… [T]his is no mere miscellany of pieces unaccounted for, but in my estimation the very pinnacle of Nigerian artistic achievement (Fagg 1963:39).
The arts of Nigeria are among the most studied of all African arts and yet there are still exceptional works about which we know very little. In this paper we offer an investigation of copper alloy works from the Lower Niger that demonstrate the extraordinary creativity and aesthetic power to which Fagg refers. These bell heads and their fantastic imagery have encouraged us to take a multidisciplinary approach, a synthesis that allows us to draw conclusions about the dating of these works and about the persistence of particular ways of thinking as embodied in ritual practices though several hundred years. Specifically, we investigate that period in sub-Saharan Africa from the inception of an art making use of copper and its alloys to the entry of coastal West Africa and its hinterlands into the Atlantic sea trade from the late fifteenth century onwards.
What we are concerned with is that provisional category of works identified by William Fagg as the Lower Niger Bronze Industry (LNBI) (Fig. 1), sometimes pluralized, and sometimes including the group of castings known as the “Tsoede bronzes,” as well as the works excavated at Igbo-Ukwu (Fagg 1960, 1963, 1970; see also Anderson and Peek 2002; Craddock and Picton 1986; Peek 2013; Picton 1995, 2012; Shaw 1970, 1977; Willett 1967). It was Fagg's hope that with the progress of archaeology the label could be dropped in favor of specific locations and casting sites, a hope yet to be realized. Moreover, in his enthusiasm for this diverse body of work, Fagg considered that the LNBI would prove more significant for the history of art in the Lower Niger region than Benin City, or even Ife, a possibility that is addressed in part in this paper.
To those ends, therefore, this paper first addresses what we know of the archaeology and metallurgy of the Lower Niger region; and secondly proceeds by way of a synthesis of the available ethnographic data in regard to bells, heads, faces, and eyes, and the species represented in the imagery in these bell heads. Thereby, we draw out some ideas about the ritual environment of these works of art. It is in this latter context that we feel able to identify these bell heads with the domain of ritual practice known in Benin City as Osun, the deity1 and its associated rituals and arts. Finally, in selecting this group of works we hope that other as yet unpublished pieces in public and private collections will be brought to light.
THE LOWER NIGER BRONZE INDUSTRIES: ARCHAEOLOGY AND METALLURGY
The Lower Niger region (Fig. 2)—from the Niger-Benue confluence southwards to the Niger Delta, with the Yoruba, Ebira, Edo, Edo-related, and Igbo peoples to its west; the Igala, Basa Nge, and Igbo peoples to its east; and the Urhobo, Isoko, Ijo, Kalabari and other Delta peoples to its south—is one of the best-described regions of Africa (Anderson and Peek 2002, Berns, Kasfir, and Fardon 2011, Cole and Aniakor 1984, Plankensteiner 2007, etc.). Yet it still presents us with many unanswered questions, as demonstrated by a survey of what we know and do not know of its various copper alloy casting traditions. Three of these are rather well described, even if still presenting disputed questions of fact and interpretation. To the east of the Lower Niger, there is Igbo-Ukwu, nineth–tenth centuries AD, a series of three archaeological sites that revealed evidence for working in beaten copper and for alloying and casting in bronze (Shaw 1970, 1977). To the west of the Lower Niger, we have Ife, where casting in brass, and also pure copper, took place probably sometime between the thirteenth and fifteenth centuries AD, although if we include the considerable corpus of ceramic sculptures,2 some making use of the imagery such as we find on the bell heads of this paper, then evidence for Ife art begins in the eleventh century (Willett 1967, Drewal and Schildkrout 2009). In Benin City copper alloy casting was probably already well developed by the fifteenth century, and the practice continues to flourish and develop today (Girschick Ben-Amos 1995, Plankensteiner 2007, Nevadomsky 2005, Gore 1997).
However, we have no archaeological evidence for origins or development of copper alloy casting within these three places and traditions, and while development is clear in Benin City, there is no evidence for change at Igbo-Ukwu and none at Ife other than the appearance in brass of forms already developed in ceramic. There is no archaeological evidence for the demise of Ife or Igbo-Ukwu, and, indeed, no evidence of casting sites or workshops in either place. There is an oral tradition documented in Benin City that looks to Ife as the source of its casting, but we have no idea what this means in any real-time sense and there is no obvious relationship between Ife and Yoruba brass-casting of the present and recent past. Igbo-Ukwu is almost a complete mystery except for the works themselves and the way in which we can draw upon twentieth century Igbo ethnography to interpret one of the three sites.
Then there is Fagg's LNBI, comprising a large number of copper alloy castings from the Lower Niger region, often without any further provenance established archaeologically or ethnographically, and as the research continues so their number increase. The variety of their forms is truly bewildering. So far, Peek's survey of private and public collections across Nigeria, Europe, and the United States has revealed over 900 works: some 376 bells, 134 bell heads, 40 figures, 3 animal heads, 12 skulls, 16 creatures, 20 face masks, 13 pedants, 10 scepter heads, 255 ovo/ofo replicas, 16 vessels, and at least 50 elaborated manillas attributed to the LNBI. Some of these castings can be grouped in terms of form and scale on a geographical basis: from villages grouped around the Forcados River to the west, villages close to the Cross River to the east, and works found in various ethnic locations; but there are a great many castings—such as the hunter with an antelope (Fig. 1) (Fagg 1960) and many of the castings east of the Lower Niger (see Peek and Nicklin 2002)—of unknown origins, some turning up among collections of material to which they do not belong.
