The Indian Ocean and Islamisation

Although separated by considerable distances, all these regions are situated along the Indian Ocean rim and were involved in long-term historical relations linking the Indian Ocean world with their respective mainland.

The acceptance of Islam by local populations took place between the eighth and fifteenth centuries when trade in the Indian Ocean was dominated by Muslims to the extent that it came to be regarded as a ‘Muslim lake’.Footnote ² Muslim trade networks and Islam represented one of the factors for the unity of the Indian Ocean, but the spread of Islam was not even or consistent; nor were Muslims the only traders in this Ocean.Footnote ³ Islam came to the northern Mozambican and Malabar coasts very early, perhaps in the eighth century, through routes that were already in place long before the Muslim faith emerged.Footnote ⁴ These routes linked the Red Sea coast, the Persian Gulf, South India, South Arabia, Persia, Southeast Asia, East Africa and China to each other and to the wider Indian Ocean web of connections and exchange.Footnote ⁵

Matrilineal Austronesian seafarers and traders from Western Indonesia and the Malay Peninsula traversed the Indian (and the Pacific) Ocean and linked India, China, Sri Lanka and East Africa to Southeast Asia, triggering major migrations of food culture, animals, musical instruments and maritime technology.Footnote ⁶ The migrants settled along the East African coast, including Mozambique, by the first century BCE and sometime between 100 and 700 CE formed permanent communities on Madagascar.Footnote ⁷ A large number of Indonesian seamen, who were not Muslims yet, had reached and raided the Mozambican coastline in the tenth century.Footnote ⁸

The Malayo-Austronesians were instrumental in spreading Buddhism and later, Hinduism from India into Southeast Asia from the seventh century BCE to the first century CE.Footnote ⁹ In the fourth century CE there was a widespread adoption of Buddhist and Hindu beliefs, Sanskrit terms and script, and of coinage, all reflecting the influences from the Indian subcontinent. However, much of this cultural adoption was done by dynastic leaders of Srivijaya and Majapahit in Sumatra and Java as ‘means by which local leaders may claim special knowledge and thereby reaffirm their claim for authority’.Footnote ¹⁰

When the ports in southern India, Arabia and Persia became Islamised, the Malay and Indonesian seafarers and traders were influenced by the new religion, and through their interaction with Muslim merchants and together with the newly converted Southeast Asian and southern Chinese Muslims, they Islamised the entire region.Footnote ¹¹ The involvement of the northern Sumatran kingdom of Samudra-Pasai within the Indian Ocean networks in the thirteenth century resulted in it becoming the first political entity to accept Islam as a court religion in the region.Footnote ¹² By the 1500s, there were several Muslim states in Sumatra, Java, Sulawesi, and the Moluccas, while other regions adopted Islam later.

Archaeological evidence suggests that Muslim sailors and merchants probably stayed through monsoon periods or used as respites the islands or the littoral of northern Mozambique on their way to the coastal ports that linked the Indian Ocean to the southern African gold-mining mainland regions probably from the eighth century onwards.Footnote ¹³ As a stopover for merchants sailing from the Red Sea to Southeast Asia, the Malabar Coast also attracted itinerant merchant communities.Footnote ¹⁴ These traders might have led some local people towards Islamisation during these short stays. But in later periods, more often than not, foreign merchants and sailors took local wives and many also settled permanently and were absorbed by local communities. The offspring of these migrant Muslims are credited with the formation of new ethnic groups, such as the Koti, Naharra, Mwani and other Swahili peoples of northern Mozambique, and the Mappila of Malabar. In the fourteenth century, Ibn Battuta witnessed the presence of local Muslim communities along with merchants from various parts of the Arab world, India and Persia along the Malabar and East African coasts.Footnote ¹⁵ Home-grown converts expanded Islam further along the coast and into the mainland. When the Portuguese arrived at the end of the fifteenth and early sixteenth centuries, they found that in both regions foreign mercantile as well indigenous Muslim communities were thriving.Footnote ¹⁶

