Gender, succession and dynastic rule

ABSTRACT

This paper considers ruling women through the lens of gender and succession, mostly between 1300 and 1800; it underlines the fundamental impact of matrilineal succession on worldwide dynastic practice. First the paper asks how women surfaced and survived as sovereigns in a world that defined paramountcy in male terms. Second, it examines how changing patterns of descent, the rise of world religions and modernization have affected regional variations in the occurrence of sovereign women. Third, it revisits the ‘matrilineal puzzle', scrutinizing divergences between matrilineal and patrilineal formats of dynastic power. Finally, the fourth section of the paper reviews the connections between matriliny, the empowerment of women, and ‘contractual' kingship. The paper as a whole connects the historical examination of matriliny to recent work by archaeologists and evolutionary biologists on ‘female biased kinship’. It rekindles the age-old debate about the status of women in early history, shows the relevance of matriliny for current research, and makes explicit the patrilineal bent in common interpretations of dynastic power.

KEYWORDS:

• queens
• gender
• kinship
• dynasty
• matriliny
• patriliny

Introduction

Chiefs, kings and emperors are ubiquitous in world history. Everywhere, men ascended the throne far more often than women; leadership, moreover, has been viewed predominantly in gendered terms, as related specifically to masculine qualities. The traditions of kingship so powerfully present in world history took shape as an unending sequence of individual kings and ruling houses or dynasties. Who would take over after the death of a king? Throughout history, numerous forms of succession have been practised. Successors could be installed by fixed rules, designated by the incumbent or elected by kingmakers; succession might be determined by martial or ritual competition. While these procedures could lead to very different outcomes, eligibility was almost invariably restricted to descendants of a real or imagined forebear. Cultural understandings of sex, procreation and heredity delineated the group of potential successors. Mnemonic instruments, performed, depicted or written, preserved and expressed these rights, connecting them to history and cosmology.

Women, men, reproduction and childrearing in one way or another shape societies everywhere; yet these universal biological phenomena form the basis of a wide range of cultural practices. Cultural understandings of family, gender and descent are profoundly variable. Remarkably, while the study of gender has thrived in recent decades, kinship seems to have lost much of its allure in the same period. From the 1850s onwards, anthropologists saw kinship as the key to understanding ‘primitive’ societies: patterns of descent and alliance, they surmised, unveiled basic social relationships. The learned terminology they devised to deal with the intricacies of kinship coined abstractions unknown to the peoples they observed. This practice has been criticised by the poststructuralist generation of the 1980s, which viewed the vocabulary of kinship as yet another way of ‘creating alterity’ by attributing a predominant role to kinship in ‘primitive’ societies, while downplaying its significance for ‘advanced’ societies – more or less like that other flagship of anthropology: ritual.¹ Matrilineal descent, strongly present in scholarship until the 1950s and 1960s, almost disappeared from the academic agenda in the 1980s.² The alertness of earlier scholars to matrilineal ‘systems’, it was argued, had arisen in part from gender bias: they had taken for granted the nuclear family headed by the father, linking fathers and sons as carriers of the family name. Thus, while Audrey Richards’ lucidly discussed the ‘matrilineal puzzle’ deriving from the separation between men as fathers and women as bearers of the line of descent, she never wondered about the ‘patrilineal puzzle’ (Peters 1997). Current anthropological research shows a renewed interest in kinship (cf. the editorś introduction), though with more reservations about the idea that it reflected the foundation of the social edifice and with a keen eye for gender. At the same time, there has been a strengthening of gender and family studies among historians: modernity is no longer necessarily equated with the erosion or even irrelevance of kinship.³

Kinship is no longer understood as a coherent ‘system’ determining all facets of social life. Consequently, it has become unfashionable to speak about matrilineal and patrilineal ‘societies’: mixed forms in fact have been far more common, and different rules pertain for various groups (women and men, social layers) or in certain domains (marriage, succession to office, property holding, inheritance). Ruling houses, more often than not, adhered to rules for kinship and inheritance that set them apart from society at large. This paper zooms in on their practices. It examines women in dynastic history through the lens of gender and succession, mostly but not exclusively between ca. 1300 and 1800. The first section traces women ruling as sovereigns and examines how they fitted into gendered views of power. Regional diversity and long-term change are discussed in the second section: how did patterns of descent, the rise of world religions and modernization affect the presence of sovereign women? The third section revisits the ‘matrilineal puzzle’, showing how forms of descent and succession profoundly modified the contours of dynastic power. Finally, the fourth section considers the connections between matriliny, the empowerment of women, and ‘contractual’ kingship. The paper as a whole rekindles the age-old debate about the matriliny in a long-term perspective, fitting recent work by archaeologists and evolutionary biologists on ‘female biased kinship’ (Knight 2011; Jones 2011; Surowiec, Snyder, and Creanza 2019).⁴

My comparative view of dynastic power does not follow the accepted tradition of bringing together examples selected according to criteria related to period, geographical location, interconnectedness, scale or complexity. Dynasties can be found in societies across the spectrum of scale and levels of economic or technological complexity, in script cultures and empires as well as in miniature chiefdoms operating through orality. Questions about dynasty, gender and kinship are as relevant for the Chinese empire as for any African kingdom – and this unorthodox juxtaposition leads to unexpected questions. While my knowledge of European princes and courts allowed me to establish several core questions, I have pursued global variety and divergence to test, stretch and amend this framework.⁵ Notwithstanding the wide scope of cases considered, my comparison does not rely on the systematic cross-cultural analysis of data assembled in anthropological atlases and reference files (See Murdock 1957, 1967; Surowiec, Snyder, and Creanza 2019). The combination of high-level scientific methodologies with the nature of the data used, it seems to me, raises questions about validity, if only because the contextual foundation of the data remains quite limited. However, by disregarding accepted methods of comparative history as well as cross-cultural analysis, this paper necessarily remains essayistic and provocative rather than conclusive.

This examination of gender, power and the diverging practices related to matrilineal and patrilineal succession, calls into question common assumptions about dynasties. It makes explicit the patrilineal bent in traditional interpretations and underlines the lasting relevance of kinship and gender for research.

