"Women are evil": personhood, gender, and disease in southern Mozambique^*

Luiz Henrique Passador^**

Translated by Henrique Mariotto
Translated from Cadernos Pagu, Campinas, n.35, p. 177-210, Dez. 2010.

ABSTRACT

Keywords: Mozambique, Gender, Kinship, Witchcraft, Illness.

RESUMO

Partindo de um caso de doença e morte atribuídas à feitiçaria, o artigo analisa as articulações entre pessoa, gênero e doença que operam no Sul de Moçambique, onde as mulheres são tradicionalmente temidas e acusadas de produzirem malefícios através de feitiços.

Palavras-chave:  Moçambique, Gênero, Parentesco, Feitiçaria, Doença.

The tile of this article reproduces a phrase that I heard in Homoíne, a district in Southern Mozambique. That statement seemed to reveal a gender configuration that imposes specific experiences on women in that social universe. In order to explain it and provide it with a context, I will have to describe a fact I observed in 2007.

José¹ passed away in October. He was a 27-year-old man from Homoíne and worked as an English teacher at a school in the district's main village. He was married to Dalva, a 23-year-old teacher from Gaza. He had a disease that left him emaciated in less than one month and unleashed a series of accusations of sorcery involving his family members, neighbors, and certain women. Connections were established with other deaths in his family group and in an extended network of friends among a group of influential young men in Homoíne.

The case involves a wide set of variables which will be summed up here in order to make it minimally understandable. José was the second child of the second marriage of his father, a polygamous curandeiro² who lived with his second wife - José's mother - and his third wife. His father had separated from his first wife, who at the time lived in the neighboring District of Morrumbene, where he was born and where he had his machambas.³ José had married Dalva in 2006 without a civil register and without paying off the lobolo⁴ - only his presentation to his wife's family had been formalized. He went to live with his wife in a house he had built far away from his father's residence. He was very skilled in making influential friends, including friends among the rare white foreigners who came to the village. He was ambitious and people commented he was not to be trusted when it came to other people's goods, not even his relatives' ones. He was associated with a group of influential young males in the village who were members of the Football Association of the district, presided by one of the district's traditional chieftains.

During the time José was ill, he avoided taking any food prepared or offered to him by members of his family, including his wife. His food intake was limited to coconut water and food he was able to swallow, but provided that they were brought and fed to him by his best friend - who he classed as a cousin. In the visits I paid him he asked me to bring him some industrialized yogurt sold in shops. He had been admitted at the Health Center of Homoíne and diagnosed with malaria⁵, but his family rapidly came to a diagnosis of a "traditional disease" caused by sorcery, took him from the hospital and brought him home. He was then submitted to treatment by a curandeiro recommended by his father. After some back-and-forth visits from the hospital to some curandeiros and vice-versa, José passed away at the house of the last curandeiro he visited. The latter only had time to come to a partial diagnosis of his disease by means of the tihlolo⁶ and found indications that the evil deed had come from somebody belonging to the deceased's family group.

His death caused consternation in the village and unleashed rumors on who had been responsible for the spell. His father, already suspected of having caused all the previous diseases and deaths within his family group, was accused even by his children. In the main village, people commented that "everybody knows it was his father" who had caused José's disease and death. The partial diagnosis given by the last consulted curandeiro only incited the rumor mongering. José's mother was also involved as a potential co-author, as she did not demonstrate sorrow during the funeral, nor had she grieved the death of her son in public - José's father had the same attitude.

While the case remained unsolved, the family had a summit with José's father in Morrumbene during the week that followed the funeral. He was aware that he would be seen as the main suspect. People suspected that he was killing members of his family in order to benefit from the use of spells, and José would have been only the last victim. Considered to be a "true" curandeiro because he had inherited the powers of ancestors who had also been curandeiros, the father would also have bought spirits from other healers in order to become more powerful, guarantee his promotion in the administrative and party structures, and sustain the high productivity of his machambas. In order to benefit from these powers, he would have to kill his relatives, which would be insignificant losses in face of the gains he was supposed to obtain. Moreover, José and his sister had challenged their father and his traditional paternal power, which was based on a system of agnatic succession and patrilocal residence. José's sister alone decided to leave her father's house before getting married, and moved to Morrumbene. José had also left the surroundings of his father's house and settled residence somewhere else. Therefore, both children had become autonomous in relation to their father.

José was a "modern" person, so to say. He earned his living without having any machambas, worked in the formal job market, had left the structures that would have submitted him to his father's sphere of influence, and was in the process of becoming a "great man" in the main village for his own extra-traditional merits, achieved individually within a network of privileged friends. His new house - with electric power and located in a good neighborhood - was the symbol of this situation of privilege and independence. People commented that the death of José and his son in the same year had been caused by his father's envy or anger in retaliation to his individual success and insubordination to his father. But even in terms of the "traditional"⁷ universe, José was a young man in the process of becoming a "great man". He had married a young, dedicated, hard-working, and fertile wife. He only needed children to start his own core of descendants, to be extended by the matrimonial alliances that his children would make in the future.

While José was ill and before the accusations against his father and mother became stronger and hegemonic, a series of speculations related the cause of his disease to other elements that were normally associated with the production of spells and evil deeds: neighbors and, especially, certain categories of women.

At first, people speculated about José's neighbor, from whom he had bought the land on which he built his house. This neighbor had only sold him the land due to the pressure of his children. When José was building his house, he found evidence of a spell buried in the lot and suspected it had been made by his neighbor, who was unsatisfied with the sale. Also, and still according to this version, José would also have made a fatal mistake: he organized a party to open his house and did not invite that neighbor. This was considered as a breaking up of the rules that establish good relations within a neighborhood. Whether caused by oblivion or a deliberate attitude taken by an individualistic person who was little attentive to the requirements of the "traditional" universe (which he frequently disdained), José had certainly created a tense plight with his neighbor and this provided the basis for the rumors that followed.

Other versions attributed his disease to a universe of "traditional diseases" associated with certain taboos. A young woman who was the daughter of a curandeira told me she suspected that José suffered from a disease called xivenze, which affects people who, after the death of a relative, have sexual intercourse or take possession of the deceased person's goods before the kutchinga is carried out.⁸ According to her, José might have disrespected the sexual abstinence period after the death of his son, or might have eaten meat of some domestic animal that belonged to someone that had died recently and were not previously purified.

