INDIVIDUAL RESEARCH REPORTS 31.
and acted generally as their intermediary with the King.
The people of Gyaman, who wanted to go to the coast (Accra) did
not pass through Kumasi. Their route lay through Nkoranza to Attebubu,
Krachi and thence to Accra.
The last Ashanti - Gyaman War 1818-1819
Osei Bonsu Panyin (Osei Tutu Kwame 1800-1824) sent Butuakwa to
demand from the King of Gyaman Adinkera, a Golden Stool/ which he was
alleged to have made and a fine of a thousand pounds for having made i t .
The people of Gyaman refused to deliver the stool and pay the fine. The
Brongs were attacked and the forces joined at the river Tano. The informant
refused to discuss the outcome of the war.
Kwame Arhin.
THE WIDER BACKGROUND OF THE SALAGA CIVIL WAR
Amongst the Salaga Papers which I have been compiling, are a
number which deal with the Salaga Civil War of the 1890s. Some
of these papers suggest that the dispute was of long standing and
throw light on a little-known period of Ghana history. Most of
the works cited in this paper are included in the Salaga Papers
now being issued by the Institute of African Studies.
The Civil War which broke out in Salaga in 1892 had its immediate cause
in a succession dispute. The Kpembe skin has, or is supposed to have a
rotating succession between the three branches of the ruling family, Affai,
Sungbum and Kanyase; these three chieftainships are known as the three
"gates" to Kpembe, whose chief had to pass through one of them. The
immediate cause of the dispute was that no Kpembe chief had been appointed
from the Kanyase family for many years.
Succession disputes in the Kpembe division are nothing new. In
1817, we hear of the deposition of an Alfai chief, and the attachment of
his stool (skin) to the "Chief of Premehinia" (?Kpembehene, i.e. Kpembe-
wura), his brother, on the orders of the Ashantihene J This arrangement
does not seem to have become permanent.
In the 1820s, there was a full-scale civil war. Kpembe tradition
remembers this in the reign of Dosi against whom Lepo and Kanyase families
were allied.2 The Qissat Salgha speaks of fighting for twelve years; 3
32. INDIVIDUAL RESEARCH REPORTS
this appears to have included the Krupe war, remembered in Krachi
tradition as taking place in the reign of the Asantihene Osei Yaw (1824-
1834):
"The cause of the war was that one Sunkpum of Gonja who was
the Sunkpumwura had wanted to claim the Kpembe stool. He
therefore went and appealed to the Asantihene to fight the
Kpembiwura and the Gonjas and that he, Sungbumwura, would
reveal the secret of the Gonja warfare, so that after the defeated
of the Gonjas he should be made the Kpembiwura. The Ashantl-
hene agreed and sent an army to fight the Kpembiwura. The
Kpembis, together with Leop, Wayan and Kenyase sent
messengers (to the Krachis).... "4
This traditional account goes on to tell of the defeat of the Ashantis and
Sungbums at Krupi (with the help of rain caused by Dente). This defeat
was followed by a punitive expedition led by Nubin or Nibire, who was
eventually defeated and beheaded by the Bagyamso chief.
This campaign may well be the same as that related to the Landers
by kola merchants who had been in "the city of Gonja" (evidently Salaga)
to buy kola in 1830. According to the Landers' informants, the first
Ashanti expedition, which was ambushed, went to punish the refusal of the
Gonja to take part in the 1824 war. According to the kola merchants,
(who do not mention any succession dispute) "the city of Gonja" was aban-
doned and destroyed in a second, punitive expedition. They regarded
Gonja as no longer part of Ashanti, but independent.5
This episode, which must be dated to 1829-30, seems to be part of
a very widespread rising against Ashanti. Accra and Akwamu, Akwapim and
Akim had all been persuaded to fight against Ashanti in 1826. Within a
few years of the end of the war, Juaben had quarrelled with Kumasi and
seceded to Akim, where the Juabens were busy selling guns and powder to
the north 0 ~ presumably to the people of Salaga and their allies, all
of whom had close relations with Juaben. These allies included part of
the Guan country, and extended south to Akroso and Pai on the Volta. The
Peki people, including many Ewe groups and the Nkonyas, broke free from
Akwamu at the same period.
INDIVIDUAL RESEARCH REPORTS 33.
The Salaga succession dispute of the 1830s must thus be seen as a
part of an anti-Ashanti rising which closely paralleled the rising after the
next major Ashanti defeat in 1874.
With the return of Juaben to the Kumasi allegiance after the death
of the Juaben chief Boaten and his brother in 1840, Ashanti control over
Salaga and Krachi was re-established. A Krachi tradition reported by
Ferguson mentions the capture and execution of a chief priest of Dente by
Dewia king of Atebubu.' Ferguson, in the 1890s, was told of the origins
I* of the later Salaga dispute:
"a long time ago, when.Salaga was under Kumasi, rival claims
in the stool of Kpembe were made by Sempe-wula/lempo-wula?7,
Kunaka-wula, in which both claimants fought with vaccilating ~ (
results. The King of Ashanti ordered Dewia, king of Atebubu,
to settle the matter, and it was arranged that the order of
succession should be, Pembi wula, Kunaka wula, Kanjase w u ! a " . °
Kunakawula is the Kanakulai wura, the chief of A l f a i ; "Pembiwula" in this
passage must be the Sungbumwura. This, and not the end of the civil war
of the 1890s, was probably the occasion of the oath mentioned by Tomlinson
(whose account of the civil war is inaccurate in a number of ways)?; by
this oath, sworn at Papitia, it was agreed that the houses of Sumbung, Lepo
(Alfai) and Kanyasi should succeed to Kpembe in turn.
It is not clear whether this settlement, which would probably have
been made in the 1840s, was the first imposition of a rotating succession,
or if a rotating succession had been imposed earlier, and had broken down
previously. It does not seem that a rotating succession ever operated
strictly in Kpembe.
If Salaga had undergone some twelve years of civil war at the time
of this settlement, it is possible that it took some time to recover; (we
know that when the town was sacked in 1892, it had not been properly
rebuilt by 1897). According to Berth's informants in about 1850 (whose
information may have been out-of-date), Salaga was at that period a town
of only some 1000 persons, where traders stayed no longer than was
necessary 10
34. INDIVIDUAL RESEARCH REPORTS
After the first civil war, the succession, as given in Qissat
Salgha runs as follows;!'
Deifu (Alfai) Sungunga Kanyase
1 . Jaware 2 . Kali
3. Sabalugu 4 . Dusi
5. Darhaman 6. Bambanga
7. Mahamadu
Count Zech gives the same list, but inserts Nabu and Shafu between Kali
and Sabaluwu. Daharman he gives as another Sabaluwu. 12
Dusi, the fourth in the Qissat Salgha list, was Kpembiwura when'
Bonnat visited Salaga. Bonnat found the people drawn up in three groups:
the "first group", whose chiefs sat on lion and leopard skins, must have
been the Alfai group, whose chiefs have this privilege. The Kenyase
group was that of the king's rival, Asumani; the king believed that Bonnat,
(who held the Ashantihene's commission) had come to replace him by
Asumani.'3 o u s i , as the Qissat Salgha list states, belonged to the Sungun-
gu family. Dusi was still king in 1882 (Lonsdale gives his name in an
itinerary as chief of Kpembe) ' 4 ; in 1884, David Asante refers to the
Alfai chief as the "successor" of the chief of Kpembe J 5 (Both Bonnat and
David Asante note that the Alfai chief is a Muslim). It seems that Dusi
succeeded at about the time of the rising against Ashanti in 1874-5; this
may explain Dusi's strong antipathy to Ashanti rule, and his fear that
Bonnat had come to replace-him; Sablugu, his predecessor, may well have
died during the rising. 16
In 1888 and 1889, the German officer von Francois visited Salaga
and made treaties with the old chief, to whom he gave a double-barrelled
gun with his name inscribed in Arabic letters on a silver crescent. 17 Von
Francois gives the name of the "Sultan" as Abu du Rahman, i . e . AbduIra-
nian, evidently the Darhaman of the Qissat Salgha. (Klose must be wrong
in stating that von Francois' chief was Mama; Mama was the next-but-one
Kpembewura, the next through the gate of A l f a i ) . Appropriately for a
chief of Alfai lineage, Abdulrahman had leopard skins.
INDIVIDUAL RESEARCH REPORTS 35.
It is possible that Abdulrahman died in 1889 or 1890; in 1890,
another German officer, Kling, presented to the Sultan another doubler
barrelled gun with the name of the Sultan in Arabic letters on a silver
crescent, apparently exactly similar to the one which von Francois had
given to Abdul rahman J 8 This is most easily explained if there was a new
Sultan. If so, this would have been Bambanga, whose reign must have been
short. In January 1892/ Kling found rr.uch new building going on in Salaga,
and in February of the same year he found a new quarter had been built in
Kpembe, with a new royal reception hall. 19 Again, the most likely
explanation is a new chief, who would be Muhammad Nafu, Klose's "Mama",
evidently an active chief.
By this time the Kanyase people were becoming very restive — they
had had no chief appointed through the Kanyase gate since the civil war,'
despite the agreement. Isafa Kabachi opposed the choice of Muhammad
Nafu, and when the.other two families joined against him to elect Nafu,
Kabachi retired to his farm. Nafu and his followers believed that Kabachi
was preparing to make war, and determined to anticipate him; they
attacked him at his farm, but he escaped, and obtained support from the
Yendi N a , king of Dagomba. The Dagomba forces routed the forces of
Alfai and Sungbum and went on to loot and destroy Salaga.20 This took
place in October or November 1892.21 Muhammad Nafu was killed or
taken prisoner during the fighting, and, according to Tomlinson, the Sung-
bum chief was killed.22 Abdul Karimu, an ex-officer of the Hausa force
at Accra, who had taken service under the Germans, took part in the
fighting, and one of those killed was Malam Ibrahima, head of the ex-
soldiers. 23 Many of the Salaga people went to Yeji under Abu Bokari, who
was probably Binger's host during his stay in Salaga in 1889.24
Some of the Salaga traders went to Krachi, where the Germans set
up a "New Salaga" with Mama's son Prince Lempo as "King of Salaga".*5
The remainder of the story is thus inextricably mixed with the Anglo-
German struggle for Salaga. The Germans made one attempt to instal this
chief in Salaga; the officer responsible was recalled.26 Later they seem
to have lost interest in Prince Lempo, (whom one of the British officer
called, not inappropriately, Prince Limbo).
36* INDIVIDUAL RESEARCH REPORTS
There seems to be no truth in Tomlinson's statement that Isafa had
his elder brother Kanyase enrobed as Kpembewura.*' All contemporaries
refer to Kabache as king; Kanyase succeeded him on his death in 1897>
Muslim moralists see in the story of Salaga's civil war a just punish-
ment for the misdeed of the Sultans — drinking, making free with other
people's property and womenfolk, addiction to soothsayers. *8 There is no
reason to doubt that the Sultans did in fact commit these misdeeds; but the
repeated breakdown of the succession system — completely in the 1830s,
perhaps nearly in the 1870s, and again completely in the 1890s — would
seem to argue some more fundamental defect in the system. If, as Ferguson
seems to suggest, the system of a rotating succession was imposed upon Kpem-
be (and presumably on Gonja as a whole) by the Ashanti overlords*9 to
prevent civil wars and also to prevent the establishment of a single strong
ruling family, then the breakdown of the system when Ashanti power was
withdrawn was not only understandable, but inevitable.
It seems that no one family was consistently supported by the Ashantis.
In the 1830s, it was the Sungbum family they supported; in 1876, the
Kpembewura expected Bonnat to support the Kanyase claimant. That civil
war did not occur in the 1870s, following the 1874 defeat, seems to be due
to the fact that the Kpembewura of that date had the support of the only
outside power which was in a position to influence Kpembe politics, Britain.
Had the Ashanti threat become more acute, no doubt the position of his
" r i v a l " would have been strengthened to a dangerous degree.
In the 1890s, the position was different. There were now two outside
powers, Britain and Germany, both anxious to exercise control over Salaga.
This is the classic situation for a civil war. It was aggravated by the
existence of two other powers, each strongly hostile to one of the European
powers — Ashanti opposed to Britain, Dagomba to Germany. With the
British at Atebubu and the Germans at Krachi, Salaga stood at the inter-
section of these forces; it would have taken a strong well-knit constitution
to withstand them. By i l l chance, the one authority which might have
presented the breakdown, the paramount skin of Gonja, was vacant at the
time, ith three chiefs contending for i t . ^
INDIVIDUAL RESEARCH REPORTS 37.
