Navigation – Plan du site

Calendar and Ritual

The Mamprusi case
Rite et calendrier. Le cas mamprusi
Susan Drucker-Brown
p. 57-86


Chaque lunaison chez les Mamprusi (nord du Ghana) est associée à une fête qui a lieu 19 jours après l’apparition de la nouvelle lune. Puisque 12 lunaisons ne correspondent pas à une année solaire, ces fêtes se déclinent par rapport aux activités de l’agriculture, de la chasse et de la pêche. Ce calendrier reflète les valeurs hiérarchiques de la société. Les 12 lunaisons de l’année solaire sont rangées en couples et hiérarchisées, l’ensemble étant subordonné à une lunaison hors couple du nom de « feu » qui peut détruire toutes les autres. Pendant les principales fêtes, la royauté depuis sa fondation est mise en scène de diverses façons, et les déplacements du roi dans l’espace qui environne son palais ne peuvent être compris qu’en fonction d’un cosmos ordonné où la personne sacrée du roi représente le soleil. Le temps se divise en périodes dont certaines sont bioo (dangereuses). Pendant ces périodes néfastes, certains interdits doivent être respectés. Ce concept de bioo lié à un système d’interdits permet d’unifier les diverses unités de temps et de mettre en valeur le roi.

Haut de page

Texte intégral


1It is common in the writings of British social anthropologists for each generation to raise anew questions studied by previous generations, as though the polemical aspect of these questions were most important and the cumulative content of our knowledge about human society were relatively insignificant. In an area of study where the idiosyncracies of individual perception are so important, it may be difficult indeed to disentangle fact from opinion or philosophy from detached observation. Nevertheless, a process of prediction at work in this subject can be illustrated by the discussion of calendrical systems. I should like to add the Mamprusi case to the discussion and to examine the notion of a « ritual calendar » in the light of that case.

2In 1927 Malinowski described the Trobriand system for counting lunar cycles and the procedure for choosing the days on which the major seasonal festival was to be performed. He noted that for the Trobrianders, most events which needed to be predicted occurred during « seasonal years of indeterminate length rather than in solar years of fixed duration ». He concluded that the Trobriand counting of moons was not a means of reckoning astronomical time. He argued that in fact people would not preserve astronomical knowledge without a practical context for that knowledge.

3Malinowski found that most Trobrianders were unsure of the names or total number of moons which comprised a complete cycle of lunations. Especially knowledgeable men could name thirteen moons, and twelve names could regularly be obtained from some informants. « As a rule the moons would be classified in two groups of five, with a moon or two regarded as intermediate ». There was no consensus on when the year began or ended, though gardening procedures and the periods appropriate to fishing and sailing expeditions were carefully timed according to seasons in the solar year. Moons were named essentially for the ecological characteristics of the periods during which they appeared in the solar year. Specialist observers consulted together in order to choose the days for celebration of the major annual festival, and this choice appears to have served to intercalate lunations with the solar year. Astronomical observations were made at three different geographical points in the Trobriand archipelago, and the major annual festival was held during four successive lunations in different districts of the archipelago, occurring last at the point where a particular sea-worm which gave its name to the festival annually appeared in the sea and was harvested.

4Evans-Pritchard (1937), looking at the manner in which « time » is conceptualized among the Nuer, argued for the contrast between what he called « ecological time » and « structural time ». Ecological time reflects observation of the natural environment. It is used to predict seasonal change and to organize practical activities. It is progressive only in that it consists of completed annual cycles, but it emphasizes cyclically recurrent phenomena. By contrast structural time reflects social organization. It is used in particular to organize the genealogical constructs by which Nuer link living clans and lineages to the deceased forbears from whom they count descent. Structural time is used to calculate periods longer than annual cycles and to conceptualize changes in social status. By contrast with ecological time, structural time is entirely progressive, encompassing a period from the origins of Nuer society to the present.

5Evans-Pritchard thus adds to Malinowski’s observation, that astronomical knowledge will only be preserved given a practical context, the notion of two different purposes for which duration may be conceptualized: first in a context of « practical ends » and short term goals; second as a means of ordering social relationships over long periods of time.

6Leach (1950 and 1954) returned to Malinowski’s concern with astronomical knowledge. In a comparison (1950) of three calendrical systems in Oceania he noted, in agreement with Malinowski, that temporal prediction occurs in these systems without people conceiving of a « solar year ». But he pointed out that the prediction of seasonal events in terms of lunar cycles is possible only if an event outside human control, which indicates the relationship between lunations and the solar year, is carefully observed. In each Oceanian case he was able to isolate such an event. He concluded that if the reckoning of lunations were « to have any practical value » some mechanism for intercalation (the adjustement of lunar counts to the solar year) would be essential. Hence, he concluded that « field records which failed to demonstrate the existence of such a mechanism were probably incomplete". Moreover, he observed of the calendrical systems he had analyzed (1954: 120), « that the system should work at all is almost a paradox; it does so only because the logic with which the natives approach calendrical time is vague in the extreme and even inconsistent ».

7In 1978 Turton, who worked among the Mursi of southwest Ethiopia, further clarified Evans Pritchard’s judgement in the paper of 1939 that « perceptions of time... are functions of time reckoning, and hence culturally determined »; he also confirmed Leach’s observation that the calendrical system could not operate without inconsistencies. Turton and Ruggles (an archeo-astronomer) emphasize that the perception of time should be distinguished from its measurement. They demonstrate that among the Mursi the manner in which lunar cycles and seasonal cycles are correlated to form a single temporal unit does indeed depend, as Leach has observed almost paranthetically, on « an agreement to disagree » (The phrase « Agreeing to disagree » is the title of their paper).

8Among the Mursi the measurement of duration (1978: 593) « is an eminently social activity, the standard of measurement used by one individual being the product of his day-to-day interaction with many others in a single local community ».

9The Mursi, they demonstrate, keep the count of lunations in phase with the solar year by a process of retrospective correction. Not one but a number of different phenomena « outside human control », of the kind Leach (1954) calls « seasonal markers », are used to make the practical decisions which require short term prediction. However the lunar count is in effect always provisional, and at any particular point in time the population is roughly equally divided in opinion as to when the moon was first sighted and as to the number of the current lunations in the total set. From the discussion of Turton & Ruggles, 1978 (see Current Anthropology 1979: 174) it is clear that « retrospective correction » is a widely used method of intercalation. See also Fortes (1975) for a description of the process among the Tallensi who are close neighbours of the Mamprusi.

