Rains Gone Bad, Women Gone Mad

This article reconsiders the argument that ‘rituals of rebellion’ be seen as womens ritual response to everyday patriarchal structures — an argument originally suggested by Gluckman, but recently evoked by Spencer, the Creiders, and others — in light of recent anthropological theorizing on gender. Using as an example one such womens ritual among the Ihanzu of Tanzania, I show how this particular formulation reduces complex notions of gender and gender practices to unnuanced, monolithic, and all-encompassing gender-systems, both in everyday and ritual realms. This it does, first, by conflating gender ideals with gender behaviours and, second, by ignoring peoples conflicting ideas about gender. By problematizing these contradictions, I demonstrate how Ihanzu womens rites are not about rebellion but gender complementarity, played out by women dancers embodying both genders simultaneously. Above all, this case compels us to rethink, fundamentally, ‘rituals of rebellion’ and ‘patriarchy’.

---

The most important of these [rain] rites among the Zulu required obscene behaviour by the women and girls.... At various stages of the ceremonies women and girls went naked, and sang lewd songs. Men and boys hid and might not go near....

[This] is my first example of a ritual of rebellion, an instituted protest demanded by sacred tradition, which is seemingly against the established order, yet which aims to bless that order to achieve prosperity (Gluckman 1963: 113–14).

For the second time on that unseasonably hot January day in Ihanzu, north-central Tanzania, the women danced naked and sang their way through the village, bellowing and gesticulating obscenities as they went. Men hurriedly removed themselves from their path, for fear of being caught, unceremoniously stripped of their clothing and ‘played with’ by the unruly mob. One hapless middle-aged man, apparently too slow to outrun the wild women, had already been captured. He managed to avoid further incident by quickly agreeing to pay a fine.

Such a fate, I was fairly hopeful, would not befall me. This was due to the hours of negotiation a few days earlier between me and the women. If I provided the dancers with a goat or, preferably, a twenty-litre bucket of beer, ‘it would please the ancestral spirits’, the women reassured me. I could thus attend the dance without myself suffering undue ridicule or abuse. Of much greater importance, from their perspective anyway, with such a gift my presence would not undermine the ritual’s principal objective, to bring the rain to this parched area of the country.

At a distance I sat, hastily jotting notes about the womens rain dance and being mildly enchanted by the realization that I was witnessing my first ever ‘ritual of rebellion’. I was familiar with Gluckman’s (1963) landmark lecture on the topic, which suggested that all such women’s rites, in spite of their manifest differences, could be explained as reaction to the prevailing patriarchal gender order. Such rites ultimately reinforced rather than undermined the gender status quo.

This seemed as good an explanation as any for the rites that unfolded before me. After all, men and women from the area had repeatedly told me that ‘men are the important ones’ (Agohá m akülü). What is more, local men’s and women’s everyday control, or lack of control, over vital political and economic resources seemed to bear this notion out. Thus the idea that these rites were women’s ritual response to men’s everyday dominance seemed perfectly sensible. Indeed, as I later discovered, a number of contemporary scholars continue to find ‘rituals of rebellion’ a useful way to think about such rites, and Gluckman’s original formulation a useful way of explaining them (e.g. Creider Sc Creider 1997; Spencer 1988; Weil 1976).

Yet the longer I remained in the field, the more difficult it became to see this particular women’s ‘ritual of rebellion’ as such. Nor, after some time, was it possible to see everyday life in Ihanzu as being wholly about ‘patriarchy’. It became increasingly clear, for example, that in certain everyday contexts women exercised more power than men and gained prestige in the process. In many ritual contexts as well, and even when discussing these women’s rain rites, informants insisted that male and female were and must be equal. As my understanding of subtleties of gender in Ihanzu grew, the seeming elegance of my previous understanding of this so-called ritual of rebellion was gradually and irreversibly undone.

The aim of this article is to explain these Ihanzu women’s rites without vitiating local understandings of gender ideals and practices. In so doing I wish to call into question the way we think about ‘rituals of rebellion’ as well as ‘patriarchy’, suggesting that both categories can be highly problematic.

There are several reasons for this. First, to talk about patriarchy or male-dominance without further qualification ignores the distinction between representations of gender and gender behaviours. This oversight, in turn, threatens to present an unnuanced, unitary picture that glosses over the subtleties of people’s daily experiences with, and ideas about, gender. Second, we cannot reasonably assume that people’s notions of gender are unitary and seamless, as ‘patriarchy’ and ‘rituals of rebellion’ demand. Ideas and ideals about gender in everyday and ritual realms alike are multiple and multifaceted. This makes untenable the still-common assertion that ‘rituals of rebellion’ are simply a reversal of the everyday gender system, discussed in singular and monolithic terms, since no such ‘system’ exists. Once we admit to the possibility of multiple and conflicting ideas and practices of gender in people’s daily lives, it is no longer clear what a ritual reversal of these ideas and practices might mean.

