AKWAMU AND OTUBLOHUM: AN EIGHTEENTH-

CENTURY AKAN MARRIAGE ARRANGEMENT

I V O R WILKS

T

HIS paper may be taken as an essay in the application of anthropological theory to other fields, in this case history. In it I am concerned t o give a detailed account of a marriage arrangement that bound together the royal family of Akwamu and the ruling house of the Otublohum quarter of Accra. Akwamu, now a small state locked in the wooded hills of the Volta Gorge in south-east Ghana, was, in the late seventeenth and early eighteenth centuries, the strongest of all the Akan powers and centre of an empire that extended for about 250 miles along the Gold and Slave Coasts of Guinea. Accra, now the capital of Ghana, was incorporated within that empire by conquest in 1681, but recovered its independence half a century later. Unquestionably a whole series of historical events involving Akwamu and Accra would be quite incomprehensible without reference to the structural implications of the marriage arrangement linking the two. The arrangement persisted through four generations, spanning the greater part of the eighteenth century. Fortunately the period is a well-documented one. The archives of the Dutch West Indies Company are especially useful. In particular, sufficient genealogical detail is recorded in the various sources for the marriage arrangement t o be recognized as involving successive patrilateral cross-cousin unions. I ,4s one would expect, eighteenth-century writings contain no clear description of the character of the marriage arrangement. It must be remembered that Bosman, astute observer that he was, considered even matrilineal succession ' so perplexed and obscure, that hitherto no European has been able to obtain a clear description of it, as I am certain they never will'.z There is, however, one contemporary and unequivocal reference to the marriage arrangement as such, by a Dane who left Accra in I 7 j o : It has already been said, that it seemed as if the Akwamu King Akonno regretted having acted so harshly against the Accras, as one sees from the following, that he let them be called back and all the Europeans at Accra had to be his guarantors that he was well disposed towards them and would no more attack and raid them. The Accras came, and as a token of his friendship he gave them back the heads of their leaders. Indeed, what is more, he gave two of the foremost Accra chiefs an Akwamu noble-woman each: Kpoti received one of them, and a chief under the Dutch fort called Dako had the other. Akonno's stratagem in this was very nice; he intended to make one folk of the Accras and Akwamus, which could have happened had the Accras given their noble-women in the same way, though a hundred years would scarcely have sufficed to achieve this. His successors, however, spoilt all.3 I For cross-cousin marriage among the Akan see Rattray, 1927, ch. xxix; Field, 1948, p p 110-12; Fortes, 1950, p p 281-2.

Bosman, 1705, pp. 203-4. Romer, 1760, p. 142. I have used the modem forms of the names. 3

392

AKWAMU AND OTUBLOHUM : AN EIGHTEENTH-CENTURY

The Marriage Arrangement: conception and inception Diagram I shows the paradigm of the marriage arrangement that was to link Akwamu and Otublohum. Shaded figures belong to one exogamous matriclan (abustla) and non-shaded to a second.1 Two matrilineal descent groups are shown, the one headed by Akonno (or earlier by his uterine brother Ado) and the other by Otu. Akonno's group was the royal family of Akwamu, and therefore resident in the state

capital. Ado was king from 1689 until 1702, when he was succeeded by Akonno, Akonno ruled until 1725 and was succeeded by his sister's daughter's son Ansa Kwao. Otu was also an Akwamu but, in circumstances outlined below, had taken up permanent residence in Accra in the area that became known as Otu's quarter, Twi Otzlbomn, the modern Otublohum.2 Otu was from an Akwamu divisional stool family, his uterine brother being an xafokeen (literally, chief of war folk, i.e. war-lord) and probably Nifahene, commander of the right wing. When young, Otu was pawned by his family to the Dutch in Accraand so became a bondman of the West Indies Company.3 In the later seventeenth century it was not impossible for such bondmen (' company slaves ') to achieve considerable power. A head trading boy, chosen from among the company slaves, would be provided by the Dutch with all the outward symbols of chiefly authority: fine cloths, valuable beads, and gold charms adorned his person; horns, drums, and umbrellas were his to use on ceremonial occasions, and his own band of armed retainers accompanied him on missions into the interior to solicit trade, keep open the paths, and conduct parties of merchants in safety to the forts.4 Otu, a person of outstanding natural ability, also had the advantage of his contacts with Akwamu, then a powerful and expanding state lying across the immediate hinterland of Accra.5 He rose to power rapidly, becoming first a head trading boy for the Dutch in For the Akan abr/sua see Fortes, 1950, pp. 2 ~ 4 - see principally WIC 97-101, 404, and misc, arch. Nuygt's Diary. T/70/5: report from Cape Coast to Royal The only published account of Otublohum is that in Field, 1940, pp. 148-9 and 189-92. It will be African Company dd. 26 Nov. 1709. apparent that I consider it in need of correction. For the expansion of Akwamu see Wilks, 1957, 3 Biographical material for Otu is scattered, but passim. 61.

