The "Hamitic Hypothesis" in Indigenous West African

Historical Thought

Robin Law

History in Africa, Volume 36, 2009, pp. 293-314 (Article)

Published by Cambridge University Press

DOI: https://doi.org/10.1353/hia.2010.0004

For additional information about this article

https://muse.jhu.edu/article/399722

Access provided at 16 Oct 2019 08:48 GMT from Princeton University

 THE “HAMITIC HYPOTHESIS” IN INDIGENOUS WEST
AFRICAN HISTORICAL THOUGHT

ROBIN LAW
UNIVERSITY OF STIRLING

This paper explores the use of versions of the “Hamitic hypothesis” by

West African historians, with principal reference to amateur scholars rather
than to academic historiography. Although some reference is made to other
areas, the main focus is on the Yoruba, of southwestern Nigeria, among
whom an exceptionally prolific literature of local history developed from
the 1880s onwards.1 The most important and influential work in this tradi-
tion, which is therefore central to the argument of this paper, is the History
of the Yorubas of the Rev. Samuel Johnson, which was written in 1897
although not published until 1921.2

II

The concept of the “Hamitic hypothesis” appears to have been coined by

the historian St Clair Drake, in 1959.3 In the historiography of Africa, it has
conventionally been employed as a label for the view that important ele-
ments in the cultures of sub-Saharan Africa, and more especially elaborated

1See Robin Law, “Early Yoruba historiography,” HA 3(1976), 69-89.

2Rev. Samuel Johnson, The History of the Yorubas from the Earliest Times to the Begin-
ning of the British Protectorate (London, 1921; reprinted Lagos, 1937, and frequently
thereafter). For assessments of this work, see Toyin Falola, ed., Pioneer, Patriot and
Patriarch: Samuel Johnson and the Yoruba People (Madison, 1993); Michel R. Doort-
mont, “Recapturing the Past: Samuel Johnson and the Construction of the History of the
Yoruba” (Doctoral thesis, Erasmus University of Rotterdam, 1994).
3See St. Clair Drake, “The Responsibility of Men of Culture for Destroying the Hamitic
myth,” Présence Africaine 24/25 special issue, 2nd Congress of Negro Writers & Artists,
Rome, 26 March–1 April 1959, 228-43.

History in Africa 36 (2009), 293–314

294 Robin Law

state structures, were the creation of people called “Hamites,” who were
presumed to be immigrants/invaders from outside, often specifically from
Egypt or the upper Nile valley, and racially Caucasian (or “white”), who
conquered the indigenous black African populations. One of the most influ-
ential proponents of this interpretation was C.G. Seligman, in a book origi-
nally published in 1930, which was reprinted down to the 1960s, and still
formed part of the background reading of the earliest generation of academ-
ic historians of Africa (including myself). Seligman declared baldly that
“the civilisations of Africa are the civilisations of Hamites,” and that these
Hamites were “European” (i.e., racially “white”) pastoralists, who were able
to conquer the indigenous agriculturalists because they were not only “bet-
ter armed” (with iron weapons, which they are suggested to have introduced
into sub-Saharan Africa), but also supposedly “quicker witted.”4 The idea
thus incorporated an explicit assumption of “white” racial superiority, and
denied historical creativity to black Africans by attributing their cultural
achievements to the impact of outsiders.
Although the overt racism of the “Hamitic hypothesis” was repudiated
by the academic historiography of Africa which developed from the 1950s,
the model of state formation through invasion and/or cultural influences
from outside continued to exercise a powerful influence. The early works of
the pioneer historians John Fage and Roland Oliver in the 1960s and 1970s,
for example, continued to posit diffusion of the institution of “divine king-
ship” from Egypt to the rest of Africa, and the formation of the earliest
states in the West African Sahel through the conquest of the indigenous
(black) agricultural peoples by Saharan (white) pastoralists—the military
superiority of the latter being now attributed to their possession of horses,
rather than (or as well as) iron technology.5 A more recent reflection of such
views is the interpretation of Dierk Lange, who posits the pervasive influ-
ence of “Canaanite-Israelite” models of cosmology and political organiza-
tion in several areas of western Africa, including Yorubaland.6
The classic racialist version of the “Hamitic hypothesis” propounded by
Seligman was, in fact, not the only, or even the original, version, but only
the last in a series of transformations. The historiographical evolution of the

4C.G. Seligman, Races of Africa (4th ed.: London, 1966), 61, 100.
5E.g., Roland Oliver & J.D. Fage, A Short History of Africa (Harmondsworth, 1962),
chaps 4-5; Fage, A History of West Africa (Cambridge, 1969), 8-10; Roland Oliver &
Brian M. Fagan, Africa in the Iron Age, c.500 B.C. to A.D. 1400 (Cambridge, 1975), 41-
2, 67-9. In a later formulation, Fage abandoned the hypothesis of the Egyptian origins of
African kingship, but still posited a critical role for conquering Saharan pastoralists in the
process of state-formation in the Sahel: A History of Africa (London, 1978), 39-43, 61-9.
6Dierk Lange, Ancient Kingdoms of West Africa: Africa-Centred and Canaanite-Israelite
Perspectives (Dettelbach, 2004).
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 295

“Hamitic hypothesis” was traced by Edith Saunders in a study published in

1969, whose general framework (if not all of its details) remains
persuasive.7 The origin and first version of the idea of the “Hamites” derives
from the Jewish Old Testament, in the story of the dispersal of the sons of
Noah after the Flood in Genesis 9–10. In this account, Noah had three sons
called Shem, Japheth, and Ham, who were held to be the ancestors of the
various peoples known to the ancient Israelites. The division among these
three branches of humanity was evidently geographical rather than racial,
with the descendants of Shem representing peoples of the center and east
(including the Israelites themselves), those of Japheth those to the north
(including the Javan, or Greeks), and those of Ham those to the west and
south—or perhaps more specifically, Egypt, with neighboring countries
within its sphere of influence. The sons of Ham thus include persons who
stand for the peoples of northeastern Africa—Mizraim (i.e., Egypt), Cush
(i.e., Nubia, or in terms of modern political geography, northern Sudan),
and “Put” (or “Phut”), which last is probably to be identified with the
“Libyans,” to the west of Egypt;8 and also Canaan, eponym of the Canaan-
ites (also known as “Phoenicians”), the inhabitants of the coastal area of
Palestine and Syria.
The account in Genesis is not merely a model of the peopling of the
world, but also a political charter, with the story of the curse pronounced by
Noah upon the descendants of Ham, provoked by the latter having inadver-
tently seen his father naked, in a drunken stupor, which condemns them to
be servants to the descendants of Ham’s brothers, Shem and Japheth. As
told in Genesis, Noah’s curse is directed (not very logically) specifically
against one of Ham’s sons in particular, Canaan, and was clearly intended to
provide a justification for the Israelite occupation of eastern Palestine, and
the dispossession and subjection of the indigenous Canaanites which this
involved.
From the Jewish tradition, the concept of the peopling of the earth
through the dispersal of Noah’s sons and their descendants (with the associ-
ated story of Noah’s curse) passed into both the Christian and Islamic tradi-
tions. As knowledge of Africa expanded, new African peoples (including
black Africans) were fitted into the genealogy as further descendants of
Ham, and especially of Ham’s son Cush (though sometimes, alternatively,
of Canaan). In time, in fact, the descendants of Ham came to be conceptual-
ized as predominantly or wholly black. Embroidery of the original story

