kabylia (and probably of all the several Kabylias) to be settled and its settlement began probably no earlier than the eleventh century CE.13 The manner in which it was settled gave rise to a unique cultural development which, so far as I am aware, has no counterpart among other Berber populations in Algeria or elsewhere.
This development was that of a mountain society which was both egalitarian and unusually orderly, fiercely independent and highly integrated but also outward looking, and of an economy characterised by intensive and highly diversified commercial craft manufacture and by an equally diversified pattern of commercial and labour migration. This society was regulated by a remarkable system of self-government which achieved its final form between the early seventeenth century and the middle of the eighteenth century CE and both constituted the framework within which a complex economic life could be carried on and simultaneously gave rise to an unusually complex and sophisticated orientation to the central power.
Pierre Bourdieu begins his account of Kabyle society with the statement that
settled in very heavy densities (267 inhabitants to the square kilometre in the arrondissement of Fort National) in regions of hilly, rugged terrain, the Kabyles are primarily arboriculturalists
and continues
the economy depends mainly on two trees, the olive and the fig, together with a few complementary crops (hard wheat and barley) and some small-scale stock raising.14
This is a very incomplete picture.
The population density of Greater Kabylia and, above all, of the central Jurjura is certainly one of the region’s most important and striking features. It has no equivalent elsewhere in the Maghrib and bears comparison with that of the Nile delta. Unlike the Nile delta, however, this superabundance of humanity is combined with an acute scarcity of arable land. It is inconceivable that this population could have been sustained had its economy depended mainly on the olive and the fig and complementary farming activities.
To ask what was the primary economic activity of pre-colonial Kabylia and what its economy mainly depended on is to ask questions which cannot be answered. The rare statistics which exist are incapable of yielding a reliable general picture.
This is partly because Greater Kabylia in the pre-colonial period was not an economy in the proper sense of the word – that is, an economic unit subject to some degree of collective accounting of resources and expenditure of the kind which generates, inter alia, accurate statistical records. There was no unified regional market. The units of measurement varied from one district to another and often from one tribe to another, as did the units of currency. The economic fragmentation of the region mirrored its political fragmentation.
But it is also because of the wide range of productive activities engaged in by the people of Greater Kabylia in the pre-colonial period and, in many cases, well into the twentieth century. It is extremely doubtful that any one of these was primary in the sense of being the economic mainstay of the population, such as was the case of potato cultivation in southern Ireland and rice paddy cultivation in Java. On the contrary, it would seem that the reproduction of the society depended upon its ability to engage in a wide diversity of activities rather than on any particular one of them. It would therefore be more accurate to say that the economy of pre-colonial Kabylia depended on five things – arboriculture, horticulture, craft manufacture, external trade and labour migration – and that all of these were indispensable, if in varying degrees, to the prosperity of each of the tribes and villages which composed the society.
Arboriculture and the various farming activities which complemented it were engaged in throughout the region. The olive and the fig were the principal crops, although in some districts almond, pear and cherry trees were also cultivated. The surplus production was sold on the market, especially olive oil. The importance of olive oil production can be gauged from the fact that in 1834 no fewer than 358 oil mills were counted in what was later known as the cercle (district) of Fort Napoléon15 (encompassing the central Jurjura and its outcrops) and 348 in that of Tizi Ouzou (encompassing the rest of central Kabylia and the western half of maritime Kabylia). In the same year, nearly one million litres of olive oil from Kabylia are estimated to have been sold on the Algiers market.16
The other farming activities were (and still are) stock raising and horticulture. The land in the immediate vicinity of every Kabyle village is given over to gardens which, like the orchards, are the private property of individual families. From these each household obtained most if not all of its requirements in vegetables: beans, chick peas, chili peppers, courgettes, green peppers, marrows, onions, parsnips, potatoes, tomatoes, turnips, and so on.
For most villages in the region, stock raising was a secondary activity. Most households would keep a few head of sheep and goats, the former supplying wool and meat for the rare occasions when it was consumed – that is, religious festivals and family celebrations (marriages and circumcisions) – the latter providing milk and hides. In the remoter villages of the Jurjura that had access to plentiful pasture, the flocks would be much larger and in some cases small herds of cattle would be kept. Chickens would also be kept and occasionally bees for their honey. The other animals kept (although not by all families) were mules, donkeys and oxen, the latter being bought at the start of the ploughing season and resold at its end.
Thus the people of Greater Kabylia in the pre-colonial period were self-sufficient in fruit and vegetables and (given their modest consumption) meat, poultry and eggs and produced an important surplus of olive oil for sale outside the region. With the exception of the narrow plains between Draa el-Mizan and Wadhia and between Tizi Ouzou and Azazga, however, the region produced no cereals and depended upon the Arabophone regions to the south for its supplies of the corn and barley upon which it relied for the staple elements of its diet, bread and seksu (cous-cous). It obtained the cereals it needed in exchange for figs and olive oil and, above all, the products of craft manufacture.
Much of Kabyle artisan production involved the exploitation of natural resources available in the region itself. Wood from the forests of the Jurjura and Akfadou and elsewhere was used in making bowls, caskets, chests, looms, paddle-wheels for water mills, ploughs, screws for olive presses, tables, tool handles and the wooden pillars, beams and doors used in house building. Much of the timber used at Algiers for boat building in the Turkish period and earlier came from the forests of Kabylia as well, as did the resin and pitch that was also required. Thus joinery and carpentry were widely practised and most villages would have a specialist or two in these crafts.
Figure 2.2: Colonial-era postcard showing an olive-oil mill in Kabylia.
The land itself supplied other materials for craft production: stone for house building, but also for the making of small flour mills and grinders as well as the large mill-stones in use in the numerous olive-mills and water mills in the region. From the earth a wide range of pottery goods was produced including most of the vessels used in each household: bowls and dishes, the huge jars known as ikufan in which grain and dried vegetables and fruit would be stored, and pots of various shapes and sizes.
A great deal of Kabyle craft manufacture relied on materials obtained elsewhere, however. Woolen goods, particularly burnouses (men’s cloaks) and coverlets and various types of clothing, were made from wool bought from the markets of the High Plateaux to the south. Some of the flax used in the production of linen cloths was grown locally, but much of the linen also had to be imported. The charcoal used in the manufacture of gunpowder was produced locally, but the sulphur and part of the saltpetre had to be imported from outside. The oak forests, especially in the Akfadou district, were the source of the tannin used in the preparation of leather skins.
In addition to this already impressive range of artisan production, the tribes of the central Jurjura engaged in three closely linked and highly specialised branches of craft industry: jewelry making, arms manufacture and the manufacture of counterfeit money. Of these three, only jewelry manufacture appears to have survived to this day and is almost entirely confined to certain villages of the Ath Yenni tribe, especially Ath Larba‘a, and the tribe of Iwadhien (Wadhia), although the craft of gun-making appears to have survived at least until the 1940s in the village of Darna of the Ath Boudrar tribe,17 and possibly elsewhere as well. In the pre-colonial period and probably well into the latter half of the nineteenth century, nearly all the tribes of the central Jurjura engaged in these branches of craft production, although the Ath Yenni were already pre-eminent before 1830. The weapons manufactured were rifles and pistols, sabres and daggers, as well as cutlery.18 Counterfeit coins were the speciality above all of the village of Ath Larba‘a; they were made to order and orders reputedly came from as far away as Morocco and Tunis, the Sahara and Tripoli.19 The importance of this activity was recognised by the village, which made provision for its regulation in its local by-laws.20
Evidently a great deal of the craft manufacture which has been described above was not for domestic consumption but for the market and for export beyond the confines of the region. Greater Kabylia depended on a wide range of commercial exchanges with the other regions of the Algerian interior. It exported carobs, figs, fruit, grapes, olive oil, red pepper, sweet acorns, arms, jewelry, wooden utensils, clothes, leather goods of various kinds, linen cloths, pottery, soap and wax. It imported corn and barley, bullocks for ploughing and also for meat, milk cows, mules, sheep, cotton, silks, wool, copper, iron, lead and tin.