I offer an alternative reading of a so-called ritual of rebellion, one that pays close attention to the complexities of local ideas of gender and transformative processes. The first such Ihanzu idea is that women are naturally wetter than men, thus making them, so we shall see, better suited than men to attract rain. The second idea holds that rain rites are only successful when the cultural categories ‘male’ and ‘female’ conjoin as equals. I shall argue that the palpable contradiction between these two propositions is overcome when women embody, through ritual action, both genders simultaneously. This they do by playing up their wet, fertile femininity while, at the same time, coopting what are thought to be more typical male characteristics of aggression, abusive behaviour, and obscenity. By combining and collapsing masculinity and femininity within themselves, women dancers create the ultimate gendered combination to bring the ultimate communal good: rain. It is thus Ihanzu men’s and women’s ideas about gendered transformative processes, not their objections or reactions to alleged patriarchal structures, that gives both shape and meaning to such rain rites.

Above all else, while the ethnography to follow is specific to the Ihanzu case, the theoretical implications are not and suggest that we reconsider some of the ways that other African societies have often been uncritically (mis)represented as male-dominant or patriarchal, and women’s rites of rebellion as being solely about women’s power.[1]

Rethinking rituals of rebellion and patriarchy

In his key paper on the topic, Gluckman (1963) used two main examples to explain his notion of rituals of rebellion, both from Southern Africa: the Zulu Heavenly Princes Cult (.Nomkubulwana) and the Swazi ncwala ceremony. My concern here is with the former.[2]

As is well known, Gluckman argued that Zulu women, by acting out certain rites surrounding the Heavenly Princes Cult, were able to reverse temporarily the gender status quo thus allowing them to become men’s equals, if not superiors. During such rites ‘a dominant role was ascribed to the women, and a subordinate role to the men’ (Gluckman 1963: 114). Furthermore,‘[t]his temporary dominant role of the women — a dominant role that was publicly instituted, indeed approved, and not exercised tactfully in the background — contrasted strongly with the mores of these patriarchal peoples’ (1963: 114). For at all other times, women were characterized by their ‘subordination and modesty’ and ‘were in every respect formally under the tutelage of men’ (1963: 115). Such rites, Gluckman argued, functioned to release underlying psychosocial tensions over men’s and women’s relative powers in Zulu society. In no way did these rites challenge, let alone undermine, the social structure. They actually reinforced it.

Though rarely billed as such, Gluckman s essay is one of the first sustained enquiries into the nature of gender relations in an African society. His claim is that Zulu society is patriarchal or male-dominant. This formulation then provides the necessary conditions for his second claim, that women’s so-called rites of rebellion are about redressing the gender balance of power by inverting it. In other words, if everyday life is wholly about gendered hierarchies and male-dominance, then ritual life is largely about gender symmetry or female-dominance.[3]

This formulation is problematic for several reasons.[4] For one thing, it has become increasingly clear from recent anthropological studies that peoples ideas about gender and actual gender behaviours are often at odds (Moore 1994; Ortner 1996; Ortner & Whitehead 1981: 10ff.).That is to say, it is no longer reasonable to assume that notions of gender directly reflect gender practices, if such assumptions were ever justified: gender practices and people’s images of those practices often chart courses quite independent of one other (Keeler 1990; Peletz 1994; Rogers 1990; Stolen 1996; Tsing 1990).The unfortunate consequence of conflating gender ideals and practices is that there is no conceptual or practical space for discrepancies between the two. With Gluckman’s all-encompassing notion of patriarchy, gender behaviours must reflect directly ideas about gender, and vice versa. Social and symbolic realms become one.

A second and related difficulty results from Gluckman’s inattention to the potentially conflictual and contradictory nature of gender representations themselves (Archetti 1996; Bloch 1987; Caplan 1989; Hatley 1990; Melhuus 1996). As Meigs (1990: 15) notes in the context of New Guinea,‘Probably there is no such thing as a single gender ideology in any society. On a topic like the relationship of males and females, each society undoubtedly has many ways of thinking — complex, subtle, and even contradictory ideological options.’

To be sure, Gluckman’s portrayal of everyday and ritual gender images among the Zulu does allow for two (and only two) contradictory notions of gender. The first links male dominance with everyday practices. The second links female dominance with the ritual realm. Even so, he does not go far enough, for if there are contradictory notions of gender between mundane and ritual domains, as there frequently are, then such contradictions may equally be found within these domains. But for Gluckman, in daily life no space remains in which Zulu men and women might consider and contest the prevailing, hegemonic gender model (Ortner 1996). By the same token, gender representations in ritual life, as portrayed by Gluckman, must be conceptually seamless, even though we have long been aware that competing and conflicting notions of gender are regularly interwoven through the fabric of rituals (Beidelman 1997; Devisch 1993; Kratz 1994; Richards 1956; Turner 1967).

In making these criticisms, my concern is less with Gluckmans original formulation than with the apparent enthusiasm with which some contemporary scholars uncritically muster Gluckman’s arguments to explain their own material.

In his exemplary study of the Kenyan Matapato Maasai, for instance, Paul Spencer gives an entire chapter over to women’s fertility rituals of rebellion, noting, like Gluckman, that these rites must ‘be viewed against the backcloth of this [male-dominant] regime to which they are subjected by older men’ (Spencer 1988: 200). During women’s collective fertility gatherings, in which women persecute certain men guilty of abusing their ‘daughters’ and wives, Spencer explains, ‘the normal social order is upturned and women take over a domain that lies outside the normal bounds of village existence’ (1988: 202). Although his argument goes beyond and adds subtlety to Gluckman’s, by incorporating Rigby’s (1968) more culturalist approach, Spencer follows Gluckman in suggesting that such gatherings have a brief reality of their own, ‘which readjusts the balance of power between the sexes’ (1988: 204) in a society where ‘[w]omen are dominated by men’ (1988: 7; cf. Llewelyn-Davies 1981).

