1Luigi Luca Cavalli-Sforza, Paolo Menozzi, and Alberto Piazza’s (1994) in many ways excellent book, The History and Geography of Human Genes, list the West-African ethnic group FulBe (sing. Pullo) by two ethnonyms as if there were two separate groups. One ethnonym derives from the English version of FulBe’s name for themselves (Fulani), the other from the French (Peulh) (MacEachern 2000). This unfortunate confusion not only marks the danger of associating biologically bounded entities with ethnic labels but also reflects the difficulties of scholars in organizing human conceptualizations of diversity into neat categories. This listing of FulBe as two separate groups1 is particularly interesting because it involves people who have long been imagined in Western thought various kinds of classifications. Imagination has in fact become one of the dominant tropes of current scholarly discussions. Over the past thirty years, scholarly conceptions of ethnicity as enclosed, static and monolithic have increasingly been replaced with conceptions of ethnic identities as fluid, flexible and contextual (see review in de Heusch 2000)—“imaginary”, to use Benedict Anderson’s phrase (1983)—indicating that they are dynamic and negotiated in various political and historical contexts. Scholars have in relation to studies on ethnicity and nationalism debated to what extent nationalities are recent phenomena—as is demonstrated in arguments between Ernest Gellner and his student Anthony D. Smith (1998). Within the various positions that seek to deconstruct ethnicity and nationalism, several recent studies have emphasized that present configurations of ethnic identities cannot be separated from colonial constructions, some claiming that particular ethnicities were “created” in colonial times (Amselle 1998; Salmone 1985).
2This discussion will focus on these issues through an example of the pastoral nomad group WoDaaBe2 in Niger, basing on ethnographic fieldwork among WoDaaBe and analysis of colonial texts. I emphasize the conceptualization of identifications of selfhood and otherness—of boundaries—in time and space, stressing that ethnicity exists in connection with other dimensions of identity. During my fieldwork in Niger from August 1996 until June 1998, I was surprised to observe WoDaaBe identification of themselves as a distinct ethnic group, because most scholarly references conceptualize them as a part of the FulBe. I was, furthermore, struck by the array of identifications, the shifting categories of sameness and difference at display in diverse situations and how people manipulated and negotiated their multiple categorizations in real life situations. Feminist theories have stressed the importance of emphasizing identity as multiple, contextual and fluid (Moore 1994) an emphasis that is still quite invisible in most theorizing of ethnicity and nationalism. My goals here can be simplified as twofold. To point out that colonial and postcolonial classifications of “others” are not necessarily coherent and can rely on contradictory and confusing ways of classifying. As my example shows, classifications are not autonomous but can be based on various actors such as informants, creating a much more complex scenario than that classifications or identities are being imposed by “outsiders”, or an elite segment of society existing outside social relations as some theorists seem to imply. My goal is also to stress—following feminist theories previously mentioned—that ethnic identifications are interwoven with various other sources of identifications and that in real life situations individuals manipulate and identify with others in a shifting ways, ethnicity having only composed one such way.
3The discussion analyzes notions of borders and ethnic identity as expressed in different texts’ classifications of WoDaaBe and then use the ethnographic information to show how the people themselves—in this case, WoDaaBe—use cosmologies and classifications of self and others. By using ethnographic information, I try to avoid emphasizing “system at the expense of agency” (Herzfeld 2001: 221), and stress that identifications of border and selfhood are dependent on contexts in space and time. Sarah Radcliffe and Sallie Westwood (1996) have, furthermore, pointed out that studies on nationalism as a lived and practice phenomena have been relatively few, the analysis of nationalism being mostly restricted to textual analysis. The discussion benefits from criticism on the culture concept and ethnicity, seeing boundaries as an “ongoing process, politically contested and historically unfinished” (Clifford 1988: 9), while simultaneously emphasizing agency and various forms of identification of selfhood. Although I stress fluidity and the politically and historically constituted nature of boundaries, I question claims of boundaries as imposed and created by outside forces. Robin Cohen (1999) has pointed out that the primordialist’ position is in it self to some extent socially constructed because probably very few scholars, if any, maintaining the position that national identities or ethnicities have gone untransformed through centuries. However, as stressed by Terence Ranger (1999), African societies have often been used as examples par excellence of primal ethnicities, as expressed in the usage of the term “tribe” in relation to Africa, thus making it important to write against essentializing notions of primordial ethnicities. In my approach here, I do to some extent agree with Anthony Smith’s position that we have to take into account “earlier ethnic ties and memories” (Smith 1999: 40) but stress simultaneously that these ethnic memories and ideas of selfhood themselves gain different meaning in different context, furthermore, as I warn against over-prioritizing ethnicity as isolated from other identifications. It is in fact somewhat surprising with theories of nationalism and to some extent theories of ethnicity, how these phenomena have often been theorized without reference to other dimensions of human subjectivities.
