5. Local and national politics

In Nigeria today it is increasingly difficult to separate local and national political processes. The power and autonomy of local-level political institutions have been progressively eroded, with the significant decisions increasingly made in the state capitals or in Lagos. Some of the indigenous political institutions do retain some of their vitality ó considerably more than they display in some other African countries including the Benin Republic ó with the 'natural rulers' apparently courted by government officials and used as channels of communication with the local population. Competition for the major titles is still intense, and often long, bitter and expensive to the contestants, and there are still benefits in terms of income and prestige in obtaining them. But the great majority of traditional offices in many kingdoms are no longer filled, and attempts to define a formal role for the traditional rulers within the constitution have not been particularly successful. With the changes in the administrative framework and the distribution of economic resources, the locus of power has shifted away from the rulers, first to the military chiefs in the l9th century, then to the wealthy entrepreneurs and professionals under the colonial regime, and finally to the politicians, the military and the administrators.

There have been four main phases in this process. In the 19th century many of the older kingdoms collapsed or were greatly changed following the decline of Oyo, or as a result of Fulani penetration in the north and European penetration in the south. The most significant development was the growing importance of the military chiefs. This led to the modification of the political systems of the kingdoms which did survive, and the evolution of completely new systems in new states like Ibadan and Abeokuta.

Secondly, following the British occupation there was the period of indirect colonial rule. This continued, with modifications, until the 1950s. The main trend was the growing involvement of literates and wealthy entrepreneurs in local politics and their opposition to, or support of, the traditional rulers.

Constitutional changes after 1945 prepared the way for the period of civilian politics which lasted from 1952 to 1966. Power shifted away from the Native Authorities set up by the British to the new regional and national governments. Events in Yorubaland became increasingly dependent on those taking place at the national level. This remained the case in the period of military rule, during which the break-up of the old political units and the growth of the financial power of the federal government strengthened central control. With the return to civilian rule under a strong executive presidency, as is now occurring, this trend is likely to continue.

There are basic difficulties in giving an account of 'traditional' Yoruba political organisation. Firstly, political institutions varied greatly from kingdom to kingdom, in size, administrative complexity, and in the ways in which the major office-holders were recruited. Secondly, presenting a static account hides the degree to which individual polities changed, often quite radically, over time, as institutions were modified, and as powerful office-holders altered the balance of power.

The major attempt to deal with these problems is the series of papers by Lloyd (1965, 1968b, 1971). In his 1965 paper he argued that in order to understand the government of African kingdoms the emphasis must be shifted from the administrative structure to the process of policy- or decision-making (1965: 73). A key variable in this is the way in which the 'political elite', i.e. elected office-holders, are recruited. In Yoruba kingdoms recruitment is 'open' in that it is not confined to a ruling aristocracy. In the northern kingdoms, chiefships are vested in descent groups, and the chiefs represent their groups' interests in the oba's council. In the southern kingdoms of Ondo and Ijebu, chiefs are members of title associations rather than representatives of specific descent groups. A number of other variables are correlated with this difference ó rights in land, the corporate strength of descent groups, the type and intensity of political conflict, and the degree of political oppression (ibid: 102ó4).

In his 1968 paper Lloyd gives a more detailed account of the types of conflict which can arise in an ideal-typical northem Yoruba kingdom, where the chiefs represent descent groups. Faced with the entrenched interests of these groups, and surrounded by ritual restrictions, the oba has a limited range of strategies to increase his own power. These include building up a retinue of slaves, gaining control of tolls and market revenue, exploiting disharmony between chiefs, and mediating between competing descent groups. In Lloyd's view, competition between descent groups tends to strengthen the power of the oba and his council as the ultimate arbiters. The outcome of conflict between the oba and his chiefs is less certain. If he wins, he may be able gradually to reduce the power of the descent groups and establish a more centralised political system.

This transition from 'tribal kingdom' with powerful descent groups to 'centralised state' is the theme of Lloyd's 1971 monograph. In fact, for a variety of reasons, the states he examines ó Oyo, Ibadan, Iwo, Abeokuta and Ilorin ó failed to centralise successfully. In Oyo the conflict between the oba and chiefs was not resolved, and in 19th-century Ibadan the power of the descent groups actually increased.

There are a number of issues raised by these papers (cf. Law, 1973a) of which two are most relevant here. Firstly, what precisely is meant by 'centralisation' and is it valid to see kingdoms like Ijebu and Ondo as being 'more centralised' than the other Yoruba states? It may, in passing, be noted that the degree of central control may vary in different areas of a kingdom. Lloyd himself points out the necessity of distinguishing between the situation in the metropolitan and peripheral areas of a kingdom and in regions within its sphere of influence (1965: 70ó1). Oyo is a case in point. Egbado was much more tightly controlled from the centre than were the metropolitan provinces, and Dahomey had almost complete autonomy, apart from its tributary obligations.

Lloyd's discussion of the concept of centralisation shifts in its emphasis. In the first paper he proposes three criteria: (a) the sphere of competence of the government, for instance in the degree to which kinship is transcended by loyalty to the state; (b) the extent to which the constituent units of the kingdom act independently of the king, and particularly whether there is a final appeal to the king in judicial matters; and (c) the degree of control exercised by the king over other office-holders (1965: 81ó2). Elsewhere, however, he relates centralisation to his model of the 'tribal kingdom' where power is vested in chiefs representing descent groups. Centralisation by implication is the degree to which a Yoruba kingdom diverges from this tribal kingdom model (1968). This is most clearly related to (a) above, and was most obviously the case in Ijebu and Ondo. But did it mean that the Awujale and Osemawe were any more powerful than their fellow rulers in the degree of control they exercised over other constituent units of the kingdom or other office-holders? The empirical material raises doubts. In an earlier paper, Lloyd described the Ijebu system as one in which the oba's council, the Ilamuren, made decisions which were then conveyed to the oba who gave them his assent (1954: 381). Ayandele makes the point that, of all the Yoruba oba, the Awujale was least affected by the upheavals of the l9th century, but he goes on to say 'In no circumstances did the Awujale . . . initiate or execute laws, although laws bore the stamp of his office' (1970: 235). In 1885, the Awujale was forced into exile by opponents of the wars, and by 1890, Ijebu policy was being determined largely by the age grades in the capital (ibid: 236). A recurring theme in Ijebu history is the difficulty the Awujale had in establishing his authority over the other crowned oba in the kingdom, particularly in Remo Division. A similar problem is apparent in Ondo, in the relations between the Osemawe and the oloja in the other towns. Lloyd's assertion that Ijebu and Ondo had more highly-centralised monarchies than other Yoruba states (1971: 3) is more difficult to accept if we adopt criteria (b) or (c) than if we adopt criterion (a).

Secondly, there is the problem of the recruitment of office-holders. ls it correct to draw a sharp distinction between the northern Yoruba kingdoms with their chiefs recruited on the basis of descent groups, and the southern kingdoms where recruitment was through political associations? Lloyd's model of the 'tribal kingdom' is derived from the example of Ado Ekiti (1968b: 34). But his fuller accounts of the Ado system (1954, 1958, 1960, 1962) show this to be a very simplified model. It is true that the most senior titles were vested in descent groups, but many of the other titles were not. Also, Ado titles, like those in Ondo, were graded: both systems may reflect the influence of Benin, where the principle of graded chiefships was extremely important in the political structure, or all three systems may derive from a common model. The presence or absence of hereditary chiefships is only one of a number of salient criteria for the classification of Yoruba political systems, and in fact recruitment from descent groups and political associations are only two of a number of ways of selecting officials which could, and often did, coexist in the same system.

Traditional institutions: an overview

(a) Sacred kingship

The most distinctive and widespread feature of Yoruba political organisation is sacred kingship. Its origins are still a matter for speculation and it did not develop in all areas. Its main features, however, were fairly similar in a number of kingdoms.

