Perspectives on the Study of Caste Systems: GS Ghurye, M N Srinivas, Louis Dumont, Andre Beteille

Four Thinkers, Four Ways of Reading the Same Institution

Sociology Optional Notes {PDF}

Complete Notes Package (Paper I & II) | 1000+ Pages | Instant Download

Discounted Price | Buy Now @ ₹600 →

• Caste has been studied through more than one method, and the method a scholar chooses shapes what they end up seeing in it.
  • The attributional approach — associated with Ghurye and, in a more dynamic form, Srinivas — identifies a checklist of defining features (hierarchy, endogamy, restrictions on food and marriage, pollution) and asks how each caste’s position follows from possessing or lacking them.
  • The interactional approach — associated with F.G. Bailey, Adrian Mayer, and McKim Marriott — studies caste rank not as a fixed attribute but as something continuously negotiated and expressed through actual patterns of interaction: who eats with whom, who accepts water from whom, who is greeted first.
  • Dumont’s structuralist-Indological approach treats caste as a system of ideas rooted in classical Hindu texts, to be decoded rather than merely observed.
  • Beteille’s empirical-institutional approach treats caste as one axis of stratification among several, to be studied through direct fieldwork rather than assumed to be India’s master institution.
• This article treats each thinker’s specific theory of caste. Ghurye and Srinivas’s broader sociological orientations — their overall approach to studying Indian society as a whole, not caste alone — are covered in the companion article on perspectives on the study of Indian society; here the focus stays on what each of the four (and, as a necessary fifth counterpoint, Ambedkar) actually argued caste is.

G.S. Ghurye: Caste as a Historically Evolved, Competitive Hierarchy

• Ghurye’s method combined historical, comparative, and Indological sources — textual evidence read alongside census data and comparative material from other Indo-European kinship systems — moving from historical and Indological grounding, through comparison across regions, to generalization about caste as an institution.
• He identified six structural features of the caste system.
  • Segmental division: caste membership is acquired by birth and fixes a person’s position in a rank order from the outset; each segment (caste) runs its own internal affairs through a caste council with legislative, executive, and judicial powers of its own.
  • Hierarchy: castes are arranged in relations of superiority and inferiority, with Brahmins conventionally at the top and untouchables at the bottom.
  • Restrictions on food, drink, and social intercourse: pollution rules govern what can be eaten, and from whom.
  • Civil and religious disabilities and privileges: unequally distributed rights of dress, speech, temple entry, and public facility use, graded by rank.
  • Lack of unrestricted choice of occupation: each caste traditionally carried a hereditary calling.
  • Restrictions on marriage: strict endogamy within the caste, with gotra (clan) exogamy operating as a cross-cutting rule within it.
• Two features of Ghurye’s specific reading are worth isolating because they distinguish him from a simple “caste is a rigid ladder” account.
  • He insisted that caste hierarchy is competitive, not absolute. Citing seventeenth-century European observer John Fryer’s writings alongside his own regional comparisons, Ghurye argued that hierarchical gradation clearly existed but a caste’s exact position within it was constantly negotiated rather than fixed: in South India, Brahminical superiority was regularly contested; Bhakti saints were sometimes treated as purer than Brahmins in practice; Kayasthas and Nayars asserted elevated status through Mughal court patronage.
  • He developed an interactional theory of food hierarchy: comparing food-exchange behaviour across Tamil Nadu, Bengal, and Maharashtra, he found a strong correspondence between how many castes would accept food from a given caste and that caste’s rank — the more castes willing to receive its food, the lower, counter-intuitively, its actual position (since higher castes accept from fewer sources without risking pollution).
• Ghurye read caste and class as operating on separate tracks under colonial rule: when men of different castes joined the same occupation (the colonial army, for instance), their shared occupational or class position did not erode caste identity in private life, because caste councils retained authority over marriage, ritual, and death regardless of what happened in the workplace. Culture, in his reading, promoted unity across India even as caste promoted segmentation within it.
• He placed particular weight on endogamy as the single most important feature holding the system together, distinguishing three types of marriage restriction: endogamy (marriage confined within caste), exogamy (marriage prohibited within specified kin categories), and hypergamy (a woman marrying into a caste or sub-caste of higher status than her own). Exogamy itself divided into sapinda exogamy (prohibited degrees of close kinship) and sept or gotra exogamy, rooted in the idea that members of a gotra shared descent, real or fictive, from a common sage (rishi) — making marriage within it symbolically incestuous.
• Ghurye also anticipated, correctly, that caste would generate voluntary associations for education and reform — the Nadars, Reddys, and Kammas of the South, the Saraswat Brahmins and Vaishyas of Maharashtra, and the Kayasthas of the North all founded caste associations that later fed into political mobilization, a development Rajni Kothari analysed at length decades later, though Kothari — unlike Ghurye — treated these associations more positively, as instruments of welfare activity and democratic participation rather than purely as threats to social unity. Ghurye himself read the backward-classes agitations of his time as a “scramble for privileges” that was fragmenting Indian society’s underlying unity, and treated caste violence as a departure from caste members properly following their own traditional dharma rather than as evidence against the system itself.
• Critique: Ghurye has been faulted precisely for this functionalist, non-critical stance. Coral Upadhyay argued he offered a classical Hindu view of caste without engaging its ideological uses of power, and was not, in this sense, a genuinely empirical sociologist. Sujata Patel made a related point more sharply: Ghurye’s account is a Brahmanical view of caste that does not hold across all regions of India, where Brahmins are not universally regarded as ritually supreme.

