Man Vol 16, Number 2, June 1981, pgs. 300-302


The authority of ancestors

Professor Kopytoff's comments (Man (N.S.) 16, 135-7) on Dr Calhoun's paper raise more issues than can be dealt with in brief response. Let me begin with the Tallensi concept of kpeem. Kopytoff lists the key definition I offered among the various 'glosses' he has extracted from the ethnography but the full definition deserves repeating. (And here let me remind Kopytoff of the linguistic commonplace that the information conveyed to the actors by a concept of such central importance, as kpeem is among all Mole-Dagbane peoples, far exceeds its semantic range. Indeed this is equally true of comparable terms in any modern European language - c.f. 'elder' in the O.E.D. or Webster.) The definition I refer to goes as follows (1945: 224). 'Kpeem literally means "senior": a kpeem is a senior, an elder' (one O.E.D. definition of 'elder' is 'One who is older, a senior, usu. in pl.)'. And in discussing this definition I remark (1945: 224) 'In all collective activities of men the places of honour are taken by the most senior men, the kpem' (plural of kpeem).
Kopytoff cites the criteria I list for attributing kpem[partialdiff]t which I translate as 'seniority', that is, the qualities that distinguish a kpeem from anon-kpeem, and infers that because social maturity and status, as of a chief, are among these criteria, the term is thus 'taken out of the lineage sphere and out of age and kinship spheres'. But the term is nowhere stated to be confined to the 'lineage sphere' in the sense of being limited to descent-regulated social relationships. The example I give (1945: 224) of seniority being recognised by a group of boys herding cattle and deferring to 'the oldest among them' should have made this clear. When a chief summons accessible kpem to come to a meeting he is seeking the assistance of men of experience and mature judgement regardless of their lineage status. When a polygynist refers to his most senior wife - most senior by time of marriage, regardless of age or lineage - as his p)ya-kpeem it is to her domestic status he is referring. And there are other usages of the same sort.
What, however, Kopytoff seems to have in mind is that the term 'elder' should always signify a status by age or generation within the lineage, or at least that this is the sense in which Calhoun has cited kpeem. But Calhoun's point,

as I understand it, is that Tallensi clearly distinguish secular eldership, the status of kpeem, from that of yaab, deceased ancestor. This is unquestionably correct. Kpeem is never used to refer to the ancestral dead or the dead in general in their spiritual or ritual mode of existence, whereas yaab is often so used. Mole-Dagbane languages do not make semantic use of tonal or pitch differences as is the case with such KWA languages as Twi, Fanti or Yoruba. Semantic distinctions are expressed phonemically. Kpeem, elder, is thus never confused with kpiim (cf. 1945: 209 fn.) the dead, or kpe'em, tough, hard, difficult in a metaphorical sense as in de kpe'em pam 'It is very hard or difficult', said of a court case.
Yaab, grandparent, like ba, father, and ma, mother, are used in reference and address for living kin of these types, in a secular sense. And they are also used for deceased parental and ancestral antecedents in the appropriate ritual contexts. Nor is there any difficulty as to knowing exactly whether the reference is to the living or the dead. Thus when Teezien explained, of a particular shrine in his possession, mba Kunkonki ma la - ['This is] my father K's mother', there was no mistaking the reference to deceased ancestors. But when Saa, explaining why his younger brother had inherited their father's divining shrine remarked tam ten)ni bi-kpeem pu deeg[partialdiff]t u ba bakologo - 'In this our country a senior son [strictly, in this context, first-born] does not take over his father's divining shrine' - the reference was patently to a shrine that had belonged to his living father. Again, when Omarra described how his yaab Boligraana had 'reached' him to become one of his Yin ancestors, there was no mistaking the reference to a deceased ancestor. But when we went on to tell how this same Bloigraana had, five generations earlier, established the settlement at Biung, it was as clear that he was referring to a then living man. The action context is decisive, not the dictionary.
Thus, on the question of translating the term kpeem as 'elder' in a strictly secular sense, I think that Calhoun is right; I have checked with Dr Moses Anafu whose knowledge of his native Tallensi language is both profound and scholarly and am assured by him that this is correct. I have also checked with him the pronunciation, orthography and translations of the other Talni words and terms I am here