AKAN MARRIAGE ARRANGEMENT

393

Crkvecceur at Accra and later Chief Broker for the Dutch settlements on the Gold Coast, a position of considerable eminence. In this way Otu, with his dependent kin and slaves, came to settle near the Dutch fort in Accra. They were securely established there when the armies of Akwamu launched their offensive against the Accra kingdom in 1677. In a series of battles between that year and 1681 the entire kingdom was conquered, its king executed, and its status reduced to that of a province of Akwamu. The victorious invaders found in Otu the obvious choice for governor of the newly won province, which office so became contingently associated with that of the Dutch brokership. Whilst Otu appears to have assumed this function immediately, nevertheless the question of the future of the governorship was one that still had to be decided. The ofice could have been allowed to become heritable within Otu's lineage, but this the Akwamu king probably considered incompatible with the necessary degree of central control over what was, in view of the commercial vitality of Accra, a most important post. Furthermore, since Otu was brother of an Akwamu divisional chief there was the strong possibility of both the divisional stool and the governorship being held by the one person. Thus, by the normal working of the principle of matrilineal succession, Otu might succeed his brother on the divisional stool, or might be succeeded by his brother in the governorship, or both stool and governorship might pass to that one person who was a sister's son of Otu and his brother. In each case the result would be an undue strengthening of the divisional stool in relation to the paramount stool itself, which was presumably just what the king wished to avoid when he chose originally to grant the governorship to Otu rather than to reward one of his war-lords with it. The problem of the future of the governorship thus turned upon the reconciliation of the interests of the Akwamu king with those of Otu in Accra. The king could negotiate from a position of strength, since Accra belonged to him by right of conquest and Otu administered the province only by appointment. Otu's interests, on the other hand, could not be disregarded, since he had not only the support of his kin in Akwamu but also a powerful ally in the Dutch, whose broker he was, and with whom the king wished to maintain close commercial ties. In these circumstances a compromise arrangement appears to have been agreed upon. Since Accra belonged to the king the governorship naturally remained the royal prerogative. The royal prerogative, however, was to be exercised in such a way that the actual management of affairs in Accra would alternate between a member of Otu's lineage and one of the royal lineage, both alike governing in the king's name. This basic arrangement in itself achieved merely the representation of the interests of the two groups while doing nothing to prevent the growth of political rivalry and antagonism between them. What was further required was an instrument for effecting the reconciliation of the two sets of interests, so making practicable the sharing of the governorship. This, as I shall hope to show, was found in the marriage arrangement entered into by the two groups, the paradigm of which was shown in Diagram I. It will readily be seen that this arrangement involved a series of patrilateral crosscousin marriages between the two groups in successive generations. Initially, the Akwamu royal family gave a uterine sister of Ado and Akonno in marriage to Otu. In the next generation the Otublohum group reciprocated by giving one of Otu's sister's daughters in marriage to Akonno's sister's son Amu. Amu's wife was thus