Although the settlement of the Lakshadweep Islands off the coast of Malabar took place in about the second century BCE, A.R. Kutty and Leela Dube suggest that Islamisation there occurred only in the fourteenth century CE under the influence of the immigrants from Malabar, while Andrew Forbes argues that it might have happened much earlier due to the impact of the South Arabian Hadrami sailors and merchants.Footnote ¹⁷

The conversion of the Minangkabau of West Sumatra probably occurred later than that of people in other parts of Southeast Asia in the fifteenth and sixteenth centuries when their region was drawn into the international trading networks.Footnote ¹⁸ As Tsuyoshi Kato underscores, it is not clear when exactly the Minangkabau accepted Islam, but an early-sixteenth-century account reports at least one Minangkabau king as Muslim.Footnote ¹⁹ The Aceh sultanate controlled the region at the time and probably spread Islam there. Around the same period in the 1500s, the Minangkabau settlements in Negeri Sembilan on the Malay Peninsula developed. Michael Peletz maintains that ‘Minangkabau males engaged in voluntary migration (merantau) navigated the narrow Straits of Malacca, sailed up the Linggi and Rembau rivers, and founded settlements in Negeri Sembilan and Malacca, just to the south.’Footnote ²⁰ There they converted to Islam under the influence of the international Islamic trading community and the emerging Muslim kingdoms.

In all regions under consideration, the Indian Ocean networks produced Sunni Muslims following the Shafi'i legal school (Ar., maddhab), a fact already noted by Ibn Battuta for fourteenth-century East Africa and Malabar. In all cases, the waves of migrations from the Hadramawt region of southern Yemen had consolidated particular high-status groups claiming to be descendants of the Prophet Muhammad, the Awlad Sharif. The sharifs in northern Mozambique and in Kerala enjoyed kinship relations with local ruling or upper class clans and their high social status resulted from both alliances with the elites and from their own sharifian descent. Hadrami sharifs are indicated as ancestors of many northern Mozambican coastal rulers.Footnote ²¹ As early as in 1517, a letter from the leader of Mozambique (probably Mozambique Island and adjacent area) to the Portuguese King Manuel was signed by Sharif Muhammad al-Alawi, a name which points to his possible Hadrami Alawi sharifian pedigree.Footnote ²² Besides the honorific titles of sharif and sayyid (Ar., lord), this group possessed additional special designations such as koya and thangal in the Lakshadweep Islands and in Malabar. The term koya denotes the land-owning Muslims who converted to Islam from higher Hindu castes of Brahmins and Nayars, but who claimed a sharifian descent simultaneously.Footnote ²³ According to A.R. Kutty and Leela Dube, ‘koya is a corrupt equivalent of the Arabic word khoja, meaning “the respected”’.Footnote ²⁴ Their position as formerly high-caste Hindus with land-owning birthrights was strengthened by their engagement in trans-oceanic trade through the centuries and by their monopoly over the ownership of sailing craft.Footnote ²⁵ They were further empowered by their sharifian ancestry which granted them exclusive rights over vital Muslim communal activities, and they usually served as kazi (qadi, the Muslim judge), imams of the Juma mosque (central mosque) and heads of the Sufi Orders (the khalifa).Footnote ²⁶

Although individual Hadrami sharifs might have migrated earlier, their widespread settling in Southeast Asia occurred later than in the Malabar and East African coasts. In the first half of the eighteenth century, significant numbers of the Hadrami sayyids resided in Aceh, Borneo and south Sumatra.Footnote ²⁷ Engseng Ho mentions that they settled among the Minangkabau in the same century, and one of them married the daughter of a local Rajah.Footnote ²⁸ Here too, the intermarriage between the Arab sharifs and local kings and other aristocracy, underlined the importance of ‘the precious gift’ of Prophetic genealogy for producing new Islamic elites.Footnote ²⁹

