The Indian caste system is conventionally explained through the Vedic varṇa model — the four-tier hierarchy of Brahmin, Kṣatriya, Vaiśya, and Śūdra, with untouchable groups outside the four-tier structure. The model is grounded in the Puruṣa-Sūkta (Ṛgveda 10.90) and elaborated in the Dharma Sūtras. The standard story treats the caste system as a development of this Vedic-Brahmanical structure.
George Hart's research on early Tamil literature offers a substantial alternative reading.1 Hart argues that the caste system as we know it in South Asia has two origin-sources, not one — the Vedic varṇa system from the Indo-Aryan north, and an indigenous South Indian ritual-kingship-and-low-caste-specialists system from before substantial Brahmanical influence reached the south. Modern Indian caste structure is the result of these two systems combining.
Samuel takes up Hart's thesis in Chapter 42 as one of the major reframings the Two-Worlds Thesis enables. Hart's reading dismantles the simple Indo-Aryan-only origin story for caste and identifies a parallel South Indian source that contributes the most distinctive feature of modern caste — the structurally-essential role of low-status polluted specialists.
Hart's reconstruction of early Tamil society (from the Sangam-period literature, roughly third-sixth centuries CE but reflecting earlier patterns) shows a political-religious system structured around ritual kingship.
The South Indian king's authority was grounded in ritual power. His drum (the muracu) was inhabited by a spirit that needed regular blood-and-liquor sacrifice. His warriors fought as ritual-religious operation as well as military operation. The system depended on a class of specialist ritualists — drummers, bards, spirit-mediums, funeral priests, priests of the deity Murugan — who handled the polluting ritual operations on the king's behalf.
These specialists were of low status. Their pollution-handling was essential to the king's ritual functioning, but it contaminated them. They occupied a structural position equivalent to what later Indian caste structure calls dalit (untouchable) — essential for the wider community's ritual functioning, ritually polluting in themselves.
Hart's key claim: this structural pattern (king + low-caste-polluted-specialists) predates substantial Brahmanical influence in South India. It is not derived from the Vedic varṇa system. It is an indigenous South Indian political-religious arrangement.
Hart argues that the modern Indian caste system results from the combination of two sources:
Indo-Aryan varṇa: the four-tier hierarchy from the Vedic-Brahmanical tradition. This contributes the basic vertical structure (high-low ordering), the categories of Brahmin, Kṣatriya, Vaiśya, the legitimation through Puruṣa-Sūkta mythology.
Indigenous South Indian pattern: the structurally-essential low-caste polluted specialists. This contributes the bottom of the caste hierarchy — the dalit groups, the elaborate pollution-distinction logic, the operational ritual function of polluted specialists for the higher-status community.
Modern Indian caste has both elements. The four-varṇa structure (or its multi-jāti elaboration) provides the upper-and-middle tiers. The South Indian pattern provides the bottom tier and its distinctive functional integration into broader society.
The thesis matters because it dismantles the simple Indo-Aryan-only origin story for caste. The pollution-distinction structure that is most distinctive about Indian caste is not Vedic. It is South Indian (or perhaps Central Gangetic, by extension), absorbed into the synthesized caste system as Brahmanical influence spread south.
The dual-origin thesis bears on several religious developments.
Tantric tradition's emergence: the kāpālika-style ascetics who become the renunciate-counterpart of low-caste polluted specialists are drawing on the indigenous South Indian / Central Gangetic pattern, not just on Vedic-Brahmanical innovation. The pollution-as-power logic of Tantric tradition has its substrate in the actual social-political reality of low-caste specialists serving structurally-essential ritual functions for kings.
Yakṣa religion: the local-deity worship that Samuel identifies as the Central Gangetic religious substrate is parallel to the South Indian pattern Hart describes. Both involve specialists handling locally-polluted ritual operations; both contribute to what later becomes Tantric tradition.
Caste reform movements: every major Indian caste reform movement (Buddhism, Jainism, bhakti traditions, modern movements) has had to address both varṇa and the pollution-distinction structures. Addressing only varṇa misses the pollution-distinction dimension; addressing only pollution-distinction misses the varṇa structure. The dual-origin thesis explains why caste has been so resistant to reform.
Connects to Two-Worlds Thesis — Hart's pattern fits the Central Gangetic / non-Vedic world type. Connects to Yakṣa Yakṣī Cult Laukika Popular Religion — the local-deity worship substrate. Connects to Witzel Kuru-Pañcāla State — the varṇa-construction origin. Connects to Kāpālika-Style Ascetics — the later Tantric inheritance.