10Turton and Ruggles conclude that « There is consequently no system of Mursi time reckoning in the sense of a self-contained set of propositions which once stated can be made to yield unambiguous determinations. »

The Mamprusi case1

The seasonal year

11The Mamprusi people occupy a region of northern Ghana around the 10°N parallel which extends west from the present border with Togo in the east to about 1°W. It is roughly 150 miles from east to west and 40 miles from north to south. The Mamprusi live by farming and their region is the northernmost extension of yam cultivation in Ghana.

12Yearly cycles of cutting, burning and clearing the savannah brush, planting, weeding, harvesting and preparing crops for storage, create basic rythms of social life, dependent on the seasonal changes in patterns of temperature, rainfall, wind and the growth-cycle of plants. The major climatic contrast in the seasonal year is between seoo, the rainy period which lasts from May to October, and wuuni, the dry season. At the height of the dry season no rain falls, and there are predictable subdivisions of this period. The harmattan (kikaa), a cold, dry, northerly wind, blowing from the Sahara, arrives between December and February. Daawqligu « days of sweat » (daari: day, wqligu: sweat), a period of extreme and at times humid heat, follows the harmattan. This heat gives way to a time of light rains in April and May (siga, pl. sigri) which heralds the true rains. But rainfall is never as predictable as the subdivisions of the dry season. The rainy season is a period of anxiety over food shortages which precede the harvests, and because the rains are both unpredictable and absolutely essential to the crops. The rains also encompass the period of most intensive agricultural labour, and the dry season is divided into the periods of farming (koobu saha) and harvest (takoori or, from Hausa: agazeeri).

13The succession of climatic changes and named seasons is known to all adults and most children. Adults link these with the life cycles of particular plants and with astronomical phenomena. The various phases of agricultural labour are timed by these seasonal markers. Certain sidereal movements of the sun, the Pleiedes, Orion and the planet Venus are well-known.

14Taken together wet and dry periods comprise a cycle Mamprusi call yumna (pl. yumni) which may be translated as « year ». The interval represented however may vary from 9 to 16 months since the word is most often used to reckon the time which will elapse between particular phases of the seasonal cycle.

  • Rituals prohibited (kyisigu – customary prohibitions and avoidances)

15During the period of heaviest farming labour (koobu-saha, roughly from mid-June to mid-August) the performance of certains rituals is prohibited. These include final funerals (the most elaborate of Mamprusi rituals), marriages, and the installation of new chiefs. Literally and in metaphor these are rites of passage, and all journeys should also be avoided. Mamprusi recognise that the season is « just for farming ». However the prohibitions are not simply utilitarian rules. Mamprusi do not speak of the avoidances which characterize the period as directly associated with farming. They explain the prohibitions by saying that the prohibited ritual (or the journey) will proceed badly if undertaken during the rains. In other words the prohibitions are regarded as linked to the successful completion of a passage.

  • 2 Mackaskie (1980) demonstrates, for the Asante, the political importance which manipulation of such (...)

16The reappearance of the same or similar prohibitions in the context of other units of time (see below) suggests that the prohibitions reflect a general view of time (or duration) as marked by auspicious and inauspicious intervals2. Activities fraught with risk, such as those prohibited in the rains, are more dangerous during periods Mamprusi refer to as bioo. This term is normally translated as « bad » by English-speaking Mamprusi; but it implies a notion of active, living force to which human beings are vulnerable and which may cause harm or even death. However the same power is beneficial under certain controlled circumstances. Rather than implying « evil », bioo implies a danger from which « prohibitions » (kyisigu) offer protection. We return below to this point but consider first the positive injunction to perform certain rituals which is complementary to the observance of prohibitions.

  • Rituals demanded

  • 3 In the Mamprusi region the site of a chiefly grave, or the grave of a Nayiiri, may be a tengbang. S (...)

17While the rains are the period of prohibitions (kyisigu), the dry season demands the performance of certain rites. The Nayiiri (king) and Mamprusi royal chiefs are responsible for performing and providing sacrifices which summon the rains. Those custodians of the earth (tendaana, pl. tendaan-dima) who hold chiefships slaughter animals provided by chiefs to particular « shrines » (tengbang/tengbanna, lit. teng-earth, gbangu-skin)3. It is the duty of the Nayiiri and royal chiefs to provide animals for sacrifice. In addition royal chiefs and the Nayiiri will sacrifice to their own ancestors for rain and the fertility of the country. Village chiefs provide sacrifices for shrines which are considered efficacious only within the area of their own villages, though some of the shrines are distant from these territories. The Nayiiri however provides sacrifices and himself sacrifices to provide rain throughout the Mamprusi region. Similarly, seasonal disasters (floods, winds, insect plagues) which affect the whole polity may be prevented by sacrifices provided or made by the Nayiiri himself.

18One may regard as reciprocal to the services provided by chiefs the small gifts of both early millet (nara) and yam (nuya) which are provided at harvest season by each village household. Neither yam nor early millet are major food crops but they relieve the stress of the hunger season which precedes the main harvests.

19The gift to a chief is made once the crop has been harvested and before it is consumed in the residential space of the village. Cooking and consumption in village space should occur only after the first fruits have been presented to the chief. The crop may be consumed earlier in the farming area.

20These ritual gifts which link chiefship to the agricultural cycle and indicate the identification of chiefs and the Nayiiri with cosmological forces are notable in that they are not accompanied by any collective celebrations. They occur privately, as in the case of the gift of first fruits made to a chief by each household head individually, or in secret, as with sacrifices to powerful shrines and ancestors. They do not take place in open court, nor are they attended by music, drumming or any other announcement. No festival accompanies them, although they may take place during festival periods. We return to this point below.

