Abstract
This article offers an interpretation of dumbbell and circular shapes in the rock art repertoire of Uganda using a contextual interpretive approach based on Pygmy ethnographies. It provides a key to unlocking how Pygmy groups experienced their cosmos and demonstrates the associations of rock art images in Pygmy thought in an attempt to explain their symbolism. In so doing, the rock art in Uganda is placed into a new conceptual framework stressing Pygmy affiliations. This approach unlocks aspects of geometric rock art hitherto not understood.
The rock art: Dumbbells and circles
In Uganda, the iconography of the geometric rock art contains repeated intriguing images such as circular shapes that occur independently at some sites, and at other sites juxtaposed or superimposed with spread-eagled shapes and images, described here as dumbbell and sausage shapes. A description of these shapes could probably suggest their meaning.
Dumbbells
Dumbbells are shapes with narrow middles and large, circular ends. In Uganda, about a quarter of the documented sites depict dumbbells, a pattern very similar to that of the broader geometric art zone in south-central Africa where there are comparatively few sites with dumbbells compared to those with concentric circles. Sausage shapes associated with concentric circles are a more frequent occurrence. The highest concentration of dumbbell and sausage shapes in Uganda is at Dolwe Island main site (Chaplin, 1974: 40). Posnansky’s re-drawing of the ceiling panel of this site shows the dumbbell shapes depicted in outline and finger-painted in red pigment. There are also ‘half dumbbells’ where the smaller end of the dumbbell is a concentric circle, a circle and dot or a circle without an enlargement (Posnansky, 1961: 107). ‘Typical’ dumbbell shapes have a dot, a circle or a concentric circle at either end. To ease discussion, I assign numbers 1 through 17 to the shapes in Figure 1.
Re-drawing of the ceiling of Dolwe Island main site (image adopted and modified from Posnansky, 1961).
It is difficult to discern the angle at which the dumbbells are intended to be viewed, if any, as they are oriented differently. In Figure 1, shape numbers 8 through 10 have a white outline and in-fill around the concentric circles (number 9). It is highly likely that red and white pigment was applied at the same time as a conceptual whole. However, white pigment faded from other dumbbell shapes although this evidence is now lost in time. A shape different from number 9 (Figure 1), but similar to numbers 1 and 2, is a motif with a long line and a small circular shape at the base and a half-dumbbell (Figure 2a) amongst a cluster of other shapes at the entrance to the Dolwe Island main site. These shapes, very similar to those on the ceiling, are oriented vertically. In Figure 2a, the circular end of the dumbbell is at the top and the narrow end at the bottom. The narrow end has a small, seemingly triangular tip suggesting a probable intended orientation of similar shapes. In Figure 1, shapes similar to Figure 2a are piercing or overlapping circles (numbers 1 and 7), or appear to be attached to a closed circle (number 13) or occur in a similar straight posture with the tip of the shape fusing with the circular one, becoming a circular outline (numbers 8 through 17).
Examples of half dumbbell and sausage shapes.
Concentric circles
Concentric circles are the most frequently painted shapes in the rock art of Uganda. They are defined here as circular shapes within circular shapes with a shared centre or axis. The circular shapes may not be equidistant in radius from the centre. A concentric circle may be an entity on its own or be part of another, such as a dumbbell, depending on the variant shape. For example, in Figure 1, the exterior of the dumbbells in shapes numbers 8 through 17 is also the exterior of a concentric circle or circle interrupted by two parallel lines as explained earlier. Concentric circles in dumbbells comprise three to four circles; often the final inner circle is filled in. The other end of the dumbbell either has a single filled-in or outlined circle, or comprises one to two concentric circles. Any understanding of the meaning of these shapes is only possible with knowledge about which relevant ethnographies (if any) may, by analogy, elucidate their symbolism.
Analogy
Archaeology is bound to the analysis of mute material remains (Buchli, 2007: 179; Thomas, 2004), making the usage of ethnography and analogical thinking important in rock art analysis and interpretation (Lewis-Williams, 1981). Analogical reasoning must rest on a logical cultural or natural association between the source of analogy and the archaeological context, where relations based on cultural continuity can be suggested, including strong relations of relevance that justify analogical conclusions and give confidence to interpretation (Hodder, 1982a: 178, 1988; Lewis-Williams, 1991: 151–152; Wylie, 1985: 94–95). Significantly, such reasoning establishes that the same relations of relevance operated in the past and that the archaeological feature is similar to the feature in the source of the analogy, holding consistency across a wide range of source contexts despite variability (Hodder, 1982b: 11–27; Lewis-Williams, 1991: 152; Wylie, 1985: 98). Using analogical reasoning, I argue that the ethnography reflects the same cognitive system as the rock art, with both drawing on common symbols or shared gender values and roles, to strengthen relations of relevance.