The distribution of Kabyle craft products was undertaken by the producers themselves, who would transport their goods on the backs of mules and donkeys to the markets of Algiers and the other towns but also further afield, to wherever there was a market for them. Sometimes this trade was conducted in groups; the tribes of the Ath Bethroun and Ath Menguellat confederations of the central Jurjura formed regular caravans for the merchants from their constituent villages who wished to sell their goods in the Arab interior.21 Sometimes, however, merchants would travel alone and so approximate in appearance at least to that other characteristic figure from Greater Kabylia, the pedlar.
The Kabyle pedlar was a familiar figure in the Algerian interior in the pre-colonial period and well into the colonial period. Indeed, it is likely that, until comparatively recently, most Arabic-speaking Algerians of the countryside formed their opinion of the Kabyles in general from their encounters with Kabyle pedlars and this opinion was a distinctly unfavourable one. Although commercial craft manufacture was widely developed in the Jurjura, it is probable that only a minority of families engaged in it. Poorer Kabyles, unable to trade in substantial commodities of their own making, relied on long-distance peddling to supplement the meagre income from farming. They would equip themselves in Algiers or other urban centres with spices, perfumes, charms and trinkets of all sorts and then travel through the Arabophone regions and among the Chaouia of the Sud-Constantinois and act as intermediaries between these regions, re-stocking their portfolios as they went.
These pedlars were known in Kabylia as i‘attaren, which means literally ‘purveyors of perfume’ (from the Arabic: ‘attar) but had perhaps already come to acquire the additional meaning that Berque notes for the word in the twentieth century, that of sellers of groceries.22 The Kabyles distinguished between the i‘attaren ou kidhoun, ‘the pedlars with tents’, and the i‘attaren ou ketsaf, ‘the pedlars with sacks’.23 The former travelled from market to market, transporting their wares on mules and passing the night under their own tents. The latter went from door to door on foot, carrying their wares in large sheepskin knapsacks, and relied on their customers to provide them with shelter for the night.
The resort to long-distance peddling was a mass phenomenon in nineteenth-century Kabylia. Hanoteau and Letourneux noted that ‘in the circle of Fort National alone, out of a population of 76,616 inhabitants, between 8,000 and 10,000 passports are issued each year to traders, of whom at least three quarters are pedlars’.24
The extent of Kabyle dependence upon this commercial migration was reflected in Kabyle law. For example, the third article of the qānūn (by-law) of the village of Koukou of the Ath Yahia tribe stipulated that ‘those absent from the village who are at Constantine or Souk Ahras have two months in which to exercise the right of shefa‘a’,25 while the second and seventh articles of the qānūn of the village of Taourirt Amrane of the Ath Bou Youcef tribe read as follows: ‘the man who disparages an inhabitant of the eastern province pays a fine of 5 reals’ and ‘the man who flees the eastern province leaving debts of more than one real pays a fine of one real’.26
It was also noted by the French administrators of the nineteenth century. Thus we read in official reports on individual tribes that ‘the Beni Bou Attaf’ (i.e. the Ath Attaf or Yattafen) ‘devote themselves to the cultivation of the olive tree and the fig; they also engage in peddling and go periodically to trade with the populations of the interior’.27 As for their neighbours, the ‘Beni Bou Drar’ (i.e. the Ath Boudrar or Iboudraren),
the inhabitants are active and industrious. A proportion of them devote themselves to the occupation of pedlar; some exercise the profession of armourer, jewelers and turners; the rest engage in cultivation.28
Perhaps the best illustration of the importance of this commerce is provided by the tribe of the Ath Bou Akkach, neighbours of the Ath Boudrar on their western side, and noted for their large and extremely beautiful villages, of whom Émile Masqueray commented in 1886,
external trade is the sole cause of this extraordinary prosperity. You will meet there, from spring to autumn, only women, old men and prodigious numbers of children. The able-bodied men are all away … Most of them are in the province of Constantine. They stock up with goods from the Jews and Mozabites of Souk Ahras, load thin mules with spices, calico, scarves, oranges – everything which may please the Nomads and, above all, their women and head south … One can meet them among the Nememcha and at Negrine, and even further afield …29
The extent to which the Kabyle pedlars exploited the susceptibilities of their women customers in their trading ventures undoubtedly contributed to the low opinion in which they were widely held dans le pays arabe. As Hanoteau and Letourneux noted,
the continual relations of the iattaren with the Arab women greatly facilitated their engagement in amorous intrigues; but these were fatal for many of them. It is very rare that there is not, each year, occasion to record several catastrophes that have befallen pedlars who have been too gallant.30
If Bourdieu in his account of Kabyle society makes no mention of Kabyle craft manufacture and commercial migration, this may be because he carried out his fieldwork in the region during the war, by which time capitalist development in Algeria, linked as it was to that of the metropolis by the customs union, had largely destroyed the indigenous artisanate, not only in Kabylia but also in the towns, and the demand for cheap labour in France from 1914 onwards had brought about the reorientation of Kabyle emigration from the Algerian interior to the factories of the metropolis. While Bourdieu’s picture of the Kabyle economy is understandable, then (although still misleading), it bears little relation to the situation in the pre-colonial period or, indeed, in the colonial period up to c. 1939.31
A very different understanding of the matter is suggested by the Algerian historian Mohammed Harbi. If Bourdieu’s account would lead us to infer that the colonial period had seen the disappearance of Kabyle craft manufacture and commerce, Harbi implies the opposite. In seeking to explain the development of a ‘Berberist’ tendency within the nationalist movement in the 1940s, he remarks
the particularism is even more pronounced in Greater Kabylia, a rural region withdrawn in a defensive posture and closed in on itself. With the advent of colonial rule, the sentiment of regional identity becomes important and acquires a new content. Thanks to the multiplication of commercial exchanges as a result of capitalist penetration, little isolated and closed universes open up to the ideas of the world outside. Sanctuary of poverty, Kabylia to survive exports its men to France and to the Algerian interior …32
In other words, Greater Kabylia was a closed world, rural in character, which only opened up in response to the stimulus of colonialism. The development of Kabyle commerce and the associated migratory activities inside Algeria, instead of antedating colonialism, were a by-product of the capitalist development it ushered in.
Again, this picture is misleading but it is understandable that Harbi should have believed it to be true. While the people of Greater Kabylia and the Igawawen above all were not merely open to the outside world but dependent upon their intercourse with it and moved freely around it, from the point of view of an Arab Algerian (such as Harbi), the region certainly appeared to be closed to outsiders. There was no reciprocal movement of Arabic-speaking Algerians into the region to balance the outflow of merchants, pedlars and labour migrants from it, for there was no economic basis for such a movement, and the society of the Jurjura, being already overpopulated and jealous of its independence, has long evinced a predominantly suspicious and sometimes unwelcoming if not hostile attitude to strangers, wherever they may come from. But the region was closed only in this sense and in other respects was permanently interested in and oriented to the outside world.