The dual-origin thesis is foundational for understanding how Tantric tradition draws on pre-Vedic substrates rather than just elaborating Vedic-Brahmanical material.
A specific case Samuel notes: the Mon — low-status drummers, blacksmiths, and other specialists in Tibetan Himalayan communities (Ladakh, Nepal, neighboring regions). The Mon retain the structural position Hart identifies for South Indian low-caste specialists: essential ritual function for the broader community, but ritually compromised and socially low-status.
The Tibetan-Himalayan case is significant because it sits outside the major Indic caste system but preserves the same structural pattern. This suggests the pattern is more general than just the Indian caste system — it is a particular political-religious solution to the problem of ritually-essential-but-polluting specialist labor that recurs across multiple cultural settings.
The pattern recurring across cultural boundaries supports Hart's argument that it is structurally indigenous rather than ideologically Brahmanical. If it were Brahmanically constructed, we would not expect to see it preserved in non-Brahmanical Tibetan Himalayan contexts.
You are writing a vault page that discusses Indian caste. Default questions:
Which dimension of caste are you discussing? The varṇa vertical hierarchy? The jāti horizontal specialization? The pollution-distinction structure? Each has different origins and operates differently.
If you are discussing the four-varṇa structure, you are working with the Indo-Aryan-Brahmanical contribution. The Kuru-Pañcāla state-construction project produces this.
If you are discussing the pollution-distinction structure and untouchability, you are working with the South-Indian-indigenous contribution. Hart's thesis applies.
If you are discussing the modern integrated caste system, both contributions are operating. The integration history matters.
Hart's thesis is one of several contending models for caste origins. Louis Dumont's Homo Hierarchicus (1972) treated caste as a unified Brahmanical-pure-hierarchy system. Later anthropological work (Berreman, Mencher, Quigley, Searle-Chatterjee and Sharma) dismantled the unified-system reading. Susan Bayly's Caste, Society and Politics in India (2001) emphasized the relative modernity of much current caste structure.
Hart's contribution is the specific identification of the South Indian indigenous source as distinct from Vedic varṇa. Samuel synthesizes Hart with the broader anthropological reassessment.
The structurally-essential-but-polluting specialist pattern has parallels across cultures.
In history, the same pattern appears in Japanese burakumin (descendants of low-status occupations including butchery, leather-work, executioner), in European medieval categories of "infamous trades" (tanners, midwives, executioners), in West African groups of caste-like specialists. The insight neither domain alone produces: many human societies develop a structural position for ritually-essential-but-polluting specialists; the position has both a functional logic (someone has to do the work) and a cultural-religious logic (the work pollutes the worker); recognizing the pattern in one case helps recognize it elsewhere.
In psychology, the analog is the projective-identification mechanism where unwanted material is offloaded onto a designated other. The vault holds material on projective identification. The insight: at societal level, pollution-handling specialists carry projected unwanted material for the rest of the community; the psychological mechanism that allows individuals to offload onto specific others is the same one operating socially. Understanding the individual mechanism helps understand the social pattern; understanding the social pattern shows how the individual mechanism scales.
The handshake to behavioral-mechanics sits in the analysis of how low-status-polluted specialists function as power-architecture. The king's authority in Hart's South Indian model is partly sustained by the existence of specialists who can handle what the king cannot ritually touch. The vault's behavioral-mechanics material on power architectures can analyze this as influence-system structure.
Treating Hart's thesis as definitive overstates current scholarly certainty. The thesis is well supported but contested. Alternative reconstructions of caste origins exist.
Treating the dual-origin reading as exoneration of caste injustice would misuse it. The structural-functional analysis explains origins; it does not justify continuation.
Strongest evidence: convergent appearance of the pattern in early Tamil literature, in Tibetan Himalayan communities, and (after Brahmanical influence) integrated with the varṇa hierarchy in modern Indian caste structure.
Open questions: How exactly did the integration happen? At what historical period did Brahmanical influence reach South India substantially enough to begin the varṇa-pattern integration? What are the implications of the dual-origin reading for modern caste-reform efforts?
Sharpest implication: Vault discussions of Indian caste should distinguish the varṇa structure (Indo-Aryan-Brahmanical origin) from the pollution-distinction structure (South Indian indigenous origin). Conflating both into "the caste system" loses important analytical distinctions.
Generative questions: