In Akropong, capital of the small kingdom of Akwapim in south-eastern Ghana, a special name given to a child by his mother, a particular patch of ground, and a chief's gold-covered linguist staff (akyeampoma) are used in ritual performances to tell indirectly of painful events of the past. This essay examines what Akropong people say without saying by means of these ‘sites of memory’. At issue are memories of past events – personal catastrophes, violent deaths and political disruptions – whose ramifications must still be confronted. These memories are accessed by verbal and visual strategies that have enhanced emotional impact but are allusive in reference. Formalized, repetitive, and felt to be obligatory,Footnote ¹ they refer to events that can neither be denied nor successfully repressed, as it would be impossible or immoral to do so. They need to be forgotten and remembered at the same time.

A vast academic literature exists on memory and ritualFootnote ² – in part, perhaps, because memories and symbols of the past are anchored and made palpable by ritual. Connerton captures the appeal of this long-standing theme when he says that in ceremonies that commemorate continuity with the past, a community is reminded of its identity as told in a master narrative and that this ‘image of the past … is conveyed and sustained by ritual performances’ (Connerton Reference Connerton1989: 70). Greene illustrates this very well in her analysis of the Anlo Ewe of south-eastern Ghana, whose identity has been ‘shaped’ by the narrative of their exodus from the early sacred site of Notsie and by its commemorative ritual enactment.Footnote ³ She examines the ‘material sites that support our mental images’ (Greene Reference Yankah, Cohen, Miescher and White2002: 8) and shows that Anlo memories of Notsie have had divergent meanings due to the power dynamics of the colonial encounter – the collective memories have been creatively ‘refurbished’ and secularized. The site may thus be considered ‘contested terrain’ (2002: 131).

The two cases I describe in this essay – I look at children's names, and then at ritual sites and regalia – are examples of such ‘commemorative ceremonies’; they incorporate multiple pasts selectively re-moulded. They centre on (1) how an individual relives or is reminded of an event in his or her past and (2) how the kingship as a whole is imagined by means of material artifacts and social practices. They do so in ways that are indirect and deniable, and it is clear in each case that the past is contested and painful. Indeed, the events might be thought of as re-enactments of painful real-life experiences for the purpose of gaining mastery over them (see Freud Reference Freud1920). The ramifications of the cases differ in temporal scale (near past versus distant past), but both allow for dissonant voices in the form of counter-narratives; both importantly involve a tension between remembering and forgetting so that the past remains unsettled and unresolved, and the future uncertain; and both concern death and separation (whether literal or metaphorical).

In Akwapim, names and regalia operate as ‘vehicles of memory’ (Barber Reference Barber2007: 24) in which the verbal and the material interpenetrate. More than that, speech that is ‘fixed’ in a proverbial name gains the ‘compact solidity and durability of a material object’, and regalia function as a ‘memory-prompt’ that ‘traps a text in material form’ (2007: 75). Barber illustrates with great clarity that oral literatures in the form of praise epithets, proverbs and poems are not only attached to material objects, but actually ‘function in the same way as material objects, as part of a … field of reminders and allusions’ (2007: 84). She argues that ‘obscurity’, as in the enigmatic language of proverbs, presents the text as something that ‘demands attention’: this makes it powerful (2007: 79). Her discussion of memory and obscurity in African oral texts (2007: 89–92) is apposite, for the stylized mode of alluding to an insecure traumatic past embedded in an individual's name in Akropong mirrors the formalized allusive references to chieftaincy affairs in carved regalia. Each, in a different genre, points to the authority and (dangerous) power of the past, and in each the actual content of the allusion is hidden or known only to a small handful of people. Such ambiguity of meaning, or ‘indeterminacy’, is useful in situations of personal tension and antagonism and in the political insecurity of a notoriously divisive and conflict-laden society such as Akwapim.

The first case concerns the pain of an individual's past distress. In the town of Akropong, until early in the twentieth century, a mother often gave her child a ‘name-response’ that referred to a traumatic event in her personal history (though she refrained from informing her child of the referent). For these mothers, all present experience resonates with memories of that intensely remembered past. They cannot speak of it, yet simultaneously seem compelled to be reminded of it.

The second case pertains to dynastic history and royal objects. It concerns the public memory of defeats and sorrows in the distant past that are commemorated in an annual ritual cycle and sited in the landscape and in a staff of office. The rite, Odwira (purification), which is actually a ‘spatial story’, neutralizes the malign influence of the past that continues to endanger the present, and displays the present-day power of the king (and his obligations to his people).⁴

A dissenting voice in Odwira is situated in the Asonahene's gold-covered linguist staff (akyeampoma: from ɔkyeame: linguist; poma: staff) that is displayed during Odwira even though its carved image alludes to a subversive historical event concerning the king that no one will openly acknowledge. The staff belongs to the Asonahene, chief of the Asona clan (abusua) and descendant of one of the first Akwapim kings. Its image contradicts the official dynastic history, the ‘master’ narrative of the ruling elite. It asserts an alternative history acknowledged in Odwira, but dismissed. Political opposition to the Akwapim king(s) is generally not ‘narrativized’: the risks are, and were, too great. This is a case of remembrance by the perceived political victim and deliberate forgetting by the victor; of a lineage commemorating what the king deliberately suppresses; of the relation between what one can say and what one can show; perhaps of the attempt to minimize or forget defeat and ignominy, and find solace in a magnified past. That the Asonahene wishes to assert his original position, but deny the reason for his banishment, silently and further enhances the unstable character of this event. Werbner asserts that buried memory produces ‘unfinished narratives’ that call out repeatedly for public resolution Werbner (1998: 9; see also 1995: 102, 106). But here we have an example of wanting to remember and forget at the same time. Again we are concerned with partial and allusive memory, here in the form of objects that say and don't say at the same time.