394

AKWAMU AND OTUBLOHUM : AN EIGHTEENTH-CENTURY

his patrilateral cross-cousin. In the third generation the royal family again gave one of its women, Amu's sister's daughter, in marriage to Otu's sister's daughter's son Dako of Otublohum, whose patrilateral cross-cousin she was. The rule governing the interchange of women between the two groups may therefore be stated in this way : A group that gives one of its women in marriage in the one generation receives back in marriage a female offspring of that woman in the next generation. The principle of succession to the governorship of the province of Accra may now be formulated thus : Succession alternates between the two matrilineal groups, but the candidates for office must be sons of the previous occupant. This is made possible by the series of patrilineal cross-cousin marriages between the two groups. I t follows from the application of these rules that a man who holds office is followed in that office by one of his own sons. By the requirements of matrilineal descent and clan exogamy, however, a son will necessarily belong to a lineage other than that of his father; thus Otu was followed as governor of Accra by his son Amu of the Akwamu royal line, and Amu by his son Dako of Otu's line. Certain political advantages would clearly be expected to derive from this arrangement of rotational succession. Since any governor had either to be a member of the royal family or a son of a member, the continued allegiance of Otublohum to Akwamu would seem guaranteed, and the arrangement to share the office workable. In this way too it inight have been thought to secure a high degree of continuity of policy and of general political stability-in contrast with the situation had there been no marriage arrangement between the two groups, the Akwamu royal family simply sending any one of its members to Accra in alternate generations. Finally, that the governorship passed from father to son was perhaps considered of some additional importance in that the Accra peoples over whom it was exercised were themselves patrilineally organized. Otu exercised his dual function of Akwamu governor and Dutch broker until his death. In European records of the time he is met with under the name of Pieter Pasop, probably given him by his first Dutch master. He is known in Accra tradition as Otu Ahiakwa-ahia is a compassion name given, for example, to a child born after his father's death; akwa means ' bondman '. In 1701 the Dutch, expecting a French attack, reported that ' in the event of fighting Pieter Pasop and his men may be esteemed as good as any white men'.' In 1703 they paid tribute to ' our Chief Broker Pieter Pasop, who has made such an impression on the Blacks by his well linown valour and knowledge of the arts of war, and by being related to the mighty king of Akwamu'.ZFrom 1703 until his death in or about 1711 he undertook on behalf of the Dutch a series of important commissions, including one to massacre all the English to be found in the Elmina area. The Marriage Arrangement: stresses Before his death Otu had already given his sister's daughter in marriage to her cross-cousin Amu of the Akwamu royal family, and from this union a son, Dako, had been born. On Otu's death Amu left Akwamu to assume office in Accra. Since by I

WIC 97.

W I c 404.

AKAN MARRIAGE ARRANGEMENT

391

Akan custom Amu could claim house-room with his patrilineal (and incidentally his wife's) kin in Otublohum, he, and those who had accompanied him from Akwamu, assumed residence there not as strangers but by right. While it might be thought that the influx of a group of one matrilineage into a quarter hitherto belonging to a group of a different matrilineage would in itself give rise to tension and conflict, there is in fact nothing to suggest that this was the case in Otublohum. Indeed, that Otu's people and Amu's coexisted in harmony within the one quarter is strongly suggested by the present dispositions of the blackened ancestral stools of Otu, Amu, and Dako, which the writer was recently privileged to see.

+-A

=

A +-

DAKo

=&AIDZA

POST 1731

W AKWAMU RESIDENCE

I

r

J

OTUBLOHUM RESIDENCE

DIAGRAM I1

All three stools are housed in the one stool-room in Otublohum, in the custody of the one stool priest, and all are ritually purified and sewed in just the same way even though the affiliation of Amu's stool is with one matriclan and of Otu's and Dako's with another. Nevertheless, that there were difficulties inherent in the situation was to become apparent in 1721 when Akonno, Amu's maternal uncle, died. The difficulties resulted from the fact that in the case of Amu's group within Otublohum, the pattern of filiation (matrilineal) had ceased to correspond with that of residence (since 1711, patrilocal).I The situation is shown in Diagram 11, in which successive kings of Akwamu and successive governors of Accra have been brought into two distinct verticals-a vertical thus corresponding to residence in the one place. Akonno died an aged man, leaving no surviving brothers and in Amu, so it seems, a lone sister's son. Whilst Amu's subsequent actions show beyond doubt that he had expected to succeed Akonno as king, in fact the kingmakers of Akwamu passed him over and chose instead a sister's daughter's son of Akonno named Ansa Kwao. Their

' For the nature of this ' correspondence ' see Homans and Schneider, p a ~ ~ i m .