Though not regarded as divine in his own right, the oba was nevertheless ekeji orisa, the companion of the divinities, set apart from his people by the spiritual powers with which he was endowed at his installation. He made few public appearances, and even during these his face was hidden by the ade, the crown with its beaded fringe. The right to wear the ade is still a jealously guarded privilege among the Yoruba rulers (Asiwaju, 1976a). Most Yoruba kingdoms have only one crowned oba or oba alade, the major exceptions being Egba and Ijebu.

The accession of a new oba was marked by elaborate rites. Often these involved retracing the route by which the founder of the ruling dynasty was said to have come to the kingdom (Parrinder, 1956: 83; Lloyd,1960, 1961). There was commonly a period of seclusion in which the oba was instructed in his new role. Finally he was endowed with the mystical powers of the office in rites which involved eating the heart of his predecessor (Lloyd, 1960: 227).

Succession rules varied. Usually the royal descent group was divided into a number of segments or 'ruling houses' which held the title in turn. Usually the new oba had to be the son of a previous ruler by a free woman, and without any physical deformity. Sometimes, as in Ijebu, there was the further provision that he had to have been born during his father's reign. Even when the choice was restricted in this way there was usually more than one candidate. The elders of the royal descent group would make a preliminary selection, and the final choice would be made by the senior chiefs in consultation with the Ifa oracle. In Oyo primogeniture appears to have been the normal rule until about 1730, after which the Alafin's eldest son, the Aremo, was expected to die with his father. This remained the rule until 1859.

(b) Palace organisation and slave officials

What is striking about the palace organisation of the larger kingdoms of Oyo, Ijebu and Ife, is the degree to which they evolved very similar systems, though it is not clear to what extent they derive from a common model. The palace organisation in Oyo was described in some detail in Chapter 2. The main features relevant here are (a) the extensive use of slaves; (b) the use of eunuchs in senior positions; (c) the requirement that some key officials should die with the ruler; (d) the principle of graded offices, ranked in seniority; and (e) the development of specialised cadres of administrators and other functionaries such as messengers, executioners, musicians, court historians, etc.

The advantage in having slave officials was that they were appointed by, and responsible to, the ruler alone. The requirement that key officials die with the ruler not only ensured their loyalty but enabled his successor to appoint his own men. Palace officials derived great influence from the greater degree of access to the ruler which they had in comparison with commoner chiefs.

In Ife the palace officials were men of modewa status. The derivation of the term, and the functions which many of them performed, make it likely that in origin the modewa were the ruler's slaves (cf. Oyediran, 1971: 69). The term is a contraction of omode owa, the ruler's 'children', and the words for child, omo and omode were common terms of address for slaves. The chief palace officials, corresponding roughly to the senior eunuchs in Oyo, were the woye, who had important roles in the state cults, as well as acting as intermediaries between the Oni and the commoner chiefs in the capital, the ihare, and between the capital and the subordinate towns. Secondly, there were the emese who acted as the Oni's messengers. Their role was similar to that of the Oyo ilari, whom they resembled even in their distinctive dress and hairstyle. The chief emese was the Samu, a eunuch who died with his master. Corresponding to the Oyo tetu or executioners were the Ife ogungbe. The power of the palace officials vis-a-vis the town chiefs seems to have grown to the point where it became advantageous for members of commoner descent groups to assume modewa status through an expensive initiation (Bascom, 1969a: 34ó5). Once achieved, the status was hereditary, and by the colonial period the modewa were the largest status-group in the Ife, population, in which they ranked above the other descent groups in prestige (Bascom, 195 1).

A similar system evolved in Ijebu. The most senior palace officials were the odi, including both eunuchs and officials who died with the ruler. The odi came to have a considerable say in the election of a new ruler. The other palace functionaries were divided into specialised groups called erinle, each under the authority of the odi. One of these groups, the agunren, combined the diplomatic and revenue-collecting functions of the Oyo ilari with the police functions of the Oyo tetu (Oroge, 1971: 30ó47).

(c) The royal descent group

The position of members of the royal descent group itself was ambiguous. Yoruba rulers usually had large numbers of wives, and the royal descent group made up a substantial proportion of the population. Normally it was administered by title-holders appointed from within it, often from among those unable to succeed to the throne. In some cases, members of the royal lineage could wield power outside it, for instance in l9th-century Iwo where they provided the ruler with his closest advisers. In Oyo the influence of officials like the Ona Isokun and Aremo was also extensive. Together with the senior palace officials they formed a liaison committee between the Alafin and the Oyo Mesi (Atanda, 1973a: 17) and the Aremo was at times the de facto ruler of the whole kingdom. But the normal pattern was to exclude royals from positions of power, particularly those able to succeed. Potential heirs were often brought up outside the palace, or even outside the capital. Sometimes they were used to administer outlying areas of the kingdom. The use of sons of the Alafin to administer Egbado had its parallels in Ijebu and Ado Ekiti (Lloyd,1962: 149, 194).

(d) Commoner chiefs

Much of the power in the kingdom outside the palace lay with the commoner chiefs, the most senior of whom often formed a council of state such as the Oyo Mesi in Oyo, the Ilamuren in Ijebu, the Ihare or town chiefs in Ife, or the Iwarefa in Ondo. The power of the Basorun in Oyo has its parallels elsewhere. The Lisa, Olisa and Orunto, the most senior commoner chiefs of Ondo, Ijebu and Ife, respectively, have all been described as the rulers of the capital ó as opposed to the Oba himself who was the ruler of the whole kingdom.[1] In many cases the senior chiefs controlled the largest wards and headed the largest descent groups in the capital. Many of them had important judicial, ritual or military functions. Collectively they were frequently responsible for the final selection of a new ruler from among the eligible candidates, as well as for his deposition if his rule was unsatisfactory. The chiefs were ranked in order of seniority and in meetings they spoke in this order, starting with the most junior.

The ways in which they were selected were varied. The most important principles were (a) free appointment by the ruler, (b) succession within a descent group, (c) promotion through graded title-associations, and (d) election >from age sets. Lloyd has concentrated on (b) in the northern kingdoms and (c) in the southern kingdoms, but the distinction is not entirely clear-cut. More than one principle was operative in most kingdoms, and the basis on which individual titles were allocated could, in some cases, change over time. The main direction of change has been for titles which were allocated in other ways, and particularly for those filled by the free appointment of the ruler, to become vested in descent groups. This has happened in Ogbomosho where the senior ilu chiefships which were not hereditary became so in the l9th and 20th centuries (Agiri, 1966). In Oyo, the military titles, the eso chiefships, became largely hereditary as particular descent groups concentrated on military careers (Law, 1977: 75). Even in Ondo some of the most senior titles have come to be monopolised by descent groups in this way (Lloyd, 1962: 106).

Titles filled from among the members of a single descent group are found everywhere. These included the Oyp Mesi titles in Oyo, most of the Ihare titles in Ife, three of the Ilamuren titles in Ijebu (those of the Olisa, Egbo and Apebi) and one of the Iwarefa titles in Ondo (that of the Jomu). These were only the most senior titles. Lloyd comments that in Ijebu most other descent groups had titles vested in them, though these were of little political significance (1962: 150).

Titles filled by promotion within graded associations were a feature of the Ondo and Ijebu systems, though, as we have seen, not all titles were filled in this way. It might also be noted that a somewhat similar system came to be adopted in Ibadan during the l9th century, and that the principles of seniority and ranking between office-holders operate in all Yoruba polities. In Ondo, title-holders were grouped into three categories, iwarefa, ekule and elegbe, with the iwarefa chiefs being appointed from the other two groups. The senior ekule and elegbe chiefs in turn headed hierarchies of minor title-holders, who may have numbered over a thousand and must have included a large proportion of the male population (Lloyd, 1962: 105ó10). The majority of these minor offices are no longer filled.

A similar principle operated in the Ifore society of Ijebu (Lloyd, 1962: 148) membership of which was open to any free-bom Ijebu on payment of the necessary fees. It had four grades of members and its leaders were members of the Ilamuren. In addition, one of the most senior chiefs, the Ogbeni Oja, was selected alternately from among the odi and the Ifore members.