One of José's friends raised some hypotheses related to sexual taboos and certain women. Firstly, José could have had sexual intercourse with some widow who had not been submitted to the kutchinga, which repeated the suspicion around the xivenze. In the second hypothesis, José might have had sexual intercourse with a woman who had an abortion and had not been submitted to purification rites. Finally, Dalva was suspected to be an "owner's wife", whose "owner" was a spirit. Widows and "wives of spirits" are defined by a similar principle: they have an "owner", that is, they have a matrimonial alliance with a man (even if he is the spirit of someone deceased), and access to them requires the permission of their "owners" through traditional ceremonies - in the case of widows, the requirements involve the kutchinga and the respect to a one-year period before they can have sexual intercourse with another man. José's friend saw greater plausibility in the third hypothesis, since two of his cousins got sick like José after marrying "spirit's wives", lost their young children and almost died themselves.

On the day after the funeral I paid a visit of condolence to the family, as I already knew that there was a consensus in the family and among the inhabitants of the village that José's father was guilty of his death. When I arrived I met Dalva with her mother and some women from José's family (her sister-in-law and José's father's wives). José's father was sitting at a table to welcome the visitors and I sat down next to him. I wanted to know his version for his son's death and so I took the risk of asking him. Then he answered that José had fallen as victim of "traditional things". He argued that he had done everything he could to save his son, but it had not been enough. I took the risk again and asked him what the most specific cause for the death could have been in "traditional" terms, or who could have done that. He told me that José "got involved with a whore" ⁹ when Dalva was traveling. When he went back home, he became ill and died. Then, in a tone of complicity, he looked at me and said in a low pitch: "we, men, are in a bad fix with things of tradition, because women are evil". And he repeated it once more, emphatically: "Women are evil!"

These versions on the causes of the evil that victimized José became more plausible as all of them made use of a consensus socially available in Southern Mozambique. According to it, there is a universe of evil deeds and spells that cause damages, diseases, and deaths among their peers which are attributed to women. I heard in Homoíne that the great majority of sorcerers and curandeiros are women, a fact that associates them with spiritual powers and damages that afflict both men and women. Thus it is a consensus that women have a certain type of power and carry out a certain type of violence related to spiritual forces or impurity. This submits them to a regime of constant suspicion and violence, conceived of as a form of counter-violence. It is this aspect of a reality that articulates personhood, gender, and diseases, that I analyze on the following pages.

Tradition

The universe of "traditional diseases" in Homoíne is closely related to the construction of the personhood and the structuring of gender relations. However, understanding this universe requires the understanding of how people think of and deal with the supposed dichotomy between "tradition" and "modernity" in that reality.

When a disease is classified as "traditional", it is presumed to be inserted in a specific field of causes, consequences, and proper procedures to treat it. The definition of a disease is always preceded by a speculative process that includes a large number of variables involving particularly its symptoms and development, and it is concluded with the diagnosis obtained from a curandeiro who uses the tihlolo to determine its causes.

A "traditional disease" is usually one whose development is considered to be atypical because of its protraction or recurrence, for the rapidness with which symptoms lead to death, or for the set of multiple symptoms and signals in the body which are also articulated with problems that are not related to one's body (low productivity of machambas, disease and death among relatives, for example). However, a "traditional" cause may be imputed to any disease, even to those with a typical development and known to be treatable with biomedicine. The "traditional" diagnosis includes a revision of the relationships that the ill person established with his/her social and physical environment in order to find in these relationships the potential causes of a disease. The causes of "traditional diseases" tend to be referred to one who, as demonstrated by Alf Helgesson (1971) in his study on the Vatswa, the majority group in the District of Homoíne.  Hence it follows that every "traditional disease" tends to be considered as a result of personal interactions and somebody else's action on one's body. That is why I asked José's father about who could have produced the cause of his illness.

Even in a social environment of available "modernity", as it was the case in the village of Homoíne, where there is easy access to state apparatus and urban resources, "tradition" is always a potential cause of diseases and this diagnosis is not necessarily abandoned when a disease is treated and cured with biomedicine. A disease may be treated simultaneously through biomedicine and traditional healing, and this joint procedure is not perceived by the subjects as a contradiction. But this does not mean that people in Homoíne do not acknowledge any type of specificity in the diseases and treatments defined as "traditional" or "non-traditional" even in the case of joint treatments, as they acknowledge the existence of distinct fields that they identify as "tradition" and "modernity".

The inhabitants of the main village refer constantly to traditional healing, sorcery, "traditional power", family, ancestors, rites, and "traditional ceremonies" as data which refer to a universe they call "African tradition" or "Africanism", discursively delimited by expressions such as "in those times", "in those areas", and "those people". That is, from the perspective of discourse, the inhabitants of Homoíne operate emically in a dichotomic register that refers "tradition" to another time (always remote), to another space (always distant), and to another social universe (always different), establishing by opposition the field of "modernity" as that one they more clearly experience in the main village - the urban character of which is perceived as closer to "modern" life. Discursive fields are thus created and constitute different subjects, temporalities, and places that may even be contradictory among themselves, establishing very well delimited alterities. This allows them to identify, classify, hierarchize, systematize and deal with a multiple set of objective and discursive data that interpenetrate randomly in the practice of daily life. Thus the village provides a space of "modernity" that is conceivable, perceptible, comfortable, and safe, as opposed to "tradition", which would be restricted to the campo (countryside) or mato (bush) - expressions used to refer to the universes of "those areas", in which "those people" still live as they did "in those times". Thereby a binary taxonomy of perceived transformations and permanencies is created, and this is how one may observe the functionality and effectiveness of the discursive strategies applied in the construction of "tradition" and "modernity" in the main village is observed. This taxonomy is constantly updated in discourses and performances¹⁰ that appropriate certain signs recognized as "modern" (clothing, vocabulary, corporality, consumption, etc.) in order to establish frontiers and ruptures with the universe of the campo or mato.