Women and sovereignty

With few exceptions, historical texts attributed paramount power to men and associated it with categories they viewed as typically male: physical strength, military prowess, stern justice. The ubiquity of royal polygyny underlined virility and fertility among the male traits attributed to kings. Women figured as mothers, spouses, daughters and favourite concubines or mistresses: roles hinging on the presence of the king at the heart of the dynastic constellation. These images do not necessarily tell us much about power. Kings in practice often depended on the circle of servants and advisers around them, and women were in a particularly strong position to exert power behind the scenes. Dynastic women had recognized spheres of activity and influence, mostly related to devotion, family, education, and charity, yet their numbers undoubtedly included prime political actors. If such female movers and shakers openly claimed domains of activity commonly ascribed to the male ruler, however, they risked being accused of meddling, conspiracy and usurpation. Their undisguised dominance challenged the ruler’s authority: how could kings who were unable to control their house govern their realm effectively? Magistrates and advisers, disconcerted by the influence of women, created a repository of clichés associating female power with illegitimacy and degeneration.

Why were women occasionally accepted as sovereigns? They stepped in most often after bloody succession violence, to prolong the dynasty in the absence of eligible male candidates. In these circumstances, the pre-eminence of their blood overruled gender bias. Indeed, with minor sons present, it was common for mothers to ascend the throne temporarily, as regents. Most women in history ruled in their capacity as mothers, in the expectation of eventual replacement by their ward. For periods and areas in history where sources are sparse it is difficult to ascertain whether a woman ruled in her own name, or temporarily in the name of her deceased spouse or minor son. Listings of famous queens in history – including the oft-mentioned queens of ancient Egypt and the candaces of its Southern successor empire Kush – comprise many queen-mothers (Kahn 2012; note the discussion about matrilineal succession in Kush). Motherhood did not necessarily make these queens less powerful, but it separated their role from the male contours of sovereignty and so did not squarely challenge gendered views: dowagers could be admired as pious, virtuous, loving and merciful educators. In imperial China, where only one woman briefly ruled as empress in her own right, many held de facto power as dowager-empresses. France and the Ottoman empire, two realms likewise noted for their outright rejection of female rule, easily accepted queen-mothers, who were conspicuously powerful in the sixteenth and seventeenth centuries.

Men on the throne who aspired to live up to the ideals of kingship found the challenge daunting; the demands placed on their shoulders were exacting as well as contradictory. Who could match the qualities of the indomitable warrior with those of the prudent sage, or liberally distribute largesse without overspending and thus exploit his people? The masculine qualities attached to kingship exacerbated the difficulties for women who rose to power. Queens-regnant had little use for qualities seen as desirable for consorts such as modesty and compliance; their success in putting into practice male princely virtues, moreover, was not always appreciated. Snubbing advisers, leading armies and ousting rivals, they risked depiction as shrews and she-males. Among the exceptional women ruling with full powers, several refashioned themselves as males. Seventeenth-century Queen Nzinga of Ndongo-Matamba smoked a pipe, wielded weapons fiercely and practised polyandry – before she went the whole way by redefining herself as male and acquiring concubines (Thornton 1991). The presence of warrior queens in myth and history underlines rather than contradicts gendered images of power: strong women needed to either take cover under the mantle of female qualities or ‘become’ men – strategies familiar from other occasions where women moved into ‘male’ positions. The first option was easier for mothers than for childless queens. Supreme power, apparently, could not be reconciled easily with accepted gender roles. Remarkably, even the most outspoken example of a ruling house with a long-term preference for women as kings, the six South African Lovedu rain queens who ruled between 1800 and 2005, indirectly confirms the connections between masculinity and paramountcy. The hidden rain queens were represented as men, maintained numbers of concubines and needed to conceal their male lovers (Krige and Krige 1943).

Notwithstanding the obstacles to female rule, it is clear that in areas where women rose to the throne as an ad hoc solution for the absence of male successors, the experiment stood a chance of being prolonged. Nzinga’s long-drawn-out battle against gender expectations inaugurated a century in which female rule predominated (Thornton 1991, 40). Farther South, queens had reigned in the various principalities of Madagascar, before the accession to power between 1828 and 1897 of four queens-regnant in the Kingdom of Madagascar.⁶ Similar clusters of female power can be found elsewhere. Four women ruled in the Sultanate of Aceh (1641–1699), seven in the Thai Sultanate of Patani (1584–1718). A Fatwa decrying female rule played a role in the restoration of male power in Aceh – but clearly, Islam had not prevented women from rising to power in these sultanates (Veth 1870, 40). Moving eastwards in the South-East Asian archipelago, we find queens as a common exception to male rule. A rare case, the Buginese kingdoms of Sulawesi even accepted women on the throne in the presence of eligible males judged to be less qualified (Amirell 2011, 2016; Watson Andaya 2006; on Aceh see Khan 2017).

Along the Eastern fringes of the Asian continent, a cluster of female rule can be found in an earlier period, with six Japanese empresses reigning for most of the period between 590 and 770, though under the tutelage of male relatives and regents (Jay 1996; Nickerson 1993). During the seventh century, two queens ruled the Korean Kingdom of Silla, where another queen emerged in the ninth century (see Jay 1996; Nelson 2003). The Chinese example shows that female rule did not invariably take root. After Wu Zetian’s exceptional reign (690–705) her youngest daughter as well as a granddaughter lined up for succession, yet their efforts failed, and no woman henceforth ever formally ruled as empress. Vilification of Wu Zetian’s legacy was actively used to prevent the rise of other women: she became a byword for women wreaking havoc by transgressing accepted moral boundaries (McMahon 2013).

Prominent women arose everywhere in mainland Asia, but only very few queens-regnant can be found, and there is no trace of a longer sequence of women on the throne. Central Asian women throughout history have been singled out for their power and independence, but Mongol queens ascended the throne as spouses and wielded power as regents or mothers rather than as queens-regnant. Princesses were important in the early modern Safavid and Mughal empires, but they, too, failed to attain sovereignty in their own name. Their West Asian counterparts in the Ottoman Empire held power only as sultanic favourites and queen-mothers. Among the early modern land-based empires, Russia stands out with a remarkable series of sovereign tsaritsas in the eighteenth century. Europe on the whole shows a mixed picture, with Mary (1553–1558), Elizabeth I (1558–1603), Anne (1702–1714), and Victoria (1837-1901) ruling in England and numerous isolated examples of queens-regnant on the continent. Yet there were restrictions in France as well as in the Holy Roman Empire. Maria Theresa (1740–1780), a figure of the same stature as her Russian contemporaries Elizabeth (1741–1761) and Catharina (1762–1796), could not formally be crowned as emperor: the dignity was expressly reserved for males. Effective rule by a woman may temporarily have alleviated gender bias, yet in the end, invariably, the default situation of male rule was restored. Everywhere, supreme power was gendered.