The arrival of the Hadrami sharifs went hand-in-hand with the launching of the Sufi Orders both in Lakshadweep and among the mainland Mappila communities, where the Rifa'i, Qadiri, and Alawi (and later, Chisti) tai'fas (Ar., branches) were founded and headed by the thangal and Koyas, who led the weekly dhikr ceremonies, annual celebration of the Prophet Muhammad's and local Sufi saints’ birthday celebrations (maulood or mawlid) and the nerchara tomb visitations.Footnote ³⁰ In northern Mozambique and among the Minangkabau, the Sufi Orders emerged much later, but were also dominated by the Hadrami sharifs entwined with local political and land-owning elites.Footnote ³¹

In all regions under consideration, Islamisation did not annihilate the pre-existing matriliny, and matrilineal kinship continued to be at the base of social relations so that families and lineages and associated with them, inheritance, succession and residence patterns were constituted through mothers, and their daughters and sons.Footnote ³² Children were raised by mothers and maternal relatives. As Evelyn Blackwood points out, among the Minangkabau, matriliny empowered women ‘as controllers of land and houses, creating an obvious dissonance with the masculinist discourses of the state, Islam and capitalism’.Footnote ³³ And as such it stood in stark contrast with the prescribed norms of Islam, which not only define kinship patrilineally and put men at the apex of power relations within families as well as wider society, but also provide detailed legal injunctions on inheritance and the rearing of children.Footnote ³⁴ Scholars have long puzzled over the reasons for the ‘paradox’ of Islam not eradicating conflicting local matriarchal and matrilineal practices in these regions.

Michael Lambek suggests one of the most compelling reasons for the concurrent existence of Islamic and pre-Islamic practices among Muslims of the East African island of Mayotte. He argues that in the context of Indian Ocean-wide Islamisation, it is more accurate to speak of ‘acceptance of’ rather than ‘conversion to’ Islam.Footnote ³⁵ Conversion ‘privileges an immediate, virtually instantaneous, unidirectional, and distinct shift in self-identification over a gradual change … insofar as it privileges a private, subjective experience over a collective process; and insofar as it privileges rationalised “beliefs” over ritual order …’.Footnote ³⁶ Acceptance, on the other hand, as David Parkin points out, does not imply an abrupt abandonment of prevailing ideas and practices; it is a process ‘which is likely to take longer and to be reciprocally inscribed in pre-existing custom and cosmology …. Yet it may well typify much Islamisation in the region [across the Indian Ocean] in allowing for Islamic and non-Islamic traits to inter-mingle steadily’.Footnote ³⁷

Lambek maintains that the local performances of the global Islamic texts provided the basis for a common religious identity shared by the worldwide ‘umma (Ar., Muslim community), but the performances themselves responded to local needs and were embedded in local kinship and community:

While this model fits well into the situation of the regions examined by this article, it is also true that not all Muslim communities along the Indian Ocean rim remained tolerant towards their pre- or non-Islamic practices. In other word, not all of them accepted rather than converted to Islam. This is particularly clear from the examples of the Southeast Asian regions other than the Minangkabau, and the Swahili coasts of Kenya and Tanzania just north of Mozambique, where matriliny was gradually replaced by patriliny or bilateral kinship.Footnote ³⁹ A closer look at these cases reveal that political power, in particular in Muslim states such as the sultanates of Java or Oman and Zanzibar, was probably essential for this kind of shift. Similar types of Muslim states were absent in Minangkabau, Kerala, and northern Mozambique, although some short-lived forms emerged here and there occasionally, and in Mozambique many coastal Muslims saw themselves as subjects of the Sultanate of Zanzibar until its fall in 1964, despite living under Portuguese rule.

Islamic reformers, colonialism and matriliny

The regions considered here were among the first to trade and later, to be colonised by the Europeans. The history of that interaction allows us to analyse three different colonial attitudes towards the coexistence of matriliny and Islam that Europeans encountered: the British, Dutch and the Portuguese. What were the consequences of these attitudes for the survival or transformations of the matriliny–Islam symbiosis?