Organised into grades like the Ifore was the Earth Cult, known as Osugbo in Ijebu and as Ogboni elsewhere. The cult has been described as a secret society, but it is arguable that this was not an apt description as entry was usually open to all free citizens (Bascom,1944: cf. MortonWilliams, 1960b). Entry and promotion were, once more, through payment of fees. The cult was particularly important in Egba and Ijebu where it formed the main judicial tribunal. In most areas the senior officials are known as the iwarefa, led by the Oliwo. They are chosen from the grade beneath in Ijebu and Egba (Lloyd, 1962: 147ó8, 233), while in Oyo the titles are vested in descent groups (MortonWilliams,1960b: 365). The cult had an important role in settling disputes in which blood had been shed on the earth, but its wider political significance lay in the fact that it provided an opportunity for the leading elders of the town to meet in guaranteed secrecy. Members were forbidden to reveal Ogboni secrets to outsiders and had to honour the decisions reached there. The ruler and chiefs were members but had no special privileges in Ogboni meetings. Morton-Williams (1967b) argues that the cult was an important third element in the politics of Oyo serving to restrain the power of the Oyo Mesi over the Alafin, while keeping the Alafin in touch with public opinion through a woman official who attended the meetings on his behalf. This was the situation after 1836: there is little evidence about the role of the cult in Oyo-Ile, or whether it existed there at all (cf. Agiri,1966; Atanda, 1973b).

Closely related to chiefships vested in graded associations were those vested in age sets (egbe) which were widely distributed throughout Yorubaland at one time, but which generally declined in importance before the colonial period. Only in Ijebu did they remain politically influential until the late l9th century (Ayandele,1970: 236) and they continued to be formed until the 1930s. Each of the three main divisions of the capital had its age sets (ipampa) whose leader was a member of the Ilamuren.[2] In Ado Ekiti and Ondo, the name of the category of junior chiefs, elegbe, appears to connect them with age organisations, though their precise role is obscure. Before the 19th century, the senior elegbe chief in Ado was the commander in war (Lloyd,1962: 192ó5). In Oyo, age organisations appear to have declined relatively early: their representative, the Ona Modeke, lost his place among the Oyo Mesi in the 19th century.

Women chiefs

A final feature of Yoruba political systems is the importance in them of female officials. Johnson's account of the Oyo court (1921: 63ó7) lists a number of them. The most senior of these was the Iya Oba, the official Queen Mother (the Alafin's real mother was expected to commit suicide on his accession). The most powerful, however, was the Iya Kere or 'little mother'. She was in charge of the royal regalia. She also had a degree of authority over the ilari, the most numerous category of royal slaves, both male and female, as well as responsibility for the major provincial towns of Iseyin, Iwo and Ogbomoso. Johnson lists a number of other high-ranking women officials with various ritual and domestic responsibilities within the palace, as well as the titles of the principal female ilari. In Oyo towns at present, the most senior woman official is usually called the Iyalode: in many instances she is a woman of considerable authority, particularly in the market with its preponderance of women traders. The office is also important in Lagos (Baker, 1974: 230) where the Iyalode is second in status only to the Oba of Lagos, in whose appointment and installation she participates as the leading female chief. She is also the highest official in a hierarchy in charge of individual markets and market sections in the city. Women are also prominent in the political system of Ondo, where there is a hierarchy of women's chiefships paralleling those of the men. Finally, it might be noted that there are references in some sources to female rulers of both Oyo and Ife, as having existed in the past.

Subordinate towns

The political institutions of subordinate towns and villages in many cases resembled those of the capital, though on a smaller scale. The Osugbo and age sets operated throughout Ijebu as did the Ogboni among the Egba. In Oyo, many of the provincial towns had chiefships vested in descent groups. Subordinate rulers were confirmed in office by the ruler. Generally they had considerable autonomy, though there was usually some form of tribute, and the oba's court in the capital was the highest judicial authority in the kingdom. Foreign policy was also determined in the capital. In Ekiti, responsibility for subordinate towns was an extension of the political system of the capital. Each of the five main chiefs in Ado had jurisdiction over the towns on the road leading out of his ward. In Oyo responsibility for subordinate towns was allocated to all categories of officials, including women and slave title-holders in the palace.

What should have become apparent by now is the kaleidoscopic nature of Yoruba political systems. Despite their variety and complexity, they represent permutations of a relatively small number of structural principles and institutional elements, as the empirical studies of Lloyd and others have illustrated. In his theoretical discussions of politics, as with kinship, Lloyd has tended to lay most emphasis on descent groups as the basis of political organisation in the northern kingdoms, and on other forms of grouping in the southern kingdoms, and he links this difference with the degree of centralisation. While acknowledging that this approach is highly stimulating, I have tried to argue that there are two difficulties with it at present. Firstly, the concept of centralisation itself needs further elaboration. Secondly, the approach obscures the degree of institutional pluralism to be found in all Yoruba polities, and also the variety of processes through which descent groups may emerge as significant elements in the political system.

A number of problems remain therefore. The most basic is the reconstruction of how some Yoruba polities actually operated during different historical periods, though this may, in some instances, prove impossible. The second is the question of the origins and diffusion of particular institutional elements, and their relationship to other variables such as social organisation and technology. With a growing body of historical material becoming available, it is to be hoped that Lloyd's pioneering interest in a comparative political sociology of the Yoruba kingdoms will be followed up by others.

The successor states

With the decline of Oyo new forms of political organisation evolved in Ibadan, Abeokuta, Ilorin and other towns which either survived the disruption of the wars, or were newly founded as a result of them. They represented attempts to deal with two sets of problems. The first arose from the heterogeneity of the populations in many of the larger towns, and the difficulties of integrating them into a single political system. The second arose from the growing power of military commanders and traders, and the decline of traditional political authorities. Descriptions of the constitutional arrangements which evolved should not obscure the fact that conditions were often anarchic: individual military commanders pursued their own independent foreign policies as their private interests dictated, while groups of traders cheerfully continued to trade with the enemy even when a war was going on.

The population was most diverse in Ilorin, with its Fulani, Hausa and Yoruba elements. The major Ilorin titles remained vested in descent groups, as had been the pattern in Oyo. The position of the Emir of Ilorin vis-a-vis his chiefs was much weaker than in other Fulani emirates to the north where the ruler could appoint his own men to key offices on his succession. In Ilorin, effective power lay with the balogun or military commanders. The Balogun Fulani and Balogun Gambari had jurisdiction over the Fulani and Hausa elements in the town, while the Balogun Ajikobi and Balogun Alanamu were both of Yoruba origin. Subordinate to the balogun were the magaji, the heads of the other descent groups. The balogun and other title-holders had control of depopulated areas of land, mainly to the northwest of the capital. These were distributed among their slaves and followers and became vested in the descent groups of their descendants. Administration of subordinate towns which had not been destroyed was in the hands of ajele, political representatives answerable to individual titleholders in the capital.

Ibadan also had its balogun, magaji and ajele but the population was more homogeneous and the basis of the political system was different. There was no oba: the most senior chief was styled Bale, a title reflecting the theoretical suzerainty of the Alafin. The three lines of chiefships were promotional rather than hereditary. Junior chiefs moved up the ranks as the senior ones died, and the positions at the bottom were filled by prominent magaji. A tendency developed for the senior titles to be held by a relatively small number of descent groups (Lloyd, 1971: 22), though there were exceptions. Bale Latosisa (d. 1885) worked his way through the ranks from obscurity, eliminating his more prominent rivals en route. Though formal central control was weak and effective power lay with the leaders of individual descent groups, Ibadan did have a succession of outstanding leaders whose careers provided its foreign policy with some coherence: Basorun Oluyole, Balogun Ibikunle (d. 1862), Basorun Ogunmola (d. 1867), and Latosisa, perhaps the most independent and aggressive of them all. He was largely responsible for the expansion of Ibadan's empire to the east, where he placed many of the towns under the control of his relatives and slaves ó a strategy tried by Basorun Gaha in Oyo a century before. He died just before the truce of 1886. No leaders of comparable authority emerged after him, and the lack of a powerful oba in Ibadan was one of the main reasons why the British transferred their headquarters to Oyo during the colonial period.