The field perceived as "modernity" is also closely related to the universe of the market and of the State and its apparatus (hospitals, courts, schools, public departments, stores, bancas, and barracas¹¹), where performances, discourses, and negotiations take place by using elements that the consensus recognizes as pertaining to a "modern" and urban universe - it is in the main village that people recognize the existence of these elements in a more constant and settled form in Homoíne. "Tradition," in turn, is where the State and its institutions do not exist, are precariously established, or subordinated to "traditional" institutions - this is the place of campo and mato. Finally, at the level of a discursive taxonomy, what is intended to be "modern" in Homoíne refers to an urban universe, and what refers to a rural universe is what is intended to be delimited as "traditional."

However, in the case of critical events such as disease and death, what is observed is a pragmatic deconstruction of this discursiveness and performativity, when the lives of the inhabitants are invaded by "tradition" - also due to the fact that they perceive themselves as vulnerable to it because the majority of them either come from the campo or are in permanent relation with it through their relatives and machambas. Such events implode social taxonomies that delimit the "modern" and "traditional" universes and unleash a set of actions that disrupts any clear frontier intended to exist between them. These events amalgamate the institutions of "modernity" and "tradition" and establish continuities - from the hospital to the curandeiro and vice-versa, which does not lead to ruptures, but to a process of pragmatic continuity. What is then revealed is the unreal and fragile character of these fields and frontiers, composed of excluding and contradictory alterities, well-defined only at the level of discourse. Therefore, to consider the variables that constitute social dramas in Homoíne, one should at least question the operationality of any analysis that affirms a clear and evident separation between "tradition" and "modernity" in processes related to disease and death, as well as such separation on other levels of the social universe and interpersonal relations.

Personhood and risk

Speaking of ontologies also means to speak of the notions of personhood. The anthropology literature on the sociocultural life in Southern Mozambique demonstrates the central character of kinship and family structures in the constitution of subjects and sociocultural systems. From Henri A. Junod to Alcinda Honwana, including David Webster and Brigitte Bagnol - to mention some of the authors with whom a dialogue is to be established here -, we find a series of studies that indicate the need to understand the forms of alliance and descent in order to explain the logic of sociocultural systems, as well as the individualities that arise from them, for they operate and build the realities observed in Homoíne.

These authors indicate, in different ways and in other terms, that in order to understand the subject in Southern Mozambique it is necessary to consider the fact that a particular ontology and personhood operate in that area, and that these remain as the basis of the relations between these subjects and the world, even after important and critical historical transformations, and in spite of the constant incorporation of new sociocultural elements. What remains is neither a set of sedimented "uses and customs", nor rigid structures¹³, but a set of logical, cosmological, and ontological principles that are present in the forms of alliance and descent that found subjects and their reality, structure the relationships between the living and the dead, and between the latter and a certain notion of nature. These principles are constantly reaffirmed and, to a great extent, allow us to understand how relations concerning gender and diseases are pragmatically established in a context of profound and constant transformations.

In Homoíne I came across some facts that indicated a certain pattern of construction of the personhood that is founded in the relationships among peers. Roughly speaking, the person is partly defined by descent relationships, as a person must be in a constant relation with ancestors, madodas¹⁴, parents, siblings, children and grandchildren, who are defined by agnatic as well as matrilateral relationships. A person is incomplete if he/she is disconnected from a group of ascendants and has not produced any descendants. Concerning alliances and matrilateral kin, a person is incomplete if he/she has not married, has no sograria¹⁵ (term used to refer to the family group of one's husbands or wives) and does not have any sons- or daughters-in-law, and the affines established through them. At a broader level, alliances are extended to non-familial forms, such as neighbors and xarás¹⁶ (when the latter are not men and women from the family from which the name was inherited), in addition to churches, circles of more or less formal friends, and work relations. This set of relationships defines a person in Homoíne, and the construction of this network is one's main investment to define one's social existence, as described by David Webster (1976) among the Chopi. My restriction to Webster arguments is that the person constituted in such a situation is not a Weberian type of individual, but a subject closer to the dividual pattern found by Marilyn Strathern (1988) in Melanesia.

This set of relations describes a life path as well as the accumulation of social capital by the person that, after his/her death, will define his/her status in the world of spirits, as well as the type of relationship between him/her and the living. It is this accumulated social capital that defines the "great man" ¹⁷ in life and will be transferred to his existence as an ancestor. This is not different in the case of women, although they are respected mainly for the descendants they have helped to generate and for the success of their machambas, being inseparable from a "great man" whose wife, mother, or relative they were - usually referred to their husbands and his agnatic group, or to the agnatic group they come from. Thus we have in Homoíne a process of construction of the personhood that is very similar to that described by the Comaroffs (2001) among the Tswana and defined by them as a being-in-process. The person is a project that is accomplished when he/she becomes an ancestor who is remembered and respected by his/her descendants, whose function is to protect his/her family group and to intervene when the prescribed order is broken by one of its members or by the action of extra familial elements. This allows to the ancestor having his name preserved by his family and subsequently attributed to its descendants, establishing a form of return and permanence in family and social life - this mechanism of nomination is another modality of the xará observed in Homoíne, and involves both male and female names. Likewise, the ancestor must also be remembered by his neighbors and by his broader community as someone who produced respectable alliances and broad relations, extending his influence and power beyond his core of descendants. The name and the lands which are expanded as the "great man" expands himself through his descent and alliances become inseparable from his person. Those who inherit them also inherit his socially acknowledged "greatness". Names are keys that open certain doors, so to speak, and they are defined by descent and alliances. Thus, by acknowledging some transformations produced by historical incorporations, what we observe in the construction of the person is a scheme that is still very similar to that one observed and described by Henri Junod (2003) concerning the life of an individual at the end of the 19^th century.

This seems to be the reason for the persistence of a pragmatism said to be "traditional" in a social environment that incorporates "modernity", as described previously: in the context of a certain "modernity" in Homoíne, the person still construct himself/herself according to the logic of a persistent and encompassing ontology, manipulating different signs to produce socialities in which that person can recognize himself/herself and be recognized by his/her peers. Although components classed as "modern" may be observed acting on reality, such reality is constructed by the "traditional" person, who recomposes his/her ontology from an apparently promiscuous mixture of "traditional" and "modern" signs. Hence, the social reality is permanently reconstructed by the logic of construction of the personhood by manipulating "traditional" and "modern" elements, which does not require the persistence of specific "uses and customs."