I interpret ‘sites of memory’ quite literally, as would the Twi-speaking people in Akropong. In Twi, the word kae means ‘to remember, recollect, remind, call to mind’, and nkae dum (memory-thing) is a monument or memorial. The dangerous, insecure history of Akropong people – full of wars, human sacrifice, illness, slavery, witchcraft accusations, and bitter political divisions – is inscribed in the geographical sites that are marked in the Odwira rite. It is displayed in regalia such as gold-covered staffs with carved ‘readable’ proverbial motifs on top, and in name-responses that were ‘scripted’ for her child by a mother to immortalize her risks and tragedies, and that are like visible proverbs and allusive metaphors. These memory-things allude to a history that should not be openly discussed; they assert a past but don't really comment on it. And their interpretation leads to other reminders and other texts (see Barber Reference Barber2007).Footnote ⁵ Readable objects – memory-things, nkae dum – are especially suited to indirection. They are infinitely multivalent and thus malleable, deniable and safe, but at the same time capable of telegraphing rebuke and causing pain.⁶

Oaths are also object-like verbal markers of painful events. The incidents they mark are alluded to indirectly by reference to a time or place. Oaths (ntam) are sworn in order to assert a truth or institute judicial proceedings; they too say and don't say at the same time.Footnote ⁷ Oaths allude to calamitous events in the life of a chief or his forefathers, but do so cryptically and allusively – often by citing the place or day of the week where a terrible event occurred. There are no oaths with ‘happy associations’ (Mensa-Bonsu Reference Mensa-Bonsu1989: 269). When swearing a family or royal oath (ɔka ntam) one deliberately refers to a tragic incident which under normal circumstances should not openly be remarked upon. When a new chief swears loyalty to the king in Akropong, it is abbreviated to ‘I say …’ (me ka) without even naming the state oath, Wukuda ne Sokodee (Wednesday and Sokodee).Footnote ⁸ And when he so swears, those around him are so pained to hear him ‘say’ it, that they pity him and reply ‘Elder, my condolences’ (Akora due).Footnote ⁹ Akropong people respond to the swearing of an oath by trying to obliterate the memory and even the event itself with pacification whose intent is to erase the memory and to free the victim from any potential repercussions for reminding the state of this evil thing.¹⁰

It is forbidden to name the terrible event hinted at in an oath, yet it is mentioned in order to show the sincerity of the one who swears. The memory of the tragic event is never allowed to disappear from the collective memory: the event is alluded to by indirection and thus made all the more powerful, and because it is condensed the details often become lost. Perhaps this reflects the perceived power of words, for to say something in Akropong brings it alive. If a chief is slandered, anyone who repeats this is himself deemed to be insulting the chief.Footnote ¹¹ Rattray (1927: 206) found similarly that in Asante it is implied that the calamity would repeat itself. Rattray also reports the reluctance of Asante ‘to disclose the “dishonourable” pages in the local histories, e.g. the slaying or defeats of Kings or Chiefs in battle … lest the spirit of his dead ancestor should rise up and take revenge because his agony was thus recalled’ (1929: 128, emphasis added). I suggest this is what lies behind the taboo (mmusu) against openly mentioning these dreadful events: it is a sacrilege to remember them, and yet they must never be forgotten.Footnote ¹² I turn now to the cases.

FIRST CASE: CHILDREN'S NAMES

Names both state and confirm membership in particular social groups and localize such membership in the individual. They are possessions, rights to which are bestowed on individuals by their guardians, as with other property. The notion of individual identity is hugely problematic. A name resonates to the person, but it is de-personalizing. Particularities are effaced in the convention of naming. The function of the name is ‘to transform the person into a character’ (de Heusch Reference de Heusch1973: 236); to give identity. Every Akropong person has a number of names: some reflect the day of the week on which he or she was born (kra din: soul name), others the group of patrifiliation (din pa), and so on. As there are a limited number of shared names, there is also a measure of anonymity and a potential for confusing rather than distinguishing identities (across time and space).Footnote ¹³ Names and their appellations are used in different social contexts to display unity and consensus, to praise or cajole, to criticize obliquely, or to call into question a person's social pedigree and/or reputation. Names, finally, are an integral part of a highly stylized reciprocal greeting formula: not only is there a choice of names to call upon, but courtesy demands that a name never be heard without a complementary and formalized response (nnyeso).Footnote ¹⁴ It is generally the father who names and chooses the major markers of identity for his child.Footnote ¹⁵ The naming ceremony generally occurs one week after birth and celebrates the infant's introduction to social life.¹⁶

Until the early 1920s, many an Akropong mother, when troubled by an unbearably traumatic experience, would demand of her child that he or she respond to her every call by reciting a particular proverb. This metaphoric response was a disguised message that concretized the mother's feelings: it named her pain and thus exposed it; it was a way to remember but also to forget and perhaps to ward it off; it maintained the presence in absence. These responses focused on maternal history: the mother was getting her child to express what she wanted to say. We are concerned here with memories that are ‘constructed, staged, and self-centred’ (Lowenthal Reference Lowenthal1985: 207).