7Edith R. Saunders, “The Hamitic Hypothesis, Its Origin and Functions in Time Perspec-
tive,” JAH 10(1969), 521-32.
8Cf. “Putaya,” the name given to the province formed in this area after its conquest by
Persia in 512 BCE: A.T. Olmstead, History of the Persian Empire (Chicago, 1948), 149.
296 Robin Law

explained the blackness of Africans as a consequence of Noah’s curse. This

story probably originated, as Saunders suggested, in post-Biblical Jewish
traditions collected in the Talmuds, compiled between the fourth and sixth
centuries CE.9 It had also entered Christian discourse by this time, and was
inherited by Islamic writers from the seventh century onwards.10
In this second version, therefore, the “Hamitic hypothesis” acquired an
explicitly racialist character. Noah’s curse condemning the descendants of
Ham to servitude was now used to justify the enslavement of black
Africans.11 This supposed scriptural justification was, notoriously, cited by
European and North American defenders of slavery from the seventeenth to
nineteenth centuries.12 It also had some currency in the Islamic world.
Although the institution of slavery was less thoroughly racialized in the
Islamic world than in the European colonies in the Americas, there was a
widespread popular prejudice linking black Africans with slavery, which
Noah’s curse was invoked to explain and justify.13 This view is cited and
critically discussed, for example, by the West African Islamic scholar
Ahmad Baba of Timbuktu, in a treatise on the legalities of enslavement,
written in 1614/15. Ahmad Baba himself refuted the justification of slavery
through Noah’s curse, together with its racialist implications, reiterating the
orthodox Islamic view that the only significant distinction to be made was
religious rather than racial, and that only those Africans who were not Mus-
lims could legitimately be enslaved.14
The classic version of the “Hamitic hypothesis,” in which the “Hamites”
became white invaders rather than black indigenes of sub-Saharan Africa,

9Saunders, “Hamitic Hypothesis,” 521-22. This view is contested by Ephraim Isaac,

“Genesis, Judaism and the Sons of Ham,” in John Ralph Willis, ed., Slaves and Slavery in
Muslim Africa (2 vols.: London, 1985), 1:75-91; but the latter’s own account in detail
concedes that the idea of blackness deriving from a curse on Ham does occur in certain
Talmudic texts (88, 90n44). See also David M. Goldenberg, The Curse of Ham: Race and
Slavery in Early Judaism, Christianity and Islam (Princeton, 2003), 102-06.
10Goldenberg, Curse of Ham, 99-102.
11Strictly, the classic “curse of Ham” represented the conflation of two stories which
were originally distinct: the Biblical story of a curse condemning the descendants of
Canaan to slavery, and the post-Biblical story of a curse of blackness on the descendants
of Ham collectively: for this development, see ibid., 170-74. Goldenberg finds the origin
of this “dual curse” in Christian and Islamic (but not, initially, Jewish) sources from the
seventh century onwards, or even more specifically (ibid., 170) “in seventh-century Ara-
bia.” However, the earliest source explicitly cited is of the eighth century.
12David Brion Davis, The Problem of Slavery in the Age of Revolution, 1770-1823 (Itha-
ca, 1975), 539-41, 553-54.
13Bernard Lewis, Race and Color in Islam (New York, 1971), 66-67.
14Bernard Barbour and Michelle Jacobs, “The Mi‘raj: a Legal Treatise on Slavery by
Ahmad Baba” in Willis, Slaves and Muslim Society in Africa, 1:125-59.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 297

emerged only in the nineteenth century. In this third version, the concept of
the “Hamites” was disconnected from its original Biblical context, and
given a modern “scientific” form, in terms of the racialist anthropology of
the period. Saunders suggests that the initial stimulus to the reconceptual-
ization of the Hamites as racial Europeans was the rediscovery by western
Europe of ancient Egypt, which was in large part a consequence of the tem-
porary occupation of Egypt by the French under Napoleon in 1798. Having
now to acknowledge the impressive historical achievement of Egypt, Euro-
peans reconciled this with their assumption of racial superiority by reclassi-
fying the ancient Egyptians (who had previously generally been regarded as
“black”) as racially “white.”15 This reconstruction of history was not, of
course, uncontested. In particular, African-American and African intellectu-
als in the late nineteenth and early twentieth centuries continued to insist
that the ancient Egyptians were racially Negroes (or at least a mixed popula-
tion containing a significant Negro element), and that their history thus pro-
vided proof of the latter’s capacity for cultural achievement—as, notably, in
the works of the Liberian writer Edward Wilmot Blyden and of W.E.B. Du
Bois in the United States, which anticipated the modern “Afro-centric”
claiming of ancient Egypt as a black African civilization.16 But the new per-
ception of the ancient Egyptians as racially “white” did become dominant in
European thought at both academic and popular levels.
The “Hamites” were thus now conceptualized as the lighter-skinned peo-
ples anciently settled in northern Africa, including the Berbers to the west
and some of the peoples of northeastern Africa, as well as the ancient Egyp-
tians, as distinct from the black Africans further south. The matter was com-
plicated by the adoption, in the same period, of “Hamitic” also as a linguis-
tic term, to refer to a group of languages—ancient Egyptian, Berber, Hausa
and related languages, and some of the languages of northeast Africa
(including Galla and Somali)—which were recognized as related to the
Semitic languages (Hebrew, Arabic, Amharic, etc.).17 The linguistic and
racial usages were in fact overlapping, rather than precisely congruent, in
their application, but were commonly conflated. One consequence of this
introduction of a linguistic criterion which should be noted is that the

15Saunders, “Hamitic Hypothesis,” 524-27.

16Hollis R. Lynch, Edward Wilmot Blyden, Pan-Negro Patriot, 1832-1912 (Oxford,
1967), 55-56; for an example of his argument, see an aside in the lecture “Philip and the
Eunuch” (1882), reprinted in his Christianity, Islam and the Negro Race (London, 1887),
154n2; W.E. B. Du Bois, The Negro (London, 1915), 30-34.
17Some other languages of sub-Saharan Africa, including Fulani and Maasai, were some-
times included incorrectly in this group. The term “Hamitic” for these languages is now
obsolete: see Joseph H. Greenberg, The Languages of Africa (2nd ed.: Bloomington,
1966), 41-65.
298 Robin Law

Canaanites, unambiguously “Hamites” in the Old Testament, but who spoke

a language closely related to Hebrew, had now to be reclassified as “Semi-
tes.”
For sub-Saharan Africa, given the persistence of models of migration
from the northeast this implied that any cultural achievement which Euro-
peans judged meritorious in this region could likewise now be attributed to
“white” initiative or influence. Saunders suggested that a “seminal” role in
the formulation of this interpretation of sub-Saharan African history was
played by the explorer John Hanning Speke, who in the 1860s attributed the
foundation of the kingdoms of interlacustrine East Africa to invasion and
conquest by pastoralists from the north, who were identified as “Shem-
Hamitic” in race, i.e., a combination of Semitic Ethiopians and “Hamitic”
Galla.18 In relation to West Africa, this was paralleled, and indeed preceded,
by European speculations about the supposedly non-Negro origins of the
Fulani, which were likewise cited to explain their role as conquerors and
state-builders, although these were not generally cast within a specifically
“Hamitic” framework.19
Since this denial of historical achievement to Black Africans could also
be interpreted to suggest a lack of racial potential for cultural progress in the
future, as long as Africans were left in charge of their own affairs, it could
function subsequently to legitimate European colonial rule, which was pre-
sented as the only means of bringing about the development of Africa. This
argument is implicit, for example, in a work of the British imperialist pro-
consul Sir Harry Johnston, published in 1899, in which the catalog of for-
eign invaders/immigrants in sub-Saharan Africa, beginning with the Phoeni-
cians and culminating in the contemporary European partition of Africa, is
prefaced by supposed prehistoric immigrations of “Hamitic civilisers,”
mainly from the Nile valley.20