Mutterrecht to patriarchy?

Mutterrecht, or matriarchy, attracted considerable attention among nineteenth-century scholars, some of whom postulated an early preponderance of women, uprooted by a patriarchic transformation around the time of the Neolithic domestication of plant and animal life. They fitted these ideas into a grand evolutionary scheme of mankind, gradually moving to modernity, to finally reach the accepted standards of the nuclear family and monogamous marriage (Bachofen 1861; Morgan 1877). The hypothesis of an early phase where women dominated – Johann Jakob Bachofen used the term Gynaikokratie, soon translated as matriarchy – was based in part on mythology, which included a great number of mother goddesses, women warriors, and founding ancestresses. How did this conspicuous presence of powerful women in cosmologies match the predominance of male leadership in recorded history? While the question remains interesting, no evidence has been uncovered for the existence of matriarchy in history. Indeed, the notion of an early phase of mankind where women ruled was discarded in the early twentieth century. Both the proponents and the adversaries of the Mutterrecht hypothesis often conflated succession or inheritance through the female line with the power of women. Bachofen himself made the distinction, but at the same time his descriptions and terminology gave rise to misunderstandings. Commenting on the discussion in the closing years of the nineteenth century, Edward B. Tylor differentiated between the two, denied a phase where women ruled, but acknowledged the marked presence of matrilineality in an earlier phase of human development (Tylor 1896). Recently, archaeologists and evolutionary anthropologists have argued that matrilineal descent may have been the predominant mould of early kinship (Knight 2011; Jones 2011).⁷ Matrilineal descent cannot be equated with matriarchy, but as we shall see it impacted the positions of men and women in multiple ways.

Can the marked regional differentiations in the presence of women near the heart of power be connected to systems of descent? The ocean world connecting Madagascar, South India, the Malayan archipelago, Australia, Polynesia and perhaps even extending to South America shares an ‘Austronesian’ background in terms of languages as well as social patterns, including the relative prominence of women. Multiple examples of descent via the female line and female property-holding can be found in these areas, coinciding with the relative frequency of queens-regnant mentioned above. Likewise, examples of matrilineal descent can be found in North Africa, West Africa and in a zone connecting Central West Africa to South-East Africa. In these areas, women on the throne and in high offices near the sovereign occurred more often than elsewhere. Spouses held esteemed positions in many places, and mothers could rise to power as regents across the globe, yet in Africa women formally and permanently complemented the powers of the adult male paramount ruler. Queen-mothers were present in the familiar form, as mothers ruling in the name of their minor son. However, in many cases the female ‘reign mate’ was not the king’s mother, nor even related to him: she might represent the women of the realm or the non-royals in the population. Reign mates were a permanent feature, counterbalancing the powers of the king and acting as ‘kingmakers’, with a particularly important role during interregnums. This dual male-female power structure could be extended to include a royal sister. Austronesian and African examples suggest that the marked presence of matrilineal traditions in these regions lowered the threshold for women to be accepted in positions of power. Moreover, the presence of matriliny and queenship appears to have had an impact on the position of women in contiguous areas where patrilineal descent obtained (Knight 2011).

Finally, in Europe as well as in the Americas, descent was rarely reckoned exclusively through the male line. In the empires conquered by the Spanish, the female line played a role. Aztec queens connected male rulers to the prestigious Toltec lineage. Inca queens were either the king’s sisters or hailed from noble lineages. The Inca royal couple was likened to the union of the goddess of the Moon and the Sun god. Forms of dual descent and the presence of the matriliny next to the patriliny can be traced throughout the Americas (Yaya 2013; Nelson 2003; Aberle 1974; note the Sun-Moon metaphor in Wunder 1992). Europe’s most marked dynastic Sonderweg, monogamous marriage as the basis for legitimate dynastic succession and intermarriage, went together with an increased stress on the queen’s pedigree. The standard of monogamous marriage engendered two characters in Europe that cannot be found elsewhere: the bastard and the mistress. Polygyny, conversely, coincided with two phenomena rare in Europe: royal harems and their typical third-gender servants, eunuchs.

Do multiple ad hoc shifts hide long-term change? Matriliny has been associated mostly with small pre-capitalist economies, organized around horticulture or fishing, and without herds or plough cultivation: cows and ploughs, it has been stated, engender patrilineal descent (Aberle 1974, 704–705; BenYishay et al. 2017; Douglas 1969). In an environment with plentiful resources, matrilineality thrives because it can easily integrate outsiders into the group; yet under conditions where groups defend scarce resources against transgressors, patrilineal descent appears much more effective. It has been argued, therefore, that the consolidation of larger political entities as well as the development of supra-regional market economies necessarily entailed the marginalization of matrilineal practices (Johnson 2016; Peters 1997; Douglas 1969). While examples of modern matriliny show that the practice persists, it does seem likely that patriliny became more dominant in the past two millennia.

Another argument points to the possibility of a gradual retreat of women from positions of power. All major religious-moral creeds consolidated existing misgivings about female rule by including them in their hallowed writings and so in the long run may have helped to strengthen these views (Watson Andaya and Ishii 1992; Watson Andaya 1994; Amirell 2016). Statements disparaging the paramountcy of women spread from learned religious tractates to mirrors for princes and the occasional angry pamphlet. In 630, when Queen Borandokht ascended the throne following violent succession strife in the Sasanian Empire, the Prophet Muhammad allegedly commented: ‘Never will succeed such a nation as makes a woman their ruler’ (Krehl and Juynboll 1862–1908, IV, 376–377). In the core lands of Islam, Theravada Buddhism and Confucianism, female sovereignty became more exceptional. Neither did Christianity leave much room for women on the throne, whether in its Orthodox, Catholic or reformed Protestant guise. All creeds to some extent accommodated pre-existing practices, in the core lands and particularly when spreading to distant territories. Thus, in areas where women had traditionally enjoyed property rights, high status and freedom of movement, the impact of new scripted creeds was mitigated by local custom.⁸ The story of queens in the South East Asian archipelago shows that the holy writ and its learned interpreters did not have the same impact everywhere. The spread of Islam in Africa, first in the North and later in Sub-Saharan Africa, suggests a more mixed picture: the persistence of local traditions coincided with a tendency towards the withdrawal of women from conspicuous forms of authority. A periodization of female sovereignty, however, will necessarily remain uncertain, if only because the sources too often remain nebulous about queens-regnant, queen-mothers, and queen-consorts.

