The Invention of Endogamy

The term “endogamy” was immediately popular from the time of its invention in the 1860s. The Scottish jurist and ethnologist John Ferguson McLennan coined the term along with “exogamy” at a pivotal moment in the mid-nineteenth century when the comparative ethnology of civilizational development was coming into vogue (Kruper Reference Kruper1988; Stocking Reference Stocking1987). As the first universalist scientific terms for the study of marriage, McLennan's concepts of endogamy and exogamy were authoritative and soon became the language of many studies of kinship and social structure. McLennan argued that this new vocabulary offered a novel universal language for the study of civilizational development through marriage practices.

Little has been written on McLennan, particularly in terms of the profound influence that his work had on the social scientific study of India, despite the importance of his concepts in modern social theory. His legacy is clear in the proliferation of concepts that he created, from exogamy and endogamy to totem and taboo—terms that shaped Charles Darwin's theory of man's descent and, later, Émile Durkheim's sociology of religion (Seth Reference Seth2016, 511). McLennan even lurks in the shadows of some of the most influential theories of colonial knowledge. Take, for example, Edward Said's Orientalism (Reference Said1978, 235–37), in which Said reads William Robertson Smith's Kinship and Marriage in Early Arabia (Reference Smith1885) as a paradigmatic text that shaped generations of knowledge about Islam and the Orient. Smith's (Reference Smith1885, v) study begins by declaring its indebtedness to McLennan's theory of primitive marriage, noting that marriage in Arabia “corresponds in the most striking manner with the general theory. . . . in the late J. F. McLennan's book on Primitive Marriage.” In fact, McLennan's Primitive Marriage would go on to influence a broad range of studies of colonized societies, including colonial Indological investigations of Sanskrit and the disciplinary study of the history and sociology of caste, race, and Aryanism by major Indian intellectuals throughout the twentieth century. Considering the pervasive citation of McLennan in foundational colonial and postcolonial texts, the absence of a robust engagement with his concepts is striking.

In Primitive Marriage, McLennan argued that matriarchy was the first form of social organization after the absolute promiscuity that he characterized as “primitivity” (Kruper Reference Kruper1988; Lyons and Lyons Reference Lyons and Lyons2004), which McLennan's contemporary, Lewis Henry Morgan (1818–81), would later call “primitive promiscuity” in Ancient Society (Reference Morgan1877). For McLennan, the origins of civilizations were not to be found solely in European mythology or ancient legal texts of the Romans or Indians, as had been the respective arguments of contemporary ethnological texts such as Johannes Bachofen's Das Mutterecht (Mother right, Reference Bachofen1861) and Henry Maine's Ancient Law (Reference Maine1861). Instead, McLennan (Reference McLennan1865, 10–11) offered that the colonial ethnographic study of nineteenth-century “barbarism” provided examples of civilization's primitive past that survived in the tribes of the colonial world. Only after “observed” social practices in remnant “tribal societies” were understood could the social scientist correlate them with histories of social institutions provided in classical Sanskrit and Latin texts.

Primitive Marriage was, in McLennan's (Reference McLennan1865, v) words, “an exercise in scientific history,” the first study to posit marriage as the primary lens through which to understand “the early history of civil society.” In the book, he took as his object the sexual practices of so-called primitive society, offering a critique of legal scholar and Indian colonial administrator Henry Maine's thesis that patriarchal monogamy was the most ancient social institution found in ancient Latin and Sanskrit legal texts (Mantena Reference Mantena2010). McLennan insisted that the marriage practices present in the surviving “tribes” of the nineteenth century, such as matriarchy and bride capture, were essential to understanding the stages of civilizational development and offered more accurate insights into primitive sociality than classical texts. Hence, McLennan analyzed the “data” from colonial ethnographies of sexual practices of “tribal” peoples of British India, which he compared with “native” peoples of central Asia, Australia, and the Americas. He made extensive use of generalizations about Indian peoples and social practices from early British colonial ethnographies and philological studies of Sanskrit text to create a universal theory of primitive marriage in which India functioned as a primary empirical referent. According to McLennan (Reference McLennan1865, 144–45), Hindu endogamy was a later stage of Indian civilizational development, in which the “advanced” nature of caste endogamy was proven in the textual evidence of The Laws of Manu.

In McLennan's theory, the control of girls and women through the regulation of female sexuality was the primary measure of a society's social evolution. He argued that at the origin of human civilizations was an open system of sexual promiscuity in which women were held “in common,” a euphemism used in ethnology to objectify female sexual promiscuity, whereby a “common woman” had sexual relations with multiple men (Reference McLennan1865, 169). So-called primitive societies evolved from complete promiscuity to become loosely organized by female polyamory and systems of bride capture.

Emphasizing the importance of a polygyny—a women-centered model of biological descent—McLennan describes a female sexuality that was unrestrained compared with the “developed” sexual restraint of women in strict structures of patriarchal marriage and descent that characterized socially evolved societies. In particular, in the section on polygyny, he cites the hill tribes of northeast India and the Nair caste of Kerala as exemplary of “primitive” matrilineal forms of familial organization that existed before more advanced Hindu patriarchal forms (McLennan Reference McLennan1865, 179, 184, 196; see also Arunima Reference Arunima2003). This narrative of the transition from the promiscuity of women to their capture is the point in the text where McLennan invents the ideas of endogamy and exogamy. The terminological definition appears in a footnote in Primitive Marriage, in which McLennan (Reference McLennan1865, 48–49) argues for the necessity of the new terms he invents:

In McLennan, both forms of marital organization are based in the control of women's reproduction. As he states, new terminology is required to describe a system of social organization that was to be differentiated from caste. For McLennan, caste is a textually dictated “social custom” that appeared in a later stage of civilization in Hindu society. On the other hand, his new terminology offers a more universalist approach to the status of marriage as the marker of any civilization. While caste is particular to India, endogamy is universal.

The terms “endogamy” and “exogamy” were created by McLennan using the suffix “-gamy,” derived from the ancient Greek word gamos. While the suffix came to mean “marriage,” its etymology was based in the botanical sciences, denoting pollination, fertilization, and reproduction. As McLennan points out in his footnote, the “parallel” science of botany offered classificatory resources to help develop a scientific vocabulary for human marriage. This relationship of botanical classification to the social study of colonized societies predated McLennan's ethnological theory. Such deployments of botanical terminology were not new to colonial social theorists. They can be found from the earliest moments of Indological study in the work of philologist and founder of the Asiatic Society, William Jones (1746–94), who cited Carl Linnaeus and coined the Latin botanical term Polyandrian Monogynian (multiple men, one woman) to explain the natural tendency of Indian women toward sexual promiscuity (Jones Reference Jones1876, 3).