Still others seem to find the argument such an obvious anthropological platitude that to evoke Gluckman by name is apparently no longer necessary. In a recent article on girls’ initiation rituals among the Nandi of Kenya, Creider and Creider (1997) call into question Langley’s (1979) earlier suggestion that these rites are primarily about gender inversion through transvestism. To the contrary, evoking Mauss, they argue that it is the act of giving clothing from boys to girl initiates that is of central importance (Creider Sc Creider 1997: 55–6).

The apparel is given by the boyfriend to the girls as part of themselves, as something precious to them, to show the girls that they care for them, and to encourage the girls to bring honour to them (the warriors) by their brave behaviour during initiation (1997: 56).

They conclude by noting that:

The real reason for Nandi girls’ initiation, we suggest, is that it is women’s way of achieving equality with men. By undergoing an experience requiring an amount of bravery equivalent to that experienced by boys during circumcision, Nandi women are saying to the men, ‘Look! We are just as good, i.e., brave, as you are!’ (Creider & Creider 1997: 58, emphasis added).

Female rites of reversal or rebellion are once again said to be about turning the mundane gendered world back to front (see also Weil 1976: 191—3). Yet such explanations imply an implausible (if not impossible) degree of coherence in people’s thinking about and experiences with gender. By problematizing the complex and contradictory nature of Ihanzu ideas about gender and gender practices, my aim is to provide a more nuanced, locally informed understanding of Ihanzu women’s rain rites. This article thus aims to contribute to a growing body of literature that seeks to explore, rather than ignore, the complex dynamics of gendered symbolic and social processes in African ‘rites of rebellion’ (Blystad 1999; Kaare 1999; Power Sc Watts 1997; 1999; Snyder 1997; 1999).

The Ihanzu of Tanzania

The Ihanzu[5] are a Bantu-speaking group numbering around 30,000 who live in the northernmost part of Iramba District, north-central Tanzania. They are mainly farmers, the most important crops being sorghum, millet, and maize. As with many African agriculturalists, it would be extremely difficult to overstate the importance of rain for these people, both in practical and symbolic terms.

Rain is scarce in this semi-arid region. When it does fall, it is often erratic and unevenly distributed. Too much or too little at the wrong time can and often does spell disaster for these agricultural people. With few working water pumps and no year-round rivers in the area, they are entirely dependent on rain for their continued survival. Famine in Ihanzu is therefore commonplace..

One means of coping with famine is, and has been since around the turn of the century, to work as migrant labourers in more fertile parts of the country. By working in the northern city of Arusha or on the tea and coffee estates that surround it, migrant labourers often manage to eke out a meagre existence for themselves and provide for their families and kin in the villages (Adam 1963d). During droughts the Ihanzu also take more active measures at home to alter positively the weather: in a word, rainmaking (Adam 1963b; Sanders 1997; 1998).

All Ihanzu rain rites are orchestrated by members of the royal matrilineage (Anyampanda wa Kirumt), and more specifically by the two royal ‘chiefs’ (atemi), one male, the other female. Although these two chiefs have not been officially recognized since the government abolished the chiefship in 1962, they hold a profound religious significance for nearly all in Ihanzu today.[6]

Everyday gender practices

During the Ihanzu wet season, between November and June, a man and his wife labour jointly on their plot(s), assisted by any children living at home. Men usually put in slightly longer hours farming than women, though women spend a good portion of their days collecting leafy greens, and fetching firewood and water to keep their households running. In the dry season, men hunt, fish, and build houses; women spend much of their time collecting firewood and relishes for the next wet season. All year round women brew and sell beer. In the dry season in particular, beer-drinking is the basis of village social life.

The Ihanzu have matrilineal descent groups. This does not mean that we can accurately sum up Ihanzu society as ‘matrilineal’ or ‘matriarchal’ and suppose it is more egalitarian, or that it holds women in higher esteem than men (pace Watson-Franke 1992). Systems of descent, after all, do not necessarily mirror political practices or ideological positions about men and women (Bolyanatz 1995).

For one, Ihanzu women as a group participate minimally in everyday politics. When important village matters arise — cattle theft, rain-witchcraft, warfare, and so forth — village meetings are called. Men are expected to participate actively on such occasions. Women rarely do. Many important political decisions are thus taken by men while women play little or no direct part in them.

Economically speaking too, women might be considered subordinate to men, since men own and control virtually all valuables in Ihanzu. Nearly all livestock, the single most important valuable, are owned and controlled by men. Even before the Iramba District Council passed a patrilineal inheritance law in 1962 which ensured that property passed directly from father to son, all valuable property including livestock, houses, and grain passed through the matrilineage from a man to his sister’s eldest son. Most valuable property was and remains today under men’s control.

Based on these specific power arrangements and, more to the point, on mens and womens differential access to vital political and economic resources, it is tempting to conclude that Ihanzu is in fact male-dominant or patriarchal. When asked to sum up, in a word, male—female relations in Ihanzu, this is the answer locals themselves often provide. However, to gloss Ihanzu society (or any other) in such a manner conceals more than it reveals. The reason becomes clearer if we delve further into gender behaviours in other contexts.