M.N. Srinivas: From a Fixed Ritual Rank to Two Dynamic Hierarchies

• Srinivas remained within the attributional tradition but gave it a dynamic twist: rather than treating caste attributes as static markers, he asked how caste structures of relationship actually operate and change — an emphasis that becomes explicit in his work on sanskritisation and the dominant caste, both of which are given fuller treatment in the companion article on his general perspective on Indian society. Here the focus is specifically on his theory of caste’s internal structure.
• Srinivas assigned caste five core attributes: hierarchy (clearest at the very top and bottom, most flexible and contested in the middle ranks); occupational differentiation, since caste is, in his phrase, “the systematization of occupational differentiation” (castes such as Lohar, Sonar, Kumhar, Chamar are literally named for their traditional trades); restrictions on commensality, dress, speech, and custom; pollution, with the most serious breaches (such as a high-caste person’s sexual relations with an untouchable) resulting in expulsion from caste pending purification; and caste panchayats and assemblies, elder-led bodies whose jurisdiction could extend across village boundaries. Every caste, in his reading, functions as a segmentary system: divided into sub-castes that are simultaneously units of endogamy, occupation, ritual and social life, shared culture, and panchayat authority.
• Caste is not Varna. Srinivas insisted these are frequently and wrongly conflated by Western scholarship. Varna is a textual, four-fold, pan-Indian ideological scheme; caste (jati) is regional, numerous (thousands of groups, not four), and empirically observed. He traced Varna’s own textual evolution — an initially two-fold Rigvedic racial division (Aryans and Dasyus), expanding to three occupational-cum-racial categories, then to the familiar four-fold scheme in the Purushasukta hymn, where each Varna is said to originate from a different part of a cosmic being’s body (Brahmins from the mouth, Kshatriyas from the arms, Vaishyas from the thighs, Shudras from the feet) — a scheme later editorially reworked (the Taittiriya Samhita shifts the Brahmin’s origin from “mouth” to “face”) in ways that further glorified Brahmanical status. Because Varna homogenizes enormous internal diversity — Shudra communities, for instance, range from long-marginalized groups to some historically wealthy and powerful ones — it gives, at best, a partial and static reading of Indian society; caste alone captures its dynamism.
  • Even so, Varna persists as a working ideological frame of reference: it lets people locate an unfamiliar regional caste within a shared pan-Indian scheme (so that inter-regional food-exchange rules can be applied), and it gives an aspirationally mobile caste a ready-made higher status to claim.
• Two hierarchies, not one. Against Dumont’s single, fixed, textually-derived ritual hierarchy, Srinivas proposed that caste rank in practice reflects the interaction of two distinct hierarchies: a ritual hierarchy (defined by birth, food practice, dress, and rites of purity and pollution) and a secular hierarchy (defined by wealth, political power, education, and occupation). Where the two diverge, the resulting rank is genuinely contested and negotiated rather than simply read off scripture.
  • Empirical cases: Lingayat Brahmins of Karnataka claim superior status even relative to born Brahmins; Bengal’s Chaitanya-following parichaskhyas observe ritual standards strict enough to claim superiority over local Brahmins; Bihar’s Bhumihars have successfully claimed Brahmin status; S.C. Dube’s study of the Rajgonds and M.S.A. Rao’s study of the Yadavs both found originally Shudra groups rising in status by acquiring land, capturing local political power, and organizing.
