Griffiths

This paper discusses issues of Popular Justice in the context of research carried out over a number of years (1982‑89) in Botswana, Southern Africa. This research was situated in the western region of the country known as Kweneng district, in a village, Molepolole which functions as the central headquarters for Bakwena, who represent one of the oldest Tswana merafe (tribes or polities) in Botswana. The research focused on family relations and in particular on sexual relationships between women and men. It examined the status of such relationships and how marital or non‑marital status affected women's claims on their partners for compensation for pregnancy, maintenance and rights to property in both social and legal terms. It is from this perspective and in terms of the village courts in Molepolole that the paper raises questions about the nature of popular justice which reflect a colonial/post colonial dimension, as well as national, regional and local domains. Running through the discussion is the issue of gender which arises through the systematic constraints which operate on women in their dealings with men.

As this paper is based on an ethnographic study I would like to give a brief account of the terms of reference under which the research was carried out. It was prompted by an invitation to prepare the foundation course in family law for the newly established law department at the University of Botswana in 1981. Botswana was subject to British colonial overrule as the Bechuanaland Protectorate until it gained its Independence in 1966. On Independence it adopted a Parliamentary form of democracy derived from the British Westminster model and which was based in theory on a separation of powers between the legislature, the executive and the judiciary.

Up until this point the model used for teaching law had been European in origin and based on statutes and cases in the Magistrates' courts and the High Court. I [page 196] considered this to be too limited and having encountered the anthropologist Schapera's A Handbook of Tswana Law and Custom (1938) wanted to look at family relations and people's use of law in a broader context. This involved doing fieldwork on the ground and looking at the interaction between what were designated 'Customary' and 'European' or 'Common law' forums.

Molepolole where the research was carried out lies at the heart of Kwena affairs in terms of the social, political and ritual life of the morafe. The population of the village, which was estimated to be 20,000 in 1980 (Botswana 1983) has grown until it was reputed in 1992 to have overtaken the Ngwato capital of Serowe as the largest village in Africa. This population fluctuates as a result of seasonal migration to the lands to engage in subsistence agriculture, or to the cattle post to deal with livestock. These activities represent an essential part of a Kwena family's livelihood together with migrant labour which is the major source of a cash income for most families in Botswana (Botswana 1982). Located close to the railway line and to the border with South Africa, Molepolole had and continues to have one of the highest rates for migrant labour in the country.

My first research visit to Molepolole lasted from January to August 1982. At that time I was concentrating on dispute forums or courts, attending cases and generally interviewing local people throughout the village. I was very fortunate to be able to work with Mr Masimega as an interpreter. In his seventies at that time he was a well respected member of the community, having been the private secretary to several Dikgosi (chiefs) and having sat as a menber of the local district council when it came into being. He was known throughout the village by the nickname of 'Mr Commonsense'.

In 1984 I went back to the village for three months. On this occasion I was following up on disputes, personnel and individuals from the previous period, but in addition was working on the life histories of people from one social unit in the village, Mosotho kgotla. A kgotla is the basic building block on which a Tswana morafe is founded. Originally based on agnatic connections through the male line (but not often today), a kgotla represents a number of households that are presided over by a headman. There were roughly seventy of these kgotlas when I worked in the village in 1982. Moving upwards in authority beyond the individual kgotla to a number of kgotlas grouped together we come to the ward which is presided over by a ward head. There are six major wards in Molepolole,^{^[1]} the most senior of which is the Chief's ward, Kgosing. It is in Kgosing ward that the Chief's kgotla which presides over the whole morafe is located.

From fieldnotes which Schapera kindly gave me, I was able to trace families and households associated with Mosotho kgotla from 1937 through to 1984. This provided me with data on men and women over several generations. In 1989 I returned for two months in order to follow up on all these aspects of my research, gaining a longitudinal perspective stretching over eight years. I also paid a brief visit to the village in 1992.

In Molepolole, as in other central villages in Botswana, the local courts are the Magistrate's court and the Chief's kgotla. Within the formal legal system of the country the Magistrate's court is designated a 'common law' forum in contrast with the Chief's kgotla which is referred to as a 'customary' law forum. Common law is defined as "any law, whether written or unwritten, in force in Botswana, other than customary law" (Customary Law (Application and Ascertainment) Act 1969, s. 2). Customary law is defined as being "in relation to any particular tribe or tribal community the customary law of that tribe of tribal community so far as it is not incompatible with the provision of any written law or contrary to morality, humanity or natural justice" (1969 Act, s. 2). This follows a formulation adopted in a nunber of African countries subject to colonial overrule (Chanock 1985). The early writings of the late '60s and the '70s concerned with the relationship between the state, law and justice tended to set up a dichotomy between what they characterized as 'formal' or 'state' justice and 'informal' or 'popular' justice.

The model of formal justice was said to reflect a rule‑based institution associated with central or state government which was dependent on experts for personnel. These experts included judicial officers and lawyers who required a legal training. The model operated on the basis of sanctions or coercion and set up a situation in which parties found themselves 'winners' or 'losers'.