There is one significant cultural arena in which Ihanzu women exercise more power than men, control over grain. Since grain is the essential ingredient for all meals and beer, control over it amounts to control over a large sphere of everyday social life. With her grain — and it is considered hers once safely stowed in the grain-store — a wife must budget from one harvest to the next. She decides how much grain she can afford to give to needy neighbours, kin, and others, and still have enough to provide for herself, her husband, and children. She must take decisions about when and how many times to brew and sell beer. Womens control over the circulation of grain is not imaginary: I know of several cases where a husband was denied access by his wife to the grain-store. Even more remarkable is that men invariably accept such refusals since, as men and women often point out, ‘only women really know about such matters’. Womens control of grain thus undermines any simplistic claims that Ihanzu women are, in all instances, second-rate persons ensnared in a patriarchal system. The waters become further muddied when considering Ihanzu men’s and women’s ideas about gender.

Everyday gender representations

Some women build considerable reputations by consistently brewing good beer. People are well aware which women are better brewers, and will often walk for miles to attend a beer party given by any such woman. It is on this basis that certain women gain prestige in women’s and men’s eyes, a fact that further erodes the possibility of an all-encompassing, patriarchal gender system.

In other contexts, too, Ihanzu ideas about the relative status of men and women are not entirely consistent. When questioned, as already mentioned, men and women alike often begin by pointing out that men are naturally superior to women. And yet, discussion on this point is more ongoing than once-and-for-all decided. As one middle-aged man from Kirumi pondered:

Men are the important ones.... If you look really hard though it might be the other way around — women are more powerful than we and we depend on them. For example, a man impregnates a woman.... Then he gets lost.The woman spends nine months preparing for the child’s birth and the man knows nothing. It’s like he was never important at all.

In still other contexts people portray male and female realms as complementary, by which I mean interdependent and equal. This notion of gender complementarity comes out most clearly when people discuss transformative processes such as making rain and children.

As I have discussed elsewhere (Sanders 1997; 1998; 1999), Ihanzu ideas about transformative processes are informed by a single underlying gendered model of reproduction. This model insists that masculine and feminine elements must unite as complementary and equal partners. Thus, the explicit sexual symbolism in Ihanzu healing ceremonies, procreation beliefs, and rainmaking rites — including rites to be described below — is elaborated in such a way as to emphasize, above all, gender complementarity. In this ideational realm of gender, men and women categorically refuse to rank the genders one above the other. Gender symmetry, not gender hierarchy, structures all transformative processes in Ihanzu.

In sum, we cannot in any straightforward or accurate way sum up everyday gender relations in Ihanzu. There is no single gender system, though there is a dominant discourse which stresses male-dominance. Given the contradictory and inconclusive nature of everyday Ihanzu gender relations and notions, the womens rain dance, to which I shall now turn, cannot be adequately explained as a straightforward reversal of everyday gender practices and understandings.

Ihanzu rainmaking and the women’s rain dance

Until the mid-1960s it was the joint responsibility of the two royal leaders and their assistants to usher in each new farming season by conducting public rites in the chiefly village of Kirumi. These ‘cutting the sod’ ceremonies centred on the royal rainshrine (mpilimo), and were intended to ensure a prosperous and fertile year (Adam 1963b). Today the male chief, named Omari, and his male rainmaking assistants (ataata) carry out such rites, with the aid of certain rain medicines and rainstones.

It is only when these annual rites fail, as evidenced by no rain for extended periods, that the two ritual leaders discuss holding a womens rain dance (isimpülya). The ceremony is always performed in Kirumi, normally either in January or February, though only in those years when the rains have failed completely. It is a method of last resort. The rituals aim, participants and others claim, is straightforwardly to bring rain to Ihanzu.

Once begun, the rain dance usually lasts two days. However, it may carry on for several, depending on whether the rain has begun to fall. As soon as the rain begins or seems likely to, the women may return to their homes confident that the royal ancestors have heard their prayers.

The female ritual leader, the male chiefs sister, is in charge of organizing and orchestrating these rain rites. Her name is Ng’welu. Throughout the rain dance participants are based at her home in Kirumi where they eat, sleep, and gossip.

Any woman who has borne at least one child may participate in the dance. Extraordinarily fertile women, those who have given birth to twins, have special roles to play that I will discuss presently. Infertile women, in contrast, are normally precluded from taking part. Finally, under no circumstances may menstruating women participate in these dances: this would threaten the wellbeing of all participants, and would likely destroy the dance by causing lightning strikes.[7]

Certain men may also take part in these ceremonies, namely, those who have sired twins. No men participated in the rites during the years I was there, but I met two who claimed to have done so on previous occasions. Several people also mentioned that the male chief, if he wished, could participate.

The rain dance to be described took place in January 1994,[8] nearly a month after Chief Omari and his rainmaking assistants had prepared their rain medicines at the rainshrine but to no avail. Rain, a topic of discussion at any time of the year, had become a constant preoccupation. When would it arrive? How and where would it fall? How much would there be?

As the situation worsened, people from several villages were publicly accused of bewitching the rain. Rain-witch trials began in Kirumi and several accused witches were fined. Some were expelled from Ihanzu. Other villagers began to leave the area in search of wage labour.