The assumption that religious change and political as well as economic modernity, in the end, did reduce the political presence of women seems plausible, though not at levels validating old ideas about a patriarchal takeover of the Mutterrecht. The rise of European commercial empires, and particularly the ensuing military and religious hegemony of colonial powers, coincided with an increased preponderance of men in political institutions. Colonial government had little patience with the intricacies of dual male-female power structures, and preferred unambiguous fixed forms of succession – typically male primogeniture – because they made more manageable the mayhem of interregnums. Political modernity had mixed consequences in Europe itself, too. The gradual marginalization of dynasties and their courts took away a domain where high-ranking dynastic women had wielded power. They had no role to play in assemblies and ministries populated by elected or appointed male office-holders. The right of women to vote was obtained only after several generations, and to date women’s access to supreme political office is still far from self-evident in most countries. The list of women first holding such offices notably includes widows and daughters of male office holders. The violent death of a father or spouse facilitated the move to supreme political office of Sirimavo Bandanaraike, Indira Gandhi, Benazir Bhutto and many others. In representative assemblies across the globe, but particularly in Asia, female representatives more often hail from ‘political families’ than their male compeers. Even in the twentieth century, apparently, family prestige could compensate gender bias (Duindam 2019a).

Dynasty: from the patrilineal standard to the ‘matrilineal puzzle’

Paramount male power was accepted as the standard by both patrilineal and matrilineal dynasties, even if the occurrence of women in leading or paramount office differed considerably. Where the preference for male power coincided with succession in the male line – unifying in one person the father and the bearer of the line of descent – father-son succession was likely to emerge as the preferred pattern. This type of succession could be fixed and restricted more easily than other forms, and therefore created the basis for long-lasting concentration of power and wealth.⁹ In the first place, sideways succession to brothers could be abolished, leaving only downwards succession to sons. In the second place, within the senior line the eldest sons might obtain a priority right to succession: male primogeniture. Restricting and fixing succession in this way had the great advantage of diminishing violent competition and facilitating the transfer of power to the next generation, always the Achilles’ heel of dynastic power. As a consequence, however, the freedom of the incumbent king to designate a successor from among his sons disappeared – and not every king placidly relinquished this prerogative.

In its extreme form, downwards fixed succession led to concentration of power in the small vertical line of the nuclear dynastic family and tended to dispossess and marginalize the extended dynasty, notably lateral junior branches. Younger sons would serve as reserves for succession, but the children sired by these junior royals tended to become a burden. Agnates might represent royal power as viceroys or generals: their dynastic charisma made them the ideal substitute for the ruler. The same quality, however, defined them as major threats, spearheading rebellions and dispossessing the king. With the incumbent king’s death looming and in the absence of an adult successor, dynastic scions stepped in. Queen-mothers could move forward as a shield against the usurpations of agnates, bridging the gap to the next generation. More incidentally royal daughters might manage to ascend the throne on the basis of the bloodline. Many kings feared the ambitions of their relatives, placed them under some form of surveillance or used more drastic measures: the dynastic record shows numerous examples of agnates who were incarcerated, exiled, incapacitated through mutilation or executed.¹⁰

With the agnates relegated to the periphery and succession fixed, other quandaries arose. The curtailment of succession rights and the elimination of rivals heightened the risk of extinction, frequently hitting ruling houses around the globe. Moreover, other tensions were more likely to arise because of fixed succession. Mature sons who awaited their turn impatiently were prone to become restless at the very time their fathers were experiencing the first signs of old age. Succession rights created tensions between fathers and heirs particularly during long reigns. Opposition in the dynastic context crystallized around potential successors, whether at court or in the provinces. The discontented among the elites sought the trust and patronage of the heir-apparent, in anticipation of his future reign. Ageing fathers, apprehensive about their declining strength and discernment, viewed the popularity of their sons with consternation. Pre-empting the possibility of conspiracy and usurpation, fathers could choose to dispossess, imprison or execute their sons (on father-son tensions see Duindam 2019b). Fixed succession rules not only complicated the relationship between the incumbent king and the heir-apparent. More fundamentally, the personal qualities of the heir-apparent now depended wholly on heredity. In the absence of the correctives of designation and competition, mediocre or incompetent figures would occasionally rise to power. Nonentities on the throne were a problem where active rule was necessary. It is therefore easy to understand that elites in dynamic empires, expecting martial leadership from their prince, hesitated about restricting rights to the firstborn. Fixed succession rules pleased leading magistrates in consolidated polities, who deemed activist leadership superfluous or even detrimental. Bureaucrats appreciated the unperturbed continuation of government, warranting the realm’s stability as well as their hold on power: the ideal prince for them was typified by his willingness to be guided by advisers.

Attempts to concentrate power in the vertical line, the wary surveillance of the horizontal family and the potential for conflict between father and son have together shaped the rather violent history of patrilineal dynasties. A diverging set of circumstances ensued in the case of matrilineal dynastic succession. This is where Richards’ puzzle comes in. A comparison of the patrilineal standard sketched above with dynasties practising descent via the female lines suggests that her analysis still has great relevance.¹¹

Separating fatherhood from the bloodline, matrilineal descent prohibits the succession of the king’s sons. The king rules because of blood transferred to him by a royal woman, but cannot himself reproduce royalty. A princess of the royal bloodline might serve as his spouse and so make downwards succession of their pure-blooded son possible. Indeed, royal incest is powerfully present in mythologies and there are examples in history. Still, these are exceptions: far more often kings in matrilineal ruling houses needed to accept sideways succession, which moved the crown to a son of one of their sisters or to a brother borne by the same royal mother. The matriliny cannot converge with the nuclear family; as a consequence, it does not have the option of limiting succession to sons or to the first-born son. In principle, the deceased king’s eldest remaining brother, the son of his eldest sister, or his eldest nephew, could be granted priority rights comparable to those of the eldest son in primogeniture; yet such fixed preferences were rare. Typically, matrilineal dynastic descent involved open sideways succession with numerous candidates entering the fray.

