Kifwebe (pl. bifwebe), a word referring to the striated masks performed by Songye and Luba peoples of the Democratic Republic of Congo, is by now very familiar to scholars, collectors, and dealers of African art. Although these masks populate many public and private collections and have even inspired designers and artists in the West, their indigenous story and time line, spilling over a vast region, is fragmentary. The object of this paper is to review and update what we know—and do not know—about the contextual life of kifwebe masquerade associations. It will deal with aspects relating to the effect, definition, and function of performances of this dynamic, interethnic tradition, mainly in Songye areas and, for comparison, in documented parts of Lubaland. In so doing, a certain amount of repetition of previous findings is necessary to place into perspective new data and interpretations which will build upon my 2012 publication on prototypical bifwebe models, especially in the little-known central Songye area, and on other aspects of performance. But a historically flawless definition and mapping of bifwebe masking is a pipe dream and not even my aim. As Sidney Kasfir has pointed to (in reference to the Ekpe society in southeastern Nigeria and adjoining Cameroon), material, performative, and ritual structures in masking are not fused together but can be used to different ends and can travel different paths (Kasfir 2019: 15). Hence, for example, the same kifwebe maskers that exploited magicoreligious ritual are known to have participated simultaneously in playful, secular events, with performance contexts shaping perceptions of the masks. Also, while masking societies throughout Songye areas1 and among the Luba played upon the gender divide, their representations and power plays were differently performed. Therein lies the malleability of kifwebe masking, the power of its creativity and adaptation to varied sociopolitical circumstances given changing local and national ideologies from the colonial era to modernity.
This material is the result of my field work in the late 1970s in the central Songye region, combined with data collected by ethnomusicologist Alan Merriam in 1959–60 among the Bala Songye; the findings of Mutimanwa Wenga-Mulayi in the Ankoro area of Luba country during his M.A. research in 1973–74; documentation by Karel Plasmans, a Belgian agronomist who worked, collected, and documented Songye material culture from 1955–70; and hitherto unpublished material on masking among the Kalebwe subgroup of the Songye compiled by Cynthia Anson in the summer of 1972 in collaboration with Congolese museum personnel.2 The latter material contributes an important piece to the complex puzzle of kifwebe performances.
What we know about the context of kifwebe masquerades derives from multiple public and private sources with varying objectives, interests, and degrees of specialization. The earliest reports and collections from late nineteenth and early twentieth century missionaries, travelers, and ethnographers provide little insight into these masquerades. The spectacular round, striated Luba mask, now in the Seattle Art Museum, is among the early examples to have drawn much attention and to have been referred to as “Kifwele” in 1913 by the collector Father Pierre Colle (Fig. 1).3 But, apart from noting that this was a dance mask, which he relegated to the “fetish” domain, and that there were two types, a larger one representing a female spirit and a much smaller male one, we learn little of the wider context (Colle 1913: 440, 676–77, ill. 12). Less well known is that Leo Frobenius, a self-made ethnographer collecting at the service of German museums, offered us an example of a striated, hourglass “kifebbe,” more typically central Songye in form but devoid of holes for attachment. In his 1905–1906 notes he referred rather sensationally to this mask as a sort of magical object empowered by human sacrifice and fire and associated with healing and funerary rites (Klein 1990: 107, 108).4 Subsequent official German expeditions, orchestrated by the Berlin ethnographic museum, provide no material evidence of such cultural practices. Most of the astounding carvings collected by Hermann von Wissmann, Paul Pogge, Ludwig Wolf, and others, is figure sculpture. There are apparently other early masks like the piece brought to Belgium by Liévin Vandevelde, allegedly acquired during the period 1885–87, but these lack contextual data (see Neyt 2004: 164).
Though this may seem disappointing, it is significant. Masquerades were performance practices to which Westerners were generally not privy and which were of little interest to them, given their preference for collecting freestanding figurative sculpture in the round or tools, weapons, and prestige objects. Turn-of-the-century collecting by Frobenius, or by Emil Torday for the British Museum in London, tended to take the form of what Johannes Fabian called “stationary ethnography” (1998: 97, 98), with locals bringing artifacts from far afield to the comfortably seated European. Fabian further notes that, “Torday and Frobenius knew that almost everything they collected was embedded in contexts that involved performance and actions,” yet they “wished that all of culture came in the form of ‘curios’” (1998: 98). Apparently Frobenius only saw his first masked dancer after nine months in the field (Fabian 1998: 99).