It was against this backdrop that, in the beginning of January, the female ritual leader, Ng’welu, suggested to her brother that the women began their rain dance. He agreed.

Day One: Anti-Witchcraft Medicines

Women began arriving in Kirumi, some from as far as fifteen miles away, in preparation for the dance. All slept at the female chief’s homestead. Men avoided the homestead.

The following morning, ten elderly women dressed in black danced and sang their way to Chief Omari’s homestead where Ng’welu was given some anti-witchcraft medicine by her brother. The women then sang and danced out of his homestead in a spirited way, down the path to a clearing on the outskirts of Kirumi. The few men who happened to be wandering about quickly went indoors. Some younger women soon joined the elderly women, bringing their numbers to about thirty. The aim of this first trip into the bush is to prepare the area, their main dance ground for the duration of these rites, against possible witchcraft attacks that might damage participants and threaten the efficacy of the dance.

I was told that the mothers of twins, using the tips of buffalo-horn drums (;mbilü), began digging small holes around the perimeter of the dance ground. They dropped small amounts of the medicine into these holes and refilled them with earth. Younger women, after disrobing, seated themselves naked atop each hole and, while singing, bounced over the holes to pack down the earth. Other women removed their clothes and danced naked. When the grounds had been ringed with medicines, the women clothed themselves and danced and sang in their characteristic single file back to Ng’welu s house. Once there, individual dancers were dabbed with medicine to protect them against witchcraft.

Day One: Anointing Drums and Snakes

Later that afternoon, the women set off for their witch-proof dance ground. Along the path they praised the ancestors in song and dance, as they did for the duration of the ceremony. They also sang many lewd songs. One man, seeing the women approaching, dropped his hoe in the path and dashed for cover in a nearby house.

At the dance ground the senior women continued beating their buffalo-horn drums and singing. Others, including all first-time participants, removed their black cloth wraps and danced naked. Still other women went to fetch water and cut firewood with which to prepare castor-seed oil. When the oil was ready, all departed for Mount Ng’waüngu.

Ng’waüngu, besides being the tallest boulder-strewn hill in Kirumi, is perhaps the most sacred place for the chiefly clan members: this is the location where their ancestral spirits meet. There are two royal clan caves on the hill and it is these sites where the royal clan section normally holds its ancestral offerings (mapolyo) when clan section members or their animals become ill, or during drought (Sanders In press).

I was later informed that the women entered both caves naked. At the first they addressed the royal spirits and anointed some ancient drums with castor oil.[9] At the second cave, they told me, they again addressed the spirits and this time anointed an enormous ancestral snake-spirit that supposedly lives there.[10] They then put on their clothes and returned to Ng’welu’s.

In the evening, after disrobing once again, the women began dancing, first at the house and later in the bush. No men were to be seen. The dancers made their way to the rain-shrine where they danced just outside, asking the chiefly ancestral spirits for rain once again. Until about four in the morning the women danced naked around Kirumi, into the bush and back again, up and down almost every path, singing their songs and beating their buffalo-horn drum.

Day One: Songs and Dances

Many of the women’s songs are blatantly sexual and, as Ihanzu themselves point out, obscene.[11] Some of the more conspicuously sexual lyrics feature super-powerful clitorises, immense penises, and copulating monkeys. One song dancers sang repeatedly, for example, consists of only one line that is repeated: ‘The penis is dried up and worthless; it stops the rain from shitting down’. Other songs contain more oblique references to male and female genitalia, fornication, and giving birth. Songs are habitually peppered with loud and spontaneous outbursts concerning penises, vaginas, and sexual intercourse.

If many of the songs are considered by locals to be obscene, so too are the dances. Several women told me that in the dark of night novices are made to dance lewdly around the rainshrine, each gripping one of its outer posts ‘like huge penises’. I was also told by a few women that some of their dances, especially those performed at night, were concerned with working female dancers up into an orgasmic frenzy. The significance of this will become apparent below.

Days Two and Three

On day two, women from the outlying areas continued to arrive at Kirumi. Men at this point stayed well away from Ng’welu’s homestead that was now full of women, perhaps sixty or seventy.

Later that day the women emerged from the compound, dropped their clothes and began to sing their obscene songs boisterously as they paraded around the area, making lewd gestures and yelling obscenities throughout the village. Men vanished. One less fortunate man however, was not fast enough and was caught: he quickly agreed to pay a goat so he could leave unharmed. (Incidents of violence against men, while pointedly recalled and perhaps greatly exaggerated, are few in number.)

Just before sundown, the women danced naked and sang their way into the bush for the last time, where they carried on for much of the night. After returning to Mount Ng’waüngu, where they again addressed the ancestral spirits and anointed the drums and snake, the women returned to Ng’welu s house, donned their clothes and slept.

The following morning the women arose and prepared for their journeys home. The weather had changed for the better: clouds were forming and there were some light showers. The chiefly spirits had apparently heard and taken to heart their requests.

Gendered bodies and personhood

All Ihanzu seem to agree that men and women are different. What those differences mean, as we have already seen, varies: either men or women may exercise greater power, depending on the context; either may be viewed as superior or inferior, again depending on the context. Since gender differences in Ihanzu are often portrayed as natural, part of men’s and womens respective bodily constitutions, it is instructive to begin with local ideas about differently gendered persons.