In contrast 'informal' or 'popular' justice was defined as having all the qualities that the formal system lacked. It took on its existence by default, by creating itself in terms of a mirror image of the formal legal system. It was community based, decentralised and did not require particular expertise for its operation which was based on broader prescriptions than those provided by formal legal rules. This decentralisation, deprofessionalisation and delegalization gave rise to increased participation and more access to justice. Those engaged in such a process, including the parties, maintained control over it and so it reflected a more collective, democratic and egalitarian mode of operation.

At a superficial level these models appear to have some resonance with the village courts in Kweneng. The Magistrate's court appears to represent a model of 'formal' justice for the following reasons. It owes its existence to formal legal rules. It is a creature of statute, administered by the state or central government. It relies on those with legal expertise to administer it. It is presided over by a Magistrate who must have a legal training. Such a person rarely has any association with local kin networks and tends to be viewed as one who stands apart from the village community. This is underlined by the fact that in the past he tended to be an expatriate from Britain or to come from other African countries such as Tanzania or Uganda. He (for virtually all magistrates were male) came into to the village on one day a week from the capital city Gaborone, which is about 70 kilometers from the village by tarmacadam road, and returned there at the end of the day.

Things have changed somewhat since 1982 and there are now a greater number of Batswana in post including the first Motswana^{^[2]} Chief Justice who was appointed in 1990 and some women. In 1989 the post of Resident Magistrate was established for Molepolole but although there is now a Magistrate working there five days a week he still comes from another morafe. While accommodation is provided he does not necessarily live in the village and often continues to commute from Gaborone. He has no connection with the community other than through his job.

This link with the capital and thus with central government is reinforced by the physical setting of the court. It was originally a low verandah type of building (built to a colonial design) and housed within district administration. It is now part of a huge complex which has the first multistorey buildings in the village built in 1986 in which nearly all local government organisations are now located. These include the Immigration, Wildlife, Welfare and Social Services departments as well as the District Commissioner's and other administrative offices. As part of this complex the Magistrate's court is very visibly a symbol of administrative power. Other aspects which link it to the formal legal system include the written documents and case records which are required for its operation, which are in English, rather than Setswana which is regarded as the predominant 'indigenous' language. There is also the provision for legal representation.

In contrast the Chief's kgotla, known as Kgosing, appears to reflect a model of 'informal' or 'popular' justice when acting as a court under the auspices of the formal legal system. That it is different from the Magistrate's court is marked by the fact of physical separation. It is located on the top of a hill ten minutes' walk [page 199] from the complex housing all the other government buildings. It has been settled in its current site since 1937. Originally this site was shared with some of the district council offices but these have all been moved to the new complex down the road. There is some resentment about this as the personnel associated with the Chief's kgotla under the heading of 'tribal administration' feel they have been neglected and ignored. They have not shared in the general upgrading and expansion of services beyond being allowed to expand into the buildings vacated by the district council.

Unlike the Magistrate, those hearing disputes have no legal training. Indeed no legal representation is permitted in this arena although the issue of whether this should be permitted and whether personnel should have legal training has been a hot topic of debate in the village. Not only are those hearing the disputes, the Chief Regent, the Deputy Chief and the Senior Chief's Representative, not legally trained, but they hold office through their connections with the community, that is, on the basis of their links with the Kwena ruling family descended from Sechele I (c.1829‑1892). This has given rise to much infighting and genealogical manipulation concerning these appointments.

The kgotla over which they preside with varying degrees of authority, lies at the heart of Kwena society and engages in in a broad complex of activity, only part of which involves the hearing of disputes. All kgotlas in Molepolole deal with disputes, but it is only the Chief's kgotla which has the status of a court conferred on it by the formal legal system under the Customary Courts Act. However, process in the Chief's kgotla (as in other kgotla hearings) involves active participation by those engaged in disputes, who are their own advocates, as well as participation by local kgotla members. This is so for all hearings and not just those concerned with family matters. The party initiating the hearing before a third party begins by making a statement. In family hearings this usually involves a woman asking for compensation for pregnancy from her male partner or where married making a statement about her husband neglecting her and their children for another woman, or seeking a division of property.

After the statement the individual who has given it is questioned by the third party handling the hearing, the other party involved and finally by any member of the public in the kgotla. The same procedure is applied to the other party who gives their version of events. At the end of this both parties are asked what they want to do and kgotla members are invited to give their opinions, after which the third party makes a pronouncement. In family cases, for example, this may be that the woman is entitled to compensation for pregnancy or that the parties are to go and live together in peace and harmony.

The proceedings remain oral except at the level of the Chief's kgotla where they [page 200] are recorded in Setswana. Where parties wish to take the matter beyond the Chief's kgotla they must appeal, either to the Customary Court of Appeal established in 1986,^{^[3]} or to the Magistrate's or High Court, in which case the written records must be translated into English. All this is taken to underline the fact that the Chief's kgotla represents local indigenous popular justice in contrast with the Maqistrate's and Hiqh court.

However a closer examination reveals that there are aspects of both courts which belie such a characterisation. In certain respects the Magistrate's court, as part of the state system, with its reliance on trained experts and with the costs which stem from this, is seen as providing less access to and participation in justice. However, while the Magistrate has a legal training, it is not necessary for the parties to be legally represented, and in family matters, such as maintenance cases, almost always the parties appear on their own behalf. There is no stigma in doing so and no perception of the individual who does so as a difficult or awkward litigant. The Magistrate will sit with the parties on their own with the clerk of court and deal with the discussion in an informal manner. Members of the public are not generally present and the atmosphere is not like that of formal court proceedings.

The proceedings are governed by rules which are seen as inflexible and of which the parties may have no knowledge. Examples of this in the maintenance field are that a woman must provide 'proof' of paternity and a claim must generally be brought within a year of the birth of the child (Affiliation Proceedings Act 1970, s. 4). These rules do not form part of the general knowledge in the community. Magistrates are aware of this and in practice make allowances, either through their interpretive skills or by ignoring the rules. Where proof is concerned many Magistrates (not just those in Molepolole) accept as corroboration the man's admission of sex with the woman about the time a child might have been conceived. The time limit is simply ignored by many.

Parties are not put off by costs as there are none, except on appeal when a minimal charge of 4 Pula (about US$ 2.40) is made. There are problems of access but these tend to be of an administrative nature, relating to a huge backlog of [page 201] pending cases and the problems of getting parties together for a hearing (which also affect the Chief's kgotla).

Similarly when we turn to the Chief's kgotla we find aspects inconsistent with the perceived model of informal justice. In contrast with the rules regulating the 'formal' system of justice those operating in the 'informal' system are supposedly more flexible. However social rules are not necessarily so. For example, where a woman comes to claim compensation for pregnancy this is denied where she has had a child with more than one man. This is not a written rule like the one governing 'proof' under the Affiliation Proceedings Act in the Magistrate's court. It is viewed as customary law because it is applied by those vested with authority in the public sphere. The rule was introduced by Kgosi ('chief') Kgari Sechle II (1931‑1962) in the 1930s in order to prevent families from capitalising on their daughters' pregnancies and encouraging a form of prostitution which was considered to undermine marriage.

Unlike the rule relating to proof people know of this rule and take it into account in their negotiations outside the kgotla as the life histories of those associated with Mosotho kgotla demonstrate (Griffiths 1989). This rule is not flexible, but rigorously applied and serves to limit women's access to support from men. As the majority of adult women have a number of children with different fathers in the course of their lifetime, and the negotiations of a customary marriage are highly complex (Comaroff and Roberts 1977), one may argue that women's access to justice is as effectively curtailed by this rule as it is potentially by rules operating in the Magistrate's court of which they are unaware.

This raises the issue as to which 'people' we are referring to when we talk of 'popular' justice. Before dealing with this issue it is important to underline the fact that the Chief's kgotla operates within the context of the formal legal system of Botswana. While those at the highest level such as the Chief Regent are appointed on the basis of their affiliation with the Kwena ruling family, they are also civil servants linked into an administrative and legal structure which has been created by central government. When in 1989 certain people in the village told me that the chieftainship was likely to die out for lack of interest I asked the Commissioner of Customary Courts whether he considered that this was a possibility. He emphatically stated that such a view was "nonsense".

What prompted this response was not so much a recognition of the strength of tradition as a recognition of self interest. He pointed out that there would always be those who took the job to acquire a certain civil service grade and the benefits that went with it. He explained that while most civil servants had to pass certain examinations and have certain qualifications to be appointed and promoted, those who relied on their connection with ruling elites could skip this process and [page 202] through their appointment as Chief or Deputy Chief acquire an automatic rank within the civil servlce.

Not only are such individuals civil servants but they also administer part of the 'Common' law with which the Magistrate's court is concerned. This is principally in respect of minor criminal law offences where their remit and sentencing powers are clearly laid down by statute. One may see the Chief's kgotla as a hybrid entity fulfilling its role as a key institution in terms of Kwena social life while at the same time fulfilling its judicial role within the formal legal system. On the one hand it deals with matters seen as forming part of its indigenous base, including those falling under the rubric of 'customary law', yet on the other it operates on the basis of 'Presiding Officers', 'complainants' and 'defenders', issuing 'orders' and 'judgments'.

This raises the issue of transformation and brings us to the critique of so called 'informal' or 'popular' justice. At the beginning of the 1980s Abel (1982) and others not only challenged the representation of communitarian, consensus‑based processes, but also attacked them on the ground that they acted as vehicles for the covert extension of state control. This critique questioned the dichotomy between the 'formal' and the 'informal', denying the 'informal' an independent existence and stressing that it must be read in the context of the formal state system. In doing so it raised questions about the nature of the 'state' with regard to national, regional and local interests.