I draw on Pygmy ethnographies to explore broad insights into their cosmology in order to provide a new understanding of the symbolism of the geometric rock art of Uganda. In this regard, I follow Smith’s (1995, 1997, 2005) Pygmy attribution and contend that some of my interpretations are closely related to his hypothesis for the Twa (Pygmy) geometric rock art of Zambia. However, I diverge from his argument that within this rock art, each gender ‘had its own artistic traditions, but with each, probably, creating pictographs for differing purposes’ (Smith, 1995: 279). For Smith, each gender practised rock art for a different ritual; the red animal forms concern molimo attributed to men and the geometric forms concern fertility and divination attributed to women, the two traditions, paired in the landscape, forming a conceptual whole (2005: 1393). I consider only geometric rock art forms, given that animal forms in this tradition have not as yet been found in Uganda. I now examine the archaeological and paleoenvironment evidence to tentatively suggest relative dates for the geometric rock art.
Archaeology and the paleoenvironment
The rock art is believed to be part of a homogenous geometric tradition, often depicted in red pigment, which spreads across east, central and south-central Africa matching the distribution of a distinct Late Stone Age (LSA) hunter-gatherer culture termed ‘Nachikufuan’ (Clark, 1950a, 1950b: 43, 1950c: 93, 1958, 1959: 211; Juwayeyi, 1981; Mgomezulu, 1978; Miller, 1969a, 1969b, 1971; Musonda, 1983; Sampson, 1974; Sampson and Southard, 1973; Smith, 1995) that may have parallels with the Tshitolian culture of Central Africa (Phillipson, 1976). Clark postulated a link between the Nachikufuan and the geometric rock art (Anati, 1986; Chaplin, 1974; Clark, 1950b: 43–5, 1950c; Garlake, 1995; Leakey, 1983; Mabulla, 2005: 37; Posnansky and Nelson, 1968; Smith, 1997) that was subsequently applied to the geometric rock art in Uganda (Lawrance, 1953: 12, 1958; Namono, 2010a). In Uganda, LSA assemblages bear strong resemblance to the Nachikufu (Ambrose, 1982: 119, 133; Brooks and Robertshaw, 1990: 151; Nenquin, 1971: 132; Sassoon, 1973: 9; Van Noten, 1971). Although rock art and excavated occupation layers are not necessarily contemporaneous, archaeological evidence indicates that the geometric rock art tradition probably belongs to a period before 14,000 BP, prior to this region experiencing significant climatic changes and shifts in vegetation.
Scant palaeoenvironmental evidence indicates that since the late glacial maximum (LGM), lowland tropical rain forests in Africa suffered drastic contractions (Adams and Faure, 1997). Recent regional overviews suggest that Lake Victoria was surrounded by a wide variety of vegetation, from savannah to evergreen forest. Between 21,000 BP and 12,500 BP, East Africa was generally cool, punctuated by periods of prolonged desiccation (Butzer et al., 1972; Kendall, 1969; Kiage and Liu, 2006: 649; Street and Grove, 1976). In the period 13,000 BP to 11,000 BP, East Africa was a cooler, drier environment with a more expansive savannah and reduced evergreen forests and woodlands. Lake Victoria Nyanza dried up completely by 12,400 BP (Stager and Johnson, 2000). However, 11,000 BP to 6000 BP experienced a significant increase in lake levels above current ones such as Lakes Chad, Victoria Nyanza and Malawi (Stager and Johnson, 2000; Thompson et al., 2002), interrupted by a brief dry spell tentatively dated to 8700 BP to 8200 BP (Muzuka et al., 2004: 294). A much drier period, although much warmer and wetter than the present, occurred around 6000 BP to 3000 BP, interrupted by drying periods around 5200 BP and 4000 BP when lake levels dropped, grasslands expanded and forests shrank. The vegetation of Dolwe Islands on Lake Victoria Nyanza changed into a coastal fringe rain forest extending into the hinterland as relic forest (Jackson et al., 1965: 38).