discussing and have his assurance that they are correct.
It is not correct to say, as Kopytoff does, that when Tallensi use terms for lineage elders this means that they 'disregard the dividing line between those who are alive and those who are dead'. What it does mean is that they recognise a form of continued existence, or rather efficacy, of those dead who have descendants in the appropriate later generational statuses. And that is documented in various institutional as well as material ways, no less than in world view, moral principles and ritual practices. But then when Christians utter the Lord's prayer they are not disregarding the line between living fatherhood and the fatherhood of God; and there is plenty of evidence that their God and other divinities are deemed to be in some way present here on earth, among the believers. Tallensi diviners, by the way, when they pass on the messages and demands of the ancestors always use only the appropriate kinship terms -ba, ma, ba-kpeem (for an actual grandfather), yaab etc. This reflects the basic rule that only genealogically connected ancestors are efficacious in the life of any person.
The verbal usages I have listed for the Tallensi are exactly paralleled in the other Mole-Dagbane languages, as Kopytoff indicates. But he has been misled by the orthography in one or two instances. Alexandre, to whose excellent dictionary he refers, uses superscripts to mark vowel length and accentual distinctions, in effect, therefore, characteristic phonemic distinctions. More kyéma, (phonetically, kyéema 'être plus âgé, plus grand, plus important ...' is directly cognate with Talni kpeem (as in a kyema man, he is older than, Tal. u kpeem man). More kyéma (pl. kyémdamba) 'aîné, plus âgé ... is the corresponding substantive as in bi-kyem, oldest child (Tal. bi-kpeem) pugh-kyema (Tal. p)ya-kpeem) 'la première femme'. Kyema, 'dur, difficile' corresponds to Tal. kpe'em and is a near homonymous but different lexeme from kyema, older etc. Kyema in Gorni is translated as 'head' e.g. of a lineage or section by Rattray (e.g. 1932: vol. i. 245), also as 'elder' (1932: 255 and passim ). Elsewhere he writes of a bride as 'the wife's eldest daughter (p)ka-bi-kyema)' ... (cf. Talni p)ya-bi-kpeem). And, incidentally Rattray provides many informants' statements passim in which yaba is translated as 'ancestor' (e.g. 1932: vol. ii, 296) and prayers and sacrifices are addressed to 'fathers', 'mothers', and other lineal kin. Thus (vol. ii, 1932: 326) 'Anereba then said "My mother ..."' and later 'Anereba ... spoke as follows "My father, Aganda ..."' Kyima (Tal. kpiim) he translates as 'spirits of the dead' (vol. I, 1932: 44).
The Mole-Dagbane dialects closest to Talni, however, are not More and Gorni but Mampruli, the dialect of the Mamprussi,
and Dagbane, the dialect of the Dagomba. According to the short but accurate dictionary compiled by Arana and Swadesh (1967) there is a Mampruli word kpeeman pl. kpeem-di-ma which they translate, in English/Spanish as 'senior': 'mayor en edad ...' Oraku (1917) in the excellent word list appended to his Dagomba Grammar cites kpema pl. kpamba which he translates as 'elder'. Arana and Swadesh cite kpeengu, strength, hardness: fuerza, dureza', which, in Okraku appears as 'kpeon, strength, power'. The Talni equivalent is kpe'on. And finally, let me quote the linguistically more up-to-date Dagbane Dictionary compiled by a committee in Tamale in 1941 under the chairmanship of then District Commissioner Mr H.A. Blair who had an expert knowledge of the language. This cites 'kpe'ma, noun pl. kpamba, elder (referring to age or rank)' and 'kpe')n, strength' as well as 'kpe'ma, adjective or verb, to be hard; to be strong.' Okraku has 'kpima, dead body' and Blair et al. have 'kpima, the dead, dead body, ghost' (cf. Talni kpiim).
I conclude that the comparative evidence fully supports Calhoun's translating kpeem as 'elder' in the secular sense and yaabas ancestor in the ritual sense. Ba, father, like ma, mother can be and often is extended classifactorily in common usage to refer to a father's brother (within the limits described in 1949: 140 sqq.) and father's father's brother's son but not to father's father's brother (as Kopytoff claims for More) who is classified with ba-kpee. Ba and ma are also extended, like the English word father, to refer to forefathers (ba-nam) and foremothers, if I might coin a term for antecedent matrikin. This is easy, among the Tallensi, since the lineage is thought of as a corporate group, 'one person' jurally and metaphorically, filially connected with an antecedent 'father' or 'mother'.
As far as I know it is universally the case that terms for 'father' and 'mother' are different - they must be, to distinguish patrifiliation precisely from matrifiliation. But parents' parents, like the reciprocal child's child phonemically distinguished by a linguistic device common throughout the Mole-Dagbane area (i.e. yaa-b, reciprocal yaa-n) are not, need not jurally be, so distinguished. However, when a distinction is made it is done by means of a gender-indicating suffix, yaab-doog, male, yaab-p)k, female grandparent. The same terms and conventions are found in More - yaba-daogha, yabpoka in Alexandre's orthography - as well as in Mampruli and Dagbane, and in every case these terms are extended to ancestors as is also the case with French aïeul, aîeule, the translations given by Alexandre.
I do not understand Kopytoff's contraposition of the 'horizontal' and the 'vertical'