The basis on which the Chief's kgotla is claimed to be an 'indigenous' institution, a repository of Tswana values and traditions which pre‑date colonial overrule associated with the state, is open to question. Benda‑Beckmann (1979), Chanock (1985), Snyder (1981) and others have emphasised the fact that what is embraced today as 'indigenous', such as the chieftainship and customary law, is in fact an artefact fostered and constructed by colonial overlords.^{^[4]}

With regard to Bakwena, we find direct British intervention, not only in shaping the role of the Chief or Kgosi, but also in choosing the incumbent. Sebele II (1918-31), for example, was dethroned and sent into exile by the British who replaced him with his younger brother Kgari Sechele (1931‑1962) (Ramsay 1991). Older people have not forgotten that the current site of the Chief's kgotla was established in 1937 amidst enforced removals of Kwena from other parts of Molepolole to that spot by the British in order to bolster the younger Kgari's authority.

We have seen the difficulties of fitting the Chief's kgotla or the Magistrate's court into models of 'informal', 'popular' and 'formal' justice, and this becomes an even more questionable exercise when we see how closely they interact with one another on an institutional level. This is not simply due to formal prescriptions requiring appeals to go from the Chief's kgotla to the Magistrate's court or for the kgotla to apply 'common' or 'European' law in the form of criminal sanctions. The personnel themselves view their role and that of their institution as part of a larger whole, an overall system in which these institutions have their place. For example, officials from the Chief's kgotla, such as clerks and the Deputy Chief, often refer women to the Magistrate's court and vice versa.

These officials see themselves as working together. They are regularly in communication. The District Commissioner (DC), who reviews decisions given by the Chief's kgotla on an ad hoc basis, regularly talks informally on the telephone with personnel from the Chief's kgotla. This was particularly the case when Mr Kgosiensho was Deputy Chief. A woman may not successfully claim compensation for pregnancy where more than one father is involved but Mr Kgosiensho and the clerks would direct her to the Magistrate's court, aware that she could claim maintenance. While women know that they may claim maintenance at the Magistrate's court some have commented that they will not go there because, as Baidkgodisi from Mosotho kgotla observes "they will see if you have discussed the matter with your family or attempted to use the kgotla system and they will direct you to Kgosing". Such officials see themselves as part of a process which encompasses a number of institutions including the family.

In such a situation women often find their access to justice constrained, whether by 'formal' or 'informal' system. This brings me back to a point made earlier and frequently articulated in the critiques of 'popular' justice in the 1980s concerning the identity of 'popular' justice and the use of concepts such as 'community'. The work of the 1960s and '70s tended to treat these concepts as given and unproblematic, with 'community' standing for common interests, and representing a homogeneous and ungendered group.

The work of Abel and others provided a devastating critique of the earlier consensus‑based, egalitarian models of informal justice. Looking at what occurs at kgotla level in Molepolole (not just the Chief's kgotla) and, beyond that institutional base, among families themselves, it is clear all are not equal and that power, social status and social differentiation play a leading role and affect people's access to hearings. They affect the space which individuals are given to make their claim and the kind of responses that they receive. This is important when considering claims for the resuscitation of popular justice on the basis that it constructs "a cultural space - an alternative justice that is more responsive to community desires" (Merry and Milner 1993: 9). Although this claim is made in [page 204] the context of debate over whether it can ever be possible to escape from or create forms of justice that are not merely state ciphers, nonetheless it contains implications about 'community desires' which it is crucial to investigate. This is especially pertinent when examining women's access to justice, however defined.

When a woman becomes pregnant her ability and that of her family to negotiate the matter is often dependent on their status and connections within the village. Where these are non‑existent they are ignored. For example, Rameijo, the eldest son of the headman of Mosotho kgotla has a relationship with a Bakalaghadi woman with whom he has two children. She is regarded as being of low status, coming from a morafe who were formerly Tswana serfs. What this means is that she has never been able to enter the normal processes of negotiation surrounding pregnancy which involve claims for compensation or marriage.

At another level an example is provided by Ninika Bakwena whom Mr Masimega and I first encountered in 1982 having a dispute with her mother‑in‑law over exclusion from the family compound. In this case Ninika was able to get as far as local kgotla hearings but was diverted by the headman of her husband's kgotla (Mokgalo ward) every time she tried to move beyond his authority to that of the Chief's kgotla. In this case nepotism was clearly at work. Her husband Moagisi is related to the headman who is directly related to the Kgosidintsi family. The Kgosidintsis are a very powerful family in Molepolole, being second in line to the ruling Sechele family.^{^[5]} Ninika's family on the other hand come from modest origins and being poor have little status. Ninika and Moagisi married without his family's consent and when it came to dealing with her in‑laws and Moagisi she was at a loss. By 1984 she had managed finally to get a hearing before the Chief's kgotla where she and husband were advised to go and live in peace and harmony. By 1987 Moagisi had divorced her and by 1989 she was in a situation where she was dependent on whatever male partners she could find for support.