From 11,000 BP to 6000 BP in north central and east Africa, several lacustrine, ravine fishing and hunter-gatherer communities emerged and thrived, such as the complex hunter-gatherer-fisher society termed ‘Kansyore’, believed to have emerged around the dry spell of 8000 BP lasting until about 3000 to 2400 BP (Lane, 2004; Lane et al., 2006, 2007). In my opinion, it is highly unlikely that the Kansyore were the authors of the geometric rock art discussed here, given the widespread distribution of this rock art tradition as opposed to the localised archaeological distribution of Kansyore material culture. Forests were present where the rock art now lies. Presently, zones with geometric rock art are not in the densely forested zones of modern Pygmy areas given the climatic changes explained above, but this cannot detract from recognising this rock art as authored by Pygmy groups probably before 14,000 BP. Absence of rock art in presently forested areas may be due to little or no exposed rock suitable for painting or engraving.
The limited and extremely fragmentary skeletal evidence within areas where the geometric rock art tradition occurs, such as Zambia and a probable skull in Uganda, exhibit greatest affinity with Central African Pygmies (Clark and Toerien, 1955: 108). Historical records attest to Pygmy presence in northern Zambia, strengthening the possibility of a previously wider distribution than today (Gabel, 1963: 39) with a heritage rooted in west-central Africa (Morris and Ribot, 2006: 24). This argument is sustained by Deacon (1984) and Wadley (2000, 2001), who cite a discontinuity in southern African Wilton lithic industries north of the Zambezi. Therefore, it is highly likely that Pygmy hunter-gatherer communities once lived in the general area of the geometric rock art sites discussed here.
Pygmy groups
Researchers (Bahuchet, 1993; Patin et al., 2009; Quintana-Murci et al., 2008: 1600) suggest that an ancestral Pygmy population diverged from the common genetic ancestor of modern Central African farming populations around 60–70,000 years ago. This ancestral Pygmy population shared physical and cultural features but split genetically into western and eastern groups about 20,000 years ago (Cavalli-Sforza, 1986: 414; Patin et al., 2009; Ramırez Rozzi and Sardi, 2010: 8; Verdu et al., 2009: 315–317). Despite evidence for a western-eastern Pygmy patterning of ethno-linguistic-subsistence diversity (Cavalli-Sforza, 1986; Hewlett, 1996: 243) and genetic divergence, material cultural debris indicates certain common beliefs and practices across eastern and western groups (Cornelissen, 2002; Mercader, 2003; Mercader and Brooks, 2001: 198, 213–214). Whilst I draw largely from ethnographies of Eastern Pygmies, I exploit common beliefs among other Pygmy groups to increase confidence in my contention that common Pygmy concepts probably existed in Uganda and have deep historical roots. Pygmy Mbuti recent territory extended to the western and south-western border of Uganda where Twa and Sua Pygmy descendants are still found (Figure 3). These ethnographies are extremely valuable in exposing the socio-cultural intricacies of a Pygmy cosmos (Lewis, 2002; Lewis and Knight, 1995, 1997; Mosko, 1987; Schebesta, 1933; Turnbull, 1960a, 1960b, 1961, 1965a, 1965b, 1978b, 1982, 1983, 1985).
Distribution of known Pygmy groups and Pygmy geometric rock art.
Pygmy cosmos
Broadly, a Pygmy cosmos revolves around the benevolence of the forest, ndura, whose centre is the most peaceful, pleasant and abundant, a place good for coitus, hunting and gathering (Turnbull, 1960a: 36, 1961: xv, 1985: 8; Mosko, 1987: 900). For the Efe and Mbuti, ‘the forest is imbued with a supernatural being called apakumandura, literally meaning “master of the forest” who controls all the life in the forest’ (Ichikawa and Terashima, 1996; Terashima and Ichikawa, 2003: 2–3). The Mbendjele divide the world into forest, ndima, a quiet, peaceful, cool and safe place; and vuli, open spaces without forest canopy, a hot, noisy and dangerous place (Lewis, 2002). The Pygmy cosmos stresses constant equilibrium, balance or harmony with ndura as the basis for behaviour. Ndura, personified as a loving parent, sibling or lover, the giver of life from whom help is ritually sought and thanks given (Mosko, 1987: 898; Turnbull, 1960a: 319–320, 330, 1965b: 252–253), is also a conceptual space, its wet and moist centre (Turnbull, 1990: 69) a metaphor for the primordial womb that engenders the Pygmies with ekimi at the centre and akami at the periphery (Mosko, 1987: 899; Turnbull, 1983: 30, 32, 44, 1984: 33). The transformative characteristic of a womb, like a hearth (Mosko, 1987: 900; Turnbull, 1965b: 264, 1978a: 202, 1984: 50) is present in the kumamolimo, the ‘vagina/womb’ fire of the molimo used during the molimo ceremony (Turnbull, 1961: 69, 148), which I explain later.