Moreover, a survey of the known centers of copper alloy casting from the present and recent past gives us no help. Benin City art is well published (Plankensteiner 2007), and so too are castings from the Yoruba city of Ijebu-Ode (Schaedler 1997:232–42; Drewal, Pemberton, and Abiodun 1989:116–31), while the casting now practiced in Abeokuta and Ibadan are nineteenth–twentieth century derivations from Ijebu-Ode. The forms of Owo casting are best described as provincial variants of Benin City work. Only the small-scale castings of the Ekiti (northeast Yoruba) villages of Obo-Ile and Obo-Aiyegunle come close to the esthetic of LNBI work (see Fagg and Plass 1964:28). There also is evidence of casting in Oyo-Ile (Williams 1974:262–63). The forms of casting in Benin City, Owo, Ijebu-Ode, and Obo are well attested and clearly distinguished from each other and from the LNBI. Benin aside, we have no evidence for the histories of casting in each place. To the east of the Lower Niger, copper alloy casting is known from at least two Igbo towns, Awka and Abiriba (Neaher Maas 1976, 1979; Cole and Aniakor 1984) though evidence relating either to the LNBI and to Igbo-Ukwu is unclear.
Casting is also reported from the Igala city of Idah; at Koton Karifi, an Egbira town north of the Niger-Benue confluence (but without precise evidence for the forms of work in either place); and by the Abakwariga, the Hausa artisans supplying Jukun and other peoples with textiles, wood sculpture, and brass casting (Neaher Mass 2011), but this really is stretching the net far too wide in the attempt to make sense of LNBI.
All that we can be certain of is that within the Lower Niger region there are two sites known archaeologically (Igbo-Ukwu and Ife), two known historically (Benin City and among the Isoko clans [see Peek 1980]), and no more than four or five casting centers at the present time across the Yoruba-, Edo-, and Igbo-speaking peoples. Yet none has any obvious relevance to the great corpus of works provisionally classed with the Lower Niger Bronze Industry, works that in formal terms point to the probable existence of casting sites and masters as yet unknown archaeologically.
One group of works, sometimes included within the LNBI group (Fagg 1970) and sometimes not (Fagg 1960), are those identified as the “Tsoede bronzes” so called because in Nupe myth they were brought by Tsoede, the founding hero of the Nupe state, when he escaped captivity in Idah and proceeded up the Niger, bringing these figures with him. These comprise ten human figures (one is probably from Ife), two “ostriches” and one elephant once located in three middle Niger, Nupe-speaking riverine villages, Giragi, Tada, and Jebba Island. Fagg also attributed to this group a female figure collected in Bida. They are all well known in the literature and need no further comment here. Even though two of them share imagery associated with some LNBI castings, the Tsoede bronzes probably do not merit consideration as part of the LNBI copper alloy mysteries. Nevertheless, they are evidence for yet more casting centers additional to those responsible for the Lower Niger Bronze Industry, those presumed archaeologically, and those known from the recent past.
So far we have used the terms “copper alloy,” “brass,” and “bronze” without explanation, but given that some of the differences of alloy might have implications for dating, it is time to be more precise. The largest element in a copper alloy is, obviously, copper, and we know from countless analyses that there was always a small amount of lead in the alloy, which improved the flow of the molten metal, as well as a number of trace elements, the purposes of which are probably only accidental consequences of the ores mined and smelted in particular places. Molten copper can be cast, using the lost-wax process, and the metal can also be worked cold by hammering. Brass is the alloy of copper with zinc (and a little lead). Bronze is the alloy of copper with tin (and a little lead). And there is also gunmetal, the alloy of copper, zinc, and tin (as always, with a little lead). The addition of tin or zinc lowers the melting point of the copper, thus facilitating the casting process.
If we do not know the precise composition of an alloy, it is safer to call it copper alloy; guesswork here is unwise, particularly when, as in the Lower Niger region, the difference between bronze and brass probably matters for any understanding of the archaeology and art history of the region. The reason for this is simple: the constituent elements for alloying a leaded bronze were all available either locally within the Lower Niger region or not too far away; and this is the alloy employed in all Igbo-Ukwu castings. Copper and lead ores have been located in the northeast of the Igbo-speaking region at a site that shows evidence of having been worked in antiquity, and we know from evidence in Nok Culture deposits that tin was smelted on and around the Jos Plateau. The castings excavated at Igbo-Ukwu must therefore be regarded as evidence for the smelting and alloying of the constituent metals for bronze in that region at that time (Craddock and Picton 1986).
The alloying of brass, however, entails a very different technology due to the fact that zinc is a volatile metal. At the temperature at which copper melts, zinc is already a gas, and the processes of smelting and alloying zinc are complex beyond the capabilities of the preindustrial technologies of sub-Saharan Africa. In other parts of the world, once the technological problems were solved, zinc became the metal more commonly alloyed with copper simply because zinc ores were more readily available than tin ores, whereas in sub-Saharan Africa, although bronze could be locally alloyed, brass was necessarily imported. We know, of course, that brass imported from Europe was the raw material for casting in Benin City from the inception of that trade in the late fifteenth century; but Ife is earlier than this, which suggests some other source for the brass used in casting. And if a local alloying technology was available in the Lower Niger region, why was it not employed at Ife? Is it possible that the local bronze alloying industry had already died out? It must have died out at some point as there is no evidence for its survival into the recent past; but when? We have no answers to these questions.