Ninika's experience was shared by Manaka Kgosidintsi. Manaka appeared in the Chief's kgotla in 1982 charging her husband Joel with neglect and seeking support. Underlying her claim for support was concern that he was involved with [page 205] another woman and was using their own family property to support this other woman. In this case the man Joel was even more closely connected to heads of the Kgosidinsti family than Moagisi and of even higher status because he was a teacher. The matter was discussed and again the rhetoric of 'living in peace and harmony' was employed. Manaka who ran a small business, a bottle store and bar, was in a better position than Ninika who had no resources to fall back on. However things did not improve for Manaka after the hearing and her husband Joel continued to appropriate property, mainly cattle, without her consent. This property was acquired from her earnings and when she complained about his behaviour he became violent towards her. When she went to the Chief's kgotla to report what was happening they simply ignored her. This went on for a number of years until by 1989 she had just given up and, driven from her home, she was trying to re‑establish her position elsewhere in the village with help from her children.

Reflecting on her experience she did not consider that she had been justly treated. She felt that "the judicial officers at Kgosing are not honest". By this she meant that when she spoke to them "face to face they would do nothing". Her explanation for this was not that she was a woman, but that her husband Joel "is almost their equal and they recognise his high rank and respect the fact that he is an educated man". In contrast another woman, Eva Makoka, who had the same grievances and experience as Manaka and whose husband was also well connected, did not attribute her treatment to personal factors, but viewed it as a general occurrence, putting it down to the fact that "they are just not interested in such cases". She knew of other women in her situation who were "just told to go and stay at home". For their part, those administering the Chief's kgotla, such as Mr Kgosiensho, and Mr Sebele, (the former Senior Chief's Representative and current Deputy Chief appointed when Mr Kgosiensho retired from office) express disinterest because of their perception of being in an impossible situation. As Mr Sebele told Manaka "there is no further action that the Chief's kgotla can take, if your husband will not come back to you. There is nothing we can do, we cannot forcibly bring him back."

Even where parties do get access to a hearing proceedings may be manipulated so that they are not given the space to make their claim and may not be heard. The most striking illustration of this for me was the Busang case. This took place in the Chief's kgotla in 1982 and concerned division of property after a divorce in the High Court. Mr Busang was well connected to the Kwena ruling family. He had been a District Commissioner in Tsabong and was close to ministerial rank, being an under secretary for the government Department of Works and Communications at the time the hearing was held. Mrs Busang's family was not nearly as well connected. In contrast to her husband she had had a minimal education and had never worked outside the domestic sphere. The inequality in [page 206] terms of social networks and status was marked.

While Mrs Busang went through the normal process in the Chief's kgotla, she was never really able to establish her claims because the third party hearing the case and Mr Busang manipulated the proceedings, cutting her short, and changing the lines of argument, until she was so brow beaten that she ceased to participate. 'Community participation' took place when a kgotla member set the tone for the discussion early on in the hearing by implying that she had been at fault. This member was the Chief Regent and although he was playing the role of an ordinary kgotla member, those present placed great weight on his contribution. At the end of the day Mrs Busang was sent away without any property because she was found to have been at fault in causing the breakdown of the marriage.

In this context the inequalities of bargaining power are not just based on individual attributes which affect the process in an ad hoc fashion, but are also structured round gender. The individuals concerned in these cases were not in the weaker or stronger position solely because of their differential access to resources conferring status, such as family connections, education and formal employment. Even where women have access to such resources, they still find themselves at a disadvantage in negotiating the property aspects of their sexual relations with men. This is because of the way in which family ownership of property is structured

Two women, Mathilda Seitshiro and Mmashaoudi Gabane provide examples of how this structuring of property relations affects women. Both were involved in long processes of dissociation from their husbands and in negotiations over property. Unlike Ninika and the other women who have been mentioned, they were in a more powerful position because of their access to resources. Mathilda was well educated, had trained as a nurse, had been overseas for further training, and was at the time of marital distribution of property, matron of the hospital in Molepolole, a highly respected position. Her husband Seitshiro did not have her degree of education or skills, and did not have the same social status. The same was true for Mmashaoudi, for different reasons. Mmashaoudi came from an influential family strongly connected with the practice of traditional medicine. This, coupled with her husband Pepere's unpopularity and exclusion from the normal male networks including his local kgotla, placed her in a more socially secure position than his.

As a result of these circumstances both women were able to press their claims and to articulate them more effectively than Ninika Bakwena and Mrs Busang. But while they were able to command a greater share of marital property, they were still unable to acquire what they felt was their fair share. There was much discussion in both hearings about 'joint' division of property and an 'equal' sharing of assets, but this was applied in the context of certain concepts of [page 207] ownership, premised on generational transfer through the male line. Women find themselves excluded from making claims on certain types of property, such as estate cattle. These are cattle which are reserved to be handed down from a father to his sons.

Women did not consider that this systematic exclusion would be unjust, given the broader family interests, if they could be sure that their children would benefit. However they were often concerned that their husbands would dissipate these assets, using them for their own interests, often to support other women and their families. This was a feature of all the cases mentioned previously. The women were excluded from control on a systematic basis to deny them access to those beasts that they were entitled to claim, because of the traditions surrounding the herding of cattle which is an exclusively male preserve. Having contributed to a herd women often found that men might sell or dispose of cattle without their consent or appropriate them by rebranding them as estate cattle. Manaka, Mathilda and Mmashaoudi complained of this and it was a common cause for grievance among those who had cattle as assets.