Usually, more than one royal woman represented the royal line, and these women were likely to bear children. High-ranking matrilineal princesses uncomfortably fitted marriage arrangements privileging the male, and they have been known for their repeated marriages, sexual promiscuity and full-fledged polyandry – this was the mirror image of the patrilineal harem. Their rank mattered more than the status of their partners, irrelevant for the royal bloodline. Promiscuous relationships not only underlined the princesses’ hierarchical pre-eminence but also concealed father­hood, and thus complicated paternal claims to power if sons ascended the throne.¹²

Candidates competing for the throne in sideways succession strengthen the hand of kingmakers. Competitors tried to find support for their candidacy among elites and leading officeholders – typically including women. Competition could take violent shape, but forms of consultation, acclamation and election were practised in many realms. Matrilineal descent engenders open sideways succession, complicates the concentration of power in the hands of a core group over time and leaves room for kingmaking (Douglas 1969, 124–127). In these ways, it is the antithesis of dynasty as it is usually understood in traditional historiography.

The contrast sketched here between matrilineal and patrilineal dynasties shows dispositions rather than inevitable outcomes. Patrilineal dynasties, too, could choose to grant succession rights to the wider royal clan, to scions of junior lines or to all descendants of a legendary founder. A shared sense of family charisma has been attributed to nomadic steppe dynasties, most particularly to the Mongols, where succession retained an element of competition and acclamation. The stunning battlefield success of a distant relative could be interpreted as divine election. It made sense to leave room for multiple successors who needed to prove their capabilities: this could bring a war leader to the throne. Conversely, however, rulers and their widows might express their preference for one of their sons, often the eldest. All candidates, finally, needed confirmation from a tribal council. Within patrilineal houses, all forms from entirely open to fully fixed are possible, yet the presence of the father-son nexus in terms of descent rights as well as fatherhood promoted downwards succession and invited the restriction of eligibility and the fixing of succession rules. This disposition is absent in matrilineal dynasties. It is no coincidence, therefore, that the most enduring dynasties have been patrilineal (Douglas 1969, point 2 on 126).

In practice, mixed forms of descent have predominated. The Asante federation (1670–1902), among the most famous and best researched examples of matriliny, shows an implicit form of patrilineal descent behind the frontage of matrilineal succession. Kings’ sons could not themselves succeed to the throne, but their sons could marry royal women, and might thus qualify for succession in another generation. Grandsons regained what their fathers had forfeited. Matrilineal descent coincided with the repeated alternation of two ‘patrifiliations’, allowing for continuity and concentration of power (Wilks 1989). Patrilineal dynasties in Africa, conversely, often expressed succession rules in terms of prohibitions. Male primogeniture, a byword for concentration of power, was prevented by barring succession for the king’s eldest son or even for all royal sons (Beattie 1960, 13, on ultimogeniture; Drucker-Brown 1975, 135–146, on the general prohibition against king’s sons and other rules; Goody 1966, 172–175, on circulating succession; Ottino 1983, 254).

Descent had an immediate impact on political connections and roles. Succession rights generate tensions among eligibles belonging to the same descent group. Connections established through marriage outside of the descent group were of particular importance for alliance building. The absence of succession rights made spouses and their families particularly important as allies: they were less likely to act as usurpers and thus more reliable as confidants and agents. Trusted advisers could be wedded to royal daughters, particularly where succession through women was ruled out. The same principle worked for princes in matrilineal settings: they were no longer their father’s rivals and would be used to cement alliances. Forms of descent dictated political roles. Examining offices and residential patterns in the Bamum kingdom of Cameroon, Claude Tardits noted the role of ‘palatial’ families attached to the dynasty through alliance in high political office at the centre. At the same time princes from the patriline dominated in regional chiefdoms. In West African Abron, similar divisions can be found, yet now with matrilineal princes holding chiefdoms and ineligible royal sons acting as the king’s confidants and agents. Government meant striking a balance between the high-profile powers of potential successors and the more pliable support of affines (Tardits 1987, 17, 29–31, 114–117).

Alliances, moreover, allowed ruling houses to attach important social groups to their dominion. Polygynous reproduction, common for matrilineal as well as patrilineal dynasties, offered plenty of opportunities to create cohesion through the exchange of women. This could be done on a voluntary basis, through violence, or with a mixture or both. In many African realms, the paramount king received daughters from chiefs who expected brides in return: the political integration of elites was contingent on the permanent exchange of women. Mughal emperors included the wives and daughters of conquered territories in their harem and redistributed women to their most loyal servants. Chinese dynasties carefully monitored the alliances of emperors, princes and princesses, calibrating their policies by considering the record of earlier dynasties. Ottoman concubines were largely of indifferent slave background, yet the daughters of the sultanic concubines were wedded to leading state servants – a practice common among Islamic ruling houses.

European monogamy soon led to the creation of dynastic networks of marriage and succession. Here, marriage first and foremost was the isogamous and exogamous alliance to scions of sovereign houses ruling other countries. This helps to explain the prevalence of international wars of succession in seventeenth- and eighteenth-century Europe, when the internal challenges to dynastic supremacy had abated. Princesses moved across Europe to become queens. These women were stuck in a ‘patrilineal puzzle’. They stood uncomfortably between loyalties of descent, alliance and motherhood. Arriving as outsiders at the courts of their husbands, they were more likely to integrate only after the birth of a child, especially a male heir to the throne. The presence of a son tied them to the interests of their acquired dynasty. Marrying dauphin Louis in 1770, Marie-Antoinette embodied for many French the unpopular alliance with their erstwhile arch-enemy, the Habsburgs. The young dauphine soon-to-be queen saw her reputation eroded further when she initially failed to produce offspring. Stories about Louis XVI’s impotence and the queen’s salacious adventures started to circulate. At the same time, Maria Theresa, Joseph and their ambassador in France scrutinized her every move from the Habsburg perspective. Whether or not the queen acted ill-advisedly, the predicament she shared with numerous other young queens immensely complicated matters for her. Wedged in between two competing royal houses, the queen’s position was necessarily awkward. European dynastic intermarriage was not without its complications for male spouses. Princes moving to other courts to marry a queen regnant, or a princess likely to succeed to the throne, tended to be received with grave concerns. How could a male prince accept a subsidiary role as consort? Should the ruling queen, as a good wife, accept the guidance of her husband? Prohibitions of female rule such as the Salic law frequently invoked in France and the Holy Roman Empire prevented this quandary; alternatively, queens postponed marriage until it became impossible – more than a few virgin queens chose this option, and thus caused their house to end with them.

