Plurality of marriage law and marriage registration for Muslims in Indonesia: a plea for pragmatism

This article discusses the law and practice of Muslim marriages and their registration in Indonesia. The central question is to what extent these accommodate the rights and needs of poor women. A historical overview of state involvement in marriage regulation demonstrates how the women’s rights discourse has been largely replaced by one based on Islamic law, but that nonetheless women’s rights in marriage have been substantively reinforced. The next part discusses the control of Muslim marriage in practice, by focusing on the registration of marriage and divorce. It shows how informal practices at the lowest level of the state and state courts serving as a safety net protect the rights of poor women in practice much more effectively than is often assumed. These findings undermine the assumptions underlying the current emphasis by national women’s rights groups and donors on extending state registration and criminalising non-registration. This emphasis on the juridification of people’s lives by casting the issue essentially in terms of human rights and remedies may in fact even lead to the opposite of what it aims to achieve: more liberty for women to dispose of their own lives.


Introduction
The past two decades have seen an irrepressible rise of women's rights to become a prime focus of human rights activism worldwide. The Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW), adopted by the UN in 1979, has been the lynchpin in this process and its associated controlling mechanisms have influenced debates about marriage, divorce, violence against women, etc., from the international to the local level. 1 In particular the obligation of states to report their advances in implementing CEDAW has become a focal point for NGO activism. If states are negligent in their reporting duties -as many are -that fact in itself will be a target for criticism; if they do oblige, the reports become the object of debate. 2 In this manner CEDAW has lived up to its objective of promoting the position of women in society in all of its aspects. CEDAW is a typical example of what Goodale calls 'sympathetic law' -'legal theories and practices understood by concerned individuals and institutions (...) to serve humanitarian, social reformist, and counter-hegemonic functions when introduced to, and used by, groups at the margins of cultural, economic, and legal power'. 3 It seems difficult to criticise such law if one is not a neo-conservative, but nonetheless, legal anthropologists have questioned the consequences of women's rights practices related to the Convention. In particular Sally Engle Merry's work on the primacy of the human rights discourse in addressing domestic violence gives rise to concerns about the appropriateness of this framework for that purpose. 4 For the purpose of this article, the following points deserve attention. First, human rights need to be 'vernacularised', i.e. adapted to local understandings and conditions, in order to become meaningful at local levels. This process is difficult and not always successful. Second, and even more disturbing, the legal categories and approach underlying the human rights discourse, with its focus on individualism, legalism, and property, may be at odds with local notions of self and community and may lead to a legalised approach that is in a way a form of legal imperialism and may actually be harmful to the interests of the women who are supposed to benefit from it. 5 In fact, issues of women's position in society that were earlier discussed under different headings are now somehow monopolised by the human rights view. Taking this perspective into account, the present article will explore to what extent the marriage law for Muslims in Indonesia and the debates surrounding its changes have been cast in human rights terms, or been influenced by the rise of the human rights discourse regarding women's rights. It attempts to assess to what extent these outcomes have actually promoted women's rights in marriage and to what extent the problems noted by Merry have materialised. To this end, outcomes in the form of marriage and divorce registration practices related to this law are considered as well. Since the result of the lawmaking process usually does not please all sides, it carries over into local disputes with registration as a main focus. In fact, the trajectory of national lawmaking processes concerning marriage and its registration are likely to predict to some extent the degree and nature of these local struggles. Marriage serves well for this purpose, as it has been a contested issue in Indonesia for many years, with widely differential views between state, religious and social authorities as well as social groups.
A state strongly committed to nation-building and promoting development, Indonesia has followed an agenda of modernisation since its independence in 1945 and has reinforced it by adopting a comprehensive catalogue of human rights in the Constitution in 2000. It also has a vibrant civil society relentlessly promoting them. At the same time, conservative Islam has gained influence on state policies and laws. 6 This may not be surprising for the largest Muslim majority country in the world (86% of the 240 million inhabitants are registered as Muslim), 7 but Indonesia is also home to many other religious convictions whose views on marriage and the family need to be taken into account if social stability is to be guarded. The same diversity exists with regard to social groups, which range from outspoken feminist ones to fundamentalist Muslims and from Jakarta's jeunesse dorée to communities that still by and large follow their own adat or custom. In short, it is hardly possible to come up with one law governing in a uniform manner all aspects of marriage and that is applied across the whole of the population. Instead, the current system is still one of various forms of legal pluralism, which continues to be the object of political struggle.
The article will start with a historical overview of the contest about Muslim marriage regulation in Indonesia, which demonstrates the differences in ideas between various groups and authorities about what constitutes a proper marriage, as well as the repertories of arguments mobilised in support of them -human rights being one of them.
We then move our focus to the local level, to analyse registration practices, in particular when and why people register their marriages and how the state authorities concerned, including the Islamic courts, respond to non-registration.