AKWAMU AND OTUBLOHUM : AN EIGHTEENTH-CENTURY 396 reasons are nowhere documented, but it seems clear that Ansa Kwao was preferred in that he had been continuously resident with his matrilineal kin in the Akwamu capital, whereas for some fourteen years Arnu had lived with his patrilineal kin in Accra, so prejudicing his claims on the stool. It must be stressed that for the Akan the principle of matrilineal succession is not a technique for deciding who shall succeed but only for determining who are the candidates. The principle yields a list of possible successors in order of the closeness of their womb kinship with the predecessor-uterine brothers, sister's sons, and so forth. Other t h n g s being equal, the actual succession should follow this order in accordance wi+hthe rule of primogeniture. Since other things seldom are equal, however, considerations of age, character, and especially of residence will all influence the decision as to which of the possible successors is in fact chosen. Amu's mortification at being left in the subordinate position of governor of Accra whilst his nephew succeeded to the stool of Akwamu was subsequently further increased by the fact that Ansa Kwao proved an injudicious ruler whose policies contributed greatly to the spread of discontent among the subject peoples. By 1728 Amu had resolved upon rebellion and, using his powers as governor, first placed the Accra quarters on a war footing, levying from them a tax that was then used to secure the alliance of malcontented subjects of Akwamu, such as the Berekuso of the hill country north of Accra.: In 1729 the rebel forces were joined by a section of the Akwamu people proper, the modern Aburi, who transferred allegiance from Ansa Kwao to Amu. The combined armies assumed the (Twi) name Akawapenl, 'the thousand companies '.Z Late in 1729 and early in 1730 a number of battles were fought between the rebel and the loyal forces, but neither side secured any clear advantage. At this critical time, however, Amu7s strength was depleted by the defection of the Accras, who were persuaded to return to their towns by Ayi Kuma, a representative of the pre-conquest Accra dynasty of Okai Koi,3 who saw in the struggle between Ansa Kwao and Arnu a chance to recover the independence of Accra. Fearing the collapse of his cause, Arnu committed what was to prove a serious error of judgement by soliciting support from Akyem Kotoku and Akyem Abuakwa, neighbouring Akan states whose long-standing enmity to Akwamu was even then a matter of history. The result was that what had started as an internal rebellion against Ansa Kwao rapidly assumed the character of a general war aimed at the total destruction of Akwamu power. In these circumstances some sections of Akwamu that had supported Arnu now returned allegiance to Ansa Icwao. The remaining rebel forces became little more than auxiliaries of the Akyem armies, and Amu found himself replaced as commander by the Akyem war-lord Ofori Dua. Late in 1730 the Akyem armies invaded Akwamu and sacked the capital. Ansa Kwao was captured and beheaded. Arnu could have gained but little satisfaction from his nephew's defeat and death. The surviving Akwamu forces abandoned the western parts of the empire and retired into their eastern territories, where, under a regent Akonno Kuma, they established a new state capital behind the barrier of the Volta river. Over the conquered Akwamu I The career of Arnu has been reconstructed from the following principal sources: KvG 43, 82-100, 291; WIC 109; VGK, letters and correspondence from Christiansborg 1727-31 ; B i ~ r n ,1788. It is quite clear from contemporary sources that

this is the origin of the modern name hkxvapim and not, as is usually stated, the Twi nkoa-apem, ' the thousand subjects '. Wilks, 1 9 ~ 7 pp. , 106-7.

AKAN MARRIAGE ARRANGEMENT

397

homeland the Akyems preferred to grant authority to a defeated Akwamu chief Kwesi Bibri rather than to Amu. A new state was created in the hills north of Accra, which took the name of the Akuwapem (the modern Akwapim), but this was given into the charge of Ofori Dua who became its first king. In Accra the Otublohum quarter, n o longer backed by the might of Akwamu, remained as an alien community barely tolerated for the contribution Amu had made towards the downfall of Akwamu. Henceforth, however, it would supply not a governor for Accra, but only a chief for its own people.

The Akwamtl Stlccession It will be appreciated that at the time of the defeat of Akwamu neither party to the marriage arrangement that linked Otublohum with Akwamu had acted in a way that could be interpreted as a specific repudiation of that arrangement. In particular, in raising rebellion against Ansa Kwao, Amu had acted in defence of his rights, whether real or supposed, as a member of the Akwamu royal family, and not as a member of Otublohum. Nevertheless, as we shall now see, Amu's subsequent activities were such as t o make impracticable the indefinite continuation of the cross-cousin marriage arrangement. Late in 1730 Amu was in Accra raising an expeditionary force with the avowed object of destroying the remaining Akwamu forces east of the Volta. The services of Akyem mercenaries were secured, as also the co-operation of an Akwapim (Aburi) chief Osae. Before taking the field Amu took into custody two of his lineage kin, first, his sister's daughter who had earlier been given in marriage to her cross-cousin Dako of Otublohum (see DiagramII), and second, the infant offspring of this marriage, also named Dako. These two accompanied Amu when, early in 173I, he led his forces out of Accra to establish a war-camp near the Volta. From the war-camp Amu announced his intention of attacking Akwamu as soon as the rains ceased, but during the delay he made an unexpected though clearly premeditated gesture of pacification towards Akwamu. Withdrawing his own warriors and the Akyem mercenaries from the war-camp, Amu marched south to Ada at the mouth of the Volta and using the chief there as an intermediary had the Akwamus informed of the location of the warcamp. An Akwamu army attacked immediately, and the unsuspecting Akwapims were annihilated. Having thus satisfactorily accomplished the sacrifice of his erstwhile allies, Amu then had the Akwamus further informed that the younger Dako and his mother (i.e. Amu's sister's daughter and her son) were in Ada. An Akwamu army, in I oo canoes, arrived there in November I 73 I, and took these two members of their royal family back to the relative safety of the new capital.= In the meanwhile Amu and his forces had crossed the Volta estuary and, having renounced their threats against Akwamu, were content to pillage the coastal towns. They sacked Keta late in 173 I, were fighting in Little Popo in 1733, and were preparing to attack Dahomey in the same year when Amu's death intervened to cause a halt in their depredations. The sequence of events is t o be understood by reference once more to the marriage arrangement between the Akwamu royal family and Otublohum. From Diagram I1 it may be seen that immediately prior to the Akwamu collapse of 1730, when Ansa Kwao still ruled over Akwamu and Amu governed Accra on the king's behalf, the R ~ m e r ,1760, pp.