Matriliny and the limitations of kingship

Dynasties were never simply a biological fact. Cultural rules defined membership, procreation and succession in diverse ways. Such hallowed conventions, however, were often circumvented or downright ignored. The ad hoc solutions resolving political crises caused by demographic mishaps or dynastic strife reflected necessities, opportunities and power balances rather than normative instructions. Conformity between rules and practices, seamlessly restored in narratives and genealogies, can never be taken for granted. The outspoken differences between ‘systems’ of dynastic descent appear gradual and flexible rather than absolute. Nevertheless, the questions that matriliny raises about the concept of dynastic power, implicitly derived from the patrilineal model, deserve to be extended one step further. Did matriliny, with its disposition towards open forms of succession and the marked presence of kingmakers, encourage the emergence limited and contractual forms of kingship?

It is necessary here to first clear the ground. Too often the elevation of kings to a semi-divine position has been equated unthinkingly with near-omnipotence. In practice, sacrosanct status engendered limitations and the conspicuous presentation of hierarchical pre-eminence burdened the incumbents. Everything related to the royal body was treated with the greatest circumspection. Prohibitions were particularly strong in Africa: kings should never touch the ground and dire consequences could result from hearing their voice or catching the royal gaze. Breaking such rules might endanger the balance with ancestors and heaven, the realm, the transgressor and the king himself. More than a century ago, James Frazer eloquently described the burden of sacrality: ‘like a fly in the toils of a spider, [the king] could hardly stir a limb for the threads of custom’ (Frazer 1894, 209–210). Recently, Marshal Sahlins and David Graeber, combining their forces in a remarkable study On Kings noted about the royal predicament:

The chief weapon in the hands of those who oppose the expansion of royal power might be termed ‘adverse sacralization’ – to recognize the metahuman status of the monarch, to ‘keep the king divine’ (Richards 1968), requires an elaborate apparatus which renders him, effectively, an abstraction, by hiding, containing, or effacing those aspects of his being that are seen as embodying his mortal nature. Kings become invisible, immaterial, sealed off from contact with their subjects or with the stuff and substance of the world – and hence, often, confined to their palaces, unable to exercise arbitrary power (or often any power) in any effective way. (Graeber and Sahlins 2017, 8; citing Richards 1968)

While I would hesitate to view sacrality as a ‘weapon’ actively used by others to restrict royal powers, it is clear that exceptional status could severely restrict freedom of action and thus powers. All kings and queens in one way or another combine two different functions: reigning and ruling; symbolic representation and political action. Some, like Japanese emperors, were pushed wholly in one direction: they retained supreme symbolic sovereignty yet were rarely involved in political action and remained distant from the public. This aloofness helps to explain the remarkable longevity of the Japanese imperial dignity (Shillony 2005). East Asia and Africa show the most outspoken examples of hidden and immobilized kingship. Elsewhere, kings more often acted as hands-on rulers, sacrificing their aura of unperturbed dignity by maximizing their actual powers: examples from Central, West and South Asia as well as Europe come to mind. Everywhere, however, political action compromised the purity of the ideal king; vice versa, it was difficult to reconcile untarnished symbolic pre-eminence with political wheeling and dealing. Reigning and ruling could be apportioned between the ruler and his leading advisers or generals. Alternatively, princes could find a middle way. Even active rulers would embrace the calendar of rituals present in all kingdoms to underline their special status – as an indispensable component of their kingship rather than to fool their subjects. Everywhere, the balance between reigning and ruling changed within the lifecycle of kings. Boy-kings started out as youngsters under the tutelage of their elders, ascended to full power during the decades of maximum physical and mental powers, and gradually subsided into the anxieties and dependencies of old age. In short: active rule cannot be taken for granted, and royal powers cannot be inferred from the elevated status of the incumbent.

For a variety of reasons most royals in history were far less powerful than modern audiences have imagined them to be. John Beattie opened an illuminating chapter about the checks and balances facing kings in Africa with the following sentence:

Older writers about primitive states in Africa and elsewhere often spoke of chiefs and kings as possessing absolute power. But it is plain from the more thorough ethnography of the past half century or so that in fact the authority of such rulers is generally restricted by a wide range of social institutions. (Beattie 1967, 35)

Replacing African kings with European royalty or Chinese emperors and ethnography with history, the statement rings true in the early twenty-first century. Representations of royal power, through high-handed writings, artefacts and buildings have helped to convey an image of strength and grandeur; revolutionary antiroyalism and the grand narrative of modernity have confirmed the overstatement of royal power, tinging it with exploitation and oppression. Moving from the heavily redacted prescriptive sources generated by literati at the centre of power to archival legacies and local voices, a different image emerges. Wherever sources are diverse and plentiful, the court resurfaces as a world of contending groups and individuals: servitors, advisers and royal intimates battled to obtain favours for their families and clients. Their activities were veiled by deferential behaviour and negated by the collective performance of royal omnipotence. The polite tone of local elites vis à vis the agents of central power, likewise, did not preclude bargaining. For them, it was important to ward off royal ire, yet at the same time maximize local rights and benefits. Describing the connections between Chinese imperial authority and local elites, Michael Szonyi summarizes their attitude as ‘neither simply representing state interests nor opposing them but rather manipulating the system to serve their own interest’ (Szonyi 2002, 205–206). This aptly characterizes the minimum, rather than the maximum case of elite resilience. At times the distant centre, notwithstanding its imposing show of power, could simply be ignored.

