Who Are the Luo?

Oral Tradition and Disciplinary Practices in Anthropology and History

Author(s): John R. Campbell
Source: Journal of African Cultural Studies, Vol. 18, No. 1, Language, Power and Society:
Orality and Literacy in the Horn of Africa (Jun., 2006), pp. 73-87
Published by: Taylor & Francis, Ltd.
Stable URL: http://www.jstor.org/stable/25473357
Accessed: 28/08/2010 19:05

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=taylorfrancis.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Journal of
African Cultural Studies.

http://www.jstor.org
Journal of African Cultural Studies, t\
Routledge ^^^0^
Volume 18, Number 1, June 2006, pp. 73-87 t\

Who are the Luo? Oral tradition and

disciplinary practices in anthropology
and history*

JOHN R. CAMPBELL
(School of Oriental and African Studies, University of London)

ABSTRACT What is oral tradition, and how can it help elucidate the past and
understand the relationship between culture, social organization and identity
today? It turns out that this question is complicated by the influence of early Euro
pean narratives that described and defined African society and which have also
indelibly marked
the methods, assumptions and forms of narrative writing used
by contemporary social science. This paper addresses this vexing issue with
respect to research on the Luo-speaking peoples of Eastern Africa by examining
how anthropologists and historians have approached 'oral tradition' and how
their approach has influenced the way they write about Luo culture, society and
identity.

This paper the production

examines of 'academic' knowledge, in the form of

anthropological and historical texts, about the Luo-speaking peoples of Eastern

Africa. Research on the 'Luo' began with early explorers, missionaries, administra
tors many of whom were influenced by the 'Hamitic hypothesis', a 'historians'
myth' that significantly influenced academic and lay understandings about 'Luo'
identity and their place in the Horn of Africa. Section 1 therefore examines the con
tribution of this early scholarship - in the -
particular significance of oral tradition
and its influence on subsequent anthropological and historical writing and on
generations of Africans in the region.
In Section 21 look at the significance of oral tradition for anthropological knowl
edge and writing, and in Section 3 I examine historical knowledge and writing on
the Luo. I conclude by noting that despite the apparent convergence in the narrative
accounts produced by anthropologists and historians, fundamental differences
between the disciplines remain based upon methodological grounds and the
approach taken to 'oral tradition/history'. Secondly, while we cannot control the
manner in which our writings are used or understood by others, we should never

*
Iwould like to thank Donald Crummey and David Anderson for their helpful comments
and suggestions.

ISSN 1369-6815 print; 1469-9346 online/06/010073-15 ? 2006 Journal of African Cultural Studies
DOI: 10.1080/13696850600750327
74 John R. Campbell

the less alert our readers that our 'claim to authority' in representing ethnographic
and historic 'others' is tenuous at best.
Two points need to be made at this point. First, within the scope of a paper there
is a limit to the amount of material that can be discussed. This paper does not
attempt to provide a definitive assessment of material about the Luo, nor have
I thought it necessary to refer to the entire written corpus of the individuals
who are cited. By and large my criticism is based on a critical reading of key
monographs and I consider that within that format the author(s) had ample space
to address the issues I discuss in this paper. Secondly, this paper should not be
read as being in any way hostile to inter-disciplinary research; on the contrary,
clarity about disciplinary methods and approaches is the foundation for inter

disciplinary work.

1. Early European scholarship, the 'Hamitic hypothesis' and 'race'

An understanding of the development of modern social science, and especially of

its methodological and 'narrative traditions', requires if not an examination then an
awareness of the influence of earlier narrative forms. For example Thornton's
examination of nineteenth century travel writing and missionary scholarship on

emerging forms of narrative in anthropology reveals the significance of work by

missionaries (via their linguistic studies and translations of African languages)
for nineteenth century 'arm chair' theoreticians (1983).
In particular, and quite apart from Malinowski's influence, early philological
research a tradition of fieldwork
(including an insistence on collecting
bequeathed
vernacular texts) to modern anthropology. A further, and often overlooked contri
bution of philological research came in the form of a professional, 'scientific' voca
that included new 'sociological' terms and an analytical focus on 'native
bulary
terms' (i.e. mana, totem, taboo, etc.). Philological research sought to classify
African and peoples and, in the process, coined new terms for large
languages
groups of 'related' peoples. Thus, in addition to terms like the 'Baganda' and
'Zulu'

... the term Bantu had even further reaching historical effects, both intellectual and
a
political, since it came to designate, ambiguously, an imagined 'race', conjectured
common history, a family of languages, a Zeitgeist or worldview, a 'stage of civiliza
tion', or a culture (Thornton 1983: 512).
While the term 'Bantu' emerged out of the study of African languages, other terms
also entered into the scholarly discourse on Africa which derived entirely from

European speculation. Thus the nineteenth century explorer Richard Speke

(1863) is said to have originated the idea that the bearers of culture and civilization
were invaders from Christian The
to Africa light skinned Hamitic Ethiopia.
'Hamitic hypothesis', which was linked to the Biblical myth of the dispersal of
the 12 tribes out from Israel, provided Europeans with a set of ideas that enabled
them to 'explain' the physical and cultural differences they encountered in the

region (Wolf 1994).

Who are the Luo? 75

A close examination of the writing of Europeans and the newly educated African
- in
elite demonstrates the power of the Hamitic hypothesis which functioned
social science as a modern myth and sociological model devoid of a factual
- to explain
basis European dominance and the savagery and barbarism of black
Africa (MacGaffey 1978).
on the - as a
For example, historical research 'Abacwezi' initially understood
-
mythic account of an ancient
dynasty of light-skinned kings in Central Africa
underwent a 'complete inversion' that turned mythic figures into 'living persons'
'who occupied the whole country then, suddenly, disappeared underground'
(Berger 1980: 67). A key contributor to this theory was the British administrator
Sir Harry Johnston who, in 1902, argued that the light-skinned Abacwezi were
in fact 'alien rulers from the north' who were the source of African knowledge
of plant domestication, iron smelting, music, etc. (ibid: 68). In the 1950s, the
Hamitic hypothesis was also used by the African elite of Uganda to present them
selves as the heirs of the 'exalted semi-European outsiders' for the purpose of
enhancing their own claims and the status of their 'tribe'/kingdom (ibid: 70).
Colonial administrators in Central Africa developed an especially pernicious,
racialist version of the Hamitic hypothesis to differentiate between African
'races' and to justify colonial forms of governance. In Rwanda and Burundi
... ... seized the occupational it
European conquerors upon categorization, imbuing
with a hierarchical racial classification. The Tutsi minority were identified as a
Hamitic aristocracy, who ruled a state with such sophistication that they could
only have originated from a place geographically, culturally and above all racially
nearer Europe (DeWaal 1994: 3).

As Eltringham shows, this mode of was systematically thought

pursued to the
extent that in the census of 1933-34 'every Rwandese was assigned an 'ethno
racial' identity (15 per cent Tutsi, 84 percent Hutu, 1 per cent Twa) and issued
with an identity card' (2004, p. 15). This racial schema was subsequently interna
lized by Rwandese and Burundians and contributed directly to widespread ethnic
violence and to the 1994 genocide. Taylor writes that

Tutsi extremists make use of their version of the hypothesis to claim intellectual
superiority; Hutu extremists employ theirs to insist upon the foreign origins of
Tutsi, and the autochthony of Hutu. No matter which side uses the Hamitic hypoth
esis, however unwittingly, it a colonial one that essentia
reproduces pattern:
lizes ethnic difference, justifies political domination by a single group, and
nurtures a profound thirst for redress and vengeance on the part of the disfavourized
group (1999: 57).
In the early 1950s when the Hamitic hypothesis was being institutionalized in
central Africa, historical research elsewhere in the region was arguing for a
'modern' conception of regional history and culture.
In The Lwoo (1954), Father H.P. Crazzolara, supported by the Istituto Missioni
Africane, sought to trace the history and customs of the (Nilo-Hamitic) Lwoo/Luo
people from AD 1000. This history of Lwoo migration from the Bahr el Ghazal area
of modern Sudan to present day Uganda and Kenya is based upon evidence derived
from oral ('native') - on
tradition the assumption that 'such narratives as present an
76 John R. Campbell

of history, are built upon historic foundations' -

outward appearance (p. 11) and
the presumed linguistic 'identity' between the languages of the Lwo, Dinka and
Nuer (adduced from comparative lists of names for different parts of the human
body).
Crazzolara's account is anchored in a conception in which distinct 'races' can be
identified as the bearers of a distinct culture and tradition. Thus,
... I do not agree with the school of ethnology which prefers to use artificial and, in
my terms such as 'Nilotic', 'Nilo-Hamitic', 'Sudanic' etc. as
opinion, meaningless
labels for groups of tribes with linguistic and cultural affinities. Here ... we are
dealing with clearly distinct racial political communities, and it is both more accurate
and more convenient to use the group names by which they have been known to one
another for centuries, that is to say 'Lwoo', 'Lango', and 'Madi' (Crazzolara 1961:

141).
The significance of Crazzolara's work for the emerging disciplines of Anthropol
ogy and History is two-fold. First he provided a conceptual focus on culture (the
concept of race was to be replaced with tribe, and subsequently by ethnicity1) as

something that was ostensibly produced and carried by discrete and clearly
bounded social
groups. Second, he suggested a methodology for comparative
research based
upon the analysis of oral tradition. While the Hamitic hypothesis
was eventually discarded as conjecture, the idea that history could be discovered
an awareness -
by collecting and analyzing oral tradition remained despite by
and his successors - that the 'memory' contained in oral testimony
Crazzolara
altered with the passage of time. Even so, the prospect that oral testimony might
unlock 'history' proved too promising a tool for anthropologists and historians to

ignore.
To summarize the argument thus far, the Hamitic hypothesis represents a form
of historiographic myth created by Europeans. The methodological issues sur

rounding its creation and perpetuation arose from European misinterpretation

and misconception of African statements about the past, and a situation in which
writers drew freely and without acknowledgement on each others' accounts,
making it impossible to independently verify individual accounts. At the same
time, socio-political pressures for decolonization, growing levels of literacy, and
nationalist political fervour gave greater salience to the production of elite
'tribal' histories that borrowed ideas and concepts from European academic

writing and discourse.