One of the reasons the Portuguese tackled the problems of sailing around the coast of West Africa was to undercut Moroccan control over access to West African gold, and one of the ways of obtaining gold in Mali was to exchange it for copper and brass. Trade between the Mahgreb and West Africa is well attested historically and archaeologically for the period prior to the coastal trade networks initiated by Portuguese in the late fifteenth century. If, therefore, the brass castings of Ife are earlier than that, the alloy must have been imported from the Mahgreb via trans-Saharan trade. Perhaps the casting technology also came from that direction. The antiquities of Igbo-Ukwu and Ife thus represent two distinct technologies within the Lower Niger region in the period from the ninth to fifteenth centuries AD, one of local inception, the other of trans-Saharan origin, which makes Ife like an island of difference within the Lower Niger region in regard to the use of copper and its alloys.
The Portuguese imported huge quantities of brass and copper into West Africa in the form of bracelets (manillas) with the result that in Benin City brass became the principal source of material for casting. This is clear from metal analyses of its works of art (Craddock and Picton 1986). However, those Benin castings which, on formal grounds, are thought to predate European trade—Fagg's “Early Period”—and for which we have analyses are either made of bronze or an alloy that includes small amounts of both zinc and tin—technically, therefore, a leaded gunmetal. Now it is well known to anyone who has investigated brass casting in the Lower Niger region that one of the sources of the brass for casting is old castings; and the presence of gunmetal suggests that, prior to direct trade with Europe, the casters in Benin City had access to both technologies, quite possibly through the recycling of older castings, i.e., both trans-Saharan brass (as pioneered by Ife) and locally alloyed bronze (as initiated by Igbo-Ukwu). However, with the new sources of brass available through coastal trade, the local technology of alloying copper with tin would surely have been rendered irrelevant.3 If the smelting and alloying technologies represented by Igbo-Ukwu were still in existence in 1485 when the Portuguese first arrived in Benin City, they would certainly have died out very soon after—but, of course, we have no archaeological proof of this.
This is where we must come back to the works of the Lower Niger Bronze Industry, that substantial corpus of copper alloy cast works of art from the Lower Niger region that do not seem to belong to Igbo-Ukwu, Ife, or Benin City. This material suggests two significant points: firstly, on the formal evidence of the works of art themselves there must be several as-yet-unknown casting centers; secondly, all the LNBI works so far analyzed for their metallurgical content were cast in bronze, not brass (or gunmetal). This includes the famous hunter with antelope already referred to (Craddock and Picton 1986). If, therefore, we are correct in suggesting that with the advent of brass the local manufacture of bronze would have rapidly disappeared, then it is necessarily probable that the works of art of the LNBI, or at least those so far tested for their metal content, were cast at some time prior to circa 1500.
Of course, it is possible that old bronze castings might have been recycled after that point, but if that were so one would expect to see the evidence of it in the analyses of castings from Benin City, which we do not. Neither do we see any evidence for the survival of the local technologies in the analyses of Ife castings. In both cases, this is unexpected given the propensity of casters to throw into the crucible anything that was available. Could the advent of trans-Saharan brass via Ife have put an end to local bronze casting? If that were so, one might expect to find more brass casting in the Lower Niger region than in fact we do. We remain at liberty to suggest that any LNBI casting in which the alloy is bronze is probably no later than the fifteenth century, thus attributing these works to a period prior to the likely emergence of the Edo state, and certainly well before modern Yoruba. We are not suggesting any direct ancestry but rather that at a comparatively early date there is (a) evidence for several as-yet-unknown casting centers; and (b) a shared set of ideas and ritual practices, as evidenced by the particular examples of the bell head form and their imagery chosen for discussion in this paper.
INTRODUCING BELL HEADS AND THEIR IMAGERY
As noted earlier, Peek has recorded over 900 LNBI castings, the majority of which are various bell types. In size, these bells vary from around 25 cm high to the small globular bells known as crotals.4 Many display human and/or animal forms. In the region to the west of the Lower Niger, which is where the majority of artworks discussed in this paper are from, bells are cylindrical, conical, trapezoid, or spherical; in section they are circular, ovoid, or square. Most appear to be locally cast in copper alloy using the lost-wax process. Some may be imported from industrial manufacturers elsewhere in the world.
The rest of this paper is directed to a small group of eleven bell heads, many of them with horns and all of them with wild animal imagery that to us looks fantastic, even phantasmagoric. It is an imagery which, in the art of Benin City, has been identified as hostile to man, as we shall discuss below. Snakes, crocodiles, and other reptilian or amphibious species appear from the nostrils or mouth, and one of them appears to be swallowing (or spewing out) a hapless human being. Although this is an imagery found in certain specific contexts in Benin City art, in some Ife works, and in two of the Tsoede bronzes, in the LNBI castings we have forms that simply cannot be contained within the accepted canon of works from Benin City, Ife, Igbo-Ukwu, or the Yoruba. Yet the imagery suggests a corpus of ideas and ritual practices of great antiquity.