In the case of Mozambique, the Portuguese did not seem to have much knowledge or capacity to deal with Muslims, except for initial confrontations in the sixteenth and seventeenth centuries, when they conquered some Swahili enclaves along the coast but left others, such as Angoche, to remain independent of their incursions.Footnote ⁴⁰

At the beginning of the modern colonial period, in the early twentieth century when the military conquests of the ‘effective occupation’ were almost concluded, the most important measures pertaining to Islam and matriliny had an administrative character aimed at organising colonies territorially and establishing a new bureaucratic system. The Administrative Reform legislations of 1907 and 1933 incorporated into the administrative machinery chiefs (wamwenye in the case of northern Mozambique) who were expected to rule according to local ‘traditions and customs’ while promoting metropolitan economic interests of collecting taxes and imposing forced labour. Significantly, women were not barred from exercising chieftaincy if that was the ‘custom’ of local people.Footnote ⁴¹ The contents of African ‘traditions and customs’ were not addressed as Portugal did not have the same approach as other European colonies, neither did it boast of politicians or intellectual elites engaged in learning and translating Islamic religious sources written in Arabic comparable to that of other European regimes of the time. Although the chiefdoms and other territorial units were reorganised anew, which in effect disrupted the pre-existing structures and ideologies of Africans, Portugal insisted on selecting and endorsing chiefs whose legitimacy and authority were based on local ‘customs and traditions’.Footnote ⁴² This attitude to a certain extent safeguarded matriliny, because the chiefs and their retinues became responsible for the application of ‘traditions and customs’, with dispute arbitrations between the clan or family members with regard to property and inheritance left firmly in their hands.

Among the Islamic reformers who emerged in the mid-twentieth century under such names as sukuti (Ar., silent; those who opposed Sufism), ‘Deobandis’ (under the influence of the Indian Deobandi school) and the Wahhabis (those who studied in Saudi Arabia), almost none addressed matriliny. However, some men with a similar outlook but living in rural northern Mozambique and contesting for the positions of chiefs did try to eliminate matriliny or female leadership there.Footnote ⁴³ They were, however, unsuccessful due to African resistance and the Portuguese preoccupation with maintaining ‘legitimate’ chiefs in order to make the peasants produce what the metropolitan state needed. During the Estado Novo (1926–74), the Salazar regime allied with the Catholic Church tried to impose Christianisation, but they realised the futility of this undertaking by 1954, when international pressure to decolonise Africa intensified. In the last decade of the colonial period, between 1964 and 1974, the regime undertook for the first time a systematic study of its Muslim colonial subjects to counteract the influence of the independence movements, but decolonisation dawned soon after so the policies elaborated from the newly-obtained knowledge were never implemented.Footnote ⁴⁴

The British, on the other hand, promoted Indirect Rule through ‘native’ administration and sought to systematise knowledge about their colonial subjects and introduce policies stemming from that process. Thus, in Malabar and Negeri Sembilan, British rule adopted a more legalistic approach to matriliny and Islam, whereby Islamic law (‘the Muhammadan law’) and ‘customary law’ were established and codified as clearly separate juridical and cultural domains. Thus, the rules of matrilineal kinship and inheritance, in the form of adat among the Minangkabau in Negeri Sembilan and the marumakkatayam in Kerala became ‘customary laws’ incorporating matriliny as opposed to Muhammadan Law, which codified what British perceived to be Shari'a and in which matriliny had no place. As a result, the British did not abolish the traditional system of clan offices and courts, but introduced previously absent Islamic courts and qadi offices leading to conflicts around the interpretation of local customs and the provisions of Shari'a.Footnote ⁴⁵ Peletz points out that adat in Negeri Sembilan, became ‘subject to processes of rationalization and secularization’.Footnote ⁴⁶ The notion of adat, as ‘customary law’ came to incorporate adat perpatih, which was based on local ‘traditions and custom’ among the Minangkabau and dealt with kinship, traditional titles, and women's prerogatives in proprietorship and inheritance. Adat temenggong prevailed among other Malays and had a more Islamic character. In general, as M.B. Hooker argues, adat in Negeri Sembilan became ‘mainly concerned with detailed land rules’.Footnote ⁴⁷ It supplemented regulations on land tenure introduced by the British in Malaya, such as the Customary Tenure Enactments of 1909 and 1926, the Malay Reservation Enactment of 1913, and a new category of property known as Malay Reserve Land, access to which was barred to non-Malays.