In view of such priorities and collecting practices, the early picture of bifwebe manifestations is patchy, and reconstruction of a wider history cannot be forced on the basis of vestigial evidence. What we find are culture-specific pockets, time capsules that reflect varying degrees of secularization and/or socioreligious practice. Alan Merriam, in his research among the Bala, estimated that the bifwebe were active among the Bapupa subgroup from 1911–1920, although what people asserted at the time of his research in 1959–60 is that it was “just a game” and not “serious magic” like among the Kalebwe and the Songye east of the Lomami River. On the basis of these declarations, Merriam therefore postulated that masking spread from east to west, losing momentum by the time it reached the Bala. He did, nonetheless, consider that the date of introduction in the village of Lupupa Ngye, where he had resided, was early, which made him question his own interpretation (Merriam 1982: 26). Marc Felix, basing his analysis mainly on visual comparisons, hypothesized that the kifwebe association started its workings in the west of Songye territory around 1880 and then moved eastward (1992: 23, 24). Admittedly, in 1910 the conical masks with painted striations from the Tempa, in the extreme western Songye sector, were the earliest masks to arrive to the Museum for Central Africa in Tervuren (Fig. 2), but these are very different and identified as “ruadi” (muadi), which is simply the generic term for the transformational state of a mask wearer (see Maes 1924: 34). About the same time, diagonally southwards from the Tempa, some rounder, monochrome masks from the eastern Luba began to appear, such as the aforementioned “Kifwele” collected and probably photographed by Father Colle.5 It should, however, be noted that Frobenius's 1905–1906 note about “kifebbe” masks is to the area of the Bena Mpassa, which he specified was a “Kalebue” group located west of the Lomami River (Klein 1990: 107); this conforms to both Cynthia Anson's research (personal communication in Merriam 1978: 63) and mine. Rev. Burton, collecting in the late 1920s and 1930s, made numerous references to Kalebwe masks and acquired several examples from the Chofwe, just to the southwest of the Kalebwe (see Hersak 1986). This emphasis on the centralized region as the hub of bifwebe masking may align perhaps more closely to what Mutimanwa Wenga-Mulayi claimed in his Luba research (1974). He proposed an origin in the village of Ngyende-Majaja among the Bena Gende, located south of Kisengwa in an area of admixture between Songye and Luba (1974: 120). He does assert that it is the Kalebwe who influenced the Luba, an idea which is also made known through the lyrics of a kifwebe song (1974: 121, 162–63). Though I have also endorsed his origin hypothesis as plausible (Hersak 1986: 42), our information is based on field work and views and ideologies upheld in the postindependence period.
The fact is that there may not be a unilinear path of dispersion of this society, its mask forms and ideologies, but rather multiple ones with continuities, ruptures, adaptations, and redefinitions, symptomatic of local historical specificities. Moreover, the distinction that is often drawn between sacred and secular practices, as if they are diametrically opposed or sequential developments, is somewhat flawed. In African societies, these two aspects can exist concurrently. Karel Plasmans enumerates the purposes of bifwebe: as playful and spectacular entertainment, as magic and sorcery, and as agents of fear (Plasmans 1965: 446). Anson reports that the ndoshi-type Kalebwe mask had the dual function “to strike terror and to entertain” (1972a: 38). In one Eastern Songye village (Ilunga Ngulu) I was even told that the same mask could be used for different purposes: It could be danced by young people for entertainment or it could be worn by elders when magic was called for. Sidney Kasfir similarly points out, with reference to her Idoma research in Nigeria, that “mask events” may differ in their goals but not in their form; hence she states that the sacred/secular distinction makes little sense, since masking is perceived as both ritual and play and may even be present in a single enactment (1988: 2, 3). While we tend to regard ritual as a serious matter and understand easily that it reaffirms and sustains an older order, the value of play and entertainment tends to be disregarded as frivolous, even though its strategies of “inversion” or “transposition” are important commentaries onf society, as we can see in Songye and Luba song narratives (Kasfir 1988: 3). References to play or to a game, as noted by Samain (1923: 95) or Himmelheber in connection to masks he collected in the 1930s (1960: 404), for example, are not to be dismissed lightly and through a Western lens.
The fragments we are dealing with point to multiple functions of the masking society that were applied variously in the different Songye and Luba polities. Like single objects that have specific biographies (Appadurai 1986), village or chiefdom associations with the same name—kifwebe in this case—may also have individual biographies and correspondingly different pathways and functions, as Ute Röschenthaler has shown in her study of the Cross River region in Cameroon (2011: 503). Through the diffusion of associations, the effect of bricolage (a concept introduced by Claude Lévi-Strauss, 1966: 16) is set into play in which individual parts change and new outside elements are always added, though difference is coupled with sameness (Röschenthaler 2011: 438, 502). As Graeme Were put it, “each creation or performance draws from the past and into the future, each is an original and a derivation and each is somehow imperfect … which endows it with the capacity to generate variation” (2010: 267–68).6 In this “fluency” of performance surprise, unpredictability, and innovation are essential ingredients; hence it becomes impossible to determine the original or single source of invention of the association (Were 2010: 502).