The days (solar observation)

21The time which elapses between dawn and dusk, combined with the hours of daylight, comprise the unit Mamprusi call daari. This unit which we call « day », is subdivided between the time of darkness (yungu) and the time of light (also called daari).The time between first cock’s crow and the sunrise (lit. wuntang buliya: sun springs up, sprouts) is called wuntang neeai or beeo. The time immediately after sunset (lit. « sun-fall » wuntang luu) is called zaanori, « dusk ». The term zaanori is applied also to the time from mid-afternoon to dark, and asuba (from Hausa) is the period before noon or midday called wuntang-saa zugu (lit. sun-on-top). Just as midday bisects the daylight, darkness is supposed to be divided by yung-tasuusi, variously translated as « deep night » or « midnight ».

22The periods of dusk (zaanori) and dawn (beeo), when sunlight is present but the sun itself cannot be seen above the horizon, are regarded as bioo — danger periods. During these times, people should not move out of their houses or villages. One might describe these periods as liminal intervals in the sun’s passage through the heavens. The spirits of ancestors and other powers are said to move about at these times. In addition to the prohibition against travel people are also warned not to wash clothes or bathe. Both activities remove body dirt (da’ari) which may be regarded as either attractive to, or containing in itself some aspect or element of the spirits.

23By contrast with dusk and dawn, midday is a period during which people also prefer not to move about or to engage in heavy physical labour but here the avoidance is simply practical and no mystical danger is associated with the time of day. Similarly, midnight is not necessarily a dangerous time to move about though only persons defended against witches or witches themselves would choose to do so.

Figure 1. Subdivisions of the day (daari) showing the danger periods at dusk and dawn

Figure 1. Subdivisions of the day (daari) showing the danger periods at dusk and dawn

24But just as the wet season is a period of growth and active power, so dusk and dawn are periods when it is particularly auspicious to sacrifice to the ancestors or at other shrines. The same forces which are dangerous are accessible at these times.

  • 4 I am grateful to Dr. A. J. Naden for observing the accuracy of these gestures which he says corresp (...)

25In addition to denoting intervals of the day as inauspicious for certain activities and appropriate for others, Mamprusi also observe the sun’s daily movements and can demarcate them with considerable precision. Mamprusi gesture to signal a particular time by imitating with outstretched arm and extended palm the predicted angle of the sun in relation to the horizon4.

26It is worth noting at this point that traditional Mamprusi architecture provides conditions in each house for an excellent astronomical observatory. The open inner circle of the compound yard, surrounded by the wall into which individual rooms are built, provides a dishlike space above which the night sky appears as in a domed planetarium. The gate providing entrance to the compound yard is located north of the large circular hall built into the compound wall. Both the gate (zaanori) and the hall through which the compound may also be entered face west. The Mamprusi word zaanori means both « evening » and « west ». It is phonetically very close to za’anori meaning « gate ». The customary westerly orientation of all Mamprusi houses makes the door of the zongo a useful frame for solar observation. The senior men of a house gather in the zongo where the household head (yiridaana) receives guests and spends his leisure. Another favourite seating place for senior men is outside the zongo facing the cleared space (samani) in front of the zongo and the gate. The ancestor shrines of commoners are normally placed just south of the zongo door. Senior men sit inside the zongo facing outwards and men sit in groups with their backs to the zongo wall also facing west (see figure 2 and photo 1 & 2).

Figure 2. Basic plan of a Mamprusi compound

Figure 2. Basic plan of a Mamprusi compound

Photo 1. External view of the west facing gate to a Mamprusi compound

Photo 1. External view of the west facing gate to a Mamprusi compound

It is built into the circular entrance hall (the zongo). Note the sharp shadow thrown by the evening sun. Logs placed in front of the zongo to the right of the gate are sitting places for men.

S. Drucker-Brown

Photo 2. The same gate viewed from inside the compound

Photo 2. The same gate viewed from inside the compound

A king’s elder is leading a procession to greet a funeral.

S. Drucker-Brown

27Another source of information about Mamprusi concept of daari is the set of words used to indicate the succession of future and past days. Although the day may be regarded as beginning with the first rays of the sun’s light (beeo or wuntang neeai) the vocabulary reflects the emphasis given to the cyclical succession of days. Thus past and future can be fixed by distinguishing « today » (dabsiri) from daylight, or day in a generic sense (comprising both night and day) which are both daari. Tomorrow (beoo) is indicated by the word « dawn ». Yesterday is specified as sosla. However the day after tomorrow and the day before yesterday (daari) can only be distinguished by context. Days may be counted by utilizing the roots of numbers as « infixes » between da- and -ri. An enumeration of past days can be specified by use of a final suffix -la.

Day names (cycles of seven days)

28A set of seven day names based on Hausa (and derived from Arabic) are in use. They also may serve as personal names for children. Although Naden (Mamprusi Dictionary M.S.) gives the translation bakwe for a seven day « week », this was not a word commonly used. The day names however were in common use. The seven day cycle contains two days which are bioo: Wednesday and Saturday (alariba and asibiri, respectively). Events occurring on these days can be expected to recur. They are therefore inauspicious for beginning projects which involve risks or performing the rituals normally avoided during these intervals. A funeral should not be greeted on Wednesday or Saturday and it is particularly dangerous to leave a funeral on a Saturday. These avoidances are always considered by courts and kin groups in setting the dates of life-cycle or court rituals. Announcement of a chief’s death should not be made on either of these days. However, these same days are chosen for performance of installation rituals. The logic of this reversal of an avoidance is impeccable. Installation emphasizes the recurrence of office (naam) in the same patriline. In Mamprusi ideology the installation marks the « return » of chiefly office to a prior incumbent in the person of his living descendant.

The lunations

29There are three Mamprusi words for « moon ». Ngmarga and namboaa seem to be used to place the moon among other heavenly bodies as in:

ngmarga moon
ngmar-wuusi stars
be-ngmarga dawn star (Venus)
namboaa-biisi stars (lit. moons’s children)

30The third word g)ri is used in the lunation-count and, combined with adjectives or verb roots, to describe phases of the moon:

g)r-palli new moon
g)r-kurgu old or full moon
g)r-kyebri waxing moon

  • 5 Nadel (1954) describes the Nupe agricultural calendar used by « peasants » as follows: « (It is) ba (...)

31Mamprusi have between 12 and 14 names for the lunations5. However it is agreed that some names appear to be alternatives for others.

32Thus 12 lunations constitute a single set.