McLennan (Reference McLennan1865, 27) argued that textual examples of exogamy through bride capture could be found in ancient Latin texts on Rome as well as in the “sutras,” with present-day “primitive capture” and “female infanticide” scientifically exemplified in Major Samuel Charters Macpherson's 1842 report on the Khonds of Ganjam and Cuttack from India (McLennan Reference McLennan1865, 27; MacPherson Reference Macpherson1865). Both endogamy and exogamy could be found in primitive societies, but, McLennan thought, endogamous castes developed later, among upper-caste Hindus who were of Aryan descent. In this evolved system of patriarchal endogamy, men could reproduce outside of and maintain endogamous lines through patrilineal descent, but endogamy required the absolute control of female reproduction within the clan or tribe through intratribal marriage. Hindu caste endogamy represented a more “advanced” stage of civilization, after primitive promiscuity and early primitive endogamy, in which women's sexuality was held in check through caste. It is in McLennan's conception of caste that philological authority and ethnological theory united, for McLennan (Reference McLennan1865, 105–6), like his contemporaries, used the claim of evidence from The Laws of Manu and early colonial ethnographies from India to demonstrate a later stage of civility in Hindus by the interdiction against intermarriage across groups.

McLennan's theory of endogamous and exogamous marriage was built on a logic of population control. In his scheme, “endogamous” primitive societies sought to control reproduction and less productive members of society through female infanticide, as women, in McLennan's view, were fundamentally less “productive” than men in a society. In his characterization of primitive societies, McLennan argued that widespread female infanticide was used to control and erase the possibility of girls who threatened the social order as unproductive bodies, an idea that McLennan took from one of his favorite thinkers, Thomas Malthus (1766–1834). Malthus in his influential 1798 essay on population had argued that abortion and infanticide emerged in evolving societies as a way to control populations to ensure access to resources. McLennan's (Reference McLennan1865, 138–39) twinned concept of exogamy and endogamy took this Malthusian reasoning of the utilitarian purpose of population control based in female infanticide as a necessary precondition for the ethnological study of marriage. According to this logic, female infanticide was necessary for civilizational evolution, as girls and women were fundamentally more resource consuming and less productive in society. Note that for both Malthus and McLennan, women's labor—whether it was reproductive, caretaking, or domestic—was insignificant in the evaluation of the productive value of women in a society.

For McLennan, it was the loss of a group's population of girls and women through female infanticide for “resource distribution” that necessitated the theory of bride capture for sexual reproduction—the marrying of women from outside of a kin group or clan. McLennan framed this evolutionary development through the notion of “bride capture,” a seemingly neutral social development that pairs the terms “bride” and “capture.” McLennan couched sexual violence in benign language, suggesting that the forced abduction and assault of women was an evolutionary novelty, when a “bride” was essential for the civilizational transition to exogamy.

McLennan (Reference McLennan1865, 306) argued that this transition to the “capture” stage of social evolution was first symbolized in the “rape of Sabine women” in the founding of Rome. The rape at the origins of civilizational development toward monogamy became what Sigmund Freud (1856–1939), building on McLennan, would describe in Reference Freud1913 as the “incest taboo” that led to the necessity of marrying outside of one's own clan. This idea of female infanticide and taboo against incestuous sexual relations would later be essential for Claude Levi-Strauss (1908–2009) in his 1949 alliance theory, built as it was on exogamous marriage as the primary social alliance across groups because of the evolutionary novelty of distant sexual relations. McLennan's theory of marriage was thus the genesis of major theories that relegated endogamy to an early stage of civilization that was followed by taboo and symbolized in the totem among more civilized peoples. In McLennan's theory, these forms of totemic symbolism and exogamous “abduction” are retained in later stages of civilization as remnants of a primitive past that demonstrated the advancement of modern civilizations from their prior forms.

In proposing the novel evolutionary development of exogamy, McLennan sought to resolve the sticky moral issue of placing “rape” at the heart of his theory of the evolution of marriage. In McLennan's theory, the act of rape was the genesis of exogamy, a novel event of social evolution. To resolve this issue, where sexual assault was cast as benign capture, McLennan argued that so-called primitive women were not exactly “raped.” Indeed, according to McLennan (Reference McLennan1865, 21–23), so-called savage women not only consented to sexual relations with men outside of their clan but were, in fact, naturally inclined to these forced sexual acts:

In this moment of his text, McLennan proclaims in his rejection of the “primitive prudery hypothesis” that women are, in their most primitive form, “depraved,” even “inured” to sexual connections with multiple men from infancy. The “primitive” female body is endowed with sexual excess from the moment of its existence, to be controlled either by interdictions against sex outside of the clan or by the sexual exploitation of exogamous abduction.

McLennan's Primitive Marriage betrays the patriarchal foundations of these concepts and the many modes of violence against girls and women that are naturalized in the making of the concepts of endogamy and exogamy. The control of female sexuality appear as a necessary precondition for evolution throughout McLennan's thought, originating in the assumption that the origins of civilization were defined by women held “in common,” where the hallmark of primitivity was female sexual promiscuity. Perhaps most strikingly, McLennan offers a definitive theory of social evolution in exogamy built on the forced abduction and socially sanctioned death of girls and women. The diminishment of womanhood, where women appear as objects of exchange on behalf of social development, is critical throughout McLennan's theory of primitive marriage: female infanticide is a natural manifestation of evolved processes of population control, the control of women's sexuality is necessary to maintain endogamy, the sexual abduction and forced marriage of women is an evolutionary novelty in the development of exogamy, and sexual violence of abduction is, in McLennan's telling, welcomed by the “depraved” girls and women of “primitive tribes.”