III

As an explanation of the nineteenth and twentieth-century European version

of the “Hamitic hypothesis” the foregoing account, which is partly summa-

18Saunders, “Hamitic Hypothesis,” 528; J.H. Speke, Journal of the Discovery of the
Source of the Nile (London, 1863), chapter 9.
19See Philip D. Curtin, The Image of Africa: British Ideas and Action, 1780-1850 (2
vols.: Madison, 1964), 2:411-12. The Fulani were variously supposed to descend from
Persians or Carthaginians, or even to have immigrated from Malaya. Note, however, the
alternative suggestion that the Fulani were descendants of Put (or Phut), son of Ham, on
the basis of the Fulani toponym “Futa,” apparently first in Gustave d’Eichthal, “Histoire
de l’origine des Foulahs ou Fellans,” Mémoires de la Société d’Ethnologie (Paris)
1/2(1841), 1-294.
20Sir Harry Johnston, A History of the Colonization of Africa by Alien Races (London,
1899; 2nd ed., 1913), chapter 1.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 299

rized and partly elaborated from Saunders, seems to me defective; but it is

not so much incorrect as incomplete. Although the idea of a civilizing input
from outside Africa was certainly often formulated in explicitly racialist
terms, and sometimes used for the specific political purpose of justifying
European domination, this was neither its only source nor its necessary
implication; rather, the dominance of the “Hamitic hypothesis” in European
thought about African history has to seen in a wider intellectual context.
First, account must be taken of the persistence into the nineteenth centu-
ry of the influence of the original Biblical version of the “Hamitic hypothe-
sis”—that is, the model of the peopling of the world by migration from the
Middle East through the dispersal of Noah’s sons after the Flood. (Apart
from the dispersal of Noah’s descendants, the alternative Biblical model for
the origins of exotic peoples was identification with the ten “lost tribes” of
Israel, who were deported and dispersed after the Assyrian conquest in the
eighth century BCE.)21 Many of the earliest European commentators on
African history were Christian missionaries, who were often scriptural fun-
damentalists, whose conception of history was predicated on a literal under-
standing of the Bible; and from such a perspective, the cultures of sub-Saha-
ran Africa had necessarily to be derived from the immigration of “Hamites,”
irrespective of whether the new doctrine of racialism was accepted. More-
over, given the anti-Biblical character of nineteenth-century “scientific”
racism, the more fundamentalist Christians were not, in fact, likely to sub-
scribe to it, being committed by their Biblical literalism to the older doctrine
of the oneness of humanity on the basis of common (and recent) descent
from a single pair of ancestors, Adam and Eve.
Moreover, the “Hamitic hypothesis” as applied to Africa has to be seen
as merely one local variant of a much more general view of the historical
process, “diffusionism,” according to which autonomous cultural innovation
was a rare occurrence, and cultural change therefore normally came about
through the diffusion of innovations from their place of origin. In such
views, the original cradle of many critical cultural innovations (and indeed
of “civilization” itself) was frequently (although not invariably) located
specifically in Egypt. The Egyptocentric version of diffusionism was pro-
pounded in an extreme form in the early twentieth century especially by Sir
Grafton Elliot Smith.22 Diffusionist interpretations were commonly applied,
not to Africa alone, but to the whole world, including the Americas, and
indeed to Europe itself; absolutely everywhere else was supposed to derive
its civilization from Egypt.

21See Tudor Parfitt, The Lost Tribes of Israel: The History of a Myth (London, 2002),
though this seems weak on West Africa.
22E.g. Smith, Migrations of Early Culture (London, 1915); idem, The Diffusion of Cul-
ture (London, 1935).
300 Robin Law

Given this wider background, it may be suggested that it is an oversim-

plification to see diffusionist interpretations of African history propounded
by Europeans during the colonial period as necessarily expressing or reflect-
ing ideas of racial superiority or colonial apologetics. A clear counter-exam-
ple is the German ethnographer Leo Frobenius, in his book Und Afrika
Sprach, published in 1912.23 Frobenius attributed the origins of political
institutions in the West African interior generally (described by him, essen-
tially, on the basis of Nupe, in northern Nigeria) to migration from Nubia in
the seventh century CE. For the particular case of Yorubaland, however, he
proposed a different origin. Frobenius was the principal discoverer of the
brass and terracotta sculptures of Ife, and he explained the existence of this
art, and more especially its stylistic naturalism, as deriving from a supposed
Etruscan maritime colonization of West Africa, attributed to the second mil-
lennium BCE, also speculating that Yorubaland might in fact have been the
legendary land of Atlantis. Although regularly dismissed nowadays as
reflecting Frobenius’ racial prejudices, this view should more correctly be
attributed to the general diffusionist mode of thought of his period. Frobe-
nius himself, it is clear, saw the historiographical significance of his discov-
eries as proving, not that sub-Saharan Africa had no record of historical
achievement, but on the contrary that it had a recoverable history which
extended (in contradiction to the then dominant assumption) back into the
pre-Islamic era—a vindication rather than denial of Africa’s place in histo-
ry.24
Even more critically, the interpretation of the “Hamitic hypothesis” as a
projection of European racialism fails to take adequate account of the fact
that essentially similar conceptions of history have often been espoused by
black Africans themselves, both in precolonial times and into the twentieth
century. It does not seem plausible, moreover, to explain this popularity of
the “Hamitic hypothesis” among Africans as merely a reflection of their cul-
tural subordination and deference to European intellectual authority, for two
reasons: first, because the African espousal of versions of the “Hamitic
hypothesis” in part predated the establishment of European domination of
Africa, but also because versions of it were often used during the colonial
era to express African cultural nationalism—that is, to contest rather than to
endorse the idea of African racial inferiority.

23Leo Frobenius, The Voice of Africa, trans. R. Blind (2 vols,: London, 1913).
24As explicitly asserted in the conclusion of his book: ibid., 2:680. I am fortified in this
reading of Frobenius by the fact that this is clearly how the significance of his work was
seen by the pioneer Afro-American historian, W.E.B. Du Bois: see Robin Law, “Du Bois
as a Pioneer of Africa History,” in Mary Keller and Chester J. Fontenot, Jr., eds., Re-Cog-
nizing Du Bois in the Twenty-First Century: Essays on W.E.B. Du Bois (Macon GA,
2007), 45-64.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 301

