Jati

Jati (Sanskrit jāti, "birth" or "kind") denotes the thousands of endogamous, hereditary kinship groups that form the functional building blocks of the caste system across the Indian subcontinent. The term derives from the verbal root jan, "to be born," signalling that membership is ascribed at birth and not acquired. While classical Brahmanical texts such as the Manusmṛti and the Puruṣa Sūkta of the Rig Veda articulate the four-fold varṇa schema—Brahmin, Kshatriya, Vaishya, Shudra—lived social organisation operates not through these four abstract categories but through an estimated three to five thousand jatis, each tied to a particular region, language, and traditional occupation. The Census of India, conducted from 1871 under colonial administrators including H.H. Risley, attempted to enumerate and rank these groups, an exercise that hardened fluid identities into fixed administrative categories and remains the documentary backbone of contemporary caste data.

The defining mechanic of jati is endogamy: marriage is contracted within the group and proscribed outside it, a rule enforced through kinship councils, family negotiation, and the threat of social ostracism. A jati is further bounded by commensality—rules governing with whom one may eat and from whom one may accept food and water—and by a traditional hereditary occupation, such as weaving (Koli, Devanga), leatherwork (Chamar), oil-pressing (Teli), or priesthood. Each jati occupies a graded position in a local ritual hierarchy calibrated by relative purity and pollution, the conceptual axis the sociologist Louis Dumont identified in Homo Hierarchicus (1966) as the organising principle of the entire system. Crucially, this ranking is local: a jati dominant in one district may rank modestly in another, so there exists no single all-India ordering of jatis.

Jatis articulate with one another through the jajmani system, the hereditary network of service-and-grain exchange that bound artisan and labouring jatis (kamins) to landholding patron families (jajmans) in agrarian villages. This created an interdependent, non-monetised division of labour in which the barber, washerman, potter, and priest each rendered customary service in return for a fixed share of the harvest. M.N. Srinivas introduced two further analytical concepts: the dominant caste, a numerically strong, land-controlling jati that wields decisive local power irrespective of ritual rank, and Sanskritisation, the process by which a lower jati adopts the rituals, vegetarianism, and customs of a higher group to claim upward mobility over generations.

In contemporary governance, jati is the operative unit of reservation and welfare policy. The Constitution of India lists Scheduled Castes under Article 341 and Scheduled Tribes under Article 342 by specific jati and tribe names notified by presidential order. The Mandal Commission (1980), implemented in 1990, identified Other Backward Classes at the jati level, and the Supreme Court in Indra Sawhney v. Union of India (1992) upheld 27 per cent OBC reservation while capping total reservation at 50 per cent and mandating exclusion of the "creamy layer." Jati mobilisation drives electoral politics—the Yadavs in Uttar Pradesh and Bihar, the Jats across Haryana and western UP, the Marathas in Maharashtra, the Patidars in Gujarat—with the Maratha and Patidar agitations of 2015–2018 and Bihar's caste survey of 2023 demonstrating the continuing salience of jati census data.

Jati must be distinguished sharply from varṇa, with which it is frequently conflated. Varṇa is the four-tier, pan-Indian textual model; jati is the empirical, regional, and far more numerous reality. One varṇa contains many jatis, and many jatis—particularly the formerly "untouchable" groups now constitutionally protected and the tribal communities—fall outside the varṇa scheme altogether as avarṇa or "fifth varṇa" populations. Jati also differs from gotra, the exogamous clan lineage within which marriage is forbidden; jati endogamy and gotra exogamy operate simultaneously, so a valid marriage must be inside the jati but outside the gotra. It is likewise narrower than the politically constructed "caste" of census and policy discourse, which aggregates jatis into broad blocs.

Scholarly and political controversy attends the very ontology of jati. The "constructionist" thesis advanced by Nicholas Dirks in Castes of Mind (2001) argues that colonial enumeration, ethnography, and law transformed plural, negotiable identities into a rigid, totalising system, whereas earlier Indologists treated jati as an ancient religious given. The decadal census has not collected comprehensive jati-wise data since 1931, a gap that fuels persistent demand for a nationwide caste census; the Government of India announced in 2025 that caste would be enumerated in the next decennial census. Urbanisation, anonymous wage labour, and constitutional bans on untouchability under Article 17 have eroded commensality and the jajmani order, even as endogamy persists robustly and matrimonial advertisement continues to specify jati.

For the practitioner, jati is indispensable analytical vocabulary. A desk officer reading Indian electoral returns, a researcher modelling OBC reservation litigation, or a journalist covering Maratha or Jat agitations cannot interpret events through the four-fold varṇa abstraction; the relevant unit is always the specific jati, its regional dominance, and its administrative classification as SC, ST, OBC, or general. Understanding that jati is regional, endogamous, occupationally rooted, and legally enumerated—rather than a uniform national hierarchy—prevents the common analytical error of treating Indian caste as a single, static, country-wide pyramid.