I lay bare the violence of these explanatory narratives for these universalist models of marriage that exist within McLennan's thinking not to condemn a thinker like McLennan for his racism and misogyny, for McLennan was not alone in his propagation of patriarchal colonial thinking, which pervaded nineteenth-century theories of comparative civilizations. Indeed, his Primitive Marriage is but one of many studies in the widely influential field of ethnology that staged civilizational progress through a diagnostic of social institutions that controlled women's sexuality. Yet even a brief sojourn into the patriarchal origins of his concepts such as primitive marriage, endogamy, exogamy, and taboo exposes the insidious logic of the control of female sexuality that sits at the heart of these concepts, which live long beyond McLennan's theory of marriage and continue to be used uncritically in theories of society in the colonial and postcolonial world. This history is significant for the appearance of these concepts in the studies of social life in India that follow.

Endogamy and the Study of Indian Society

It is from McLennan's Primitive Marriage that Risley, colonial administrator and author of Tribes and Castes of Bengal (Reference Risley1891), took the categories of endogamy and exogamy and made them the primary framework to understand Indian society. In Tribes and Castes, Risley utilized a form of circular reasoning in which colonial ethnography from India functioned as both the citational apparatus for universalist theories of ethnology and as local exemplars of the universal applicability of ethnological theory in colonial contexts. McLennan utilized a wide range of colonial administrative ethnographies of India, and then Risley and the many studies that followed cited the same ethnographic studies as specific examples of McLennan's theory that proved the local manifestations of primitive capture, female infanticide, and endogamy.

In his definitive study, Risley argued that endogamy, and therefore marriage, must be the primary way to ethnographically study Indians. As he introduces the concept in his Tribes and Castes,

Here, Risley asserts that special attention is to be paid to marriage in the study of Indian society, because it is the foundational form of organization at every level of Hindu society. If caste defines Hindus, and caste and tribe can be measured through structures of endogamy and exogamy, then marriage, and therefore endogamy, are necessarily the primary descriptive categories and subjects for analysis for any study of Indian society. Thus, Risley, building directly on McLennan, establishes that marriage was the sole way to describe and comprehend all facets of Indian society.

Bringing together philological studies of Aryan society, the ethnology of Aryan migration through his historical account of caste, and hundreds of responses to his questionnaires from European and largely upper-caste Indian administrators, Risley characterized the division between endogamous and exogamous groups through ideas of caste and racial difference. The questionnaires structured the responses to conform to Risley's narrow vision of caste as wholly organized through the structure of marriage, which defined the difference between Aryan and non-Aryan descent (Mitra Reference Mitra2020, 62). As Risley (Reference Risley1891, xliii) argues early in the text, “[a]t the bottom of the social system, as understood by the average Hindu, stands a large body of non-Aryan tribes and castes.” These peoples were representative of an earlier stage of civilization, prior to the Aryan invasion of India, which he describes using the language of McLennan as the “totemic exogamous septs,” while the next “totemic” group, higher in this hierarchy than the lowest “non-Aryan” groups, were the tribal peoples of West Bengal and Orissa, who represented “an early stage in the course of development by which a non-Aryan tribe transforms into a full-blown caste” (Risley Reference Risley1891, xliii).

A rigid adherence to endogamy, for Risley, defined Aryan caste society, and strict dictates of endogamy among caste Hindu society explained the uniquely important place not only of caste difference but also of hierarchy in the very making of Hindu society. Unlike exogamy among “totemic” peoples, upper-caste endogamy was explained entirely through the hypothesis of Aryan migration, which created the first totalizing hierarchy in India: “Intermarriage with other ranks of the Aryan community was first discouraged and then wholly prohibited, and thus by degrees was developed the rigid law of endogamy which distinguishes the Indian caste system from other apparently similar forms of social gradation” (Risley Reference Risley1891, 143).

Three interlinked connections made in Risley's declaration were to become widely influential in the studies of caste that followed: Aryans practiced endogamy, endogamy was constitutive of caste, and endogamous castes were arranged in a social “gradation,” a hierarchy based in sexual control. While McLennan had differentiated the specificity of caste in Hindu society from the generalized, universal phenomenon that he labelled as endogamy, Risley argued that McLennan's concept of endogamy was, in fact, the best social scientific description to explain the existence and perpetuation of caste among Hindus. In the “tribes and castes” that he studied, Risley (Reference Risley1891, lv) found multiple exemplars of all of the features of McLennan's model of matriarchal descent, as “wives were captured, female infants slain, and all the incidents of primitive society as sketched by Mr. McLennan were in full force.”

Later, in Tribes and Castes in Bengal, Risley, again citing McLennan, explains the natural evolutionary appearance of female infanticide in the evolution of “primitive” peoples. He asks, why did primitive tribes in India, instead of remaining endogamous, go to the trouble of killing their own girl infants and then capturing other girls and women? His answer is Darwinian “natural selection,” which he explains made “it … easy to understand that in certain states of society a tendency to female infanticide would be a natural consequence” (Risley Reference Risley1891, lxv; emphasis added). For, if natural selection placed a taboo on marrying women within the clan, “girls would be useless to the men of the tribe as wives.” Risley argues that the “uselessness” of girls is proven in the case of India through Major Macpherson's 1842 ethnography (Risley Reference Risley1891, lxv–lxvi). Risley thus utilizes the very same “evidence” to localize Indian “custom” within McLennan's universalist theory of primitive marriage that McLennan had used to create his universalist theory from the “local” example of India.

Risley's innovation in adapting McLennan's concept of endogamy can be found in his elaboration of the idea of hypergamy, the constitution of intracaste hierarchy through exaggerated forms of endogamy. Risley took the term “hypergamy” from colonial administrator Denzil Ibbetson (1847–1908), a close collaborator who had conducted his own ethnographic questionnaires that influenced Risley's methods (Fuller Reference Fuller2017). Ibbetson cited colonial civil servant William Coldstream as the inventor of the term “hypergamy” in the Panjab Census Report (Risley Reference Risley1891, lxvii). Hypergamy accounted for the “excess” women of an endogamous caste through polygamy. Risley (Reference Risley1891, lxvi–lxvii) argued that hypergamy was a novel innovation on endogamy found in upper-caste Hindu society, exemplified by the practice of Kulin Brahman polygamy in Bengal. He theorized the example of the Kulin Brahman subcaste, suggesting that the scarcity of Kulin men, the excess of Kulin women, and strict adherence to endogamy created a system of expansive and exaggerated polygamy used to maintain the Kulin bloodline. Risley and his contemporaries would characterize this social structure, hypergamy, as persistently plagued by the problem of women's sexual deviancy, as Kulin marriage in their view led to the marriage of dozens of women to a single Kulin man. The Kulin women did not, by “custom,” reside with their husbands, and they were inadequately controlled because of their sanctioned existence outside of the domain of the conjugal home.