One gender distinction the Ihanzu commonly make is that women are wet (atotu) while men are dry (akalamüku) (cf. Beidelman 1964: 377; 1973: 138, 151; 1993: 39). In Ihanzu eyes, these gender differences are biological and therefore non-negotiable. These qualities are said to be observable in any men or women: the former are often said to be lean while the latter have more soft fat. People often instance womens vaginas as evidence of their watery composition. The interesting question, naturally, is how these differences make a difference (Moore 1994). Or, to put it another way, how people relate wet and dry bodies with notions of fertility, moral personhood, and gendered behaviours.

Having a watery body in Ihanzu is a good thing. Water equals fertility, a proposition that is hardly surprising in this arid region where, without water, life in all forms rapidly grinds to a halt. Water is said to be cooling, a life force and something without which women could not give birth. In the form of rain, water allows crops to grow. As semen it produces children.

Conversely, having dry bodies, as men allegedly do, is not necessarily a good thing at all. This point comes into sharp focus when we recall the ‘dry’ (kalamüku) penises, mentioned above in song, that impede the rain from falling. ‘Kalamüku implies a certain waterless (and often lifeless or useless) quality and frequently has negative connotations. It may be used to describe barren or parched land (ihi nkalamüku). It may also be used to describe a phenomenally filthy person (müntü nümükalamüku), one who never bathes, one who remains at a distance from water. When applied to the entire male universe it is always negative, or at least it implies a water-deficiency, a lack of a certain life-giving quality, somehow falling short of womens natural moist make-up. If women are cool and thus fertile for the water they contain, then men are comparatively hot and infertile for their lack of water.

Blood is another bodily substance that is sometimes evoked as part of this dichotomy of wet/cool/female vs. dry/hot/male. Blood is thought to pass through the patriline. Those with whom I spoke regularly suggested that men possessed more blood, or perhaps hotter blood, than women. In Ihanzu blood is inimical to water. Blood is commonly associated with the colour red, heat, disorder, lightning, brawling, and warfare. It is often inauspicious.

Locals sometimes opine that these supposed biological differences incline men to be more physically and verbally abusive than their ‘cool’ female counterparts. For this reason men are ‘natural’ aggressors, in sexual liaisons as in warfare. Women, on the contrary, are said to be naturally relaxed, passive and followers. People recognize that men and women do not always behave in accordance with their bodily dispositions. None the less, all I spoke with felt rather certain that many gender tendencies derive ultimately from natural differences between men and women. These particular ideas about gendered personhood, when combined with certain cosmological notions we shall now explore, help us to understand the transformative power of womens rain rites; how it is, from a local perspective, they are thought to bring the rain.

Ihanzu rituals of transformation

There are two cosmological principles that inform all Ihanzu rituals. The first, which has been widely reported across Africa, is ‘like attracts like’ (Herbert 1993: 85–6; Klima 1970: 47; Ten Raa 1969: 50–1; Wilson 1957: 10).

In Ihanzu, black (-dwalu) is ritually auspicious. It is the preferred colour for clothing worn during rain rites and for sacrificial animals. The reason for this, all seem to agree, is that rain clouds are also black. Black is cool and brings life-giving waters which cool the land. One black thing invites another.

The colour red (-kaüku), on the other hand, is ritually inauspicious. It threatens to attract other undesirable ‘hot’ (-pyu) and ‘fierce’ (-taki) things like lightning, also said to be red. Wearing red during an ancestral offering or menstruating during a female rain dance would, people say, cause lightning to destroy people and livestock, crops and homes. Red animals are thus wholly unacceptable as sacrificial ones. Heat summons more heat.

Other ritual items follow the like-attracts-like logic. For example, trees are sorted into two mutually exclusive categories that might be glossed as ‘cool’, ‘gentle’ trees (rríifi ríi rnipolo) and ‘hot’, ‘sharp’ trees (rríifi mitaki) (cf. Jellicoe 1978: 80; Rigby 1966: 9; Spencer 1988: 205; Ten Raa 1969: 41). The former are desirable for rainmaking and appeasing the ancestral spirits, the latter are not. The noteworthy point, once again, is that ‘cool’ trees attract cool rains while ‘hot’ trees do not, but are more likely to attract ‘hot’ lightning.

Thus, in ritual contexts, black clothes and animals bring black rain clouds; red clothes, menstrual blood, and other types of blood bring lightning and destructive weather. Gentle, cool trees bring like qualities in rains. In all ritual instances, whether for good or for bad, similar things attract each other.

A second major cosmological principle evident in Ihanzu rituals is what I have elsewhere called gender complementarity (Sanders 1997). In Ihanzu eyes the joining of gender categories as equals provides a local model for transformative processes, whether for the production of children, pots, iron, rain, or almost anything else. In this respect male and female together can make things happen. Things gendered include people, as with the male and female chiefs, but also gendered rains, rain-stones and other aspects of the Ihanzu cultural universe. To join the genders is to generate, to create and transform by activating the cosmic and divine powers of the Ihanzu social and natural worlds. Just as men and women conjoin and bear offspring, male and female chiefs reputedly engage in royal incest to give birth to the people of Ihanzu and bring the rain. It is on this basis that the two gendered rulers exercise power and ‘naturalize’ their rule (Sanders 1998). As this implies, the Ihanzu cultural categories ‘male’ and ‘female’ prove particularly salient for thinking about their world, and their complementary combination is a powerful means of transforming it. Cosmic power comes in differently gendered pairs.