Indonesian marriage and marriage registration law
Before 1974 the Indonesian population was subject to a variety of marriage regulations inherited from the colonial state. In its typically pragmatic manner the colonial Government never attempted to bring all citizens under one statute, but instead only intervened in family matters if this was required by external pressures -for instance, from the church in the Netherlands which wanted a special regulation for their fellow Christians in the Indies. From the 20 th century onwards the administration's moral convictions played a more important role and led to some limited intervention. The reasons were basically the same as those being currently cited by the Indonesian Government: preventing child marriage, controlling polygamy, countering the birth of illegitimate children, and protecting women against unfair divorce proceedings. 8 Unification never took place, however, and the population remained subject to differing legal regimes.
The first was that of Book I of the Civil Code that applied to the European part of the population (including the Japanese) and those who had been equated with the Europeans in certain civil matters -mainly the Chinese. The second was the Ordinance for Christian Indonesians (S. 1933 no. 74), for whom special records were kept by the colonial state. The third and most important one was the regime for the Indonesians: either adat (customary) or Islamic law applied to those Indonesians and non-Chinese Foreign Orientals who were not equated with the European group. This concerned more than 90% of the population of the Dutch East Indies. Finally, there was a special regulation on 'mixed marriages', for those who were subject to different regimes but still wanted to marry. 9 The three regimes above ranged from implying fairly extended Government control to hardly any. In particular the validity of Islamic and (unwritten) adat law for non-Christian Indonesians allowed the Government only very limited intervention in marriage practices. Outside of Java these were highly diverse, corresponding to the bewildering variety in family and clan systems. On most of Java marriage was ruled by the teachings of the Shafi School of Islamic law. 10 Marriage registration for non-Christian Indonesians was only made obligatory at the end of the 1920s and early 1930s. It was intended as a tool to supervise whether the substantive Islamic law or adat law was properly applied. 11 Non-registration was a criminal offence, but the marriage was left intact and the punishment was a rather insignificant fine. 12 This system of substantive rules was complemented by institutions that exercised some control. The most important were the Islamic courts, which had jurisdiction over all matters concerning marriage and divorce, but were linked to the general state court system for Indonesians by the need for an exequatur of their judgments. 13 On the outer islands adat courts played a similar role, with the difference being that their judgments were subject to appeal to the state court system for Indonesians.
At the grassroots level practices were highly diverse, 14 but if we confine ourselves to Java, certain uniformities can be discerned. Marrying very young was common, as was getting a divorce and remarrying. Generally speaking, women were relatively free to select their spouses and could obtain a divorce rather easily. 15 One can say that under Islamic law Javanese women were rather well off, which was mainly due to the widespread use of marriage contracts. These provided protection against whimsical unilateral repudiation and financial abuse. 16 Moreover, under Shafi Islamic law the provisions on child custody generally favour women until the children are considered to be adults, which is around their twelfth birthday. In short, the protection of women's interests was not a matter of state regulation, as indeed in practice registration by local village heads or religious persons seldom occurred. 17 After Independence some procedural changes were made with regard to the system to reduce pluralism, yet substantive pluralism remained unaltered. The three colonial huwelijksordinanties were replaced first on Java and Madura by Law 22/1946 on the registration of marriage, divorce and reconciliation (for Muslims) which after the establishment of the unitary republic in 1951 was made applicable to all Muslims in Indonesia by Law 32/1954. Another significant formal change taking place during and immediately after the Revolution (1945)(1946)(1947)(1948)(1949) was the abolition of adat courts, which led to a situation where marriage disputes were immediately subject to the newly unified state court system. In practice, however, many adat courts remained in place and continued to function as forums for 'mediation'. 18 Yet, with regard to substantive family law almost no unification attempts were made.
This lack of intervention is not surprising. While the newly established State of Indonesia immediately started to arduously pursue the twin goals of nation-building and economic growth, the quest for modernisation did not immediately extend to marriage practices. This certainly had to do with the fact that anything related to religion and deeply engrained social practices must be treaded carefully by a power which is not fully in control -and this certainly applied to the Indonesian Government. After the breakdown of democracy in 1959, moreover, President Soekarno promoted a policy of population growth, which would be hindered by marriage regulation rather than the other way round.
A major change therefore only followed under Soeharto's New Order regime . From the perspective of the Soeharto Government, introducing a unified Marriage Law served several purposes at the same time. First, the central concern of the New Order Government was to promote economic development. Demographic policies with the objective of reducing the population growth were a central element and controlling marriage -and in particular marriage age -was one of the vehicles to achieve it. 19 This coincided with the second purpose, which was nation-building by a policy of cultural unification. The new ideology projected a model of the happy, 'modern' family (two parents with two children) as the nation's foundation. 20 Replacing pluralistic colonial regulation by modern, unitary legislation also became a core concern of the Government. A new Marriage Law would contribute to both objectives. Ironically, the New Order Government thus continued the capitalist project supported by Dutch conservatives before independence, instead of the humanitarianoriented, more culturally-relativist oriented project of diversity promoted by Van Vollenhoven and his pupils. 21 Finally, feminist groups had long been pushing for a law that would better protect women's rights in marriage. 22 Now that their goal coincided with the new Government's policies, they seized the opportunity to start an effective lobby, probably helped by the fact that Soeharto's wife -Ibu Tien -was sympathetic to their quest. 23 The draft law was one with a 'general' secular form as found in most countries. It made divorce before the general court obligatory. Jurisdiction over Muslim marital affairs was for the largest part transferred from the Islamic courts to the general courts, which in practice would mean that the powers of Islamic courts were abolished. With regard to substantive law marriage became primarily a monogamous state affair, with state registration as a core condition. Polygamy was made conditional and required permission from the general court. Women and men held equal rights to initiate a divorce and women and children held rights to alimony. The bill was presented to Parliament on 21 July 1973 and given Soeharto's firm control on the legislative process, a smooth enactment seemed likely.