2. Anthropology of the Luo

While a growing number of books have delved into the history of British social

anthropology (Kuper 1993; Stocking 1983; Goody 1995; Parkin 1990), they
focus on specific or on the 'discipline'. For this
primarily individuals/institutions

1
In anthropology, attempts to conceptually delimit clearly bounded social groups (i.e.
races, tribes, ethnic groups, etc) has been heavily criticized (cf. Southall 1970, Lentz
1995).
Who are the Luo? 77

reason they are of little help looking at interdisciplinary research or work that was
influenced by linguistics and history.
Broadly speaking, the discipline developed out of the speculation of Victorian
'armchair' ethnologists whose theories were heavily dependent on research con
ducted by missionaries and colonial administrators. The 'facts' generated by
such 'men on the spot' (James Frazer's term for those who sent data to the metro
politan theoreticians) was constrainedby current theorizing, the absence of sub
stantive data, evolutionist thinking, and a disregard for the study of language and
linguistics (which, in British anthropology, lasted until the early 1960s; Henson
1974). Gradually 'metropolitan theoreticians' turned their attention to the methodo
logical problems involved in soliciting 'facts' from 'natives' and, between
1841-1912 they produced a sophisticated questionnaire to be used by their 'men
on the spot' (i.e. 'Notes and Queries in Anthropology'; Urry 1972). Toward the
end of this period, and following on from the fieldwork undertaken by the 1898
Torres Straits Expedition, and by W.H.R. Rivers and Malinowski, the emerging
discipline of social anthropology embraced fieldwork as its sine qua non.
Alongside the institutionalization of fieldwork there emerged a 'new paradigm',
functionalism (first as a method and subsequently as a philosophy; Kuper 1993: 85
ff). One of the first post-war statements reflecting this paradigm is found in African
Political Systems (Evans-Pritchard & Fortes 1940) which set the agenda for the com
parative study of Africa by focusing largely upon lineages as universal social insti
tutions (i.e. that regulated marriage, religion, ritual, economic relations, etc).
As it happens, Evans-Pritchard was the first professionally trained anthropolo
gist to study the Luo in the 1930s (in the area around the Kavirondo Gulf in
western Kenya). Illness shortened his visit (to less than 6 weeks) and restricted
him to interviewing - in English -
Luo mission converts (Evans-Pritchard
1965a: 228). His principal purpose was to survey the political structure of the
Luo whom he believed to be closely related to other Nilotic such as the
peoples
Nuer. His account takes the form of a realist description of the area, its people,
culture, and social structure (1965a, b).
In 1949 Aidan Southall conducted extended
fieldwork among the western Luo of
Uganda and wrote the first monograph length account of a Luo-speaking people,
Alur Society (1953). Southall argued that the Alur were intermediate between the
Luo of - to were
acephalous Kenya whom they related by history, migration,
and culture - and
language the centralized political states of Bunyoro and
Buganda. His objective was to describe and analyze the process which the
by
Alur had come to dominate other the use of force. In effect,
peoples without
Southall sought to define a fifth form of political system to those identified in
African Political Systems, namely the segmentary state.
Alur Society differs from other ethnographies written in this period. First and
foremost it is concerned with 'political theory'.2 The monograph focuses primarily

Its concerns derive from the work of Max Weber whose work was in
just appearing
English.
78 John R. Campbell

on socialstructure at the expense of social relations and beliefs, and there is

relatively little use of vernacular terms and categories. In other ways, however,
Alur Society does conform to the genre of ethnographic writing (see Marcus &
Cushman, 1982). Southall developed a specialist vocabulary (a jargon or meta
language) to tease out the significance of fieldwork findings and convey meaning
to an audience of anthropologists. The text takes the form of a generalized, some
what static account of the organization and function of Luo lineages; notably the
anthropologist is invisible (in the sense of being absent from the text) though his
perspective defines the manner in which the Luo are represented. Furthermore,
the text's validity is largely dependent upon the anthropologist's claim to have
lived among the Luo.3
Okot p'Bitek, a British-trained
Ugandan anthropologist, undertook fieldwork in
1962. A poet, dramatist and writer of fiction, Okot wrote and published primarily
for African audiences. His Religion of the Central Luo (1971) argued against nine
teenth century European theories, and in particular the Hamitic theory which, he
believed, devalued the contribution of Africans to the history and culture of the
Great Lakes. Okot also sought to refute the work of Christian missionaries, who
he argued had mistranslated Luo into English to evangelize Luo-speakers. They
had, he argued, misinterpreted Luo religion by turning it into the 'High God' of

Christianity.
Religion of the Central Luo differs significantly from other ethnographies. The
introduction merely defines the book's objective as describing 'the religious ideas
-
and practices of the Luo people' resident in east and central Uganda. The book
which looks at theories of jok or spirit, ancestral shrines, witchcraft and sorcery,
etc. - observation and draws instead from
cursing, disavows participant insights
myth, ritual, song and poetry. In so far as he identifies his sources, information
comes from family members.4 Religion self-consciously stands outside the estab
lished genre of ethnographic writing in the form it takes (i.e. as a rebuttal of

European speculation), in the data it utilizes, and its concern with the contribution
of the eastern Luo to African culture. Significantly, Okot is acutely aware of the
limitations of oral tradition and myth as a basis for generalization. Not only does
he that memory
argue (recollection of the past) is strongly affected by the
passage of time and by religious conversion etc., he notes that all myths 'are
about the foundation of existing institutions and political groups'. Indeed, Okot
argued that