But the degree of political fragmentation was greatest in Abeokuta, where refugees from 150 Egba settlements retained their separate identities, interests and political leadership. It was estimated in 1853 that there were 4000 people engaged in running the town (Phillips, 1969: 118), including Ogboni officials and military leaders for each of the constituent communities, as well as the parakoyi, representing trading interests. The major conflicts tended to be between the Ogboni and the military chiefs, the ologun, but the position was complicated by the arrival of the missions and the Saro. Both required peace and political stability ó the missions in order to consolidate their position, and the Saro in order to pursue their commercial interests. Both were aware of the lack of a central authority in the town. In the 1850s, the CMS missionary Henry Townsend had some influence with the Egba chiefs, particularly where relations with Lagos were concerned, and it was at his suggestion that the office of Alake was revived in 1854. This failed to provide the necessary leadership. In the 1860s Townsend's influence declined as relations between Abeokuta and Lagos deteriorated, and pressure for political modernisation now came from the Saro, and particularly from G.W. Johnson. With the support of the leading military chief, Basorun Somoye, he set up the Egba United Board of Management in 1865, an attempt at modern administration financed by customs dues. Somoye soon found that his patronage of the Board clashed with his interests as a military commander (Phillips, 1969: 128) and in any case he soon died. Johnson was unable to find another powerful patron. His candidate failed to be elected as Alake, and the EUBM died a gradual death, despite his repeated attempts to revive it. Leadership reverted to the shifting coalitions of Ogboni and ologun and disunity remained a major problem.

In 1898 there was a second attempt to establish a modern administration, this time with the assistance of the British. The E,gba United Government had a governing council consisting of the oba together with Ogboni, military, Christian and Muslim representatives, all under the chairmanship of the Alake, and with a Saro as secretary (Pallinder-Law, 1974: 74). An elaborate administration developed, employing 350 people by 1908, and divided into numerous departments. Its weakness proved to be its dependence on the British. Its establishment of law-courts deprived the Ogboni and township chiefs of income and prestige, and agitation against the government was only put down with the help of British troops in 1898, 1901 and 1903. The government relied on customs dues which it could only collect with British support, and it was only able to finance its larger projects with loans from Lagos. Another crisis in 1914 was again put down by the British with force, resulting in the so-called 'Ijemo massacre', but this time the Lagos government demanded the end of the degree of Egba independence guaranteed by the 1893 treaty (Pallinder-Law, 1974).

The colonial period and indirect rule

Once British rule had been consolidated, the aim of the colonial government was the creation of cheap, viable administration by relying as much as possible on the existing political institutions. But the institutions which they chose to recognise were in many cases those which had lost much of their authority to the military chiefs during the l9th century. In particular the powers of the oba were 'restored' giving them a degree of independence from the other chiefs which they had never enjoyed before. In Northern Nigeria the British encountered strong rulers, administering through a hierarchy of district and village heads. The application of this model to the Yoruba kingdoms created problems. The Yoruba oba were not Islamic autocrats, and their new, increased, powers were often resented. Principal oba, including the Alafin, Oni, Alake and Awujale, were made Sole Native Authorities, with sweeping powers to act on their own without consulting the chiefs. At times, the oba and the British Resident were able to work together closely. In Oyo, Captain Ross, the long-serving British Resident, was a close friend of Alafin Ladugbolu (Atanda, 1973a). He had access to the Alafin's private apartments, and major policy decisions tended to be settled there. The Alafin's court had wide powers. Appeal was possible to the provincial court, but the Resident who presided over it tended to support the decisions of the Alafin.

The relative powers of the oba and chiefs were also affected by the new salary structure. The British abolished customary tribute and paid them with money derived from direct taxation, which was introduced from 1916 onwards. The salaries of the chiefs were much lower than those of the oba. In the 1920s the Alafin received £4800 per annum. The Bale of Ibadan received £2400 and the Oni £1400 (Atanda, 1973a: 143ó6). Minor oba received between £300 and £600 and most chiefs received less than £100. Many title-holders had no official salary at all. Some chiefs were paid for acting as judges in the local courts: titles which offered neither a salary nor a place on the bench tended to be left vacant.

The new hierarchy of district and village heads often bore little relation to the situation which existed before the British arrival. Some oba now had jurisdiction over towns which they had never controlled before, and subordinate rulers who were traditionally answerable to the oba now found themselves answerable to a local district head. There was serious violence in Nigeria, just as there was in French territory. Risings in Okeho and Iseyin in 1916 were put down by force, and the leaders, including the Aseyin of Iseyin, were publicly executed (Atanda, 1973a: 173). In the Egba riots of 1918, one of the oba, the Osile, was killed. The railway was attacked and stores were looted. The troops were brought in and 500 were killed before order was restored.

The causes of the trouble were complex: the growing powers of the oba and district heads, the imposition of direct taxation, the free labour demanded by the British for road construction (Agiri, 1972: 157) and discontent inherited from the Egba United Government period (PallinderLaw, 1974). Other oba at times only managed to stay in power with the help of the British, including the Oni and the Awujale (Oyediran, 1973a; Lloyd, 1977).

Discontent with the system of indirect rule centred on the relations between the oba and four other groups: the chiefs, the subordinate rulers, the British and the growing educated elite. In selecting a new oba, the British were often in a dilemma. They wanted to follow traditional procedures, but they also wanted to make sure that a 'suitable' candidate was chosen. At times it was impossible to do both, which meant supporting a 'progressive' but unqualified candidate.

Their record in Ijebu was particularly inept. The Awujale had to be the son of a previous oba, born of a free woman during his father's reign, and without physical deformities. In 1916, the British supported a candidate with a mother of slave descent and with a toe missing. He had to be replaced. It had been the practice for the title to rotate between three ruling houses: Tunwase, Fidipote and Ogbagba, but some of the educated elite wanted to revive the claims of a fourth house, Gbelegbuwa, which had not held the office for a century (Ayandele, 1970: 244). This became possible when the requirement that the Awujale should be the son of a reigning monarch was waived, because of a shortage of candidates. In 1929, the literate and popular oba Adenuga was deposed by the British on a counterfeiting charge. His successor died in 1933, and there was support for Adenuga's restoration. But the British turned to the Gbelegbuwa house, and Daniel Akinsanya was appointed. He was a literate tailor, related to the house only through his mother. From the start he was unpopular both with the chiefs and with the educated elite. He survived an early assassination attempt, and reigned until 1959, thanks to the support of British administrators, despite numerous petitions to have him removed (Ayandele, 1970; Lloyd, 1977).

The new subordinate status of previously independent rulers was a major issue in Oyo where many now found themselves answerable to Alafin Ladugbolu and Captain Ross. The best example was Ibadan, the most powerful of the Yoruba states in the l9th century. The Bale was deposed by the Alafin in 1914 for supporting a petition demanding the removal of Ross (Atanda, 1973a: 135ó6). The Bale of Ogbomoso received similar treatment in 1916, and a second Bale of Ibadan was deposed in 1925 (ibid: 155ó65). Discontent in Ibadan was finally defused in the 1930s after Ross's retirement. The 'New Oyo Empire' was split up, and a new Ibadan Province created, with the Olubadan as the senior traditional authority. This produced new demands for autonomy from the Osun towns like Iwo and Ogbomoso. Osun Division was given administrative autonomy in 1954. A similar split had taken place in Ijebu in 1937 with the creation of Remo Division.

In the early days of indirect rule, the balance of power had shifted sharply away from the chiefs in favour of the oba. In the 1930s newly compiled intelligence reports on many areas led to a re-evaluation of the role of the chiefs. Belated attempts were made to give them greater influence. The Sole Native Authorities were abolished one by one, and by 1950, all oba were required to take notice of the opinions of the chiefs and other council members (Brown, 1950: 17). But the differentials between the salary of the oba and his chiefs remained. Chiefs found it increasingly difficult to perform their role without additional sources of income, and few literates were attracted by the offices (Lloyd, 1958: 174ó80).