The person is not defined by the materiality and objectivity of "uses and customs" in themselves, but by how he/she orders, reorders, invents, and reinvents them, recomposing through this performativity of multiple elements an ontological and pragmatic unit that is coherent with the presuppositions and principles that define him/her and the reality that he/she recognizes. In view of this reality, an analysis focused on the presupposition of duality, contradiction, and complementarity between "tradition" and "modernity" would not be appropriate just because one observe the mere presence of elements externally classified as "traditional" and "modern". At the pragmatic level, subjects do not differentiate them - except a posteriori, when they impose this taxonomy on themselves as a discursive need to delimit fields of experience and knowledge they want to circumscribe. Therefore, what matters is what is done with these elements and what they refer to for the subjects that manipulate them. I have watched some ceremonies identified by the subjects of my research as "traditional":  mhambas, lobolos, funerals, and depositions of flowers¹⁸. They were all pervaded by "modern" elements, as were the lobolos observed by Brigitte Bagnol (2006). However, these elements were there only as distinguishing markers of a certain urban status of the participants and were not perceived as anachronistic and contradictory with the "traditional" character experienced in those ceremonies. Therefore, they reaffirm what Bagnol indicates regarding the lobolo: these ceremonies are of an ontological character and thus restate and reaffirm principles of the construction of worlds and persons that are based on persistent foundations related to kinship and cosmology. That is why they make explicit some crucial aspects to be discussed here.

During a mhamba I watched in July 2007 in Homoíne, carried out by a family that was reconstructed after the massacre of 1987¹⁹, I asked the head of the household whether the curandeira that mediated the consultation with ancestors was an inhabitant of that area. He was astounded with my question and denied it vehemently. He said they could never hire a curandeira who lived close to the family or was acquainted with them. Putting her into contact with the protecting ancestors would be a great risk, as she could break their protection and trigger evil deeds against the family. This indicates something that I noticed as something generalized in that field: evil deeds do not come from far away, but from very close - from the family group, from neighbors, and more rarely, from the other social circles and networks in which people are inserted. The case of José's disease and death makes it explicit. The fields of sociality which offer more potential of aggression and risk are always those which are closer to the subjects. It is as if relatives and neighbors were permanent potential enemies, and it is precisely for this reason that most of the accusations of sorcery and preventive attitudes are directed towards them. And there are many accusations of sorcery in Homoíne.

Throughout my research, I found a social environment in which people are afraid of each other and express this in the form of reciprocal discourses and acts. This fear is usually expressed in conversations about violence unleashed by spells or by actions of bandidos (bandits) - this is how criminals are named, who usually steal small things but also commit physical aggression and eventually kill someone. There are discourses about a real or presumed violence and there is a social environment of insecurity that results from the perception of a permanent vulnerability to the actions of bandidos and sorcerers who, although different in categories and processes, operate in a similar way in the construction of fear and suspicion. What unites these categories is the notion of ambition, perceived as the motivation of their actions. But discourses are expressed differently: when they refer to bandidagem (banditry), people speak objectively, directly, and explicitly; when they refer to sorcery, this is done with half words and ellipses, in a fragmented way, as rumors. While bandidagem threatens goods, houses, and bodies through direct actions, imposing preventive and punitive measures which are also direct and objective against their agents, sorcery, on the other hand, threatens the same elements and expands them to the family and to other circles of personal and social relations by means of indirect, capillary, silent actions, making it more difficult to identify, prevent, and punish, and rendering them as the cause of most "traditional diseases". This rhizomatic²⁰ nature of agency turns sorcery into a more diffuse and amplified, less objective and punctual element than bandidagem and, therefore, more fearful.

In the speeches about violence that I heard in Homoíne, sorcery was closer to the descriptions of unpredictable attacks by matsangas²² (RENAMO'S bandidos armados) than to the current bandidagem, despite the noticeable connections that these speeches produce between war, criminality, and sorcery as violence - the speeches on these themes were very similar structurally. The indetermination of sorcery, always inserted in a speculative field of obscure causes and consequences, continued and non-structured processes, lends it a character of greater power and danger than that recognized in the contemporary bandidagem, and much closer to the assigned to matsangas²³. It is important to point out that sorcery was associated with representations of violence during the civil war.²⁴ Sorcery renders a form and an expression to a diffuse and permanent insecurity in view of the dangers, unpredictable events and imponderabilia of daily life that put people in risk.

Another aspect that causes insecurity and triggers speeches on an imminent threat of deconstruction of the person concerns the management by ancestors of the individual and family life as well as of nature. Protectors on the one hand, the ancestors are also object of anxiety for the people, since their discontent results in punishment for the living. Breaks in family and kinship bonds, disobedience to rules, taboos, and disrespect to the natural elements under their guard revert into evil deeds for the living. Owners of the land and of the natural cycles and elements, ancestors are also owners of certain women, who are given to them as wives - a fact that has been previously mentioned and has also been described and analyzed by Honwana (2002) and Bagnol (2006). This was part of the speculations concerning José's disease. Access to these elements and people require the authorization of ancestors. Inadequate use of their belongings and the failure to follow their orders result in punishment. Protectors at first, ancestors frequently transform their descendants into enemies in a given sense, and as a consequence they also become enemies and predators of their descent.

Still regarding the social predation scheme observed in Homoíne, I found a remarkable equivalence and even an indistinction between curandeiros and sorcerers for people in the village. Instead of integrators - as Honwana (2002) describes -, the inhabitants of the village see curandeiros as potential disintegrators of the social life and of the personhood. As disintegrators as the sorcerers, and the more frequently the closer they are to the village - from this perspective, to the local inhabitants, curandeiros from the mato would still heal and would not produce evil deeds. People speak of curandeiros that produce evil deeds, carry on intrigues among relatives and neighbors, turn their clients into slaves because of interminable and unpayable debts, and are consulted in order to cast spells to the benefit of their "ambitious" clients. Curandeiros are accessible only to those who have belongings, for consulting them is expensive, which constitutes a phenomenon of social class in Homoíne. Those who do not have any belongings, besides seeking shelter against curandeiros and their "ambitious" clients in churches, also flea to larger urban centers, where they feel less vulnerable to "tradition" and away from family conflicts that may lead to evil deeds. Noticeably women seek churches to protect themselves against evil deeds and accusations of being sorcerers.