Some of these bell heads are known from shrines in the Lower Niger region, while others arrived in European collections from the late nineteenth century onwards without any indication of their origin. A number of works are associated with the flood of material taken from Benin City in 1897 and may or may not have been collected there, even though they are not part of the Benin canon as established by memorial heads, figures, and plaques. Some appeared due to accidental discovery (such as from farming) or possibly illegal excavation of burial sites, judging from the quality of their patination.
The most exhaustive survey of LNBI imagery to date is by Denis Williams (1974:218–60; see also Lorenz 1982), although the imagery on the bell heads discussed here is wilder than anything he treated. Indeed, Williams does not discuss the bell-head form at all.5 He concluded that LNBI castings were probably from Ijebu-Ode, a southern Yoruba casting center that was also known to European trade. There are at least three problems with this attribution. Firstly, the prototypical works of cast copper alloy from Ijebu-Ode are a series of bell heads identified as eighteenth–nineteenth century (e.g., Drewal, Pemberton, Abiodun 1989:122, Fig. 126). Given the vast scholarship on Yoruba arts, it is curious that there has been no focused study of these Omo bell heads, of which at least fourteen have been located in private and public collections (Fig. 4).
These are notably different in form and imagery from the bell heads under discussion here. Nevertheless, the very fact that the Ijebu works are bell heads does suggest that we have to take some account of Yoruba ideas and practices in our consideration of the LNBI. Secondly, there is a consistency of form amongst all known Ijebu castings that simply does not embrace the variety of the LNBI. Thirdly, those few LNBI pieces for which we have analyses were cast in bronze which, as we have already suggested, places them in a period prior to direct European contact and access to the copper and brass that came with that trade. Ijebu-Ode casting is without any archaeological data or metallurgical analyses that might support or deny Williams's hypothesis. This is true also of the other known casting centers of the recent past to the west of the Lower Niger, e.g., Obo and Owo; but at least we can say that there are no formal continuities between the known works from these places and the LNBI.
All we can do at this stage is to reiterate the proposition that there must have been more casting centers in the region and, as Vansina (1984) has suggested, all of these places would have formed part of a network of trade inevitably leading to a complex series of interchanges of forms, ideas, technologies, and works of art throughout the Lower Niger region. If we discount Williams's attribution of the LNBI to Ijebu-Ode, then the works discussed here, as with the LNBI generally, might be attributed to one or more unknown and early Edo, Edo-related, or Yoruba-related (as they are termed today) communities. These could be anything from a city state (comparable to Ife, Ijebu-Ode, Owo, or Benin City) to a small village (such as Obo) in the period of the eleventh–fifteenth centuries AD. Obviously, this depends on the hypothesis that anything cast in bronze is likely to predate European coastal trade (but if a casting were shown to be in brass, that would merely confirm the persistence of a tradition of considerable antiquity).
Before proceeding to a discussion of these bell heads’ imagery, we need to consider ideas and practices surrounding bells,6 as well as heads, eyes, and mouths, in the ritual environment to the west of the Lower Niger. This is a complex region in terms of languages, ethnicities, and so forth (Picton 1997, 2012), yet these differences can be balanced against the many ideas and practices shared across the communities in this region. This is evident in ideas about bells and human heads, especially the significance of faces, eyes, and mouths, and how the properties of things can be actualized and directed through appropriate ritual treatment. When all this ethnographic data is brought together with the imagery of the bell heads, a sense of purpose is clearly suggested. Of course, we can have no idea whether today's interpretations are necessarily relevant in earlier centuries. In Benin City, for example, we know that new circumstances demand new interpretations of extant forms (Blackmun 1988, Bradbury 1959, Gore 2007, Picton 1997). All we can do is to make the most of the data we do have in the hope that more and better data will emerge to substantiate or disprove our conclusions.
We take it for granted that bells are meant to be rung, and yet we know that there is more to a bell in its ritual environment than the sound it makes. During extensive research on Isoko religion, Peek realized that although he had seen many bells in shrines, he had never heard them being struck or rung. Rather, they served only as inert, “soundless” shrine objects, although one priest wore his collection of ancient copper alloy objects attached to a wooden carving (see Anderson and Peek 2002:43). This was also the custom among Igbo titled elders and seemed to serve the same ritual functions of warning others of the priest's approach (Neaher Mass 1979, Cole and Aniakor 1984:53). Among the Isoko, the war leaders (the iletu, still an active position in village government) wear bells around their waists as part of their military gear, as do ivie (priest chiefs) as part of their ritual dress.
In 1969, Picton was in the village of Kuroko, north of Okene in the Ebira-speaking region of what is now Kogi State. A young mother passed by carrying a child around whose neck was a long rawhide thong with a small bell attached. Just above the bell, sewn back-to-back around the thong, were two small squares, one of leopard skin, the other of python skin. Picton asked those around what it was for, and was told: “When the leopard and the python wrestle, the bell will ring and the witch will run away.” The witch in Ebira tradition was always a woman of your own household or kin, perhaps secretly jealous of your achievements, who would fly at night to attack your children. Divination, sacrifice, and especially traditional medicine were the available remedies. In this example, the hostile energies of leopard and python were directed against a witch intent upon afflicting the child, the bell ringing out as the mother moved about, evidence of the protective aura with which she had surrounded her baby.