Second, Harbi’s claim that it was the colonial impact and the development of the capitalist market economy which gave rise to Kabyle migration into the Arabophone interior ignores the pre-colonial trading and migratory traditions of the Kabyles. But there is nonetheless a germ of truth in his statement. For, whereas the long-term effect of the development of colonial capitalism was to undermine traditional craft manufacture and the associated trading circuits, it is likely that, as Morizot has suggested, the initial short-term effect was to stimulate them both, in so far as the enhanced security in the Algerian interior consequent upon the establishment of the colonial order combined with the development of communications facilitated and stimulated Kabyle commerce and the artisan production associated with it.33 Indeed, Hanoteau and Letourneux cite instances of new productive ventures developing in Kabylia in response to the European market as early as 1867 (within ten years of Icherriden), notably the cultivation of grapes on a much greater scale than before for sale to colons for wine-making.34
But, while Kabyle enterprise and commerce may well have expanded under the stimulus of the colonial impact, they were not created by it. A capacity to respond to new market opportunities when these appear is characteristic of a culture which is already oriented towards commercial and productive enterprise. Such things cannot be created or implanted in the space of a decade. And the existence of the most important kinds of Kabyle craft manufacture, notably jewelry, arms manufacture and coining, and of Kabyle migration in the Algerian interior, is well documented for the Ottoman period.
Commercial craft manufacture and long distance trade and peddling were not typical of Greater Kabylia as a whole, but only of the tribes of the central Jurjura, the Igawawen. Maritime and western Kabylia were less densely populated and their populations were better able to support themselves by their farming activities. These were certainly not sufficient, however, but the activity by which they were supplemented was labour, rather than commercial, migration. The main form was seasonal migration to assist the harvest in the areas of cereal cultivation. The men from western Kabylia would mostly head for the Mitija plain to the south of Algiers and to the Titteri, while those from other districts would head south to the Bouïra area and much further afield to the high plains of the Setifois, the Constantinois and the Sud-Algérois. In addition, many Kabyles, especially from maritime Kabylia, found work at Algiers or in its suburbs, as masons, navvies, gardeners, domestic servants and day-labourers. The bath attendants in Algiers were traditionally from the Ath Jennad confederation to the north of Azazga.35
Finally, the men of Greater Kabylia also found employment and notoriety as mercenaries. The French word zouave, meaning originally a native light infantryman, is a derivation of zwawi, but here as in so many other respects France merely took over an established tradition. In his Annales Algériennes, Pélissier de Reynaud noted that ‘the Zouaves or Zouaoua are independent Kabyles of the province of Constantine who sell their services to the Barbary powers as do the Swiss in Europe’.36 As Charles-André Julien has observed, the Ottoman rulers of Algeria
soon came to realise that their authority stood in greatest danger not from their subjects, but from the janissaries. Accordingly they endeavoured to form an equally valiant but more reliable army from Kabyle, principally Zwawa, contingents.37
This policy, first tried under the beylerbeys in the sixteenth century, was also adopted by the penultimate Dey of Algiers, Ali Khodja, on his accession in 1817, when he quickly formed an ‘honour guard’ of 2,000 Zwawa ‘which was charged with both guarding his person and seeing to the strict execution of his decisions’.38 Following his death in 1818, his successor Husseïn Dey retained the services of the Zwawa contingent he had formed.39 Ahmed, the last Bey of Constantine, followed the same policy, his infantry being composed partly of Turks and partly of Kabyles.40
The mercenary activities of the Zwawa were not confined to Algeria. The Ottoman rulers of Tunis also employed Zwawa, and the Tunisian case gives a striking indication of the importance of the Zwawa forces. According to L. Carl Brown, there were between 1,500 and 2,500 Zwawa in the Tunisian army, in the infantry, in the period 1837–55. (Since these figures refer only to the number actually registered, it is possible that the true number was higher.) They were drawn both from Zwawa immigrants and from the permanent Zwawa community of Tunisia, which numbered some 20,000 at this time.41 This community had its own patron saint, Sidi Mohammed El-Bachir, who died in 182742 and Husseïn Bey built Sidi Mohammed El-Bachir a zāwiya (religious lodge). As Brown remarks, ‘the beys understandably courted the man (and his successors) who commanded the religious fealty of such an important part of the army’.43
The people of Greater Kabylia were far from being ‘primarily arbori-culturalists’ in the pre-colonial period. They engaged in a wide range of productive and commercial activities and these activities took them in large numbers to all parts of Algeria and even further afield. There can be no doubt that it was their success in developing a skilled and diversified craft industry and in exploiting the range of commercial opportunities open to them in the interior of the country which enabled them to sustain the astonishing density of population which already characterised the region as a whole and the Jurjura in particular by the time the first French troops disembarked at Sidi Frej in 1830.
The development of this remarkable economy was uneven, however. As we have seen, the people of western and maritime Kabylia relied largely on labour migration to supplement their farming activities, whereas the Igawawen of the Jurjura relied on commercial craft manufacture and peddling and the much greater returns from these trading activities supported a correspondingly higher density of population. Thus the cultural dichotomy between the Igawawen and the inhabitants of the rest of Greater Kabylia had a clear economic content.
But if the villages of the Jurjura could not have supported their populations without the recourse to external trade and emigration on a large scale, it is also true that they could not have sustained these activities had it not been for the particular qualities of their socio-political organisation.
Morizot emphasises the role of the patriarchal family in this context.44 Within the family there was a strict and elaborate division of labour: the first division was between those who stayed at home in the village and those who were delegated by the family to go abroad. The latter were uniquely males and a young man setting out on his first trip was obliged to marry beforehand to ensure his eventual return, the woman being on no account permitted to leave the village. In addition, the austere sexual mores of the Kabyles and the constant surveillance of each woman by her family and by village public opinion allowed her husband to remain abroad for long periods in total confidence that, on his eventual return, he would find his household as he had left it. But there are grounds for reconsidering Morizot’s view of this matter, for this feature of the division of labour only developed in the colonial period, essentially from 1914 onwards, in the context of the reorientation of Kabyle labour migration to France, when it was grafted onto an underlying and far older sexual division of labour which Morizot does not discuss.
Similarly, Morizot also emphasises the significance in this context of the traditions of mutual aid and village solidarity, the extent to which Kabyle traders and pedlars in foreign parts could count on the support of fellow villagers should they ever find themselves in difficult circumstances, as not infrequently occurred.45 Since men from a particular village tended to migrate to the same regions of the interior and of France, such solidarities would frequently be invoked. Here again, it is only one, external, aspect of the role of the village, the solidarity it engendered among its members when abroad, that is mentioned; yet this solidarity had its origin in the character of the village itself on its own ground.
Morizot has, with reason, described Kabyle society as ‘a village society’ (‘une société villageoise’),46 classifying Kabylia together with, among others, the Aurès, the Mzab and the Wad Ghir in this respect, in contrast to those other regions of the Algerian countryside where dispersed settlement or transhumant pastoralism have been the rule. But it is not only a society of villages. A condition of the existence of the Kabyle village is its membership of a larger whole.