By the 1930s the spread of education and the growth of the cocoa industry had resulted in the development of a group of literates and wealthy businessmen in many towns, and their influence on local politics started to grow. The oba and chiefs were no longer the wealthiest men in the towns, and the new elite began to demand a formal role in native authority politics. In some cases oba were elected who were themselves members of this group, like the present Oni of Ife, who came to the throne in 1930. In other towns, they formed an articulate and well-organised opposition to the oba, as in Ijebu and Ogbomoso (Lloyd, 1977; Agiri, 1966). They provided the leadership of the new progressive unions which were being formed during this period. In Ijebu, T.A. Odutola, one of the wealthiest Yoruba industrialists, led the opposition to the Awujale from the 1930s onwards. In 1953 he took the title of Ogbeni Oja. In Ogbomoso the local progressive union was engaged in a long struggle with the oba over control of the courts and the district council. After the death of oba Oyewumi in 1940, the union campaigned for the appointment of a literate successor. A rival candidate was confirmed by the Olubadan, but the union appealed the case all the way to the Privy Council, and won in 1944. The struggle in Ogbomoso had many elements in common with the later conflicts in Ibadan: the oba and chiefs, mainly illiterate Muslims, were opposed by wealthy literates, mainly Christians.

After 1945, rapid changes took place in many councils. In Oyo, for instance, the Alafin had been Sole Native Authority. By 1952, he was President of the Oyo Divisional Council which now had a majority of elected members. One of the first acts of the Action Group government in the 1950s was to formalise these changes, and to create a local government structure in which three-quarters of the councillors were elected, rather than traditional, members. But by this time the constitutional changes in the country as a whole had created a much more important political arena in which Nigerian politicians could compete, and power had shifted away from the Native Authorities to the new regional govemments.

Political parties and constitutional development

In the decade after 1945, rapid changes took place at both the local and the national levels: a succession of new constitutions, the strengthening of the powers of the regions, the transfer of many powers to Nigerian legislators, and the development of party politics.

Nationalism was not a new phenomenon in Nigerian politics. There are several studies of early nationalist sentiment (Coleman, 1962; Ajayi, 1961; Ayandele, 1966), but for a long time this was confined to Lagos. Here the key figure was Herbert Macaulay (Baker,1974: 88ó94; Cole, 1975: 110ó 16), though during his career he was mainly concerned with Lagos issues like the exile of the Eleko, the Lagos oba, and the imposition of water rates on the island. His mass support came from the Lagos Muslims and the market women, and his party, the Nigerian National Democratic Party (NNDP) won all the elected seats reserved for Africans in the Lagos Council between 1923 and 1938 (Baker, 1974: 286ó7). It was not until the 1930s that the nationalist movement developed a wider base, with the foundation of the Yaba College of Higher Education, and the return of growing numbers of Yoruba and Igbo graduates from Europe and the United States. A group of the new iptellectuals formed a political party, the Nigerian Youth Movement (NYM), and this was strong enough to capture the Lagos seats from the NNDP in 1938. The Youth Movement fell apart in a disastrous split over the election of a new secretary in 1941. One group, led by Dr Azikiwe, formed the National Congress of Nigeria and the Cameroons (NCNC) in 1944. Herbert Macaulay was the first President (Sklar, 1963: 48ó59). The party inherited the supporters of the NNDP in Lagos, and Azikiwe, although an Igbo, represented Lagos in the Western Region House of Assembly for a time. It also became the major party in the Eastern Region of the country, and gained considerable support in the Western Region, in the predominantly non-Yoruba Benin and Delta Provinces, and in areas which for one reason or another were in conflict with the ruling party in the west: the Action Group led by Chief Awolowo.

Awolowo had also been a member of the NYM. He had tried to revive the party in Ibadan, before leaving for England to study law in 1944. In London he founded the Egbe Omo Oduduwa, a pan-Yoruba cultural organisation, which later gained the support of the Oni of Ife and other leading Yoruba figures (Sklar, 1963: 233ó5). This was explicitly founded as a counterweight to the Ibo Federal Union with which Azikiwe became involved. Ethnic nationalism was becoming a major force in Nigerian politics. The Action Group, which developed out of the Egbe was founded in 1951 to fight the approaching elections. Awolowo had great success in gaining the support of the traditional rulers. Nearly every year, from 1951 to 1962, there were local, regional or national elections, and in the Western Region these were AGóNCNC confrontations. Local disputes rapidly tended to become party-political conflicts.

These took place within the framework of a rapidly developing constitution. The Richards Constitution of 1947 proposed regional assemblies, while elected representation outside Lagos came with the Macpherson Constitution of 1951. This provided for an elected House of Assembly in each region, together with a central legislature and a cabinet including members nominated by each of the regions. The first elections in the Western Region for the regional and national assemblies took place in 1951ó2, and the AG emerged with 49 of the 80 seats in the Western House of Assembly.

The 1951 constitution soon had to be replaced. Because of disunity in the NCNC, Azikiwe found himself as opposition leader in Ibadan rather than a minister in Lagos. The politicians from the north and south of the country were deeply divided over the issue of self-government, and communal rioting broke out in Kano in 1953. In the 1954 constitution, the powers of the regional governments were strengthened, and they were given control over education, agriculture, justice, local government, and the funds of the marketing boards. Lagos was separated from the Western Region and became a federal territory.

Local government reforms

The first piece of legislation passed by the AG government in the west was the reform of local government. In fact, the system had been changing even before the advent of party politics, with moves to involve the chiefs and the educated more fully in the native authority system, and to replace the Sole Native Authorities. In all the councils, a number of educated members had been recruited, and the council finance committees included only literate members.

Often the traditional office-holders and the new elite came into conflict. In 1949ó50 there was agitation in Ibadan against Chief Salami Agbaje, the educated and wealthy Otun Balogun, led by the illiterate Muslim chiefs. The charges they made against him were rejected by a government commission, but the affair only highlighted the problems of administering the largest city in the country through a political system whose senior members were often senile (Post and Jenkins,1973: 55ó78).

The new local govemment legislation set up a complex and flexible system, in which a number of different arrangements were possible in different areas. A hierarchy of divisional, district and local councils was set up. In some areas, all three types could be found, with different powers divided between them. The crucial control over taxation could be located at any of the levels, depending on local conditions. In the large towns, there were single all-purpose councils. Traditional chiefs could now only hold up to a quarter of the council seats, though some oba like the Oni could continue to wield considerable influence through their wealth, political contacts and support of the ruling party. The chiefs who supported the NCNC came under increasing pressure.

In 1951, the response to the arrival of the national political parties at the local level had been apathetic. The AG gained its initial support through canvassing traditional rulers. Initially party loyalty and ideology meant little, and in 1952, elected legislators in other parties openly defected to the Action Group once it was in power (Post and Jenkins, 1973: 1 10).

What probably did more than anything to create interest in party politics was the rise in taxation under the new govemment. The local authorities were responsible for carrying out the government's ambitious programme for expanding primary education, and taxation effectively doubled in many areas. There were tax riots in Ogbomoso, northern Oyo and Egbado, while in Ibadan the Mobolaje Grand Alliance, led by Adegoke Adelabu and allied to the opposition NCNC, won control of the new Ibadan City Council.

Adelabu was the most successful Ibadan politician of the 1950s (Post and Jenkins, 1973). After a period working as a manager for the UAC, followed by unsuccessful ventures into business, this largely self-educated man helped organise the opposition to Chief Agbaje. After 1951, he became, in rapid succession, a member of the Western Region House of Assembly, Chairman of the Ibadan Council, and a federal minister. He lost both the ministry and the chairmanship after a government enquiry into the affairs of the Ibadan Council, but soon made a comeback. By the time of his death in a road accident in 1958 he had become Leader of the Opposition in the Western House.