33The lunation count of the Mamprusi is paradoxical. The Mamprusi subsist on the basis of a primitive agricultural technology requiring very careful observation of climatic conditions, yet they measure duration in order to organize festivals with a lunar count which appears to be totally detached from the seasonal year. Because the twelve lunations constitute an interval which is shorter than the solar year, the festivals move over time through the seasons of the solar year. Roughly over 20 years each festival recurs in the same season of the solar year. Taking together my field records 1963-65 of the festival celebrations and those provided by Dr. A. J. Naden (1976-84) it appears certain that the festivals are dated by a lunation count made with no intercalation of the lunar count and a seasonal year. Indeed early accounts of the Mamprusi calendar emphasize that no intercalation occurs (see Chart 1, also Blair 1932, J. A. Braimah pre-1963).

Chart 1. The Dates of Mamprusi Festival Celebrations










« end » of July


5 Oct.

20 July

26 Nov.

1 Jan.

28 Jan (?)

24 April


12 Jan.

13 Mar.

(25 Sept.)


22 Dec.

(20 Feb.)

(15 Sept.)

(22 Nov.)


20 Dec.

(10 Feb.)

5 Sept.

12 Nov.


30 Nov.

30 Jan.

13 Aug.

1 Nov.


11 Nov.

5 Jan.

23 July

(28 Sept.)


27 Oct.

28 Dec.

13 July

18 Sept.


17 Oct.

(17 Dec.)

(The dates of festivals 1963-65 are from my fieldnotes. Dr. Naden has kindly provided me with his calendar of festival dates from 1976 to Dec. 1984. Dates in parentheses are extrapolated on the basis of informant’s reports. Other dates are of festivals whose performance he witnessed.)

34This is all the more extraordinary since the Tallensi, who are close neighbours of the Mamprusi, celebrate their major festival (da’a), which closely resembles the major annual Mamprusi festival (damna) at the time of the first harvest. Adjustement of this festival moon and the preceding lunation provide retrospective correction which keeps the Tallensi lunar count in phase with the solar year (Fortes, 1975).

  • 6 Trimingham (1949, 1959) gives descriptions of Islamic festivals which parallel those of the Mamprus (...)

35The Tallensi, like the Trobrianders, name the moons with reference to ecological characteristics, barring only the moon of the main festival season. The Mursi number the moons which make up a seasonal year. By contrast to both Tallensi and Mursi, the taxonomy of lunations in Mamprusi is based entirely on references to a succession of festivals (kyuu-daari) clearly or obscurely associated with Islam6.

36This is not the place to discuss the role of Islam in the Mamprusi polity. It would seem that Muslims have always formed a very small proportion of the total Mamprusi population. They are not regarded as having participated in the founding of the polity. However there have been, in Mamprusi history, a few important literate Mamprusi Muslims. (In 1963-1965 no Muslim, literate in Arabic, resided in the capital). Literacy in Arabic, combined with long-distance trade and other islamic practices do however link the Mamprusi Muslims to a wider social universe than is available to most Mamprusi (Drucker-Brown: 1975).

37The senior Muslim court official, Nayiiri’s Limam, who resides in a neighbouring village, is responsible for fixing the dates of the mamprusi lunar calendar after consultation with the king’s court. This has meant, in effect, that the beginning of the new lunations is locally determined. The new moon of each festival month is announced at the king’s court, usually by the firing of muskets and rifles; and the festival months of damma and kyivmsi are celebrated at the capital courts. In certain festival months drumming at dusk and dawn in the king’s village occurs throughout the lunations.

38As among the Mursi, the sighting of a new moon is always an occasion of disagreement. However disagreement is limited to within 2 lunations (at most) and the festivals move slowly over time through the seasonal year (see Chart 1). A particular festival moves from the height of the rains to the height of the dry season roughly every ten years (or would do so were the seasonal changes more predictable). The movement is sufficiently slow that the lunation-count serves for prediction within a twelve to sixteen month period.

39It should be emphasized however that the entire set of lunation names is used in conjunction with a calculation of when, in the seasonal year, a particular festival will recur. Though the total set of lunations is regarded as occurring in a twelve moon period, it is the reappearance of particular festival moons which Mamprusi recognize as the beginning of a new year. In fact, the appearance of each new festival moon is the occasion of the greeting ni ti yum-palli, literally « our new year ». The Mamprusi thus observe a number of concurrent, annual cycles (similar to our own school year, fiscal year, sports year..) which occur within the seasonal year.

  • 7 I am grateful to M. S. Abudulai for this observation. During 1981 he remembers news reports of figh (...)

40In theory the Limam and members of the Muslim community keep the lunar calendar in phase with the Islamic year. But given the vagaries of Mamprusi lunar observation, this correlation is by no means precise. In recent years the lack of concordance between dates in the international Islamic year and the celebration of Islamic festivals by Muslim communities in Northern Ghana has led to conflict between local Muslims and representatives of the National Muslim Council whose headquarters are in Accra. The latter has sought to impose on indigenous Muslim communities the calendar of international Islam7.

41The Gregorian calendar has also had considerable impact on the measurement and perception of time. School holidays, the payment of government employees and the celebration of Christian festivals are all dated by the Gregorian calendar. The Islamic calendar, however, has been a source of conflict between those who see Ghana as part of the political universe of nation-states and those indigenous Muslims who belong to the Mamprusi or other indigenous polities. By contrast Christianity and the Gregorian calendar are more simply seen as part of « Ghana times » (Ghana-saha), the new and current political epoch.

Festivals and the calendar

42Festivals give their names to particular lunations, and each festival moon is paired with a « junior » lunation which precedes it. The pairing of lunations is an important feature of the calendar both as a mnemonic device and in terms of the division of the seasonal year into periods of prohibition and danger as well as periods of enjoined ritual performance and essential agricultural labour.

43Mamprusi say there are twelve named moons (g)ri / pl. g)ya) which correspond to the seasonal year. All but one lunation are said to occur in pairs. The exception, bugum (see below), is regarded as bioo and is said to be uniquely powerful. In fact, though Mamprusi do not mention it when speaking of the lunations, there is actually another unpaired lunation; but this period is regarded almost as an appendage of the pair which precedes it, and the festival for which it is named is relevant only to the Muslim community which observes a fast during the preceding lunation (see chart 2).