Consulting with curandeiros is considered a way to obtain benefits for social climbing and capital gains in a labor market that is extremely competitive because of the extremely low job offer.²⁵ This search for benefits is even expressed by a specific verb: kukhendla, in Citswa - in Portuguese, the verb becomes a neologism: khendlar. Kukhendla traditionally means to look for magical powers produced by curandeiros in order to obtain goods, physical force, defense against enemies, and personal qualities that make the person socially well-considered and positioned. During the civil war, soldiers used kukhendla to obtain protection against the weapons of their enemies. Today, in times of peace and free market, the demand switched to the obtaining of job and material benefits, and the availability of money in circulation causes inflation in the prices of curandeiros. Therefore, a constant historical updating of this phenomenon is observed.

For the kukhendla to be effective and lasting it requires the sacrifice of animals and people death in return - people that are usually part of the family. The spirits of the dead are given to the curandeiros, who use them to increase their own powers. Kukhendla is one of the phenomena associated with the current notion of spell I found in Homoíne, and it is the main element in the construction of the equivalence between curandeirismo and sorcery expressed in its inhabitants' speeches. Moreover, it is the explanatory principle of several of the accusations addressed to José's father. The current demand for curandeiros in the village has gone from healing to obtaining personal benefits connected with the requirements of a "modern" life. Meneses (2004) had already mentioned that the afflictions of "modernity" referred to in the market eventually promote the search for benefits with curandeiros. In Homoíne, the consequences of this search were made explicit in the subjects' speeches: the benefit of some requires the harm to others, which results in generalized insecurity and suspicion concerning those who are successful in their lives. Curandeirismo is thus referred to in the field of discourse on violence against the person. While the search for well-being and success results from the very process of construction of the person, this process happens in a constant experience of potential risks and dangers, involving permanent disputes and negotiations with elements that may hamper its accomplishment.

In the fields defined by descent and matrimonial alliances, there is a recurrence of accusations of deconstruction of a person due to casting spells, as exemplified in José's case. In most cases, women are accused of causing tension, diseases, and deaths among their affines and descendants. This articulation between the construction of the person, the action of women inside the family core, and the production of diseases will be discussed further ahead.

All these phenomena concerning the construction of the person and the risks of its deconstruction may be interpreted from a common standpoint that allows us to understand them from the perspective of an analysis used contemporarily in studies of Amerindian groups.²⁶ All these facts indicate that the social system at issue is based on the conflict that certain social mechanisms aim to pacify. In this sense, the conflicts unleashed or represented by sorcery, bandidagem, matsangas, curandeirismo, and ancestors are not anomic disruptions of a social order that is functional and harmonious by principle. The conflict is a permanent possibility of this very social order. This order is constructed based on the acknowledgement of the conflict and predation as permanent and virtual foundations of relationships with primordial enemies whose origin date back to the pre-colonial wars for occupation of that area (cf.  Passador, 2009b), which requires mechanisms and process of constant pacification through socialities that produce people and social networks. In this case, it is possible to rethink the kinship system in the investigated area and the alliances that extend beyond that system, assuming that they operate as a matrix of permanent and systematic pacification of potential enemies inserted in the fields of relations that constitute the person. Social networks create fields of relations in permanent risk of deconstruction of the pacification processes that constructed them. The enemy is not an element constructed by a punctual conflict that disrupts previous orders of solidarity. They arise from the deconstruction of the relations that had pacified them when they were introduced into fields of socialities under a regime of hierarchical power. Therefore, the social construction of enemies takes place through their ontological replacement, by deconstructing a field of agencies and socialities intended to pacify them. This explains why the enemy is always very close and why accusations of sorcery are usually addressed to relatives and neighbors - especially women. The example of José's case and the speculations around his disease and death may be understood according to the terms of this proposition.

More objectively, the proposal here is to think of matrimonial alliances in that area as mechanisms for the incorporation and pacification of potential enemies in a core of descendants that has already been pacified by an agnatic hierarchy. The perpetuation of a system of agnatic descent requires the incorporation of women and affines by means of matrimonial alliances. Women and affines come from the outside, from another descent group, as foreigners or vientes²⁷ - expressions used indistinctly in Homoíne to refer subjects that come from outside a certain universe considered to be the "inside". Vientes are always subject to suspicion and fear, for they are - by principle - potential enemies that require pacification. These "outside" elements, women and affines incorporated by alliance, are the ones that cause the most fears of aggression and deconstruction of a person.

On the other hand, alliances and pacifications operated by ascendants can be interrupted by their descents by interrupting the investments of the person that started the descent core. Ascendants, on their turn, due to their generational position in the hierarchy, may punish insubordinate descendants and subject them to their interests and ambitions - which is the case in the relation between José and his father, as well as in the more general case of punishment by ancestors. Thus, women and affines are potential enemies, as well as ascendants and descendants. This is the reason for constant ambiguity in kinship relations: if the person is constructed in a network of relations of alliance and descent, there is also a great potential of predation and deconstruction of the person by the same relations.

The result of all these is a social matrix that operates through reciprocal and solidary exchanges, and depends on a hierarchic power to be preserved, as their agents are basically pacified enemies that may potentially return to their original predatory condition. This is how the person is constructed and deconstructed, and this is how health, disease, and death by sorcery may be understood in terms of the "traditional" system, as in the case of José's death.