Indeed, Risley theorized that hypergamy was the origin of the custom of female infanticide itself in India. It was from hypergamous upper castes that Indian society as a whole had learned the practice of female infanticide to “control” the population of girls and women: “when the custom of infant marriage had once been started, under pressure of social necessity, by the families of the highest group, who had the largest surplus of marriageable daughters, a sort of fashion would have been set and would be blindly followed through all the grades” (Risley Reference Risley1891, xci). In this moment of Risley's argument, female infanticide appears as the primary way to maintain endogamy among the highest castes and becomes a replicable structure “blindly followed” by all other castes. The excess of girls and women, those who “disturb the balance of the sexes,” continuously haunt the strictures of endogamous marital structures with their sexuality (Risley Reference Risley1891, xci). Through his position as colonial ethnographer and census administrator, Risley institutionalized the use of ethnological terms such as “endogamy” as both descriptor and explanation for Indian social practice.

Following Risley were generations of colonial studies of castes and tribes, including influential studies by Indian social scientists on castes and race. An influential elaboration of Risley's theory in early formal studies of caste by Indian social scientists can be found in the work of a student of ethnology and Indology, Shridhar Venkatesh Ketkar, a Marathi Brahman scholar educated in Bombay. Ketkar completed his master's and doctoral degrees at Cornell University in the United States in 1911 and was perhaps most famous for his novels and his publication of the first encyclopedia in Marathi. Ketkar eventually returned to India to teach for years as professor of economics and universal jurisprudence at the University of Calcutta and eventually at the University of Pune. He became a key thinker in the complex intellectual milieu of early twentieth-century Maharashtra, which by the 1920s was the center of the expanding discipline of sociology in India, particularly between University of Pune and the University of Bombay.

Caste was a central issue for the Marathi intelligentsia, and Ketkar joined a network of upper-caste thinkers who utilized social science to produce claims of a reformed upper-caste Hinduism based in caste subjugation and exclusion. Upon his arrival back in India, Ketkar would publish his master's thesis as part of his book, The History of Caste in India, volume 1, Evidence of the Laws of Manu on the Social Conditions in India during the Third Century A.D., Interpreted and Examined, with an Appendix on Radical Defects in Ethnology (1909). Ketkar's History of Caste in India exemplifies the complex integration of ethnological terminology, Indology, and social exemplars offered in defense of conservative Brahmanical visions of a Sanskritized sociology that characterized much Indian social scientific writing in the decades that followed McLennan and Risley.

In his book, Ketkar proposed a novel theory of caste based in the Manusmritis to identify the “deficiencies” in American ethnological theories of race that distinguished civilizations through a hierarchy of physical characteristics like skin color. Ketkar critiqued Risley's reliance on racial phrenological and phenotypical observation, instead arguing that scholars should return to the pure exegesis of Sanskrit text to understand caste and race in India. The History of Caste in India builds on definitions of caste by French Indologist Émile Senart (1847–1928), British colonial administrator John Collin Nesfield (1836–1919), and Risley. These European conceptions of caste resulted from an intertwined history of colonial administrative exchange of ideas about caste and race. Indeed, Senart's theory of caste built on ethnographic studies by Nesfield and Risley. Risley's own understanding of endogamy was derived from Ibbetson and Nesfield's ethnographies of early nineteenth-century India. The study of endogamy thus depended on the circulation and repeated citation of a set of colonial ethnographic generalities about Indian society based in Aryan race theory and ideas of primitivity derived from colonial reports, Indological studies, and ethnological theories. These administrators all borrowed and cited one another, inventing and elaborating classifications from the footnotes of administrative reports and creating sociological generalities from the “customs” described in colonial censuses.

In offering his own definition of caste, Ketkar redeployed Risley's equation of endogamy with caste and caste with marriage, taking marriage as the primary object of study for Hindu society. In his study, Ketkar (Reference Ketkar[1909] 1979, 15) argues that there are two interdependent features of castes in India: “(1) membership is confined to those who are born of members and includes all persons so born; (2) the members are forbidden by an inexorable social law to marry outside the group.” The social organization that resulted from these principles could be described by its single distinguishing feature: endogamy. Endogamy created the possibility of another characteristic of caste, hierarchy, for caste groups were not only distinct from one another but could then be organized by rank. Strictures of endogamy produced these major features of caste, particularly hierarchy and descent. Endogamous hierarchy ensured blood lineage and the prohibition against intermarriage prevented any forms of descent outside of the clan or group (Ketkar [Reference Ketkar1909] Reference Ketkar1979, 27). For Ketkar, the principle of endogamy structured prohibitions of sexual relationships that crossed the boundaries of hierarchies of lower-caste men and upper-caste women.

In Ketkar's History of Caste, the hierarchy that defined caste existed only because endogamy created static divisions between groups and “social customs” established groups as higher in status. Groups were valued differently and reflected the unique history of India's Aryan past. Like Risley and many of his Indian contemporaries, Ketkar's theory argued that the lowest orders of the caste system were peoples who were closer in proximity to non-Aryan tribal peoples. As he notes, practices of exogamy found within castes were a clear indication of tribal primitivity: “The greatest number of castes which do not marry with each other are simply tribes converted into castes” (Ketkar [Reference Ketkar1909] Reference Ketkar1979, 28). Using the work of ethnologist Edvard Westermarck (1862–1939) in his 1891 study, History of Human Marriage—a study that built on McLennan's theory of marriage—Ketkar [Reference Ketkar1909] Reference Ketkar1979, 27) argued that endogamy was a natural mental state among higher orders of civilization.

As Ketkar explained, Manu laid out detailed violations of endogamy on the basis of female sexuality. The higher the caste of the woman involved in an intercaste sexual relationship, the more serious the punishment of a lower-caste man who violated the bounds of endogamy, whether the sex was consensual or not. In extended sections on “Varna and Marriage” and “Varna and Illicit Intercourse,” he laid out a theory of caste as a social institution enforced through the control of sexuality. Sex with an upper-caste woman could lead to a punishment of death for a lower-caste or outcaste man, whereas a Brahman man would only be fined with no corporal punishment if he had a sexual relationship, whether forced or not, with a lower-caste or outcaste woman (Ketkar [Reference Ketkar1909] Reference Ketkar1979, 148–52).