Taking into account the viewpoint of this mechanism, it is easier to come to terms with the kaleidoscopic story of bifwebe practices. But, from a highly general overview, one can say that the associations functioned as regulatory bodies enforcing obedience and social responsibilities in multiple contexts. They appeared at the investiture and death of a chief, at new moon rites, initiations of new members to the closed bukishi society (particularly among the Eki), circumcision proceedings, and general communal work efforts such as the maintenance of roads. These are patchwork glimpses of their past twentieth-century activity, but in the late 1970s postindependence period during which I conducted field work, bifwebe matters were twofold. Among the Kalebwe, Chofwe, Eki, Bala, and also Luba, bifwebe featured in entertainment spectacles already in the colonial period, as we can see from reports about their playful role from postcards (Fig. 3) and from photographs taken by Karel Plasmans in the 1950s (Fig. 4; see also Hersak 1986, 2012). These may have been staged to divert the eye of the colonial administration, highly suspicious of bifwebe practices. I was told that this coincided with the Belgian prohibition of the poison ordeal in 1924 (see Hersak 1986: 42, 44). Rev. Burton touched upon this explicitly in 1932 in a handwritten note of a letter addressed to Mrs. Winifred Hoernle of the Bantu Studies Department, University of the Witwatersrand, for whom he was collecting pieces (Fig. 5). He stated: “[O]wing to the use of masks in blackmail, the Kabinda authorities took a very strong stand against them some years ago.” He expressed the carvers' great fear of making even duplicate pieces.7
Upon independence and the presidency of Mobutu Sese Seko—who upheld the ideology of authenticité, that is, the return to “traditional” values and practices—a renewal of masquerades occurred.8 The blue, green, and orange colors of masks that had been popular during late colonial times disappeared, returning to the red, white, and black prevalent in earlier periods, and sackcloth costumes were replaced by woven ones (Figs. 6–7; see Hersak 2012). This was particularly conspicuous east of the Lomami River among the small chiefdoms of the Eastern Songye, where sociopolitical matters were different. In this rather secluded place, a rotational, more democratic political system, quite unique in Central Africa, had survived and was interacting with the bwadi bwa kifwebe society. Supreme chiefs were elected into power for a period of three to five years and occupied the distant sacred grove known as eata/ehata. As Jan Vansina noted, “Eata is an original variant on the more common pattern of centralized states where several branches of the royal dynasty rule in succession” (1990: 182). He estimated that this institution existed already in the fifteenth or sixteenth century and was based on a strict division between the ruling class of aristocrats and the plebeians. In this interesting aristocratic republican blend, political rivalry was probably fierce, as was the requirement of significant payments by the elected leader to his peers.9 Although Vansina claims that eata was ousted after 1820 (1990: 182)—due to incursions of the neighboring Luba, and the devastating consequences from about 1870 to 1900 of Arab raiding for ivory and slaves, Belgian colonial penetration, and the centralization imposed by Lumpungu Kaumbu—the system survived among the Eastern Songye (see Merriam 1974: 7–11). It is even mentioned in the 1880s by Tippu Tip, one of the key Zanzibari traders in the region (Brode 2000 : 41–43).
Given this context of political competition among potential contenders for ehata rule and the need to accumulate substantial resources, the bifwebe association played a significant role. They helped to sustain the polity, but seemingly as an independent body. Maskers exercised social control and financial redistribution, as confirmed by Burton (1961 ) and Merriam (1978: 94) by imposing fines and collecting initiation dues; however, their appearance for festivities was not excluded. Therein lies the unpredictable power of these spectacles. Many people I spoke to, among the Songye in DRC and more recently among the Songye diaspora in Belgium, noted the twofold functions of the bifwebe as well as differences in regional effect. Within the dual potential of action, what was regarded with most distrust and apprehension was the magical power commanded by the maskers.
During my field work in the late 1970s, it was well known that among the Eastern Songye the bifwebe were agents of witchcraft (buchi)—considered an inherent and inherited mystical power used for protective and aggressive purposes—and the dreaded and much-feared sorcery of masende. The latter was accessible to anyone through initiation, the sacrifice of family members, and the use of activating magical ingredients. Witchcraft was probably an aspect of bifwebe masking early on, as Samain indicates in his reference to the feared medicine of the ndoji (1923: 95), but it seems that masende was a postindependence introduction to reinforce the power of this association. Given the antiquity of ehata rule and the progressive divide and conflict between elders and youth, as in many parts of Africa, masende may have been a key tool of the gerontocracy to maintain old social and political values and ongoing ehata rule.