... it is not possible to reconstruct the histories of the Northern, Central, and
Southern Luo groups separately, it is not possible to trace the history of all
the+Luo peoples to the first man, Luo. This is because oral traditions are

3
For example, there is no discussion of his fieldwork nor about his research methods and
the data they generated.
4
As the son of teacher, he attended prestigious local schools prior to attending Bristol
University, obtaining a law degree at Aberystwyth, and undertaking a doctorate in
anthropology at Oxford (entitled 'Oral literature and its social background among the
Acoli and Lang'o'; Heron 1976: 3).
Who are the Luo? 79

concerned not so much with the ultimate and historical origins, but with the
foundation and maintenance of existing institutions. Crazzolara's attempt to trace
the history of the Luo people as far back as A.D. 1000 is thus a futile exercise
(1971:3).

Religion of the Central Luo is an account of a cultural 'insider': Okot was an Acoli,

quite possibly descended from a chiefly lineage, the son of a Church Missionary
Society teacher, and among the first generation of university educated Ugandans.
His elite background raises issues about his interpretation of Acoli/Luo culture.
How might the ethnography differed if it had been written by a commoner,
have
a Ugandan from a different group, or by a woman?
ethnic Indeed, Okot's outspoken
attack on the influence of European culture on African society suggests that his per
spective was quite distinctive.5
A related issue concerns the anthropologist's competence in the vernacular
language of the society she/he researches (Owusu 1978). For example, while
Evans-Pritchard interviewed his Luo informants through an interpreter, Southall
gained basic competence in Dholuo midway through his fieldwork (i.e. after 9
months), and Okot's is partly demonstrated
competence by his use of vernacular
concepts though he
says nothing about his linguistic abilities. There is an
ongoing debate in anthropology about the role of language in fieldwork between
those who argue for a basic grasp of a language (quite possibly based on a
limited grasp of phrases/locutions) and those who argue, pace Malinowski, that
the ethnographers - 'to -
task grasp the natives point of view' implies the need
for fluency in the vernacular (Owusu 1978: 313-4).
While Owusu berates anthropologists for their lack of facility in the vernacular
-
indeed he argues for a need to 'speak and understand several local relevant ver
naculars' tominimize and informant error -
data/translation (his emphasis, p. 313)
his emphasis on language competence is at the expense of other skills that are also
important (cf. Henson 1974). Thus to focus on an ethnographer's ability to 'speak
and understand' one or more vernacular valuable this undoubt
languages, though

edly is in comprehending life, is to undervalue

social the linguistic training required
to comprehend language use (e.g. language shift, the ability to deal with differences
of dialect/argot, etc.) and the ability to accurately collect and transcribe vernacular
texts. Even so, the anthropologist's task is not limited to accurately recording and
reporting what her informants say or do, we are also tasked with the problem of
interpreting and writing-up our fieldwork in a manner which is transparent and
which addresses issues other than language.6
To conclude this section, we have seen that the ethnographic accounts of the Luo
vary considerably from realist accounts to a dialogue between the ethnographer and
her/his sources. Just as important, there is a conspicuous absence of any discussion
of fieldwork, of research methods, and of the linguistic competence of the ethno
grapher. Furthermore, the contribution of social theory in interpreting fieldwork

5
See his interview in Lindfors (1980).
6
See Altheide and Johnson (1994) for a recent discussion about the validity of ethno
graphic accounts.
80 John R. Campbell

findings is either conspicuous by its absence, or it tends to exclude detailed

accounts of culture, beliefs, etc.

3. The 'new' African historiography of the Luo

Historical research in the region took a different route from that followed by anthro
pology. Historians initially argued that in the absence of documentary records
('facts') historical analysis was not possible (cf. Evans 1997), which meant that
in sub-Saharan Africa historiography was based on European sources, archives,
the work of missionaries and administrators, or fieldwork. An example of the his

toriography of this period is Roland Oliver's 'Discernible developments in the

Interior, c. 1500-1840' which deals with Uganda and Lwo history (1963). With
one exception (Oliver 1959), his account is based entirely on secondary sources
- and 'literate Africans'
the accounts of missionaries, historians, anthropologists,
and colonial district books. Indeed, while critical of certain claims, e.g., the 'doubt
ful etymological evidence' provided by Crazzolara (fn. 1, p. 172), he nevertheless
relies on'genealogical evidence' found in clan/tribal chronologies and 'special'

(Nilo-Hamitic) clan names to explain the survival of groups incorporated by the

Nilotic Lwo.
At roughly the same time a 'new African historiography' was developing based
on the work of Jan Vansina (Oral Tradition, 1965). As Cohen noted, Vansina
offered a way to 'join science to the study of African voices' (2001: 50), and it
was not long before historians took up the challenge. In East Africa the new histor
a Kenyan and a Joluo, in his History of the
iography was initiated by Bethwell Ogot,
Southern Luo (1967). Ogot was the first professionally trained historian to engage
with oral tradition in the region. He distanced his research from
systematically
work that blurred the distinction between collecting and analyzing oral tradition

(particularly Crazzolara). Such accounts were problematic because they did not
leave behind
... the vernacular texts, the circumstances under which the material was collected,
and the manner of transmitting oral evidence in the particular societies they
studied. Consequently, they presented us with accounts which are difficult to inter
pret. Oral traditions need to be transformed into written texts before the historian's
art can be applied (my emphasis, p. 19).