But it was not only the British who stripped adat of its cultural meaning and made it more functional in terms of administering the territories and justice.Footnote ⁴⁸ In West Sumatra, the Dutch intervened in the early 1800s in the Padri wars, a religious conflict between Muslim groups initiated by clerics returning from the hajj in Mecca, where they witnessed the Wahhabi conquest, and declared jihad against Minangkabau adat back home.Footnote ⁴⁹ In particular, Syeh Achmad Chatib refused to accept matriliny among the Minangkabau.Footnote ⁵⁰ The Dutch intervened in favour of the adat proponents as opposed to the pilgrims. Thus, in addition to responding to conflicts related to Islam, the Minangkabau continually reconstituted adat in their interactions with the Dutch colonisers, who began codifying it after the 1830s. In particular, the Leiden School's legal historicism influenced by British approaches was instrumental in codifying adat.Footnote ⁵¹ Malay intellectuals contributed to that process by responding to Dutch requests to write about adat, which they did in ways that made it seem more in line with Dutch legal thought.Footnote ⁵² As Evelyn Blackwood argues, the resulting Minangkabau adat texts emerged in dialogue with the Dutch, Islam, and local oral tradition.Footnote ⁵³ In both Negeri Sembilan and West Sumatra the scope of adat was reconstituted and reduced, however, and the flexible and fluid oral tradition was codified by the colonial administrators and thus, ‘frozen’ in writing.Footnote ⁵⁴ Adat, which hitherto incorporated oral tradition on origins, customs, beliefs, and ‘the proverbs that serve as guiding principles and rules for ceremonies, behaviours, and kinship relations’, became confined to a concern with its implications for inheritance and control of harta pusaka (ancestral land).Footnote ⁵⁵

Similar conclusions could be drawn from the case of Malabar, where the British principles of Indirect Rule and Native Administration meant that local social and political hierarchies and land ownership remained largely in place, which enabled the Hindu higher castes to maintain the status quo with regards to their ownership of land. As in Malaya, the British set to codify local ‘customary laws’ in the nineteenth century by introducing the Malabar Marriage Act in 1896 and the Malabar Will Act in 1898, which enforced prevailing matrilineal customs, in particular the tharavadu (matrilineal corporate kinship property/land) as its central principle.Footnote ⁵⁶ And as in Southeast Asia, the legal reforms and the crystallisation of ‘customary law’ arose with the participation of local elites.

An important caveat to British and Dutch codifications of matrilineal rules was the colonial preoccupation with a father's self-acquired property, which often caused conflicts between his children and his matriclan. Dutch colonial courts in Indonesia consistently validated those adat rules that required the consent of the matrilineal heirs for donations of self-acquired property to children.Footnote ⁵⁷ In 1930, a father was allowed to donate or will his self-acquired property without his matrilineal relatives’ consent, but the intestate inheritance of self-acquired property remained subject to matrilineal principles.Footnote ⁵⁸ In Negeri Sembilan, today only harta pusaka is under adat perpatih, while self-acquired property is considered private property more often owned by men than women.Footnote ⁵⁹

The Malabar Marriage Act allowed members of any matrilineal caste to ‘register’ marriage, in which case laws governing bigamy, adultery, divorce and guardianship of children could be enforced, and a self-acquired property could become a bequest.Footnote ⁶⁰ The Malabar Will Act also allowed any matrilineal people to will any property which they could have given away while they were alive, especially self-acquired property; but in the absence of a will, matrilineal rules applied. In general, matrilineal principles governing the joint-family tharavadu property remained in place.