44The pairing of lunations however is a reflection of a powerful emphasis in Mamprusi social life. In both kinship terminology and the ideology of political hierarchy, individuals are grouped in pairs which share a single essence and a common subordination to a more powerful force. Juniors precede their seniors. The logic of this view is most clearly expressed in the Mamprusi belief that the second-born of twins is senior in birth-order to the first-born.

Chart 2. Lunation names and festivals in the Mamprusi calendar showing paired lunations

common names

alternative names occasionally substituted where shown below:


no alternative


no alternative


no alternative






no alternative


no alternative


no alternative


no alternative






no alternative

45All festivals are accompanied by sacrifices which Mamprusi « land-lords » (yiiridan-dima) and lineage elders make to the ancestors. On such occasions Muslims give offerings of food in the name of their ancestors to children in their local neighbourhoods. Where conversion has occurred in traditionally non-Muslim families, the lineage or household head will continue to perform sacrifices.

46Following are the descriptions of the major festivals.

Bugum (fire)

47Muslims regard bugum as the beginning and end of the complete cycle of lunations. Other Mamprusi frequently begin the count with « damma-bira, damma »; but any pair of festival lunations may be used to start the count.

48The rites performed at bugum include:

1. The throwing of fire, that is, throwing elephant-grass torches. These are lit from a single torch held by the chief (naba) or the Nayiiri in the capital. The chief’s torch is lighted inside the palace from a cotton wick. This occurs in all villages. I was told that the throwing of fire can become a mock battle and people may be injured. The festival I witnessed in 1964 was entirely peaceful, but violence is prohibited in the Nayiiri’s presence.

2. The drinking and sprinkling of water. The water used contains the chalk of Islamic prayers written on slates and then washed into large containers of water, distributed by the Nayiiri’s servants. The Nayiiri pays the Muslims for writing the prayers. This part of the festival may be performed in the provincial capitals and some villages.

3. At the capital the Limam joins the resident court of Nalerigu. He reads a prophesy containing a list of disasters to be expected in the coming year, and prayers are said. This occurs only in Nalerigu.

5. In all Mamprusi villages gifts are exchanged. Senior joking partners (see Drucker-Brown 1983) give gifts to their juniors. Women demand « vagina-money » (pen ligri) from their husband’s brothers. Husbands should make gifts to their wives. Nephews may demand gifts, or wives, or even seize the daughters of their mother’s brothers in this festival.

49The aggressive nature of bugum is emphasized when it is sometimes said that bugum is alone because « he has eaten his companion ». -biraa, the suffix which distinguishes the companion from the festival month, may be variously translated as « companion », « slave » and « little brother ». Bugum, as one might expect, is dangerous (bioo); and the prohibitions characteristic of the farming season should be observed during this lunation.

Damma-biva + damma

50I have been told that damma-bira continues the prohibitions begun in bugum. Since these lunations occurred during the farming seasons in my fieldwork I am not sure they are observed if the two lunations fall during the dry season. The period of prohibitions ends with the sighting of the damma moon.

  • 8 The predominance of damma is reflected in the fact that both the preceding lunation and the followi (...)

51Damma is the most important annual festival celebrated by Mamprusi. It is celebrated universally in Mamprusi villages by resident foreigners and visitors as well8.

52In Mamprusi Muslim theology the damma lunation celebrates the prophet’s birth. Non-Muslim Mamprusi, however, appear to be ignorant of this; and even among Muslims the prophet’s birth is regarded as a separate event from the focus of the damma itself. The central performances of this elaborate ritual commemorate the foundation of the polity, celebrate the Nayiiri and the cult of naam.

53Outside the capital the damma festival is celebrated locally in villages presided over by royal chiefs (nabia / pl. nabiisi) and in the provincial capitals. The most elaborate performance occurs in the Nayiiri’s village which is roughly the centre of the polity.

54I mention here only a bare outline of the festival and certain relevant characteristics. Figure 3 (plan of the Nayiiri’s palace) will help to follow the description of distinct phases of the ritual.

The basic elements of the damma festival

  • 9 A. Adler (1982) has commented on the manner in which a pact between indigenous hosts and immigrant (...)

551) A narrative ritual, almost a mimed performance of the pact between the first Nayiiri (Na Gbewaa, founder of the royal clan) and his indigenous hosts9. Two priest-chiefs (tendaana) perform warrior’s dances. A riderless horse is brought out of the palace and returned. The priest-chiefs threaten the Nayiiri in another dance and he weeps. The mime evokes the founding by the first Nayiiri and the relationship between the Nayiiri and his people or by extension the royal clan and the commoners.

562) Combined with the narrative ritual is the singing of a list of royal names and those of the ancestors of priest-chiefs who perform the warrior dance. The senior King’s drummer, accompanied by a phalanx of royal drummers, sings a list of the names of chiefs and their kingly fathers which begins with the founder and ends with the reigning king.

Figure 3. Plan of the Nayiiris palace

Figure 3. Plan of the Nayiiri’s palace

57This is followed by a similar performance of the names of successive priest-chiefs. The royal names are sung in the presumed order of succession of Nayiiris, and separately, the order of succession of priest-chiefs is sung. The list thus manifests a conception of progressive, cumulative, « historical » time. However the drummers insist that the list can never be exhaustive because such a performance could not be encompassed in the evening period (zaanovi, in the sense of afternoon till dusk) allotted to the singing.

583) damma is ordered by the sun’s movement throughout the day and by its disappearance at night. The Nayiiri is greeted by guests and resident elders in the bitter-hall (zong-tuua) of the inner palace yard (dindon-go). He emerges in processions three times during the daylight hours: first he moves, surrounded by the court, from the bitter-hall to the « inner hall » (zong-ni) and then to the outside yard of the palace (samani). These two processions take place during the « morning damma » (asuba damma). The afternoon damma (zaanovi damma) which is the most elaborate ceremony requires the Nayiiri to proceed from the bitter-hall, through the inner-hall, across the outer yard, and into an outdoor enclosure (leenga) opposite the palace gate. On each occasion the skins on which the Nayiiri sits are spread on special platforms (zii-ziya) located in a semi-circle from northeast to southwest of the palace gate. Thus over the course of daylight hours the Nayiiri proceeds clockwise in a semi-circle, moving farther west each time he is separated from the skins laid in the bitter-hall. He returns to the skins in the bitter-hall at the end of each phase of damma. The bitter-hall of the palace adjoins the sleeping room (piia) which will become the Nayiiri’s tomb when he dies, hence its name (see figure 3).