~d

279-22.

AKAN MARRIAGE ARRANGEMENT

399

was only because it still held the privilege of the Dutch brokership. Effective control over Accra affairs had passed to descendants of the Accra king Okai Koi who had been executed by the Akwamus in 1677, and especially to Ayi Kuma and his younger brother Okaidza. Despite an attempt to effect a reconciliation between the true Accras and the locally resident Akwamu people by the marriage of Amu's daughter Amonoa to Okaidza, Dako's short reign was especially marked by the deterioration of his relations with his brother-in-law. For a few years Dako and Okaidza did indeed maintain an uneasy peace, but this ended abruptly in 1737, when the Dutch officially confirmed Dako in the brokership which he had in effect exercised since Amu's death. Okaidza immediately entered into negotiations with the Danes in Christiansborg and with Owusu, the heir apparent to the Akyem Abuakwa stool, for a joint attack upon Dutch Accra to exterminate ' Dako and his following '. Okaidza's claim was that Dako was in league with the Ashantis, who were at that time threatening an invasion of Akyem and the coast, one of the avowed objects of which was ' to set the Akwamus up in their country again '.I In 1738 and 1739 there were several minor battles between Okaidza with Danish and Akyem support, and Dako with Dutch. In 1742, when the long expected Ashanti invasion finally materialized, Dako happened to be in Akyem. He sought refuge in Akwamu. There a fellow fugitive revealed his identity to an Akwamu chief Popisaa who, despite Dako's offer of 200 ounces of gold for his life, had him beheaded so that ' he should pay for what his father had done, who was the chief instrument of the killing of the Akwamu country ' . 2 Thus in 1742 perished the Otublohum manth Dako, son of Amu the rebel governor of Accra, and father of Dako the future king of Akwamu. Shortly after his death it was reported that ' Okaidza had obtained two of the upper teeth of Dako for eight dollars and had caused holes to be bored through them, and wore them with similar objects around his neck '.3 Although the conflict between Dako and Okaidza was primarily a result of the political rivalry between the Accras and the Akwamu section in their midst, pet its peculiar acuteness is to be understood by reference to the relationship between a man and his sister's husband. This, for the matrilineal Akan, may yield the classic case of conflict between the mother's brother and the father.4 Thus, even granting the prevalence of the matrilineal rule in Accra, Dako's exercise of his rights (as mother's brother) over Okaidza's children by Amonoa could well disturb the relationship between the two. Since, however, the Accras were in any case patrilineally organized, Okaidza could refuse to recognize that Dako had any such rights whatsoever. The matter was a singularly important one, for, if the marriage arrangement between Otublohum and the Akwamu royal family was to continue, it was essential that Dako should be able to give his sister's (and Okaidza's) daughter in marriage to the younger Dako who, not having succeeded to the Akwamu stool at this date, was still in line for the Otublohum mantJc-ship. Something of the confusion is reflected in the Dutch attitude towards Okaidza's progeny. In 1742 his eldest son Okanta was captured by the Fantis but was ransomed through the Dutch, who considered him one of their broker Dako's people.5 In 1766 Okanta committed adultery with one See Rattray, 1927,p. 322.

KvG

293.