However, it soon became apparent that the Government had underestimated the sensitivity of the subject. Pluralism appeared to have deeper roots than the Government had assumed, but resistance against the New Order's project of state formation was also fuelled by another concern. According to Bowen, 24 the symbolism of the loss of control over the last stronghold of Islamic law, family law, was at least as important as the substance of the law. If state courts should decide upon Islamic family law issues at all, then it had better be Islamic courts, by judges whose authority was founded at least in part on the training they had received in Islamic law. Resistance from Muslim organisations was unexpectedly strong and in the end it culminated in a 'walk out' from Parliament by the Muslim party -an action unheard of at the time. 25 As we will see, the importance of the symbolism of being able to control the sphere of intimate relations between persons has influenced marriage regulation politics until today and at every level of the state.
In response to the 'walk out', the Government decided to review the draft and remove its most controversial provisions. Thus, the Islamic courts retained jurisdiction in all Muslim marital matters, polygamy would still be allowed -albeit under strict conditions and with Islamic court permission only 26 -and through Article 2 the Law abandoned the idea of a separate civil marriage by requiring a religious ceremony for a marriage to be valid. The same article did state that a marriage needed to be registered, but it was beyond doubt that the religious nature of marriage was acknowledged as its core. The provision that divorce required court permission was retained, yet for Muslims the Islamic court would be competent. 27 At the end of the process, with polygamy and divorce having become judicial matters, the new marriage law did not mean a decrease, but an increase of powers for the Islamic courts and thus somewhat accommodated the concerns of Islamic political groups resisting secularism. That in this manner pluralism in state law had not disappeared is clear. Muslim marriages are moreover registered at another office (the Office of Religious Affairs, (KUA)) than marriages based on other religions. 28 Many of the law's provisions, like the ones on polygamy, apply to Muslims only and significant bodies of substantive rules have remained in place. Thus, to Christians important sections of the Civil Code still apply, for Muslims Islamic law and adat law are still valid and Hindus and Buddhists have remained subject to their own bodies of religious or adat law. If anything, this last part shows that even in its original form the Marriage Law would not have abolished all pluralism, as such bodies of substantive rules were projected to remain in place anyway. The Government was realistic enough to understand that in order for the new law to have any legitimacy it needed to be able to accommodate, to a certain degree, the enormous diversity in marriage practices and the normative notions underlying them. However, that the state could not establish its primacy in concluding marriages was a serious setback for its ideology of modernity.
In order to avoid such open resistance in the future, the New Order Government decided to take a more inclusive approach to reforming marriage. It now involved Islamic law scholars into the process of law reform, thus preventing controversies from spilling into the open and accommodating Islamic scholarly views earlier in the lawmaking process. This led to women's rights groups experiencing a loss of influence on the lawmaking process, in particular in the sense that much of the debate was no longer waged in the women's rights terms of the modernity project, but in those of Islamic law. This remained unaltered by Indonesia's ratification of CEDAW in 1984, while the increasing authoritarianism of the New Order worked against NGOs generally and their human rights discourse in particular. 29 Still, the next stepping-stone in marriage reform, the Law on the Islamic court (no. 7/1989), further improved women's rights. This law provides a special procedure which is unique for the Islamic court, 30 stipulating that a divorce request or divorce suit must in principle be filed at the Islamic court where the wife resides, regardless of whether it is the husband or the wife who initiates the divorce. 31 This provision considerably facilitates access for women to the Islamic court in marriage and divorce matters. It shows a pattern that has dominated new marriage regulation for a considerable amount of time: cast in patriarchal Islamic law terms, legal reform has been 'inching towards equality' between men and women. 32 Unlike some other parts of the law, this provision hardly received public attention, the debates having been waged almost completely internally within the Government and with its Islamic advisors.
The reform of substantive marriage law continued in this manner in 1991, when the Government introduced the so-called Compilation of Islamic Law. This is a restatement of Islamic family law, as formulated by prominent Islamic law scholars in a process driven by the Ministry of Religion. The Compilation elaborates and complements the Marriage Law and derives its validity not only from the state, but also from Islamic law doctrine itself, as it can be seen as being based on ijma -the consensus among (Indonesian) Islamic scholars that is generally recognised as a source of law. By translating it into the discourse of Islamic law itself, the Compilation's authority among more conservative Muslims has been considerably reinforced. 33 Together with the Marriage Law, the Compilation is the main statute applied by the Islamic courts and although its language is still pervaded with patriarchy (just as the language of the Marriage Law, and -for that matter -the Civil Code) it is a remarkably liberal statute, amongst other things by granting women an almost identical legal position as men in initiating divorce proceedings.