Beattie’s statement, however, did not refer to such general limitations of power: he specifically points to institutions checking the abuse of power by rulers: king-making during interregnums, by councils, relatives and queen-mothers; oath-taking during accession ceremonies; the enduring possibility of admonition or deposition by elite bodies and sometimes even by the populace. The preceding paragraphs pointed out the correspondence between matriliny, open sideways succession and the marked presence of high female office-holders acting as kingmakers. Contesting candidates needed to generate support for their ambitions, whether by assembling groups of armed supporters or by mollifying elites. In such cases, elite rights were confirmed and advantages were granted. Not unlike modern election promises, these pledges could be withdrawn silently after power had been secured. Nevertheless, the institutionalized presence of kingmakers made the emergence of a horizon of expectations regarding the choice of the king as well as the performance of kingship likely. Eventually, it could engender contractual kingship, where dysfunctional kings might face the risk of dethronement. Asante political practice reminded the British anthropologist R.S. Rattray of ‘Greek democracy’, with consultations at all levels of the political edifice and the accepted practice of ‘destoolment’ (Rattray 1929, 401–406). Yet councils of elders and queen-mothers acting as kingmakers can be found in patrilineal kingdoms with open forms of succession. Outspoken examples of dethronement, likewise, are not limited to matrilineal dynasties. The Sultanate of Buton, in Sulawesi in the East Asian Archipelago, allowed hereditary succession only in rare cases, and usually selected a new sultan from among the patrilineal descendants of the first royal couple (the kaomu). Interestingly, the Wakala, kingmakers, as well as dignitaries who checked the sultan’s behaviour, were recruited from the second-ranking estate, who were ineligible for the sultanate. The accession ceremony made it painfully clear that the new sultan needed to respect conventions: on the festive parasol used in the ceremony a tasselled cord was hung, signifying the possibility of execution through strangling.¹³ An additional explanation can be given in this case: Buton’s sultans were ‘stranger-kings’, outsiders who had ascended to the throne and in the process accepted checks on their power (Henley and Caldwell 2008, 278). Not all stranger-kings accepted limitations: military force and ruthless violence, not uncommon among the qualities attributed to stranger-kings, cannot easily be aligned with contractual and consensual government.

Contractual kingship was certainly not limited to matriliny; neither were all stranger-kings tamed by contracts and institutions. Sideways, open and competitive succession, characteristic for matrilineal succession but present in patrilineal kingdoms, appears as the decisive factor. The concentration of examples in Austronesia and Africa can be interpreted in several ways. Matriliny may have been far more widespread here in an earlier phase of history, which helps to explain the persistence of typically matrilineal customs among patrilineal groups. Alternatively, this aspect of matrilineal succession simply appealed to contiguous patrilineal groups eager to prevent the concentration of power in the hands of a single family. Finally, the presence of institutions limiting royal powers may be seen as typical for smaller-scale chiefdoms and kingdoms, where distances were limited and connections dense. None of the larger consolidated empires practised matrilineal descent at the level of the ruling house. Such empires often did have ‘ombudsman systems’ as well as institutions exerting some control (the Chinese censorate; religious authorities), yet limitations on the ruler’s power arose as the ad hoc outcome of balances with elites and advisers rather than as a consequence of fixed institutional procedures. In Europe, corporate groups cherishing their hereditary rights and privileges were markedly present, and in several realms sovereigns came to power through election.

Open succession leads to institutionally circumscribed royal power most particularly in cases where interregnum violence remains limited and where the peaceful involvement of kingmakers is accepted. Open but violent succession, as in the case of competing princes in the Ottoman and Mughal empires, does not necessarily dispose the victor to respect the rights of competitors and counterforces – not, in any case, beyond the level of paying gratifications to the soldiery. Only after paying accession donatives to the janissaries and other household troops could the new Ottoman sultan comfortably sit on the throne – and although the veiled threat of violence might prompt him to take into account the interests of these groups, few rules or institutions explicitly circumscribed his powers (Murphey 2008, 86).

Might a reduced level of violence be related either to women on the throne, or to matrilineality? Consensual attitudes have been ascribed to women rulers in South East Asia (Amirell 2016). Elsewhere, too, queens on the throne have been described as eager to placate elites and please their peoples. Again, it is difficult to accept such generalizations: the annals of dynastic power include ferocious women as well as gentle men. Taking into account the negative bias towards women in power prevalent among literati chroniclers, the sources still strongly suggest that Wu Zetian acted ruthlessly towards her servants as well as her potential rivals, many of whom were purged and executed (Twitchett 2003). Her record may not have differed much from those of male emperors who escaped criticism and vilification, yet it hardly matches the cliché of the merciful queen. While the numbers of women queens like Nzinga who actually led their troops into battle may be limited, we have numerous examples of women as astute political operators and war leaders, remembered for their competence in organizing defence.¹⁴ As we have seen, princes acting as sage-kings and figureheads can be found in many places. Active male rulers, too, could take the normative guidebooks seriously and strive to establish consensus with their peoples. It makes no sense to ascribe pacific attitudes and a disposition for consensual decision making to all women on the throne or to assume that all men tended to be violent and irresponsible. More than men, women were confronted with a legitimacy deficit: the assumption of a mismatch between women and supreme power complicated the acceptance of women rule by elites and the populace. This predicament could lead queens to either prove their equivalence to males by conspicuously exerting power and using violence, or, conversely, by showing eagerness to placate elites and please their peoples. Even queens striking a careful balance between toughness and compassion could expect to be perceived in terms of one of these extremes.

The different dispositions of matriliny and patriliny can fruitfully be connected to earlier discussions about the factors stimulating and limiting violence. In response to Norbert Elias’ model of civilization as Affektbeherrschung, the Dutch anthropologist Thoden van Velzen expounded the elaborately ‘courtly’ behaviour of the Surinam Ndyuka (or Okanisi). He showed that commonly accepted explanations (state formation, religious mission, colonial government, etc.) could not account for this pattern of behaviour and proposed a far more precise and contextual alternative. Matrilineal descent, combined with polygyny and matrilocal residence, forced men to shuttle back and forth between their village of birth and the villages of their spouses, catering to an unsettling variety of demands in these different places. In the process, they were forced to become diplomats, communicating in ways that bring to mind the hypocrisy, or ‘unaufrichtige Kommunikation’, often ascribed to courtiers (Winterling 2004). These intricate social patterns broke up the concentration of agnates in the ‘fraternal interest group’, and this, Thoden van Velzen states, can be accepted as one of the main factors in reducing violence within (rather than between) societies (Thoden van Velzen 1982 referring to Paige and Paige 1981). Likewise, the impact on violence of matrilocal residence patterns in a patrilineal society has been suggested for the Heian period in Japan (Nickerson 1993).¹⁵ The criss-crossing effects of descent and residence moved apart agnates and so helped to reduce conflict. This can be seen not only in patterns of residence and land administration, but also at the highest levels of political power, through the entanglement of Fujiwara regents and emperors.