The Mbuti believe individuals exist in a sphere of energy, like a womb (Finnegan, 2009/10: 33; Turnbull, 1978b: 97–98). Moving ‘too quickly or violently’ within this sphere to its periphery causes ‘noise’ akami, synonymous with a poor hunt, death, sickness and social aberrations (Turnbull, 1961: 239, 1984: 57; Woodburn, 1982: 198), with being ‘on the edge’ or ‘off balance’ and vulnerable to disembodied spirits that may enter the capsule, enabling movement from one realm to another. When ‘off balance’ a person may be replaced as in death by a mirror image, one’s essence, a constant phenomenon within Mbuti cosmos (Turnbull, 1983: 123, 1990: 61). For the Pygmies, preferred movement should be ‘silent’ ekimi, synonymous with joy, peacefulness, awareness, order, quiet and gentle, the absence of conflict, tension and anger rather than absence of sound (Turnbull, 1985: 12). Given the variation and diversity amongst Pygmy communities in Africa, I focus only on two major Mbuti rituals, molimo and elima, that elaborate gender synergies and the vibrant relationship with ndura amongst almost all Pygmy groups. Pygmy rituals function at individual and societal levels to channel and express emotion, to restore harmony and balance, and to reinforce and maintain fertility of the environment and people. Sanctity is constituted by beliefs that the centre is inherently sacred. Sanctity is negotiated by both genders in the molimo and elima, rituals to call out the spirit of the forest which I now elaborate.
Molimo
During the molimo ceremony performed at night in the camp, hunters sit around a special molimo fire, kumamolimo, clapping hands and occasionally dancing. Molimo, the spirit of ndura, is also a trumpet kept in the forest and blown to awaken ndura to a plight (Turnbull, 1983: 51–52). When ndura awakens, it negotiates, restores ‘balance’ and ‘centralises’ the individual and the camp, ‘curing’ akami (Turnbull, 1965b: 263, 1984: 57, 1985: 11–14), or it shares the ‘joy’ ekimi of ndura (Turnbull, 1961: 87). A molimo death song captures the strong desire for ‘balance’ and belief in the benevolence of ndura: ‘There is darkness all around us; but if darkness is, and the darkness is of the forest, then the darkness must be good’ (Turnbull, 1961: 88).
The molimo ceremony starts when a hunter with embers leads molimo out of the forest and into the camp where it criss-crosses the kumamolimo in ritual coitus (Turnbull, 1960b: 313). During the ceremony the molimo trumpet, a metaphor for the primordial penis, fiery, wet and phallic, is used by hunters to mime intercourse as they dance around kumamolimo (Zuesse, 1979: 50–51). Pygmy groups recognise ritual/real copulation as an axis of transferring and containing life (Mosko, 1987: 899) without explicitly referring to vaginal fluids and semen as pepo. Although the molimo has a masculine character, it is associated with a female component (Moore, 1998: 142; Turnbull, 1961: 133, 1965b: 151, 263, 1983: 85, 1985: 15): Fire is primarily connected with women; the hearth is often referred to as the vagina; the association of blood and fire is considered particularly dangerous. Thus, the association of the male Molimo trumpet with the fire and the ashes gives the trumpet phallic significance in this context. (Turnbull, 1965b: 264)
Molimo equates to masculinity and femininity, its polarity is echoed in sexual contest and gender reversal (Zuesse, 1979: 50–53). Sometimes, a molimo ceremony is held at the end of an elima.
Elima
Elima is a ritual of menarche and a celebration of female maturity, of being ‘blessed by the moon’ (Turnbull, 1984: 57). The elima girl is secluded in an elima hut (Mosko, 1987: 907) – a symbolic ‘womb’, the female counterpart of the deep forest guarded by women. Hunters battle to enter the hut/womb that is ‘with blood’, equating to ritual coitus, deflowering the girl(s) for whom the elima is held. Elima is a symbolic hunt (Namono, 2010b: 187, 193), a festival of courtship (Turnbull, 1984: 57) that also celebrates the hunter (Turnbull, 1965a: 224, 1965b: 256, 1983: 47).
Subsequently, I consider that, metonymically, molimo represents a state of decay or death; metaphorically, it is associated with growth or regeneration. Elima and molimo ritually reinforce social solidarity and cohesion and are principles of social organisation. These spirit rituals are very fluid and have diverse characteristics amongst Pygmy groups (Tsuru, 2001: 103), ones best understood in the cosmological model I propose below.