Consequently, the bifwebe were effectively instrumentalized not only through overt masende action but also through their ambivalent identity of Otherness. Conceptualized as bizarre creatures from a mountainous wilderness beyond Songye territory (Fig. 8), they defied familiar classification as either human, animal, or spirit, though imbued with attributes of all three. Those who wore the masks were referred to in the first place as muadi, a generic term for a transformational state of the familiar into something other. (Maes also used the term ruadi in reference to the masks collected among the Tempa [1924: 34]).10 Specifically, however, they were buadi bwa kifwebe, that is, buadi of the kifwebe striated species as opposed to other types of masks, such as the raffia fiber kalengula, or the lumachecha, which was often made of a gourd. Another interesting term for the mask wearers among Eastern Songye—and, according to Anson, also among the Kalebwe (1972a: 8, 9)—was ngulungu, meaning antelope and in particular the bushbuck. This species (Tragelaphus scriptus), common in the region around Kisengwa, is robust, pugnacious, and potentially dangerous, like the bifwebe. But, as Roberts points out, ngulungu refers to a state of being rather than to outright resemblance (1995: 92); nonetheless the bushbuck is characterized by certain physical anomalies like the white striations on its brown to reddish coat and the conspicuous dark band between its eyes, both of which relate to salient features on bifwebe masks (see Hersak 1986: 50). Anson emphasizes this associational link by saying that, among the Kalebwe, all important masks were “identified as horned animals.” The antelope was the most significant because it was considered to have a close connection to the supernatural world, which is the reason its horns were frequently used as containers for magical substances (Anson 1972a: 8, 9). Writing in 1930, Rev. Burton also mentioned the “‘Tengu‘ or roan antelope dancers [around] Pyana Mbayo,“ south-east of Ankoro (1961: 177) but even among the more recent, striated bifwebe, especially those of the Luba, horned examples have come to light (Fig. 9; see also Mutimanwa 1974: pl. 12; Hersak 1986: 106, fig. 62; Felix 2003: 16, 68, 69). And indeed, all kifwebe masks have a hornlike extension attached to the costume at the back of the head—or to the top of the cranium, in the case of Luba pieces (Fig. 10). An interesting Eki proverb emphasizes this importance of horns, saying ”ears come first and horns follow,“ which is a reference to the extraordinary wisdom acquired by the elders (Tshiluila 1973: 58).
As anomalous creatures, the bifwebe would energize the public arena with the unpredictable, unusual, and spectacular but also, paradoxically, with regulating effects. This they achieved through the different roles of the maskers, whose visual, auditory, and behavioral characteristics brought into play publicly recognizable gender and power distinctions. The gender divide was noted by early observers such as Colle (1913); however, the hierarchy of power may have been expanded and brought into view with reinforced masende practices. Broadly speaking, masks identified as female were predominantly white, with their more contained features outlined in black and red and, in particular, a flat nose-forehead delineation (Fig. 11). Symbolically their color, form, and role were linked to physical, biological certainties and to cooling and controlling effects. They were associated with the moon and, by extension, to the cyclical nature of the moon's potential for renewal and continuity. Male maskers, identified by their red, black, and white striations and prominent crests, were typified by erratic, hot-tempered behavior and unpredictability, ensuring change and transition as in initiation rituals.11 Female maskers communicated their power to solicit the aid of ancestral spirits through the idiom of dance (Fig. 12), whereas male maskers were apt to rush about in uncontrollable fashion emitting unusual, nonstructured sounds and performing miraculous feats (Fig. 13–16). This striking contrast in the behavior and corresponding identity of the two genders of masks is underpinned by local conceptions of music and, correspondingly, dance. Songye aphorisms collected by the ethnomusicologist Alan Merriam tell us that, “When you are content, you sing; when you are angry, you make noise. A song is tranquil; a noise is not” (1973: 272).
According to this, singing and dancing are structured modes that were a part of female masks' identities. They brought pleasure, whereas male masks simply created noise and evoked fear. Merriam tells us further that the bifwebe (at least among the Bala) were recognized by a characteristic sound called mukungulo, which was recognized as one of the five great classes of sounds, described as deep growling like that of a bullroarer, an airplane, a motorcar, or the roar of a lion, leopard, or crocodile—all of which excited and terrified the auditor (1978: 93). There is more to this interesting analogy. Music, like dance, is distinguished by the Songye from nonmusic (and probably nondance) on the basis of three factors: It must be produced by human sources, it must always be organized, and it must have minimal continuity in time (Merriam, 1964: 64–66). This would apply to the dances of the female masks, whose interactions with ancestral spirits and witchcraft powers relate to those who once were, or are, human, whereas the random, irregular movements of the males point to nonhuman beings or at least those transformed into dangerous creatures of the wild. Alfred Gell's study of Umeda dance (West Sepik Province, Papua New Guinea) echoes this notion that dance is a human, stylized deformation of nondance mobility, which has meaning that always refers to the world of mundane action (1999: 143, 156). In sum, dance and music are worldly expressions, whereas the antics of male bifwebe are interludes of the supernatural.