While anthropologists used oral tradition, Ogot argued that

... they too lack the necessary historical training; and on the whole, they tend to con
centrate on those aspects of oral traditions which relate to their theoretical models,
especially myths, legends and genealogies (p. 20).
to
In addressing 'historical questions' Ogot pursued a radically different approach
and analyzing oral tradition. To begin with, considerable effort went into
collecting
and translating histories written in Dholuo, and in convening
collecting lay/popular
with clan elders (18 clans of Luo origin in Uganda, and 4 Luo clans in
meetings
western with the aim of producing official clan histories. Discussions
Kenya)
with assembled clan elders took place over several days and culminated in 'one
named the official story to Ogot who tape-recorded it. Clan
expert' relating
Who are the Luo? 81

elders were also asked to comment on the 'official' history of other clans as ameans
of cross-checking accounts.
Overall, Ogot collected oral evidence in the form of

'genealogies, migration stories, clan songs, histories of clan and tribal cults, his
tories of social organization, i.e. changes in the social structure, and evidence on

place names' (p. 23). All oral testimony was transformed into text for analysis
and was examined against other 'evidence'.
History of the Southern Luo is an 'ethnic'7 history recounted in the third person
as a 'factual' academic history. The following excerpt provides an indication of

Ogot's method8 of interrogating clan histories:

Up to this point, the traditions are unanimous, and indeed from Acholiland itwas the
most likely route a migrating group of Pastoralists would follow. But there are con
flicting accounts as to where the Luo went when they abandoned their temporary
settlements in North Ugenya. According to one account, they first went toManga
inWanga, and thence to Boro in Alego; and according to another account, they
went straight from North Ugenya to Alego. But these accounts, both of which are
given by experts on Luo traditions, are contrary to the accounts of some of the
Joka-Jok themselves ... (Ogot 1967: 149).

Interestingly, Ogot eschews a search for absolute dates in favour of a 'relative'

chronology defined as the estimated length of a generation, i.e., the time that
elapses between the birth of a man and his first surviving child. Based on inter
views, Ogot a 'generation'
calculated as 26.5 years, which enabled him to date
Luo migration to western Kenya back 16 generations (to AD 1490-1517 plus or
minus 52 years; p. 28). In this manner Ogot sought to put historical research on
a scientific basis.
A closer examination reveals
that Ogot, like Crazzolara and some anthropolo
gists, assumed the existence of clearly bounded and socially discrete 'clans' that
were believed to exist through time as culture-bearing units.9 In this way, clans
were linked with the concept of 'generation' enabling a chronology to be calculated
for Luo migration. Just as significantly, Ogot uncritically accepted the claims of the
elders as the custodians of historical knowledge, thereby lending academic support
to elite male views which were subsequently recorded as undisputed 'facts'. It is
worth noting that 'traditions' are deployed by contending parties to contest and
influence the outcome of disputes/litigation; as such tradition is better understood

7
'Ethnic' history is one of several historical genres of writing/representing the past, in
this instance it is produced by a professional historian. See Atkinson's discussion of aca
demic reliance on the 'stylistic conventions of realist fiction' and the way in which
different disciplines invoke specific criteria to 'partition' the field in an effort to
situate it from other texts and disciplines (1992: 30).
8
In the words of his successors, 'Like the larger and less honed work of J. P. Crazzolara,
Ogot's History of the Southern Luo is essentially nomenclatural. The central technique
... was and the central mode of reconstruction was
genealogical name-linkage' (Cohen
& Odhiambo 1989: 17).
9
'Clan' boundaries appear to correspond with contemporary speech communities, an
assumption that privileges existing Luo dialects, traditions, etc. In perceiving ethnic
boundaries as fixed and timeless, this assumption fails to deal with historical change
i.e. in and the movement of ideas, etc.
language/dialect, people,
82 John R. Campbell

as situated knowledge that reflects contemporary socio-political processes rather

than as a historical truth.
In addition Ogot's primary data, the transcriptions of officially agreed traditions,
is 'archived' and is not available to his readers; instead he references the transcripts
via footnotes (e.g. as Padhola Historical Texts vols. I, II, etc). This practice
sidesteps concerns testimony was transcribed
about how (Emerson, Retz & Shaw
2001) and about how
linguistic and other problems involved in analyzing
such material were handled. Interpretative problems are further compounded
because the transcripts are subjected to the same type of analysis as other types
of texts, i.e., they are scoured for consistency and 'nuggets' of data rather than
to understand them as statements reflecting the socio-political status of 'clans' or
as culturally encoded statements indicative of Luo perceptions of themselves and
their world.
The successor toHistory of the Southern Luo is Siaya, The Historical Anthropol
ogy of an African Landscape (1989) written by David Cohen (an American) and
Atieno Odhiambo (a Kenyan and a Joluo). Siaya takes up the challenge of provid
ing a voice for the Luo by contrasting the views of anthropologists (who are said to
see the Luo as an exemplar of a segmentary patrilineal society) with those of his
torians and differently positioned Luo who are represented as active social agents
involved in producing their own culture and history.