A fourth central law which has a bearing on marriage, and which was passed almost unnoticed and hardly debated in public, is the post-New Order Registration Law no. 23 of 2006, which deals with all major forms of civil registration in Indonesia. The Law does not constitute a break with New Order marriage policies. It stipulates the obligation for citizens to report their marriage to the Islamic or the civil registry, and threatens non-registration of a marriage by a maximum fine of Rp 1 million (approximately € 75), which more or less equals two months' wages for the poor. The law has a typical procedural quality, which probably explains the relative disinterest among the wider public concerning its enactment. Thus, the Registration Law grants no authority to the registrar to evaluate whether the substantive provisions of the Marriage Law have been fulfilled. The married couple should go to the registry and only need to bring a report by the person who has concluded the marriage. The check as to whether a marriage is lawful is thus in the hands of the person in charge of the religious ceremony. 34 Article 35 and its elucidation provide for the registration of a court decision on mixed marriages in a likewise manner. The message is clear: registration offices only register and have to defer to the judgment of religious authorities and courts.
These four laws -the Marriage Law, the Law on the Islamic Court, the Compilation of Islamic Law and the Registration Law -are all legal reflections of the continuous struggles about control over marriage norms and practices. As elsewhere in countries with a Muslim majority 35 the arguments in these struggles are presented in a way so as to gain Islamic legitimacy, 36 which is politically more expedient than casting them in human rights terms. From a legal point of view the argument can simply be based on human rights principles like those included in CEDAW, but CEDAW and its terminology seem to have little currency. The processes leading to the four laws mentioned above show that the Indonesian Government cannot ignore traditional conceptions of Islamic law, 37 and have thus to some extent sidelined groups promoting women's rights in a 'traditional' feminist way or in human rights terms. As indicated in the introduction, the issue does not stop here. The controversies about what constitutes a proper marriage have not been resolved by this set of laws and they have continued to surface in local administrative practices and court decisions. They all have to do with marriage registration, but centre on different problems. The first of these problems is religiously mixed marriages. 39 Until 1989 the colonial mixed marriage regulation of 1899 was used as a basis for concluding such marriages at the civil registry, but according to a Supreme Court ruling of that year this statute would no longer be applicable. As a consequence, marriage candidates of different religious convictions have since had serious difficulties in finding a proper forum to be married and a civil registry that is willing to register the marriage. 40 , 41 The second issue has been the recognition of religions. Although Indonesia's Constitution guarantees freedom of religion, since 1966 a practice has developed on the basis of Presidential decree 1/1965 to only recognise six of them (Islam, Catholicism, Protestantism, Hinduism, Buddhism and more recently Confucianism/Taoism). 42 Those most afflicted by this policy are 'mystic' groups, whose members are generally considered to be Muslims, but who refuse to have their marriages registered by the registry for Muslims. Several cases have been taken to the administrative court by prospective marriage partners in order to gain recognition as a separate religion, but results have been mixed. 43 The third issue is less focused and affects a much larger number of people: it concerns the actual implementation of the rules of the Marriage Law and the Islamic Law Compilation on Indonesia's Muslim population by means of registration of marriage and divorce. As we will discuss in the next part of this article, many divorces and marriages remain unregistered, with 183 potentially disruptive effects on the protection of women. 44 However, a variety of practices and strategies by spouses, local government institutions and religious courts, based on norms of differing origins, has led to surprising outcomes. They will show that in order to guarantee a fair result concerning divorce for poor women, a principled approach based on a human rights discourse and focusing on state control may not be necessary and may perhaps even be counterproductive.

Marriage, divorce and registration
In 1978 Katz & Katz wrote that: 'Almost four years after its promulgation and two years after its implementation, the new Indonesian marriage law has succeeded in dramatically affecting Indonesian society -a success which was not anticipated by respected Indonesian legal experts.' 45 Basing themselves on figures from the Department of Religion and interviews the authors claimed that the divorce rate had decreased by approximately 70%, that the procedure before the Islamic courts had reduced polygamy and that child marriage was now 'sufficiently rare to be reported by the press as an oddity'. 46 To all those involved in legal implementation studies this will sound too good to be true, and indeed it is. It took a number of years before thorough research into this matter was carried out, but the findings where quite different from those of Katz & Katz. First, Cammack, Young and Heaton 47 reported on the basis of a demographic survey that the number of underage marriages -in violation of the law -had continued to decline steadily over the years, but still stood at more than 20% of the total. They perceived no impact of the Marriage Law on the downward trend recorded. While some of these underage marriages were actually registered by 'sympathetic' registrars, most were not registered at all.