At the beginning of the twentieth century, the British delineated Malabar rules for specific ethnic or caste groups which included the 1912 Travancore Marumakkatayam Act, the 1918 Mappila Succession Act, the Nayar Act of 1925, and the Mappila Marumakkathayam Act of 1939. These laws also left the tharavadu basically unaltered, but non-registered traditional marriages were legalised allowing the wife and children the right to half of a husband's self-acquired property.Footnote ⁶¹ Although G. Arunima claims that the senior male, karanavan, was transformed by the British into the pater familias of Roman law and undermined the position of senior women, the kandoti and female-headed households, in practice this seems to have remained largely unchallenged.Footnote ⁶² Due to prevailing matriliny among them and in accordance with the British principle of ‘customary’ laws, the Malabar Muslims, in particular the Mappila, were barred from the application of the Anglo-Muhammaddan or Islamic laws being applied in northern India with regards to their land and inheritance rights, unless they renounced matriliny in favour of Islamic law. Otherwise, the customary laws encapsulating matriliny (the marumakkatayam) as codified by the British in Mappila Succession Act of 1918 and Mappila Marumakkathayam Act of 1939 were consistently enforced.Footnote ⁶³ As with other matrilineal peoples, the laws enabled the self-acquired property of a Mappila man who died intestate to bypass Muslim law and allowed equal sharing of the matrilineal property among male and female members of the family. However, in the Lakshadweep Islands, according to Dube, the difference between the tharavadu property (the so-called ‘Friday property’) and self-acquired property (‘Monday property’), was already entrenched historically.Footnote ⁶⁴ The Friday property was expected to be inherited matrilineally; while the Monday property, which was usually acquired by purchase, either was inherited through the Shari'a provisions or when a man was the sole heir, governed by the Shari'a rules.

Similar to northern Mozambique and the Minangkabau, in Kerala Muslim reformers appeared in the nineteenth and early twentieth centuries. The Mambran and Kondotti thannals had been politically active since the nineteenth century, leading or significantly influencing the Mappila revolts and jihads against the British reforms and the Hindu communities, but they did not challenge matriliny or Sufism.Footnote ⁶⁵ In the 1920s, several thangal critical of local practices and conceptions of Islam surfaced under the influence of the Egyptian reformers Muhammad Abduh (1849–1905) and Rashid Ridda (1865–1935). The most important was Vakkom Maulavi (1873–1932). He and his followers were staunch opponents of what they perceived to be shirk (Ar., idolatry) and bid'a (Ar., abominable religious innovations), including Sufism and matriliny.Footnote ⁶⁶ The Nadvat-ul-Mujahideen movement that emerged in the 1930s and formally organised in 1952, continued in the same anti-Sufi and anti-matriliny vein as the earlier reformers.Footnote ⁶⁷ But they were unable to alter matriliny to a great extent, at least in the form codified by British rule.

Matriliny compared: Post-colonial context

My own research and that of Isabel Casimiro, Signe Arnfred, Daria Trintini and Carla Braga indicate that matriliny is still alive among Muslim communities of northern Mozambique despite post-colonial attempts to modernise, secularise and even eliminate related ideas and practices.Footnote ⁶⁸ The same could be said about the Mappila and Koyas of Kerala and Lakshadweep, and the Minangkabau. This does not mean that matriliny has not been altered or has not been undergoing continual transformation, but it is more and more centred on so-called ancestral land, rather than other types of property. As in the colonial period, the ‘escape’ so to speak, from the control of the matrilineal kin group leading to individualisation of rights comes through the rules for so-called self-acquired property. This kind of property is owned mostly by men; it can be disposed of individually and it can be inherited by direct descendants rather than the matriclan at large, either in accordance with the statutory laws or the Islamic Shari'a.