Wet bodies, wet rains and embodying gender(s)

In Ihanzu eyes, I shall argue, women are physiologically better equipped for carrying out these rain rites because they are wetter than men. Their bodies make this so. Since like things attract each other, women’s wet bodies attract wet rains. Men’s dry bodies do not: they threaten only to attract more heat. One elderly man made these connections much more concisely than I have:

Women are wet like rain. Men are barren, dry. This is easy to explain. Women, their vaginas, never dry out; they are always wet. We [men] can be dry for years on end. Even if we have sex only a few times in one day, we become really dry. A woman, however, never dries out. Water comes out of them all the time that keeps them wet. For this reason you might say that women are like rain.

During the women’s rain dances they dance naked, even in the middle of the afternoon! This is because they are displaying themselves, showing that they have lots of water and fat. The ancestral spirits rejoice and the rains come. You don’t see those [male] rain-making assistants wandering about outside naked, do you? They don’t have anything wet [to show].[12]

On balance, women’s bodies are superior to men’s for such rites. This accounts for the fact that the women require no rain medicine for their dance, only anti-witchcraft medicine. Their wet bodies are quite enough to get the job done. For this reason women, not men, are the main actors in these rites. What is more, by allowing only those who have given birth to participate, and by disallowing menstruating and barren women, dancers present women’s ‘natural’ embodied fertility in a dramatically exaggerated fashion.[13] Women dancers are in a sense hyperfeminine.

But this is only half the equation. As with all Ihanzu rituals of transformation, these rites are premised on the notion of gender complementarity. In order to do what they supposedly do, male and female must combine as equals. This is where the manifest violence, obscenity, and sexuality in the women’s rain dance come to the fore.

If women’s bodies are superior vessels to house wetness and hence fertility, then they are also better equipped to house both genders at once (cf. Broch-Due 1993; Jacobson-Widding 1985: 10–11). Men, no matter what they do, are physiologically impoverished. They are dry Thus it is impossible for them to attract fertile wet rains. Yet women have a decided advantage: they have no difficulties coopting typical male traits through ritual action. By becoming aggressive, violent, and generally obscene — all traits commonly associated with men and their hot composition — women dancers can successfully embody some of the more stereotypical features of Ihanzu masculinity.[14] But as before, these coopted gender traits become a positive if not perverse exaggeration of the actual gender they supposedly represent. Dancers thus display, in a highly ritualized and exaggerated form, some of the more typical characteristics of both genders as they perceive them, and furthermore push them to their logical conclusions.Through ritual action women dancers therefore embody hyperfemininity and hypermasculinity.

This accounts for the heavy emphasis on sexuality during these Ihanzu rites. All Ihanzu rites of transformation are replete with sexual symbolism, playing on themes of gender complementarity. What I am suggesting here is that these rites do the same, but in a rather striking manner: by embodying both genders simultaneously, dancers bring male and female into a relationship of perfect and equal union within themselves. In times of dire need, as a method of last resort, what better way to bring the rain than for dancers to combine sexually, en masse, hyperfemininity with hypermasculinity within themselves? The ultimate combination of masculine and feminine principles brings the ultimate communal good: rain.

Conclusion

It is worth noting that Gluckman s thesis could be applied to explain the Ihanzu womens rain rites, almost. If we are prepared to take dominant discourses about male dominance at face value, to ignore the power and prestige some women hold, and to dismiss local ideas about gender complementarity and transformative processes, then it would take little imagination to argue that the ritual portrays daily gender relations back to front. Many scholars, wittingly or unwittingly following Gluckman, have argued this point.

But these Ihanzu womens rain rites take the shape they do, not because they are responding in ritual and inverted form to everyday gender relations, but rather due to people’s ideas about transformative processes that privilege gender symmetry over gender hierarchy. The ritual logic, in other words, is linked to but is ultimately independent of what men and women do and do not do, think and do not think, in their everyday worlds. Hence the only sensible conclusion to be drawn is that this particular Ihanzu ritual of rebellion, quite simply, is not a rite of rebellion at all.

While it may be tempting to identify certain womens rituals as being primarily about rebellion against patriarchal orders, I have tried to show that such arguments necessarily rely on an (over)simphfication of gender notions and practices, both in ritual and daily contexts. Rather than allowing us to explore and problematize local gender representations and behaviours, such arguments positively hinder our ability to do so. This they do by demanding that a single, monolithic everyday gender system be identified in any given society that, when ritually reversed, constitutes an equally monolithic mirror image of that system. In offering an alternative perspective on women’s rain rites, my aim has been to accommodate some of the complexities of people s ideas about gender and gender behaviours, rather than simply to pave over them.

Acknowledgements

For funding different portions of the research on which this article is based I thank the US National Institute of Health, the UK Economic and Social Research Council, and the London School of Economics and Political Science. Also I thank Albert Schrauwers, Katherine Snyder, and the JRAÍs anonymous reviewers for their comments. I alone am responsible for any shortcomings that remain. Most of all, I am indebted to the people of Ihanzu with whom I lived between August 1993-May 1995 and June-September 1999.