4The discussion begins by looking briefly at early twentieth-century classifications of FulBe, paying attention to the classification of WoDaaBe in these references, while also consulting more recent texts regarding the relationship of WoDaaBe to other FulBe groups. The classifications of WoDaaBe in different texts have been far from consistent, typological thinking has been dominant, stressing involvement or non-involvement in pastoralism. Ethnographic classifications of WoDaaBe appear, as the example of FulBe above shows, to reflect the problem of boxing diversity, turning “names into things”, as phrased so eloquently by Eric Wolf (1982). Studies of nomads, as argued by Neville Dyson-Hudson (1980), have tended to essentialize their societies, even though more recent studies focus more on pastoralists’ relations with the outside world and the broader economy (Fratkin 1997). The classifications of boundaries between FulBe and WoDaaBe are particularly important because contemporary WoDaaBe claim not to be part of the FulBe while historical and scholarly references claim otherwise. These classifications demonstrate that characterization of the boundary between FulBe and WoDaaBe are inconsistent, and frequently based on essentialist and reifying categorizations. These imposed ethnic classifications during the colonial period demonstrate how those colonial classifying WoDaaBe based in some cases on pre-existing categories thus refuting that WoDaaBe ethnic identity was just “invented by colonial” authorities. The focus then turns toward WoDaaBe conceptualizations of their own ethnicity and relationship with the FulBe, both as expressed in their origin myths and by using ethnographic data collected in the Tchin-Tabaraden area in Niger. The ethnographic study shows that the ways in which WoDaaBe characterize their boundaries with FulBe today are contextual, further underlining the importance of looking at other dimensions of identity which interact with ethnicity, asserting that even though ethnicity constitutes an important part of identity, different kinds of boundaries are emphasized in different contexts.
5The German explorer Paul Staudinger’s 1889 list of FulBe ethnonyms reflects to some extent the various names that have been used in texts to refer to this ethnic group: “Fulde, Fulbe, Fulla, Fullan, Fellani, Fillini, Fullani, Fillani, Fellata, Futa Jalon, Futa Toro and Pullo” (Staudinger 1990: 46)3. These different names have in some instances been seen as constituting different groups, as in the case of Cavalli-Sforza, Menozzi, and Piazza’s distinction between Fulani and Peul. WoDaaBe are, as previously discussed, generally classified as a sub-group of FulBe in much scholarly texts. The term WoDaaBe does not usually appear in older texts, they presumably being included in the more mobile section of the FulBe and as my discussion goes into later, the boundaries of FulBe and WoDaaBe have been somewhat incoherent and unclear. To understand WoDaaBe definitions and characterizations as an ethnic group it is important to explore briefly how the FulBe in general have been characterized in different texts.
6Continuing briefly with the various terms used for these ethnic groups, it can be pointed out that the term Fulani derives from the concept FulBe (the plural form) and the French Peulh from the same concept in singular (Pullo). Fulani and Peul are thus ethnonyms given to the same group of people by speakers of two different languages, English and French. These terms were, however, not coined by those nationalities: the concept Peul is the Wolof term for the FulBe, and the concept Fulani a Hausa term (McLaughlin 1995: 159). German writers have used Foulbe, while English writers in Gambia and Sierra Leone have generally adopted the term Fula(s), which is a Mande term for the FulBe (Hagberg 2000: 160; Delcalo 1997: 144). This indicates that labeling of FulBe by colonial authorities was influenced by neighboring communities. The term Fellata has also been used, but at present Fellata refers to FulBe immigrants in Sudan, whose migration seems to have started in the early sixteenth century. The term Fellata derives from Pulata, which is the Kanuri name for FulBe, and is, according to Shun’ya Hino, often used in Sudan to refer to all immigrants from West Africa (Hino 1993: 73)4.
7In focusing on the different terms given to WoDaaBe and FulBe in various texts, it has to be noted how many of these terms derive from other languages. The confusion regarding the labeling of FulBe is thus to some extent derived from the uncritical adoption of these various names in texts. The ethnonyms used in different languages for the same group are, in some cases, assumed by the authors of the texts to constitute a new group altogether. Several examples support this. One document from the colonial time classifies FulBe as Peulh Farfarous and Peulh Bororo (as: Tahoua Archives n.d.: 5). According to my WoDaaBe consultants, Farfarou is simply a Tamaseck (the language of Tuaregs) term for FulBe. Thus, according to these informants, Peulh Farfarou is a repetition of the term FulBe in two different languages. Another text classifies the FulBe into Peulh and Bororo Dassawa (Brouin 1944: 25), combining Bororo with the Hausa ethnonym for WoDaaBe (Dassawa). Marguerite Dupire (1962: 38) division of the FulBe into three groups uses the terms Fulbe, Fulbe Farfaru, Ndowi’en5, and Bororo, but as noted above, Farfaru is a Tuareg term for the FulBe while Ndowi’en is a WoDaaBe term for the FulBe6.
8
Table 1. — Terms used for WoDaaBe and FulBe in Niger7
Ethnic Group Using the Term
|
Term Applied to WoDaaBe
|
Term Applied to FulBe
|
French
|
Bororoji
|
Peulh/Peul
|
English
|
WoDaaBe/Bororo
|
Fulani
|
Hausa
|
Dossa’a
|
Fulani
|
FulBe
|
WoDaaBe
|
Fulbe
|
WoDaaBe
|
WoDaaBe
|
Ndowi’en
|
Tuareg
|
Folan
|
Farfaru/Farfaruutan
|
Zerma
|
Fuula
|
Fuula
|
9
10It is possible that some authors adopted Hausa or Tamaseck terms because their consultants were from those ethnic groups. Marie B. Perinbam’s discussion indicates (1997: 115), however, that the French generally recruited local hires from those who possessed extensive language and writing skills, and that due to an emphasis on literacy in Islamic communities many of the local hires were from the literate Islamic and elite communities of which the FulBe were often a part. It is thus interesting from this perspective that so many of the ethnonyms used for the FulBe in texts derive from other languages, though this could to some extent be due to elite FulBe communities sometimes adopting languages of neighboring groups and being generally competent in many languages8.