Thus, in all regions examined in this article, matrilineal descent groups are constituted around ancestral land, owned by corporate descent groups, but inherited through female lines, with overlapping rights of disposal vested in matrilineal kin groups. Individuals within these groups only have usufruct rights. Among the Minangkabau, both in West Sumatra and Negeri Sembilan, such land is classified as harta pusaka belonging to a village (nagari) that consists of several ideally exogamous matriclans, suku, divided into several matrilineages (kaum) with ceremonially instituted male head called the penghulu or mamak.Footnote ⁶⁹ The suku are defined by the claim of a common ancestress (bundo kanduang) and sa pariuk (one womb), and they usually own a longhouse, belonging to the women, and possess the land. The land (its use, distribution, and income) is controlled by the senior woman along with the penghulu (or mamak within the smaller units), usually her brother or maternal uncle.Footnote ⁷⁰ The ideas and practices of matriliny among the Minangkabau currently are related to the old social and cultural system known as adat perpatih, which according to Peletz, ‘was grounded in pre-Islamic (that is, Hindu–Buddhist and pre-Indic) beliefs and practices in a rich corpus of oral tradition’, but which was codified under pressure from Islamic reformers and Dutch and British colonial rulers in the nineteenth century. Footnote ⁷¹

Land is central to the Muslims of Kerala too, especially for the land-owning Koyas, who having emerged from the upper caste Hindu Brahmins and Nayars, held views similar to them in this regard. Among the Nayars, land which ‘passed from generation to generation through women, was the foundation of matriliny’.Footnote ⁷² Similar to the case of the Minangkabau, for Nayars as well as land-owning Koyas of the Lakshadweep Islands and the Mappila of Kerala, the central kinship group has been a maternal lineage or clan termed tharavadu, consisting of a group of men and women tracing descent from a common ancestress (thus they are the ‘children of one womb’), and living under the control and management of the eldest male, called karanavar, the oldest woman's, or karnoti’s uncle, brother or son.Footnote ⁷³ Karnoti, the eldest woman, enjoys a special status; she is consulted by the karanavar in all important matters.Footnote ⁷⁴ The smallest unit is called thavari, which could be regarded as a segment descended from each woman of a tharavadu. It owns the joint family house also known as a tharavadu. Similar to the case of the Minangkabau's adat perpatih, largely pre-Islamic Hindu-based matrilineal traditions were codified under pressure from the British colonial regime in legal corpuses known as marumakkathayam, matrilineal customary laws, and in particular for the Mappila, as the Mappila Marumakkathayam.

In contrast to the Minangkabau and Kerala, the northern Mozambican matrilineal peoples did not possess a joint family house, but rather individual family units lived in separate residences. Today as before, the rural population lives in dispersed communities with individual houses scattered around some distance from each other, unless it is a village organised by the state. Otherwise, the perceptions pertaining to kinship and land have a lot in common with the Minangkabau and the matrilineal peoples of Kerala. Here, the ancestral land belongs to a kinship group called maloko (pl., sing. n'loko) descending from a common female ancestor symbolically identified as errukulo (‘a womb’) or nipele (‘a breast’).Footnote ⁷⁵ Each n'loko consists of several smaller kinship groups – nihimo (sing., pl. mahimo). The male head of each n'loko is mwenye n'loko; and of each nihimo is humu or mwenye nihimo. A senior woman, the elder uterine sister of the mwenye, is called piya-mwenye (literally, the female mwenye) or nunu. Each clan has its own senior man, a maternal uncle or brother of the senior woman called tata or halu.Footnote ⁷⁶ As with the Minangkabau and Kerala, senior women and men overlook and control the distribution and use of land, regulate relationships between members of the original kin groups and the newcomers, and settle disputes and oversee internal family matters as well.