Every Kabyle village without exception belongs, or at any rate in the pre-colonial era used to belong, to a community known as an ‘arsh.47 This term is conventionally translated as ‘tribu’ in the French and ‘tribe’ in the English literature, a usage which, for brevity’s sake, I have myself followed thus far. But this convention encumbers the word with connotations that are out of place and misleading.48
The word tribe is used for populations in North Africa and the Middle East that are quite unlike those designated as ‘tribes’ in, for example, sub-Saharan Africa or Amazonia, where a number of salient features such as habitat, ecology, language, religion, socio-political organisation, diet and even physiognomy usually distinguish one population quite sharply from another (e.g. the Kikuyu and the Masai in Kenya) in a way and to a degree that is self-evidently untrue of the ‘tribes’ of North Africa and the Middle East. Such features as distinguish tribes elsewhere distinguish much larger and looser categories of population in North Africa and the Middle East – Arabs, Berbers, Kurds, Turks, Persians, Pashtun, etc. – and also, if less sharply, the regional or other historically formed identities into which these population categories may be sub-divided (e.g., for Berbers: Rifians, Imazighen, Chleuh, Kabyles, Chaouia, Mzabis, Tuareg, etc.), but not the so-called ‘tribes’ into which these are further sub-divided. It is, of course, widely understood that, in respect of the features (ecology, language, religion, etc.) instanced above, one Kabyle or Rifian (or Beduin or Kurdish etc.) ‘tribe’ is generally very like another. But, while ‘tribe’ does not, therefore, usually connote a distinctive and exclusive ethnic or cultural identity in the North African context, it does connote at the very least a specific kinship identity, and this is the problem. It is in part because it has been taken for granted that the Arab and Berber populations of the Atlas mountains and the high plateaux and deserts of the Maghrib consist of ‘tribes’ and that these are essentially extended kinship groups that the segmentarity thesis of Berber (but also Beduin49) politics possessed its initial plausibility.50 But, in Kabylia at least, the ‘arsh is not a kinship group and this fact matters.
The Arabic word ‘arsh (Arabic plural: ‘urūsh or ‘arūsh; Berber plural: ‘aarsh) is derived from the root: ‘-r-sh which, as a verb (‘arasha), has the meaning ‘to erect a trellis’ and a secondary meaning of ‘to roof over’.51 One of the nouns derived from this root, ‘arīsh, has the meanings of ‘arbour, bower, hut, trellis’,52 and is the name of a town in the northern Sinaï, El Arish, close to Egypt’s border with Gaza. Another, which probably also subsumes the notion of a deliberately created structure, has given rise to the name of the town of Larache (El-‘arash), at the other end of North Africa, on the Atlantic coast of northern Morocco.53 In the Arab Mashriq, the substantive ‘arsh derived from this root means ‘throne’, the seat of sultans and kings, sovereign rulers. It is accordingly a mystery why this word should have come to be used in Algeria to refer to the discrete communities of which the populations of the countryside have traditionally been composed.
This mystery can been seen to be all the deeper in the light of the fact that the Arabic term translated as ‘tribu/tribe’ in the Moroccan case is not ‘arsh but qabīla (plural: qbā’il),54 which is the standard term for ‘tribe’ in the Mashriq as well; the Beduin ‘tribes’ of Libya, Egypt, Sudan, Jordan and Saudi Arabia are all called qbā’il.55 In addition, the word qabīla has certainly long been known to and employed by Algerians; as already mentioned, the standard term used by the townsfolk of pre-colonial Algeria to refer to the populations of the mountains was el-qbā’il – ‘the tribes’ – whence the derivation of the words Kabyle and Kabylia. Finally, as if this were not enough, the word qabīla was not only known to the Kabyles themselves but employed by them in its Berber form, thaqbilth, to refer to the highest level of political unity which the scholarly literature has called ‘the confederation’. Why, then, should what most observers call the ‘tribe’ in Kabylia and, for that matter, the rest of Algeria, be known as ‘arsh?
To answer this question we should break it down into two parts. Why ‘arsh and not qabīla? And what is the etymological significance of ‘arsh in this context? The answer to the second question will indicate the answer to the first.
Qabīla is derived from the root q-b-l which, as the verb qabila, has the basic meaning of accepting, receiving, admitting, but the derived meanings of ‘to stand exactly opposite, face, confront or counter’.56 In Algerian Arabic and in Thaqbaylith (which has borrowed it), the adverb qabel can mean ‘face to face’, ‘frontally’, ‘directly’, ‘straightforwardly’, ‘frankly’ etc. and carries a positive value in the context of the code of honour.57 Thus the idea at the root of the word qabīla is that the populations in question are divided by a number of oppositions between them, that it is the opposition between Groups A and B and between Groups C and D (and so on) that gives them their stable identities.58 We should note that there is no reference to kinship or shared descent in this root meaning. What defines a qabīla is not the degree of real or putative (or entirely fictive) descent its members may (or may not) share from a distant ancestor, but its relations – primarily of facing, opposing, competing – with other similar groups. Thus it is the political aspect that is constitutive of the community. Since it is precisely my argument that it is equally the political that is constitutive of the Kabyle ‘arsh, it remains to consider why this is not, like its Moroccan (etc.) counterpart, also called a qabīla.
Let us rephrase this question. Why should the Arabic word for throne have been employed to indicate a particular kind of community in Algeria?
A throne is the seat of a sovereign ruler. What materially characterises a throne in the North African context is not merely the distinctive splendour of this item of furniture, its size and weight and the quality of the materials used to make it and so forth, but also two features which distinguish it from ordinary chairs: the fact that it has a high back and that it also, at least usually, has a kind of quasi-ceiling or canopy protruding from the top of its back above the person sitting on it that at least partially covers and protects him. Thus the idea of a throne incorporates two ideas that belong to different registers: the basic idea, intrinsic to the root ‘-r-sh, of a deliberately erected superstructure and the derived idea of a seat of sovereignty. Both of these ideas enter into the way the word ‘arsh has historically been used in Algeria to refer to a particular kind of community.59
The ‘arsh in Algeria in general and in Kabylia in particular was a sovereign political community collectively possessing and controlling a definite territory.60 The community was sovereign and its territory was its seat. The territory of an ‘arsh could not be intruded upon with impunity; outsiders needed a laissez-passer, the ‘anāya (guarantee of protection) of a leading member of the ‘arsh, in order to enter and travel across this territory in safety. The leadership of the ‘arsh, whether an egalitarian assembly as in most of Kabylia or an aristocratic dynasty as obtained elsewhere, was the sovereign government of the community, the upholder as well as partial source of the laws which regulated the social life of its inhabitants and the allocation and enjoyment of rights in its land. The territory of a Kabyle village, both its commons – pasture, woodland and watercourses – held collectively (meshmel) and the arable land held in individual private ownership (lemluk), was part of the collective territory of the ‘arsh and access to it presupposed membership of the ‘arsh. The collective capacity of the population of the ‘arsh to defend its territory as a whole was the guarantee of the security of its constituent villages and of the integrity of their individual territories. It is here that the idea of the superstructure can be discerned, in that the ‘arsh existed over and above its constituent villages and provided protection for them all.
Sovereignty was not the exclusive monopoly of the ‘arsh; the latter did not have absolute authority in relation to its constituent villages, for each village also enjoyed its measure of sovereignty and so its authoritative capacity to make its own law. Rather than concentrated in a single locus, sovereignty was diffused between the two levels of ‘arsh and constituent villages. But the ‘arsh was normally the upper level and limit of sovereignty; while some ‘aarsh might belong to a higher-level grouping, the thaqbilth, this was optional and many did not. In Lesser Kabylia there have never been any thiqbilin.