11FulBe are generally divided in the texts into FulBe si’ire (also FulBe gidda)9 and FulBe na’i (sometimes translated as “town Fulani” and “cattle Fulani”), or into Fulani and Bororo10, Bororo and FulBe na’i being identified as the more nomadic section of the FulBe (Alexander 1908; Morel 1911; Burns 1929; Wilson-Haffanden 1930; Nadel 1942; de St. Croix 1945; Urvoy 1949). In present day Niger, Bororo is used to describe WoDaaBe in the Hausa language and in French, but is generally not used for FulBe groups (Burnham 1996: 12).
12Many texts from the early twentieth century racialize the distinction between FulBe and the more nomadic groups, viewing Bororo as constituting a purer racial category (e.g., Alexander 1908; Viellard 1932). Francis Nicolas (1950: 42), for example, divides the FulBe into “peul blanc” and “noir”, the latter identified as Bororodji. Urvoy (1949: 96) emphasizes a distinction between FulBe more or less fixed and frequently living in villages, and Bororo, whom he regards as preserving their pure blood and the morals of the ancestors, and being less Islamicized. Wilson-Haffenden (1930: 92), classifying FulBe into Town or Settled Fulani and Cow or Bororo Fulani, characterizes the latter as keeping their “blood pure by marrying only within their tribe and retain[ing] for the most part their pagan beliefs”. The term WoDaaBe is, as previously stated, used in very few references, with the exception of Brackenbury and Reed, who both give a rather detailed classification of the WoDaaBe lineage organization. Brackenbury (1924: 210) divides the FulBe into Jafumen and Fulbe, classifying WoDaaBe under the former name but stating that “they are not recognized as true Jafumen”. He further explains that the Jafumen see the term Bororo as insulting and apply it “to a clan of FulBe in Bornu called Bodado in the singular and Wodabe in the plural” (ibid.: 209). Reed (1932) similarly classifies WoDaaBe as a one of seven major groups of Fulani. From these sources it can be inferred that the older references generally use Bororo as a broad term referring to the more nomadic section of the FulBe, including WoDaaBe in that category. The term Bororo11 is thus used to signify a much greater variety of groups than is the term WoDaaBe.
13Lugard (1904): Writes interchangeably about “Fulah” (22), “Fulain” (21), “Fulani” (81), “Foulahs” (375). Mentions, however, that, “there seems to have been always a distinction between the purely pastoral shepherd, or Cow Fulani [. . .] and the aristocratic or ruling Fulani” (383).
14Alexander (1908): Uses Cow/Bush Fulani, terms seen as referring to the “purest representative of the old race” (193) as well as stating that they are never found in cities (194).
15Kisch (1910): Uses the terms “bush” and “cattle Fulani” (as referring to the same people) presumably to distinguish them from other Fulani (139, 149).
16Morel (1911): Uses the terms M··bororoji and Cow Fulani as referring to the WoDaaBe does not specify how they are different from Fulani generally (119).
17Falconer (1911): Gives a vague separation of bush-Fulani, who are of “purer blood” than the Fulani of the towns (139).
18Temple (1922): “Filani” are classified into two main groups: Cattle Filane, or Borroroje, who have retained the Semitic physical characteristics, and Filanen Gidda, the product of a mixture of the original Semitic immigrants and various tribes (397).
19Brackenbury (1924): Argues that Bororo is not a self-designation, but is applied by town-Fulani and Hausa.
20Wilson-Haffanden (1927): Uses the term Borroro, seeing Borroro as a Sub-group of Fulani.
21Burns (1929): Separates the Fulani into Fulani Gidda and Cow Fulani (53). Cow Fulani refers also to Borroroje, and is seen as referring to the purest type, and retaining of pastoral habits (53).
22Wilson Haffanden (1930): Classifies Fulani into town or settled Fulani on one hand and cow or Bororo Fulani on the other. The latter are characterized as keeping their “blood pure by marrying only within their tribe and retain for the most part their pagan beliefs” (92).
23Viellard (1932): Separates WoDaaBe from Fulani, does not define Fulani but defines WoDaaBe as “nomades de language peul, pasteurs de brousse qui ont peut-être mieux conservé le type physique et moral du Peul” (92).
24Reed (1932): Classifies WodaaBe as a part of the Fulani, but divides Fulani into seven major groups (WoDaaBe being one of these). Explains that Fulani are known by WoDaaBe as Ndowi’en.
25Nadel (1942): Uses the separation of town Fulani from nomadic cattle people, who are known as Bororo (71).
26Urvoy (1942): Lists the names Foulas, foulbés, fellatas, along with Peuls (29). Bororodjis are mentioned as heathen nomads (29).
27de St. Croix (1945): Uses the distinctions Cow Fulani and Bush Fulani to apply to nomadic Fulani (10). Uses the term WoDaaBe without defining it.
28Urvoy (1949): Distinguishes between Fulani, who are more sedentary, and Bororo, who have purer blood and greater attachment to their traditional morals. The former are described as Muslims, the latter as heathens (96).
29Tylor, F. W. (1953): Uses Mbororo’en to refer to “the best-known clan of the ‘Cow Fulani’”(xi).
30Kirk-Greene (1958): Talks about Mbororo’en as a clan within the Fulani (22).