Apart from gender, the male maskers I witnessed among the Eastern Songye were ranked into two power grades according to their mastery of inherent and acquired mystical expertise. These I have designated broadly as “youth” and “elder.” The latter are the most powerful bifwebe, recognizable throughout by their prominent features and especially their highly protuberant crests, an indication of their superlative control of masende sorcery. As noted earlier, this fanning out of power distinctions may be a later development, since one of the well-known kifwebe in the collection of the Museum for Central Africa in Tervuren (Fig. 17), collected prior to 1928 in the Songye/Luba area of Katompe in Kabalo territory, was identified unanimously both east and west of the Lomami River by those I interviewed as the most powerful masker and yet named ndoshi, or witch. I was puzzled by this identification, expecting a reference to masende according to the classification I had learned, but concluded that this was simply a veiling of its true nature. I wondered if this appellation did not confirm the context prior to the reinforcement of masende, when maskers were distinguished visually mainly by gender rather than power. I also hypothesized that this was a prototypical model that probably marks a period of transition and subsequent diversification in style and usage in western and eastern Songye regions.
According to Cynthia Anson's field research, the term ndoshi did not refer to a particular type of masker or to the nature of its mystical power. It was rather the generic appellation for the main male maskers (kifwebe kia mulume) in former Kalebwe practices. She states, “In Bekalebwe thought, Ndoshi is the personification of malevolent and evil intent” (1972a: 19) and is the sorcerer capable of the most potent level of masende (1972a: 23). Ndoshi combines the diverse mystical powers at the highest level. With this new information in view, some of the early literature now comes into focus and supports Anson's view. In particular, the first mention of the word kifwebe in A. Samain's Kisongye dictionary and grammar clearly emphasizes the gender and role distinction between the “harmless” kifwebe kia bakaji (female or wife) and the dangerous male ndoji possessed of supernatural powers he referred to as manga (1923: 95).12
In contrast to Eastern Songye practices—where a single female mask was the requirement, surrounded by a multitude of male maskers of varying degrees of power—among the Kalebwe there was a fundamental difference. Kifwebe groups were composed of a single powerful male performer (although it was possible to have up to three), named and individualized on the basis of his solo performance, and an unlimited number of anonymous female maskers who constituted a “corps de ballet” that performed subdued and graceful dances bringing male action to a finale (Anson 1972a: 18). The interesting aspect here is that the male power distinctions are not represented visually as among Eastern Songye, yet they are indeed present in three echelons of buadi membership. These comprise junior initiates, boys around the age of ten to early teens, known as milengi (sing. mulengi), who accompanied the bifwebe unmasked and prodded them on (Fig. 18; see also photos of youth participants in Hersak 1986).13 The two other levels of initiatory achievement and knowledge included the badi bwa ntoshi, who generally wore the white female masks, and the senior badi bwa nkula elders, transformed through progressive masende revelation into the superlative ndoshi (Anson 1972b: 3, 7, 14, 16). I should point out here that the two levels of maskers correspond somewhat to my classifications of “youth” and “elder.” The nomenclature of the two levels recalls (and may be borrowed from) the two initiatory stages in the Bukishi society but, more significantly, ntoshi (meaning white) and nkula (red) are the two signaling colors that define the female and male masker, seemingly both in this context as well as that of the Eastern Songye. Furthermore, this color designation seems to hint at cosmological connections and the distinction drawn by Anson between nighttime and daytime performances, the former occurring during new moon rituals associated with human, crop, and animal fertility and the latter, daytime masquerades that were more sporadic and initiated when community crises arose and social order was endangered requiring the regulatory powers of ndoshi (Anson 1972a: 31, 32).
In performance, whether among eastern or western Songye groups, an overt display was readable visually and experienced emotionally by the villagers. Anson's, Plasmans's, and my informants all referred to the very same miraculous feats of the male maskers (e.g., cutting themselves in half, spewing bees or wasps out of their mouths or topknots, making snakes appear, and flames of fire on their bodies) which Anson attributes to clever trickery rather than imaginative lore (1972a: 21). Indeed, the esteemed Kalebwe chief Kitumbika Ngoy also gave me a few examples confirming such practices. For example, he revealed that ndoshi maskers sometimes carried a rolled piece of fibrous banana stalk, stuffed with cotton and a piece of glowing coal, which they attached around their necks inside the costume. When the masker blew into it, smoke would rise from the top of the horn, much to the bewilderment of the spectators. In confronting wrong-doers, ndoshi would blow smoke directly at an individual through the facial perforations, thus causing illness to the person in question. Kitumbika further explained that bees were manipulated by being enclosed in a handheld gourd attached to a stick; if someone angered or dared to insult ndoshi, the masker struck the person with the gourd, which would break, releasing the angered bees into action.14 But Anson also adds that among the Kalebwe, the male ndoshi masker demonstrated his exceptional Otherness by giving birth, all of these acts showing ndoshi's extraordinary power to defy normal human capacities (1972a: 22). Another major show of strength was the reversal of gender roles: he would mimic labor, fall on the ground with exhaustion, and then bring into view a “baby mask” (Anson 1972a: 21). These miniature bifwebe, without eye holes, breathing holes, or mouth perforations, are like examples seen in the Woods Davy private collection in Los Angeles and others.