Epilogue

Women in supreme power have been the exception in the historical record. They rose to full sovereignty mostly because of the absence of male candidates, and invariably had to cope with the male qualities associated with kingship. While queen-mothers could veil power with maternity and hide behind the presence of a male heir, queens-regnant faced difficult choices between stereotyped gender roles. This seems to have been a timeless predicament, yet the levels of acceptance of female sovereignty and its frequency show regional variation. Queens-regnant were markedly present in Austronesia and Africa, figured repeatedly in Europe, yet remained a rare exception in Asian land empires. A connection with patterns of descent in the male and female lines appears plausible, and this hypothesis is strengthened by recent publications on early ‘female-biased kinship’. The regional and temporal prevalence of matriliny and its consequences for the presence of women in power deserves further research.

Patrilineal descent has greater potential to concentrate power in the hands of the dynasty; hence it facilitates the rise of long-lasting large-scale empires. Political consolidation, coupled with the rise of world religions and the globalized economy, has reduced the presence of ruling houses practising matrilineal descent. Matriliny cannot concentrate wealth and consolidate power in quite the same way as patriliny. Up to this point the nineteenth-century view of matrilinearity as a ‘phase’ is correct. Matrilineal descent moves sideways and strongly tends towards diffusion of power; this can engender strong social cohesion and may reduce internal violence.¹⁶ Moreover, the connection between matriliny and the diffusion of office and wealth makes the emergence of consensual forms of power more likely. The criss-crossing tensions created by matriliny against the background of persistent male dominance – at the heart of Richards’ puzzle – show that there was more at stake here than the benign influence of women.

Among numerous other classics of anthropology, Jack Goody’s acute examination of succession and Audrey Richards’ puzzle can still help students of dynastic power to understand the dynastic tensions and alliances arising across the globe. The functional-structuralist bias of this generation of anthropologists may at times seem crude. Yet much can be achieved by combining their enviable clarity with the nuances and doubts acquired since the 1980s, both for anthropology and history. Moreover, current genetics and neuroscience are rapidly redefining our understanding of heredity and gender. It is high time for historians and anthropologists to add their share to this ongoing recalibration of biology and culture, acknowledging the impact of genetics while questioning the cruder forms of sociobiological reasoning.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 See in particular Schneider (1984) and a recent reappraisal: Sahlins (2013).

2 Key works on matrilineal descent: Richards (1950), Schneider and Gough (1974), particularly Schneider’s ‘Introduction: The Distinctive Features of Matrilineal Descent Groups’, and Aberle’s chapter; see also a lucid and concise discussion by Douglas (1969), Peters (1997); see a balanced overview by Johnson (2016).

3 See e.g. several volumes edited by David Warren Sabean et al. (2007, 2011, 2013) and also the contribution by Lutter et al. 2021; see a wide-ranging historical overview of the anthropological discussion by Gillian Feeley-Harnik (2019, 51–87).

4 Matrilineality and matri-locality/focality are grouped here as female biased kinship or FBK.

5 Jeroen Duindam (2016) presents these issues in four concentric circles, moving from the prince, via the family and the court to the realm; a brief outspoken statement on dynasty in world history can be found in Duindam (2019a).

6 With the brief tenure of two kings between the long reign of Ranavalona I (1828–1861) and those of her two namesakes Ranavalona II and III, reigning respectively 1868–1883 and 1883–1897.

7 See Knight et al.’s (2011) discussion of the early positive responses to Bachofen and Morgan by Marx, Engels and Durkheim, followed by a backlash among early twentieth-century anthropologists. Knight sees evidence for the predominance of matrilineal descent in early kinship. Another perspective likewise stressing the presence of matriliny is given by Jones (2011), who connects matrilocality not only to strong inner solidarity, but also to waves of expansion, a model based mostly on Austronesian and African/Bantu examples – and clashing with the interpretation offered by Knight. Jones sees matriliny as a new arrival, making possible the expansion waves, yet subsiding at the point the tribal format was pushed out by emerging states. The discussion continues, see Surowiec, Snyder, and Creanza (2019).

8 Note that Veth (1870) opens with the parallel of independent South Asian women with Berber women in Muslim North Africa.

9 Douglas (1969) lists a number of characteristics for matriliny and patriliny conforming to the image of concentration and diffusion sketched here. She also stresses the relatively open nature of descent and its consequent potential to attract and recruit others. Conversely, matriliny made it more difficult for ruling houses to persist and maintain their status as well as their grip on resources.

10 See numerous examples in Duindam (2016): incarceration in Solomonid Ethiopia or in the Seventeenth-century Ottoman empire; exile throughout Africa, blinding as a Safavid custom, chopping off hands among the Solomonids, execution in Buganda as well as in the sixteenth-century Ottoman empire.

11 See Peters (1997, 142): ‘The matrilineal puzzle is worth revisiting, but this time with a more explicitly gendered and historical approach’.

12 There is some resonance among patrilineal princesses, likewise known for a variety of partner arrangement uncommon among other women in society, see e.g. the discussion of Ottoman sultans’ daughters in the dissertation by Dumas (2013). On related themes see Peirce Leslie (1993); Parkin, 2021, gives a broad overview of historical and contemporary marriage arrangements.

13 Schoorl (1994) on strangling cord at 25–26; see more generally on contractual kingship and stranger-kings in Henley and Caldwell (2008), notably the editors’ introduction: ‘The Stranger Who Would Be King’, and in the same volume their ‘Kings and Covenants’ 269–291. See Beattie (1967, 372) on the Nupe and Buganda stranger-kings.

14 Amirell (2016) cites several examples; for Europe the list of Russian Tsaritsas plus Maria Theresa can be extended to include many other examples, see e.g. Helfferich (2013); note the changing interpretations of Queen Ranavalone I of Madagascar’s attitude and policies: Ayache (1975), Campbell (1987).

15 See an attempt to resolve these issues through larger-scale cross-cultural comparison, perhaps relevant mostly for the pertinent questions it raises: Ross (1986).

16 Conversely, see the association between matriliny and expansion expounded by Jones (2011), where matrilocality is not only connected to strong inner solidarity, but also to waves of expansion, a model based mostly on Austronesian and African/Bantu examples.