In all the regions under consideration, it is incumbent upon the eldest man, be it a  mamak, karanavar, or atata to take care of his sister's children, as well as play the role of guardian (wali) for his nieces in Islamic weddings (adat, nikah or chuo), and arrange and oversee their marriage partners as well as their divorce and remarriage, and negotiate their dower (mahr).Footnote ⁷⁷ Thus, children belong to the matriclan, and are under the care of the eldest woman and her brother/maternal uncle. The father has no responsibility towards them, although nowadays he often pays for their schooling and other expenses. Upon divorce, which is quite common in all four regions, children are left with the mother or her matriclan; upon the death of the husband, his matriclan has no responsibility towards the children (although the attitude varies nowadays according to individual families).

The land rights of a lineage or clan among the Minangkabau and coastal northern Mozambique are derived from an idea of a first-comer status, while among the land-owning Koyas through the Hindu principle of birthright, janmon, which, broadly-speaking, also stands for the first-comer status. Thus, in all regions under consideration, most of the land is owned by the traditional elites, and access to land comes through the elites. What Blackwood underscores for the Minangkabau is true for the other two regions as well: those who are not considered descendants of the original settlers are clients of these clans with regards to the land, and ‘although bracketed within kinship ideology, their relationship is one of patron and client’.Footnote ⁷⁸ In the case of the Minangkabau, suku are frequently differentiated internally according to whether they were direct descendants of the original settlers of the nagari, or of latecomers.Footnote ⁷⁹ In wider geographic terms, people from the Minangkabau heartland mountains are called darek, the first-comers, ‘who feel superior to merantau, the inhabitants of the shoreline, rantau’.Footnote ⁸⁰ The arang asa (descendants of original settlers in West Sumatra) and the waris (‘heirs to the soil’ in Negeri Sembilan) are considered kemanakan beneath the chin, those whose ancestors were the original settlers, while kemanakan beneath the navel are considered to be their relations by favour, and the third group kemanakan beneath the knees allegedly descended from debt bondsmen, prisoners of war, or purchased slaves.Footnote ⁸¹

In coastal northern Mozambique, exemplified here by Angoche, the alleged first-comers are called the Anhapakho.Footnote ⁸² This group is constituted of four major clans (nhandare, nhamilala, nhatide and m'bilinzi), descendants of four sons of the mythical woman founder. The Anhapakho are considered to be the putative first-comers who ‘own’ the land, thus, they are the wamwenye (the landowning elites). In this quality, they oversaw its distribution to the later arrivals to whose allegiance they had special claims through marriage and kinship relations. The latecomers were expected to receive a portion of land from the first-comers in return for tribute and loyalty. Nowadays this relationship is expressed by payment of cash, crops or other goods.

The situation of the Mappila Muslim clans of Malabar, referred to as ‘children of the same womb’, differs from Minangkabau and northern Mozambique in that Hindu social organisation and perceptions prevailed before the conversion to Islam, and the caste system was transported into the nascent Muslim community. While the upper caste Brahmins and Nayars maintained their janmon land rights after conversion but were a minority, the converts from the lower castes, who were the majority, remained landless. This, according to Stephen F. Dale, led to various revolts and jihads by the Mappila that punctuated Malabar history between 1489 and 1922, and even after Independence and the partition of India.Footnote ⁸³

In all cases, the duolocal residence pattern of the earlier periods, when a married man did not reside with his wife but visited her daily while residing with his own matriclan, has been transforming into a more uxorilocal pattern since the mid-twentieth century.Footnote ⁸⁴ Of course, this varies according to the situation of the individual families, especially when they migrate to the cities or other regions. Thus, features of the idealised matrilineal kinship system are not always strictly followed nowadays. Modern Western education and employment, transnational experiences, such as in Gulf countries where stricter Islam is followed, and economic transformations in agriculture and in societies at large, have been engendering many changes. Inheritance and ownership of other property that is not the ancestral land could have become more patrilineal, the residence pattern might vary from uxorilocal to patrilocal or even independent, and the power of the nominal heads of clans, be they women or men, could be reduced to overseeing the ancestral land or sorting out family disputes. But so far, the ancestral land itself remains the common point that sustains matriliny in all the regions examined in this article.