IV

Versions of the “Hamitic hypothesis” occur in early local West African

Islamic historiography, written in Arabic, at a period earlier than the rise of
the classic European racialist version. This should be entirely unsurprising,
since West African Muslims took over the general Islamic (and ultimately
Old Testament) historical framework of the peopling of the world by the
dispersal of the descendants of Noah after the Flood. The dominant tenden-
cy was to affiliate West African peoples to Ham through his sons Cush or
Canaan, although links were often also posited with other figures from
Islamic/Arabic historical tradition. As a typical example one can cite the
historical work of the noted scholar and political leader Muhammad Bello
of Sokoto in Hausaland. His Infaq al-Maisur, written in 1812, attributes
Middle Eastern origins—from Egypt, Palestine, the Yemen, and even Per-
sia—to a number of West African peoples, though for the most part without
reference specifically to Ham or Noah. In the case of the Yoruba, however,
Bello declares them to be descendants of Canaan, and more particularly of
his son Nimrud, who is traditionally linked with Babylonia/Iraq.25 He com-
bines this conventional piece of speculative genealogy with an equally spec-
ulative venture in historical etymology, deriving the name “Yoruba” from
that of Yar‘ub ibn Qahtan, the legendary ancestor of the Himyarite kings of
Yemen in pre-Islamic times, hypothesizing that the ancestral Yoruba had
been expelled from their original homeland in Iraq by Ya‘rub.26
Equally, in West African oral traditions (in some cases, recorded in early
Arabic writings), the origins of local peoples (or at least of the ruling dynas-
ties of local states) are frequently attributed to the arrival of
immigrants/conquerors from the Middle East, most often specifically from
Mecca; but sometimes also from the Yemen, or from Iraq (sometimes,
specifically, from Baghdad).27 As early as the twelfth century, the kings of
Ghana were claiming descent from a man called Salih, who was represented
to be a great-great-grandson of the Caliph ‘Ali, the son-in-law and fourth
successor of the Prophet Muhammad, who ruled in Iraq, although this claim
was dismissed by mainstream Islamic scholarship.28 Later, in Songhay, the

25In the Islamic tradition, Nimrud is considered a son of Canaan; rather than, as in the
Old Testament, of Cush.
26E.J. Arnett, The Rise of the Sokoto Fulani, Being a Paraphrase and in Some Part a
Translation of the Infaku’l Maisuri of the Sultan Mohammed Bello (Kano, 1922), 16.
27Reference to Baghdad probably reflects its eminence as the seat of the Caliph, the
(nominal) supreme ruler of all (Orthodox) Muslims, between 762 and 1258.
28al-Idrisi, in N. Levtzion and J.F.P. Hopkins, Corpus of Early Arabic Sources for West
African History (Cambridge, 1981), 109, referring to Salih b. ‘Abd-Allah b. al-Hasan b.
al-Hasan b. al-Hasan b. ‘Ali b. Abi Talib. Ibn Khaldun later commented that “this Salih is
not known among the descendants of ‘Abd Allah b. al-Hasan:” ibid., 320.
302 Robin Law

founder of the first royal dynasty of the kingdom, the Zuwas (who suppos-
edly ruled until the thirteenth century), is named as Zuwa-Alayaman, which
name was explained (already in local Arabic scholarship in the seventeenth
century) as meaning “he came from the Yemen.”29 Likewise among the
Hausa, the mythical dynastic ancestor Bayajidda is usually presented as
coming from Mecca, although sometimes from Baghdad.30
Although sometimes (as in these cases just cited) the individuals claimed
as founders, although said to have come from the Islamic/Arabic world,
appear to be local West African creations,31 in other instances, prominent
figures from Islamic historical tradition were appropriated by West Africans
as ancestors. In the case of the kingdom of Mali, for example, the royal
dynasty claimed descent from Bilali, and the caste of griots from Surakata.
As has been shown by David Conrad, the first of these represents Bilal ibn
Rabah and the second Suraqa ibn Malik, two of the early companions of
Muhammad—the former being his muezzin and the latter an initial enemy
of Muhammad, who had sought to kill him before being converted to
Islam.32 The Fulani similarly traced their descent to one “Ukuba,” who is
identifiable with ‘Uqba ibn Nafi‘, one of the commanders of the Arab con-
quest of North Africa in the later seventh century—a claim recorded already
in the early nineteenth century.33
In other cases, the claimed ancestors are figures from pre-Islamic Arabic
tradition. The best-known example is Borno (in northeastern Nigeria),
whose royal dynasty claimed descent (at least as early as the thirteenth cen-
tury) from Sayf ibn dhi Yhazan, last scion of the Himyarite royal dynasty,
and hero of Yemeni resistance to invasion from Christian Ethiopia in the
sixth century (the “War of the Elephant” alluded to the Qu’ran).34 Sayf,
although living before the time of Muhammad, can be thought of as a proto-
Islamic hero, as a defender of Mecca against Christian imperialism. In other
cases, however, West African origins were traced to more ambiguous, or
even explicitly anti-Islamic figures. In Borgu (which, in terms of modern

29John Hunwick, ed., Timbuktu and the Songhay Empire: Al-Sa‘di’s Tarikh al-sudan
down to 1613 and Other Contemporary Documents (Leiden, 2003), 5-6.
30W.K.R. Hallam, “The Bayajidda Legend in Hausa Folklore,” JAH 7(1966), 47-60.
31Although it has also been suggested that “Bayajjida” should be identified with Abu
Yazid, leader of an Islamic sectarian uprising in Tunisia in the late eighth century CE:
H..R. Palmer, Bornu Sahara and Sudan (London, 1936), 273-74.
32David C. Conrad, “Islam in the Oral Traditions of Mali: Bilali and Surakata,” JAH
26(1985), 33-49.
33David Robinson, The Holy War of Umar Tal: the Western Sudan in the Mid-Nineteenth
Century (Oxford, 1985), 82-84. The claim is already found in Muhammad Bello’s Infaq
al-Maisur.
34First recorded by Ibn Said, in Levtzion/Hopkins, Corpus of Early Arabic Sources, 188.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 303

political geography, straddled northern Bénin and adjacent areas of Nigeria

to the east), the royal dynasties of the various kingdoms trace their descent
from Kisra, who represents Khusraw (or, in a Greek form, Chosroes), a Per-
sian king of the early seventh century, a contemporary of Muhammad, who
reputedly rejected an appeal by the latter to convert to Islam.35 In the West
African traditions, Kisra is usually presented as having left Mecca (sic) after
refusing to follow Muhammad.36
In the case of Yorubaland, the origins of the principal royal dynasties are
conventionally traced to a common ancestor called Oduduwa, who in the
usual (and presumably original) version of the story descended by a chain
from heaven, when the earth was covered with water, to create dry land in
the middle of it, at Ife. But in the version told in the northern Yoruba king-
dom of Oyo (and recorded in Samuel Johnson’s History in the 1890s)
Oduduwa is presented as the son of one “Lamurudu,” a king of Mecca who
renounced Islam, clashed with a pious Muslim called “Braima,” and was
killed in a Muslim uprising.37 “Lamurudu” and “Braima” in this story evi-
dently represent Nimrud and Ibrahim (i.e., Abraham), and the story is based
on an Islamic saga of the confrontation between these two figures. Muham-
mad Bello, as cited earlier, in tracing Yoruba origins to Nimrud, may have
been drawing on this tradition, although alternatively, the tradition may be a
more recent development, itself derived from Bello’s written account.38
In the early years of academic study of African history in the 1950s and
1960s, traditions of this sort claiming origin from the Middle East were
often interpreted essentially literally, as recording real historical
migrations.39 Alternatively, although the literal truth of the supposed migra-
tions was discounted, the existence within West African oral traditions of
what were recognized as elements derived ultimately from Biblical or other
Middle Eastern sources was seen as evidence of ancient cultural links with
the Middle East. John Fage in the 1960s thus spoke of the transmission of
“ideas and influences,” not directly from the Middle East but through the
intermediate agency of Saharan pastoralists (here identified as “Hamites”);