A later study by Cammack, Donovan and Heaton looked at divorce, combining the results of a large family survey across 13 provinces with the official figures on divorce. It found that more than 50% of divorces remain unreported, 48 something Katz & Katz did not take into account as they based their claims on reported divorces only. 49 The authors did recognise a downward trend in divorce, but just as in the case of the decrease in marriage age, the reasons seem not to be in the Marriage law. Research in Cianjur, a district in West Java, indicates an even higher number of unregistered divorces. In 2009 Van Huis interviewed 120 divorced women and found that more than 75% of the divorces remain unregistered. 50

184
Although national studies on illegitimate polygamy are not available, the results may be even more striking. 51 For the entire district of Cianjur Van Huis found that in the years 2006-2008 only eight persons officially obtained court permission for a polygamous marriage, whilst 42 of the 120 interviewed women (35% [!]) stated that the polygamy of their husband was the main reason for divorce. This figure contrasts sharply with the eight persons who obtained court permission to enter into a polygamous marriage, in particular if we take into account that Cianjur has a population of 2.1 million persons. 52 The obvious question is why spouses do not register their divorces and marriages more often. Most studies exploring this subject in Indonesia have been made by donor organisations and claim that registration would be beneficial to divorced women. For instance, a report on a female-headed households empowerment project sponsored by AusAID in Cianjur, West Java said that: 'The legality of marriage and divorce affects the inheritance rights of children, as well as the legal responsibility for the financial care of former spouses and any children of the marriage. Divorce, like the death of a breadwinner, is thus often a tipping point for families living on the edge of the poverty line.' 53 The account then continues to focus on issues of costs, distance, knowledge, etc. to explain why these poor people do not register their marriage.
Women's rights NGOs and the National Commission on the Elimination of Discrimination and Violence Against Women (Komnas Perempuan) generally take the same view and have advocated the further criminalisation of the perpetrators of unregistered marriages. They argue that women become victims of unregistered marriages as they are unable to bring their husbands to court if the latter violate their rights, including post-divorce rights and the right to be free from domestic violence. 54,55 Another often heard complaint in women's rights circles is that tolerance towards unregistered marriages facilitates illegitimate polygamy by the husband, as it is easier to keep it secret from the official wife.
While they may sound convincing, these arguments are not self-evident. Recent research carried out by Van Huis 56 and Parikesit 57 in the same region as AusAID (Cianjur) shows that the situation poor women face is usually different than assumed and therefore the registration and rights-based approach which these international and national organisations advocate may be of little use.
We will first deal with divorce. The main point is that the rights and registration-based argument fails to take into account that the main reason for divorce is an economic one: the husband does not contribute much to the income of the household, or even the opposite. Hence, alimony and inheritance rights hardly play a role for the poor and neither do rights to maintenance, or even marital property -of which there is usually almost none. Most poor women 58 Jones, supra note 15. 59 Van Huis forthcoming, supra note 50. 60 Nurmila, supra note 51. 61 Suherdjoko, 'Plan to ban "unofficial marriages" gains support', The Jakarta Post , The Archipelago, 18 February 2010. 185 therefore feel that they have little to gain from an officially recognised divorce and do not resort to the Islamic court. A majority of women can cope with the post-divorce situation very well, as is reflected by the finding of Van Huis that 75% of the 120 divorced mothers surveyed stated that their financial position had not become worse after divorce. Some 25% even argued that they were financially better off. Support comes from their parents or other family members in the form of housing and childcare. Around 50% of the divorced mothers are working. Moreover, as mentioned before, in this part of Indonesia remarriage is common and without any stigma being attached 58 and thus constitutes another safety net. In short, for many women post-divorce financial rights are simply irrelevant.
Still, 25% answered that their financial situation had become worse after the divorce, so this seems to support the argument by AusAID. However, it is unlikely that the court can do anything for these women. Women who do apply for a divorce at the Islamic court seldom request alimony -overall only about five percent. Usually the court does grant alimony if requested -at least for any children resulting from the marriage -but this established monthly amount of alimony is seldom paid and the judgment is not enforced or is even thought to be unenforceable. 59 So court involvement seldom leads to the result desired.