Bells were a frequent component of medicines because “nothing can stop a bell from ringing.” The special buffalo tails carried by lineage chiefs to signify their status as keepers of the ritual properties inherited from lineage founders often had bells attached. The preserved tail of this most dangerous animal was evidence of hunting prowess, accumulating power in the hope of protecting the hunter's descendants; and just as nothing would stop a bell from ringing so the ancestor would remain present in his descendants, his lineage would prosper and always be heard about as the bells on his buffalo tail would ring, spreading the aura of his protecting presence. Bells might also be worn by masked performers as part of the medicines hung secretly around their necks for protection from witches and rival performers as well as serving to emphasize their movement and advertise their presence.
Bells command attention, emphasize movement, protect individuals and lineages, and announce a presence (Lorenz 1982). In many places titled men, elders, or priests carry bags of ritual significance with bells attached, which serve to warn of the approaching bearer of power and possible danger. In Benin City square-sectioned brass bells, usually with faces modeled in relief, were placed upon ancestral altars and worn by titled men and warriors, described as “… for protection in battle and to announce their victories on returning home …” (Girshick Ben-Amos 1995:84).
Among the Ijo, even wrestlers wear bells. On ancestral altars bells might signify the continuing presence of the deceased, but they might also announce to the ancestor that sacrifices are about to be performed. In the southeast Yoruba city of Owo, bells were also placed on ancestral altars, worn as chiefly regalia especially by those involved in military activities, and attached to ceremonial swords. In Ekiti, square-sectioned bells, often with faces, were attached to ceremonial swords. The idea that nothing can stop a bell ringing rings true in all these contexts, even when the idea of the bell does not demand that a bell must be sounded. Among the Isoko, where bells are often “silent” shrine ornaments that are never sounded, serving instead to honor the shrine's ancestor or deity, the idea of the sound of a bell remains.
Whether the quality of the sound a particular bell makes, depending on whether it is forged of iron or cast in a copper alloy and depending also on its size (the larger the bell the bigger the sound) has any significance, we do not know. In some parts of southern Nigeria, people banged manillas together to test the quality of their metal by their sound. We must also remember that the bells we see today would have been highly polished, unless they were the recipients of blood and other sacrifices. The patination that either comes from burial in the ground (or from simply not keeping them polished, as in Western collections) is not appropriate in a ritual environment where polished brass, bronze and copper, carry two key significances. Firstly, shiny things are apotropaic, i.e., they have the capacity to reflect evil back to its sender. Secondly, in all languages of the lower Niger region, the range of colors from yellow through red to brown are within the local definitions of the category “red.” For “brass is red in color and this is considered by the Edo to be ‘threatening,’ that is to have the power to drive away evil forces” (Girshick Ben-Amos 1995:88, 108). It is the association between red and warfare, blood, and fire that makes it hostile to the enemies of the person, the king, the empire, and thereby protective for all those within its aura.7
Copper and its alloys intensify these powers, making them ever-present—now represented in metal, they are made permanent. Therefore, ritual items made of copper alloys (and those copied in copper alloys) carried special significance by being bright, hard, and red—just like the coral associated with the Oba of Benin, who controlled the production and ownership of coral or copper alloy artifacts.
HEADS, FACES, EYES, AND MOUTHS
One is immediately struck by the diversity of features and forms of the 134 bell heads Peek has recorded. While a few are clearly human, an extraordinary number are barely “humanoid,” with horns, dramatically furrowed brows, bulging eyes, and myriad facial decorations including various creatures. But the clear presence of basic facial features remains a constant.
A person's head is identified as somehow the image and representation of both character and destiny, although the details may vary from one place to another (Abiodun 2014, Bradbury 1961). When Picton asked an Ebira elder why he was forever wishing someone a “good head,” the reply came in the form of a question: “Which part of a baby is [usually] born first?” Sight, hearing, and speech are all located in the head. In other words, the head is, more than any other part of the body, that with which we encounter and know someone in the worlds beyond ourselves.
When we turn to lower Niger ideas about eyes and faces, we note first of all that in many of the languages of the region the word for “eye” is also the word for “face” (see, for example, Abraham 1958, Agheyisi 1986). In the ritual environment of a Yoruba shrine, the place where sacrifice is made is known as oju'bo, “the eye of the sacrifice.” It is not thereby the eye or face of the deity to which the sacrifice is made. Rather, it is the notion of the eye (or face) as point of mediation between deity and supplicant, and between character and fate. Moreover, if these bells were intended to mediate a power into the world that was both protective and hostile, it may be significant that Benin City Osun heads in London, Berlin, and Dresden collections have snakes emerging from the inner corner of their eyes. Given the dangers of such creatures, this representation surely highlights the power of the “gaze” of artifacts, deities, and priests.
Most easily observed on masks, open or closed mouths play a major role in characterizing the entity being portrayed. As we shall see, this distinction is a major feature of Benin court arts. An open mouth, often displaying large teeth, is most associated with wildness, with danger and impropriety, whereas a closed mouth demonstrates civility, decorum, even wisdom (Peek 1994). Formal contemporary portrait photography also reveals a preference for a closed mouth. Just as the closed mouths of Benin heads reflect the calm and wisdom of the Oba and titled elders, so the open mouths of our featured head bells portray wildness and danger. In fact, all of our featured head bells have open mouths. Nevertheless, a closed mouth might also imply temporary control of destructive powers, as seems to be the case with the Osun heads we will discuss below.