59At each procession the Nayiiri’s gowns increase in size, becoming bulkier and brighter. For the evening ceremony the Nayiiri wears a hooded, white gown (bulmusu) made of shimmering white cotton and silver silk. The gown is said to cover his face completely. It does in fact allow him to see but makes his face invisible unless he is viewed directly. However no Mamprusi will look directly at the Nayiiri.

60The Nayiiri’s separation from the skins and his return suggest an analogy with birth and death. The skins themselves symbolize naam (Drucker-Brown 1981), the quality transmitted from the founder of the polity and embodied in each successive Nayiiri. But Mamprusi also say that every child sits on skins in the mother’s womb, and the placenta is « the skins ».

61In effect, the prince who becomes Nayiiri, once installed, is tied to the skins from his installation (a kind of birth) to his death and burial in the sleeping room, adjoining the bitter-hall.

  • 10 This identification can be supported by other evidence in court ritual and behaviour. I did not dir (...)

62I regard the processions of the Nayiiri, his dress, and the direction in which he moves, as a metaphorical identification of the Nayiiri with the sun10.

63Mamprusi do not say explicitly, as do their neighbours the Mwba, that the sun disappears into the underworld to re-emerge each morning (de Surgy 1983: 38-39). However in the context of the entire damma ritual the Nayiiri’s separation and return to the skins may be regarded as a theatrical performance of just such a passage.

644) night damma (yungu damma), the final stage of the damma performance, emphasizes the Nayiiri’s role in maintaining normal order. It can be described as a period of egalitarian chaos. The Nayiiri appears at midnight without ceremony. He wears a dark gown and a black hat with a silver buckle. Royals who normally dance alone, on this occasion participate in a circle dance. The circle moves in a counterclockwise direction. Royals and commoners, men and women dance simultaneously. Women may appear in men’s clothes, dancing and drumming as men do. The Nayiiri may join the dance, again with no special ceremony. The Nayiiri’s wives, normally secluded, appear among the dancers and serve beer provided by the Nayiiri. The space in front of the palace (samani), normally avoided, is totally transformed. It becomes a public dancing ground. A small market springs up around the dancers where refreshments are sold. The dance ends at dawn with the unique appearance of the Nayiiri accompanied by his wives. Inside the palace a sacrifice is performed to the royal ancestors, and later the meat is shared among the participants.

655) Cumulative time is linked to spatial movement. At the singing of names in the evening damma, the musicians move slowly forward in a westerly direction some 80 metres from the palace gate to the enclosure where the Nayiiri and court witness the performances. The musicians arrive at the entrance to the enclosure just as they begin to sing the names of the reigning Nayiiri’s own « gate », the lineage segment founded by his paternal grandfather (see figure 3 for location of king and musicians).

666) Cumulative time as expressed in the name-singing is also associated with genealogical manipulation and status change. The names of all living male children of deceased Nayiiris (potential successors of the reigning king) must be sung, as must the grandchildren of Nayiiris who hold chiefship. Living daughters of Nayiiris are also named. Like the Nayiiri’s sons, all these individuals are potential founders of new gates. Deceased royals who held no office are omitted from the list, and those Nayiiris who failed to produce numerous chiefly sons may also be unsung.

Gaani and gaani-bira, or obanna + obanna kyena11

  • 11 Nadel (1949) regards the gaani festival celebrated by the Nupe as « clearly alien in origin...impor (...)

67This pair of lunations has no special festival, but damma may continue during these periods, and in some accounts the pair is described as « slaves of damma ».

Kpiini-bira + kpiini (also called kpiini-bila + kpiini)

68The festival occurs on the tenth day after the kpiini moon is sighted. It may also be called kpiini feibugu (lit. beating -feibugu, the guinea fowl -kpangu/pl. kpiini). In villages of the outlying districts the guinea fowl is plucked, beaten with twigs, then cooked and eaten. In Mamprusi legend guinea fowl are often women, particularly royal women. The delicately speckled grey and white feathers of the guinea fowl are similar to certain locally dyed cloth which women wear. Furthermore, guinea fowl will return to the wild if abandoned or ill-treated. Women, and also the commoner followers of a royal chief, are said to share with guinea fowl the tendency to run away from their « owners ».

69The kpiini celebration is a domestic one; it occurs simultaneously in different households and kin groups. Gifts are given by senior to junior joking partners (Drucker-Brown 1983). Mother’s brothers give gifts to sister’s sons’, grandparents to grandchildren. In outlying districts, I was told that husbands should give their wives fowl to sacrifice at their destiny (wuni) shrines.

Nol)ri-biraa + noloori/konyuuri-kyuu-daari, lasiyaa-biraa/lasiyaa

70After kpiini there is a period of taxonomic confusion. In this period the rituals performed exclusively by Mamprusi Muslims take place.

71In the lunation called nol)ri Muslims observe a fast during the hours of daylight (daari) which is broken after dark. Nol)ri (lit. mouth -nori, locked -l)’ari) is followed by konyuuri-kyuu-daari (lit. water -konyuuri, drinking day -daari). Unlike other festival days which occur on the tenth day after the moon is sighted, konyuuri-kyuu-daari, the day of drinking water, begins the very day the new moon is sighted. Thus the nol)ri fast ends with the beginning of the new lunation. However, konyuuri-kyuu-daari may be omitted in the list of lunations and the pair lasiyaa-bira + lasiyaa will then be named. I do not know the content of the festival associated with lasiyaa. I believe I share this ignorance with many non-Muslim Mamprusi. Because nol)ri continues for an entire lunation and fasting is regarded as curious behaviour when continued for a long period, the fasting which occurs at nol)ri is well-known to non-Muslims. Similarly the end of the fast at konyuuri-kyuu-daari is known.

Kyiimsi-biraa + kyiimsi

72If the lasiyaa lunations are both named, kyiimsi-biraa must be omitted from the lunation list. However in all lists, kyiimsi precedes bugum (see Chart 2). The kyiimsi ritual, like damma, is divided between a morning and an afternoon section. Throughout the day, as at damma, the Nayiiri’s dress changes and he moves in processions to and from the « bitter-hall » of the palace.

