Printer-friendly format  Article citation        


The Dance and Status in Swahili Society

Marjorie Franken


In this paper, I illustrate how dancing, as one aspect of ngoma, is one of many status-laden activities in Swahili society. Through a brief introductory analysis of distinctive structured movement patterns, the reader can see exactly how understanding of the uses of the body itself and the space around it contributes significantly to our understanding of the nuances of social roles and the differentiation of Swahili social status on the coast of postindependent Kenya.

     The Swahili are a Muslim, African people characterized by a stratified town culture of many centuries duration. Their society illustrates the integration of Islamic cultural codes into indigenous Swahili urban culture, providing a synthesis that has been incorporated into a particular set of performances known as ngoma. Swahili ngoma, as a category of activities that includes music, poetry and dancing, is one of many activities in Swahili urban society that mark social class.

     This paper will examine some of the ways in which the most elite class, the aristocrats of Swahili towns usually referred to as waungwana, strove to establish and maintain this distinctive, recognizable, important place in town life through ngoma. As a category of activities, ngoma has frequently been overlooked in the literature, yet it can be shown to shed considerable light upon social roles and class distinctions in Swahili society.


The Kiswahili word ngoma means "drum" or "music." The verb form is kiatezea ngoma, "to play a dance," and I will use it throughout this essay instead of the phrase "music and dance event." I wish to avoid the word dance because I don't want to limit readers' conceptions of these performances to Western notions of "steps" or "step patterns" and to the notion of movements performed only while standing on the feet. Some of the performances I will describe take place while the participants are seated, but, in spite of that, they fit into the Swahili category ngoma.

     The status markers that were most obvious were tall, white-washed stone houses. These were so distinctive, set amid the small, brown mud-and-thatch houses of a town that they were mistaken by Europeans as part of a foreign cultural heritage. Inside those houses were still more material goods and also activities that marked the inhabitants as upper class. The familiar description of imported porcelain from China and Persia comes immediately to mind.

     The stone houses can be thought of as showcases for a wide assortment and great variety of status-marking objects and activities. One activity possible in a large house with an enclosed courtyard is the complete seclusion of women. Devotion to religious observance and study were perhaps second only to pedigree in marking waungwana status, and seclusion of women is a highly noticeable act signifying religious purity and wealth. A separate room into which women retreat when male guests are present (a "harem" or ndani) and an enclosed area for domestic tasks were only affordable for wealthy upper-class people.

     Another function of large houses was their use as a framework for waungwana rituals and ceremonies. Indeed, it has even been suggested that the zidaka on the back wall of the ndani (women's quarters) were used primarily as a backdrop for the most lavish ceremonies of all—those done for a new bride (Allen 1979). In the more public rooms at the front of the house and in the courtyard ngoma, (music and dancing events) could be held. Poetry composition was another activity that required time and tutelage to study and perfect, which was another marker of high status.

     The poetry competitions between a women's team and a men's team were possible in such a house, when a purdah curtain across the first great room could separate the competing teams. What better way to display all one's status-markers at once than in a stone house, richly appointed, surrounded by one's elite friends, performing in poetry and dance recitations? We will now focus on these ngoma occasions and their forms of activity as another kind of status-marker in Swahili culture.

     We begin by discovering two ngoma that, in the past, were done in this setting of a stone house as exclusive waungwana affairs. The first was ngoma la hazua, a men's dance done for weddings and circumcisions. The dancers were adult men dressed in their best kanzu (a spotless white and freshly pressed prayer robe) and their best kofia (a richly embroidered prayer cap). In addition to drums and music, a singer intoned poetry, recounting the noble deeds or distinguished ancestors of the sponsoring family.

     The men moved in unison, in a line formation; they stepped at a diagonal forward on the right foot, thrust the sword they carried to a fully extended arm position to the upper right. They fell back on the left foot while lowering the sword, stepped to the side right while moving the sword across the body and wiping it with a handkerchief held in the left hand, after which they stepped left while lowering the sword, point to the ground (see Figure 1). This pattern was repeated as long as the performers wished to keep on. After the performance, a feast was served.

Figure 1
Figure 1. Waungwana Dances—Ngoma La Hazua. The performance space is a rectangular room in a house or kiwanja (area between houses).

     Foregoing analysis for a moment, I will describe a second dance that was also done in private homes at the turn of the century. This dance is called lelemana, an "adult" (that is, married women's) dance done usually as a rite of passage.