35Richard Kuba, Wasangari und Wangara: Borgu und sein Nachbaren in historischer
Perspektive (Hamburg, 1996), 98-148.
36Another version makes Kisra clash with “Anabinuhu,” i.e., the Prophet Noah: Frobe-
nius, Voice of Africa, 2:617.
37Johnson, History, 3-4.
38For discussion, see Robin Law, “How Truly Traditional is Our Traditional History? The
Case of Samuel Johnson and the Recording of Yoruba Oral Tradition,” HA 11(1984),
195-221, esp. 202-05.
39E.g. Hallam, “Bayajidda Legend,” 49-50, accepting the identification of “Bayajidda”
with the historical Abu Yazid (see note 31 above), supposed that remnants of his army
might have fled south into West Africa, following his defeat and death in Tunisia.
304 Robin Law

while Dierk Lange more recently suggested that the crucial conduit was a
pre-Islamic trans-Saharan trade, supposedly conducted by Phoenician
colonists in North Africa.40
The more general view nowadays, however (at least among academic
historians), is that these traditions of migration from the Middle East are to
be explained by the expansion of the influence of Islam in sub-Saharan
Africas in relatively recent times. The concern of African peoples to claim
origins from the Islamic world reflects no more than a desire to relate them-
selves to what was seen as a prestigious world civilization—comparable, for
instance, to the concern of the ancient Romans to relate themselves to
ancient Greek historical tradition by representing themselves as descendants
of the defeated Trojans; and indeed, the concern of many European societies
in medieval times to link themselves to Classical Antiquity—again, very
often through stories of the dispersal of defeated Trojans. Britain, for exam-
ple, was commonly said to have been colonized by, and to have derived its
name from, a Trojan refugee called Brutus.41
In cases where descent was claimed from specific figures in Islamic tra-
dition, it is sometimes possible to suggest reasons for the choice of these
particular persons as ancestors. In the case of Mali, for example, the claim-
ing of Bilal as dynastic ancestor evidently reflected the fact that the original
Bilal was, according to tradition, in origin a black African slave, and thus
the first black African Muslim. In that of Suraqa/Surakata, as Conrad has
argued, his appeal as ancestor derived from his status as a converted oppo-
nent of Islam, which had a particular resonance for the griots, as the intel-
lectual establishment of the pre-Islamic order.42 Likewise for the Fulani, the
claim of descent from ‘Uqba, as one of the pioneers of the spread of Islam
into Africa, functioned to underwrite their claim to pre-eminence in the dis-
semination and consolidation of Islam from the eighteenth century
onwards.43 On the other hand, anti-Islamic figures such as Kisra and Nim-
rud were appropriate choices of ancestors for peoples who resisted conver-
sion to Islam.
It has been argued that some cases of the claiming of Middle Eastern
ancestry cannot plausibly be put down to Islamic influence, since such sto-
ries are not limited to Islamic societies in West Africa, but occur also

40Fage, History of West Africa, 9-10; Lange, Ancient Kingdoms, esp. 277-85.
41First recorded in the ninth century: John Morris, ed., Nennius: British History and
Welsh Annals (London, 1980), 18-20. Note that this work also supplies a genealogy trac-
ing the ancestry of Brutus to Javan, son of Japhet, son of Noah (and beyond to Adam and
Eve).
42Conrad, “Islam in the Oral Traditions of Mali.”
43For the context see Robinson, Holy War of Umar Tal, 81-89.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 305

among “essentially pagan peoples,” as for example the Kisra story in

Borgu—and, one might add, the Nimrud story among the Yoruba.44 This
seems a doubtful inference, however, since Borgu and Yorubaland, while
still indeed mainly “pagan” down to the nineteenth century, had certainly
been involved in Islamic trading networks, and had resident communities of
foreign Muslims since at least the seventeenth century.45 As has been
argued by Philips Stevens, the appeal of figures like Kisra and Nimrud for
such societies was precisely their status as heroes of anti-Islamic resistance.
They represented ancestors who were figures of stature in Islamic tradition,
and thus conferred an antiquity and status comparable to that of Islam, while
simultaneously legitimating their supposed descendants’ resistance to con-
version to Islam.46 A similar argument, it may be suggested, applies to the
Roman claim to descend from the Trojans, which gave Rome a place within
the Greek scheme of world history, but which also emphasized their dis-
tinctness from the Greeks themselves. A likely historical context for the
adoption of Kisra and Nimrud as legitimating ancestors is the early nine-
teenth century, when both Borgu and Oyo were resisting the militant Islam-
ic jihad led by Usuman dan Fodio from Hausaland.47
The existence of versions of the “Hamitic hypothesis” in early West
African Islamic writings and oral traditions deserves emphasis, not only
because these were evidently independent of (since earlier than) the classic
European versions of this model, but also because they clearly exercised
some influence on European thinking about African history. It may be sug-
gested that stress on the racialist/imperialist purposes served by the
“Hamitic hypothesis” tends to obscure the role in its evolution and consoli-
dation that was played by the nature of the available evidence. It seems
clear, at least, that one of the major reasons why Europeans attributed Mid-
dle Eastern origins to West African peoples/cultures was, straightforwardly,
because this is what they were told by their African informants. The classifi-
cation of the Fulani as racially “white,” for example, was not simply a pro-
jection of European prejudices, but was a claim which they themselves

44Fage, History of West Africa, 9.

45For Islam in Borgu see Nehemiah Levztion, Muslims and Chiefs in West Africa
(Oxford, 1968), 173-78; for Oyo see Robin Law, The Oyo Empire c.1600-c.1836
(Oxford, 1977), 75-76, 255-60.
46Philips Stevens, Jr, “The Kisra Legend and the Distortion of Historical Tradition,” JAH
16(1975), 185-200. It should be noted that the defense of the historicity of the Kisra tradi-
tions by Marjorie Helen Stewart, “The Kisra legend as Oral History,” IJAHS 13(1980),
51-70, although cast in part as a critique of Stevens, does not relate to the alleged Persian
connection.
47For the Oyo case see further Law, “How Truly Traditional?,” 204-05.
306 Robin Law

made.48 Likewise, in positing exotic origins for pre-Islamic West African

kingship, Frobenius drew upon the local legends of Kisra.49 Intellectual
influences did not flow in one direction only.