boundaries between living and dead lineage members. What these represent is not opposed dimensions of Tallensi social structure but complementary domains of social action, neither of which can be regarded as secondary to the other. The lineage boundary is decisive in, for instance, the politico-jural sphere of competition for high office, or in the domestic sphere of marriage negotiations. But one cannot sacrifice to a living lineage elder. In all ritual activities, the horizontal dimension is the critical one. And the transactions in the vertical dimension can never, according to traditional custom, be finally settled without the appropriate ritual sanctioning. Semantics alone cannot account for this. Where, to my mind, Kopytoff's semantic analysis of 1971 did make a most important contribution to our - or at least my - understanding of African ancestor cults was in a different direction. It was in emphasising, as I have already noted, the continuous presence of the ancestors among their living descendants in states of existence that represent a transformation, through the agency of ritual, of aspects of normal, secular parenthood. The dead are not fathers and mothers and grandparents in any sense that is coterminous with the exercise of living parental and grandparental roles and capacities. But they are deemed to perpetuate in a transcendental mode as Calhoun argues, the components of parental and grandparental roles and capacities that represent the authority vested in their status.
Meyer Fortes
Alexandre, R.P. 1953. La langue möré. Dakar: IFAN.
Arana, Evangelina y Mauricio Swadesh 1967. Diccionario analitico del Mampruli: con una introducciòn etnogràphico por Susan Drucker Brown. Mexico: Museo de las Culturas, Instituto Nacional de Antropologia e Historia.
Blair, H.A. et al. 1941. Dagomba (Dagbane) dictionary and grammar. Accra, Gold Coast: Government Printer.
Calhoun, C.J. 1980. The authority of ancestors: a sociological reconsideration of Forte's Tallensi in response to Forte's critics. Man (N.S.) 15, 304-19.
Fortes, Meyer 1945. The dynamics of clanship among the Tallensi. London: Oxford Univ. Press.
Kopytoff, Igor 1971. Ancestors as elders in Africa. Africa41, 129-42.
Okraku, John A.S. 1917. Dagomba grammar: with exercises and vocabularies. Cambridge: Univ. Press.
Rattray, R.S. 1932. The tribes of the Ashanti hinterlan. v.2. Oxford: Clarendon Press.

Go to the main Ancestors in Africa page Go to the main Experience Rich Anthropology page Go to the CSAC Anthropology Pages - online