These proverbial responses (tete nnyeso) resemble the metaphoric names that were formerly bestowed on pets, slaves and villages in Akwapim, and they correspond to a type of personal name, noted by Levi-Strauss (Levi-Strauss Reference Levi-Strauss1966: 181),Footnote ¹⁷ in which an individual ‘classes’ or identifies her own subjective state under cover of giving a name. They also resemble the names of patterns on wax-print cloth worn today in Akwapim by men and women in their daily life to indirectly communicate anger, hostility, lack of commitment and so on.Footnote ¹⁸ In all these genres, proverbs (or pithy expressions) are used to signal incidents concerning individual identity and emotions; they provide neutral authority, but, being metaphors, convey meaning with a degree of indeterminacy and uncertainty that allows for disavowal by the principal.¹⁹

In Akropong people take care not to tread upon the emotions of others; they celebrate consensus and harmony. Saving face is important.Footnote ²⁰ In public, much dissension is expressed indirectly: the object of the criticism is aware of it, but cannot take offence and thus future alliances are left open. The concepts bɔ akutia (to chide, scold or slander someone without mentioning his name) and di nkasaguaa (kasa: speech, guaa: to drive away; speech to provoke or openly chide without naming) refer to speech meant to say something indirectly, publicly, without mentioning the person's name, though wanting him to know (Christaller Reference Christaller1933: 276; Yankah 2001: 232–3). They are devices to control the potency and danger of the spoken word (Yankah Reference Yankah1995: 50–1). Lakoff and Johnson call this ‘deniable’ speech.Footnote ²¹ It is similar to the use of proverbs (ebe) as names and as a response: conscious, ritualized, denial of what is known in order to maintain an illusion of something other.²²

None of the various names that each Akropong person possesses indicates his matriclan. But the mother/child relationship is intimate and the allusive fusion was until early in the twentieth century often marked by a special kind of name that was actually a proverbial response (tete nnyeso). This stylized response was demanded of a child, usually a girl, by her mother (or on rare occasions by her father) and was used by the child all her life. Most of these responses alluded to a traumatic experience suffered by the mother (who chose the response), an event which ‘agitated’ her mind, such as a death, extreme poverty, hardship, envy, neighbourly spite, and so on. The child would be told: ‘When I call you, you must respond by saying [for example] “Death has taken my honour (Owu agye me nim)”; or “Domestic problem/issue! (Afis ɛm)”; and you reply in this way to everybody.’ Most of the older women, and a few of the older men, whom I knew when I first visited Akropong in the mid-1970s, had been given such a response, though very few knew their meaning: the response was simply given without explanation, and the child was told to use it automatically. This is important. It suggests that the meaning was of direct relevance to the mother as a means to deal indirectly with a disruptive event in her life.

One way of looking at this, a superficial one, is that the mother herself could appear publicly subservient, submissive and seemingly respectful while simultaneously protesting her lot in life through her child. But there is more to it than this. The child's response ‘scripted’ by the mother is being used to maintain the mother/child connection through a kind of verbal rapprochement or merging, while simultaneously it allows for separation, individuation and privacy of thought.²³

The affectively charged proverbial response that a mother gets her child to voice may thus be considered an example of early mother/child symbiosis. It is both a projection and a disavowal of the mother's anguished emotions. The mother seems to deny what she knows to be unacceptable in order to maintain the illusion of general compliance; this, in turn, masks a resolve to maintain some form of self-determination and expression, and so the latter is deflected onto her child. The child is voicing what the mother wants to voice but cannot. It is a form of splitting.²⁴

When the mother uses her child as an instrument to recount an experience of ‘unspeakable’ pain or anguish in her past, this nullifies and de-exoticizes it and yet simultaneously keeps it alive by incorporating it into the present (if only for herself). ‘Nomination is an evocation of presence and maintains presence in absence’ (Lacan 1954–5: 297 in Richardson Reference Richardson and A.1985: 108). The mother partially deflects or projects her sorrow or anger; she ‘uses’ or restores history or the memory of a terrible event and by means of a very stylized, condensed and repetitive use of words that apply to the disaster, she recreates or recalls the experience through her child in a way that is safe (banal). This is ineffective, or she wouldn't have to keep recalling it. So it is not catharsis. But it is a way to handle the experience, to manage it.²⁵

The proverbial response assigned by a mother to her child is also a way to ward off something that the mother fears or dreads; a highly conventionalized distancing mechanism to protect her or perhaps her matrilineal descendants magically. It is a culturally standardized way for the mother to set aside her upsetting life: it gives her a way to function. By demanding of her child that he or she repeat back her own tabooed forbidden thoughts, the mother is turning passive into active. But if these traumatic thoughts are loaded onto the child, it turns the child into an unwitting source of uncomfortable memory and misery for the mother. The child is made a participant; he or she becomes part of the tragic event that is memorialized for the mother. The children so-named are burdened with tragedy and blame, and often they don't even know the specific source of the tragic event. The mother, in effect, is rejecting her child. It is hostile on her part, but she cannot get gratification out of this – it is not successfully cathartic.

It is fitting that it is the mother who demands these responses, for while Akropong people believe that women possess wisdom – knowledge of descent and by extension of events, both good and bad, within the ‘house’ (yafunu: literally ‘womb’) – women, in a collective sense, are associated with death and they express sorrow more directly than do men (Gilbert Reference Gilbert and Nooter1993, 1994b). By bestowing a proverbial response on her child, the mother is mixing mothering with death. I suggest that the fact that the mother does this makes her guilty – it is her life, her past she is attaching to her child; thus if her children die (and in the past infant mortality was high), death is her fault, so she is the mother of death.

Although people are not all named after those who are dead, one might say that all names are tinged with death, for the ‘out-dooring’ and naming rites only occur when it is clear that the infant is viable and will not ‘return to god’. The newborn is vulnerable and must be protected by herbalists’ ‘charms’. A pregnant woman is said to have climbed a tree and on delivery to have descended; labour is tinged with death and referred to as war. In a way the ‘out-dooring’ (tan fi) and naming rites (abadinto) are tinged with victory over death. But there is more to it than this.