Apart from this public display, there were more specific codings and rules that were secret knowledge shared only by members of the kifwebe closed association. During initiation, in-group members learned a series of esoteric words and associated multireferential terms to nature, culture, and cosmology which defined every detail of the mask and costume. According to my findings, in this coded imaging, the face of the kifwebe was defined as the concentration of its power (eiba/eyilu), with the right side referred to as the sun (nguba) and the left as the moon (mweshi). This duality, like that of day and night, east and west, and male and female, and their daytime and nighttime appearances, emphasized the essential complementarities, like the ongoing linear pattern on the masks. Other features of the mask emphasized its mystical power, with the eyes referring to sorcerers, obvious agents of extraordinary vision; similarly, the mouth was called the beak of a bird or the flame of a sorcerer, the painted striations and design evoked something transformed, a transmutation, and reference the striped bushbuck antelope and the less familiar and consequently anomalous zebra, specifically Burchell's zebra, Equus (Hippotigris) burchelli (see Hersak 1986: 49, 60, 62). The latter, like the crocodile's snout associated with the chin and the lion's mane related to the beard of raffia fibers, are references to large, powerful, and ferocious animals, although insects and rodents like the porcupine (Hystrix sp.) are also recalled in view of their potentially fatal stinging and piercing capabilities. The bodysuit is similarly coded, down to the cords used to sew the soles to the leggings.
I have documented and analyzed this in much greater detail in an earlier publication (see Hersak 1986: 58–66), and although many symbolic parameters can be identified, it is a code that is obviously, and of necessity, specific to every bifwebe group or locality. So, for example, while the mortar and pestle, a life-sustaining implement clearly reflecting gender complementarity, is associated with the legs, in Anson's documentation there is a relationship to the neck of the masker. Both references can be explained from a symbolic perspective, but what I see here as significant is the pivoting of a similar nomenclature that reflects the dynamic dispersal of bifwebe activity.
The coding I have referred to may, in fact, not be essential to the overall understanding of this closed association. It is simply a means of separating initiates and nonmembers as in any such society. However, what I have questioned in a more recent publication is why this coding refers to a singular male kifwebe masker (Hersak 2012: 16–19). Cynthia Anson's research provides a likely answer. It is the verbal image of the all-powerful ndoshi mask of the Kalebwe that is the defining model. This confirms its overwhelming notoriety and recognition throughout the Songye region. The extraordinarily exaggerated facial features of this mask type (Fig. 19), with its bulging eyes and highly protruding mouth, nose, and crest, emphasize acute sensory perception as that of sorcerers like the sha masende, who see, feel, and smell like animals rather than humans.
In contrast to this bizarre and frightening creature whose image may reflect the psychological dramas of initiation (Hersak 2012: 17, 18), the morphology of the white female mask (Fig. 20) exudes containment, calm, and the “metaphoric rubric of coolness,” a concept coined by Robert Farris Thompson in observation of other African sculpture (1973: 43). From Eastern Songye sources, the white mask was the first to be carved in a new bwadi faction. It was the precursory model from which other variables developed. And, in view of this practice of replication, coupled with the role of the white mask as the guardian of continuity, its image became a recognizable icon seen on shields, stools, and other objects (Figs. 21–22). Given its emblematic status, free of any verbal code and concern with “copying or reiterating an existing order,” such representations were formal, fixed, and timeless, as McLeod would say (1976: 101–102).
LUBA MASKING IN THE ANKORO AREA
Luba masquerades differ from the Songye contexts of bifwebe, at least as described by Mutimanwa for the region of Ankoro, although some similarities are shared with Kalebwe practices. Principally, while Songye maskers exercised control by using sorcery, among the Luba their main function was to eradicate such practices and to cleanse villages of evil spirits of the living and the dead. They referred to themselves as banganga (sing. nganga), a term used for diviners and healers that emphasized their positive social role. Nonetheless, it was common knowledge that antisorcery practitioners were of necessity endowed with immense mystical powers able to match their anti-social opponents. Mutimanwa states explicitly that it was a society based on majende (1974: 147), like among the Songye. The fact that the carver of the masks chose to distance himself from the finished piece and to remain anonymous also points to the ambivalence and potential danger of his clients' use of this mystical power (Mutimanwa 1974: 149, 178).