Ketkar's study, even as it claimed to reassess the caste system in the framing remarks, made a defense of an exclusionary Brahmanical Hindu hierarchy by arguing that caste stricture through endogamy was a highly developed system of social organization based in Aryan migration and verifiable through the Laws of Manu. For Ketkar, Manu's dictates on punishments for sexual transgression revealed the importance of sexual hierarchy based in caste endogamy. Ketkar's theory and the many studies of caste that followed from upper-caste Indian intellectuals utilized McLennan's universalist ethnological concept of endogamy to theorize the Indological study of caste and race, marrying racial theories of Aryanism with theories of marriage and hierarchy.

As Brahman scholars like Ketkar produced the study of caste as a social scientific object, they created theories that systematically normalized and naturalized caste difference and discrimination through the language of objectivity based in evolutionary science and Indological origins. European and upper-caste elite Hindu intellectuals systematically obfuscated the political questions of caste, gender, and sexual violence through the language of social science, thereby producing caste hierarchy and upper-caste patriarchy as objective and unchanging facts. In doing so, they disappeared practices of domination and subordination by theorizing endogamy as a textual dictate and ethnological stage of civilizational development, mindlessly enacted and adhered to by all of Indian society.

Sexuality and the Political Imposition of Endogamy

One of the most powerful critiques of the Indology of endogamy can be found in the work of Bhimrao Ramji Ambedkar in his essay “Castes in India: Their Mechanism, Genesis and Development” (Reference Ambedkar1917). It is one of the earliest publications of Ambedkar, renowned Marathi Dalit intellectual and anticolonial political leader, jurist, economist, and primary architect of the Indian Constitution. Strikingly, in this early foray into ethnology and social theories of value, Ambedkar offered foundational critiques of the gendered power of caste domination and the hidden bias of Brahmanical epistemologies in the social scientific study of India. This framework critically shaped Ambedkar's later critiques of caste based in the subjugation of women and his deconstruction of Brahmanical claims to objectivity in social sciences and the law.

In “Castes in India,” Ambedkar synthesized studies of endogamy with his own historical and political critique of caste hierarchy. He presented his research, in 1916, in a course taught by ethnologist Alexander Goldenweizer (1880–1940), a lecturer at Columbia University with whom Ambedkar took several ethnology courses. Goldenweizer himself wrote extensively on primitive marriage and totemism based in the works of McLennan. Ambedkar's essay appeared in Indian Antiquary, a popular Indology journal of the early twentieth century. In his essay, Ambedkar (Reference Ambedkar1917, 81) reframes the concept of endogamy, which he describes as the “key to the mystery of the Caste system,” as an institution of enforcement. Through an exegesis of the unresolved problem of excess female sexuality, Ambedkar argues that caste hierarchy was a perverse ideology of dominance based falsely in the Laws of Manu.

Ambedkar framed his argument by crediting the discipline of ethnology for explaining the phenomenon of caste. Building directly on Ketkar's own definition and citations for his definition of caste, Ambedkar repurposed Ketkar's citational apparatus of key ethnologies of caste to settle on his own explanation of endogamy by rereading the work of Senart, Nesfield, Risley, and Ketkar himself. Revisiting Ketkar's 1909 defense of Brahmanism, which up to that moment was one of the most important studies of the “history” of caste by a Marathi intellectual, Ambedkar reconceptualized endogamy as a political concept essential for an anti-Brahman critique of Indian social science. Reframing McLennan's attempt to distinguish endogamy from castes in India and Ketkar's notion of the prohibition of intermarriage, Ambedkar (Reference Ambedkar1917, 84) succinctly describes the ethnological development of caste as one of institutional imposition: “Thus the superposition of endogamy on exogamy means the creation of caste.” Endogamy, in other words, is enacted and enforced, rather than a natural evolutionary product.

Ambedkar's study of endogamy came after his extended coursework in economics, philosophy, sociology, and history before and at Columbia University, where he arrived as a student in 1913. Influenced in his early years at Columbia by his courses in economics, Ambedkar wrote extensively on economics in his first master's thesis, “Ancient Indian Commerce” (Reference Ambedkar1915) and in his second, “National Dividend of India–A Historic and Analytical Study” (Reference Ambedkar1916), a study of economic policy in the East India Company that formed the basis of his dissertation and book, The Evolution of Provincial Finance in British India (Reference Ambedkar1925). Economic theories of value, particularly Karl Marx's conception of “surplus” as the basis for exploitation, shaped Ambedkar's understanding of the “human institution” of endogamy. Also critical to Ambedkar's understanding of endogamy were a wide range of fields of thought from pragmatic philosophy to early sociology that linked abstract evolutionary theory and philosophy to situated observations of social practice.

It is from Ambedkar's declaration of endogamy as social imposition that he described as the most dangerous result of this social interdiction, what he termed the “Surplus Woman.” Surplus Woman, and the social position that Ambedkar called “Surplus Man,” were problems of people who resided outside of monogamous marriage, an excess that Ambedkar diagnosed as “surplus,” produced as a result of forced interdictions against intermarriage between castes through the enforcement of endogamy. Economic theories of value, the subject of his two master's theses on precolonial and colonial India, shaped Ambedkar's theory of the “human institution” of endogamy, reflecting the intimate relationship between ethnological thought and political economic theories of social relationships such as caste for India.

As Ambedkar (Reference Ambedkar1917, 85–86) establishes after declaring endogamy as imposition, the Surplus Woman is antithetical to caste reproduction, a persistent threat to the social imposition and perpetuation of endogamy:

The Surplus Woman, in other words, was an irresolvable excess, a “double danger,” a failure that defied the bounds of heterosexual reproduction and undermined the very replicability and boundedness of a group. In this logic, according to Ambedkar, in order for a clan or group to be endogamous, the sex ratio of man to woman needed to be 1:1 in order to create a reproducible system of heterosexuality without any spare members. Without a 1:1 ratio of man to woman, endogamy was persistently under threat.