Mixing mothering with death is literally the case if all of a woman's children die within a week of birth, for then the mother will bestow a name on her next child that refers to a social death. Until the early twentieth century she would give him a slave name and his face would be cicatrized like that of a northerner (that is, a slave). In so doing, the mother metaphorically took away her child's identity: she ‘killed’ him so that he would live and not ‘wish to go back to god’ or would not be attractive enough for death (owu mpo soro). In such a case, no proper name (din pa) was given to the child; it was hidden or reserved. The ‘name’ was scarification at birth; and the child was told to answer to a proverbial name – generally one with a troubled referent. By bestowing a name on her child that refers to a social death, the mother was referring to it in miniature: she was giving what Sir James Frazer (Frazer Reference Frazer1935) called a ‘mock calamity’, a mock death, which was then returned to her so as to resolve the death from the child. It is a mock calamity that protects the (vulnerable) child by naming the calamity. It is magical thinking. The mother wards off something she fears – that her child will die and she'll be left alone – by referring to it. In all these cases, the risk is put on the mother. It is a mock calamity if the child lives, but a real calamity if the child dies. As infant mortality in the past was high, the implication is that the mother is guilty. She is the bringer of death, not the nurturing mother one runs to for protection. Should one ask why the newborn or young child is included in this unhappy story, the answer must be to make sure the mother is blamed at the outset for whatever goes on.

I suggest that a mock calamity can be spoken and that these names are possessions, objects to be exchanged. The proverbial name is like an image, and words are being used as things (cf. also names of slaves, pets and villages).Footnote ²⁶ While the processes of naming and the name-responses I have described are deeply psychological, they are not about promoting individuality in the child. Rather, the child is socialized by the content of the message and simultaneously – as there is an indirectly implied fusion with the mother whose words and thoughts have become those of the child – the child is identified with his or her matrilineage.

A few examples follow. Explications vary, much as do interpretations of proverbs. While the ‘owner’ of a response often does not know the specific incident to which his/her response referred, the general meaning can be understood by anyone.

1. (1) Owu agye me nim: ‘Death has taken my honour; or death has seized my glory’. At our village, my elder brother had his own house: he was old enough and he had a wife. His younger brothers lived in our father's house. My elder brother was the leader of all the young men in Akropong. He was very popular: they called him Omanma (child of the state). All the young men followed him in dances; he was very courageous. When he died suddenly, my younger sisters were told to respond like this because the family thought their glory was gone. My elder brother died when I was very young. My mother died from fasting when she mourned this son. She thought her honour was gone, so she refused to eat for three weeks, and then she had stomach trouble and many herbalists came but they could not save her. When this important son died and my mother died too, my father started to drink. So when my elder brother died, my sister when she was called replied Owu agye me nim: ‘Death has taken my honour’ or εnsa me yam: ‘The pain will never leave’, or Owu yɛnya: ‘Death makes people ashamed’. My sisters and nieces responded this way as long as they lived. I too was told to say this when I was a child. It is not done today. All these expressions centred on the death of my mother and brother. All are about death. This man was very promising and he died suddenly. Before his death, we were somebody; after it we were nothing. In those days everyone respected the sisters and brothers of this man. Everyone loved and admired him. We were in a good position with such a man as a brother. When he died, ‘Death snatched my honour.’ (Elderly man, Akropong, 1977)

2. (2) Owu y ɛ nya: variously glossed as ‘Death gets hold of you’, ‘Death makes people ashamed’, or ‘Death makes people laugh at you, belittle you, ridicule you’. (Nya: get hold of; manya wo: I have got you. Nya has the implication of slavery; see Christaller 1933).

   Suppose your father has money, so you behave as if you have nothing to fear: you are arrogant; you know you have a support. Then he dies. So you are broken, without a support and people can laugh at you. In this expression, death is nya. Before the death of this person, people thought you were someone important, valuable. You were important because he was there, but now he is gone and all the benefits which would have come from him had he lived have stopped. Nya means to catch or trap someone, an animal or man. Death is like a trap: it lets your enemies get you. (Elderly man, Akropong, 1977)

3. (3) εnsa me yam: literally, ‘It will never leave the inside of me’, that is, ‘It will never be finished; I will never forget it’ (εnsa: negative of sa: finished, exhausted; yam: womb, or stomach, a shortened form of yafunu).

4. (4) Yε aboa yie: ‘Do good to an animal' (they are more grateful than men)’.

In the 1970s a number of elderly men and women who had been given these name-responses were still alive, but younger Akropong people no longer relived their past and managed conflict in this stylized and indirect way.Footnote ²⁷ What remained and in some ways took its place was the wearing of wax-print cloth with pithy or proverbial names, such as ‘Ingratitude of man’ (Onipa nni ayɛ), ‘Patience results in victory’ (Ntoboase ma nkunim), or ‘There is no medicine for hatred’ (Ɔtan nni aduru). Cloth designs have names, whether or not a written text is included in the design, and many such names are obliquely hostile. Thus a woman (or sometimes a man) can wear a cloth whose pattern indirectly states her feelings of anger, jealousy or sorrow, knowing that it will be understood, but will still be deniable speech. ‘Oh, did you think it was addressed to you? I just liked the pattern’, she could say. More aggressively, during the Akropong stool dispute in the 1980s, the supporters of Nana Addo Dankwa III, the reigning king (Omanhene), took cloth named for the opposition candidate (Akuffo ahenni: Akuffo's reign), dragged it on the ground from their taxis when they left the courthouse, and then used it as a doormat. By the first decade of the twenty-first century wax-print cloth was still being worn but its use as a stylized communication strategy was beginning to diminish. Proverbial expressions in the palace were also on the wane and more and more people wore inexpensive second-hand European clothes for everyday, and expensive eyelet and brocade for weddings and church – cloth that is incapable of expressing subtle indirect messages.