400

AKWAMU AND OTUBLOHUhl : AN EIGHTEENTH-CENTURY

of the wives of his cousin, Dako king of Akwamu, but when Dako claimed compensation from Otublohum the Dutch equivocated and pointed out that Okanta belonged t o the house of Okaidza as well as the house of (the senior) Dako.1 Finally, however, Okanta succeeded matrilineally, as the senior Dako's sister's son, to both the Otublohum man([&-ship and the Dutch brokership in 1790, and died eight years later. As we have seen, the Otublohum fjzantk Dako was executed in 1742 in the course of the Ashanti invasion of Akyem. Despite reports that he had been detained in Akyem against his will, the victorious Ashantis took the view that Dako had been there to oppose them. An Ashanti embassy was sent to Accra to demand that Dako's people, and Amu's and Otu's, should all be handed over to them in pacification. The Dutch, however, represented the interests of their brokers and the embassy finally agreed to a financial settlement. Despite the unexpectedly early death of Dako any crisis in the Otublohum succession was averted by the choice of Ofei as next mantJ~. Earlier sources describe him as a brother of Dako senior, later ones as of Dako junior. There was probably a simple confusion of names, since only the former relationship is chronologically possible. For twenty-nine years Ofei ruled Otublohum, for the last twenty-four of which his brother's son, the younger Dako, was king over Akwamu. Under Dako, and in close alliance with Ashanti, Akwamu became powerful once more,z and as a result Otublohum too began t o recover something of its former importance in Accra affairs. The death of Ofei in 1771, however, was t o raise in an acute form the whole question of the relationship of Otublohum to Akwamu, since at that date the sole person satisfying the full requirements of the Otublohum succession was Dako who, as king of Akwamu, was scarcely likely to advance his claims. The difficulties in the situation may best be appreciated by considering how it must have appeared at the time to the representatives of the two parties to the original marriage arrangement, namely, the Akwamu royal family and Otu's section of Otublohum. T o Otu's matrilineal descendants in Otublohum it must have seemed that in removing the younger Dako and his mother from that quarter in 173I Amu, as a member of the Akwamu royal family, had in effect repudiated the cross-cousin marriage arrangement between the two groups, even though the effects of that repudiation had not been felt fully until forty years later. Otu's people might therefore have claimed that Ofei's successor should be chosen from, and by, them. The Akwarnu king on the other hand, intent upon restoring Akwamu influence in Accra, was scarcely likely to concede such a claim, and in any case probably took an uncompromising view of the matter, since the Otublohum succession had already passed from Dako t o Ofei, both members of Otu's lineage, at a time when Akwamu had been powerless to uphold its interests in Accra. Furthermore, it was arguable that in 173 I Amu had acted as an individual and not as a member of the royal family, from which he had been estranged by twenty years' residence in Accra and with which in any case he had been in armed conflict. It must be borne in mind that although Otublohum was no longer politically subject to Akwamu in 1771 yet nevertheless the active hostility of Akwamu could have serious consequences, especially in the commercial sphere, since much of the trade from Ashanti to Accra passed through Akwamu middlemen. It is therefore KvG. 127.

Dupuis, 1824, p. 234; B i ~ r n ,1 788, p.

21 I .

AKAN MARRIAGE ARRANGEMENT

401

perhaps not surprising that Otublohum finally accepted a successor to Ofei from the Akwamu royal family. Ofei's successor was Oto.1 He was the son of Amu's sister's daughter, the mother of the younger Dako, by a second marriage contracted after her return to Akwamu in 1731 (see Diagram 111). He was therefore a uterine half-brother of Dako, king of

J -

RESIDENCE IN AK'AIAP!U

RESIDENCE IN DADEDANA SECTION

RESIDENCE IN ATIFI SECTICN

OF OTUBLOHUM

OF OTUBLOWUH

Akwamu, and in this lay the strength of his claims on the Otublohum succession; the weakness, that he was not a son of the senior Dako. That Oto's succession necessarily involved the end of the series of cross-cousin marriages that had linked the two groups for almost a century may be seen from the fact that even had Otu's lineage given the senior Dako's sister's daughter in marriage to Oto, nevertheless she would not have been Oto's real or classificatory cross-cousin.2

Dadebuna and At@:dissociation It will be remembered that Amu, upon leaving Akwamu to assume the governorship of Accra in 171I, had been able to claim house-room with his patrilineal kin in Otublohum. When Oto left Akwamu for Accra sixty years later, however, he was unable to advance any such claims since his patrilineal kin were not of Otublohum. Oto was therefore obliged to set up his own house in Otublohum, which quarter so Oto is the same name as Otu, the latter showing G2 influence. The variants have been used here to avoid confusion. Similarly the Twi name is Amo, and Amu the GB form.