73At the morning performance the same spears carried at damma by the priest-chiefs are carried by a resident elder. The riderless horse, led by a servant, heads the procession of the Nayiiri and court to a cleared field northeast of the palace. The princes follow the procession mounted on horseback, their horses elaborately dressed. The entire court, with many visitors, gathers there. The Limam, surrounded by junior members of his lineage, each holding a section of his robe, faces the court and performs prayers. After the procession returns to the palace, the princes race their horses, rifles are fired by the warrior officials and damma dancing is performed.

74As at damma the Nayiiri and court witness a performance of the royal names in the afternoon. This performance is identical with the damma ritual, excepting that the horse is not present and the priest-chiefs do not participate. Note however that the spears with which the priestchiefs dance are carried out in the morning to the prayer-ground and the riderless horse leads the procession. The princes make great display of their horses in the morning phase of kyiimsi.

75As at damma a sheep is sacrificed in most Mamprusi houses at dawn of the following day. The Nayiiri provides Muslims with their special food, cooked rice and meat. At kyiimsi meat is exchanged among all the resident elder’s households. At damma however, all households receive meat from the palace.

76The kyiimsi festival repeats essential elements of damma. It also balances damma in that it focuses on the contribution of the Limam and Muslims rather than on the priest-chiefs. The Islamic prayers on the Nayiiri’s behalf are analogous here to the tendaana’s dance at damma. At no point however, does the Muslim performance physically menace the Nayiiri in the way the priest-chiefs do.


77In societies where the seasonal year and a count of lunations are merged to create a single interval of duration, collective, public ritual may be directly linked to the changing seasons. In such a calendar, public ritual may be distinguished from rituals which mark life-cycle passages, though these may also be merged with collective seasonal celebrations or timed to coincide with changes in the seasonal year.

78Where the lunation count is detached from the seasons, festivals – that is, the collective rituals which emphasize social cohesion – cannot be directly correlated with seasonal change. In the Mamprusi case, the damma festival, a celebration of the particular constellation of forces that give cohesion to Mamprusi kingship, is segregated from rites regarded as affecting the seasons by use of the lunation-count. The participation of the kings and chiefs is essential for seasonal rituals and those marking life-cycle passages, but such rituals are given no public performance at predictable intervals of time. Thus separated in time from the seasonal year and distinguished in their public character from all other rites, the festivals marked by the lunar count emphasize the Nayiiri’s unique connection with the daily solar cycle and the social order.

79Leach’s conclusion that the lunation must be merged with a solar year in order for the calendrical system to have practical use, does not hold true for the Mamprusi. Short-term prediction is made within the context of the lunation-count; for despite the movement of festivals over time through the solar year, the cycles emphasized within the lunation correspond roughly to a seasonal year.

80Evans-Pritchard’s contrast between ecological and structural time is relevant. In the Mamprusi case the set of lunations is a sequence measured by observation of a single feature of the natural environment. It is progressive only in that it consists of completed annual cycles. Nevertheless it is a measured sequence which reflects certain Mamprusi views of social organization. The pairing of lunations, though it may aid primitive techniques of measurement (and a pairing of lunations short + long occurs vestigially in our own lunar-count), also reflects more general Mamprusi conceptions. At another level however, the singing of names in the damma performances shows that progressive, « historical » time and structural time can be used simultaneously. The name-singing links the temporal interval between the founding of the polity and the present with the spatial progress of the singer from the palace gate to the enclosure where the living Nayiiri sits in court. This may be regarded as a proclamation, a kind of demonstration, of progressive time. However the same rite is clearly « structural ». It serves to rearrange and to publicize the claims of individuals and lineages on office (naam). In this respect Mamprusi history of course ressembles our own. These observations require greater elaboration and have major implications for understanding the Mamprusi polity and the history of this region. However these ideas cannot be developed in the context of a mainly ethnographic paper.

81The characterization of intervals of time as dangerous (bioo) and the demarcation of these intervals by specific prohibitions unify the different intervals of measured time. The overlapping of such intervals inevitably creates complexities which make divination essential for the Mamprusi. The concept of more or less auspicious time also provides diviners with a tool for re-making order of the resulting disorder.

82The conceptual identification of the various intervals of measured time by means of a common concept of danger coincides with the emphasis Mamprusi place on the Nayiiri as a unique force in the polity. The lack of coordination between the seasonal year and the lunations, suppressing as it does a public, collective celebration of any other forces « beyond human control », highlights the importance of the Nayiiri. Measurement of lunar cycles as well as the measurement of other time sequences thus serve the dominant concern of the Mamprusi with courts and kingship.

Haut de page


A. Adler (1982) : La mort est le masque du roi, Paris, Payot.

E. Arana & Swadesh, M. (1967): Diccionario analitico del Mampruli, Museo de las Culturas, INAH, Mexico.

H. Blair (1929): Unpublished manuscript on file at the Balma Library, Legon, Ghana.

Braimah (pre-1963): Unpublished manuscript on file at the Balma Library, Legon, Ghana.

S. Drucker-Brown (1975): Ritual aspects of the Mamprusi Kingship, Cambridge, African Studies Center, African Social Research Documents, vol. 8.

S. Drucker-Brown (1981): « The structure of the Mamprusi Kingdom and the cult of Naam » in The study of the state, Claessen, H.M. & Skalnik,P., eds., The Hague, Mouton, pp. 117-131.

S. Drucker-Brown (1983) « Joking at death » in Man, vol. 17, n° 4, pp. 714-738.

E.E. Evans-Pritchard (1939) « Nuer time-reckoning » in Africa, vol. 12, pp. 189-216.

M. Fortes (1970) « Ritual festivals and social cohesion in the hinterland of the Gold Coast » in Time and social structure, London, Athlone Press (reprinted from original in American Anthropologist, vol. 38, 1936).

M. Fortes (1975) « Tallensi ritual festivals and the ancestors » in Cambridge Anthropology, vol. 2, n° 2, pp. 3-32.

E. Leach (1950) « Primitive Calendars » in Oceania, vol. 20, pp. 245-262.

E. Leach 1954 « Primitive time-reckoning » in History of technology, Ch. 5, edited by Singer C, Holmyard E.J. and Hall A.R.