In one respect, the relationship of village and ‘arsh appears to be circular. A Kabyle was a member of ‘arsh X because a member of a constituent village of this ‘arsh and membership of ‘arsh X was simultaneously a political condition of the existence of the village in question. No village might exist outside an ‘arsh. Proof of this is furnished by the perfect limiting case. There is, in Greater Kabylia, a large and well-known village that is unquestionably a Kabyle village (as distinct from an Ottoman-sponsored settlement of outsiders) and which exists in both spatial and political isolation, all on its own. This is Ighil Imoula, where the historic Proclamation of 1 November 1954 was printed. Instead of being an exception to the rule, however, Ighil Imoula confirms it, for it has been its own ‘arsh: that is, Ighil Imoula is an ‘arsh consisting of a single village of the same name and the sole constituent village of an ‘arsh of the same name.
The notion, implicit in the root meaning of ‘arsh, of a deliberately erected superstructure reflected the fact that the community in question was not constituted by kinship, by shared descent from a common ancestor, but by political agreement. While many ‘aarsh in Kabylia, as elsewhere in Algeria, had names that suggested a shared kinship: Ath Yahia (‘the sons of Yahia’), Ath Mahmoud (‘the sons of Mahmoud’) and so on, at least as many did not. Often an ‘arsh name evoked the most salient feature of its territory: Ath Boudrar (‘the people who possess the mountain’), Ath Wasif (‘the people of the river’), Ath Oumalou (‘the people of the shadow’ – that is, of the north-facing slopes), Ath Ousammeur (‘the people of the sunlight’ – that is, of the south-facing slopes); Ighil Imoula means ‘the ridge of the forest’. And where the collective name might be interpreted as implying shared descent, this was not to be taken literally. At most it commemorated an individual credited in local legend with establishing in the dim and distant past the original nucleus of the ‘arsh in question in the shape of the founding settlement, around which the other constituent settlements subsequently gravitated or coalesced.
A corollary of the fact that the Kabyle ‘arsh was not a kinship group, that shared kinship was not constitutive of the community even where the idiom of kinship was vaguely employed to provide a given ‘arsh with its collective name, is that, for each of its villages, membership of the ‘arsh was voluntary or, at any rate, the result of a political decision (whether strictly voluntary, as opposed to coerced, or not) and therefore conditional. Although most Kabyle ‘aarsh have displayed impressive continuity and stability in their size and membership, it was not at all unheard of for villages to switch their allegiance to or be annexed by another ‘arsh. It is well known that the villages of Tassaft Waguemoun and Ath Eurbah in the central Jurjura originally belonged to an ‘arsh (now defunct) called Ath ou Belkacem before joining the Ath Wasif, as did the village of Taourirt el-Hajjaj (then called Thakhabit) before joining the Ath Yenni and the village of Ath Ali ou Harzoun before joining the Ath Boudrar. Among the Ath Mahmoud, the oral tradition records that the villages of Taguemount el-Jedid, Tala Khelil and Tighilt Mahmoud all belonged to the ‘arsh before being ‘surrendered’ at different moments to its hostile neighbours, the Iwadhien, Ath Douala and Ath Zmenzer respectively.61 And the remote village of Ath Waaban in the heart of the Jurjura is another case in point. Originally founded by migrants from the Ath Wakour south of the watershed, the village subsequently severed its connection with its ‘arsh of origin and secured incorporation into the Ath Boudrar.62 It is likely that other cases could be identified if the relevant research were to be conducted.
Precisely how Kabyle political organisation at the level of the ‘arsh was articulated with that at village level will be considered in a later chapter. But the fact that this extended form of human settlement was – and still is – known as ‘arsh, not qabīla, reflects a fundamental way in which Algeria has differed historically from Morocco and, for that matter, most other countries of North Africa and the Middle East, namely the remarkable degree of independence (if not preponderance) of the society of the mountains in relation to the towns, an independence that the Ottoman Regency struggled against with difficulty, which French colonial rule began only partially to modify and contained rather than abolished and which was subsequently reflected in the fact that the revolutionary movement which fought the war of national liberation and constituted the independent state was based overwhelmingly on the society of the countryside and of the mountains above all.63
To refer to the extended forms of settlement of the mountains as el-qbā’il is to employ the vocabulary and adopt the frame of reference of the urban population. The conception of these communities as qbā’il, as groups constantly opposed to and confronting one another, was the corollary of the negative conception of the society of the hillsmen ‘beyond the pale of civilisation’ as essentially anarchic. The existence since the ninth century CE of a comparatively stable state tradition in Morocco based on and controlling the society of the Atlantic plan and promoting the development of flourishing cities (Fès, Meknès, Marrakesh) ensured that urban conceptions of the society of the mountains have long prevailed there. But the configuration of the relationship between mountains and high plateaux on the one hand and coastal plains on the other in Algeria has been very different, and there was no tradition of continuous state formation in the lowlands until the consolidation of Ottoman power in the sixteenth and seventeenth centuries CE. A consequence of this very different history is that the hillsmen’s own, essentially positive, conception of their extended forms of settlement has historically prevailed over that of the urban populations.
This conception – of sovereign communities – reflected the fact that, however forbidding it might appear to the town-dwellers, the society of the mountains was not anarchic. The territories of the ‘aarsh were governed spaces.
Generalisations about the economic life of Greater Kabylia have had the effect of obscuring the dichotomy which existed between the Igawawen and the people of the rest of the region. Generalisations about Kabyle social organisation have had the same effect. This is especially true of the numerous attempts to explain that enigmatic abstraction, ‘the Kabyle village’.
The ‘aarsh of Kabylia come in a wide range of sizes but they have in common the fact that they are composed of units which are conventionally referred to as ‘villages’. This usage, which was established by the nineteenth-century French observers, abstracts from an important difference between the two kinds of constituent unit in question, the village proper, in the sense of a self-governing and highly integrated settlement, thaddarth, and the association of scattered hamlets, tūfīq. While most writers have noted the existence of these two forms, they have invariably taken the thaddarth as the standard form and based their interpretations of Kabyle social organisation upon it and have for practical purposes ignored the tūfīq altogether. This has been a great mistake.
In order to rectify this mistake, we must first discuss with more precision than we have needed hitherto the terms we are using. Three terms and their meanings need to be explained: thaddarth, tūfīq and thakhelijth.
Thaddarth, plural: thiddar, is a Berber word which means a place of habitation or residence. In parts of Morocco it is used to mean ‘house’ but it is not used in this sense in Kabylia. This is because, in Kabylia, houses as a rule are not isolated places of habitation, but parts of larger agglomerations. It is the agglomeration or settlement as a whole which is designated by the term thaddarth. The Kabyle word for house and, by extension, household or family, is akhkham.64
Thiddar can in principle be of any size. A thaddarth may consist of a few houses totalling a hundred people or even substantially less. Or it may consist of two or three hundred houses belonging to as many as twenty different lineages which in turn belong to several clans, giving a total population of a thousand or even, these days, several thousands. The Kabyle language makes no distinction between these two cases, yet they are very different in important ways. In the first, the thaddarth is almost invariably a kinship unit; in the second, it almost invariably is not. This important difference is connected with other differences as we shall see.