Even more critically for present purposes, versions of the “Hamitic hypoth-
esis” also occur regularly in the writings of western-educated Christian
West Africans, writing in European languages. An early example is the
Afro-British ex-slave (and Abolitionist activist) Olaudah Equiano, who in
his autobiography (published in 1789) discussed the origins of his own peo-
ple, the Igbo of southeastern Nigeria, and suggested that they were derived
from the ancient Jews—the argument being based mainly on similarities of
customs, notably circumcision. The difficulty that the Jews were white and
the Igbo black was explained on the supposition that blackness was an
acquired rather than an inherited characteristic, attributed to the West
African climate.50 The idea was subsequently elaborated by James
Africanus Horton, born in Sierra Leone but also of Igbo ancestry, in his
book published in 1868, which suggested more specifically that the Igbo
were remnants of the “lost tribes” of Israel.51
Equiano and Horton were not strictly historians, but offered these histori-
cal speculations casually while writing on other subjects. A more self-con-
scious and systematic interest in local history developed, however, from the
late nineteenth century onwards, among communities of Anglophone
Africans in Sierra Leone, the Gold Coast, and Yorubaland. This rise of the
study of local history was part of a wider movement of “cultural national-
ism,” which sought to overcome the alienation of western-educated Christ-
ian Africans from their indigenous African societies.52 Several of these local

48As noted already in the 1790s: “[t]hey evidently consider all the Negro natives as their
inferiors; and when talking of different nations, always rank themselves among the white
people:” Mungo Park, Travels in the Interior Districts of Africa (London, 1799), 46. This
statement has to be understood in its local (Senegambian) context, in which “white” (Ara-
bic bidan) denoted (patrilineal) Arab descent and free status, rather than skin color.
49Frobenius, Voice of Africa, 2:617-20.
50Olaudah Equiano, The Life of Olaudah Equiano, or Gustavus Vassa, the African, writ-
ten by himself (2 vols, London, 1789), 1:38-41.
51James Africanus Horton, West African Countries and Peoples (London, 1868), 167-71.
This example is missed by Parfitt, Lost Tribes.
52See Robin Law, “Local Amateur Scholarship in the Construction of Yoruba Ethnicity,
1880-1914” in Louise de la Gorgendière, Kenneth King, and Sarah Vaughan, eds., Eth-
nicity in Africa: Roots, Meanings and Interpretations (Edinburgh, 1996), 55-90, esp. 56-
63.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 307

historians put forward arguments for the origins of West African peoples
from the Middle East, often citing local oral traditions in support. In the
case of the Yoruba, for example, Samuel Johnson, writing in 1897, claimed
that the ancestral Yoruba had migrated to West Africa from “Upper Egypt,”
meaning Nubia. In support of his argument he cited not only the local tradi-
tion, mentioned above, of descent from “Lamurudu” of Mecca—arguing
that the reference to Mecca should not be understood literally, but symboli-
cally, as representing “the East” more generally. He also adduced supposed
parallels between Yoruba and Middle Eastern customs, and even (anticipat-
ing Frobenius) archeological evidence in the form of the stone sculptures of
Ife, which were held to be similar to those of ancient Egypt.53
This claim for the Yoruba of Egyptian origin was reiterated and elaborat-
ed in several later works, notably by Archdeacon J. O. Lucas in his study of
Yoruba religion published in 1948.54 Lucas argued for the ancient Egyptian
origin of Yoruba religion, and supported his argument not only by citing
supposedly parallel customs (the Yoruba practice of wrapping corpses in
cloth for burial, for example, being compared to Egyptian mummification),
but also with linguistic evidence, suggesting Egyptian etymologies for the
names of Yoruba gods, and indeed for the Yoruba language more general-
ly—he claimed that over half of Yoruba vocabulary could be traced to
Egyptian origins. Alternative versions of Middle Eastern origins for the
Yoruba were also canvassed. For example, the Sierra Leonian historian
A.B.C. Sibthorpe, in a book published in 1909, claimed that the Yoruba
were the “lost tribes” of Israel.55
This way of looking at Yoruba origins, it may be noted, survived into the
earliest stages of the development of academic historical writing in the
1950s. The pioneer Yoruba academic historian, S.O. Biobaku, in a series of
radio lectures on Yoruba origins in 1955, synthesizing earlier speculations,
argued for ancient Egyptian, Jewish, and Etruscan influences on Yoruba
civilization, and explained these on the assumption that the ancestral Yoru-
ba had migrated from somewhere in the Middle East, probably the Nile Val-
ley, between the seventh and ninth centuries CE.56 But Biobaku avoided the
implication of “white” initiative by positing that the medium through which

53Johnson, History, 3-7.

54Archdeacon J. Olumide Lucas, The Religion of the Yorubas (Lagos, 1948).
55A.B.C. Sibthorpe, Bible Review of Reviews: the Discovery of the Ten Lost Tribes,
Yorubas or Akus (Cline Town, Sierra Leone, 1909), for which, see Christopher Fyfe,
“A.B.C. Sibthorpe; a Tribute,” HA 19(1992), 327-52. This example is also missed by
Parfitt, although he does allude to other theories of exotic origins of the Yoruba (referring
to Frobenius, Samuel Johnson, and Biobaku): Lost Tribes, 199, 203.
56S.O. Biobaku, The Lugard Lectures–1955 (Lagos, n.d.); reprinted as The Origin of the
Yorubas (Lagos, 1960).
308 Robin Law

Middle Eastern elements were transmitted to the Yoruba was the kingdom
of Meroë in Nubia, which was emphasized to be itself an “all-Black king-
dom.”
The idea of the “Hamitic hypothesis” as a rationalization of European
claims to racial superiority cannot, obviously, be very easily applied to
cases such as Samuel Johnson, who was writing from an explicitly “nation-
alist” (or in his own terminology, “patriotic”) perspective, seeking to assert
the value of African culture and its historical achievement.57 For Johnson,
the claim to Middle Eastern origins was a way of attaching Yoruba civiliza-
tion to what he saw as the mainstream of history, to claim prestigious ori-
gins for it rather than to dismiss it as marginal or derivative.
In Johnson’s case, the thesis of Middle Eastern origins has an additional
twist, which added to its “nationalist” appeal. He argued, as has been seen,
that the ancestral Yoruba came not from Arabia (as local tradition claimed),
but from Nubia, and on this basis inferred that their original religion had
perhaps been, not paganism, but Christianity, in its specifically Egyptian
(Coptic) form.58 (Lucas, by contrast, sought to derive Yoruba religion from
ancient Egyptian paganism.) Johnson supported his argument by suggesting
that versions of Biblical stories were preserved in Yoruba oral tradition—
interpreting the Yoruba story that the earth was covered with water prior to
the descent of Oduduwa from heaven, for example, as a derived variant of
the Biblical story of the Flood.59 For Johnson, therefore, Yoruba religion
was a degenerate or corrupt form of Christianity. The appeal of such an
interpretation for an intellectual in Johnson’s position is evident. There was
on the face of it a fundamental contradiction in the project of “cultural
nationalism,” in that the value of African culture and history was asserted
by people who themselves remained Christians, and committed to the con-
version of their fellow-Yoruba to Christianity—that is, effectively to the
subversion of indigenous historical and cultural traditions. The representa-
tion of the ancestral Yoruba as Christians resolved this contradiction. Chris-
tianity was after all not alien, but traditional, and conversion to it would rep-
resent simply a restoration of the original and pure form of the indigenous
religion. Hence, it was after all possible to be simultaneously a true patriot,
a champion of Yoruba culture, and a Christian.60

57For Johnson’s “purely patriotic motive,” see History, vii.