     The lelemana costume in those times was suruali ya njiwa, so named because the fitted leg ended in a ruffle like the fluffy feathers above a dove's foot. The body and sometimes the head were covered by a pair of leso that were bright rectangles of printed cloth. On the fingers, the dancers wore mbiu, little gazelle horns tipped in silver, with the hollow end fitted over the fingertips, perhaps as finger cymbals are worn. Sometimes a single mbiu was struck with a small stick. The dancers stood in a line or in ranks and followed a leader in a sort of drill. The motions were arm extensions and finger flexions done in a smooth, slow, graceful manner (see Figure 2). A singer also intoned poetry to the musical accompaniment. Again, the dancing was followed by feasting.

Figure 2
Figure 2. Waungwana Dances—Lelemama. The performance space (women covered) is a tent.

     These two ngoma are still performed today, but there have been some alterations over the last eighty or ninety years. Ngoma la hazua is most often now seen outdoors in the town of Lamu, performed for weddings or during Maulidi week.

     Maulidi is a name derived from the Arabic term mawlid or mulid, meaning a poetic recitation of the events surrounding the birthday of the Prophet. Maulidis are recited on any important occasion but most frequently on the anniversary of the Prophet's birthday. This recitation form is pan-Islamic, but the Swahili have added some variations and adaptations of their own. Maulidi can be sung for small family occasions as a devotional exercise or can be public events as parts of weddings. Characteristic movement patterns frequently form an integral part of the performances (Franken 1987).

     The largest maulidi performed today is held in Lamu, one of the most traditional Swahili towns and distinguished for both religious learning and famous poets. Maulidi week on the coast is somewhat like Christmas week in America, in the sense of being a time for parties, traveling, visiting, fun, and some serious religious activities. The centerpiece of the week is a maulidi recitation at the biggest mosque in Lamu on Thursday night. Hundreds of people attend, traveling from all over East Africa (Franken 1986).

     In this context, men use bokora, thin wooden canes instead of swords. Lelemana is still sometimes performed in Mombasa for weddings, and the popular dress is a military uniform style, such as sailor-collared dresses. The mbiu are gone.

     We can now ask, what are the waungwana status emblems in these two historical performances as I have described them? First (and indispensable) was the setting of a stone house. Next were the guests—one's family and friends, also from the upper class. Fine clothing is a familiar status symbol. Men used both quality materials and a religious style of clothing to signify high status. Women displayed affluence by having special dancing clothes and paraphernalia. Poetry singing was a mandatory part of the ngoma, and, finally, food was offered. These are the familiar emblems of high status: wealth, as shown in houses, clothes, food, fine friends and refined pastimes. It is easy to see counterparts in our own and many other societies.

     What is not so obvious is that the actual danced movements are also valued as waungwana kinds of movement. By way of contrast, we will now look at two ngoma of very humble classes of people in Swahili culture in order to make the contrasting movement styles dear.

     The contrasting dance I will describe is called uta. This is another men's dance that is one of the showpieces of Maulidi week in Lamu. It is known as the "coconut-cutters dance," and it has always been performed out of doors in a kiwanja, a cleared sandlot area between houses. The name uta means "bow" as in "bow and arrow," pointing to the principal instrument plucked while the player also intones the uta song.

     More vital for dancing purposes are the leg frames worn by the dancers. These are called msewe and are woven on a strip of palm fiber rope, about every six inches. The rope is wrapped around the dancer's legs from ankle to knee. The dancers thus make their own music, for no drums or other instruments are used. They wear typical Swahili men's work clothing; a kikoy (like a sarong), a shirt and sandals, and a prayer cap (kofa), perhaps.

     The dancers form a rough circle and leaning on canes placed before their feet, stamp the uta rhythm in unison, producing a percussive rustling sound of the fiber rattles (see Figure 3). Occasionally, a dancer breaks out of the circle, raises his cane over his head, and twirls through the center to take a new place on the opposite side of the circle.

Figure 3
Figure 3. Watwana Dances—Uta. The performance space is out of doors in the kiwanja.

     Another men's dance, mwaribe, can also be seen in Lamu during Maulidi week, and it is then done out of doors by young men. The instruments are msondo (also an alternate name for the dance), which is a long narrow drum somewhat similar to a conga drum. The dancers wear smart street clothes—trousers and new shirts—"Western wear."

     The dance begins with one person inviting a partner into the dance space by hopping onto one foot to a position directly in front of the partner, who then responds by hopping forward in unison as the first person hops backward.

     The sequence continues: hop forward on the right, put feet together, hop back on the left, (again, feet together), step in place, and step in place again. The arms and hands are held chest high, and the hopping motion gives them the appearance of flapping like birds' wings (see Figure 4). This dance is the only one I ever heard of that is sometimes performed by mixed-gender groups.