31Hoben (1958): Bororo are seen as a separate group within the Fulbe na’i, which does not refer to a specific clan or tribe but a general category of pastoralists, having different values and customs (1-2).
32Lhote (1958): Bororo is seen as referring to the more nomadic section of Peuls.
33Séré de Rivières (1965): Peul seen as referring to two ethnic groups, Farafarou (or Foulbé) and Bororo (or Bodaado, in plural Wodaabé) (54).
34Last (1967): Distinguishes between Torokawa, Sullebawa, and cattle Fulani.
35Even though the individuals quoted above had probably many scholarly goals, modern day scholarly research of FulBe is usually seen as starting with ethnographic studies conducted by Derrick Stenning (1957), C. E. Hoben (1958), and Marguerite Dupire (1962). These more scholarly classifications of FulBe continued to stress the distinction between nomadic and sedentary. Marguerite Dupire (1962: 38) divides the FulBe into three groups: sedentary Fulbe, Fulbe Farfaru, which she states are also called Ndowi’en, and finally the nomadic Bororo. Stenning (1957: 57) categorizes FulBe into four major groups: the ruling dynasty, settled Fulani, semi-sedentary Fulani and finally pastoral Fulani, which he points out are often called Bororo in the literature. He makes it clear that WoDaaBe are a part of this fourth group. His classification thus corresponds to Dupire’s with the exception that he distinguishes between the more elite sedentary FulBe and the general sedentary FulBe population. What is interesting about Stenning’s classification is, however, that contextualizes his information by stating that it applies to FulBe in Nigeria, thus implying that within other geographical areas another set of categorization could be used. A. S. Johnston (1967) classifies Bororo and FulBe na’i as separate groups, stating that the division of FulBe into two groups is an oversimplification and that in fact there are at least four. These are the “true” nomads, Bororo’en; the semi-sedentary pastoralists called FulBe na’i; FulBe belonging to the ruling class, TorooBe; and finally, pastoralists who have lost their cattle, called FulBe siire. His classification is obviously quite similar to Stenning’s.
36The review of these sources indicates that FulBe ethnonyms have not been particularly coherent, the same concepts being often used in different languages in attempts to distinguish between FulBe groups. Colonial and postcolonial classifications of “others” can thus base in contradictory and confusing ways of classifying. Colonial classifications (which have to great extent be continued by scholars) did, furthermore, privilege a distinction into more-or less nomadic group, including WoDaaBe as a more mobile section of the FulBe.
37Prior to turning to discussion of contemporary WoDaaBe ways of identifying themselves, I explore this issue of colonial classifications and ethnic identifications a little further by referring a more detailed example of a lineage group of WoDaaBe. In my dissertation research, I paid special attention to the classification of WoDaaBe in the Tahoua district of Niger, where a part of my ethnographic study took place. It is interesting to observe briefly how this group of WoDaaBe was classified in older texts, especially considering that the ethnographic information upon which I base the following discussion comes from present-day WoDaaBe in the same area. FulBe and WoDaaBe have only been a small minority in this part of the country. In 1907, Lt. Leignot claimed that 2000 FulBe were to be found in the Tahoua region, comprising approximately 2% of the total population. However, as Fuglestad points out, the nomadic FulBe in Tahoua were greatly underrepresented in all such data (Fuglestad 1983: 84). One archival document states that no census of the FulBe had been conducted in Niger prior to 1927 (Guilland 1933: 2).
38Texts from colonial times in Tahoua are scarce on the subject of FulBe and WoDaaBe. Some official colonial documents stored in Niger’s archives12 show a clear ethnographic interest in the FulBe, but express, interestingly enough, no attempt to classify FulBe groups in the area in any detail. In other documents, a distinction is simply made between sedentary and nomadic FulBe, just as in the published sources previously cited. Notably, the lineage organization of another ethnic group, the Tuaregs, in contrast, is carefully discussed and elaborated on in these texts13. The lack of classification of FulBe groups, including the rather vague identification of WoDaaBe as a group, indicates a lack of interest in the lineage organization of these groups and/or a lack of knowledge of their internal differences. Classifications of WoDaaBe and FulBe in Tahoua were probably of minor importance to the politics of the Niger territory, for example, in comparison to the Tuaregs. An archival document from 1936 has, for example, detailed ethnographic information about the social structure and customs of the Tuaregs but no distinction of FulBe groups beyond dividing them into sedentary and nomadic (Anonymous 1936). In another colonial document, M. Brouin distinguishes between two “races” of FulBe: “The Fulani in this country seem to belong to two different races, the white Muslim Fulani, and the Bororo of the Dossawa race that are non Muslims, who migrate and have a different way of life than the former”14 (as: Brouin 1944; as: Genderault 1957: 5). Brouin careful phrasing indicates that he is not certain of his information. Another document, listing people paying taxes in Tahoua, carefully identifies Tuaregs according to lineage affiliation, while the classification of FulBe is inconsistent, in some cases simply associating them with the place where they were found and in others classifying them by lineage name (Durand-Veil 1947). The list, according to my WoDaaBe informants, blends together various categories of identification, sometimes actually using the label of a lineage group, but at other times referencing a body of water, a well, or a town where a group probably stayed. The text does not identify what kind of category is being referred to, but presents all of these different classification categories as lineage names. Even though there was a little emphasis on classifying FulBe in any details in the department of Tahoua—probably due to their minimal political influence at that time—and WoDaaBe thus being considered as a part of the FulBe in official documents, that did still not make them emerge into the FulBe ethnic group. My informants tell me that the WoDaaBe in Tahoua had tried to gain political reorganization as a separate ethnic group for a long time, which is supported by an archival document from 1957 (Genderault 1957). In this archival document, the author states that WoDaaBe request for their own lamido (head-chief who is in charge of tax collection and other governmental affairs) should be granted instead of continuing recruiting a FulBe chief that represents them to the government. The author suggests that WoDaaBe did not receive a separate lamido earlier because the French failed to see that they were dealing with two different groups (1957: 5), thus indicating that WoDaaBe had been trying for some time to get their lamido accepted by the colonial government. Genderault recommendations were, however, not followed and it was not until 1973 that Bayre bi Tuka’e elected as the chief over the “9e groupment” (i.e. the WoDaaBe of Tahoua), the state thus formally acknowledging WoDaaBe as a separate group (Diverse documents).