If, then, we bring the ideas about bells, heads, faces, eyes, and mouths together, the questions are obvious. Firstly, could there be something of the mediation between seen and unseen worlds, and something about the destinies of the individuals and social groups to whom the bell-heads considered here belonged embodied in these forms and their imagery? Secondly, in a world in which divination, sacrifice, and magical medicines have such obvious force and effect, how might these forms and their imagery also shape the destinies of those to whom they once belonged?
THE RESONANCE OF OSUN
The eleven phantasmagoric bell heads we have so far identified, and which are the subject of this paper, could be described as “monumental” in appearance; yet, in the context of Lower Niger casting as a whole, they are relatively small, between 13–25 cm in height. They include two key historical examples still in Nigeria, the “Avbiama” and “Enowe” bell heads to be discussed later (see Figs. 13–14).
Others are located in the Museum of Fine Arts Houston, Yale University Art Gallery, Krannert Museum (University of Illinois at Urbana-Champaign), Musée Dapper, Monika Wengraf-Hewitt in London, Charles Davis in New Orleans, and other private collections in the US (Figs. 4–12).8 On the basis of form and style it is not hard to see that each is the work of an individual caster, possibly located in different towns or workshops. On the other hand, they share a number of features suggesting that each is part of a common “universe” of ideas and ritual practices.
Seven of the bell heads have a pair of what appear to be short animal horns. Such small horns are often used as containers of magical medicines, although Curnow considers these to represent braided hair decorations (2007:176–77) and extends this interpretation to the “humanoid” bell heads. Nevadomsky notes that in Benin City all bells with horns are called “Igara,”9 but whether this is a reference to Igara in Akoko-Edo, a town of Ebira origin that chose to identify itself with Igala, or directly to Idah, the Igala capital, is unclear.
Eight of the bell heads have something at the center of the forehead which in some cases might represent a prehistoric stone axe head, found all over West Africa and habitually interpreted as marking places where lightening has struck the earth. In Benin City, stone axes are associated with Ogiwu, the deity of death (see Girshick Ben-Amos 2007) and relate here to the uses of medicines to both kill and cure. Alternatively, on many of the bell heads a small calabash, as used for containing medicines, is shown on the forehead.
Most of the bell heads have heavily furrowed or otherwise emphasized brows over often bulging eyes without pupils, and all but one (Musée Dapper) have a decorative bar or strap across the nose bridge. All but two have flared nostrils with snakes emerging, which curve across the cheeks towards the ears. In the art of Benin City, snakes signify the ritual energy directed towards enemies by representing “the actual curses sent by the herbalist against enemies” (Girshick Ben-Amos 1983:51). As well, snakes are the “warriors of the night people” sent by Osun, the Edo deity responsible for the powers of medicines to kill others (Girshick Ben-Amos 1976:250), and even more dramatically, “The representation of snakes issuing from orifices refers to the belief that those who have occult powers vomit snakes when wishing to destroy their enemies” (Girschick Ben-Amos 1999:71). Nevadomsky offers detailed notes on the uses and meanings of various creatures in Benin ritual traditions, with representations of poisonous snakes symbolizing their dangerous potency as messengers and that “most dangerous snakes represent Osun, the god of medicine” (1988:82) whose devotees must know poisons in order to know antidotes. As an Isoko priest once announced to Peek, “I am Kill and Cure!”
In all but one example (Yale; Fig. 6) mouths are open, framing large bared teeth, four of them biting or vomiting a frog. On the Wengraf-Hewitt example (Fig. 9), there is a human form at this point holding what appears to be a mudfish in one hand and a sword in the other; and on each side of the mouth there is a crocodile.
With regard to the amphibians and reptiles emanating from facial orifices, we habitually interpret such creatures as liminal beings which move between worlds, between land and water, between observed and hidden domains, thus carrying great potency. Benin Osun shrines are known for their large bubbling pots of medicines full of leaves and creatures from the forest and watery domains, including frogs.10 Gore reports the presence of frogs in the Osun pots in which sacrifices were offered for ceremonies at which the priest danced the “frog dance of power.”11
Frogs are frequently depicted in Benin City arts and they are often found on other LNBI bells (not discussed here).12 According to Girshick Ben-Amos:
… frogs are considered representatives of the ihen, those famous warriors and magicians who transformed into rivers and other natural features. When a king is installed or must prepare for war, frogs are brought to the palace from every river in the realm and are placed on the royal shrine of war magic, Osun Okuo. The water in which they are kept is used by the warriors for ritual bathing (1999:108).
On the bell heads perhaps it is not so much that the frog as such has any power, but rather that, placed in or around the nostrils and mouth, the images represent specific ideas about the powers associated with the bell heads. In these representations emerging from facial orifices we are meant to understand “the supernatural power inherent in a powerful man's very breath” (Curnow 2007:177).
In other words, the bell heads of this paper fit well with the ideas and practices associated with the Edo-Benin cult of Osun, in which powerful substances are brought together as directed by the priests of Osun, thus providing their devotees with protection and power and the abilities to harm and destroy their enemies. Perhaps reflecting the centrality of Osun's powers, amphibian and reptilian species are found on many different cast objects in Benin City. For example, Willett (1973:11) illustrates two small face masks which share many features with the bell heads discussed here.