The principal sources for the careers of Dako of Otublohum, Dako of Akwamu, Ofei, and Oto are KvG 10-67; WIC 141-50; Romer, 1760; Bierm, 1788; Isert, 1793.

402

AKWAMU AND OTUBLOHUM : AN EIGHTEENTH-CENTURY

became divided into two sections, Dadebana and Atifi, as shown in Diagram 111. Oto is known in tradition as Oto Braffo, a word that in the eighteenth century had the sense o f ' general '. Under his leadership Otublohum maintained its recovery, and in 1784 Oto was chosen to command a combined force of Accras and Akwapims in a war against Arjlo, in which capacity he is to be met with in the writings of Isert, a Danish physician, who himself accompanied the armies. Although mention has already been made of Okanta's succession to Oto in I 790, later developments are beyond the scope of this paper, though something must be said of the progressive dissociation of Dadebana from Atifi, the course of which has been marked by a series of land and stool disputes. The Atifi section rightly regards Oto BrafTo as its founder, and still maintains especially cordial relations with Akwamu : the Atifi stool is described as zvofase, ' sister's son ', of the Akwamu paramount stool. This section, although in the midst of Accra, has not adopted the Accra custom of circumcision and still adheres to fairly well-defined principles of matrilineal succession and inheritance. The Dadebana section, on the other hand, practises circumcision and admits a compromise between matrilineal and patrilineal principles. That Dadebana should show a higher degree of assimilation to Accra custom is of course to be expected, since it was originally founded by Otu some 280 years ago, whereas the origins of Atifi are to be sought a full century later in its foundation by Oto Braffo. Nevertheless signs are not lacking of positive dissociative processes working to increase the distinctness of each section: Dadebana could be said to welcome, Atifi resist, acculturation. Such dissociative processes are particularly well exemplified in the Dadebana claim to be not of Akwamu, but of Denkyera, origin. The view that Otu Ahiakwa was Denkyera was first advanced by Reindorf in 1895,' but did not achieve immediate currency in Dadebana. It was recorded from there neither by Crowther in 1907 nor by Quartey-Papafio in 191I . ~By 1931 it h2d become official Dadebana tradition and, in Reindorf's version, was stated in evidence before the Supreme Court in AJee Ankrah Qzlansah v. Manche Aqbonsah. il curious chance made it possible t o transform Reindorf's error into official tradition. A cominon Akwamu emblem is that of a bird looking backwards, with which is associated the saying : T9 w'ahiri a fa, which may be rendered, ' If he (the enemy) falls on your rear, seize him '. This emblem is also known and used in Otublohum which, lying on the coast, had once been regarded as the rear of Akwamu. One of the Otublohum horns blows : Oto e Oto e, Ayer~w'aAyiri o, with the sense of ' Oto, Oto, face your rear '. An Otublohum war-song tells of Akoto king of Akwamu who, when he went to war, 9de a b i r i gyawyn, ' left us in the rear '. In Otublohum the device of the backward-looking bird is associated with the legend danhirifa, ' the rear-facing people '. It would appear that at some time the legend was misinterpreted as Danhirafo, ' the (nrgjlii) Denkpera people ', in order to make acceptable the view that while the Atifi section was unquestionably of Akwamu origin the older Dadebana was of Denkyera, and so to assert in a fundamental manner the distinct identities of the two sections. That these dissociative processes developed only after the arrival of Oto Braffo in Otublohum is interestingly confirmed by the present locations of the blackened ancestral stools of Otublohum. Thus whilst Otu's, Amu's, and Dako's are, as we have Reindorf, 1895, p.

28, but

see also p. 104.

Crowther, 1907; Quartey-Papafio, 191I, p. 435.

AKAN MARRIAGE ARRANGEMENT

403

noted, all kept together in Dadebana and ritually served without distinction of clan, Oto's stool is not with them but is kept in Atifi and has its own stool priest even though its clan affiliation is in fact that of Amu's. In this paper we have examined in detail the utilization for political ends of a marriage arrangement based upon successive cross-cousin unions. It may be mentioned in conclusion that there are clear indications that this was no isolated phenomenon, but exemplifies a standard feature of Akwamu administrative practice. The several Adangme states which lie between Accra and the Volta, for example, all appear to originate from administrative units of the Akwamu imperial period; they had earlier been integral parts of the old kingdom of Ladoku which was overthrown by Akwamu in 1679 and subsequently partitioned into a number of small administrative units. With the collapse of Akwamu power in 1730 these units became independent states over which in most cases the Akwamu administrators succeeded in retaining control and so in founding the present ruling houses. Old documentary sources and modern social structure alike suggest that marriage arrangements were extensively used by Aliwamu t o make its administrators acceptable to those whose traditional authority they were t o appropriate.