Figure 4
Figure 4. Watwana Dances—Mwaribe. The performance space is out of doors in the kiwanja.

     My informant told me that, forty years ago, on Kiunga Island (north of Lamu), mixed groups of men, women, and children did this dance together, even single and married women dancing with the men. This dance is also said to be especially liked by the Pokomo, a people who live adjacent to the Swahili area on the coast. They and other related inland groups, the Miji-Kenda people, for example, are said to have other, similar dances. In larger Swahili towns, however, mwaribe is never performed by mixed groups. Young girls dance mwaribe indoors, at strictly secluded girls' parties, and young men perform it outdoors, as described here.

     These dances, the mwaribe and the uta, may be said to be characteristic of the watwana class in Swahili society. The watwana are the lowest class. In fact, one meaning of the word is "slaves." These people are said to lack all waungwana attributes. That is, they are recent arrivals in town life, they have no pedigree or distinguished ancestors, they practice a slack sort of Islamic religion, and, of course, they are economically poor. Uta is a dance of a very humble occupational group: coconut cutters. As a type of work, it is dirty and hard, therefore, only for poor and unrefined men.

     Mwaribe betrays its humble status by its similarity to Miki-Kenda dances which are done by nonbelievers and, in relation to town dwellers, lower-class people. It is low in status because of the tendency to perform it in violation of purdah structures.

     It is interesting to compare the two pairs of dances: ngoma la hazua and lelemana, which exemplify waungwana ngoma, on the one hand, and uta and mwaribe, which are watwana dances, on the other. What are the contrasting attributes of each pair? Before answering that question, a clarifying point must be made. The reason that watwana dances are "low status" in Swahili society is simply because lower-class Swahili people dance them. From the point of view of the outside observer, there is nothing intrinsically better or worse about the movements themselves, but Swahili society assigns class meanings to these dances; therefore, they are talked about as if there were something in the nature of the dances themselves that was "lower" or "poorer" than the ngoma la hazua dances.1

     The relatively "easy-to-see" contrasts between the two types of dancing can be thought of in this way: the waungwana dances are distinguished by place, music, poetry, dress, and feasting. They are not mixed gender dances. Groups of men and women are kept strictly separated. Only ngoma la hazua dances can be performed in the waungwana context described in detail earlier, notably, the great stone houses. Men have an option also to perform these dances out of doors, but women must maintain their elite seclusion standards. If they do dance outdoors (as I saw in Mombasa in 1984), they do so inside a tent that conceals both dancers and audience.

     Waungwana styles of dress also distinguish aristocratic dances from all others. Men display their religious status by dancing in their finest prayer dress, and women show conspicuous consumption by making special costumes and paraphernalia. Waungwana music is relatively complex and refined. It uses a full complement of Swahili percussion instruments: drums and utasa (a metal tray played with brushes); a melodic instrument, either a zumari ya mtapa (like the Arabic mizmar) or a Western trumpet. Both mark the players as sophisticated, cosmopolitan people. The lyrics of the songs consist of poetry composed especially for the occasion by the best poets of the town, who are members of the performing class.

     The watwana dances are marked by being held out of doors, even in mixed-gender groups. People dance in clothes that do not advertise piety or wealth. Their musical accompaniment is confined to percussion only; their singing is identifiable mainly by its volume, rather than refined or witty poetry composed especially for the event. These contrasts are tabulated below in Figure 5.

Figure 5
Figure 5. Table of some contrasting features of the Waungwana, Watwana and Chakacha/ Ngomi Ya Ndani dances.

     The first column (which remains to be discussed), contrasts the movement patterns of the dances of the two extremes of Swahili class structure. The elements in this column show that the Swahili identify their status by moving differently in dances and by using different body parts. Specifically (in a rather oversimplified manner), waungwana danced forms can be characterized as "arm dances." Watwana dances can be characterized as "foot dances."

     Waungwana forms use the feet only to support and give spatial direction to the dance and to the important features of the dance—the arm extensions and hands, adorned with swords or mbiu. Movements are up and away from the torso. Watwana forms use the feet as part of the percussion and as a focus of the danced movements. The dance takes place primarily in a vertically defined space, the dancers' weight moving in an up-and-down pattern.

     If the reader finds this contrast surprising, so too, I think, would the Swahili. I did not ask them about these things during the research period because I didn't notice them myself until I had returned to the United States and was in the process of analyzing my data. I think this aspect of dancing is largely out of awareness, just as the differences in upper-class and lower-class speech in many societies are salient defining features of class, but they are not attended to much of the time. I do not think that the dance movements that differentiate classes are deliberate inventions, any more than class speech differences are deliberate inventions. I do, however, think that they are a result of a process of participation, feedback, and observation that a dance community, like a speech community, takes part in as it sees itself dancing.