Gender, power, and disease

If this interpretation is correct, then we have a pathway to understand the relations of gender and generation that pervade this matrix and that turn into recurrent conflicts associated with women and elders - especially older women. In Homoíne, women are feared because people believe they have power to cause ruptures, produce aggressions, and disintegrate sociality fields. Therefore, they are potential enemies submitted to constant procedures carried out to prevent evils as well as to hierarchical pacification on the part of men. As they always come from the outside in the matrimonial alliances scheme, they remain as vientes, potential enemies, just as their family group of origin - the affines. Hence, the power to which they are submitted in this matrix operates as a male counter-power that seeks to invalidate the evils they may cause in the universe of the family. This is necessary for the construction of the male person, because descent, alliances, and masculinity are not possible without women.  But that is not all.

Women are seen as having a power of ambiguous and specific nature. As they are the sorcerers and curandeiras, they are the ones who have a special relationship with the world of the spirits and the capacity to make them act on the world of the living. When they are curandeiras, they are the ones who enable the agencies of spirits in social life, and this makes them vital to the integration and maintenance of a desirable order. When they are mothers and wives, they make descent and alliances possible. Therefore, they stand as the possibility of constructing persons. But they are also the ones who cast spells on relatives and neighbors and carry on intrigues and conflicts as curandeiras, deconstructing persons. This ambiguous power results from the very position they have in local schemes of descent and alliances, for these mechanisms operate through them and in them.

Strathern (1988) stated that the ideal relations of alliances between men, to which she refers as same-sex relations, are possible only through actual relations of alliance between men and women, or cross-sex relations. The relationship of virtual and potential affinity between family groups headed by men can only be accomplished in concrete terms through the matrimonial alliances between men and women from these groups. Thus, women are the ones vested with potential to establish and ratify alliances.

The same phenomenon is observed at the level of spirits and curandeirismo. Honwana states that the nyamussoro²⁸ is originally a Vandau's category, from Central Mozambique, and this was an extremely women-only category among that group, that was later redesigned in the Southern region as a result of the pre-colonial wars. When we pay attention to the author's description of the emergence of this category in Southern Mozambique, we notice that it is only possible when constituted by women and their role in descents and alliances: the incorporation of male Vanguni and Vandau spirits to Tsonga groups occurred through cross-sex alliances between these spirits and the women given to them as wives (the nsati wa svikwembo), turning them into affines and transmitting them by descent to the next generations - which allowed men to become vanyamussoro at a later moment through the transmission and possession of spirits of their female vanyamussoro ancestors. Thus the power of healing is guaranteed by the power of establishing matrimonial alliances, incorporating spirits who become affines of a descent group. This is a female power and turns the nyamussoro into an originally female category by definition. This incorporation of spirits takes place through pacification processes, for before becoming healing spirits, they were warrior, revengeful spirits who made a group of agnatic descent vulnerable. Hence, there is a parallel between alliance and pacification and the family field appears as a field that is defined and produced by strategies to pacify potential enemies from the outside to the inside, and this can only be accomplished through women in cross-sex relations.

These facts once again point out the ambiguity of the female situation in Southern Mozambique: as integrators and pacifiers, women remain as vientes that are incorporated and pacified through matrimonial alliance in order to make descent possible in a group to which they do not originally belong. On the one hand, they intermediate male spiritual forces and reproductive forces that make possible the existence of a person by means of descents and alliances. On the other hand, they remain as potential enemies that can break the peace agreement that the alliance establishes. This situation gives them the acknowledged and feared power. Male power becomes effective only through the female power, for the latter guarantees the male agency in the social and spiritual fields. However, the female power can also cause disruptions and conflicts that hinder the maintenance of an order governed by male elements. This is the type of ontological female power to which accusations of sorcery addressed at women try to give form and intelligibility in the social universe observed, and the one that creates fear of women and the diseases they can produce.

Many are the cases of women accused of sorcery in Homoíne. Among these many cases, there are cases of spells cast by mothers-in-law against their daughters-in-law, which indicates that gender and generation operate to delimitate the most predatory categories - and their vulnerability to accusations. An objective indicator of this situation was the existence of a group of approximately 30 old women living in the Catholic mission close to the village and known by the local population as the "mission's old women". According to the missionaries who gave them shelter, those "old women" had been expelled from home by their own children after having been accused of sorcery by their daughters-in-law. Their daughters-in-law accused them of being responsible for their being infertile, of causing diseases and death of their babies, or of causing decrease of productivity in their machambas. This is to say, accusations refer to the incapability of daughters-in-law to generate descent and produce subsistence for their husbands, and therefore fail to comply with the alliance agreement established in the lobolo.

This tension between daughters- and mothers-in-law seems to be a structural datum: if daughters-in-law do not produce children and descendants, they become unsuccessful women and incomplete persons. As a consequence, they cannot make their husbands complete as a male person. The same logic applies to their machambas: if daughters-in-law cannot guarantee the subsistence of the family group, they become unsuccessful and incomplete women. When daughters-in-law are unsuccessful, the marriage is usually disrupted, they are returned to their family of origin along with the goods negotiated in the lobolo. On their turn, mothers-in-law reiterate their presence in the group to which they were incorporated when they got married and reaffirm their ascent on their sons. Sharing the domestic space with the daughters-in-law in the general scheme of virilocal residence, preferably patrilocal residence, the mothers-in-law submit their daughters-in-law to a hierarchical power relationship. The tensions between them are notorious and their roots seem to be the women's constant need to reaffirm their membership and subordination to the agnatic group to which they have been incorporated by alliance, and to which they remain legitimately connected only if they guarantee its continuity through descent and subsistence - children and food are the fundamental goods women are supposed to produce. Disruptions in descent and subsistence turn them into the main suspects of causing them.

Another example that associates women with evil deeds inside their own circuit of descent and alliances is the also common accusation of spells made by wives against their husbands in order to benefit from the family properties that he belongs. Every widow knows she will be a suspect of having caused the death of her husband. The foundation for this type of accusation is that wives, moved by ambition, would take advantage of their influence on their children to benefit from the inheritance entitled to them after the death of their father, in a scheme of adelphic succession. Sorcery appears here as a possibility to make intelligible and give form to the causes of the problems related to predation within the family group, problems which are usually attributed to women incorporated by alliance and affinity.