But to maintain the paradoxical illusion of positive action, all the postindependence masks, both male and female, were predominantly white, a color associated among the Luba with benevolent spirits of the dead and with healing (Figs. 23–29). Moreover, unlike the fiery and anomalous bifwebe creatures of the Songye that were said to originate from a mountainous bush, a Luba myth speaks of their masks as spirits that emerged from a lowland ditch near a large supernatural lake (Mutimanwa 1974: 116–18; Hersak 1993: 154), generally an abode of ancestral spirits. There were two males and one female, called Kyeusi, which corresponds to the dominance of male masks to the single female in Luba groups. The female decided to go to live in the village, whereas the males preferred the bush and all its nourishment, until one day the villagers solicited the bifwebe to come to dance among them. All the inhabitants were so enchanted, they (presumably the men) asked to be initiated on agreement to guard their secrets and, as Mutimanwa states (1974: 118), that is how the bifwebe became an affair of real men rather than spirits and how bifwebe masking was transferred from one village to another.
This myth points to the greater interaction between bush spirits and the village and perhaps to the spread between sacred and secular activity. But fundamentally, bifwebe affairs of these spirit entities are orientated towards the beneficial interests of men in the hands of the nganga, antisorcery practitioners of the bifwebe. It is therefore not surprising that dream images of bifwebe experienced by Luba women in pregnancy were welcomed as positive omens rather than dangerous birth signs, as among the Songye. Luba women who had these nocturnal symptoms received their newborn ritually as a kifwebe child, named it after the mask the mother dreamt of, and kept the child for two years in the mutamba, a special enclosure adjoining the house. The mother wore a miniature mask around her neck resembling the mask she saw to announce her special status (Mutimanwa 1974: 126). It also follows that small-scale images of the bifwebe were used in diviners' paraphernalia, on emblems of status and symbols of prestige. Though these are postindependence findings, an example of a miniature mask was acquired by the Museum Fünf Kontinente in Munich as early as 1911 from F. Mitchell (Fig. 30; Kecskési 1987: 345).
Luba bifwebe of the zone in question in the 1970s danced at new moon rites and at the death or investiture of member of the bakasanji association, hunting down evil spirits and sorcerers. They also participated in the kyeusi, a cult attributed to the first female kifwebe which was devoted to healing through spirit possession (Mutimanwa 1974: 30–34). In addition, they made their appearances during investitures and funerals of chiefs and members of their own association, during conflicts requiring judicial action, and at the birth of a bifwebe child and upon its exit from the mutamba enclosure. Performances similarly occurred on the occasions of national celebrations, visits of important personalities, and village celebrations (Mutimanwa 1974: 141–42, 127); Mutimanwa also reports at length on the use of round, striated bifwebe masks called katotosi, during key moments of male circumcision rites among the Luba Samba (1974: 127–40). He notes that some of these examples were field collected by the Institut des Musées Nationaux du Zaïre (Mutimanwa 1974: 49), but for the large part they are unknown, having been burned with all the other ritual paraphernalia at the end of the initiation (Mutimanwa 1974: 140). While there is an unfortunate gap in our knowledge about practices involving round masks, Mutimanwa's report is probably correct, but only in part. The round mask katotosi may have been used in varied contexts, and among these, it appeared at crucial moments of boys' initiations, somewhat like the kakungu masks known among the Yaka and Suku peoples (Bourgeois 1980). But, as among the latter, the masks and paraphernalia that are likely to have been burned were those worn and used by initiates in keeping with the cycles of ritual death, rebirth, and generational renewal. A little-known source, namely that of Father Peeraer, published in 1932 under the name P. Servatius (and referred to by Theodore Theuws almost thirty years later) confirms the previously noted role of the katotosi (katotoshi) among the Luba Samba (1969:136–45).15 Like kakungu, referred to by Bourgeois as a “mask-charm,” a “lineage nkisi” (1980: 42, 43), Peeraer also describes katotosi as a “fetish (mukishi)”(1932: 539). Both were powerful personae, greatly feared by the neophytes, as they were there to reprimand and to protect, both within the initiatory context and beyond (Bourgeois 1980: 42; Theuws 1960: 139). The field photograph published by Volper (2012) of the round Luba mask now in the Seattle museum could well be one of those maskers, also categorized by the collector Colle as a fetish (1913: 440). The imposing figure stands before a row of young men and boys, three of which hold the wooden slit drums (nkumvi in Kiluba; lunkufi in Kisongye) used in important ritual occasions as well as the bifwebe performances I witnessed (Merriam 1973: 254; Mutimanwa 1974: 176; Hersak 1986: 46–47). There is a conspicuous absence of women, which suggests a closed male ritual context, quite different from maskers' appearances in village spaces with wider community presence.16
The lives of round Luba masks, documented individually and sparsely, if at all, pose further challenges. Can we assume that they are part of the kifwebe complex because of their striations? Did the use of the name kifwebe hold the same meaning and application throughout the region? We are fortunate to have some data from the 1930s that corresponds to oral accounts in the 1970s, as this is a fragment that extends the patchy interpretative fabric of both Luba and early Songye pieces.