Dancing, Status, and Role

Dancing communicates. It can communicate a narrative plot, as in many classical ballets, or emotional and ideational affects, as in much early American modern dance. In a society such as that of the Swahili, where there are no professional choreographers or ballet masters, dancing is intended to communicate something entirely different, and it is not the result of a single individual's creative endeavor.

     Dancing in Swahili society reiterates to the society at large (the extended audience) something about the dancers' status and role. For example, this dancer is a young working-class male; that dancer is a high-status bride. These positions in society are usually indicated by many other forms as well, that is, dress, housing, paraphernalia, and so on. If the danced action itself has anything to add to this picture of status and role, it will have to utilize the medium of movement itself to formulate the message.

     Phrased in another way, dancing for the Swahili (and many other societies) is an important part of class-differentiated behavior because it uses another medium—another form of social actions—to designate different social groups. Dances are at once redundant in actually defining groups, but they are unique in isolating movements of different parts of the body as markers of contrasting categories of people.

     I must add here that there are no formal rules for Swahili dance participation or sponsorship, nor are there fines for doing the "wrong" dance. Anyone is free to dance in any way they like, to sponsor dance types of any kind they like. The trick is to get all the right people to attend. Assuming one's pocketbook can stand the strain, one can only ask, not command, the best musicians and poets to perform. Likewise, one can only invite the best people or the traditional performers of a certain dance form. One can even join a dancing group of a different social category, and no one will forcefully object. They may cut you dead, but they will not call a "bouncer."2 The "rules" are unwritten and consensual. The movements are prescribed and recognized, and the dances are advertisements for, and reaffirmations of, the dancers' social position.

     To conclude, I want to look at another dance, probably the most well-known of all and perhaps the most ambiguous—but not, I hope, confusing. This dance is chakacha, the women's hip dance that all women, regardless of class, participate in. Chakacha appears to be a new name for a very old dance, formerly called ngoma ya ndani, meaning "dance of the private women's quarters," that is, a harem dance.

     Because it is intentionally erotic and celebrates female sexuality, it fits the Western stereotype of a harem dance. Chakacha could be and still is done at any time when a group of women are relaxing together at home. It was also done as sexual instruction for a bride during wedding festivities. The new name seems to denote its outdoor version, still maintaining purdah, however, by the use of gunny sack tent walls with a fancy ruffled awning above.

     When the dance is held indoors, women wear their most glittering finery: jewelry, flowers, hairdos and makeup. Out of doors, most of this opulence is partly concealed by black bui-bui veils. The instruments for the dance include a full range of drums and percussion, a zumari, usually a Western trumpet and sometimes an electric organ. So far, this dance—this ngoma form—has attributes that make it seem like a waungwana display of status. But, what of the movements? Do they fit the pattern of differentiated movements by social rank in the same way as the other dances we have examined? Here is the anomalous part of chakacha (ngoma ya ndani).

     The focus of body movements for chakacha is neither with the arms/hands extended horizontally in space (like the lelemama and ngoma la hazua), nor are [the movements] with the feet in the vertical, under-the-body space, (like the uta and mwaribe of watwana). Chakacha focal moves consist of two forward hip drops to the right, followed by two on the left, repeated throughout the dance (see Figure 6).

Figure 6
Figure 6. Women's Harem Dances—Chakacha/Ngomi Ya Ndani (variation 2).

     A meter-length cloth (leso) is tied tightly around the dancer's hips to accentuate the movements. The arms are relaxed, flexed at the elbow, hands horizontal, perhaps consciously framing the movement of one hip. The feet do very little, perhaps shuffle forward in tiny steps as alternate hips drop.

     There is no necessary spatial formation. Sometimes the dancers move about following one another as in the familiar ‘conga' line.3 Sometimes the dancers perform solo; sometimes they pair off and dance toward each other in a friendly competition. If we were to diagram chakacha movement, it would look like the bottom entry in the first column of Figure 5, putting it midway between the others, in terms of the body space utilized. The movements do not require the arms in extended spaces. They are vertical, but there are no percussive elements—no impacts of feet to the ground. Can we explain the median position and movement of this dance by other contextual clues?

     One is tempted to think of other female dances involving hip drops and hip sashes—Middle Eastern ‘belly dances' come to mind immediately. However, this genre as an origin is probably not likely for several reasons. First, the most similar movements are from the Saidi region of Egypt, which is an area with far less contact with the Swahili than other Middle Eastern areas. Second, the most recent (and most intense) contact was with Oman, but women in these areas do not dance primarily with hip movements, as they wear voluminous dresses when dancing that conceal everything but the head, the neck, and lower arms, and most of their characteristic dance movements involve these body parts.