  • Adrian Mayer, studying Ramkheri village, found Jats economically dominant, Rajputs politically dominant, and Brahmins ritually dominant — three separate hierarchies coexisting in one village. Oscar Lewis, studying Rampur, found a comparable pattern: Rajputs and Jats dominant in the secular sphere, looking down on Brahmins in practice even while Brahmins retained ritual precedence in theory.
  • Srinivas listed the concrete mechanisms of secular mobility feeding back into ritual status: taking up new occupations, adopting new customs, constructing a fictive higher-caste genealogy, mounting a jati campaign, securing support from lower castes, benefiting from new legislation or political patronage, and migration. Secular mobility, on his reading, is never simply an end in itself — it fertilizes and eventually pulls along ritual mobility too.
• Caste and politics: the growing secular role of caste. Srinivas directly contradicted the modernization theorists of his time — Myron Weiner and Gunnar Myrdal — who expected the spread of the constitution, modern education, rural development, and democratization to dissolve “caste India” into a class-stratified “modern India.” Srinivas argued the reverse was happening: the more India modernized, the more old caste identities were being repurposed to capture the benefits of that modernity. He termed this the growing secular role of caste alongside the declining ritual role — caste’s composition and functions were changing, but caste itself was not being replaced by class.
  • He called the phenomenon of castes temporarily setting aside internal differences to pursue a shared political goal, then dissociating once that interest is secured, the “varnisation” of caste — mirrored by what Marxist sociologists Yogesh Atal and Ghanshyam Shah termed “classization,” arguing that a dominant caste coalition is, in substance, closer to a class than a caste, bound by shared economic and political interest rather than ritual kinship.
  • Contemporary examples of this consolidation include the Patidar agitation in Gujarat, the Maratha agitation in Maharashtra, and Jat mobilization in Haryana — all instances of a numerically and economically strong caste group pressing collective demands through modern political channels.
  • T.K. Oommen extended Srinivas’s insight, describing caste as functioning as a “power reservoir” and caste leaders as “power exercisers,” producing a contemporary “caste elite” through this political mobilization. Rajni Kothari captured the same reciprocal relationship memorably: “politics needs caste as much as caste needs politics” — caste supplies a ready-made vote bank, while political competition gives caste new reasons to consolidate.
  • Pauline Kolenda traced the specific intellectual lineage of “dominant caste” to Evans-Pritchard’s study of the Nuer of Sudan, where “dominant clan” described local, area-level, or regional superiority — a debt Srinivas acknowledged while insisting his own concept had emerged independently from his Coorg fieldwork. K.L. Sharma pushed back against overgeneralizing the concept, noting there is no single, all-India “dominant caste,” only locally or regionally dominant ones. Deepankar Gupta separately challenged the numerical-strength criterion directly: in western Uttar Pradesh, Jats are only around 9 percent of the population against Dalits’ 25 percent, yet political and economic power remains with the numerically smaller Jats — evidence, in Gupta’s reading, that dominance cannot be reduced to Srinivas’s checklist of attributes.
• Srinivas’s caste theory, taken as a whole, performed three functions for the discipline: it validated fieldwork as sociology’s central method against purely textual approaches; it offered a “field view” that directly challenged the colonial “book view” of caste as static and unchanging; and it insisted on the regional and “little tradition” dimensions of caste against scholarship that assumed a single, uniform, pan-Indian system. Notably, despite documenting caste’s continuing vitality so thoroughly, Srinivas himself did not support caste-based reservation as a strategy for altering unequal caste relations.