Given the differential power relations and the way they operate in Molepolole, and the differing perceptions of what is fair or just, it is meaningless to talk of 'popular' justice in the terms outlined earlier in this paper. The critique of the 1980s is validated. We have swung, as one commentator has put it, from the heights of optimism to the depths of pessimism (Matthews 1988: 1‑24). Is this the end of the discussion? For some scholars such as Abel (1982) and Fitzpatrick (1988) the power of the state is so pervasive that to talk in terms of 'popular' justice is meaningless. While sharing many of the views of those who have criticised concepts of informal or popular justice, I believe that those involved in cases before the Chief's kgotla have some room for manoeuvre which allows for a more complex account to be formulated. This involves taking on board the criticisms that have been made which in themselves require some refinement. Two aspects of this critique which require further discussion concern the depiction of the state and of inequality.

The early 1980s presentation of informal or popular justice as covert extensions of state control presupposes a central, monolithic entity with clear and coherent aims. Just as scholars have attempted to deconstruct such an image of law (attributed to the legal centralist paradigm) so much recent work has involved [page 208] rigorous scrutiny of what it means to talk about the state (Rose 1986; Santos 1992). What we are in fact dealing with is a plurality of agencies and institutions with contradictory agendas and interests, which interact in such complex ways that they undermine any simple notion of coherence.^{^[6]}

This complexity is illustrated in the Kweneng materials. To say that the Chief's kgotla is an agent of the state does not tell us very much. It is undeniably such, at the very least, because of its position within the formal legal system of Botswana. The Magistrate's court is also such an agent. Both are there to handle local and regional interests. In the case of the Magistrate's court the interests of central government are served by ensuring that court officials are appointed on the basis of a national system and that they are sent to work in areas where they have no linkages with local kin networks which might create conflict between national and local allegiance. In contrast it is precisely these local links which create the power base of a morafe which operates on a regional basis. Central government recognises these regional interests in so far as they do not pose a threat to the national interest. Thus 'customary law' is endorsed within the national legal system of Botswana.

The official personnel of both institutions are all civil servants. However while sharing certain features in common they also diverge in ways which relate to the ways in which they have been constituted and the sources on which they draw. The majority of civil servants acquire their positions through training, which requires the passing of certain examinations. Magistrates for example are required to have a law degree. The top personnel in the Chief's kgotla, on the other hand, acquire their position on the basis of kinship ties which they establish with the Kwena ruling family and on the basis of acknowlegment by the morafe through a process of consultation which involves all the male headmen.

In this context what occurs in the Chief's kgotla can be viewed as upholding kinship as an important aspect of social relations. Thus families may find that when it comes to dealing with pregnancy, compensation, marriage and rights to property, who their kin are has an impact on on the kind of hearing they receive. So it is not surprising to find that authority is accorded to certain kinds of [page 209] individuals such as the Chief Regent who participated in the Busang hearing as an ordinary kgotla member, while others, such as Ninika Bakwena have difficulty in makinq their voices heard.

The Chief's kgotla and Magistrate's court operate from different power bases and while they may see themselves as part of a broader scenario their interests do not necessarily coincide and may come into conflict. Such conflict is clearly visible in Mochudi, which represents the central village of the Kgatla, another Tswana morafe. Here the current 'chief', Kgosi Linchwe II is at constant loggerheads with the District Commissioner and local Magistrate. The conflict in Mochudi is political in that it involves a struggle by the Kgosi to retain his power which he maintains has been systematically undermined by central government. The political stand which Kgosi Linchwe II is adopting is rooted in the period of colonial overrule when the power of the chiefs during the later years of the Protectorate was curtailed to serve the interests of colonial administration. These controls put in place at that time were adopted at Independence and continue to be exercised by central qovernment.

Conflict also arises in Molepolole but not with the same political resonance. An example is provided by Teko, a member of Mosotho kgotla, who raised an action of compensation for pregnancy in the Chief's kgotla against her former male partner. She was supported in her claim by her grandfather Tshitoeng, an influential man within the circles of the Chief's kgotla, who often participated in hearings there as a commentator and was accorded the honorary rank of court councillor. Teko was unsuccessful in her claim^{^[7]} and lodged an appeal on the basis that the Deputy Chief had "already made his decision" without giving her a proper hearing.

The reason is simple and single. Tshitoeng Mere appears as the complainant's [Teko's] witness and also as a court member. This has never happened anywhere. One cannot testify against a defendant and later ask him questions as a kgotla member.

We see here a conflict of values, based on differing views of what constitutes justice. The Deputy Chief, Mr Kgosiensho did not agree with the District Commissioner. He discussed the matter at length over the telephone with the District Commissioner, who was a Motswana like himself. Mr Kgosiensho upheld the participation of Tshitoeng

because I considered him as the parent of the child who normally is the person to claim eight head of cattle in accordance with Tswana practice. I felt there was nothing wrong with him participating in the case and being a court councillor.