B. Malinowski (1927) « Lunar and seasonal calendar in the Trobriands » in Journal of the Royal Anthropological Institute.

T. Mckaskie (1980) « Time and the calendar in nineteenth century Asante: an exploratoty essay » in History in Africa, vol. 7, pp. 180-200.

S.F. Nadel (1949) « The Gani ritual of Nupe » in Africa, vol. XXIX, p. 183.

S.F. Nadel (1954) Black Byzantium, London, Routledge & Kegan Paul.

A.J. Naden Mampruli Dictionary (English-Mampruli), unpublished manuscript.

A.J. Naden The Mamprusi calendar, unpublished manuscript.

A. de Surgy (1983) La divination par les huit cordelettes chez les Mwaba-Gourma, Paris, L’Harmattan.

J.S. Trimingham (1949) Islam in The Sudan, Oxford University Press.

J.S. Trimingham (1959) Islam in West Africa, Oxford University Press.

D. Turton & C. Ruggles (1978) « Agreeing to disagree: the measurement of duration in a southwestern Ethiopian community » in Current anthropology, vol. 19, n° 3.

K. Voller (1910) « Calendar (Muslim) » in Encyclopedia of religion and ethics, Hastings, J., editor.

Haut de page


1 This paper is the result of field work carried out in the Mamprusi region from June 1963 to April 1965. I am grateful to Dr. A. J. Naden of the Wycliffe Institute, resident in the area from 1975, for his unpublished article, The Mamprusi Calendar, his letters describing the damma performance of 1980 and his unpublished Dictionary of Mampruli. It is a pleasure to thank Dr. Naden for making his notes available to me. He is not responsible for the conclusions I have drawn. I must also thank Dr. Keith Hart, M. S. Abudulai and Dr. L. M. Brown for helpful comments on this paper.

2 Mackaskie (1980) demonstrates, for the Asante, the political importance which manipulation of such intervals gave the Asante, and documents the prevalence of this conception among the Asante.

3 In the Mamprusi region the site of a chiefly grave, or the grave of a Nayiiri, may be a tengbang. Such shrines appear to have a variety of origins.

4 I am grateful to Dr. A. J. Naden for observing the accuracy of these gestures which he says correspond to within half hour of time measured by a modern clock.

5 Nadel (1954) describes the Nupe agricultural calendar used by « peasants » as follows: « (It is) based on lunar months and starts with the first rains, i.e. roughly in April. From the beginning of the rains people count twelve, thirteen, sometimes fourteen rains until new rains break. The last month of the old year (i.e. the thirteenth or fourteenth month) is then identified with the first month of the new year ». Nadel notes that this calendar was being superceded by the Nupe adaptation of a Muslim calendar. The two might coincide « where the Mohammedan New Year falls in April, but if the lunations differed by more than one, the peasant’s time reckoning was thrown into serious confusion. »

6 Trimingham (1949, 1959) gives descriptions of Islamic festivals which parallel those of the Mamprusi. Thus aspects of bugum, damma, kpiini feibugu and kyiimsi may be identified respectively with muharram, Ramadan (?), isfahan and id al-adha. Iskutun, the tenth night of the ashuva month in the Nupe calendar, Nadel identifies as "the feast of the beating" (1954).

7 I am grateful to M. S. Abudulai for this observation. During 1981 he remembers news reports of fighting in northern Ghana over the date of Ramadan. Local communities of northern Muslims refused to celebrate the festival on the date announced by the National Muslim Council and preferred to use the dates fixed locally. The situation is reminiscent of the conflict between northern and southern branches of the Christian church in Britain over the date on which Easter should be celebrated. This conflict was resolved by the Synod of Whitby in 664.

8 The predominance of damma is reflected in the fact that both the preceding lunation and the following pair of lunations are said to be « slaves of damma ». Thus, of twelve lunations, four are linked to this festival. Morover, elements of the damma festival, notably the dance performed exclusively by royals, and the performance of royal names, reappear in other festivals.

9 A. Adler (1982) has commented on the manner in which a pact between indigenous hosts and immigrant founder is commemorated in the festivals of the Moundang. This is, of course, a typical pattern in « origin myth ».

10 This identification can be supported by other evidence in court ritual and behaviour. I did not directly ask Mamprusi if the Nayiiri was « like the sun » nor did I ask « what happens to the sun at night ». However my evidence emerges from behaviour rather than verbal explanations.

11 Nadel (1949) regards the gaani festival celebrated by the Nupe as « clearly alien in origin...imported into Nupe by Moslem immigrants from the north... probably about 1760-1770. The main theme of the gaani ceremonial is the periodical reconstitution of village age-grade associations ». Though Mamprusi « young men » in larger villages may constitute themselves into an informal group which may have a leader (nachinaba), no age-grade structure exists.

Haut de page

Table des illustrations

Titre Figure 1. Subdivisions of the day (daari) showing the danger periods at dusk and dawn
Fichier image/png, 228k
Titre Figure 2. Basic plan of a Mamprusi compound
Fichier image/png, 203k
Titre Photo 1. External view of the west facing gate to a Mamprusi compound
Légende It is built into the circular entrance hall (the zongo). Note the sharp shadow thrown by the evening sun. Logs placed in front of the zongo to the right of the gate are sitting places for men.
Crédits S. Drucker-Brown
Fichier image/png, 835k
Titre Photo 2. The same gate viewed from inside the compound
Légende A king’s elder is leading a procession to greet a funeral.
Crédits S. Drucker-Brown
Fichier image/png, 617k
Titre Figure 3. Plan of the Nayiiris palace
Fichier image/png, 527k
Haut de page

Pour citer cet article

Référence papier

Susan Drucker-Brown, « Calendar and Ritual », Systèmes de pensée en Afrique noire, 7 | 1986, 57-86.

Référence électronique

Susan Drucker-Brown, « Calendar and Ritual », Systèmes de pensée en Afrique noire [En ligne], 7 | 1986, mis en ligne le 05 juin 2013, consulté le 26 mai 2017. URL : ; DOI : 10.4000/span.569

Haut de page


Susan Drucker-Brown

Faculty of Archeology and Anthropology, University of Cambridge

Haut de page

Droits d’auteur

© École pratique des hautes études

Haut de page
  • Logo EPHE
  • Logo CEMAF