Without some superior authority, a rotating succession of this type
is almost bound to break down; if two sections join together against a third
they can monopolise the succession — until the third loses patience and
calls in an outside power. This is a situation which suits an overlord quite
well; he can always intervene to redress a balance. Without an overlord,
the result is a civil war with foreign intervention, which can benefit no one
but the foreigner.
References
1 . T.E. Bowdich - Mission to Ashanti p.396 and 4 0 1 .
2. Notes to list of Kpembe imams, J . A . Braimah.
3 . Qissat Salgha Tarikh Gonja, IASAR/1; 6; 1$; 2 6 1 .
4. Krachi tradition, collected J.E.K. Kumah, IAS unaccessioned
5. J . & R. Lander Journal of an expedition to explore the
Niger, V o l . II p. 191.
6. A . Riis Diary of journey to Akim in Magazin fur
die neuste Geschichte der. Evangelischen
Missions - und BibelgeselUchaften, Basel
TJ54UI
7. G . E . Ferguson Memorandum on Atebutu, 1893, para. 3.
8. Ibid. Letter to Adjutant of Atebubu Expeditionary
Force, C7917of 1896, p. 151. ^
9. H . H . Tomlinson Customs, constitution and history of the
Gonja people, 1954"
10. H. Barth: r Travels and discoveries in North and
Central Africa, Centenary edn. 1965.
Vol.111 p.645.
38. INDIVIDUAL RSEARCH REPORTS
11. Qissot Salgha
12. Count Zech; in Mitteilungen
g ous den deutschen
Schutzgebie ten 1896,
13. M . J . Bonnat Diary in I'Exploroteur, 1876
14. R. la T. Lonsdale Itinerary In Affairs of the Gold Coast
C3386 Enclosure to No.42.
15. D . Asante Journey to Salaga and Obooso in Geog.
Gesellschaft zu Jena, 1886.
16. Hausa account of the rising in Salaga (translation in this number of
. Research Review). I had not seen Dr. Goody's notes on this when'
this paper was written.
17. Von Francois * In Mitteilungen ou» den deutschen
Schutzgebieten, 1888 and 1889.
18. Kling - Ibid, 1890.
19. Ibid, ibid, 1893.
20. Qlssat Solgha; Zech, op. cit; tradition from Ma lam Alhasan Bisamah
., ^ ^ ^ Krachi, collected by Kumah,
IAS unaccessioned.
2 1 . The date must He between October 8, when Krause wrote a letter from
Salaga to the Royal Geographical Society of London about the geogra-
phical position of Salaga, and December 6, wrote a letter to Kreuz
Zeitung (which we have not yet been able to obtain) describing the
civil war Sn Salaga.
22. Tomlinson, op. cit.
23. Qissat Salgha
INDIVIDUAL RESEARCH REPORTS 39.
24. Armitage Report on Yeii in C O . African (West)
No.538 p. 157.
25. H. Klose Togo unter deutscher f lagge p. 362.
2,6. Insp. Parmeter Letter dated March 31 1897 in C O .
African (West) No.529 p.206.
27. Tomlinson, op. cit.
28. Al-Hajj Umar: Poem on the causes of the Civil War
IASAR/27.
29. G . E . Ferguson Letter to Adjutant in C7917 of 1896.
* . * •
30. G . E . Ferguson. Memorandum of interview at Salaga in
1892, in C O . African (West) No.497 of
1895. "The king of Gonja died about a
year ago, and there is anarchy in the
country. The kings of Kosoo, Gun and
Boniape were claimants to the stool".
Marion Johnson.
SIERRA LEONE ARCHAEOLOGY
In February and March 1966, I conducted a short survey of the
af£haeologyof Sierra Leone. This was sponsored by the Institute of African
Studies, Fourah Bay College, with the kind support of the Ministry of Lands,
Mines and Labour, the main purpose being to discover whether the archaeo-
logical potential of the country was sufficient to warrant, or demand, the
appointment of a Research Fellow in Archaeology.
It is surprising that so little work has been done there. The miniature
Museum is continuously crowded with people anxious to know of their
cultural heritage; and in addition to its variegated recent material culture,