A cosmological model
From the above discussion, I propose a model of Pygmy cosmologies that underpin their rituals and philosophy (Figure 4) and condenses the social dynamics that probably created the rock art. The model rests on the ethnographies that show all Pygmy rituals call out the spirit of the forest (Finnegan, 2009/10; Lewis, 2002, 2008; Tsuru, 1998, 2001; Turnbull, 1961, 1965b). I consider the relationship between the Pygmy groups and the forest as fluid. According to Finnegan (2009/10: 35) gender does not belong to nor does it constrain individuals in Central African societies. Gender conflict is averted and mediated through ritual. Molimo and elima engage the central ritual elements – water and fire, metaphors for coitus, hunting and social cohesion: The association with the festival of all the basic elements – fire, water, air and earth – and the central importance of food and sex, all combine to create a highly direct, personal and intimate relationship between the Mbuti and the forest, while at the same time binding them together as a band, the one individual dependent upon the next. (Turnbull, 1965b: 267)
Cosmological model of Pygmy worldview.
Hunting and coitus are mediating contexts, both involving sharing (Finnegan, 2009/10; Turnbull, 1965b: 155–156; Zuesse, 1979: 24–25). A hunt joins and preserves the realms of primeval power and everyday life (Zuesse, 1979: 35). I suggest that the molimo is a real and metaphorical hunt; the molimo ceremony, the interface between the two realms. Movement between the realms, I argue, is facilitated through song, dance, mime, myth, food, drink, water, fire, sex, humour and rock art. As transformative agents, water is emphasised in myth; fire, in daily life and ritual. These transitional processes, I suggest, occur in the primal sphere, the cosmic womb (the deep forest) embracing the human womb. The supernatural and real are conceptually similar and continuously interact.
For Pygmy groups, ndura is an integral part of coitus (Mosko, 1987: 899; Turnbull, 1978b: 167–170, 1984: 33); it moulds the foetus, hence ‘child of the forest’ (Mosko, 1987: 899). Coitus during menses equates to ‘sleeping with the moon’ (Mosko, 1987: 899; Turnbull, 1960a: 191, 1981: 169) and, during pregnancy, is believed to make a child grow faster (Mosko, 1987: 899; Turnbull, 1965a: 178, 1965b: 123, 1978b: 168). I contend that the Pygmies do not explicitly refer to bodily fluids (menstrual blood/vaginal fluids and semen) as pepo, but recognise copulation as a means of transferring it (Mosko, 1987: 899; Turnbull, 1965b: 122, 254). A child born in the forest (Duffy, 1984: 134), bathed in waters of the forest and wrapped in soft bark cloth, ‘womb of the forest’ (Mosko, 1987: 900; Turnbull, 1965a: 213, 1965b: 129), symbolically represents maternal (bark cloth) and paternal (plant fluids) dimensions of the forest (Mosko, 1987: 900).
Metaphorical structures like these are fundamental to all belief systems (Tilley, 1999: 10). I have made frequent yet cautious use of metaphor as an interpretative tool to contextualise Pygmy thought, aware that what I may interpret as a metaphorical association may actually be a real and powerful connection for Pygmy groups. Using the cosmological model in Figure 4, I now propose that, at the time of deposition, the geometric rock art of Uganda was an interface between spheres in the context of the rituals at the heart of the Pygmy cosmos. Within this cosmos, the rock art reflects a widespread shared cognitive system with an overriding variety and consistent recognisable method and manner of depiction. This recognition indicates that any shifts in iconography probably reflect regional variation or change in pattern through time. I argue that aversion of gender conflict and the need for balance and harmony are at the heart of Pygmy social organisation and ritualisation (cf. Finnegan, 2009/10; Lewis, 2009) from whence the symbolism of the geometric images derives. I concentrate on a forest-gendered relationship as a key metaphor for cosmological symbolisation (Turnbull, 1965a: 257, 1965b: 252–253, 1978a: 164–165), focusing only on broad societal meanings and motivations. Although knowledge of individual meanings and motivations in the construction of the dumbbells and circles in the rock art is obviously irretrievable, I argue for their interpretation as phallic/womb symbols that harness a life-force linked to fertility, harmony and balance with the forest that engenders Pygmy groups individually and collectively.