The District Commissioner felt this prejudiced the case. However Mr Kgosiensho tried to explain that this was not so, because although Tshitoeng gave evidence "a court councillor has to follow proper procedure and will do this regardless of whether it is his own child or not [that is involved]". This conflict of values arises within a state based system of justice.

This kind of analysis leads away from a consideration of justice in terms of constituencies that are defined by state and/or non‑state interests. This requires that within communities, such as those in Molepolole, a more complex approach be adopted that takes account of the multi‑stranded connections involving kin, gender, local, regional and national sites of power as they coalesce, dissolve and reconstitute themselves in particular instances.

We turn now to the issue of inequality raised by the 1980s critique in relation to the extension of state control. The image here is one of domination. Given inequality between two parties the more powerful dominates the other to such an extent that the weaker party is rendered powerless. Where such a relationship exists it is futile for the weaker party to engage in challenge.

At first glance the Molepolole materials concerning women making claims in the family sphere seems to bear this out. What is the point of trying to raise a compensation claim for pregnancy given the problems of access and its limited application? What is the point of continually bringing marital matters up for discussion when experience demonstrates that the inevitable response is to tell the parties to live in peace and harmony even when it is obvious that this state is unobtainable? Such actions have little effect in controlling men from beating their wives, consorting with other women or appropriating family property. Is it not all a waste of time and effort?

A closer look at those who are in the weaker position reveals that they need not be completely powerless and that they have some room for maneuvre, albeit in terms constructed by the more powerful. Most women find themselves excluded from making claims for compensation for pregnancy in the Chief's kgotla because they generally start having children young and continue to procreate throughout their [page 211] lifecycle in circumstances which involve different men as fathers. However for that brief moment when they have their first child they are in a more powerful position than the man who is named as the father when it comes to making a claim. The power of the kgotla is behind the woman, and her word is given precedence over that of the man if he denies paternity. The onus is on him to establish that he is not the father and he does so under pressure from the kgotla.

Even where a claim cannot stand, as in Teko's case where she had too many children involving different fathers to qualify for compensation, it may have served its purpose. In Teko's case it served to expose the man to public comment, to make their private affairs public and to mark publicly the end of their relationship. Teko could not force him to keep his promise to marry her, or to support the children, but she could make him acknowledge her existence by making his actions public knowledge. One of the reasons she brought the case was because she was angry.

He deceived me, he told me that he was going to marry me. We were together eight years and he stopped supporting me and the children.

A hearing may operate as a form of shaming a man or subjecting him to public humiliation. Manaka may not have been able to affect her husband Joel's behaviour but she was able to subject him to public humiliation, by making public, in front of all the men in the kgotla, the fact that he had been unable to prevent her from hitting him on the head with an iron bar. In addition to that, while told to go and live in peace with Manaka, he was admonished by Mr Kgosiensho in his judgment when he said:

It is very shameful for a man of his status, an educated, respectable man in his position of employment that he allows himself to be exposed to a dispute of this nature which is a sheer matter of concubinaae.

Even Mrs Busang had her moment. While unable to participate effectively in the proceedings, by initiating the hearing she was making matters public. This was commented on by kgotla members, who in expressing their disapproval of her actions, indicated that it placed Mr Busang in an awkward position, even though he had most of the support on his side. Such hearings often form part of the ongoing process of negotiation in a relationship which need not always end in dissolution. While Mrs Makoka felt her marital troubles had not been adequately handled by the Chief's kgotla, because of their refusal to go beyond the rhetoric of 'peace and harmony', she and her husband Patrick did manage to sort out their problems outside the kgotla and were by her account living together quite happily [page 212] in 1989.

The models formulated at the beginning of this paper, of 'popular' and 'state' justice, in the context of local courts in Molepolole, cannot be sustained for a number of reasons. The image of 'popular' justice in terms of the Chief's kgotla, as representing something 'other' than state justice, as more egalitarian and from which considerations of power and status were absent, cannot be upheld. The depiction of 'state' justice in terms of the Magistrate's court, as representing an inaccessible and inflexible rulebased system of justice, is also untenable. Despite differences in their constitution both these courts share characteristics which cut across the models, leaving them between paradigms. So, for example, while written rules relating to proof were flexibly interpreted in the Magistrate's court, social rules concerning pregnancy were rigidly enforced in the Chief's kgotla. The two courts cannot be set up in opposition to one another in this way and indeed both fall within the ambit of the 'state' as representing interests which are endorsed by the constitution and which fall within the national legal system of Botswana. They can be seen as representing 'state' interests at different levels.

The fact that the 'state' is always present in some form has led a number of scholars to abandon any concept of 'popular' justice. For some scholars, such as Abel (1982), it represents the dominant discourse which overpowers all others. For others, such as Fitzpatrick (1984), it may allow some form of resistance to take place, permitting a space in which other claims may have a voice, but this can never truly represent popular justice because the claims are articulated within a framework that is created by the state.