Tūfīq, plural: tuwāfeq, does not mean a place of residence or settlement or agglomeration. The Arabic word tawfīq has a number of closely related meanings, including ‘adaptation, adjustment, reconciliation, arbitration, peace-making, re-establishment of relations’. In Kabylia, tūfīq designates a group of usually small thiddar which are dispersed in space but associated with one another to constitute a political unit, this association being expressed in the institution of an assembly, jema‘a, of the tūfīq as a whole, in which each constituent thaddarth is represented. (It should be noted that, although the thiddar are dispersed in space, the territory they collectively inhabit forms a continuous whole and is inhabited by no groups outside the tūfīq in question.)
Thus the word thaddarth can refer either to a large village inhabited by numerous lineages divided into several clans which are rarely if ever descended from a common ancestor, and which is entirely autonomous subject to the conditions of its membership of the ‘arsh, or to a much smaller settlement, composed of closely related families, which usually forms a dependent part of a larger association of such settlements and is subject to the collective decisions of this association. In the second case, the settlement in question, although a thaddarth in the sense of agglomeration of houses, is more generally known as a thakhelijth.
Thakhelijth, plural: ikhelijen,65 is a Berber derivation from the Arabic khalīj, one of the meanings of which is ‘canal’ or ‘channel’. The metaphor of a channel is employed in Kabylia to designate a social group which is a dependent part of a larger whole. (In Morocco, anatomical metaphors are often used: bones, limbs and so on.) The small thiddar which compose a tūfīq are its ‘channels’ in the sense that the collective life of the association and the collective decisions of its jema‘a are transmitted to the individual households which form the base of the community through the intermediary of the constituent ikhelijen. The English word which best corresponds to this meaning (without, however, wholly doing so) is ‘hamlet’.
But thakhelijth is also used to refer to the component parts of a large autonomous thaddarth. While this kind of thaddarth is composed of several clans, themselves composed of a number of lineages, it is divided spatially into a number of areas which correspond broadly to the spatial distribution of the clans, although this correspondence is not absolute. Where these areas are clearly separate from each other on the ground, they may also be referred to as ikhelijen. Thus the thaddarth of Ath Waaban is composed of two ikhelijen, thakhelijth n’Ath Tighilt and thakhelijth n’Ath Wadda,66 each of which is occupied by two clans (iderman). The English word which conveys this meaning is ‘quarter’. Thus the meaning of thakhelijth, like that of thaddarth, depends upon the context. It should be noted that there are thiddar which have taken the process of integration to the point where the buildings form a virtually continuous whole, such that there is no clearly visible division of the village into two or three quarters, merely narrow streets and alley-ways. At Taguemount Azouz, for instance, which is traversed by two long narrow streets (azniq oufella, ‘the upper street’, and azniq bouadda, ‘the lower street’), the inhabitants do not speak of this or that thakhelijth, but of successive sections of each azniq, each section being associated with and bearing the name of a particular lineage – e.g. azniq Ath Lmessaoud, azniq Ath Bouzid, azniq Ath Chemloul – while passing their houses.67
Like thaddarth and thakhelijth, the term tūfīq also has acquired two meanings. In Kabylia, tūfīq is, as I have stated, an association of a number of small hamlets, roughly equal in size, at least to the extent that no one constituent settlement dominates the tūfīq as a whole. However, both Hanoteau and Letourneux and, following them, Masqueray apply the term tūfīq on a number of occasions to the very different situation where a large complex thaddarth possesses outlying smaller settlements. In these cases, the tūfīq in question is unequivocally dominated by the principal thaddarth, and it is more appropriate to regard the social unit in question as a large thaddarth plus satellite settlements than as a tūfīq in the normal sense of the term.68
There are thus three main kinds of autonomous unit below the level of the ‘arsh to be found in Greater Kabylia: the small thaddarth, comprising often less than a hundred and rarely more than two hundred people who all belong to the same lineage or clan and which is therefore a kinship unit; the tūfīq, a group of small thiddar associated for political purposes in one unit of which they are the dependent hamlets; and the large thaddarth, comprising many hundreds and sometimes more than a thousand people who belong to distinct clans and do not pretend to be descended from a common ancestor, and which accordingly derives its unity from something other than kinship.
The internal political organisation of Kabyle ‘aarsh varies considerably, depending on which of these forms of settlement prevails. This can be illustrated by comparing an ‘arsh from north-western Kabylia, the Ath Slegguem, an ‘arsh from north-eastern Kabylia, the Ath Adas of the Ath Jennad confederation, and an ‘arsh from the Jurjura, the Ath Wasif of the Ath Bethroun confederation.69
Table 2.2: ‘arsh Ath Adas (confederation of Ath Jennad)
Table 2.3: ‘arsh Ath Wasif (confederation of Ath Bethroun)
These examples have not been chosen at random. In north-western Kabylia, the small isolated thaddarth prevails. In the Jurjura, the large complex thaddarth prevails; the thiddar of the Ath Wasif are on average nearly ten times larger than those of the Ath Slegguem. In north-central and north-eastern Kabylia, neither the small isolated thaddarth nor the large complex thaddarth is to be found, except occasionally; here the prevalent form is the tūfīq. But the tūfīq actually prevails in all parts of Greater Kabylia other than the western and north-western marches at one extreme and the Jurjura at the other.
Hanoteau and Letourneux provide detailed data concerning the distribution of the population in the four cercles (administrative sub-divisions) of Greater Kabylia c. 1868, by confederation (thaqbilth), tribe (‘arsh) and village (thaddarth and tūfīq).70 The cercles were those of Dellys (western and north-western Kabylia), Draa el Mizan (south-western Kabylia), Tizi Ouzou (northern, north-eastern and part of central Kabylia) and Fort Napoléon (renamed Fort National after 1870, covering the central and eastern Jurjura, its northern outcrops and the southern Akfadou district). If we examine this data in aggregate form in the light of our distinction between the three zones characterised by different settlement patterns, we can arrive at a rough estimate of the relative predominance of the two kinds of thaddarth and the tūfīq in the three zones in question. This is shown in Table 2.4.
Table 2.4: Regional distribution and relative significance of thiddar and tuwāfeq in Greater Kabylia c. 186871
Thus in the central and eastern Jurjura in the second half of the nineteenth century, thiddar outnumbered tuwāfeq by two to one and accounted for 60 per cent of the population as against 40 per cent, whereas in the rest of Greater Kabylia, apart from the western and north-western marches, tuwāfeq outnumbered thiddar and accounted for over twice as much of the population – 70 per cent as against 30 per cent. Moreover, the thaddarth of the central Jurjura was on average four times larger than its counterpart in the western and north-western marches, and 32 per cent bigger than in the rest of Greater Kabylia. Furthermore, since, as has already been stated, Hanoteau and Letourneux classify as tuwāfeq many large thiddar in the central Jurjura possessing satellite hamlets, the above figures seriously understate the prevalence of the large thaddarth in the central Jurjura and the peculiarity of the central Jurjura in this respect.72
Twenty-one of the thiddar listed by Hanoteau and Letourneux were recorded by them as possessing populations of 1,000 or more c. 1868. Sixteen of these belonged to the ‘aarsh of the central and eastern Jurjura or its northern outcrops and these figures are also underestimates, for the same reason.