References

Quand les pluies tournent mal, les femmes virent á la folie: un nouveau regard sur les rituels de rébellion entre les sexes et la patriarchie

Résumé

A la lumiére des avances théoriques recentes sur les relations entre les sexes en anthropologie, cet article reconsidere l’argument selon lequel les ‘rituels de rebellion’ seraient une réponse ritualisée des femmes aux structures patriarcales dans la vie quotidienne. Originellement suggéré par Gluckman, cet argument a été evoqué récemment par Spencer, les Creider et d’autres. En prenant comme exemple un de ces rituels féminins chez les Ihanzu de Tanzania, je montre comment cette formulation particuliére réduit les notions complexes de rapports entre les sexes et des pratiques s’y rapportant á des systémes sans nuances, monolithiques et totalisants tant dans le domaine de la vie quotidienne que dans le domaine rituel. Ceci est effectué tout d’abord par une conflation des relations idéales entre hommes et femmes avec les comportements actuéis et ensuite par une méconnaissance des idées conflictives du public sur les rapports entre les sexes. En problématisant ces contradictions, je démontre que les rites des femmes Ihanzu ne relévent pas de la rébellion mais de la complémentarité entre hommes et femmes, mise en scene par des danseuses qui incarnent simultanément les deux sexes. Avant tout ce cas nous force á repenser, d’une maniere fondamentale, les notions de ‘rituels de rébellion’ et de ‘patriarchie’.

Department of Anthropology, The London School of Economics and Political Science, Houghton Street, London WC2A 2AE. D.TSanders@lse.ac.uk

[1] None of this is to deny that in many, and perhaps even most, African societies it is men, not women, who control vital economic, political, and symbolic resources. I only wish to indicate that the complexities of people’s ideas about gender and gender behaviours in ritual and everyday contexts have not always been fully appreciated. In this respect, more subtle explorations of African gender representations and practices are long overdue.

[2] There are probably two reasons Swazi ncwala rites have been widely scrutinized while the Zulu womens rites have not (e.g. Apter 1983; Beidelman 1966; Kuper 1972; Lincoln 1987; Makarius 1973). This has to do, first, with the richness of Kuper s original data on the ncwala on which these analyses, Gluckman’s included, are based (Kuper 1947; also Kuper 1973). Second, Gluckman’s own concern was to explain why such rites did not fundamentally challenge the status quo and thus lead to political revolutions. For this, exploring ncwala ceremonies rather than womens rites is perhaps more relevant, or at least this is what the weight of his analysis seems to imply.

[3] Harrison (1985) and Hill (1984) provide cases in Melanesia and Venezuela respectively which offer precisely the opposite of this proposition, but which none the less maintain the secular—sacred dichotomy. All of these works, but especially Gluckman’s, resonate strongly with Turner’s (1969) work on structure and anti-structure.

[4] Gluckman’s argument has been repeatedly challenged and criticized on theoretical, logical, and ethnographic grounds (Krige 1968: 184; Norbeck 1963; Rigby 1968).

[5] ‘Ihanzu’ is what locals call the land on which they live, while the term used to describe themselves, Anyihanzu, means simply the ‘the people of Ihanzu’.

[6] In fact the female chief was never recognized by the government.

[7] During menstruation women farm, cook, and mind the children and house as usual. They are never confined during menstruation, and in mundane contexts menstrual blood itself is unthreatening. Allegedly, however, it can be used by witches to make the woman from which it came infertile.

[8] I attended both the 1994 and 1995 dances. The rites and sequence of events were nearly identical both years.

[9] There are numerous caves in the area with massive, ancient drums in them (see Hunter 1953). Some insist that these drums were made and played by giant savages (washenzi warefu) who inhabited the area prior to the Ihanzu themselves. Living off a patently odd diet of wild animals and trees, these savages reputedly hid their drums in the caves during war. Whatever the case, it is the caves that are of primary importance for these offerings, not the drums.

[10] Everyone in Ihanzu has heard about the snake spirit that lives in this royal cave. This snake is invariably said to be a python, a rare snake in Ihanzu and the only auspicious one. I have never seen this snake spirit nor have I seen any other alleged to five in other clan caves around the country.

[11] The prominence of obscene songs in various fertility and other rites has been noted across Africa (e.g. Beidelman 1964: 382–4; 1973: 137–9; Evans-Pritchard 1929; Krige 1968).

[12] It is interesting to note that Schoeman was given an identical explanation for Swazi rain-making rites, though he failed to make any sense of it. In a footnote he writes: ‘On asking my informants why females play such an important part in the rain-ceremonies, one of them ... answered: [...] it is done for the sake of the earth, so that the rain clouds will gather, and get soft, i.e., rainy, because it is a female person, and she (a female) is wet (has a uterus). When I asked him to explain what he meant [...] he just shrugged his shoulders and walked off’ (Schoeman 1935: 172 n.).

[13] Menstrual blood is not spoken of as ‘wet’, nor is it associated with fertility as is water.

[14] In an essay rarely cited, Gluckman noted that the Zulu women’s rites had elements that were symbolically male and female. The former he explained as women’s reversal of everyday gender roles; the latter, as those tasks women ritual participants would carry out later in life (Gluckman 1956: 121). An alternative explanation for the symbolic presence of both genders in the Zulu rites might be that both are required to effect transformations, as in the Ihanzu case.