These principles of social containment were required to deal with the “surplus” created by the potential failure of the “ruling ratio” that resulted in Surplus Woman and Surplus Man (Ambedkar Reference Ambedkar1917, 86). Institutions that protected conjugal reproduction, in the form of women's social and actual deaths, were the only avenue for “repairing the disparity between the marriageable units of the two sexes within it” (81). In this economics-based sociological theory of surplus sexuality, Ambedkar offered a critical diagnostic of “ratio” akin to Risley's theory in 1892 of caste and the “balance of the sexes.” That principle later would be called “sex ratio,” which by the second half of the twentieth century would be the “index” of social and racial development across much of the world, perhaps most significantly so in postcolonial India. Indeed, the heterosexual reproductive imperative of the measure of “sex ratio” in development policy was grounded in the colonial study of proportionate sexes for caste endogamy.

For Ambedkar, the founding principle of caste as endogamy was premised on the sexual control and elimination of girls and women. In his exegesis of the exploitation of the Surplus Woman in the essay, Ambedkar proposes three supplemental principles to endogamous marriage that regulate the Surplus Woman before and after marriage: sati, enforced widowhood, and female infanticide.

First, in a trenchant critique of upper castes, Ambedkar notes that the Surplus Woman was best resolved by death, as a woman “dead and gone” creates the equilibrium necessary for 1:1 ratio that was essential for endogamy. Thus, as Ambedkar explains, in the logic of caste endogamy, the Surplus Woman of the endogamous group must commit sati at the moment of the death of her husband to eliminate the surplus. Yet it is not so easy to throw oneself onto a burning pyre.

Ambedkar thus proposes a second possible sociological resolution of the Surplus Woman through other social institutional mechanisms. If the woman who is no longer legitimate through marriage was not to die on the funeral pyre, she must socially die through enforced widowhood and remain permanently outside of the realm of marriage and reproduction. As Ambedkar emphasizes, enforced widowhood is a delicate, often failed social institution because the woman continues to live and be human, and that humanity endured, however circumscribed by social exclusion and shame. Note his sarcastic tone. Indeed, as Ambedkar (Reference Ambedkar1917, 87) insists, it is difficult to keep a woman within a society as a degraded apparition, a ghost, not fully human, permanently outside of reproductive kinship.

Third, according to Ambedkar, a final way to deal with the potential of the Surplus Woman is to prevent the possibility of a woman ever being outside of marriage in the first place. Thus, a girl must be married before puberty to avoid the possibility of moving beyond the limits of caste interdiction.

In other words, upper-caste endogamy required the perpetual creation of boundaries in a girl/woman's life that centered on marriage and her control under different patriarchal regimes. From the moment of her birth, the girl was a potential threat to endogamous caste, for she had the potential to fall outside of the bounds of reproductive patriarchal marriage. When a woman did exceed the bounds of marriage, through widowhood or otherwise, her surplus womanhood was a threat to social reproduction (Ambedkar Reference Ambedkar1917, 87).

In Ambedkar's theory, the Surplus Man assumed a secondary and more easily resolvable role. As in McLennan, men were ultimately more productive members of an endogamous society as all patrilineal societies primarily valued men. When a man lost his wife and became a widower, he was in the position of the Surplus Man. He was only able to resolve his status as a surplus by either becoming an ascetic, outside of the bounds of reproduction, or marrying a child bride to reestablish his legitimacy within the caste through endogamous marriage. In Ambedkar's theory, the Surplus Woman was irresolvable in life, and could only be solved through a woman's death, whereas social institutions in caste allowed for a man to transition back into endogamy from the position of Surplus Man through child marriage between child brides and adult men. Yet the resolution of the Surplus Man through child marriage in endogamy perpetuated the problem of differential ratios of man to woman, by allowing men to remarry girl children, while barring women from the same privilege (Ambedkar Reference Ambedkar1917, 86–87).

For Ambedkar, the boundaries of endogamy required an equal population of the sexes to sustain the ideal of normative heterosexual reproduction. The imposition of endogamy on exogamous practices, rather than a purely textual dictate, was a sociological and political phenomenon, an institutional interdiction that defined social trespass through the enforcement of caste. As Prathama Banerjee (Reference Banerjee, Paik and Raoforthcoming) demonstrates in her close reading of Ambedkar's essay, Ambedkar argued that caste reconstituted Indian society into a hierarchy of groups enforced and sustained through an economy of difference and social control. Endogamy, for Ambedkar, rather than being an essential characteristic of the whole of Hindu society, was an antisocial institution. It was the antithesis of a unified society, the primary mechanism for social exclusion and hierarchy.

So why do people abide by this wretched system? In this moment in the essay, Ambedkar (Reference Ambedkar1917, 88) answers with a cutting remark: “This high-flown and ingenious sophistry indicates why these institutions were honoured, but does not tell us why they were practiced. My own interpretation is that they were honoured because they were practiced.” That is to say, for Ambedkar, it was Brahmanical ideology that made sacred, honorable, and transcendent those practices that banned widow remarriage but perpetuated child marriage, those institutions held as sacred that were enforced for the sake of maintaining Brahman power. The “social custom” of caste, in other words, was an ideology of dominance and social imposition, not a result of social evolution, transcendent religious piety, or honor. In other words, in “Castes in India,” Ambedkar brings the insights of generations of anti-caste thinking to bear on the foundational place of a Sanskritized social science of India, critiquing the way the social science of India had naturalized the biased and violent structures of caste domination through a claim to evolutionary science.

In her introduction to Ambedkar's essay, Rege (Reference Rege2013, 61) argues that Ambedkar was “highlighting a double maneuver by which Brahmanical ideology both preserved and eulogised the very practices that degraded women.” As Rege demonstrates through her feminist reading of Ambedkar, there were three mechanisms for the institutionalization of endogamy, “intragroup organization of reproduction, violent control of surplus woman's sexuality, and legitimating control practices through ideology” (61). These three structures worked in tandem: the ideology of Brahmanical superiority functioned to justify and perpetuate the social imposition of endogamy and the violent disappearance of the Surplus Woman.

It is after laying this base of the three mechanisms that constituted endogamy in his essay that Ambedkar suggests that the ethnological dependence on Sanskrit text as evidence of civilization was the most perverse of all. In his most direct refutation of the Brahmanical Indology of caste, Ambedkar (Reference Ambedkar1917, 89) argues that the progress of the caste system was not to be found in the institutionalization of Manu and authoritative Sanskrit text:

For Ambedkar, Manu is not an epistemic event memorialized in text at the origins of the development of dictates of Hindu society. Manu is simply a man, from a long time ago, who created a philosophy of a social hierarchy that already existed in his society's own social world, a society defined by discrimination and hierarchies of political power. Ambedkar suggests that these textual strictures of enforcement for caste held sway in modern India only because lower castes sought to attain the position of upper castes.