Most importantly, there are at least three cast copper alloy heads replete with the same motifs we have reviewed on the bell heads representing Osun which were created presumably for altars within the Oba's palace. For the British Museum example, see Figure 15; others are located in Berlin and Dresden. They have many of the features of the bell heads: snakes, frogs, axe heads, and so forth; but they are also surmounted by the birds known in Edo as ahianmwen-oro, the bird of prophecy or evil omen (Nevadomsky 1988:14), that relate to a mythic episode in the reign of the sixteenth century king Esigie. Prehistoric stone axes are found on most shrines in Benin City, in honor of a deity or ancestor, honoring Ogiwu as well as Osun, the deities of death and of medicine, and the ability to both kill and cure. Intriguingly, while these heads share most of the bell heads’ motifs, their mouths are closed and facial features are rather similar to Benin court art in their being more “naturalistic.”
It should by now be evident that in their features, Osun cult objects and the LNBI phantasmagoric bell heads share imagery such that we can infer a shared range of ideas and ritual, and we feel confident in assuming that the bell heads were parts of cults closely related to what we know of the Benin City cult of Osun in the present and recent past. Working from fewer examples than we have been able to identify, both Lorenz (1982) and Curnow (1997, 2007) have previously proposed links of art works with these reptilian and facial features to the service of Osun in Benin City. In that case, this imagery, taken together with the antiquity of these bronze castings, presupposes the antiquity of a ritual environment now represented by the Edo-Benin cult of Osun, in which powerful substances are brought together by Osun priests to provide their devotees with protection and power as well as the abilities to harm and destroy their enemies even though none of these Osun bell heads have been found within the pretwentieth century core of Benin City.
In addition to the Osun heads, there is another group of cast brass works from Benin City that may offer support to our hypothesis linking the bell heads to Osun traditions: the Ododua masks performed during the Ague Osa Festival (Curnow 1997). These masks date from the late seventeenth–early eighteenth century revival of the economic and military success of Benin, marking the onset of William Fagg's Late Period. Ododua is the Edo version of Ododua, the founding hero of Ife civilization, and a new set of masks came into existence to reiterate the royal dynasty's link with Ife. These masks incorporated reptilian imagery and bared teeth—hitherto a very un-Benin-like feature—and, indeed, Curnow (1997, 2007) and Nevadomsky13 both note the suggestion that these masquerades had their origins outside Benin City in Edo villages. Such a reorientation also correlates with Osun's outsider/forest associations.
The argument in this paper for the antiquity of the ideas and practices represented by the bell heads and by the aspects of Benin City art already discussed is further supported by two features of the technology of these bell heads: their metallurgical content and their core forms. We only have analyses for the Houston and Wengraf-Hewitt bell heads via the X-ray Fluorescence method (Figs. 5, 9). Both are almost pure bronze with no more than 1% made up of several trace elements. The Wengraf-Hewitt bell head is 89.75% copper and 9.53% tin, while the Houston example is 84% copper and 15% tin. This, if the argument presented earlier is accepted, places them in the period between the Igbo-Ukwu inception of bronze casting and either the trans-Saharan brass casting of ancient Ife or the arrival of European scrap brass with coastal trade from the late fifteenth century. It is possible that they are contemporary with Benin City casting in the period before European coastal trade, yet the relative purity of the alloy suggests that they are much earlier owing nothing to ancient Ife.
The second technological feature is the careful modeling of the internal cores of each bell head around which wax was applied in the lost-wax technique. While decorative aspects of LNBI works are relatively delicate, the relationship to types of wax remains unclear (see Williams 1974, Neaher Mass 1979). The careful modeling suggests a relative scarcity of alloy driving the skill of the caster, much as we can see in the earliest Benin City castings. Indeed, the bronze in some of these bell heads is so thin as to be beyond the skills of modern-day casters. In other words, both the visual and technological data presented by these bell heads suggests a very great antiquity for the ideas and ritual practices associated with the Benin cult of Osun as it has been described in present-day research.
Weights are not available for most bell heads, but there seems to be a large variation, from rather heavy pieces to those very thinly cast. None have clappers but most have holes created for such. Five of these bell heads have the same form of looped handle with which they could be rung and/or attached to something else.
Although it appears that these bell heads must have come from several as-yet-unknown centers in southern Nigeria, we cannot rule out the possibility that one of those centers might have been Benin City, perhaps before access to gunmetal (see discussion above) and the evolution of early period court casting. After all, we hardly need to leave the city to find evidence for a LNBI casting still in its local ritual environment. One of the bell heads included here was first recorded by R.E. Bradbury in Benin City (Fig. 14) in a shrine dedicated to “a legendary giant named Enowe.” This may refer to “General Enowe of Ughoha from Isha clan; a very powerful warrior, fought many wars with Ozolua about 1481–1504 AD and Oba Esigie about 1504–1550 AD” (www.edoworld.net).14
Another bell head, still in its ritual environment, at least until recently (it is now in the Lagos Museum; see Fig. 13), once belonged to the Onogie of Avbiama (Willett 1973:13, Gillon 1986:266), originally a village outside of Benin City now incorporated into the ever-expanding city. It is one of the two janus-headed examples discussed here (see also Fig. 8). Of course, the location of a work of art in a given place proves nothing, in the absence of further data, about where it might have been made: artists and works of art do travel independently of each other. As with the Ododua masks, Fagg did wonder if these objects indicated influences from the margins or even beyond the Edo kingdom (Fagg and Plass 1964:123).