58Ibid., 7.
59Ibid., 9; for other claims that Yoruba traditions contained “garbled forms of scriptural
stories,” explained as “showing that the ancestors of the Yorubas were acquainted with
Christianity in their land of origin,” cf. ibid., 148, 154.
60Robin Law, “Constructing ‘a Real National History:’ a Comparison of Edward Blyden
and Samuel Johnson” in P.F. de Moraes Farias and Karin Barber, eds., Self-Assertion and
Brokerage: Early Cultural Nationalism in West Africa (Birmingham, 1990), 78-100, esp.
96.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 309

VI

Given the overt nationalist purpose of much of local Yoruba historical writ-
ing, it does not seem helpful to dismiss the espousal of versions of the
“Hamitic hypothesis” by Johnson and others as merely a case of intellectual
colonization “imposed by the West” and “uncritically accepted” by Euro-
pean-educated Africans, as A.E. Afigbo has suggested in the case of the
Igbo.61 A more nuanced analysis is offered by Philip Zachernuk in a study
of Nigerian historical writing published in 1994. Zachernuk rejects the sug-
gestion of “uncritical acceptance” of European ideas by local intellectuals,
in favor of that of “critical dialogue” with them. He acknowledges that
African historians selectively appropriated and adapted, rather than simply
repeated, the “Hamitic hypothesis,” and used it as a means to vindicate
rather than denigrate African historical achievement. However, he maintains
that although West Africans had their own agenda, they were nevertheless
“constrained” in their thinking by the colonial situation; even though they
sought to subvert the racialist and imperialist intentions of European litera-
ture, they “had to work within European concepts of history.” In a felicitous
formulation, he says that they “appropriated the authority but undermined
the intent of imperial literature.” Zachernuk concludes by invoking the idea
that under colonialism West African thought was “de-centred,” in the sense
that it had to work within a framework of ideas shaped by European, rather
than African, needs.62
This analysis, however, although a substantial advance on earlier treat-
ments, is not wholly persuasive. It is, first, arguable that it by implication
exaggerates the monolithic character of “European” thought. Although
Zachernuk recognizes successive transformations and competing construc-
tions of the “Hamitic hypothesis” by African intellectuals, he gives less
attention to variety within European thinking, which in effect gave Africans
a choice of which European models to follow.63 Even more critically, I
would suggest, this perspective exaggerates the degree of “dialogue” with
European thought in which African writers in fact engaged. There were cer-
tainly particular African (or African-American) writers who pursued the sort
of “critical dialogue” with European thinking which Zachernuk describes,

61A.E.Afigbo, “Traditions of Igbo Origins: a Comment,” HA 10(1983), 1-11.

62PhilipS. Zachernuk, “Of Origins and Colonial Order: Southern Nigerian Historians and
the ‘Hamitic Hypothesis’, c.1870-1970,” JAH 35(1994), 427-55, quotations from 430,
454-55. For an application of the argument in a wider context, cf. Philip S. Zachernuk,
Colonial Subjects: an African Intelligentsia and Atlantic Ideas (Charlottesville, 2000).
63Zachernuk himself does make this point in general terms in Colonial Subjects, 6-7, but
he seems to lose sight of it in relation to the “Hamitic hypothesis.”
310 Robin Law

and, indeed, who in the process found themselves to some degree entrapped
within the categories of the very European thought which they were seeking
to subvert. A clear example is E.W. Blyden, who in seeking to contest Euro-
pean imputations of African racial inferiority, nevertheless internalized
many of the assumptions of European racialist thought.64 But the tradition
of specifically Yoruba historiography appears much more locally focused
and autonomous than Zachernuk’s analysis allows.
Samuel Johnson, for example, does not show much sign even of being
aware of, far less of being influenced by or seeking to contest, European
racist models of African history.65 The only point at which his History
explicitly engages in “critical dialogue” with imperial literature is, not in its
examination of Yoruba origins, but in its account of British attempts to
negotiate peace among warring Yoruba states in the 1880s, in which John-
son himself had served as interpreter to the British mediators, where extend-
ed citation of British official records is combined with critical commentary
on the conduct of British policy.66 In his discussion of Yoruba origins, the
only “European” source cited is an English translation of Muhammad
Bello’s Arabic account, cited earlier: an African Muslim authority, that is,
even though accessed through a European intermediary.67 The most obvious
influence on Johnson’s historical thinking was not the “imperial” version of
the “Hamitic hypothesis” (which, indeed, at the time of his writing in 1897,
hardly yet existed in published form, in relation to the Yoruba),68 but rather

64See Law, “Constructing,” 84-88, 98.

65For a contrary view see Philip S. Zachernuk, “Johnson and the Victorian Image of the
Yoruba,” in Falola, Pioneer, Patriot and Patriarch, 33-46. In particular, Zachernuk
argues that Johnson was responding to the work of A.B. Ellis, The Yoruba-Speaking Peo-
ples of the Slave Coast of West Africa (London, 1894). This connection is confessedly
“inferred,” since Johnson nowhere explicitly cites Ellis. But in any case, the argument of
Ellis that Johnson supposedly sought to refute was not the “Hamitic hypothesis,” but
rather his “evolutionist” interpretation of Yoruba religion.
66Johnson, History, 538-60. Most, and perhaps all, of these critical comments seem to be
the work of Johnson’s posthumous editor, his brother Dr Obadiah Johnson.
67Ibid., 5-6, citing Dixon Denman and Hugh Clapperton, Narrative of Travels and Dis-
coveries in Northern and Central Africa (London, 1826), Appendix XII, 165.
68Zachernuk, “Johnson,” 40-41, argues that rudimentary versions of the “Hamitic” theory
of Yoruba origins can already be found in T.J. Bowen, Adventures and Missionary
Labours in Several Countries in the Interior of Africa from 1849 to 1856 (Charleston,
1857), and Richard F. Burton, Abeokuta and the Camaroons Mountains (2 vols.: London,
1863). The former does indeed posit “large immigrations of white people into Africa,”
producing a partly “mulatto” population in much of West Africa, including Yorubaland
(pp. 267-69, 276-80), but although he stresses the “somewhat civilized” character of the
Yoruba, he does not relate this explicitly to the supposed admixture of “white” blood;
while the latter work likewise hypothesizes a “stream of immigration from the lands near-
er Arabia,” but explicitly regards the Yoruba as “a race of pagans” displaced by this
immigration, rather than deriving from it (1:231). In any case, here again, Johnson does
not cite (or show any evidence of acquaintance with) either of these works. Another
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 311

the older Biblical version of diffusionism, which it is misleading to regard

as specifically European, since, as has been seen, it was already naturalized
in West Africa through the influence of Islam, long before any European
missionaries or administrators arrived.
Archdeacon Lucas in the 1940s was much more engaged with wider cur-
rents of thought, mainly because his book was based on a doctoral thesis (in
Divinity, not History) written for a British university (Durham, 1942). But
his main orientation was not in fact towards the “Hamitic hypothesis” as
such, but rather towards the debate within the anthropology of religion over
whether apparent convergences between different religious traditions were
best explained through common origin or parallel evolution—in the specific
terms in which Lucas posed the question, was Yoruba religion a case of
“degeneration” (from a “higher” religious tradition from which it was
derived) or “arrested development” (at a “lower” evolutionary stage)?69 It so
happens that the first of these options, which is the one which Lucas himself
espoused, was diffusionist. But it does not seem all that helpful to “explain”
his adoption of this position as reflecting deference to European authority,
when there was also a non-diffusionist alternative on offer which was equal-
ly “European.” Moreover, Lucas, in adopting a “Hamitic” version of Yoru-
ba origins (overland from the Middle East), was explicitly rejecting the
principal European diffusionist interpretation of the origins of Yoruba civi-
lization then on offer, which was Frobenius’ “Atlantic theory,” in favor of
something closer to that propounded by the African Samuel Johnson (and
beyond him, indeed, by the Muslim African scholar Muhammad Bello).70
Although Yoruba scholars after Johnson certainly regularly cited Euro-
pean “authorities” (alongside indigenous traditions, and indeed earlier
Yoruba historians, including above all Johnson himself), this was more a
matter of plundering such European sources for usable detailed “facts” than
of taking over their intellectual frameworks. Ironically, indeed, with regard
to discussion of Yoruba origins, one of the major areas of such borrowing
was in assimilating the oral traditions of other African groups, as repro-
duced in European colonial literature. By the 1960s local historians in
Yorubaland had incorporated versions of the Kisra legend into their
accounts of Yoruba origins.71 They knew of these not directly from the tra-