     It seems reasonable to assume that chakacha indicates something else about Swahili society because it was (and is) danced by all women regardless of class. Chakacha danced movements represent women's special role (and rank) in Swahili society. Their role was reproduction. Their rank was in one sense ambiguous. In another sense, their rank was high, although tightly limited, contained, and controlled.

     That all women, regardless of social class, could perform chakacha is not surprising because all women in Swahili society could (and do) produce legitimate offspring. Female household slaves often achieved manumission by producing a child fathered by the head of a household. A woman's occupation, then, no matter what her pedigree, was primarily child-rearing and domestic work. Unlike men who were divided into occupational statuses of high or low prestige and who indicated rank by dance types, women were united in the kind of labor they performed. They also danced their occupation, so to speak, with their hips.

     The tension between Islamic ideas and African heritage also affected women's position in the Swahili context. As many authors have noted, Swahili women—as part of their African heritage—have more rights and higher status than their Arab sisters. Property ownership and management and independence after divorce are two important rights Swahili women enjoy. But the Islamic ideal puts women in a totally dependent and subservient condition. Swahili women had more latitude in their society, with the option to act either as autonomous African women and/or subservient Muslim women. Swahili women danced a dance that was all their own, that celebrated their clear and vital role in society as child-producers and bearers, which left behind the high-status dance of the upper body and the low-status dances of foot movements. Women were squarely in between, in religion and rank—and in their danced movements.


1 This is an important theoretical point which illustrates the essentially social and linguistically based nature of human behavior with reference to assigned meanings. [Editor's note]

2 An American idiomatic expression for a person in a public or private venue such as a bar or club who forcibly ushers unwanted persons from the premises.

3 Visually, the Conga line is a "follow-the-leader" line of dancers, true, but there the resemblance ends. The Conga is a West Indian freedom dance (the form to which I refer originated in Trinidad), and the spatial form is a line because former slaves were chained together so they had to dance following one another. We make a point of the issue because it is just here that the kinetic or kinological (roughly comparable to phonetic and/or phonological) elements of danced movements (even some spatial relationships, as we illustrate here) may appear to be the same, but semasiologically, the elements are not the same at all. [Editor's note]

References Cited:

Allen, J. de Vere.
1979. The Swahili House: Cultural and Ritual Concepts Underlying Its Plan and Structure. Art and Archaeology Research Papers 9: 1–32.

Franken, Marjorie A.
1986. Anyone Can Dance: A Survey and Analysis of Swahili Ngoma, Past and Present. PhD diss., University of California, Riverside.
1987. Women's Dances on the Swahili Coast. UCLA Journal of Dance Ethnology 11: 30–38.

Additional References on Swahili Dances

Allen, J. de Vere.
1974. Town and Country in Swahili Culture. Symposium on Leo Frobenius, Seminarbericht de Deutschen. UNESCO Kommision No 25: 293–316.

Bakari, M. Mm
1981. The Customs of The Swahili People. Trans. J. W. T. Allen. Berkeley: University of California Press.

Nicholls, C. S.
1971. The Swahili Coast. London: Allen & Unwin

Nurse, D. and T. Spear
1985. The Swahili: Reconstructing the History and Language of an African society, 800–1500. Philadelphia: University of Pennsylvania Press.

Prins, A. J.
1967. The Swahili Speaking People of Zanzibar and the East African Coast (Arabs, Shirazi and Swahili). Ethnographic Survey of Africa, Part X11. London: International Africa Institute.

Ranger, T. O.
1975. Dance and Society in Eastern Africa: The Beni Ngoma. Berkeley: University of California Press.

Salim, I. A.
1973. Swahili Speaking Peoples of Kenya's Coast, 1895–1965. Nairobi: East African Publishing House.

Strobel, M.
1979. Muslim Women in Mombasa, 1890–1975. New Haven, CT: Yale University Press.

Trimingham, J. S.
1964. Islam in East Africa. London: Oxford University Press.



Content in Journal for the Anthropological Study of Human Movement (ISSN 1940-7610) is intended for personal, noncommercial use only. You may not reproduce, publish, distribute, transmit, participate in the transfer or sale of, modify, create derivative works from, display, or in any way exploit the Journal for the Anthropological Study of Human Movement database in whole or in part without the written permission of the copyright holder.
© 2012 by the Board of Trustees of the University of Illinois
Terms and Conditions of Use.