39Analyzing WoDaaBe own ways of classifying themselves and create boundaries, I make use of statements by contemporary WoDaaBe and also of origin stories, both those recited today and those documented in the past, but WoDaaBe origin stories frequently evolve round explaining difference between them and the FulBe. Stories of origin are intrinsically contextual, as Richard Henderson (1972: 76) points out in his discussion of the Onitsha Ibo; people emphasize different versions of their history according to what elements they find meaningful to stress. WoDaaBe stories of origin can be seen as belonging to two categories: mythical-historical, aiming to understand the origin of WoDaaBe as an ethnic group (thus focusing on important ethnic identity markers such as boundaries to other groups and the acquisition of cattle, an integral marker of their ethnicity); and internal origin stories, aiming to explain the lineage division of WoDaaBe society and thus the historical and hierarchical relations of different lineage groups.
40The mythical-historical origin stories are quite interesting for the purpose here because many such narratives address the boundaries of WoDaaBe/Bororo and FulBe. Brackenbury (1924: 211) asserts that origin stories either claim that the Bororo descended from the incestuous marriage of a FulBe woman and her son or that Bororo descended from a woman of Arab descent called Bajomagu. Around the same time, Wilson-Haffanden (1927: 276) collected a narrative claiming that a woman had given birth to a daughter and a son who spoke a new language, Fulfulde, and then later had two additional children with a water spirit, who became the ancestors of the Bororo. M. Dupire (1962: 34) argues that one of the major recurring themes in WoDaaBe origin stories is the association of the origin of WoDaaBe and their cattle with a water spirit, as well as the condemnation or rejection of the Bororo. She retells an origin story involving the birth of two children (from an incestuous relationship), one becoming the ancestor of FulBe, the other of WoDaaBe. De St. Croix (1945: 8) retells a story that addresses the origin of the nomadic FulBe15, telling of a FulBe child becoming lost, growing up in the bush, and later acquiring cattle from a water spirit. In a more recent study, Bonfiglioli (1981: 114) presents an origin story containing some themes similar to those reported above, wherein the separate ancestry of FulBe and WoDaaBe can be traced to two siblings. A. G. Adebayo (1991: 5) points out that the more sedentary branch of the FulBe emphasizes different kinds of stories of origin than those discussed here, focusing largely on Anabi Ukuba (also called Okba or Ykubu), a disciple of Prophet Muhammad.
41I find it noteworthy that so many stories of origin collected at the beginning of the twentieth century specifically sought to distinguish the origin of the more nomadic Bororo/WoDaaBe from FulBe. It is of course impossible to determine whether these story tellers would have identified themselves as WoDaaBe (except in Dupire’s research where as her informants identified themselves by the term WoDaaBe). Nonetheless, it is clear that the more nomadic branch of FulBe has understood its origin as somehow different from the origin of other FulBe, using origin stories to explain in a meaningful way how that difference originated. In contrast, an origin story told to me by a WoDaaBe man of high authority within WoDaaBe society was relatively similar to the stories of origin discussed previously (involving two children and cows arriving from water), but did not mention the FulBe at all. This indicates a minimal need to characterize WoDaaBe separation from FulBe or even to explain the similarities in the social structure and culture of WoDaaBe and FulBe (Loftsdóttir 2001b).
42Interestingly, the origin stories relating to the kinship structure of WoDaaBe have no relationship to the origin stories previously discussed. In these stories, two brothers are identified as the ancestors of all the different WoDaaBe lineage groups, a theme that serves to highlight unity within the WoDaaBe while emphasizing differentiation between different lineage groups identifying themselves as WoDaaBe. The relationship between the two maximum lineages is an ambiguous one, as has been discussed elsewhere, characterized by competition and hostility (Loftsdóttir 2000). Within WoDaaBe society, seniority is quite important in terms of authority, and this is clearly emphasized in the lineage system. Interestingly, within the narratives of origin of the lineage groups and structure, there is no mention of FulBe or of the relationship between WoDaaBe and FulBe. Belonging to a lineage group is crucial for contemporary WoDaaBe sense of identity, and is in some ways inherent in their lineage structure, breaking down into opposite segmentary units. That indicates that lineage affiliation has been quite important for quite some time.