SECOND CASE: RITUAL GEOGRAPHY AND REGALIA

Akropong dynastic history is fiendishly complex and has no single agreed form. Confusion results when chieftaincy positions are left vacant for political reasons and other people, from different families, step in to perform the duties of the absent ones. Even the king-list is contested: the officially accepted version omits the first five incumbents (see below). But while the missing kings are not mentioned, they are not forgotten.

The Akwapim kingdom, sited on two high ridges in south-eastern Ghana some 35 miles from the capital Accra, is said by local officials and historians to have been established in 1730 when the indigenous Guan people invited Akan warriors from the neighbouring kingdom of Akyem to help fight off their oppressive Akwamu overlords. The Akyem warriors then took the opportunity to rule over the unwitting inviters. This is the Akropong version; the Guan have their own version of rapacious interlopers. Akropong people say they first settled in the valley below the escarpment at Amanprobi. Because the land was swampy, they moved up to the ridge, first to Nsɔrem (once a sacred grove on land owned by Abiriw, a Kyerepong or Okere Guan town) and then to the present location at Mpeniase (also once a sacred grove, on land that was the boundary between the indigenous Guan and the Okere Guan, between the towns of Larteh and Abiriw). Each of these three sites is marked in one of a number of processions during Odwira when the history of the town is recapitulated (and made viscerally real). Amanprobi and Nsɔrem are also sites of danger and power that commemorate fallen kings and warriors who died in the war of independence from the Akwamu (1730) and in the Akantamansu war (1826), in which Akwapim was liberated from the Asante and Akropong gained powerfully sacred booty. Like oaths, the place recalls the incident.Footnote ²⁸ And as the sites are an integral part of the local landscape, the past is experienced by the viewers as a zone to which they are seamlessly connected.

Amanprobi, the first settlement, is now the site of the royal mausoleum. It was formerly the slave village of the earlier kings of Akwapim.Footnote ²⁹ During the Odwira rite, palace elders (ahenemma: abrafo, banmufo, adumfo, nkonguasoafo …)Footnote ³⁰ visit Amanprobi, leave offerings, remove ancestral grave dust, and then return to town in silence, dressed as warriors signifying the seriousness and danger of the occasion. The chief of the royal mausoleum (Banmuhene) then uses this ancestral grave dust to anoint and thus protect and empower the king (who at the time is symbolically dead). The Banmuhene brings death to make new life. Amanprobi is thus intimately identified with the king (and through him with the people). Contradictory historic events are attributed to this place but each adds to the power of the site. It is said that Ofei Boa, an early ancestor of the Asonahene and one of the first Akwapim kings, murdered his servantFootnote ³¹ near Amanprobi at Odum Anim with dire political consequences – their line was banished and the stool was transferred to another branch. It is also said that Nana Saforotwe (r. 1802–16) was betrayed by Nana Addo Dankwa II (r. 1816–38) into the hands of the Asante and that he was ambushed and shot at Amanprobi.³²

Nsɔrem, a mile or two away from Amanprobi and a half mile from the present Akropong palace, was once a dense and sizeable sacred grove on the edge of the town. People say it was always misty, as if it were drizzling; they feared to cross it. In the 1940s a large public market was constructed there and a small area, around six by eight feet, was demarcated with a high protective wall for ritual events. Today the towns of Akropong and Abiriw have grown together to meet there. Nsɔrem (sɔre: to part, leave)Footnote ³³ is said to be a place where the ancestors lived. Once a year during Odwira, three women from each of the major Asona lineages of Akropong walk in procession to Nsɔrem carrying offerings of palm wine, water and mashed yam (ɔtɔ). There, on tiny shrines (pata: shed) constructed of odwen (Baphia nitinda) twigs and leaves, they place their offerings for the ancestors who fell in the wars against the Akwamu (1730) and the Asante (1826). At another time during Odwira, at night and in great secrecy, the remains of sacrificial offerings to Odosu are left at Nsɔrem (presumably in the past at another part of the grove), by the abrafo and adumfo, dressed appropriately in blood-stained war smocks.Footnote ³⁴ Odosu, a powerful suman (a ‘small’ deity),Footnote ³⁵ was captured from the Asante in the 1826 Akantamansu war. It protects people from the revenging spirit of those sacrificed at the behest of the king and state.Footnote ³⁶ The final site is the present location of the Akropong palace, under the large shady tree (a species of ficus) called Mpeni. Mpeni is an ɔbosom (deity). Every day for one week during Odwira offerings of mashed yam are sprinkled from the palace ancestral stool roomFootnote ³⁷ up to the roots of Mpeni, which are said to stretch to every house in Akropong. The shade provided by Mpeni Kofi (Friday-born Mpeni) is compared to the protection of the king; it is viewed as an assault on the state for a leaf to be cut from this tree and a gun shot into Mpeni would imply the destoolment of the king.