Thus, in his theory of the imposition of endogamy, Ambedkar demonstrated that this dependence on Indology as the evidentiary correlate to ethnological theory naturalized the history of caste supremacy as scientific fact. From McLennan onward, Manu had been enshrined as the source to understand the origins of endogamy, taken to be an unchanging and irrefutable fact of caste in India. Through the twinned sciences of ethnology and Indology, colonial administrators like Risley and upper-caste Hindu social scientists such as Ketkar had presented caste hierarchy based in the control of women's sexuality as a natural development of modern Indian society, in which Manu was placed at the origin as the very genesis of caste endogamy. Upper-caste Hindu scholars naturalized caste difference and upper-caste power premised on the control of female sexuality as an Indological fact based in Sanskrit text and utilized concepts from ethnological theory to justify the primacy of textual dictates in defining the elementary structures of Indian society.

In arguing against the correlation of ethnology and Indology through a theory of endogamy as a social and historical imposition, Ambedkar undid theories of caste endogamy that saw it as the pure, natural, and inevitable foundation for the holistic text-based study of an unchanging Hindu society. Instead, Ambedkar retained the concept of endogamy as a structure of caste but outlined it as a Brahmanical ideological imposition and hegemonic political symbol. Yet Ambedkar did not critique endogamy as an idealized structure or call for the abandonment of endogamy as a concept altogether, despite the countless examples of social practices among a wide range of people, including Brahmans themselves, that defied strict adherence to endogamy in the workings of caste. Rather, shifting from the logic of Indological origins to an analysis of social power, Ambedkar revealed the systemic social and sexual violence against women based in an ideology of endogamy that undergirded caste supremacy. Indeed, in this genealogy of thought on endogamy, Ambedkar was the first to explicitly name and make visible the sexual control and violence that sits at the heart of the political enforcement of endogamy.

Endogamy was a potent term from its invention as a universalist category by McLennan to its reinvention as a political system of imposition and dominance by Ambedkar. Theories of endogamy, exogamy, and hypergamy were all predicated on a logic that saw the socially mandated elimination of “excess” women outside of monogamous conjugality as a necessary and natural biproduct of endogamy, from female infanticide in the lowest and highest orders of society to enforced widowhood, child marriage, and sati. Yet Ambedkar, unlike the British colonial administrators and Indian social scientists who came before, utilized the concept of endogamy to reveal caste as a political and historical system, rather than an inevitable result of evolution or ancient textual dictate. In doing so, Ambedkar reframed the concept of endogamy through a trenchant critique of upper-caste power based in sexual control.

Endogamy and the Control of Female Sexuality

The legacy of the concept of endogamy can be found in its continued uncritical use of across social scientific studies in India through the twentieth century. Endogamy continued to have a vibrant life as an indispensable concept in the social theory of South Asia throughout the twentieth century. Yet, as Rege notes, modern sociology in India entirely ignored Ambedkar's argument in “Castes in India” about the Surplus Woman and endogamous marriage as enforced sexual control based in false ideologies of caste supremacy. Even as Ambedkar sought to declare the end to the Indological justification of endogamy by publishing in a specialist journal, Indian Antiquary, his ideas about endogamy as imposition remained on the margins of modern social theory. Disciplines as diverse as Indology, anthropology, sociology, and history continued to exalt Manu and create new social theories about caste and race through the uncritical marrying of the concept of endogamy with Sanskrit Indology.

The disappearance of Ambedkar's argument about the irresolvable problem of the Surplus Woman becomes apparent in even a brief survey of the scholars who took up the study of endogamy that followed. These social scientists went on to establish major social science departments and organizations in colonial and postcolonial India. For example, Indologist and sociologist Govind Sadashiv Ghurye (1893–1983), a professor at the University of Bombay, offered in his Indological/sociological study Caste and Race (Reference Ghurye1932) a detailed reading of Manu as an originary sociological text to understand caste and racial difference. Ghurye (Reference Ghurye1932, 18), founder of the Indian Sociological Society, argued that hierarchy and endogamy were the primary features of this text-based system of caste, citing Westermarck to demonstrate that strict endogamy was the organizing principle of caste. Ghurye's theory of caste and race naturalized the control of female sexuality through endogamy as a necessary and natural result of the evolution of the caste system. At stake for Ghurye was a vision of early Indian sociology that could displace colonial ethnographies with a holistic vision of the methods and textual origins for Indian sociology. In Ghurye, endogamy was best investigated with methods that linked ethnology to the Indological study of Sanskrit texts like Manu.

Ghurye's most well-known student at the University of Bombay, M. N. Srinivas (1916–99), would go on to offer a novel, transformative theory of Sanskritization based in a theory of the fluidity of caste among middle castes through the emulation of “custom” and the Brahmanization of “polluting” practices among lower castes (Srinivas Reference Srinivas1952). Indeed, Ambedkar had made a parallel argument to Srinivas in “Castes in India,” critiquing lower-caste emulation of Brahman ritual and Sanskrit principles decades before and condemning lower-caste “Sanskritization” and exaltation of Manu (Jaffrelot Reference Jaffrelot2005, 33; Banerjee, Reference Banerjee, Paik and Raoforthcoming). Caste emulation for Srinivas was one of mimicry and the adoption of upper-caste ritual, not through the transformation or radical defiance of sexual interdictions against caste intermarriage by lower castes. Thus, in Srinivas's conception of caste, lower-caste emulation of upper-caste customs was not based in reformed dictates of sexual prohibition but rather the opposite: more stringent control of female sexuality to enforce endogamy among all castes. These sociologies of caste and race do not theorize the structures of power that organized and controlled women's sexuality. Instead, these key sociologists used endogamy as a concept that explained caste stricture while leaving the control of female sexuality unmarked, even as normative models of sexual prohibition and marriage undergird every aspect of their models of caste society.