43If asked to identify themselves most contemporary WoDaaBe would probably use the term WoDaaBe and then refer to their lineage affiliation. WoDaaBe agnatic lineage system is the basis for their political organization, land use, household composition and marriage patterns and thus can be seen as holding a crucial role in their identity. Similarly to E. E. Evans-Pritchard’s (1940) model of segmentary lineage organization, WoDaaBe lineage fractures into smaller and smaller segments, each lineage fraction having a specific identity, tracing themselves to a common ancestor16 (Stenning 1957: 58). Individuals see themselves as belonging to every one of these segments, leading to the two brothers previously mentioned.
44The ambiguity reflected in the various colonial texts and origin stories in regard to WoDaaBe—FulBe boundaries is evident in classification of these boundaries by contemporary WoDaaBe. WoDaaBe refer to the FulBe using the term Ndowi’en, and not including themselves in that category, underlining that the WoDaaBe and the FulBe constitute different groups17. One informant said, when asked whether they were not one and the same ethnic group: “Ndowi’en and WoDaaBe are different; their Fulfulde appears perhaps to be the same, but it is not” (Interview September 1997). During one group discussion, a man remembered hearing a story of two men who had a dispute, each going their own way, one becoming the ancestor of WoDaaBe, the other of FulBe. Another man in the group then made a comment which the others seemed to agree with: “This we don’t know, this must have been a very long time ago. That time people were self-aware, things were as they are today” (i.e., WoDaaBe and FulBe were separate groups; Interview September 1997).
45The literature on FulBe has strongly underlined pulaaku as a key concept in FulBe self-identification. The concept pulaaku is derived from the same root as FulBe (sing. Pullo), and according to Riesman (1977: 131) refers to “qualities appropriate to the Fulani” (Zuboko 1993: 202). WoDaaBe generally do not use the concept pulaaku to refer to their own value system. I, for example, seldom heard people use the concept in natural speech. Instead, WoDaaBe use the concept mbodagansi, which is seen as only applying to WoDaaBe, not to FulBe groups. It is possible that the concept mbodagansi, which, like the term WoDaaBe (sing. Bodaado), derives from the root mboda (taboo), refers not to mboda per se so much as to “qualities appropriate to WoDaaBe”, to paraphrase Riesman. Bonfiglioli (1988: 63) writes that the WoDaaBe formed or enforced a notion of difference between WoDaaBe and Ndowi’en only after the nineteenth century jihad, forming their own code of moral conduct, mboDangaaku at that time.
46Another facet reflecting the ambiguous relationship between WoDaaBe and FulBe is the disparate interpretations of the term “WoDaaBe”. WoDaaBe explain their name as signifying “those who respect taboo” or “people of the taboo”. FulBe, however, explain the name to me as “those who are excluded” or “those who are tabooed” (Dupire 1962: 33). The Historical Dictionary of Niger gives two translations of the term WoDaaBe: “People of the taboo” (p. 144) and “the rejected” (p. 72) (Delcalo 1997), mirroring this ambiguity. These two explanations are important because the former tries to establish the WoDaaBe as morally superior to the FulBe, while the latter interprets the FulBe as the morally superior group. These different explanations alone indicate some tension and differentiation between the two groups. Bonfiglioli’s discussion on mbodagansi as forming only after the jihad seems to suggest that prior to the jihad, WoDaaBe did not regard themselves as a distinct group, which partly explains why it is hard or even impossible to trace their history separately in older texts. In addition, Dupire (1962: 285) claims that the WoDaaBe narrative of their lineages originating from two brothers is not backed by any historical evidences. Furthermore, most written references published in the early twentieth century use a vague reference to more nomadic and less nomadic FulBe when categorizing FulBe ethnicity. That could either indicate that the WoDaaBe had not then developed a strong separate identity as they have today, or simply reflect a lack of interest on the part of these authors in differentiating the various FulBe groups.
47I will now focus more on ethnicity as a lived, acted-up-on phenomenon, stressing that ethnicity exists in conjunction with other aspects of individual identity, such as gender, age, status and religion. Furthermore, it is necessary to ground ethnicity as simply one of many ways in which societies have historically socialized and explained difference. When ethnicity is regarded as part of a general identity formation, it becomes almost self-evident that ethnicity, like other aspects of people’s identities, is not emphasized in all social situations (i.e., belonging in other kinds of categories can be important as well). When ethnicity is understood as only one aspect of people’s identities, it can be renegotiated and understood differently at different times and circumstances. It can perhaps be said that ethnicities are fluid because the individuals embodying them have complex and unique life histories.
48The relationship between WoDaaBe and FulBe, and WoDaaBe ethnic boundaries, cannot be determined merely from labels or from visual symbols such as clothing (which have not been explored here), or from individuals’ statements regarding their own views. It must also be considered in the context of experience, of the practices of individuals in their everyday settings. The relationship between FulBe and WoDaaBe is not clear-cut in such a context, but rather ambiguous and contextual, meaning that in certain social situations WoDaaBe tend to minimize the difference they perceive between themselves and the FulBe. The emphasis placed on the connection between FulBe and WoDaaBe depends to some extent on the context in which the discourse or interaction takes place. Even though as previously shown WoDaaBe emphasized in formal interviews and within a general dialogue that they and the FulBe constituted separate ethnic groups, they would state precisely the opposite in certain situations. On several occasions, WoDaaBe explained to me that FulBe and WoDaaBe were really the same—some of the same individuals having previously stated firmly that WoDaaBe and FulBe were not the same ethnic group. To give just one example of many, when coming out of the petit marché in Niamey, I gave money to two Jelgobe Fulani women who were begging. When the women had passed us, my WoDaaBe companions informed me, almost as if excusing the women, that I did a good thing by giving them alms; these women were after all FulBe, being the relatives (bandiraaBe) of WoDaaBe. Even though the term bandiraaBe to some extent has fluid meaning in Fulfulde, this phrasing was not used to describe other ethnic groups, and in this context, it seemed to imply an affinity between FulBe and WoDaaBe.