With Luba and Songye post-1970s masks, a similar time frame and more extensive research allows for a more solid comparative basis. But while morphological similarities are apparent, iconographic codes, performances, and applications come across like mirror images with refractions, sharing sameness and difference. As a result of direct confrontations and exchanges, these masking societies constructed opposing discourses about their origins and purposes. Luba claimed that the bifwebe came from the Songye, particularly the Kalebwe (Munimatwa 1974: 162–63) and that they speak Kisongye, whereas the latter maintain quite the contrary (Hersak 1993: 154, 155). Similarly with Kalebwe and Eastern Songye masks, familiar elements are exploited and recomposed into different scenarios, as this paper has shown. These divergences are a marvelous strategy of differentiation which shows creativity with borrowings of forms, iconographies, performances and “powerful” names that bring into being reconstructions and ongoing adaptations. These “creative ingredients” are what Jordan Fenton aptly calls “expressive currencies” that are employed to keep masquerade “arts fresh, new, and relevant” (2019: 20).
This article was completed in early April 2017, adapted for African Arts and submitted on July 5, 2018.I must thank Philippe de Formanoir for kindly restoring my old field images and some of the studio shots included in this article. Costa Petridis deserves special mention and thanks for his thoughtful and constructive remarks.
The name Songye has been used since the colonial period as a generalized ethnic label to designate linguistically and culturally related groups that include Kalebwe, Eki, Ilande, Lembwe, Bala, Chibenji, Chofwe, Budia, Sanga, and numerous small chiefdoms east of the Lomami River. The latter are, in fact, the only peoples to have referred to themselves as Songye during my field research, whereas the others identified according to the aforementioned names. In this text I capitalize Eastern Songye to distinguish those whom Merriam also calls “the pure Basongye to the east” (1973: 250).
I am in possession of the notes and images of Karel Plasmans and honored to have them. I am also most grateful to Cynthia Anson for engaging with me since 2015 in exciting scholarly exchanges and for sharing with me her unpublished field findings.
This is not the earliest round, striated mask but it is aesthetically quite exceptional and associated with the well-known Father Colle. See the Museum Fünf Kontinente, Munich, for examples acquired in 1905 from H. Deininger (Kecskési 1987: 347, 348).
The statement provided by Frobenius may contain a word or two of truth but not in relation to the mask in question. It sounds rather like selling jargon offered to an insufficiently informed Westerner. In fact, the absence of peripheral holes, like other examples in major public collections (such as masks AE.0334 and AE.0335, acquired from the Antwerp collector and dealer Henri Pareyn now at the Museum aan de Stroom, MAS, etc.) suggests that the pieces were purchased prior to use or preparation for performance, in which case holes for the attachment of a costume would not have been necessary.
Were's statement relates to an exploration of prototypes especially as conceptualized by Alfred Gell in his 1998 anthropological theory of art.
University of the Witwatersrand Art Galleries, Johannesburg; W.F.P. Burton archive.
For an interesting study of Mobutu's ideology of authentcité and its political deployment, especially in relation to collections at the IMNZ, see Van Beurden 2015.
In his reference to ehata, Jan Vansina overemphasized the role of the bukishi “secret association” in maintaining this political system (1990: 182, 355). The bukishi was principally an Eki institution, documented in the mid twentieth century by G. Wauters (1949) with information obtained from a prisoner seeking liberation in exchange for data, according to Mutimanwa (1974: 35).
Merriam noted that a special magical mixture of unknown composition, known as mulawe, was rubbed on the mask and to activate it and to facilitate the wearer to enter in the special state of bwadi (1974: 144).
Zoë Strother has discussed the issue of physiognomy relating to gender distinctions with Pende masks and has shown a wider pattern of correspondence that also includes the Songye (1998: 155–67).
Samain noted specifically the following: “… c'est plutôt un jeu, un amusement, aussi on l'appelle kifwebe kia bakaji parce qu'il est inoffensif; mais il y a aussi kifwebe kia ndoji qui est craint parce qu'il y a question de manga” (1923: 19).
Ansen notes that the level of milengi did not exist around Kabinda and that the boys were instead organized in their own organization called Kalengula (1972b: 3).
Kalebwe chief Kitumbika Ngoy, interview with author, Lubao, October 1977.
These two sources have been brought to light by Constantine Petridis (2005: 7).
The photographer's framing of this field shot could have omitted features contradicting my interpretation. But Volper's comparison of this image of the round mask to a 1928 photo of an oblong white masker from the Katompe chiefdom emphasizes, in my view, the difference between the two contexts rather than what he sees as “a similar environment” (2012: 28, 29; Figs. 14–15).