A number of other sites in the Akropong landscape are also memorialized annually during Odwira. When returning from Amanprobi during Odwira, the elders pause to pour libation at three sites that were former palaces where the Omanhene's royal ancestral stools were once kept: they thus memorialize the ancestors’ power. On the way to Amanprobi the elders stop at a small sacred grove called Odum Anim and leave offerings under a tall odum tree (chlorophora excelsa) called Odum Kyeame (in other words, Odum is viewed as a ‘linguist’ because it is believed he relays messages to ‘the big house’: ofie kɛse). Some believe that in the distant past a Mamfe chief was killed here. Over time these different historical events have become intertwined and indistinguishable. Some allege that one of the early kings might have been buried at Nsɔrem; others that one of the kings (Asa Kurofa, r. 1866–75) was found dead at Mpeniase. There are three other sites on the edge of the town that are marked by the elders as they make their way to and from Amanprobi during Odwira. These are really crossroads: Fofie Nkwanta, Asen Nkwanta, and Ehensua. It was difficult to learn the significance of these places.³⁸

Akropong people do not like to discuss these things and, in fact, most do not know much about them. Events such as these would be shameful to forget, but they are also events that are taboo to mention, and as such are only alluded to indirectly or memorialized in a condensed way, as in oaths. Indeed, one never refers to the ‘death’ of a king except euphemistically. Dead kings ‘go to their village’, ‘go for treatment’, and are compared to a ‘large tree falling’. It is thus to be expected that sites linked to the deaths of kings and important personages, to sites of utmost chaos and confusion, are sacred, powerful and dangerous, or become sites where taboo events of many sorts are condensed and thus merge. This then is the master historical narrative, whose remembrance (by means of performance) is crucial for Akropong people.

I turn now to the missing kings and their rival history and to the way in which objects indirectly alluding to this history are publicly flaunted during Odwira by the participants of the suppressed Asonahene's lineage. Diverse perspectives of a fragmented historical tradition are re-presented in the commemorative ceremony of Odwira. The king and Asonahene share a tradition and a collective self-awareness, but each also possesses contradictory memories and competing interests concerning his own identity: each is trying to show who he is. There is also an issue of power. The Asonahene asserts his ‘unsaid and unsayable’ identity, and his message is ignored by the king and the royal line, who are the audience for this display.

This is how it works. The finial on top of the Asonahene's gold-covered linguist staff depicts a man carrying another man over his shoulder (Figure 1). This illustrates a proverb: Wo huruw tra wo panyin a wo siaw naben mu ana wo siaw ne kɔn mu: If you try to jump over your elder, you'll get stuck on his horn (literally, around his neck). With that proverb-embedded staff and a similar phrase that is ‘spoken’ on the horn (abentia, see below), the Asonahene, chief of the ruling Asona clan, warns the Omanhene to watch his step – it is a cautionary reminder, a threat containing a powerful unspoken reference to history. The past that is alluded to is tangible, palpable and potent for both the performer and the one – indirectly and thus safely – referenced.

Verbal eloquence in Akropong is a highly valued art, a sign of high status and wisdom. Proverbs (mmɛ bus ɛm) and allusive metaphors (kasa koa: bent speech: kasa: to speak; koa: to bend, or akwa: a roundabout way) are aesthetic devices for benign or malicious purposes. They appropriate the past and are powerful strategies for indirectly voicing controversial or offensive topics when direct expression might be taken as criticism or as open aggression.Footnote ³⁹ Proverbs are distancing devices. They are a relatively risk-free strategy of communication; a strategy of indirection, ascribed to an impersonal third source, a neutral authoritative source.Footnote ⁴⁰ Proverbs allow Akwapim people to present a public face of ‘cool’ control when referring to delicate or forbidden matters. For all these reasons language in the palace tends to be suffused with proverbs and allusive phrases. Because Twi is a tonal language, drums and horns may be said to ‘speak’; much of Akwapim history is remembered by means of the appellations drummed on the talking drums (atumpan). Regalia – umbrella finials, tops of gold linguist staffs and swords – are also fashioned to make non-verbal statements about the history and political position of chiefs and the king. They do so by taking a resource, a ‘text’ that already exists, and recontextualizing it within another genre (see Barber 2007: 78–9). The underlying ambiguity in palace oratory and regalia allows for subtlety and flexibility of communication. The objects tied to proverbs can ‘say without saying’.Footnote ⁴¹ Added poignancy and power may be attributed to the motifs depicted on regalia precisely because their compositional repertoire of proverbs and aphorisms is shared with the children's name-responses of everyday life described above.

Much of the alternative history of Akropong focuses on the person of the Asonahene. His ancestor, Ofei Boa, was the first king of Akwapim: he followed Safori (see below), the Akyem warrior who led the defeat of the Akwamu and established the state in 1730.Footnote ⁴² The Asonahene, the elder of all the Asona royals (adehye mpanyin), is the uncle (wɔfa) to those Asona royals (adehye nkumaa) who are now the occupiers of the Omanhene stool. He is also one of the king makers. The Asonahene, however, is not allowed to see the royal Asona black stool ‘for fear he will cause trouble’, and the talking drums that call out the appellations (praise poems) of all the kings ignore the first five kings – including Ofei Boa.⁴³

The ‘official’ history is that Ofei Boa's black stool (Amanpongua: big state stool) was taken back after he allegedly murdered his servant, the Akyeamehene. Ofei Boa and his line went into exile. It is said they swore if they were not allowed to occupy the stool then the others should not make a replacement of the original. There was confusion. So Akropong people sent to Akyem to ask for another ruler and Fianko Betuafo (see below) came to occupy the stool, and at this point Akwapim history begins.