The second reason for women to initiate a divorce is unlawful polygamy by their husband. As mentioned above, Van Huis found in Cianjur that for 35% of the self-reported divorced women one of the main reasons to divorce was that their husband was polygamous. However, considering the low number of official polygamy permission cases at the Islamic court of Cianjur, the bulk of those polygamous relations must have remained unregistered, since state-recognised polygamous marriages require prior court permission. Although more research is necessary, it is likely that many of these polygamous marital relationships resemble extramarital affairs. Nurmila gives some colourful insights into polygamous marriage practices. 60 To generalise, the first wife is betrayed, whilst the second wife in most cases knows about the existence of the first wife. When the first wife finds out about the second wife, she will typically be forced to take a decision whether to accept or -through divorce -to escape a polygamous marriage. The unregistered married persons, the second wives, are thus not the main victims here, since they are aware of the consequences of their action. The aggrieved are the first wives, who have often registered their marriage, especially if it concerns their first marriage (see below). As an Indonesian gender and human rights expert put it: 'Many men have misused and taken advantage of nikah siri [unregistered marriage]. They have legalized adultery on the pretext of religious law. They have deceived God '. 61 In fact, what we see is a largely semantic issue: what is discussed in terms of polygamy rather resembles adultery and a lack of fidelity. Polygamy serves as a way to legitimise this behaviour in religious terms. In other words, the grievances of first wives in unregistered polygamy cases do not concern the registration issue, but the adulterous behaviour of the husband. It seems rather naive to expect that criminalising those who enter into unregistered marriages will end the practice of men (and women) being adulterous. Prohibiting unregistered marriages will probably decrease the possibility for men (and second wives) to legitimise polygamous marriages socially and religiously and force them to be more secretive -although religious legitimacy based on the fact that polygamy is allowed in Islam will probably remain strong in conservative circles. However, it seems to be a fact of life that many wives will keep becoming victim of their husbands' adulterous behaviour, no registration law can change that. 62 Domestic violence presents a more complex case. Law 23/2004 makes domestic violence a criminal offence, but the effects of this law remain limited. The little research available indicates that although the number of cases is rising, women are still reluctant to bring criminal charges against their husband 63 and prefer to escape violent marriages via a court divorce. 64 Women in unregistered marriages are disadvantaged as they cannot file criminal charges on the basis of Law 23/2004. Whether women in unregistered marriages -who cannot divorce via a court decision -have more difficulties in seeking a socially acceptable divorce from their violent husbands is not known and deserves further research.
Having discussed the registration of divorce, we will now turn to registering marriages. Perhaps remarkably, marriages are far more often registered than divorce, especially first marriages. Muslim marriage registration takes place at the Office of Religious Affairs at the municipality level, and compared to divorce is relatively cheap and easy. In Cianjur Van Huis 65 has found that more than 80% (103 of 120) of the divorced women interviewed had registered their first marriage. However, only a small number (17 of 103) of them had gone to court to divorce. According to the Marriage Law these women simply cannot register a remarriage as long as their divorce from their first husband is not processed by the court. This indicates the social practice, as is also indicated in a World Bank report, 66 that people often register their first marriage, but fail to go through the somewhat cumbersome and relatively expensive process of a court divorce and thus according to state law are still married to the spouse from whom they divorced in a socially accepted way.
Practical interests are also involved for many poor people in the decision not to register marriages. These people mostly live in faraway places where they do not need a marriage certificate to obtain services. As one respondent mentioned to Parikesit, 67 in urban areas one needs such certificates to get a loan or even to stay in a hotel, but in the area where she lived this was not an issue. If one would need documentation or proof of possessions, the village head would supply it. The role of lower administration is often vital. For example, in order to be eligible for a state food and fuel support programme people do not need to have their divorce registered, or their marriage registered, since the legal proof required is mainly an identity card (KTP) and a document by the vicinity head or village head explaining the situation of the 'poor' family concerned. 68 An important reason for people to register their marriage has to do with the birth certificate for any offspring. Not in the substantive sense, as one can obtain a birth certificate anyway and neither does the law attach any negative consequences to being the child of a single parent. 69 Note that this registry is recognised by the state and performs tasks of state registration. 70 An important point is that the registers in various offices are not connected. This means that one can conclude many marriages and register them all over Indonesia without the KUA knowing this.

187
However, here we see for the first time that social norms follow state law, in the sense that women do not want to register their child as being one from unmarried parents. However, apart from this issue, state recognition of a marriage is considered unimportant from a social perspective. Living together outside wedlock is a serious taboo in most of Java and Indonesia in general, so marriage is the social norm, but whether a marriage is valid depends on whether it has been concluded in conformity with the Islamic law as perceived locally -and this does not need to involve the state. This takes us to the crucial role of the lowest levels of state administration. These officials often apply the rules in a way that is beneficial to those seeking documentation, but one that actually goes against official state policies. We have already mentioned the village head who would provide documentation to a person who did not fully qualify. There are even officials who are still closer to the 'grassroots' -for instance the vicinity head (Kepala RW) who in his turn may serve as an intermediary to the village head in order to explain why the particular conditions of a case make condoning an attractive option to obtain a just outcome. While this is often not only a matter of compassion but involves some petty corruption as well, the price of seeking certification at higher levels is much higher. Moreover, the local knowledge of the officials will also set limits to the favours obtained. A similar role can be played by the Office of Islamic Affairs (KUA). 69 In fact, in some remote areas the KUA has largely taken over the role of the Islamic court in processing divorce, by registering second marriages without an official divorce having taken place. Although this is against the law, it does ensure that the data at the registry are in conformity with the de facto situation and that a second marriage can at least be registered. The advantage for locals, again, is more flexibility and lower costs. Ironically, the Islamic registry has its representatives in outlying hamlets and within these hamlets the representative may register second marriages outside the purview of the KUA. From the citizen's perspective the only concern is that this registration is recognised by the state agency that provides the service he or she requires, for instance that it is accepted by the civil registry which provides birth certificates. 70 The main question remains of course to what extent this system produces certainty for those who for one reason or another need formal documentation to prove their legal status. While the involvement of grassroots officials in marriage and divorce registration offers some guarantees, one can imagine that it is not enough if people get into the higher echelons of the state to obtain state services. For instance, the Islamic court may refuse documentation provided by the Islamic registry that a person has remarried, if it appears that a previous divorce was not processed by the court.