Acknowledgements and sources I have received much help and courtesy from Nii Amu Nakwa, Otublohum Mantfi, and his Elders, and from Nii Ankrah of Dadebana and his Elders. The assistance of my colleagues, Mr. E. F. Collins and Mr. P. C. Gibbons, has been invaluable in discussion, as has that of my wife with interpretation. I am grateful to the University College of Ghana for a grant towards the costs of research. The major sources of the historical material used in this paper are the archives of the Dutch Possessions on the Coast of Guinea (KvG) and those of the General Chartered West Indies Company (WIC) in the State Archives, The Hague. I have also used the records of the West Indies and Guinea Company (VGK) from the Danish Archives, and those of the African Company (T/70) from the Public Record Office, London. Extensive series of photostats of, and transcripts from, these records were made by the late Mr. J. T. Furley, and are now in the library of the University College of Ghana. The following are also cited: BIBRN,A. ' Beretning 1788 om de Danske Forter og Negerier ', in Thaarups Archiv, iii, Copenhagen, n.d.

BOSMAN, W. A New and Accurate Descr$tion of the Coast of Guinea, London, 1705.

F. G. Inquiry into Accra, 1907.(MS. in the National Archives of Ghana.) CROWTHER,

J. Journalofa Residence in Ashantee. London, 1824. DUPUIS, FIELD,hl. J. Social Organization of the Gd People. London, 1940. -Akim-Kotoku. London, 1948.

FORTES,hI. ' Kinship and Marriage among the Ashanti ', in African Sysfenls of Kinship and hhrriage (ed.

Radcliffe-Brown and Forde). Oxford, 1950. HOMANS, G. C., and SCHNEIDER, D. 31. i\.larrtage, A u t h o r i ~ and , Final Cawes. Illinois, 1955. ISERT,P. Vqages en Guinb. Paris, 1793. QUARTEY-PAPAFIO, A. S. ' Kative Tribunals of the Akras ', in Journ. A f r . Soc. x, 39, 40, and 41, 1911. RATTRAY, R. S. ReligiDn and A r t in Ashanti. Oxford, 1927. REINDORE, C. C. History ofthe Gold Coast and Rrante. Basle, 1895. RBYER,L . F. Ti(forlade1igEfierretning om Kjsten Guinea. Copenhagen, 1760. WILKS,I. ' The Rise of the Akwamu Empire, 1650-1710 ', in Trans. Hist. Soc. Ghana, III. ii, 1957.

AKWAMU A N D OTUBLOHUM

AKWAMU ET OTUBLOHUM: U N SYSTEME MATRIMONIAL AKAN nu DIX-HUITIEME SIECLE Le mariage entre cousins a Ctk Ctudik chez les Akan du Ghana, par Rattray, Field, Fortes et d'autres. I1 est ici CtudiC en dktail un exemple historique de ce systkme matrimonial bask sur des unions successives entre cousins. Lorsque Akwamu conquit Accra en 168 I, la personne qui devait manifestement Ctre dksignke commc gouverneur de la nouvelle province ktait le chef d'un clan Akwamu dCja domicilik a Accra, et qui appartenait a une lignke Akwamu de tabouret divisionnaire. Desireux de garder le contrble d'Accra et peu dispose B laisser le tabouret divisionnaire devenir trop puissant, le roi Akwamu ktablit un systkme de succession rotatif, de sorte que les affaires d7AccraCtaient alternativement gkrkes par des membres de la lignke royale Akwamu et par des membres du clan Akwamu domiciliks dans la localiti. Au cours du dix-huitikme sikcle, ces deux groupes matrilinkaires se sont lies par des mariages patrilatkraux entre cousins dans les gtnkrations successives, de telle sorte que, malgrk le fait que l'autoritk sur Accra ait CtC exercke successivement par les deux groupes, c'etait toujours son propre fils qui succidait a un gouverneur dans ses fonctions. Toute une skrie d'evknernents qui ont exerce une influence sur Akwamu et Accra seraient complktement incomprehensibles pour l'historien s'il ne tenait pas compte des implications de cette structure matrimoniale: il est demontrk qu'elle a CtC largement exploitee par Akwamu pendant sa pkriode impkriale comme instrument d'administration politique.