writer of the 1850s identified the Yoruba as descendants of Canaanites dispersed from
Palestine by the Israeli conquest (in the thirteenth century BCE), and more explicitly
linked this to their relative “civilization.” But this work was still unpublished when John-
son wrote: W.H. Clarge, Travels and Explorations in Yorubaland, 1854-1858, ed. J.A.
Atanda (Ibadan, 1972), 287-92.
69Lucas, Religion, 3-4, 344-45.
70For his critique of Frobenius, see ibid., 347-52.
71Chief S. Ojo, Bada of Saki, Iwe Itan Oyo, Ikoyi ati AFIJIO (Oyo, n.d. [ca.1961]), 16-
17.
312 Robin Law

ditions of neighboring Borgu, but at second hand through European

sources.72
It may be suggested, in fact, that the dialogue in which local Yoruba his-
torians engaged, at least prior to the emergence of academic history in the
1950s, was more with each other than with the European colonial master.
Johnson’s work, for example, was explicitly addressed to fellow Yoruba
who were ignorant of the history of their own nation, rather than to Euro-
peans who denied or denigrated it.73 It should also be stressed that, although
Johnson himself wrote in English, a significant minority of other early local
historians wrote in Yoruba: evidently, these were writing for a local rather
than a European readership. Indeed, this internal orientation remained to
some degree true even in the early period of academic history, with Biobaku
in the 1950s. Although Biobaku was certainly concerned to relate his work
to, and legitimate it within, a European tradition of academic historiogra-
phy, this related more to its methodology than to its content. He gave atten-
tion to Yoruba origins, in a series of lectures explicitly addressed to a wider
non-academic audience, not in order to contest colonial European versions,
but because the question of origins had been a central issue of local interest
since the nineteenth century. If one wants to characterize the development
of Yoruba historical thinking in terms of “critical dialogue,” there is a much
stronger case for saying that this dialogue was with Johnson’s History than
with any European literature.
Moreover, insofar as there was an interaction between African and Euro-
pean thought, it did not flow only one way.74 Local historians such as John-
son, as well as the older Islamic scholarship and oral traditions, were regu-
larly cited by Europeans writing on African history. For example, the stan-
dard textbooks on Nigerian history written in the 1920s and 1930s, by the
British colonial officials Alan Burns and Rex Niven, both posited Middle
Eastern origins for the Yoruba. But they did so on the authority of the
African Samuel Johnson (and, here again, in preference to espousing the
alternative “Atlantic” theory of the European Leo Frobenius).75 African

72The pioneer of the integration of the Borgu legend of Kisra with Yoruba traditions of
origin (as reported by Johnson) seems to have been H.B. Hermon-Hodge, Gazetteer of
Ilorin Province (London, 1929), 115-21.
73“Educated natives of Yoruba are well acquainted with the history of England and with
that of Rome and Greece, but of the history of their own country they know nothing
whatever! This reproach it is one of the author’s objects to remove:” Johnson, History,
vii.
74Again, this point is made by Zachernuk in general terms, in Colonial Subjects, 66, but
he does not apply it to the case of the “Hamitic hypothesis.”
75A.C. Burns, History of Nigeria (London, 1929), 32-33; C.R. Niven, A Short History of
Nigeria (London, 1937; 3rd ed., 1948), 64-65. Also C.R. Niven, A Short History of the
Yoruba Peoples (London, 1958), 6-7.
 The “Hamitic Hypothesis” in Indigenous West African Historical Thought 313

constructions of the “Hamitic hypothesis” thus reinforced the dominance of

this paradigm in European thought, as much as vice versa.

VII

Zachernuk concludes his study by referring to the decline in the popularity

of the “Hamitic hypothesis” among Nigerian historians since the 1950s, and
links this to decolonization: “the Hamitic Hypothesis and the colonial order
decline together . . . the Hamitic Hypothesis was a colonial cultural artefact,
not suited to post-colonial times.” However, he acknowledges that there was
in fact “not a clean break,” and that versions of the “Hamitic hypothesis”
have continued to have some currency in modern Nigeria.76 In my own
reading, the abandonment of the “Hamitic hypothesis” was clear enough
among academic historians—although this perhaps had less to do with the
rejection of colonial paradigms than with a shift of scholarly interest away
from “origins” towards other (more recent) subjects. But it still seems alive
and well in local amateur historiography.
In Yorubaland, it is true that the most substantial recent treatment of the
“origin” question, by Chief M.A. Fabunmi, published in 1985, rejects theo-
ries of migration from outside, in order to restate the traditional local (Ife)
account that the world (and therefore civilization) was created at Ife. On
Fabunmi’s view, any similarities between ancient Egyptian and Yoruba
civilisation are to be explained by diffusion from Yorubaland to Egypt,
rather than vice versa.77 However, he adopts this position in explicit repudi-
ation of contemporary as well as earlier literature which has continued to
maintain the foreign origins of Yoruba culture. Indeed, only a year previ-
ously (though too late for Fabunmi himself to take account of it), the migra-
tionist thesis had been restated in another local publication (in this case,
claiming specifically Jewish origins for Yoruba religion), by Canon R.A.
Fajemisin, writing from a perspective outside Ife (in the eastern Yoruba
town of Ilesa).78
Explicit in this recent revival of the controversy over Yoruba “origins”
are disputes over the contested claim of the Oni of Ife to seniority of status
among Yoruba rulers. The creationist story self-evidently underwrites the
primacy of Ife, while the migrationist alternative can be employed (as it is
explicitly by Fajemisin) to challenge it. Zachernuk registers this sort of use
of “Hamitic” links in the service of local micro-nationalisms (citing, among
others, the case of Fajemisin), and suggests that this represents a further
transformation of the Nigerian use of the “Hamitic hypothesis,” reflecting a

76Zachernuk,“Origins,” 453-54.
77ChiefM.A. Fabunmi, Ife, the Genesis of Yoruba Race (Lagos, 1985).
78Canon R.A. Fajemisin, Primacy in Post-Oduduwa Yorubaland (Ilesa, 1984).
314 Robin Law

fading of concern with contesting “imperial” ideology and a reorientation

towards internal Nigerian issues.79 It is questionable, however, how new this
really was. There had always been an ambiguity about whether supposedly
prestigious foreign origins were claimed for Africans as a whole, or for par-
ticular peoples in contrast with others—an ambiguity which goes back at
least to Johnson, who certainly claimed a special status for the Yoruba
among Africans, as well as the dignity of Africans vis-à-vis Europeans.80
Already by the 1900s, the study of Yoruba history had become implicated in
contests over seniority of status and territorial disputes between different
Yoruba groups.81 Rather than being a specifically post-colonial phenome-
non, this was an element of continuity with the colonial past: the principal
dialogue in which Yoruba historians engaged, it may be suggested, had
always been an internal one.

79Zachernuk, “Origins,” 451-52.

80Johnson, History, xxi-xxii; for the claim of Johnson and other early local historians to
the distinctiveness of the Yoruba, see more generally Law, “Constructing,” 91-92;
idem.,“Local Amateur Scholarship,” 79-82.
81Law, “Early Yoruba Historiography,” 78.