By the 1960s, another of Ghurye's students from the University of Bombay, a woman and self-declared feminist anthropologist Irawati Karve (1905–70), argued in her 1961 study Hindu Society: An Interpretation that endogamy was the central principle of caste (Das Gupta Reference Das Gupta1962, 17). Karve, founder of the anthropology department at Deccan University (today the University of Pune), was committed to studies that combined Indology, ethnology, archaeology, and participant observation (Sundar Reference Sundar, Uberoi, Deshpande and Sundar2007). Karve's theory of caste as endogamy, unlike those of Ghurye and Srinivas, remarked on women's nonconjugal sexuality, and in doing so, stands out from the disciplinary sociological studies of her time. Yet Karve quickly sought to resolve the problem of nonconjugal female sexuality for a structuralist theory of society based in endogamy. For Karve, there was but one exception to endogamy and hypergamy of castes, which was “an exception in a biological sense.” It was not a social exception and did not change the underlying structure of caste: “while leading to sexual relations across caste, has little effect on the family organization or the caste organization as such. In this category are to be mentioned practices of concubinage, housemaids and prostitution” (Karve [Reference Karve1961] Reference Karve1968, 16–17).

Karve thus recognized the “biological” exception to endogamy in a typology of “deviant” women outside of conjugality that were akin to the problem of the “Surplus Woman” of Ambedkar (but, like her contemporaries, without any reference to Ambedkar). Yet rather than suggesting, as Ambedkar had, that the presence of these forms of sexuality and nonconjugal reproduction represented a fundamental failure of a structuralist theory of caste as endogamy, Karve argued that the “biological exception”—those concubines, housemaids, and prostitutes—had little effect on endogamy. For Karve, men participated in reproductive endogamous caste-based marriage alongside their relationships with these “surplus” women, and the multiple mechanisms to deal with these nonconjugal forms of female sexuality worked in tandem with a coherent and replicable system of caste endogamy.

Debates about caste as endogamy further expanded as a result of the publication of Louis Dumont's widely celebrated, debated, and critiqued Homo Hierarchicus: Essai sur le système des castes (Reference Dumont1966). In Dumont's totalizing theory of Hindu society, caste was fundamentally based in social hierarchy. In prioritizing the concept of hierarchy in the study of India, Dumont offered the strongest rejoinder against decades of theories that posited caste as endogamy, arguing that the structure of endogamy was subordinate to hierarchy. For Dumont, stratification explained the true nature of caste—depicted as an essential and unchanging institution by Dumont, just as it had been for Risley almost a century earlier.

Dumont's study reinvigorated the “tropological hegemony of hierarchy” in the study of caste, but also inaugurated field-changing critiques by South Asian scholars who sought to shatter the primacy of Dumont's holistic object of “Hindu society” through detailed studies of regional communities and critique of structuralist claims to the study of caste as a universal and unchanging “essence” of Indian society (Appadurai Reference Appadurai1986, 757). No longer could the sociology of India be based in theories of caste as the singular underlying structure that was the primary essence of Indian society (Madan Reference Madan1971).

Indeed, this genealogy of the concept of endogamy also reveals how Dumont's supposedly novel argument, caste as holistic hierarchy, had been made decades before in the study of endogamy and exogamy. Hierarchy was nothing new, and the main principle of Dumont's theory of the wholeness of Indian society relied just as readily on the naturalized control of women's sexuality within and between castes. Like other sociologists in the twentieth century, Dumont's theory of hierarchy as the primary structure of caste, and thus of Indian society, was premised on the normative regulation of women's sexuality through marriage or death, for no hierarchy could be maintained without strict divisions between castes and the control of female sexuality within caste structures. The social enforcement of sexual strictures and the control of women is again left unremarked upon by Dumont, whose own theory of a holistic hierarchy depended on the naturalization of caste divisions through patriarchal authority over women's sexuality and reproduction. Indeed, in Dumont, the control of female sexuality through heterosexual conjugality was nothing more than a natural fact in a romantic vision of the totality that was Indian society, past and present.

Totalizing sociological studies of “Hindu society” declined in the aftermath of critiques of Dumont and others (Appadurai Reference Appadurai1986; Dirks Reference Dirks1992; Madan Reference Madan1971). Yet endogamy continues to have a vibrant life as a prevalent concept in the study of caste. Today, endogamy has been reinvigorated as a site of knowledge in a new scientific field that takes ethnology and Indology as its origins: the Hindu nationalist genetics of caste difference and racial Aryanism institutionalized in new large-scale genetics projects undertaken by the Indian government (Mukharji Reference Mukharji2014). The continued importance of endogamy in twentieth-century social science merits further study and may offer insights into the origins of other concepts in modern social theory that have endured in postcolonial sociology in South Asia, despite their dubious origins.

So how do we kill the ghost of Manu, as Ambedkar so strongly urges us to do? It requires a critical reassessment of the foundational disciplines, frameworks, and concepts that have come to organize how we study South Asian societies. It mandates further reflection on Ambedkar's insistence on the historical, sociological, and phenomenological contours of that superposition of caste as a politically imposed form of sexual control, what Sharmila Rege, Aniket Jawaare, and other scholars of caste, gender, and Dalit subjectivity have theorized as the enforcement of caste through controls of the body, touch, sexuality, public space, and intimacy (Jawaare Reference Jaaware2018; Mokkil Reference Mokkil2018; Ramberg Reference Ramberg2014; Rao Reference Rao2003; Rege Reference Rege2013). Indeed, it may mean a radical dismantling and reimagining of the very terms used in the social scientific study of caste where the superposition of endogamy is studied not as an already given structure used to describe a society, but as an unstable and incomplete act of caste enforcement, a form of imposition based in the ontological life of caste domination. Displacing endogamy as the primary descriptor of caste altogether, theories such as Jawaare's conception of touching and not touching offer a radical reconceptualization of the study of Indian society outside of all-encompassing concepts like endogamy, accounting for the multisensory processes and experiences of discrimination, violence, and resistance that constitute the making and unmaking of caste domination and subordination.

My argument has been that the social scientific concept of endogamy effectively concealed the control of female sexuality in major theories of Indian society based in the social scientific study of marriage. In doing so, I have utilized methods in the history of sexuality to reveal the critical place of normative ideas of conjugal sexuality, sexual violence, and caste subordination in the modern study of Indian society. Through the excavation of these ideas of sexuality at the heart of the invention of endogamy, we may better understand how modern social theory kept the ghost of Manu alive while disappearing the problem of the Surplus Woman.