The movement in space from the western and north-western periphery of Greater Kabylia to the region’s core in the central Jurjura was a movement from less integrated forms of settlement and socio-political organisation to more integrated ones. The tūfīq represents a higher level of integration than the isolated small settlements of the marcher tribes, and the large thaddarth of the central Jurjura represents a higher level of integration than the tūfīq. And at the same time the movement from north-west to south-east was one from populations which were broadly incorporated into the political system of the Ottoman Regency through the agency of local dynastic leaderships to populations which were intermittently dissident or, in the Jurjura heartland, permanently independent of the Regency in respect of their political arrangements and governed themselves by means of representative assemblies and so without recourse to self-perpetuating hierarchies or the dynastic principle.
There is an evident irony in the fact that the nineteenth-century French ethnographers and their successors, in concentrating their attention upon the Jurjura, took the large thaddarth they found there to be the typical form of settlement and the standard political unit, below the level of the ‘arsh, of Kabyle society in general. For the large and complex thaddarth was not at all typical of Greater Kabylia as a whole. It was typical only of the central Jurjura – that is, of the society of the Igawawen.
Morizot is himself a victim of this oversight. In stressing the role of the patriarchal family in the society of the Jurjura as one of the key factors underlying the remarkable migratory activities of the Igawawen, he emphasises an aspect of their social organisation which was not at all peculiar to them but, on the contrary, common to the whole of Kabylia. In emphasising the traditions of mutual aid and village solidarity, on the other hand, Morizot gets closer to the heart of the Igawawen enigma. There is no doubt that these traditions were enormously important to the commercial and migratory enterprise of the Igawawen. But what Morizot overlooks is the fact, which we have now unearthed, that the traditions in question here are those of a village which is all but peculiar to the Igawawen, the large and complex thaddarth.
Pour décrire convenablement un village Kabyle des Igaouaouen …, il faudrait être peintre, poète …
(Henri Genevois, Tawrirt n’At Mangellat)73
To understand the Igawawen thaddarth it is essential to recognise that it is not a kinship unit.
A very large number of social groups in Kabylia, from the topmost level of the confederation to the village or even the constituent hamlet of the tūfīq or quarter of the village, do not have genealogical names but topographical or territorial ones. This fact has been quite simply ignored by certain authors. Bourdieu, for instance, has written
if genealogy is used in more or less arbitrary fashion every time that it is important to create or justify a social unit, it is because it allows a kind of relation of kinship to be created, through the fiction of the eponymous ancestor, between individuals joined together as a result of the operation of quite different forces; it is as if this society could not conceive of any type of relationship existing within a social body other than that which exists between relatives …74
Figure 2.3: Ath Lahcène, the largest of the Ath Yenni villages, in 1920.
Bourdieu is quite right to suggest that such ancestors as are invoked are often entirely fictional and that kinship provides an idiom through which relationships resulting from ‘the operation of quite different forces’ may be expressed. But it is simply untrue that Kabyle society invariably employs the idiom of kinship to denote social units and cannot conceive of other types of relationship.75 Many Kabyle tribes and villages do refer to themselves as ‘the sons of X’, but at least as many do not, employing names which are simply topographical or geographical expressions. Only two of the seven villages of the Ath Wasif referred to above, for example, have genealogical names (Ath Abbas and Ath Bou Abderrahmane). Similarly, only two of the seven villages of the neighbouring Ath Boudrar tribe have genealogical names (Ath Ali ou Harzoun and Bou Adnane); the names of the other villages are ‘the ridge of the middle’ (Ighil Bouammes), ‘the ridge of the lionness’ (Ighil n’Tsedda), ‘the fountain of the dried figs’ (Tala n’Tazart) and so on. And Ath Wasif and Ath Boudrar themselves mean only ‘the people of the river’ and ‘the people who occupy (or possess) the mountain’, as we have seen.
Of those Jurjura villages which have been investigated, all have been formed by families and lineages which had been previously unrelated and had arrived at different times and from different points of the compass, some from neighbouring ‘aarsh, some from maritime Kabylia, some from the Wad Sahel and Bouïra districts south of the Jurjura, some from Lesser Kabylia, and even some from much further afield, the Arab regions of the Constantinois and the Sud-Algérois and the Cheliff valley. According to local traditions, the important villages of Taguemount Azouz, of ‘arsh Ath Mahmoud, and Taourirt, of ‘arsh Ath Menguellat, were constituted in this way.76 So too were the villages of Ath Lahcène, Ath Larba‘a, Taourirt Mimoun and Agouni Ahmed of the Ath Yenni.77 Aucapitaine details the distant origins of many of the families of Jema‘a n’Saharij and of the other villages of the Ath Fraoucen.78 And I know from my own fieldwork that the villagers of Ath Waaban, of the Ath Boudrar, take a similar view of their own history.
Anthropologists coming across legends which assert a common ancestry are often rightly inclined to discount these as convenient fictions. But where local traditions regularly assert, on the contrary, the absence of such common ancestry, there is no reason to discount them and every reason to take them seriously, both as factually reliable and as sociologically significant.
The extent to which the thiddar of the Jurjura are composed of families and lineages of diverse origins is inconsistent with the segmentarist thesis. As such, the constitution of these villages is to be distinguished from the more general phenomenon of migrant or refugee families being incorporated into entirely different kinship groups. One of the reasons why Gellner felt that this general phenomenon did not affect his argument in the case of Moroccan Berbers is that it was very marginal in the central High Atlas.
Berber society of the central High Atlas may not be totally rigid – birth and kinship do not hold the individual in an iron vice – but it is certainly not very fluid … In any one village, the proportion of male immigrants will be under ten per cent, usually well under this ratio.79
In Kabylia – at any rate, in central Kabylia and in the Jurjura in particular – the truth is the exact opposite of this. For all the villages for which evidence is available, it is clear that those families able to claim an origin going back to the foundation of the village are a minority and that the majority of families are immigrants from elsewhere and, moreover, are well aware of the fact and see no inconvenience in acknowledging it. (The only villages in the Jurjura which are a regular exception to this rule are the exclusively saintly settlements of the imrabdhen, which only rarely constitute autonomous thiddar and whose populations even more rarely exceed two or three hundred.)
Kabyle society may well give the appearance of a certain rigidity. ‘Fluid’ is not the first word which springs to mind when one gropes for adjectives to describe it. But the rigidity in question is not that of kinship structures. All the evidence suggests that pre-colonial Kabylia was the theatre of unusually intense movements of population in all directions, at any rate from the early sixteenth century onwards. The rigidity which one may observe there is the rigidity of political arrangements which were evolved in the context of this turmoil and in response to it.
The large, complex thiddar of the Igawawen were not invented out of thin air. The evidence suggests that the first of them were formed through the concentration of scattered hamlets previously united only loosely, if at all, in the tūfīq form. This is what local traditions attest for Taourirt n’Ath Menguellat and the three core villages of the Ath Yenni: Taourirt Mimoun, Ath Lahcène and Ath Larba‘a. According to local traditions, the consolidation of the last three dates from the early seventeenth century and was effected under the aegis of a local amrabedh, Sidi Ali ou Yahia.80 The formation of Taourirt n’Ath Menguellat is also credited to the influence of a local saint, Sidi Lhadi Bou Derbal, and occurred at the same period.81 The same processes are likely to account for the formation of some of the large, complex thiddar which exist elsewhere in Kabylia, outside the Igawawen district proper. The leading village of ‘arsh Ath Mahmoud, Taguemount Azouz, for example, was formed, at an unknown date, out of ‘primitive farm hamlets’ situated in the immediate vicinity of the future thaddarth, but in this case without the prompting of a saint.82
I believe that further research would find that many of the other large thiddar of the Jurjura were formed in the same evolutionary manner. But I doubt that all of them were. It seems more likely that it was the earliest thiddar which developed out of tuwāfeq and scattered hamlets and that those formed in the second half of the seventeenth century or later were in many cases constituted from the outset as integrated villages, in imitation of a model of social and political organisation which was by that time well established locally and had demonstrated its merits. This certainly appears to be the case of Ath Waaban, for example; although no one knows precisely when it was founded, it is almost certainly the youngest of the seven Ath Boudrar villages, constituted most probably in the second half of the eighteenth century if not later, and there is no memory of hamlets or a tūfīq preceding the formation of the present village.