Phallic symbolism
Commenting on the dumbbells in Figure 1, Posnansky drew on specific formal similarities between the painted shapes and a proposed subject, stating, ‘[t]he designs defy interpretation unless a phallic significance can be attached to the dumbbell like designs’ (1961: 107). If Posnansky is correct, the circular or oval shape at the end of them (at the top in the case of the one in Figure 2a) is best explained as the scrotum. The meaning of shapes in a pattern is contextual of other shapes (Douglas, 2003: xxxii). I note that a phallic identification is plausible, and while not diagnostic, it explains the pattern in Figure 1 where dumbbells are depicted piercing or overlapping circular shapes.
Figures 2b and 2c have diagnostic depictions of phallic-like shapes. Figure 2b is suggestive of a flaccid penis with scrotum. Figure 2c is a sausage shape with a triangular extension and distinct morphological similarities to a penis. The penis shape overlaps or pierces a concentric circle in a manner similar to those in Figure 1. The difference between the dumbbell in Figures 2a and 2c is their orientation. There is regularity in the manner of depiction – a single concentric circle with two straight vertical parallel lines and an irregular triangle at the base. If half dumbbells are indeed representations of male genitals, we may expect to find further physiological details depicted such as parallel lines suggestive of erectness. Characteristic features of dumbbells in Figures 1 and 2 are the divisions that seem to demarcate a scrotum. A bird’s eye view of Figure 2b would appear as a dumbbell shape with a phallic attachment. I therefore argue that dumbbells and the shapes in Figure 2 represent the same thing and are compellingly phallic.
Smith noted that phallic shapes, such as in Figure 5, occur at several sites in Zambia (1997: 46). However, these shapes were not depicted as frequently as the majority of other shapes within the geometric rock art repertoire. At a site in Zambia, the phallic shape is positioned at the cave entrance opening into an elongated ovoid space, highly suggestive of a penis penetrating a vagina (Figure 5). Depiction of this phallus with an emission suggesting ejaculation lends credence to recognition of coitus.
Ovid entrance to the cave showing location of phallic shape (image: RARI 2007).
From my synthesis of the Pygmy cosmos, I suggest that coitus is a powerful relationship of ritual potency. Male pepo is transmittable through coitus, underpinning the symbolism of dumbbell, sausage and phallic shapes. Pepo is incarnate in the concept ‘penis’, ‘vagina’ or ‘womb’. Female pepo is a crucial agent of nourishment and growth in the womb, the receptacle for the phallus. Naturally, phallic shapes conjure ideas of human procreation. The phallic molimo trumpet ritually copulates with kumamolimo, restoring fertility to barren maidens and virility to impotent hunters.
Womb and vulvae symbolism
Having dwelt on the symbolism of phallic shapes, I now consider the circular shapes with which they are often associated. Chaplin postulated that in many ritual contexts circular shapes are prominent (1974: 41). Collinson (1970) argued that circular shapes are depictions of the sun, moon and stars. Although these ideas are possible, circular shapes are open to a number of other possible interpretations. Many societies that refer to circular shapes consider the sun a male symbol and the moon female, so I do not discount the possibility, but such insight requires prior recognition of the subjects depicted. In the geometric rock art, repeated contextual associations of circular shapes with probable phalli condenses subject options. If identification of phalli in the geometric rock art is correct, then the shapes they penetrate or overlap should naturally be vaginas. A concentric circle pierced by a phallus is far more likely to represent a womb than the sun, and probably carries broader symbolic connotations. I have shown the deep forest, hearths and bark cloth as symbolic wombs in Pygmy iconography, but only move to these deeper levels of symbolism on recognition of the subject, womb. With the strong possibility that the concentric circles or circular shapes depict wombs, I draw on the cosmological model and propose symbolism of the womb within Pygmy societies.
For Pygmy groups, distinct asymmetrical substances, semen and menstrual blood/vaginal fluids, ‘male’ and ‘female’, are transformative and regenerative. In the womb, biological substances become gendered substances: blood and semen, fire and meat (Smith, 1995: 284; Zuesse, 1979). Amongst these groups, a woman’s fertility originates in the source of her pepo: menstrual blood and vaginal fluids. If ndura is involved in conception and birth then pregnant or menstruating women are interacting with the supernatural, and hence their pepo is likely to interfere with that of hunters. According to Lewis (2008), Pygmy groups believe most animals flee from humans involved in the procreative process because coitus, childbirth, pregnancy and menstruation have human ekila offensive to animals. Hence, expectant women or parents of newborn babies must not engage in a hunt (Turnbull, 1965b: 156, 317; Zuesse, 1979: 32). In speech, ekila may refer to menstruation, blood, taboo, a hunter’s meat, animals’ power to harm humans and particular dangers to human reproduction, health and sanity (Lewis, 2008: 298). A hunter receives his first hunting net from his mother (Turnbull, 1982: 148), symbolising a conceptual link between women and animals, both hunted by men, wherein relations with one prevent relations with another (Zuesse, 1979: 31). There is a symbolic relationship between the world of the life-producing, food-producing forest and the life-producing, food-cooking women, between the child transforming in the womb and fire transforming raw into cooked food (Turnbull, 1965a: 177–178). A successful hunt depends on clarity between human flesh and hunted game (Zuesse, 1979: 32–35). This link between sex, menstruation, pregnancy and hunting is a common thread throughout hunter-gatherer ethnography in many other societies (Knight et al., 1995; Lewis-Williams, 1981: 51–52). In rock art research, the link between ritual power and fertility is a prominent theme (e.g. Hays-Gilpin, 2004).