49This fluidity in categorization is not only restricted to FulBe—WoDaaBe boundaries, but a more globalized world bringing people into various social networks and groupings have made WoDaaBe identify with a larger social group than FulBe, such as community of Islam, Nigeriens and Africans. References have tended to minimize the importance of Islam to WoDaaBe, and while it is correct that many WoDaaBe do not follow all ritualistic aspects of Islam as closely as many other Muslims, it must be emphasized that the definition of a “Muslim” was one of the politically contested issues during the nineteenth century jihad (Loftsdóttir 2000: 431). In any case, WoDaaBe today define themselves as Muslims, and the religious practices of many individuals have become more structured, probably due to an increased involvement with urban centers and thus larger religious communities. I heard WoDaaBe men identifying themselves in various contexts as belonging to the Islamic community, and creating borders and community on that basis. Observing men of different economic class and ethnicity, standing side by side, praying in the multi-ethnical Niger gives a striking image of how religion can create cohesion and unity. This brotherhood of Islam becomes much more significant in an urban context, where individuals become part of a larger community of Muslims, than in the bush where religion becomes more a private matter.
50In a similar fashion, WoDaaBe (probably increasingly) see themselves as Africans; a unity that often exists as a counter-identification to Euro-Americans. Thus, even though in certain situations WoDaaBe would draw a sharp line between themselves and other ethnicities, the same individuals who had stated that WoDaaBe were completely different from their neighboring groups stressed to me on other occasions that assistance and kindness from individuals from other ethnicities was due to the fact “that he was an African like me”. This identification overlaps with belonging to Islam, and, similarly, exists primarily in an urban context. Finally, WoDaaBe increasingly see themselves as subjects of the nation state—especially younger WoDaaBe males. Younger WoDaaBe men often situated their claims to resources such as land, and entitlement to social services, in the context of their status as subjects of the Nigerien nation state, having “nationalité nigérienne”. Giving a more specific example, WoDaaBe males referred to their right as nationals of Niger to utilize the land, while criticizing the State’s possession of valuable land for state run ranches. WoDaaBe increased engagement with globalized world, also contributes toward new and shifting identities. A number of WoDaaBe travel to northern European countries in order to sell artisanry objects, staying there for few months each time. I was surprised when I meat there with some of the WoDaaBe who I had befriended in Niger, to hear them refer to a Hausa man as “one of my people”, the phrase referring to people of Nigerien nationality in this context. The affinity they felt with other Africans vice verse Europeans was also obvious in much more generalized sense, expressed in bold statements of the nature of Europeans in contrast with similar totalizing references to the nature of Africans. I am not implying that being a part of the Islamic community or subjects of the Nigerien state involves the abandonment of WoDaaBe ethnicity. On the contrary, these are dimensions of identity that can coexist meaningfully with a sense of membership in the WoDaaBe community. As Burnham (1999: 280) has shown, in relation to the FulBe in Cameroon, narratives of Islam were incorporated into ideas of pulaaku among some FulBe groups.
51The ambiguities of ethnicity can also be related to internal differentiation and gender of WoDaaBe. Many internal distinctions exist within the WoDaaBe lineage system, and in certain situations, especially those involving only WoDaaBe, the notion of WoDaaBe unity loses meaning (Loftsdóttir 2000). To give an example, when meeting a group of people from the opposite lineage to the group which I was living, one WoDaaBe friend told me that these were not his people (wana dum duunia amin) stating that they were really a different kind of WoDaaBe, thus drawing boundaries between himself and them. Gender identification is also a very important dimension of identity in WoDaaBe society. Although constructions of manhood and womanhood exist alongside the notion of ethnicity, men and women are still generally seen as inherently different beings, with different roles and capabilities. Thus, after being in Niger for a long while I frequently experienced WoDaaBe women referring to our common “women-ness” in certain social situations, such as by emphasizing that “all men were the same”, or by referring to common experiences relating to our sex, posing us in essentialized way against the WoDaaBe males around us.