The talking drums ‘speak’ the king's appellations early in the morning before meetings are held in the palace. As they trace the king's ancestors’ feats, they link the past to the present. Fianko Betuafo's appellation addresses the deleted rulers:

Asuan is a plant that grows wild over abandoned houses and deserted villages. The drum asks, ‘Where have you been to allow this plant to cover the house?’ Akwapim is likened to an abandoned village. The implication is that what happened before Fianko Betuafo never happened, and that non-royals (non-adehye) have taken over. This means that after Safori, the first and legitimate ruler, those five people who ruled were not true members of the royal family. The Akwapim people wanted members of the ‘real’ royal family (ɔdehye kronkron)Footnote ⁴⁴ to come, so Fianko Betuafo was brought from Akyem. Fianko Betuafo was a latecomer, but the one they yearned for all that time. He was ‘Safori's sister's son’; he was from Akyem; he thus restored the true line. This ‘official’ perspective views the intervening chiefs as caretakers – when it acknowledges them at all.

Members of the Asonahene's family (abusua) claim that because of literacy (nhoma nim: book knowledge) and money, the Asonahene's position has been pushed aside and all the tradition (amanne) that shows the real position of the Asonahene has gone. They say that they are the real royals (adehye kronkron). There is a dynamic tension between their present position and what once existed. In an effort to keep their position strong and remain forever indelible, the Asonahene uses regalia such as gold-covered umbrella finials, swords and linguist staffs with condensed and powerful proverbial meanings to depict his position.

The gold-covered staff contrasts the Asonahene to what he mockingly calls the subordinate Asona (Asona nkumaa: the younger/junior Asona line, that of the present Omanhene). It cautions the ruling Asona not to try to surpass their elders/seniors. Proof of the Asonahene's superiority, they say, is that the amanpongua (big state stool) is still held by the Asonahene – and they think they should be ruling Akwapim with that stool. Every time they display this linguist staff, the friction between the Asonahene and the ruling Asonas is silently renewed. The issue is that while the Asonahene once occupied the Akwapim ruling stool and possesses the Akyem stool, he can no longer rule.

The contested ‘histories’ I now describe are at the heart of the Akwapim past and the link between past and present. The details, while complex, are essential; although they concern issues that are not openly discussed, most Akropong elders are aware of them to a greater or lesser extent. Peel's examination (1984) of the way the Yoruba manipulate rival oral historical and written records to alter the consequences of fixed tradition, and his careful and important analysis of the problems of historical reconstruction from oral tradition and the way societies reconstitute their past ‘as part of their self-production’, are relevant. The material I draw on for Akropong comes primarily from oral historical narratives supplemented by a few jealously protected hand-written books compiled by particular chiefs or by the children of kings; the latter comprise what Barber (2006) has aptly called ‘tin trunk texts’.Footnote ⁴⁵ Most of the books that I was privileged to see were begun in the early twentieth century and later continued by various descendants. The entries, much as in a diary or a private journal, are extremely varied. They include financial accounts, genealogies, lineage histories, excerpts copied from court cases now in the national archives, excerpts from early published histories, and so on. Cherished possessions, they are guarded carefully by their owners in the awareness that knowledge is power.

There are two different king lists in Akwapim: a long ‘hidden’ one and a short official one. In the short revised version, five kings (numbers 2–6) are deleted. Time is telescoped. Inconvenient history is denied. The problem concerns the immediate successors of Safori. After the war with the Akwamu, some of the warriors died, some returned to Akyem, and some remained. They had a war stool that they brought from Akyem. The leaders of the Akyem included Safori (or Ofori Kuma)Footnote ⁴⁶ who was a war captain. He stayed at Amanprobi and died (after the war). It is said that the Akyem warriors decided to return to Akyem, but the indigenous Akwapim people said no; and so the Akwapim kingdom was formed with its capital in Akropong as constituted by the 1731 Abotakyi Concord. In the long king list, Safori is followed by five rulers who are said to have occupied the Akwapim king's stool before the accession of Sakyiama Tenten in 1731. These rulers are (1) Safori; (2) Okyerema Manukure; (3) Ofei Boa; (4) Ofei Amanyopa; (5) Offei Ntoakyerewa; and (6) Amaniafem.

Note the collapsing of time.Footnote ⁴⁷ Fianko Betuafo, the seventh or eighth king, follows Amaniafem and is remembered because he showed the Akwapims the use of hand-woven cloth, which they substituted for bark cloth, and because during his rule black stools were given to the indigenous priest-rulers of Akwapim. Symbolically, if not in fact, Safori was the first founder. Fianko Betuafo introduced ‘civilization’ or ‘culture’ – that is, clothing and chiefship. The latter's name reflects this: obetuafo means ‘he came to advise’. On the talking drum, he is referred to as Safori's sister's son.

When I first asked about the ‘missing kings’ I was told these people were caretakers (nhwɛsofo): they did not look after the state properly so Akwapim cried out for the proper people to come back.⁴⁸

Akropong palace elders identified the five kings who preceded Fianko Betuafo as Asona from AkyemFootnote ⁴⁹ who came from the houses of the present-day Kodomasehene and Asonahene, who at the time were one and the same person.Footnote ⁵⁰ Kodomase elders confirmed that Okyerema Manukure, Ofei Amanyopa and Ofei Ntoakyerewa (2, 4, 5) were indeed from the Kodomase clan/lineage (abusua). Kodomase had two people after them called Asa and Asa who should have ruled after the Ofei's but both were murdered. Though there were other strong warriors in the family, the female head of the matrilineage (aberewatia) said she had no other man to succeed; instead she would herself accept to be Omanhene, but the people refused because she was a woman. She then decided to leave but was given a compensatory position as one of the king makers.Footnote ⁵¹ She accepted this and took her stool away from the palace.