His career was one of brilliant opportunism and an ability to exploit the complex issues in Ibadan politics. Firstly, there was the conflict between the new elite and the others chiefs and magajis. Adelabu drew on the support of the chiefs and magajis against the wealthy, literate and predominantly Christian Ibadan establishment. When a prominent Christian, Chief l.B. Akinyele, was elected Olubadan in 1956, Adelabu tried to have his own candidate installed. Secondly, there was the conflict between the indigenous Ibadan people and the Native Settlers' Union, which represented immigrant (mainly Ijebu) interests. Thirdly, there was the question of Osun separatism, the demands for administrative autonomy of the northern areas of Ibadan Province. Finally, there was the rural discontent which resulted from the swollen-shoot epidemic in the cocoa industry and the administration's policy of cutting down infected trees without compensation. This led to the formation of the militant Maiyegun League. The position taken by Adelabu was to support the Muslim chiefs and the rights of the Ibadan indigenes, and to oppose Osun separatism and the administration's treatment of cocoa-farmers. All this brought him into conflict with the Action Group. Chief Awolowo, himself from Remo, had been legal adviser to the Native Settlers' Union, and his deputy. Chief Akintola, who came from Ogbomoso, supported the Osun claim for autonomy. Despite the fact that it was the regional capital, Ibadan remained an NCNC stronghold until after Adelabu's death, and his funeral triggered off serious riots both in the city and in the rural areas.

Party conflict was also intense in Oyo, where it led to the exile and deposition of Alafin Adeniran. The issues were similar to those in Ibadan. Again, there was antagonism between the traditional office-holders and the new elite. This was marked by the personal conflict between the Alafin and Chief Bode Thomas, one of the founder members of the AG and Chairman of the Oyo Divisional Council. Secondly, there were the inequalities between the northern and southern parts of Oyo. The southern part lay on the edge of the cocoa belt: it was wealthier and had a higher level of education than Oyo town with its predominantly Muslim population. While the southern areas supported the AG, Oyo town supported the NCNC. After tax riots in northern Oyo, the Action Group government suspended the Alafin's salary. He was given support by Adelabu. Rioting broke out at an NCNC rally, and six people were killed. A commission of enquiry put the blame on both sides, but the AG government refused to accept its report. The Alafin was exiled and later deposed. The other oba took the hint. In 1958, only one out of the 54 members of the Western Region House of Chiefs was an NCNC supporter.

In Ife, on the other hand, the Oni had been a staunch member of the AG since its formation. His chief opponent in Ife politics was Chief Fani-Kayode, chairman of the Ife Divisional Council. In 1954, the Ife District Native Authority (of which the Oni was Chairman) granted a concession to a timber company partly owned by the Oni to exploit the Ife Forest Reserves for 25 years (Oyediran, 1972). After the creation of the Ife Divisional Council in 1955 a struggle for control of the timber revenues began. It was complicated by the 1959 federal election in which Fani-Kayode, the official AG candidate, was defeated by an independent candidate supported by the Oni. After this incident, Fani-Kayode joined the NCNC instead. In 1962, the position changed again. Fani-Kayode was now Deputy Premier of the Region and Minister for Local Government. The Oni was forced to give up the timber concession, but he received it back after the 1966 coup.

Parallels can be drawn between the events in Ibadan, Ife and Oyo. In all three cases, local issues turned into party conflict. There was an element of class conflict, with Adelabu opposing the wealthy Ibadan establishment, the Alafin supporting the poorer areas of Oyo against the administration, and Fani-Kayode attacking the commercial interests of the Oni. Rut there was also the question of communal loyalties, raised by Adelabu's campaign against the Native Settlers' Union and Fani-Kayode's success with the Modakeke voters in Ife.

What all three cases showed was the increasing interdependence of regional, local and national politics. The Oni was able to retain his local influence, despite his loss of Sole Native Authority status, through his links with national politicians. When he was temporarily defeated in the battle over control of the forest reserve, it was because his opponent had been able to do the same. Other oba, however, had surrendered much of their political independence. Whereas Alafin Ladugbolu had been supported by Captain Ross, his successor, Alafin Adeniran, was exiled by the Action Group. The threat of deposition or a salary reduction could be used openly to force recalcitrant rulers back into line, and their position became even more difficult during the more serious political conflicts which began after Nigerian independence in 1960.

National politics and party conflict

By 1960, the Action Group had gone far in wearing down the opposition in the Western Region by persuasion, the use of patronage, or sheer force. The serious NCNC challenge had disappeared after the death of Adelabu. The climax came with the 1961 local elections in Ibadan. By the time of the poll, a number of opposition candidates had either been arrested or forced to flee the town. The NCNC leaders, faced with a hopeless situation, called a boycott, and the AG won the election with 97 per cent of the vote (Jenkins, 1967: 229).

However, there was already a split developing in the Action Group leadership. In 1959, Awolowo had resigned as regional Premier to fight the national elections. Akintola took over as Premier, but Awolowo remained as party leader. The party lost the national election: Awolowo was now leader of a tiny opposition in a legislature dominated by a coalition of the NCNC and the Northern People's Congress.

Relations between the AG leaders deteriorated, over issues of personality, ideology and cooperation with the other parties in the country. Awolowo and party intellectuals were moving towards an ideology of 'democratic socialism' and opposed cooperation in a national government. Akintola, supported by the businessmen and some of the chiefs, were against fighting expensive elections in other regions which could not be won. The businessmen wanted a greater share in federal government contracts which were largely allocated on party lines. At the 1962 party conference in Jos, the Awolowo supporters voted to sack Akintola as deputy leader, and a majority of the members of the Western House of Assembly signed a petition to the regional Governor, the Oni of Ife, asking for his removal as Premier. The Oni supported Awolowo. He dismissed Akintola and invited Alhaji Adegbenro, an Awolowo nominee, to form a new government. The Assembly met to debate a motion of confidence in the new government. Akintola's supporters, aided by the NCNC members, started a fight and the meeting broke up in disorder (Mackintosh, 1966: 448). The federal government declared a state of emergency in the region, appointed an administrator to run it, and placed restriction orders on the political leaders. The order on Akintola was soon lifted, and he was able to form a new party, the United People's Party, from his own faction of the AG. Together with the support of the NCNC, he then formed a new government.

This gave the federal government an opportunity to discredit Chief Awolowo. An investigation into the conduct of the regional government revealed corruption on a massive scale, financed largely by the marketing board reserves (Mackintosh, 1966: 434 40). In November 1962 a long treason trial began in which Awolowo and other AG leaders were accused of plotting to overthrow the government. To this end they were said to have accumulated arms and arranged for men to be trained in Ghana. Many details of the plot and how far Awolowo himself was involved remained unclear, but certainly the use of strong-arm tactics was by now institutionalised in Nigerian politics.

But events in the Western Region cannot be understood without looking at Nigerian politics as a whole. During the 1950s, each of the three regions hadhad its own dominant party: the Northern Peoples' Congress in the north, the NCNC in the east, and the AG in the west. The population of each region consisted of a large dominant ethnic group, together with minorities which, >from time to time, mobilised to demand autonomy. The dominant party in each region became allied with separatist movements elsewhere, but only the NCNC had much success in more than one region.

Within their own regions, the dominant parties were able to consolidate their positions through the distribution of patronage, through control of the police and the courts, and through violence. They were united in a national government between 1957 and 1959, but during the 1959 election relations between the Action Group and the NPC became particularly bitter. Afterwards, the NPC formed a coalition with the NCNC, leaving the AG in opposition (Post, 1963).

After the split in the AG in 1962, the NPC gradually came round to supporting Akintola who was initially able to remain in power only with the help of the NCNC. In 1962, the Midwestern Region was created, consisting of the mainly non-Yoruba areas of the west. This also had an NCNC government. The NPC was now faced with the prospect of NCNC administrations in all three southern regions, and this would have created constitutional difficulties for their control of the country as a whole. In 1964, a controversy over census results created a rift between the NCNC and the NPC. Akintola formed a new party, consisting of his own UPP and the majority of the NCNC members of the Western House of Assembly, led by Fani-Kayode. The new party, using the old NNDP label, formed an alliance with the NPC to fight the 1964 federal elections (Post and Vickers, 1973: 107ó218), the Nigerian National Alliance or NNA.