It is at this point that the Islamic courts have demonstrated the kind of pragmatism necessary to deal with such a situation in the interest of the weakest party instead of sticking to their doctrinal guns. We are referring here to the institution of isbat nikah, which allows people -in practice always women -to have their marriage recognised retroactively by the court if they have failed to register it and if they are not in a situation to follow the 'normal' road to do so.
Isbath nikah is based on the Marriage Law (Article 7) and was meant to be a transitional article for those who had not yet registered their marriage according to the Marriage Law at the time of its enactment. However, the wording of the provision is ambiguous and has been interpreted so as to also apply to the retroactive registration of marriages contracted after 1974.

188
At least in West Java, but most likely in other parts of Indonesia as well, the article is often applied in this way. 71 Thus, the use of isbath nikah allows non-registration to be rectified. For instance, in Cianjur Parikesit found that in several cases in which a woman wanted to divorce her husband before the court, the latter discovered that the marriage certificate had been forged by the Islamic registry and therefore would not accept it as legal proof. Typically, in such cases a woman concerned will be advised by the Islamic court that in order to divorce her husband according to the state rules -that is before the court -she has to drop the case first and then use the isbath nikah option to have the marriage recognised. Only in this way can she divorce her husband officially and have her post-divorce rights recognised. Obviously, only those people for whom considerable material gain is at stake will go through such a cumbersome process, for instance when there is substantial joint marital property.
This pragmatic interpretation of the transitional isbath nikah article by judges of the Islamic courts also allows divorced mothers to obtain an official birth certificate for a child who would otherwise be deemed to have been born outside wedlock. The court will ask for evidence that all the ritual requirements rendering a marriage valid according to Islamic law must have been fulfilled. This evidence is not difficult to produce as statements by two witnesses -typically one family member and one more neutral witness such as the head of a hamlet -will be accepted by the court as sufficient.
There are several advantages to this situation from the perspective of the women involved. The protection it offers to children and single mothers is clear. Furthermore, isbath nikah is invoked successfully by widows who seek recognition of their rights to the pension of their deceased husbands. A less obvious outcome is that the possibility of an unregistered 'religious' marriage and isbath nikah allows adolescents in rural areas more sexual liberty than they would otherwise be likely to enjoy. 72,73 Parikesit relates how young couples -if caught -will give in to societal pressure to marry unofficially with the advantage that they may split up again without much ado. If a pregnancy follows, the marriage can be registered officially and thus become valid in the eyes of the state. However, if the man runs off and leaves his former 'wife' alone, she may apply for isbat nikah at the court and get a divorce, but still have her child recognised as legitimate. In this way, shameful extramarital pregnancy, often resulting in risky illegal abortions and in a considerable amount of cases even in the eventual death of the mother, 74 can be avoided.
The downside of this practice may be that it becomes more difficult for the state to conduct its anti-polygamy and anti-child marriage policies, since the need to officially register marriages and process divorces is not obvious from the start. Nurlaelawati in her research on the implementation of the provisions of the Compilation of Islamic Law by the Islamic courts clearly blames the lack of obedience by Indonesian society with regard to state rules on registration for this pragmatism by the Islamic courts. 75 While it seems that at least in Cianjur both polygamy and child marriage are common, 76 it seems highly unlikely that a more strict approach, as proposed by women's rights activists, 77 donors like AusAid 78 and the World Bank, 79 but also by a radical Islamist like Abu Bakar Ba'asyir who want to counter un-Islamic marriages, 80 will remedy this problem. Rather on the contrary, at the end of the day pragmatism allows the state to exercise more control. By means of retroactive registration it can better guarantee the rights of women and children than with formalist policies that deny and criminalise the normative plurality found in Indonesian practice.