Tracing Ofei Boa (3) to the Asonahene's family was difficult: one elder after another told me not to ask for these painful stories about terrible events. Finally I was told that Ofei Boa was brought from Akyem as a direct replacement for Safori. On the talking drums he is called Ofei Boa a odii ahenkan: Ofei Boa the first chief. His position is also displayed when the Asonahene swears allegiance to the Omanhene stool, for he then calls upon his ancestors – all these former kings (2, 3, 4, 5, 6) – to witness his action. Ofei Boa allegedly used a knife to murder someone from the Akyeamehene's family – and as a result was exiled. Asonahene Anim Bardieh II (r. 1990–2007) would only say that they ruled too harshly.

After Ofei Boa and his followers were driven from the state and the Akwapims were unwilling to depend on the Kodomase aberewatia for protection in wars, they sent for Fianko Betuafo from Akyem. Ofori Panyin of Akyem Abuakwa then deputed Amoa Batafo (Amoa the trader) to collect his war stool. I was told that two stools were brought to him: the Omanpon agua that was carried by Safori from Akyem to fight the Akwamu war was handed over to the rejected Asonas and they were told that in the annual festival the male stool should be given food last to mark its seniority over all the other stools. And indeed during Odwira the Asonahene feeds his stool last – on a separate day.⁵³

Members of the Asonahene family (abusua) told me that after the stool went to the other Asonas (the subordinate Asona: adehye nkumaa) the royals began to die and so they sought advice from the ɔbosom at Krakyi. After this, presumably, the stool was returned to the Asonahene. In any case the Akwapim king now rules with the Okuapehene stool and the stool that was brought from Akyem is with the Ofei Boa line, with the Asonahene. The Asonahene thus alleges that ‘it [the present Omanhene's stool] is a “duplicate”; the “real” stool, we have – we brought it from Adanse, it is with us’.Footnote ⁵⁴ The appellation on the Asonahene's horn alludes to this. It says Asafo Kwakye nea onim nim … ammen: ‘Asafo Kwakye, he who knows knows … it is not well cooked’ (that is, he knows that the stool in the palace is not the original; ammen is the negative of aben: to be cooked, powerful, strong, fortified). Another appellation is Adaworantu adawnantua (adaworan: something removed; toa: tie it); I was told this indicates that ‘when the right thing is away, a duplicate or fake is in its place’.

So the ‘real’ Asona people continued to rule on the Ofei Boa stool ‘which was the highest of all the Nsona [singular: Asona] stools of Akwapim’. And later the occupants of the Ofei Boa stool were brought back from exile and asked to enstool Asona royals on the Ofori Kuma stool (‘the second male stool’) and the Queen Mother stool.Footnote ⁵⁵ As for the stool that was made for the Omanhene in Akwapim, the descendants of Ofei Boa (the Asonahene people) are to this day barred from seeing it – they are barred from even entering the ancestral stool room (nkonguadan) in the palace. Akropong people say this is because of the murder they committed. Asonahene Anim Bardieh II denied this: he said they came for war and should have met gratitude, but met ingratitude instead. He says the Asonahene's stool is the Asona stool from Akyem and is thus superior to the Okuapehene stool in terms of when it came and how it came – but Akwapim people now recognize the other one. The Asonahene sees the Omanhene's stool as a ‘replica’ of that which should have ruled Akwapim.

During Odwira, the Asonahene people do not go to Nsɔrem on the same day as do the Asona nkumaa; they celebrate and feed their ancestors separately on the following day, Friday. This makes them last (and symbolically reasserts their seniority over the adehye nkumaa – the present ruling Asona royals and Okuapehene). A critical passage in one of the hand-written ‘histories’Footnote ⁵⁶ suggests an additional reason:

The implication is that the ‘subordinate’ Asona royals who now rule were those who fought the Asante and captured the powerful deities Oboaman and Odusu as war trophies. But as they came later, they were not there to fight in the Akwamu war and therefore do not feed the ancestors who originally came to make Akwapim. It was the 1930 Akwamu war that originally brought the Akyems to Akwapim and their ancestors fought and fell. The Asonahene people who came from Akyem are the ones who feed those ancestors at Nsɔrem on Odwira Friday (they were in exile during the 1826 Asante war, but they recognize the Akwamu war when their fallen ancestors are fed on Odwira Friday). They go on separate days – each for a different reason.Footnote ⁵⁷ The Asonahene thus belittles Odwira Thursday when the Omanhene and his contingent go to Nsɔrem, rather than face the fact that he is being ignored. The procession to Nsɔrem with offerings for the Asonahene's ancestors on Odwira Friday is led by an attendant carrying the gold linguist staff described above that ‘says’ this silently – in an image rather than in words. In contrast, the Omanhene is saying ‘We banished you because of what you did. We don't want to do things with you, and therefore you must do your rites separately, and not with us.’

ACKNOWLEDGEMENTS

This article was presented in 2006 at the Sites of Memory Colloquium organized by Suzanne Blier at Harvard University, Cambridge. Some of the material was discussed at the XIIth Satterthwaite Colloquium on African Ritual and Religion (1996). Gratitude is an inadequate expression for the gracious help and friendship given to me by Mr B. E. Ofori, Nana Anim Bardieh II, Asonahene; Aberewatia Yaa Asantewa of Kodomase, Opanyin Adu Darko, Joe Banson, Bossman Murrey and a host of other Akropong people too numerous to list here. I owe a great debt to Barbara Bianco, Gillian Gillison, J. D. Y. Peel and John Middleton, who generously read and criticized an earlier draft of this essay (and will be glad I finished it). Special thanks go to the anonymous readers.