In terms of the cosmological model, there is a ritual/cosmic relationship between hunting (and the hunter), animals, the deep forest and women. Penis-womb symbolism ensures fertility and regeneration in all realms. The symbolism of dumbbell, sausage and phallic shapes has a natural relationship in Pygmy traditions with concentric circles. It would be fallacious to regard the symbolism of the womb in isolation from that of the phalli with which they are juxtaposed since they form a conceptual pair. This conceptual pair is often associated with U-shapes. Smith (1995: 285) argues that in Zambia and Malawi, some U-shapes could represent vulvae, and in the context of my suggested subjects of wombs and penises, this is a real possibility for the geometric rock art of Uganda. Concentric circles may be a metonym or a metaphor (cf. Tilley, 1999: 5) for female, mother and womb. Circles may also be a scrotum filled with pepo, so that a dumbbell may symbolise male and female linking concretely in coitus and metaphorically to the categorised synergy between male and female. Circular shapes are also associated with the moon.
Moon symbolism
The moon has a strong presence in Pygmy cosmology. During molimo and elima ceremonies, the Mbuti evoke ‘mother moon’ (Turnbull, 1965b: 253). Among the Efe, chameleon, the sacred animal, is in rapport with the sky and the moon, with rainstorm and lightning (Turnbull, 1965a: 189; Zuesse, 1979: 20). According to Towles (1993: 72), the fallen logs of the araumbi ‘creation’ tree are a source of mushrooms (buuwoo, ‘stability/growth’) that are said to come out on moonlit nights. The female creation and growth aspect of the araumbi and its association with the moon (hence blood and sex) liken it to the womb (Towles, 1993: 72). The patterned monthly phases of the moon resemble the monthly menstrual cycle, a metaphor for life, death and immortality. A rotund full moon simulates a womb. Wholeness is conceived as female: the arc space of a hut is female, the ‘outside’ space where bachelors sleep is male. Rhythmic changes and fruitfulness are concepts associated with the feminine compressed into a single visible object, the moon. The moon is part of a communion with the forest. Menarche is a gift of the forest and Mbuti females attempt to synchronise their menses with each other and hence with the forest (Turnbull, 1965b: 133). A solitary erotic dance in the forest equates to coitus with the forest (Turnbull, 1961: 245), a rhythmic encounter between male (penis) and female (moon). The Mbuti word for dancing is the same for coitus (Grinker, 2000: 2). Aka net-hunters refer to songé (moon) as creator of all things and of man as created by menstrual blood (Turnbull, 1965b: 253).
I suggest that the moon as a manifestation of pepo probably accounts for circular or crescent-like shapes in the geometric rock art. My reading of the ethnographies suggests that rock art is a part of an intimate relationship with the forest, with molimo as the communal expression of this intimacy (Turnbull, 1960: 319). The forest as the moon is a manifestation of puberty and the menstrual cycle and an agent of transformation (like fire). Females, as mothers, are life-givers and bring ekimi; males, as hunters, cause death and bring akami (Turnbull, 1982: 153). The forest realm, conceived of as being like a woman, is wooed by the masculine hunter (Zuesse, 1979: 32). Sex emphasises the equal potency of male and female. Sexual potency manifests in blood and semen. Sex during menstruation enables the direct mixture of blood and semen. My reading of the manifestations of the forest as moon leads to the conclusion that various circular shapes may symbolise the moon. Often, circular shapes are juxtaposed with concentric circles, sets of parallel lines, dumbbell and phallic shapes. They occur as a conceptual whole and are thus well explained as the moon because this symbolism fits perfectly with that of the other shapes already identified. A solid circle evokes a full moon in its basic form, the most important phase of the moon cycle. Moon symbolism nestles with my interpretation of dumbbells and concentric circles.