Figure 2.4: Igawawen villages today: looking north along the ridge of the Ath Bou Akkach and Ath Wasif, with the Ath Yenni villages perched on their hilltops on the horizon.
The development of the Igawawen thaddarth – that is, of a form of settlement and community which was not a kinship unit, immediately created a fundamental problem, because it meant that order could no longer be (if it ever had been) maintained on segmentarist principles. In a segmentary system, order of a kind can be preserved through the balance and opposition of the segments precisely because the segments conceive of themselves as segments – that is, as related parts of a larger whole which is itself a kinship unit at the next level of size and so unites its constituent segments at that level. But, since the large and complex thaddarth of the Igawawen was not a kinship unit, and its constituent clans were well aware of the fact and in no way disposed to pretend the contrary, how was its cohesion and internal order to be secured?
It may be objected that the Igawawen thaddarth was not unique in this respect. It is certainly true that amongst Berber populations elsewhere one also finds village settlements which are not kinship units. In southern Morocco, for instance, many villages contain lineages belonging to quite unrelated clans, such that these villages also are not kinship units. But there is a fundamental difference with the Igawawen case. For the clans in southern Moroccan villages are rarely confined to a single village; in general, a given clan will be settled in several villages, often at considerable distances from each other. In the internal organisation of the large transhumant tribes, notably the Aït Atta, a virtue, if not an obligation, was made of this territorial discontinuity and reduplication of the clan: the fact that a clan had members in various far-flung parts of the tribe’s territory was a major factor preserving the subjective unity of the tribe, while also affording each clan access to the various types of natural resources – pasture, woodland, water, oases, etc. – contained within the tribe’s territory.83 In this system, village settlements are not the main constituent units of the tribe’s political structure, the clans and, above all, the ‘five fifths’ (khams khmas)84 into which they combine are, so the fact that a settlement is not a kinship unit does not create a significant political problem.
Figure 2.5: Looking beyond Taourirt el-Hajjaj (Ath Yenni) to the south-east of the Igawawen district, with in the middle distance the Aqbil villages of Beni Mahmoud and Aourir Ouzemmour (right) and those of the Ath Bou Youcef (left), with the Jurjura beyond.
In the wholly sedentary society of the Jurjura, matters were very different. There is no territorial discontinuity and reduplication of the clan; a clan does not have members settled in several villages, it is invariably confined to one village.85 And, pace Bourdieu, the village in Kabylia is not merely a territorial unit, a settled space, but a political unit of the first importance. Below the level of the ‘arsh, it is the village – thaddarth or tūfīq – which forms the constituent political unit, not the clan. It is the village which was and often still is the ultimate owner not only of common land (meshmel) but also of all cultivable land, in that land previously in the possession of a family that has left (or been banished from) the village or died out reverts to the village for the jema‘a to reallocate, unless there are agnates of the last owner still present in the village to exercise the right of shefa‘a in respect of it. It is the village which organises at intervals the mobilisation of collective labour for work done in the collective interest (thashmelith), such as the annual overhaul of the irrigation system for the village’s gardens; it is in effect the village which mobilises to build a new house for a young family that has finally moved out of its earlier home in the house of the man’s parents;86 and it is the village which decides upon and organises the occasional sacrifice of cattle or sheep for a thimshret, in which the meat is distributed with meticulous fairness to every family, the funds for buying the animals coming from the village chest and the allocation of meat being conducted by the temman under the watchful eyes of the rest of the adult male population. The fact that, amongst the Igawawen, this crucial political unit was usually not a kinship unit was therefore of the greatest significance, and had no parallels elsewhere in Berber society.
The problem which arose with the constitution of the large and complex thaddarth was unique. Moreover, it was not merely (so to speak) an abstract problem of political order, of the principles upon which the self-government of the thaddarth was to be based; it involved at least one very practical problem, to which a practical solution had to be found early on, if not immediately, that of the physical proximity of unrelated clans, lineages and families.
The establishment of the original complex thiddar of the Jurjura was effected through the voluntary amalgamation of the constituent hamlets of the pre-existing tuwāfeq. The complex thaddarth was a development out of the tūfīq which occurred when the hamlets which composed the latter agreed to regroup in one place – often, although not always, on the summit of a ridge – and thereby form a single, concentrated settlement. This process of concentration obliged clans and lineages and families which had previously dwelt in quite distinct settlements at some distance from one another to exist henceforth cheek by jowl. Where before a kilometre or more, plus a belt of woodland and orchards, as well as pasture and gardens, might have separated the houses of clan A from those of clan B, now only a narrow alley would separate them, and perhaps nothing more than a stone wall.
Figure 2.6: A thimshret in the Jurjura at Ath Anzar (Ath Itsouragh).
This unprecedented contiguity of families recognising no common descent introduced an entirely new element into the daily social life of the people concerned. This change was clearest in relation to the women: whereas, in the tūfīq, the women of a given clan could visit their gardens, collect firewood and fetch water from the well, as well as visit one another’s homes, without once leaving the territory of the clan to which they belonged, in the complex thaddarth they were obliged to traverse parts of the village inhabited by unrelated clans on most of these errands and to do so, moreover, perhaps several times a day. In addition, many houses in a thaddarth overlooked the courtyards of their neighbours, so that the privacy of the women would be in jeopardy even when they were at home.
Figure 2.7: An alley in an Igawawen village: Taourirt Mimoun of the Ath Yenni.
The way in which these problems were handled was through the development of a strict code of proper male behaviour, which has remained in force to this day.87 It was not done for a man to look at a woman when passing her in the street or alley-way, unless she was a relative (in which case a brief greeting was in order); the rule was to ignore her presence entirely and so enable her to pass by as if she was invisible and so unflustered. The fountain, the women’s public meeting place par excellence, was entirely out of bounds for the men. And for a man to post himself in a strategic spot in order to look at the women as they passed by was the ultimate in anti-social behaviour and widely condemned as such.88 In the same way, it was not done to look out of a window into a neighbour’s yard. In the old days, to break this rule was to risk an immediate riposte from the outraged menfolk next door, namely a bullet through one’s window.89
More generally, this intense physical proximity of unrelated lineages was bound to provide endless occasions for friction between them which would simply have been absent from the social routine of the tūfīq. The preservation of the unity of the village as a whole therefore required its inhabitants to develop a degree of civic consciousness expressed in a kind of social behaviour and an array of polite manners embodying to an unusually developed degree the value of consideration (‘anāya) for others.
Such a development would have been unlikely to occur of its own accord and in the case of the Igawawen thaddarth it occurred on the basis of an underlying innovation of a wholly unspontaneous and deliberate character. This innovation was a development in Kabyle law.