Though the contrast

overly simplifies disciplinary difference, it provides an

interesting entree
into differing discourses about what it means to be a Luo. For

example, separate chapters address the issue of social and geographic boundaries;
the meaning of 'Siaya' as a homeland (for run-away fathers, migrants, and the
elite); the problem of 'land hunger' (for the urban rich vs. the rural poor); and
the paradox faced by rural widows whose efforts to gain control over their lives
are frustrated by men.
In the text Siaya is simultaneously:

a historical locale in western Kenya where Dholuo-speaking peoples have settled

and from which many have emigrated;
an administrative district;
an (ethnic) 'homeland'; and
a socio-cultural landscape central to the reproduction of contemporary Luo
culture in which individuals confirm their ethnicity (e.g. through building a

house, burial, performing ritual, etc.).

The Luo are said to constantly redefine their history through a process of 'rehearsal'
in which current themes and concerns are reworked by the elders who, in effect,
rewrite 'custom' (the past) and in the process validate their authority as 'custo
dians' of clan history, land, etc.
Cohen and Odhiambo take their argument a step further in Burying SM: The
Politics of Knowledge and the Sociology of Power in Africa (1992) which chron
icles the use of 'custom' in litigation and in particular the role that academic and
other texts play in disputes as a 'factual' record of the past. While the case of
SM is unique, the political use made of 'custom' in resolving disputes should
Who are theLuo? 83

remind academics that disciplinary accounts are also read by laymen who apply
their own criteria of 'validity' when they read our work.
While the narrative device of contrasting disciplinary and Luo perspectives
makes for
interesting reading, it obscures important issues. Perhaps the most
obvious concerns the basis on which specific ethnographic examples are selected
by the authors and inserted in the text to validate their argument. For example,
in chapter five there is an extended account of a visit by a jogam (the intermediary
sent to negotiate for the hand of a man's intended wife; pp. 103-9). Not only is it
the only account of jogam provided, very little information is given about the
context, the participants, etc. (i.e. information that would allow readers to assess
the extent to which the event is 'typical'). Furthermore, the account is a partial
and partisan recollection of one man.
The issue of perspective is complex. Whereas Ogot's history is written from a
perspective that purports to be above and outside individual/clan concerns, this
is not the case inasmuch as both books incorporate transcribed field data in the
form of 'stories'
and first person narratives from informants. As Atkinson has
remarked of such practices, not only has 'a great deal of tacit and invisible authorial
work in collating, editing and rewriting the personal narratives' taken place, the net
effect is to produce a 'highly conventionalized' representation that supports the
authors' perspective as against demonstrating the views of other social actors
(1992: 25).10
Finally, Cohen and Odhiambo do not discuss their fieldwork, the methods
they employed to gatherdata, nor their competence in Dholuo (though poems,
proverbs and popular songs are cited in the vernacular and in English).
Interestingly, such material does not seem to be drawn from extended observation
nor is it discussed in relation to specific settings or events; instead it is cited to
support the authors'
arguments regarding the significance of particular kinds of
discourse. the narrative
Finally, format of Siaya and of Burying SM shows
to ethnographic accounts - in terms of the of the
greater similarity invisibility
author in the text, the realist writing conventions adopted,11 and the implicit
assumptions about ethnicity, social boundaries, etc - than with History of the
Southern Luo.

4. Conclusion: disciplinary methods, oral tradition and explanation

Over the past 40 years there has occurred a remarkable convergence between
anthropology and history: both disciplines undertake fieldwork, both rely to
varying degrees on working in local vernaculars, and both make use of oral

10
Stamp, for example, looks at the burial of SM in terms of its implications for women
(1991). The issue of perspective is complicated when an author is writing about his/
her own ethnic
group.
11
A further issue concerns the intended audience for a publication. David Anderson has
suggested that both books by Cohen and Odhiambo were primarily intended for a literate
Luo audience, and specifically for a group of Luo 'professors' (comments made during
the discussion of the paper).
84 John R. Campbell

testimony in writing their narratives (Faubion 1993). Indeed, Krech has argued that
the convergence is driven by shared theoretical concerns and that disciplinary
boundaries and genres have become 'blurred' (1991: 350).
However, while it is clear that the narrative/textual forms of writing employed
by the two disciplines converge in their reliance on realist conventions of represen
tation, considerable differences remain. A notable difference is found in the manner
in which each discipline 'partitions' the field: historians adopt certain criteria to
establish or define
a specific genre of writing such as 'ethno-'/'ethnic' history
(Krech 1991); in anthropology one finds 'regional' ethnographic traditions
(Fardon 1990). Just as significant has been the manner in which each discipline
addresses, or fails to address, key methodological issues including: linguistic com
petence in fieldwork; problems of translation and transcription of texts; the link
between theory, method and data in the production of disciplinary knowledge;
and thevexed question of textual writing strategies in representing ethno

graphic/historic 'others'.
I suspect that the very different way which the two disciplines have made use
of oral tradition is a reflection of the recent turn to fieldwork by historians
and of deep seated differences regarding how to find and interpret 'facts'.12
In this regard Evans' claim
(1997: 77) that in history 'facts thus precede

interpretation conceptually, while interpretation precedes evidence' obscures a

deep division between historians regarding the reliability of certain types of

data and a stubborn reliance by those who prize 'oral tradition' over other
sources.13