Fire symbolism
In Pygmy cosmology, the poignant relationship between molimo and fire is synonymous with a hunter and his wife. Fire, like sex, is a node of transformation or transition, the association between fire and sexuality being that ‘they are all liminal and generative, and their symbolisms would be likely to recur at times or places of transition’ (Zuesse, 1979: 22). A hunter perpetuates human (and animal) mortality by a ‘necessary act of violence and aggressivity’ minimised through the hunting fire (Turnbull, 1982: 148). In the cosmological model, hunt-fire context is best understood in terms of the deep forest concept.
Depictions of fire in the rock art are variations of concentric circles and circles with radiating lines or complex circular shapes referred to here as ‘rayed shapes’, previously described as sun bursts, a stylised sun figure, concentric rings with appendages (Chaplin, 1974: 19, 28) and sigmoid forms (Sassoon, 1973: 8). Pygmy ethnography does not refer to the sun as central to their cosmos so the sun is a very unlikely subject. The rayed shapes show particular attention to the distal parts, suggesting that these shapes differ from basic concentric circles and the distal parts are important. However, the close association of these shapes with dumbbells and concentric circles suggests that they are an integral part of the penis-womb symbolism. Concentric circles symbolise the womb/fire. Equally, fire is a symbolic womb. Around the molimo fire are an array of vines and baskets that fit well with the shape and the arrangement of objects appended to the concentric circles. These rayed shapes probably symbolise fire.
Water symbolism
In Pygmy thought, molimo is associated with water. The trumpet, sometimes called the water animal, symbolically drinks water in the forest (Turnbull, 1960a; Zuesse, 1979: 24). Water may evoke semen or vaginal fluids (Turnbull, 1982: 146–147). This understanding models the significance of water symbolism in the rock art probably depicted in sets of parallel lines or dots.
Conclusion
This article builds on Clark (1950a, 1950b: 43, 1950c: 93, 1958, 1959: 211) and Smith (1995, 1997, 2005), who attribute the red geometric rock art to Pygmy authorship. Smith argues that each gender practised a form of rock art for some personal ritual, wherein the red animal forms concern molimo and were authored by men and the geometric forms concern fertility and divination authored by women. Following Turnbull (1983), Smith noted that among Pygmy groups almost all body and bark cloth geometric decoration is done by women. Although there is a close resemblance with dumbbells and concentric circles depicted in bark cloth and rock art, it is difficult to confirm whether geometric shapes stand for the same set of ideas wherever they appear (Namono, 2010b: 214). Smith (1995: 277–288) suggested some geometric shapes were a product of Pygmy diviners employed in rain-calling rituals for farming communities as part of a symbiotic relationship. He noted that concentric circles and joined concentric circles are almost always associated with fertility. I argue that these and other shapes in this geometric rock art tradition are symbols associated with fertility by Pygmy rather than Farmer groups. These symbols are probably polysemic or may have focused polysemy deriving from contextual associations (Figure 6).
The relationship between nodes of transition, mediators and the rock art.
In terms of my cosmological model (Figure 4), the symbolism of gender opposition in the rock art evokes coitus, mirrors male and female aspects of the forest and is the basis of most ritual action. Coitus merges opposites whose procreative powers are conceived of in semen that symbolises male aspects and vaginal or menstrual fluids that symbolise female aspects. In the rock art, phallic and womb shapes symbolise fecundity, regeneration and potency of the forest and the male-female juxtaposition of procreative powers. Contra Smith, molimo is concerned not only with men’s issues, but also with women's, and thus the geometric rock art concerns molimo relevant to both genders. Within this framework, I suggest that Pygmy rock art sites are mediating points and the images depicted may be recognition of and an attempt to harness pepo. This geometric rock art probably functioned in different ways for individuals and for groups with a more active role within society rather than as a passive reflection of male or female synergies.
Footnotes
Acknowledgements
This article is adopted from my Doctoral thesis (Namono, 2010b), Department of Archaeology, University of the Witwatersrand. I am extremely grateful to Professor Ben Smith for his academic guidance and Professor David Lewis-Williams for his mentorship. I thank the National Museums of Kenya, Uganda National Research Council (permits) and Local Councils in Uganda for logistical support; and Michael Robert Okwi, Dismas Ongwen and Alex Tamale for survey assistance. I thank the anonymous reviewers of this paper for their valuable comments.
Funding
Research was funded by Jimmy Kutosi, a Susan Ward Xharo Award, the British Institute in Eastern Africa, University of the Witwatersrand and the Rock Art Research Institute. I thank them for the generous support.