It is believed that when women get old, are no longer fertile and have lower productivity in their machambas, they become sorcerers in order to make their living and obtain benefits, a fact that cause husbands to fear their old wives. Old and lonely wives are always much feared and accused of being sorcerers, for their only way to make a living would be ambition and predation.

Women are always seen as being potentially evil: by means of supposed voluntary evil deeds created through sorcery, they would put women and men in risk for believing that these may also put their integrity as a person at risk. This is the reason why women are a permanent object of fear and suspicion. More than structural or functional, this recognition of evil as inherent to women - particularly older ones, but not only - assumes an ontological character, as it is not restricted to the position of women within a kinship system. It expands to a condition of potential female impurity and danger. Contact with the feminine elements and their ambiguous power involves presumed risks of deconstruction of the person, although they represent the possibility of its construction and persistence at the same time. The fact that this power is attributed to women causes them to be subordinated to a male counter-power that is made legitimate by the social spheres of the family and kinship, which makes it possible to keep them pacified in order to produce the person and the social matrix that has produced them. They are the object of actual violence for being the perpetrators of a presumed violence in a system that is built around a permanent and potential conflict, placing enemies very close to their victims.

The consequences of this situation are reflected in the field of diseases, as previously discussed. In this context, diseases have the potential to become symptoms of contact with certain categories of persons marked by gender, or of the voluntary actions of these persons. For this reason, diseases are frequently taken for more than mere physiological events to be treated by biomedicine - they constitute the category of "traditional disease". Therefore, diseases provide concrete and critical experiences that make it possible to articulate at the pragmatic level the ontological principles that operate in the construction of the person and genders.

For instance, in a scenario of HIV/AIDS epidemics with high rates of prevalence and death, such as the case of Southern Mozambique, "traditional" experiences with diseases ultimately define relations with its prevention and treatment (cf. Passador, 2009a). The association between Aids and xivenze that I found in some speeches of subjects in my research is very illustrative. This is made possible by conceptions of disease elaborated around cosmologic precepts and ontological conflicts that define people, bodies, nature, gender, sexuality, and a whole lot of pragmatic experiences based on them, which then redefines the experience with the HIV/AIDS.

Hence, it is possible to observe a constant elaboration of experiences with diseases in terms which the notion of personhood and the gender relations are built, pragmatically incorporated to the logic of a persistent and "traditional" ontology. In this universe of "tradition", women are always rendered as problematic and linked to pathologies in specific ways that corroborate to stigmatize them as powerful "whores" and "evil" sorcerers, as attested by José's father.

References

Bagnol, Brigitte. Gender, self, multiple identities, violence and magical interpretations in lobolo practices in Southern Mozambique. Thesis for Ph.D. Degree in Social Anthropology. Department of Social Anthropology, University of Cape Town, 2006.

Butler, Judith. Gender trouble: feminism and the subversion of identity. New York, London, Routledge, 1990.

Comaroff, John L. & Comaroff, Jean. On personhood: an anthropological perspective from Africa. Social Identities, 7(2), 2001, pp.267-283.

Deleuze, Gilles & Guattari, Félix. Mille plateaux: Capitalisme et schizophrenie. Paris, Editions de Minuit, 1980.

Douglas, Mary. Powers and dangers. In: Purity and danger: An analysis of the concepts of pollution and taboo. London, Routledge, 2002 [1966], pp.117-140.

Fausto, Carlos. Inimigos fiéis: história, guerra e xamanismo na Amazônia. São Paulo, EDUSP, 2001.

Geffray, Christian. La cause des armes au Mozambique: anthropologie d'une guerre civile. Paris/Nairobi, Karthala/ CREDU, 1990.

Goodman, Nelson. Ways of worldmaking. Indianapolis, Hackett, 1978.

Helgesson, Alf. The Tswa response to christianity. Dissertation for Master Degree, Faculty of Arts, University of Witwatersrand, Johannesburg, 1971.

Honwana, Alcinda Manuel. Espíritos vivos, tradições modernas: Possessão de espíritos e reintegração social pós-guerra no Sul de Moçambique. Maputo, Promédia, 2002.

Jardim, Marta. De sogra para nora para sogra: Redes de comércio e de família em Moçambique. Cadernos Pagu (29), Núcleo de Estudos de Gênero - Pagu/Unicamp, 2007, pp.139-170.

Junod, Henri A. The life of a south african tribe. Vols. I and II. Montana, Kessinger Publishing, 2003 [1926].

Meneses, Maria Paula. "Quando não há problemas, estamos de boa saúde, sem azar nem nada": para uma concepção emancipatória da saúde e das medicinas. In: Santos, Boaventura de Sousa & Cruz e Silva, Teresa. Moçambique e a reinvenção da emancipação social. Maputo, Centro de Formação Jurídica e Judiciária, 2004, pp.77-110.

_________; Fumo, Joaquim; Mbilana, Guilherme & Gomes, Conceição. As autoridades tradicionais no contexto do pluralismo jurídico. In: Santos, Boaventura de Sousa & Trindade, João Carlos. (eds.) Conflito e transformação social: Uma paisagem das justiças em Moçambique, 1° Volume. Porto, Edições Afrontamento, 2003, pp.341-425.

Passador, Luiz Henrique. "Tradition", person, gender and STD/HIV/AIDS in Southern Mozambique. Cadernos de Saúde Pública, 25(3), 2009a, pp.687-693.

_________. "Eating alone" or when modernity feeds tradition: Money and magic in Southern Mozambique. VIBRANT, 5(2), 2009b, pp.100-129 [http://www.vibrant.org.br].

_________. & Thomaz, Omar Ribeiro. Raça, sexualidade e doença em Moçambique. Revista Estudos Feministas, 14(1), 2006, pp.263-286.

Strathern, Marylin. The gender of the gift: problems with women and problems with society in Melanesia. Berkeley, University of California, 1988.

Viveiros de Castro, Eduardo. A inconstância da alma selvagem - e outros ensaios de antropologia. São Paulo, Cossac & Naify, 2002.

Webster, David J. Agnation, alternative structures, and the individual in Chopi society. Dissertation for Doctor Degree of Philosophy, Faculty of Arts, Rhodes University, Grahamstown, 1976.