Louis Dumont: Caste as Ideology, Not Stratification

• Dumont’s central claim, developed across his career and stated most fully in Homo Hierarchicus (1967; English translations 1970, 1972), is that caste is not a form of social stratification comparable to class — it is a distinct system of ideas and values, and the sociologist’s task is to decode that ideological structure rather than simply catalogue its behavioural attributes.
• His method combined the Indological (close reading of classical Hindu textual sources) with the structuralist (borrowed from Lévi-Strauss, whose method of understanding a system through the binary oppositions structuring it Dumont applied directly to caste) and drew explicitly on Hegel’s idealism — the premise that what is ideal (an idea, a value) is what is ultimately real and determinative of social behaviour — and on Célestin Bouglé’s earlier definition of caste as involving three characteristics: separation (rules of marriage and contact, direct or indirect through food), interdependence (a hereditary division of labour, each group tied to a traditional occupation), and gradation (hierarchical ranking of groups as superior or inferior).
• The pure–impure opposition is, for Dumont, the single organizing principle beneath the entire system. Hierarchy expresses the superiority of the pure over the impure; separation exists to keep the pure and impure apart.
  • Brahmins, occupying priestly functions, sit at the summit of this hierarchy as maximally pure; untouchables, assigned to inherently polluting tasks, are placed outside the village and denied well access and temple entry.
  • Dumont distinguished temporary impurity (the kind attached to specific events such as birth, death, or menstruation, found across many world religious traditions — he drew a direct parallel to a mother’s temporary exclusion from the church in some Christian traditions) from the massive and permanent impurity attributed to untouchable castes as a matter of birth. P.V. Kane’s History of Dharmashastra is cited for the observation that even a person’s closest relatives become temporarily “untouchable” to them around such events — showing that impurity as a category is not confined to untouchables alone, even though its permanent form is.
  • Even objects carry graded purity in this framework — silk is purer than cotton, gold purer than silver, silver purer than bronze, bronze purer than copper — and this is a cultural prescription, not an individual one.
• Varna over caste. Where Srinivas insisted caste (not Varna) captures empirical Indian reality, Dumont reversed the emphasis: Varna is the pan-Indian ideological framework structuring the whole system, while caste is merely its regional, empirical manifestation, “born not of occupation” but organized fundamentally around access to and distance from ritual power. Studying India from the caste standpoint alone, Dumont warned, risks over-glorifying social change that has not actually touched the underlying ideological structure; the Varna standpoint reveals what he considered the system’s genuinely rigid and static core.
• A second defining claim: caste is distinguished from other stratification systems by the disjunction between ritual status and secular (economic and political) power within a single social system — status (the Brahmin/priest) and power (the King) are structurally separated, with power formally subordinate to status even though the two remain practically interdependent.
• Reading post-independence change: “substantialization.” Dumont read the traditional interdependence of castes as increasingly replaced by what he called a “universe of impenetrable blocks, self-sufficient, essential and identical, and in competition with one another” — the substantialization of caste. He catalogued the sources of this shift (judicial and political change, social-religious reform, Westernization, growth of modern professions, urbanization, spatial mobility, market economy) but read the dominant contemporary pattern as a mixture or combination of traditional and modern features, rather than a wholesale transformation. Significantly, he treated the rise of inter-caste competitiveness he acknowledged in the twentieth century as a behavioural change, not a change in the underlying values or ideological principles of the system.
• Critique. Dumont’s comparative frame — Homo Hierarchicus (hierarchical India) against a projected Homo Aequalis (egalitarian Europe) — has drawn sustained criticism for Eurocentrism.
  • Andre Beteille was Dumont’s most persistent critic on precisely this point, arguing that “the ideology of equality and resistance to inequality is a universal phenomenon,” not a Western peculiarity — invoked from the Renaissance through J.S. Mill in Europe, and equally from the Buddha through Gandhi in India — and that “no society is absolutely open, and no society is absolutely closed: openness and closedness are a matter of degree, not kind.”
  • Yogendra Singh argued caste should be understood as a structural reality, not a cultural fact unique to Hindus; Harold Gould challenged Dumont’s assumption that Brahmanic values were internalized by everyone, including Dalits, for whom this plainly does not hold; M.S.A. Rao noted that Brahmins are in fact despised in parts of Haryana and Punjab, directly contradicting a universal internalization thesis; Veena Das argued the king is, in practice, at least as powerful as the priest, complicating Dumont’s status-over-power ordering.
  • More broadly, critics have argued Dumont’s account reflects a dominant, internal viewpoint — that of priests and princes — legitimized by precisely the classical Sanskrit texts he relied on, without adequately interrogating the social conditions under which those texts were produced or the relations of power and domination they encode and partly obscure.
  • Even granting these criticisms, Dumont’s Homo Hierarchicus remains, in Beteille’s own assessment, “a rare coherent sociological enterprise that cannot be ignored” by anyone seeking to understand how contemporary India has been socially constituted.