52Analysis of FulBe and WoDaaBe ethnonyms reflects the strong emphasis that has been placed on typologies in these classifications, especially the resilient binary typology of herders and agriculturists; involvement in herding has often been the basis for categorization of FulBe. Additionally, as Philip Burnham (1999: 270) has pointed out, recent theorizing on FulBe ethnic identity has been characterized by an impulse to essentialize and reify FulBe ethnicity. Although such emphases do in some respect reflect the typologies that FulBe invoke when characterizing their own ethnicity (ibid.: 271), studies on the FulBe have shown that FulBe ethnicity is in practice quite fluid. While some groups emphasize pastoralism as important to their ethnicity even in cases where they have become fully sedentary (Waldie 1990), other FulBe have emphasized different aspects, not related to pastoralism, in new conditions (Shimada 1993). Although contemporary WoDaaBe strongly emphasize cattle as a part of their ethnicity, my ethnographic study among migrant workers in Niger indicated that even when herders had lost all or most of their animals and stayed in the city for a long time, they continued identifying themselves as WoDaaBe (even as herders) and were identified by others as no less WoDaaBe than those living in the pastoral area (Loftsdóttir 2000, 2001a). This is consistent with a point that has long been made by scholars: that the boundaries between agriculture and pastoralism are fluid and constantly changing (Johnston 1967: 12; Swift 1979: 1-2), making it common for groups to shift from one occupation to another as a response to environmental fluctuations or a changing social-political environment (Horowitz & Little 1987: 72). Studies have, for example, shown that WoDaaBe have historically cultivated during times of difficulty. My point here is that the WoDaaBe do not stress the classification of different groups according to their involvement with the livestock, and possibly such classification has never been important to them. The classification of their own diversity into lineage groups is, on the other hand, more meaningful to them, which is not to dispute that lineage groups have, as discussed by Dupire (1970), often been reconstructed and manipulated, thus undergoing historical modifications. I find a comment by Burnham (1999: 279) extremely valuable in this context:
“Ethnic group identity [. . .] among mobile pastoral FulBe is a function of a complex historical process in which concepts of descent, inter-group marriage relations, the trajectories of pastoral movements (themselves subject to ecological and political opportunities and constraints), and competition for political ascendancy and control of space all interact to ethnically differentiate or merge FulBe pastoral groupings.”
53While the information discussed in this article indicates that FulBe and WoDaaBe ethnic identities have been quite fluid, it does not mean that their current ethnic identifications are merely the result of colonial processes. Jean-Loup Amselle (1998) claims, in his otherwise excellent book Mestizo Logics, that “only with colonial conquest was this ‘chain of societies’ [. . .] disarticulated so as to give birth to distinctive ethnic groups”, taking FulBe as an example of such a colonial creation. Such an instrumentalist view of ethnicity as invented can be quite different from claims that address ethnicity as fluid and contextual, as well as questionable in its location of agency as belonging only to the colonial powers. Fluidity implies that people recycle and make use of pre-existing categories in different ways, in terms of both social situations and history. WoDaaBe ethnicity seems clearly not to be a colonial invention; as my discussion has shown, the colonial authorities adopted names that were already applied to FulBe and WoDaaBe by other ethnic groups. Their confusing classification of FulBe and WoDaaBe can be seen as a reflection of the ambiguity of the relationship between FulBe and WoDaaBe, as well as of the views of their informants and their own biases. Ethnicity’s recent importance in academic studies may be regarded as partly due to its increasing significance as a force of political mobilization in the contemporary world (Appadurai 1998). At the same time, scholars’ theorizing of ethnicity has changed, increasingly grounding the concept in dynamics of history and politics, a position often identified as social constructiveness. As Harri Englund (2001: 91) has noted, studies emphasizing colonialism as a force in the stiffening of ethnicities have, despite some limitations, been important in substituting historicity for primordialism. I have nonetheless warned against emphasizing colonialists as dominant in the creation of WoDaaBe ethnicity, preferring to emphasize WoDaaBe instrumentalism, their choosing and underscoring different aspects in different contexts. Focusing on the ethnic boundaries of WoDaaBe and FulBe, both as characterized in texts and through analysis of their own conceptions of them, shows that these boundaries cannot be reduced to colonial constructions, ethnic identity and identifications being so much more complex than that. As noted in Sten Hagberg and A. B. Tengan’s (2000: 10-11) discussion of ethnicity, boundaries and borders are conceptualized not only on an ethnic level but in various ways, integrating cosmological and geographical notions of space. Even though my discussion has stressed that ethnicity is in real life multi-layered concept that can be played with and negotiated in real life situations, I am not ignoring that identities are simultaneously pre-imposed and used for discrimination purposes by various agents, such as the nation state.
54Amselle (1998) identification of counter-identification as crucial to identity formation, thus echoing structuralistic and post-structuralistic positions on identity and constructions of the self. We still have to be careful not to assume that the counter-identification is always the colonialists, as well as to avoid reductive claims that the colonial system created these ethnicities. One of the important aspects of fieldwork is that we cannot assume beforehand the texture of people’s concerns (Englund & Leach 2000: 236). Prior to the colonial period, WoDaaBe and FulBe counter-identification was strongly based on the Hausa as a point of differentiation (Burnham & Last 1994), relating not to the colonial system but to West African processes of identification and ethnicity politics. As my discussion shows, there can be diverse counter-identifications during the same time period, identity being a contextual phenomenon. We cannot view colonial classifications as being the only factor affecting how people classify themselves, because colonial constructions, for example in relation to the construction of FulBe by other non-FulBe ethnic groups, existed in relation to indigenous constructions.
55Ethnographic examples have shown that people renegotiate their ethnicity when facing new contexts, which is clearly demonstrated by the literature on FulBe. Throughout their history, FulBe have emphasized language, religion or moral qualities, according to what was most appropriate in certain circumstances. That does not necessarily mean that these things are invented, but that people often use pre-existing meanings and symbols, combine them, and understand them differently in new environments. The analysis of FulBe and WoDaaBe ethnonyms reflects, furthermore, the inconsistencies in how these groups are referred to and demonstrates the dominance of typologies in their classification (i.e., involvement in herding often being used as the basis for categorization of FulBe). Even though such typologies are to some extent generated by WoDaaBe (and probably FulBe as well), ethnographic studies have demonstrated that their economies and ethnic identities are flexible, in addition to their identity being contextual, people emphasizing different ways of identifying in different kinds of circumstances.