The fluidity of casting practice is evidenced by the association of bells and warriors as already discussed, and the tradition of casters traveling with armies to repair weapons and armor (Bradbury 1961:133); perhaps there were casters who chose to stay outside of Benin City. And much of the potency of the Osun cult also depends upon elements from the forests and elsewhere beyond the ordered city!
Vogel (1987:102) illustrates a cast copper alloy pendant mask with a frog emerging from the mouth, and horns and reptilian figures on the sides, suggesting that the more we know about the art of Benin City, the more we are reminded of all those tantalizing references in Benin history of other and earlier settlements around the Niger-Benue confluence, all of which work against any simplistic contrast between royal court and village art traditions that has informed our interpretation of southern Nigerian arts.
The interconnections then are numberless. There are not simply one or two threads running through the history of bronze-casting in the Niger Valley which need only to be unraveled. We have rather a piece of woven cloth with many threads in both the warp and the weft which will take a long time to sort out (Willett 1967:173).
As if following on from this, Vansina (1984) suggested there must have been several political, trading, and art-making centers in the region west of the lower Niger, each with its forms and traditions, each interacting with others through the medium of trade and warfare. The diverse forms of the works of art classed within Fagg's provisional category of the Lower Niger Bronze Industry confirm that (a) there were many artists probably working in many centers, none of which is yet known archaeologically; and (b) these traditions owe nothing to the archaeologically better-known city of Ife. Precisely what is the relationship between the Lower Niger Bronze Industry and the emergence of casting in Benin City, Owo, Obo, and Ijebu-Ode remains entirely speculative. Moreover, (c) the fact that those few castings within this corpus of material for which we have metallurgical analyses are cast in tin-bronze suggests an antiquity that precedes the advent of European brass through coastal trade from the late fifteenth century onwards. Then, (d) we turned our attention to one small group of these casting, bell heads displaying phantasmagorical imagery. By bringing together what we know ethnographically from the lower Niger region about bells, heads, faces, eyes, and animal imagery, we were able to suggest that we have evidence for ritually intense protective strategies in an era of political and mercantile conflict that lie behind the emergence of modern states and ethnicities such as Edo and Yoruba. Finally, (e) in a comparison of the imagery of these bell heads and the imagery associated with the Edo cult of Osun, we can see a continuity of forms that might suggest a continuity of ideas and ritual practices through a period of considerable antiquity, justifying the title of this paper: the resonance of Osun through a millennium of Lower Niger history.
We hope these conclusions, though entirely tentative, will prompt the revelation of further bell heads and other castings in private and public collections. The Lower Niger Bronze Industry has for too long remained peripheral to discussions of copper alloy works from Igbo-Ukwu, Ife, and Benin City. As this research continues, this once-miscellaneous grouping, though having William Fagg's enthusiastic blessing first bestowed in the 1960s, may well be a precursor to these other, now better-known traditions. As we began with Fagg's thoughts on the LNBI, so shall we end:
At least one more important centre remains to be identified—more probably several, for although all bronzes in the group show affinities with some others in it, they cannot be said to form a single homogeneous style. What they have in common is an imaginative freedom not found in Benin or even Ife work … (1960).
We very much appreciated the aid of our colleagues in this project, especially Charles and Kent Davis, Bernard de Grunne, John Kowalchuk, Barbara Plankensteiner, Julie Hudson, Doig Simmonds, and Paula Girschick Ben-Amos. Peek wishes to thank the Smithsonian Institution for a very helpful Research Associateship.
When spelled without proper diacritical marks, this name appears the same as the Yoruba river deity, Osun; but they have nothing to do with each other.
The commonplace use of “terracotta” is confusing given that the clay used for sculpture in West African antiquity is the same coarse-textured material used for domestic pottery, quite unlike the refined clay used by sculptors in Europe.
Much as the local technologies for smelting iron from local ores was overtaken by the availability of scrap iron after the inception of colonial rule.
Denis Williams (1974:269ff) extensively investigated the crotal which he believed was introduced in West Africa in 1590. Although it is one of the few attempts to look systematically at bells in Nigeria, this aspect of his study is not relevant here.
Although Williams (1974) was familiar with Fagg's LNBI, many of the works discussed here were simply not known about at that time.
Too often studies of African rituals and arts remain visiocentric and fail to appreciate that no ritual activity takes place silently. There are always sonic elements, especially when initiating sacred acts at altars and these usually involve bells or gongs (see Peek 1994).
The color red is important throughout the region, though its precise value may differ from place to place.
In a time of extreme concern over “authenticity,” we must allow that we cannot absolutely confirm these bell heads to be “authentic,” although we suspect they are valid works of art and undoubtedly represent the cultural complex of Osun.
Joseph Nevadomsky, personal communication, 2014.
Joseph Nevadomsky, personal communication, 2014.
Charles Gore, personal communication, 2014.
Joseph Nevadomsky, personal communication, 2014.
Egharevba (1973:12) in his Benin Museum catalogue mentions “pieces of [a] bronze girdle” which Enowe wore for protection in war.”