Andre Beteille: Caste, Class, and Power as Three Independent Dimensions

• Beteille’s sociology of caste is explicitly empirical, reflexive, and comparative, positioned as a deliberate departure from what he read as a shared assumption running through Ghurye, Srinivas, and Dumont alike: an (explicit or implicit) treatment of Brahmanic ritual superiority as the taken-for-granted centre of the whole caste order.
• His conclusions rest on intensive fieldwork in Sripuram, a village in Tamil Nadu’s Tanjore district, first studied in 1964 and published as Caste, Class and Power: Changing Patterns of Stratification in a Tanjore Village (1966).
  • Sripuram contained three major caste blocs: Brahmins, Non-Brahmins, and Adi-Dravidas, with a wide cultural, symbolic, and relational gap separating Brahmins from Adi-Dravidas at the two extremes of the local hierarchy.
  • Crucially, Beteille found the Brahmins of Sripuram themselves sharply divided — Srivaishnavas and Smarthas, distinguished by ritual practice, doctrinal affiliation, and residential separation, each practising endogamy toward the other, with the Smarthas further subdivided into four groups and beyond. (This attention to internal segmentation owed an explicit debt to Evans-Pritchard’s analysis of segmentary structure among the Nuer, transmitted to Beteille through his teacher.) A single Brahmin identity claiming timeless, undiluted superiority, in other words, does not survive contact with the field data — Brahmins are as segmented and internally contested a category as any other.
  • Non-Brahmins, meanwhile, were themselves stratifying internally by wealth — upper, middle, and lower economic strata within the broadly artisan-and-peasant Non-Brahmin bloc. Beteille observed that even Non-Brahmins who achieved substantial economic upliftment, spent that wealth on modern education for their children, and adopted the self-identification of “Mudaliar” (a peasant-caste title), did not seek to become Brahmin even when the option existed — direct evidence that sanskritisation is not a universal aspiration and cannot be assumed as the default trajectory of caste mobility everywhere in India. Under the influence of the Dravidian movement, some Non-Brahmins instead treated secular mobility as an end in itself, deliberately distinguishing themselves from Brahmins framed as Aryan outsiders rather than emulating them.
• Caste, class, and power as three separate structural sources of inequality. Following Weber more than either the structural-functionalist or the Marxist traditions, Beteille argued that village India’s social structure rests on three genuinely distinct axes: unequal distribution of land (generating class), unequal access to political power, and unequal access to status via caste identity — none of which reduces to, or is fully explained by, either of the other two. His first book’s title, Class, Status and Power, states this thesis directly.
• From harmonic to disharmonic social structure. In earlier, more traditional village society, caste, class, and power tended to move together — the “summation of status” pattern in which high ritual rank correlated tightly with land ownership and political influence, producing what Beteille called a harmonic social structure. Modern change (education becoming genuinely open rather than a Brahmin monopoly, land entering the market and being sold off by traditional Brahmin landowners, non-Brahmins and Adi-Dravidas gaining access to white-collar employment and political participation) has progressively pulled these three axes apart, producing an increasingly disharmonic structure in which a person’s caste, class, and political position no longer reliably track each other. Beteille summarized this shift not as caste being replaced by class, but as “caste old” being replaced by “caste new” — caste persisting, but reorganized around new bases of value and competition (existentialism competing with spiritualism, materialism with moralism) rather than simply dissolving.
• Beteille on reservation policy. In Backward Castes and New Social Order, Beteille mounted a sustained critique of extending reservation to Other Backward Classes.
  • Defining class in Weberian terms (shared income and occupation), he argued that caste groups combined into a single “backward class” category are neither conceptually coherent nor empirically accurate, since OBCs vary enormously among themselves in historical and present access to income and occupation.
  • He argued reservation policy had become driven by vote-bank politics — the politics of reservation superimposing itself onto, and distorting, any defensible sociology of reservation — and that using present-day reservation to compensate for historical discrimination is, in his view, “anti-meritarian”: those who currently benefit were not personally the ones discriminated against in the past, nor are present-day non-beneficiaries personally responsible for that past discrimination.
  • He drew a sharp distinction between equality of opportunity (in education, healthcare, and employment access, which he actively supported) and equal distribution of benefits (which he opposed as a policy goal), arguing that citizenship, not caste identity, should be the basis on which India organizes access to public goods.
• The 2012 debate: “India’s destiny not caste in stone.” In a widely discussed Hindu op-ed, Beteille argued that caste consciousness is dying out slowly across most domains of life — inter-dining restrictions have become nearly impossible to enforce in colleges and offices, and while caste endogamy in marriage persists, there is no clear evidence it is increasing. He located caste’s continued political salience specifically in the mechanics of universal adult franchise and, above all, in television media, whose commentators have a professional incentive to foreground the “caste factor” in every election.
  • Joseph Tharamangalam responded that Beteille understated how caste persists through a material base — deeply unequal control over resources and the means of coercion — that has simply changed its instruments, no longer enforcing purity/pollution rules directly but instead restricting Dalit and OBC access to resources and opportunity. He cited India’s poor comparative social indicators (its 2010 Human Development Index rank of 119, well below China’s 89; its Global Hunger Index standing; a Multidimensional Poverty Index in which 65.8 percent of SCs and 81.4 percent of STs were poor against 55 percent nationally) and the fact that India’s roughly three lakh manual scavengers are drawn almost exclusively from Dalit communities, as evidence that caste’s material consequences remain acute even where its ritual prescriptions have eroded.
  • A second, independently published response argued that what actually sustains caste voting is not the prescriptive aspect of caste (the purity rules Beteille shows eroding) but its positive, internal culture — the values, distinctive to each caste, that its members take pride in and vote to protect. As illustration: nine of India’s ten wealthiest individuals (per Forbes) come from mercantile castes, reflecting an internalized culture of capital management the argument’s author attributes to caste values rather than to caste discrimination as such; honour killing, conversely, is disproportionately practised by landowning peasant castes of North India (notably Jats of Haryana and Punjab), whose caste culture confers honour specifically for defending family “honour” through violence.
  • Beteille’s own rejoinder held firm that he never claimed caste consciousness would vanish within fifty or a hundred years — only that new middle-class formation, driven by demographic, technological, and economic change, is genuinely altering how caste operates, with real (if unevenly distributed) examples of both continued reproduction of advantage and genuine upward mobility, citing Harish Damodaran’s India’s New Capitalists as documenting caste groups moving into business ownership for the first time.