This was opposed by the United Progressive Grand Alliance, which brought together the NCNC in the east and midwest, the AG and a few loyal members of the NCNC in the west, and minority parties in the north. The campaign saw even greater refinements of the techniques of intimidation. A number of NNA candidates were returned unopposed in the west because their opponents were unable to file their nomination papers. A further factor helping the NNA was a boycott called by the UPGA which was only effective in Lagos and the east. In the west, enough votes were cast to allow the government to declare the election valid. The NNDP won 36 of the 57 seats.

The result only increased the political isolation of the Eastern Region, and led to a consitutional crisis when the President, Dr Azikiwe, refused to ask Alhaji Abubakar, the federal Prime Minister, to form a government. A compromise agreement was worked out which provided for elections to be held in the Western Region in 1965, and for the formation of a 'broad based' federal government. In the event, this still excluded the AG.

Of the diverse elements which made up the Action Group before 1962, Akintola could draw on the support of the businessmen whose futures depended on access to government contracts, and of some of the poorer areas of the region which felt that they had been neglected in the previous decade. His strongest support came from Osun Division, and especially Ogbomoso, his home town, where he is remembered with great respect. The oba and chiefs were brought into line through government control of their salaries, and the remaining elected councils were dissolved and replaced by reliable management committees or sole administrators.

But the popularity of a Yoruba politician ultimately depends on his ability to deliver the goods to his constituents, and after 1960 Akintola was finding it increasingly difficult to do this. A new waterworks and a shoe factory were built in Ogbomoso and NNDP supporters elsewhere remained affluent, but the marketing board surpluses were exhausted, and the administration was unable to fund the projects which would have increased its popularity elsewhere. Cocoa prices were low and were reduced still further after the election campaign of 1965. During the campaign, the marketing board price was kept artificially high at £110 a ton: it then fell abruptly to £65.

The campaign itself was even more corrupt and violent than its predecessors (Post and Vickers, 1973: 219ó38). UPGA candidates found themselves once more unable to file nominations, and NNDP candidates were returned unopposed in 16 of the 94 seats. Ballot papers were distributed to NNDP supporters before the polling opened. The government controlled the announcement of the results. Several UPGA candidates with a majority of the votes in their constituency believed that they had won, only to hear on the radio that they had lost. The two parties announced completely different results. UPGA claimed to have won 68 seats, while the official figures gave 73 to the NNDP.

On voting day itself, violence broke out at Musin, north of Lagos, and after the election it spread to other UPGA strongholds, including Ilesa Ijebu, Ekiti and Ondo. Looting, burning, killing and pitched battles with the police continued until the middle of January 1966. In many areas the violence was well organised (Anifowose, 1973). It gained the name of 'Operation West-It' from the way in which the property and persons of political opponents were systematically sprayed with petrol and set on fire. Despite hundreds of deaths, the federal government stood by Akintola and refused to declare a state of emergency. Akintola had a meeting with NPC and army leaders on 13 January 1966, probably to discuss bringing in the army to control the situation. Before this could happen, he had been killed himself in the military coup of 15 January.

Yoruba politics and military rule

The coup was probably more welcome among the Yoruba than anywhere else in Nigeria, as it brought relief from the chaos and violence of the previous years. But the question was now whether Nigeria itself could survive. General Ironsi, who assumed power after the January coup, abolished the regional structure, but his regime was overthrown in a second coup, the following July. After several days of confusion, Lieutenant-Colonel Gowon emerged as the new Head of State. The position of the new administration was initially very weak. To gain wider support, it started a series of constitutional talks, and the western delegation was led by Chief Awolowo newly released from jail, together with oba and rehabilitated Action Group politicians.

The Yoruba appeared divided over their future position in Nigeria. Some wanted to see an independent Yoruba state, to include Lagos and Ilorin. Others favoured participation in a loose confederation, while a third group favoured a strong federation, with the Western Region split up to give a greater degree of local autonomy (Dudley, 1970). In the period leading up to the Biafran secession of May 1967, it was not at all clear whether the west would follow suit. In the event a number of factors combined to keep the west in the federation. Firstly, there was its military vulnerability. The Yoruba were poorly represented in the rank and file of the army. The northern troops stationed in Ibadan, who had been a bone of contention between the west and the Lagos government, were finally removed in May 1967, but there were still federal garrisons in Ilorin and Lagos. Secondly, Gowon persuaded Awolowo to become Deputy Chairman of the Federal Executive Council and Federal Commissioner for Finance ó a position nearly equivalent to that of Prime Minister. When the country was divided into twelve states in May 1967, the west remained intact, apart from the loss of Colony Province, in which most of its industry was situated. There were demands for the creation of a Yoruba Central State, including Oyo, Ibadan and Osun Divisions, and a Yoruba East State, including Ondo and Ekiti (Panter-Brick, 1970: 267ó76) though these were not met until 1976.

On its arrival in power, the military lacked any political experience, and left policy-making in the hands of the civil service. Later civilian commissioners were brought in, some of whom were former politicians. The civil service was now freed from the pressures it had experienced under NNDP government (Murray, 1970), but it was now exposed to the sorts of pressures from local interest groups which had previously been directed at the politicians. As Dare says, the success of a local delegation to a ministry was directly correlated with the number and seniority of officials from the locality from which it came (1973: 104). Civil servants came to be seen as supporting the interests of their home towns, and as being divided into competing groups, based on their areas of origin.

The position of the oba and chiefs was more ambiguous. After the 1966 coup, many old scores were settled. A number of rulers who had supported the NNDP were attacked or had their property destroyed. Many were exiled and a few were killed. In 1970, the Onigbeti was still in exile in Ogbomoso and the remains of his burnt-out car still stood in the garage in the ruins of his Igbeti house. Other chiefs who had had their salaries withdrawn under the NNDP now had them restored.

In the absence of elected representatives, the oba were courted by the government as leaders of public opinion, and some had close contacts with the Military Governor, Governor Adebayo (Dare, 1973: 112). Conferences of oba and chiefs were called, but increasingly the traditional rulers were seen by the government as mouthpieces for the demands of their own communities. By 1972 under Governor Rotimi, the chiefs' conferences were discontinued (Dare, 1973: 111).

Nevertheless the military government in the west continued to intervene in the selection of traditional rulers, just as its predecessors had done. Perhaps the clearest example of this was in the election of a new Alafin of Oyo in 1968ó 70. The previous Alafin had died in January 1968. The candidate put forward by the Oyo ruling house and the Oyo Mesi was not accepted by the government, and they were told to think again. In 1969, the Oyo Mesi were relieved of their traditional task of selecting the Alafin, and another committee was appointed. The government's candidate was duly installed in 1970. The new Alafin fitted well with the trend towards younger and more educated obabut the affair illustrated well the extent to which chiefship was now controlled from the state capital. This was symbolised at the Alafin's installation in 1970, in which prominent roles were taken by the Military Governor, the Commissioner for Local Government and Chieftaincy Affairs, and the wealthy Lagos businessman who had been backing the winning candidate.