Despite the problems surrounding the concept of 'popular' justice it has been valuable in pushing us towards a redefinition of law and state. In relation to the field discussed in this paper, the classic approach to 'European' or 'Common' and 'Customary' law, represented by the Magistrate's court and Chief's kgotla, representing separate and autonomous legal systems - an approach embedded in the 'dual theory' of law put forward by legal pluralists such as Hooker (1975) - has to be abandoned. We become aware of the range of agencies and institutions which the state incorporates, through the Magistrate's court and Chief's kgotla. Such bodies have interests that overlap in some respects but which may also come into conflict or prove contradictory in others. This undermines any view of the state which presents it in terms of a monolithic structure.^{^[8]}

Such refinement of concepts forms part of an ongoing process. The insights acquired from the study of women pursuing claims against male partners in the village courts raise issues of power and gender which encourage us to move beyond the confines of legal institutions (however defined), to other bodies and agencies which construct social relations. In this case we are led to look at the structures which underpin the family in everyday life; at the conditions under which families come into being, the resources that they have at their disposal and the effect that this has on the roles that women and men adopt in relation to one another. This kind of analysis provides insights into how communities are formed and transformed, with all the implications that this entails for individual actors. To look at popular justice from this perspective would provide another set of insights to complement a courtbased perspective or dispute‑oriented perspective. This is a project that I am currently undertaking in my forthcoming book on Botswana.

1982 The Politics of Informal Justice (2 vols.). New York and London: Academic Press.

1979 Property in Social Continuity: Continuity and Change in the Maintenance of Property Relationships through Time in Minangkabau, West Sumatra. The Hague: Nijhoff.

1982 Migration in Botswana: Patterns, Causes and Consequences, Final Report, National Migration Study, Vol. 3. Gaborone: Ministry of Finance and Development Planning, Central Statistics Office.

1983 Census Administrative/Technical Report and National Statistical Tables: First Part of Population and Housing Census, 1981. Gaborone: Ministry of Finance and Development Planning, Central Statistics Office.

1985 Law, Custom and Social Order: The Colonial Experience in Malawi and Zambia. Cambridge: Cambridge Universitv Press.

1977 "Marriage and extra‑marital sexuality: the dialectics of legal change [page 214] among the Kgatla." Journal of African Law 21: 97‑123.

1988 "The rise and rise of informalism." Pp. 178-198 in R. Matthews (ed) Informal Justice. London and California: Sage.

1989 Negotiating Pregnancy: Kwena Women's Strategies Concerning Marital and Extra‑marital Status in Botswana. Edinburgh: Centre of African Studies.

1975 Legal Pluralism: An Introduction to Colonial and Neo‑Colonial Laws. Oxford: Oxford University Press.

1988 "Reassessing informal justice." Pp. l‑24 in R. Matthews (ed.) Informal Justice? London and California: Sage.

1993 The Possibility of Popular Justice: A Case Study of Community Mediation in the United States. Ann Arbor: University of Michigan Press.

1991 The Rise and Fall of the Bakwena Dynasty of South Central Sotswana 1820‑1940, Ph.D. thesis. Boston, Mass.: Boston University.

1985 "Introduction: some notes on 'African customary law'." Journal of African Law 29: 1.

1986 "Law, rights and psychiatry." In P. Miller and N. Rose (eds.), (eds) The Power of Psychiatry. Cambridge: Polity Press.

1992 "State, law and community in the world system: an introduction. Social & Legal Studies 2: 131‑142.

1937 A Handbook of Tswana Law and Custom. London: Oxford University Press for the International African Institute.

1981 Capitalism and Legal Change: An African Transformation. New York and London: Academic Press.

1991 A Little God: The Twilight of Patriarchy in a Southern African Chiefdom. Hanover and London: Wesleyan University Press.

^{^[1]} These are Maunatlala, Ratshosa, Ntloedibe, Mokgalo, Borakalalo and Kgosing wards.

^{^[2]} The term Motswana (plur. Batswana) is applied to an individual who is a Tswana speaker and lives in Botswana.

^{^[3]} This was done under s. 36 of the Customary Courts Act (04:05) which empowers the President to appoint a Customary Court of Appeal and s. 44 which gives the Minister of Local Government and Lands power to establish the rules under which such a court is to operate.

^{^[4]} There are others, such as Roberts (1985) and Wylie (1991), who deny that this represents the whole account.

^{^[5]} Kgosidintsi and Sechele were sons of Motswasele II (c.1807‑21) by different mothers. Sechele was the son of the senior wife and therefore born into the first house, while Kgosidintsi was the son of a more junior wife and born into the second house.

^{^[6]} Fitzpatrick (1984) points to these complexities by showing how state law both constitutes and is constituted by the normative orders of which it is composed. These normative orders represent a plurality of social forms which both shape and are shaped by law. However even although he accepts that the process is not all one sided and that law should not be viewed simply in terms of domination over all other social forms he still considers that it has a preeminence which derives from the state in whatever form it takes which undercuts any possibility of 'popular' or 'informal' justice (1988).

^{^[7]} For a more detailed account of this case see Griffiths (1989).

^{^[8]} It is just such a view which underlies a legal centralist model of law, where [page 213] the state is conceived of in unitary terms and where the relationship between law and state is presented in uniform and unproblematic terms.