Ambedkar’s Counter-Perspective: Caste as a State of Mind

• Though not among the syllabus’s four named thinkers, B.R. Ambedkar’s theory of caste is essential precisely because it rejects the shared premise running through Ghurye, Srinivas, and Dumont alike — that caste can be adequately explained by its cultural or ideological content — and locates it instead in a single structural mechanism.
• Ambedkar’s foundational principle is graded inequality: Hindu social order does not simply divide people into different categories, it ranks them unequally, one above another, with unequal claims on the good things of life, reinforced by prescribed, hereditary, father-to-son occupations and strict limits on inter-dining and intermarriage across categories.
• Reviewing and rejecting prior scholarly definitions of caste — pollution (Senart), absence of inter-dining (Nesfield), and prohibition of intermarriage combined with birth-based membership (Ketkar) — Ambedkar concluded that prohibition of intermarriage is caste’s one truly critical, distinguishing element; everything else (including the “idea of pollution,” which he noted operates even in contexts with no caste system at all) is either derivative or non-essential.
  • He traced a striking structural implication: enforcing strict endogamy on a group produces numerical imbalance between marriageable men and women within it whenever normal mortality is uneven, since a normal population has roughly equal numbers of both sexes at similar ages. Societies had, historically, only a few ways to manage this “surplus” — burning surplus women on their husbands’ funeral pyres (sati), imposing compulsory widowhood and prohibiting widow remarriage, imposing celibacy on surplus widowers, or marrying surplus widowers to girls below marriageable age. Ambedkar read the customs of sati, enforced widowhood, and child marriage in Hindu society as direct social consequences of the mechanics of maintaining endogamy — not free-standing customs in their own right.
  • He explained the spread of endogamy through Gabriel Tarde’s law of imitation — imitation flows from higher to lower status groups, and its intensity varies inversely with social distance. Because Brahmins held (and were scripturally assigned) the highest prestige, non-Brahmin groups closest to them imitated endogamy fully (including sati and enforced widowhood), groups somewhat further away imitated only enforced widowhood and child marriage, and groups furthest away retained only the core belief in caste as a principle — producing endogamous closure across the social order by imitation rather than by any single legislative act.
  • Ambedkar explicitly rejected the argument that a specific lawgiver (traditionally Manu) created caste, doubting Manu’s very existence and noting in any case that caste predates any such figure; he was equally sceptical of theories tracing caste’s origin to race, occupation, or migration alone, since these forces exist in many other societies that did not develop a caste system — the closure into “self-enclosed units” was, on his account, the genuinely distinctive step.
• Annihilation of Caste. Ambedkar argued that caste is fundamentally “a notion, a state of mind,” which is why it cannot be dismantled through material or economic reform alone. Real change requires attacking belief in the sanctity of the shastras themselves — the scriptural texts that supply caste’s claim to religious legitimacy — since, in his assessment, reformers including Gandhi had failed to recognize that people’s conduct flows from their beliefs, not the reverse. His proposed institutional reforms included a single standard religious text acceptable to all Hindus, abolition of hereditary priesthood in favour of a state-licensed, examination-qualified priesthood open to any Hindu, and legal limits on the number of priests — measures aimed at breaking caste’s religious sanction rather than merely its economic consequences. He also directly challenged socialist strategies that assumed economic revolution alone would dissolve caste, arguing that the poor are not a homogeneous economic category and remain internally divided by caste and creed — social reform, in his view, had to precede or accompany economic reform, not follow from it automatically.