Significantly, nearly all the writers discussed in detail look principally at the
verbal element/aspect of oral testimony and in doing so tend to adopt a literalist
- reflected of an account onto
approach partly in an unproblematic transcription
- and to or the social context in
paper gave little attention allegory, metaphor
which 'tradition' was/is orally performed/ transmitted. Nor do they appear to be
aware of the significant differences between different 'genres' of oral performance
- i.e. between the verbal 'arts' (poetry, song, narrative recitations) and dance and
'theatrical' performances, much less a concern with metre and different types of
audience participation.
Interestingly, the key issue here is not, strictly speaking, disciplinary. As Vail
and White (1991) have argued, research on oral tradition has with few exceptions
the same approach. By affirming a problematic distinction between
adopted
literature and oral performance, academic research has tended to
written/printed
focus on the narrative element of oral performances without examining the
context in which they were performed which embues a narrative with meaning

12 - that
The extent to which this reflects earlier disciplinary orientations anthropologists
documents - remains to be seen.
people, while historians study
13 study
Miller's The African Past Speaks (1980) argued for the unequivocal value of oral tra
dition above all other sources, while Spear (1981), Waller (1999) and others have
argued that oral evidence should be integrated with data from comparative linguistics,
etc. and not treated as something unique.
archaeology, ethnography
Who are the Luo? 85

(e.g. as satire, elegy, social commentary or praise). Equally problematic is the lack
of attention to issues of power involved or reflected in the oral performance, which
results in a reductive approach to the words/narrative as reflecting some historical
'truth' rather than as an evolving aesthetic or poetic tradition (p.71). Their ability
to invoke license', however,
'poetic depends critically not merely on what is said
but also on how it is said and performed to his/her patron, ruler and audience
(Furniss & Gunner 1995).
It should be clear, therefore, that transcribing and transforming oral testimony
into a written text is hugely problematic. First, the task is methodologically
fraught and may divorce verbal elements from their cultural context. By becoming
fixed in print, a commentary/tradition may loose its flexibility and its social value
as an independent commentary on life. Just as important, some forms of oral tra
dition may be secret and access to it may be controlled because the knowledge
encoded in it is power.
At the same time, academic texts circulate alongside other written (written and
oral) accounts; not surprisingly academic
writing may inform
lay understandings
and behaviour in unpredictable ways. The longevity and salience of the 'Hamitic
hypothesis' was noted, as was the use made of published ethnographies in Kenyan
courts. Similarly, Shetler (2002) draws our attention to the recent spate of written
'tribal' histories in south central Tanzania that purport to tell the history of an
'ethnic' group. Shetler's account is a useful reminder of the pitfalls that confront
academics, and his proposed solution is worth serious consideration. He reminds
us about the importance of the wider political context in which - it is
laymen
almost always literate men, not women - use writing to 'translate the past into a
useable idiom for the present', one which often promotes particular political ends
(ibid: 424). Secondly, he reminds us that we cannot control the way our writings
are used or understood by others. Even so, Shetler suggests that the problems
posed by laymen who use academic texts can be met in part by writing and publishing
accounts in a format (e.g. an edited volume) that forces readers to compare preferred
accounts (i.e. an 'ethnic' history) to other accounts in an attempt to 'reveal the
constructed nature of tribal histories and to bring to light their common cultural
roots' (ibid).
Iwould only add, that in view of the many limitations of academic research and
of the myths that it has helped promulgate, we need to recognize that our claims to
- in terms of our
authority ability to represent other cultures and/or to capture and
define an authentic African 'voice' - are limited. These limitations should be expli
citly acknowledged in our accounts by discussing problems of research method,
language competence, etc. and by stressing how comparative Study is important
(e.g. to understand shared culture, and that 'historical' accounts are constructed)
so that lay readers are not left with the impression that research is only or even pri
marily about discovering cultural difference.

John Campbell can be contacted Department of Social Anthropology, School of

Oriental and African Studies, Thornhaugh Street, London WC1H OXG; email:
[email protected]
 86 John R. Campbell