While the Western State government did manage to establish a modus vivendi with the civil service, the oba and the former politicians, it had less success in its relations with the masses. The result was the so-called Agbekoya Rebellion (Williams, 1974; Beer and Williams, 1976). During the period of the civil war, taxation was the most pressing issue. Enforced austerity, combined with inflation, low cocoa prices, and frequent (and often brutal) tax raids exacerbated the situation. Disturbances started in Ibadan in 1968 with an attack on the local government of fices, and they spread to Egba, Oyo, Ede, and elsewhere. Tax-collection was suspended and a commission of enquiry was appointed to look into the reasons for the trouble. Its report concluded that the riots had been spontaneous and were due primarily to the high levels of taxation. There were complaints from all over the state about the sole administrator system, corruption in local government, and the failure to provide local amenities, despite the high taxes (Lloyd, 1974: 205ó7). The government accepted these complaints in principle, but refused to lower the flat rate of income tax to the level demanded. In July 1969 the government gave an ultimatum to tax defaulters, and the raids started again. The Sohun of Ogbomoso was killed by rioters in July, and in Ibadan the rioters freed all the prisoners, including tax defaulters, from Agodi Prison in September. Pitched battles with the police and the army took place in the rural areas. The leaders of the rebellion in Ibadan were mainly illiterates, small-scale farmers, and men who had not been involved in politics before. The most prominent leader to emerge was Tafa Adeoye, a farmer from Akanran, which had also been one of the main centres of opposition to tree-cutting during the swollen-shoot epidemic. In October 1969, Chief Awolowo made a well-publicised trek through the bush to negotiate with Adeoye, and many of the demands of the Agbekoya were met. The flat rate of taxation was reduced to £2 a year, an amnesty for tax defaulters was declared, and Agbekoya members were soon out helping local tax officials in the task of collection.

Beer (1976) argues that the roots of the Agbekoya lay in the failure of the political system to represent peasant interests successfully. Once again, the farmers had resorted to direct political action in the (justified) belief that it would succeed, but only after their other efforts had failed. Their aims were limited to eliminating defects in the taxation and administrative systems, rather than altering them fundamentally. Once they had achieved these aims, the movement disintegrated.

Since 1970, the structure of local government has been under constant review. After 1971, the 114 existing local government bodies in the west were replaced by 39 new councils, and in 1976 a more uniform system was adopted throughout the country. At present, with the transition back to civilian rule still in progress, the precise structure of local government in the future is still uncertain, but the direction of change will be related to four major questions. The first is that of the scope of local government and its relationship to the state govemment. Should local government be simply an extension of the state government, or should it be relatively autonomous? Local government changes in Nigeria are commonly justified as 'bringing government closer to the people', yet since the days of indirect rule and the Native Authorities the tendency has been for more and more of its functions to be taken over by state and federal agencies. With the creation of smaller states in 1976, and the return to elected representation at the state level, this trend seems likely to continue.

But whatever powers local government is left with, the second question is that of financial viability. In the past the larger urban-based local authorities have had a large enough tax base to be economically viable: many of the rural councils have not. The 1976 constitutional proposals recognised this problem, and suggested the creation of local government units with populations of 150,000 or more, to include towns wherever possible. The position in Yorubaland is complicated by the very strong desire in many areas for local autonomy. Administrative autonomy means control of funds in the hands of the local elite, jobs for local people as opposed to outsiders, contracts for local entrepreneurs, and the provision of amenities locally rather than somewhere else. Amalgamation with a larger political unit is seen as meaning that a disproportionate share of resources will go to the larger unit. The pressures both for the formation of more states at the national level, and for increased autonomy at the local council level, reflect this definition of the situation.

A third problem concerns manpower. During the colonial period the problem was one of attracting enough educated men into local government to allow it to run efficiently. In the early 1970s, this appeared to have been achieved. Oyediran (1974) found that the members of the management committees appointed to run the local councils in 1972 were a highly educated group. Ninety per cent of them had secondary education or above, compared with 16 per cent of the councillors before the period of military rule. The new men tended to be lawyers, teachers, businessmen and administrators while their predecessors had been predominantly self-employed entrepreneurs. The present problem is that, with the increased powers of the state governments, the functions and powers of the local councils are no longer wide enough to interest recruits of this calibre. As Oyediran puts it, 'It seems fruitless to call for a better calibre of political and administrative leadership if local councils will be required to devote their time, energy and talent merely to clearing motor parks and local cemeteries' (ibid: 406). Whether the newly elected councillors will be able to alter this situation remains to be seen.

Finally, there is the position of the traditional rulers. Although their power and prestige have been gradually undermined, both by government intervention in the selection process and the growth of the bureaucracy, the oba are still popular and influential figures. The trend towards the selection of younger, more educated, rulers has continued. The present Alafin formerly worked for an insurance company, and the new Ataoja of Osogbo was a chartered accountant. Others, like the Sohun of Ogbomoso, are wealthy and literate businessmen. The continuing importance of the offices is indicated by the vigour with which they are still contested. Long gaps between reigns are common, while the rival candidates fight it out in the houses of the kingmakers, the courts and the corridors of the ministry buildings in the state capitals.

But why have the institutions retained this degree of vitality? It may be ascribed partly to their prestige, though prestige has not prevented more radical governments sweeping traditional rulers aside in other parts of Africa. It also suits the national leadership to maintain these institutions. Traditional rulers have two characteristics which both civilian and military governments in Nigeria have tended to lack: legitimacy and permanence. In the early days of nationalist politics, Chief Awolowo's initial success was based on the support of the senior rulers. In the constitutional crisis of 1966, Gowon turned to the Oni and other prominent rulers as 'leaders of thought' for their regions, and in 1976, the Nigerian government was again turning to the local rulers to persuade people to register and vote in the local government elections. Some of the most senior oba were appointed in the days of indirect rule. They have survived the colonial administration, two civilian regimes, and three changes of military leadership, to say nothing of a long series of local government reorganisations. Oba like the Oni of Ife, have retained their influence despite these changes.

Chiefs, on the other hand, are in a far less fortunate position. During the colonial period, the powers of the rulers increased at their expense, and their incomes grew steadily smaller in relation to those of other sections of the community. In pre-colonial times, the chiefs were the wealthiest men in the community. In the three towns where I carried out my own research, the chiefs were generally elderly, illiterate and poor. The more effective title-holders appeared to be retired entrepreneurs. They were still active in dispute settlement in their own wards and some retained a seat in the local court, but they had little wider influence. The only exceptions to this were the senior chiefs in Igbeti and Ogbomoso who had taken over some of the ruler's responsibilities, during his exile in the one case, and during the interregnum in the other. When there is no oba, the next most senior chief acts as spokesman for his community, but when the new oba is elected, he returns to comparative obscurity.

What of the future? The oba and chiefs will presumably retain their judicial role, unless the government decides to appoint professional magistrates to take over the local courts, in which most of the cases concern divorce and land tenure. They may retain a residual control over land allocation, though most strategically important land has either become private property or been taken over by the government. But the future of Yoruba chiefship ultimately depends on two other factors: the stability of national political institutions, and continued loyalty on the part of the average Yoruba to his town of origin, despite increasing geographical and social mobility. If there is a long period of stable civilian rule, then many functions which are now performed informally by the chiefs on behalf of their communities will be taken over by the elected politicians. At the moment, an influential oba is able to bring considerable benefits to his town. He is able to go directly to the top within the civil service to lobby bureaucrats on behalf of his community, and contacts of this sort are an essential qualification for office. From this point of view, it does not matter whether the oba is highly educated, as in Osogbo, or a wealthy entrepreneur, as in Ogbomoso. Both types of men are likely to have social networks useful to the town. Neither Igbeti nor Ogbomoso had a resident ruler in 1970, and informants in both towns thought that the government was unlikely to do much for them unless the right sort of ruler was appointed, in the absence of effective representation in the army or the civil service.

A corollary is that if home-town loyalties are eroded by education, migration and industrialisation, then the position of the traditional rulers, like that of many of the settlements over which they rule, will become increasingly peripheral. Political instability, on the other hand, will tend to reinforce communal and home-town loyalties once more, and enable the traditional rulers to continue to play an important informal role. But the major question is how traditional rulers can be given a forrnal role in politics. If they are given a degree of real power, either in local councils or in a State Council of Chiefs, they run the risk of being discredited once again by involvement in party politics, and subject to arbitrary manipulation by the party in power. If they are insulated from decision-making and relegated to a purely symbolic role, they are also likely to be insulated from the control of resources, particularly money and patronage. Denied access to these, chiefship in Yorubaland would become an empty anachronism. In either case, the prospects of the Yoruba oba and chiefs in a Nigerian Second Republic do not appear particularly good.