Reading the Perspectives Together

• Placed side by side, the disagreement among these thinkers is not really about whether caste hierarchy exists — all five accept that it does — but about what kind of thing it is and what holds it in place.
  • For Ghurye, caste is a historically evolved institutional complex, contested and competitive in its details but functionally integrative overall.
  • For Srinivas, caste is a dynamic, empirically observable structure of relationships, best studied through fieldwork rather than texts, in which ritual and secular hierarchies interact and secular gains eventually translate into ritual claims.
  • For Dumont, caste is, at bottom, an idea — a religiously grounded ideology of pure and impure — and behavioural change in modern India leaves that underlying ideological structure largely intact.
  • For Beteille, caste is one of at least three independent structural dimensions of inequality (alongside class and power), historically correlated but increasingly decoupled, and best studied through the same empirical methods used for any other form of stratification.
  • For Ambedkar, caste is sustained by a single institutional mechanism — enforced endogamy — given religious sanction by texts whose authority must be broken before caste itself can be.
• This disagreement is not merely academic: it has concrete, present-day stakes. India’s 2025 Cabinet decision to include caste enumeration in the next national census — the first such exercise since 1931 — reopened exactly the empirical question Srinivas and Beteille pressed against Dumont’s textual approach: how many castes actually exist, in what numbers, and with what real access to land, income, and office, as opposed to what classical texts say should be true of them. The Supreme Court’s 2024 ruling permitting states to sub-classify Scheduled Castes and Scheduled Tribes for reservation purposes — reversing its own 2004 position that the SC/ST list must be treated as a single homogeneous bloc — turns on a related question: whether “backwardness” within a legally reserved category is itself internally graded and unequal, a proposition far closer to Srinivas’s and Beteille’s emphasis on internal differentiation than to any account that treats caste categories as uniform blocks.

Key Terms

• Attributional approach — studying caste through a checklist of defining features (hierarchy, endogamy, pollution, and so on), associated with Ghurye and Srinivas.
• Interactional approach — studying caste rank as continuously negotiated through actual patterns of interaction, associated with Bailey, Mayer, and Marriott.
• Dominant caste — Srinivas’s concept for a caste combining numerical strength, landownership, and political power within a local area.
• Sanskritisation — Srinivas’s term for a caste’s adoption of the customs, ritual practices, and way of life of a locally higher caste to claim upward mobility.
• Substantialization — Dumont’s term for the post-independence shift from interdependent castes to self-sufficient, mutually competing caste “blocks.”
• Harmonic / disharmonic social structure — Beteille’s terms for, respectively, the traditional convergence and the modern divergence of caste, class, and power.
• Graded inequality — Ambedkar’s term for Hindu social order’s ranking of groups in unequal tiers, as opposed to a simple horizontal division.