* Text received for publication in September, 2009, and accepted in August, 2010. This text is the result of the research developed for my doctoral studies (PPGAS-Unicamp) on "tradition" and the universe of diseases in Southern Mozambique, with special focus on the local experience with the HIV/AIDS epidemics. Fieldwork was conducted in the District of Homoíne, in the Province of Inhambane, between June 2007 and February 2008, funded by CAPES through the Doctorate Intership Program Abroad. The research was initially supervised by Dr. Omar Ribeiro Thomaz and then supervised by Dr. Mariza Corrêa (PPGAS-Unicamp), being co-supervised by Dr. Teresa Cruz e Silva (CEA-UEM). ]]> ** Ph.D. in Social Anthropology, PPGAS-Unicamp. lhpassador@gmail.com
1  All the names of people that appear in the text are fictitious.
2  The expressions "curandeiro" (fem. curandeira) and "curandeirismo" are currently used in the Portuguese language by the inhabitants of Homoíne to refer respectively to traditional healers and traditional healing. The expression for "curandeiro" in the local language (Citsua) is ‘nyanga'.
3  Machamba is the local expression used to refer to crops.
4  Lobolo is the name given to the traditional process of matrimonial alliance in Southern Mozambique (see Bagnol 2006 and Junod 1996). This expression is also currently used to refer to the specific ceremony that celebrates the alliance. The lobolo is traditionally acknowledged as marriage, although there are civil forms of marriage defined by law.
5  Shortly before that, José had already had malaria, which may be supposed that he was facing a relapse. A few months earlier, he had also had a respiratory disease for which he was not able to inform the exact diagnosis. Its main symptom was prolonged coughing.
6  Tihlolo is the divinatory process used by curandeiros to interpret the causes of an evil deed. It is also known as cuxa-cuxa, expression created by the Portuguese colonizers.
7  I use the expressions "tradition" and "modernity" between quotation marks herein because I refer to the emic uses of these categories made by the inhabitants of Homoíne.
8  Kutchinga or kutxinga is the name given to the ritual for the purification of people and goods associated with a recently deceased person when he/she was alive. His/her goods may only be accessed after purification. Widows and widowers must also be purified in order to have sexual intercourse with other partners. Until they are purified, the deceased's relatives and widows must refrain from having sexual intercourse and may not use the goods that belonged to the deceased are completely purified. Traditionally, the kutchinga also involves sexual intercourse of the widow with her brother-in-law. The basic idea is that the semen "washes away" the woman's impurities. When her brother-in-law refuses to do so, another member of the family is chosen or a man out of the family group is hired to do it.
9  Refers to women who have many sexual partners but is not necessarily a sex professional. ]]> 10 I use the notions of performance and performativity to point out that identities and personhood result from a pragmatic doing, taking the same theoretical perspective used by Judith Butler (1990) to analyze genders.
11 Banca is the term used to indicate small houses where small trade activities are pursued. Barraca is the name of the covered huts observed in public markets and streets.
12 I use the notion of construction of worlds refering to Nelson Goodman's (1978) notion of worldmaking.
13 David Webster (1976) drags our attention to the fluidity of these structures among the Chopi, who inhabit an area adjacent to the one I studied.
14 Madoda is the term used to refer to "great" men, usually elders, who are respected and become a reference for communities and families.
15 Regarding the idea of sograria, also used among Hindu Indians in Southern Mozambique, see Jardim (2007).
16 On the uses of the term xará (namesake) and the role of agnatic and matrilateral relatives, neighbors, and formal friends, see Webster (1976). Xarás are individuals who share the same name. The exchange of names is a form of alliance that establishes a xará as a sort of one's relative. The same is true for neighbors and formal friends.
17 On the notion of "great man", see Junod (2003 [1926]) and Webster (1976).
18 The deposition of flowers is the last phase of the funeral process, carried out eight days after burial.
19 On July 18, 1987, the main village of Homoíne staged the most significant massacre that occurred during the civil war in Mozambique (1977-1992). On that occasion, according to official statistics, 428 people were killed during the ten-hour attack. ]]> 20 I use the notion of rhizome as proposed by Gilles Deleuze and Felix Guattari (Deleuze & Guattari 1980), understanding that it can explain how unstructured and unpredictable the actions and processes attributed to sorcery are.
21 Acronym of the Associação dos Médicos Tradicionais de Moçambique (Association of the Traditional Healers of Mozambique), entity acknowledged by the Mozambican S tate. During my fieldwork I could observe the prevalence of its legal action in the solution of cases of sorcery accusations, including some of them that were sent to AMETRAMO by the very court of the District of Homoíne. On the history and action of AMETRAMO, see Meneses (2004), Meneses et alli (2003) and Honwana (2002).
22 Expression used to refer to soldiers of RENAMO (Mozambique National Resistance), which opposed the government of FRELIMO (Mozambique Liberty Front) during the civil war. The etymological root of the word matsanga is the name of the first leader of RENAMO, André Matsangaíssa. People also usually referred to RENAMO soldiers as bandidos armados (armed bandits) to qualify them as criminals.
23 According to Douglas (2002), the notions of danger and power are associated with interstitial, undetermined, and non-structured spaces.
24 On the uses of sorcery, traditional healing, and religion during the civil war, see Geffray (1990) and Honwana (2002).
25 On relationships between sorcery, curandeirismo, and market, see Passador (2009b).
26 Especially the works of Viveiros de Castro (2002) and Fausto (2001).
27 The etymological root of the word viente is the Portuguese verb vir (to come) and refers to people who come from somewhere else. It operates as a marker of alterity regarding the original location and/or residence of a person.
28 Nyamussoro (pl. vanyamussoro), according to Honwana (2002), is the most powerful category of practitioner in the traditional healing system of Southern Mozambique. A nyamussoro operates Tsonga lineage spirits (tinguluve, sing. nguluve) together with Nguni (pl. vanguni) and Ndau (pl. vandau) spirits, respectively from the Zulu and from Central Mozambique areas. According to the author, this category emerged in Southern region as a result of the invasion of the Vanguni, who subordinated the Tsonga and the Vandau and produced matrimonial, spiritual, and hierarchical alliances between these three groups.