REFERENCES

Altheide, D. & J. Johnson. 1994. Criteria for assessing interpretative validity in qualitative
research. In Handbook of Qualitative Research, ed. by N. Denzin & Y. Guba, pp.
485-99. London: Sage.
Atkinson, P. 1992. Understanding Ethnographic Texts. (Qualitative Research Methods 25).
London: Sage.
Berger, I. 1980. Deities, dynasties and oral tradition: the history and legend of the Abac
wezi. InMiller (1980), pp. 61-81.
Cohen, D.W. 2001. African historians and African voices: Bethwell Alan Ogot and the
changing authority of the 'African Voice.' In African Historians and African Voices,
ed. by E.S. Atieno-Odhiambo, pp. 47-55. Basel: P. Schettwein.
Cohen, D.W. & E.S. Atieno Odhiambo. 1989. Siaya. The Historical Anthropology of an
African Landscape. Oxford: J. Currey.
-. 1992. Burying SM: the Politics of Knowledge and the Sociology of Power inAfrica.
Oxford: J. Currey.
Crazzolara, J.P. 1954. The Lwoo. Verona: Editrice Nigrizia.
-. 1961. Lwoo migrations. Uganda Journal 25 (2): 136-48.
DeWaal, A. 1994. The genocidal state, Hutu extremism and the origins of the 'final sol
ution' in Rwanda. Times Literary Supplement (July 1).
Eltringham, N. 2004. Accounting for Horror: Post-Genocide Debates in Rwanda. London:
Pluto.
Emerson, R., R. Retz & L. Shaw. 2001. Participant observation and fieldnotes. InHandbook
of Ethnography, ed by R. Atkinson, A. Coffey, S. Delamont & L. Lofland, pp. 352-368.
New York: Sage.
Evans, R.J. 1997. In Defence of History. London: Granta.
Evans-Pritchard, E.E. 1965a. Marriage customs of the Luo of Kenya. In The Position of
Women and other Essays, pp. 228-244. London: Faber & Faber.
-. 1965b. Luo tribes and clans. In The Position of Women and other Essays, pp. 205
227. London: Faber & Faber.
Evans-Pritchard, E.E. & M. Fortes. 1940. African Political Systems. London: International
African Institute & Oxford University Press.
Fardon, R. (ed.). 1990. Localizing Strategies. Regional Traditions of Ethnographic Writing.
Edinburgh: Scottish Academic Press; Washington D.C: The Smithsonian Institute.
Faubion, J. 1993. History in Anthropology. Annual Review of Anthropology 22: 35-54.
Furniss, G. & L. Gunner (eds.). 1995. Power, marginality and African oral literature.

Cambridge: University Press.

Goody, J. 1995. The Expansive Moment. Anthropology in Britain and Africa, 1918-1970.
Cambridge: Cambridge University Press.
Henson, H. 1974. British Social Anthropologists and Language. Oxford: Clarendon Press.
Heron, G.A. 1976. The Poetry of Okot p'Bit ek. London: Heinemann.
Kuper, A. 1993. Anthropology and Anthropologists. The Modern British School. London:
Routledge.
Lentz, C 1995. 'Tribalism' and ethnicity inAfrica. Cahiers Science Humaine 31 (2): 303
28.
Lindfors, B. 1980. Mazumgumzo. Athens, Ohio: Ohio University Centre for International
Studies, Africa Program.
MacGaffey, W. 1978. African history, anthropology, and the rationality of the natives.
History in Africa 5: 101-21.
Marcus, G. & D. Cushman. 1982. Ethnographies as texts. Annual Review of Anthropology
11:25-69.
Miller, J. 1980. The African Past Speaks. Essays on Oral Tradition and History.
Folkestone: Wm. Dawson & Sons.

Ogot, B. 1967. History of the Southern Luo. Nairobi: East African Publishing House.
Okot p'Bitek, J. 1971 [1979]. Religion of the Central Luo. Nairobi: Kenya Literature
Bureau.
Oliver, R. 1959. Ancient capital sites of Ankole. Uganda Journal 23: 51-63.
 Who are the Luo? 87

-. 1963. Discernible developments in the interior c. 1500-1840. In History of East

Africa, ed. by R. Oliver & G. Matthew, vol. 1, pp. 169-210. Oxford: Clarendon Press.
Owusu, M. 1978. Ethnography of Africa: the usefulness of the useless. American Anthro
pologist 80: 310-34.
Parkin, D. 1990. Eastern Africa: the view from the office and the voice from the field. In
Fardon (1990), pp. 182-203.
Shetler, J. 2002. The politics of publishing oral sources from theMara region of Tanzania.
History inAfrica 29: 413-26.
Southall, A. 1954. Alur Society. Cambridge: Heffer & Sons Ltd.
-. 1970. The illusion of tribe. Journal of African and Asian Studies 5: 28-50.
Spear, T. 1981. Kenya's Past. An Introduction toHistorical Method. London: Longmans.
Speke, J.H. 1863. The Journal of the Discovery of the Source of theNile. London: Denton.
Stamp, P. 1991. Burying Otieno: the politics of gender and ethnicity in Kenya. Journal of
Women in Culture and Society 16 (4): 808-45.
Taylor, C. 1999. Sacrifice as Terror. The Rwandan Genocide of 1994. Oxford: Berg.
Thornton, R. 1983. Narrative ethnography in Africa, 1850-1920: the creation and capture
of an appropriate domain for anthropology. Man (N.S.) 18: 502-20.
Urry, J. 1972. 'Notes and queries inAnthropology' and the development of field methods in
British anthropology, 1870-1920. Proceedings of the Royal Anthropological Institute
(1972): 45-57.
Vail, L. & L. White. 7997. Power and the Praise Poem. Southern African Voices inHistory.
London: J. Currey.
Vansina, J. 1965 [1961]. Oral Tradition: A Study in Historical Methodology. London:
Routledge, Kegan & Paul.
Waller, R. 1999. Pastoral poverty in historical perspective. In The Poor are not us: Poverty
and Pastoralism, ed. by D. Anderson & V. Broche-Due, pp. 20-49. Oxford: J. Currey.
Wolf, E. 1994. Perilous Ideas. Race, Culture, People. Current Anthropology 35 (1): 1-12.