In the criminalising discourses of many Indonesian women NGOs, typically all women engaged in unregistered marriages are depicted as passive, powerless victims of their husbands. This is a distorted image of the actual practice on the ground. Women who do not register their marriage are mainly women in second marriages who did not go to court to divorce and as a consequence have no divorce papers. 81 They are not 'tricked' by their husbands into an unregistered marriage, but cannot register their marriage as long as they do not take the initiative to divorce at the Islamic court. This is mainly a matter of choice, not a matter of a lack of access. This argument is supported by Buttenheim and Nobles, 82 who have found that the freedom for women in marriage decisions increases significantly in subsequent marriages after a first marriage has failed. Of course, there are also women who are pressured, betrayed or deceived by their husbands in entering into an unregistered marriage, but most data indicate that most women enter into unregistered marriages because they have freely chosen to close the registration option by not divorcing at the Islamic court. And women do not divorce at the court because there is little to gain from it. 83 Furthermore, one must consider that an opposite image of present-day Muslim women in Indonesia also exists. In contrast to the view of women as passive victims in divorce matters presented by donors and women's NGOs, in this image it is men and the family who are the victims of women's growing 'individualism': 'Men, and sometimes women, at seminars and public gender education programs or forums expressed fear of women's no longer needing men. Women's not needing men was a very real and growing concern towards the end of Suharto's rule, and remains so in present-day Java. It related to rapid social and economic shifts and increasing awareness of being part of a global economic system that is bringing new material goods and ideas to Indonesia, alongside new educational and work opportunities for women. (…) concern about women's potential new roles were thus tied to fears of western influence and the potential for a subsequent breakdown of the family.' 84 Women thus find themselves somehow squeezed between two discourses: one denying them agency and branding them as victims of the state's incapacity to regulate their lives as wives and mothers, the other ascribing them too much agency and thereby uprooting Indonesian tradition.
These two discourses seem mutually incompatible, but ironically policies based on the first may reduce the fears of the latter. Strict enforcement may easily lead to unregistered marriage becoming further stigmatised and the 'individualism' of women being once again curtailed -especially in non-urban areas. Even if a more strict enforcement would lead to more registration, this in its turn will probably lead to more restrictions on female movement in society, because the fears of unwanted pregnancies leading to children born outside marriage is likely to lead to more surveillance by parents and siblings.
Most Muslim women's rights activists in Indonesia seem unaware of this danger. But maybe what Budiman 85 states about the good intentions of the Western feminists towards their 'developing world' sisters, to some extent also applies to middle-class and upper-class Muslim women's rights activists' view of their rural 'sisters': 'There is (…) a tendency among Western feminists to view Islam as "hopelessly misogynist", and Muslim women as "oppressed" and, therefore, needing to be rescued by their Western "sisters"'. As we have seen in the first part of this article, the use of a women's rights discourse is not indispensable to promote women's rights, rather on the contrary. The vernacularisation of women's rights in Indonesia must accommodate Islamic concerns and fears in order to be effective. The second part of this article concerns the practical equivalent. The further 'juridification' of marriage and divorce may lead to the opposite of what Muslim feminists would like to see -less freedom and less 'individualism' for rural Muslims.

Conclusion
Just as many other countries in the world, Indonesia has to deal with a situation of normative or legal pluralism regarding marriages. This has resulted in unresolved debates about the proper way to regulate marriage and the position of women in particular. Since 1975 these debates have been waged in Islamic terms rather than in human rights terms, but this has not prevented legislation from offering increasing protection to women. It would be going too far to say that 'vernacularising' women's rights in Indonesia was meant to translate them into Islamic law terms, but progressive Muslim women do draw their inspiration from the women's rights discourse while pleading their cause in Islamic legal terms.
As a result, the potential disadvantages from the human rights approach as coined by Merry and others seem to be absent. The Marriage Law does not mean a clear break with the past, by introducing a secular, same rights type of marriage and divorce. The practices developed on the basis of this law and the typically pragmatic way in which religious courts and officials close to the grassroots have responded to local situations have further reinforced a practice that to a large extent accommodates women's rights. This applies at least if one focuses on outcomes instead of on the patriarchal terms and the nominal differentials between men and women used in the Marriage Law and the Compilation of Islamic Law.
Still, the registration of marriages and the processing of divorce have become focal points of contestation. A remarkable finding is that such protest comes from unusual bedfellows: women's rights organisations 86 and some donors aligned with radical Muslims who want to address the licentious practices condoned under the unregistered marriage regime. This produces a difficult dilemma for the state. By following this approach it runs the risk of giving up control over such marriages altogether, in particular if it prohibits isbath nikah by the religious courts as a way to legalise unregistered marriages. Those who want to draw the benefit from state protection will have to address an institution that is expensive in terms of time and money, while there are no clear advantages to be obtained from doing so in the short run. This is not acknowledged by the upper-middle-class women who represent women's rights at the national and international level, and who find themselves in entirely different circumstances than the poor women whose concerns they assume to voice.
At the same time, it is difficult for the state to resist this pressure, since the road suggested squares with Indonesia's still prevailing quest for modernity. In this sense the government is caught in the same logic as women's organisations and donors -in a way similar to how many are caught in the human rights paradigm as criticised by Merry and which also requires the further juridification of social life. The result will be that social practices which are not in line with state rule may go completely underground and hence evade any state control -except perhaps for highly repressive and expensive forms of it. As a result freedom of movement for women could become the main victim of such policies, as the socially acceptable safety net of extramarital relations, that is an uncomplicated unregistered marriage, will disappear. This is why we argue that in fields with strong normative pluralism, such as family law, a pragmatic approach will work better to promote human rights values than simply imposing a human rights framework.