Ask what Hinduism is and the answer usually starts with a list: many gods, many scriptures, many festivals, one civilisation. That is a description of a landscape, not of an institution. The useful question is narrower — what would you have to agree to in order to belong? — and the answer is: nothing in particular.

That is not a defect of the traditions involved. It is the consequence of a category that was assembled from the outside, named after a river system, and then gradually adopted by the people it had been applied to. Reading the traditions as a family rather than a church explains both what holds them together and why almost every attempt to speak for them on behalf of everyone fails.

The word arrived as an address

Hindu began as a geographical word. Persian and later Arabic usage applied it to the people living beyond the Sindhu — the Indus — and by extension to the land. It said where someone was from, not what they believed. The same term shows up in Greek as Indoi for the people of the same river basin. Nobody in the subcontinent described themselves as a Hindu in the sense of belonging to a religion called Hinduism; they described themselves by lineage, by region, by sect, by the deity they were devoted to, and by the community into which they were born.

The shift from address to category happened in the nineteenth century, in stages, and both parties to it had reasons. British administrators needed countable units; the decennial census from 1871 onward asked people to tick a religion, and the tick-box created the thing it was measuring. At the same time, Indian reformers arguing for social change needed a name for the tradition they were reforming — a name that could be contrasted with Christianity and Islam, and could therefore make parallel claims about age, coherence and dignity. The category that emerged was useful to both, and it has never been a neutral description since.

What a church is organised around, what these traditions share, and what actually holds them together A CHURCH THIS founder one, dated none — no founding figure defines membership creed one, recited none — no statement anyone must assent to authority one seat many, competing, none binding on all canon closed list open, layered, regionally different WHAT DOES BIND narrative repertoire Rāmāyaṇa, Mahābhārata, Purāṇas — told everywhere ritual grammar mūrti, pūjā, prasāda, pilgrimage — recognisable service economy priests, temple trusts, festival contractors A family shares a repertoire, a way of doing things and a set of obligations. It does not share a creed — which is why nobody can speak for it. The legal category is tighter than the theological one: statute needs a boundary, doctrine does not.
Four things a church is organised around, none of which is present here — and the three things that do hold these traditions in the same conversation.

What actually binds them

If there is no creed, something else must be doing the work of making these traditions recognisably related. Three things are.

The first is a shared narrative repertoire. The Rāmāyaṇa and the Mahābhārata are known from Tamil Nadu to the Himalayan foothills, in retellings that differ enormously and in a few cases reverse the sympathies entirely, but which are all retellings of the same story. Versions of the same epic can serve as royal propaganda in one kingdom, as devotional text in another, and as village theatre in a third. Sharing a story is not sharing a doctrine; it is sharing a set of reference points that arguments can be conducted with.

The second is a ritual grammar: the mūrti in the shrine, the sequence of pūjā, the distribution of prasāda, the circumambulation, the vows and the pilgrimage. A visitor from Maharashtra can follow the shape of a service in a Tamil temple without understanding a word of it, because the structure is the same even where the language, the deity and the theology are not. This is the strongest single piece of evidence that a family resemblance exists: it is visible in behaviour, not in belief.

The third is an economy. The traditions are served by hereditary priests, funded by temple endowments and land, staffed by cooks, flower sellers, garland makers, drummers, sculptors and printers, and connected by pilgrimage routes that move money and people across state borders. Anyone who wants to understand why a particular practice survived at a particular place should look at who was paid to keep it going. That is entry 08 of this pillar, and it is the least mystical and most explanatory chapter in it.

The law is more precise than the theology

Since no authority can rule on who is a Hindu, the practical definition has ended up in statute — and the statute is remarkably clear. The Hindu Marriage Act of 1955 opens by stating that it applies to any person who is a Hindu by religion in any form or development, and then explicitly includes Buddhists, Jains and Sikhs, and any other person domiciled in India who is not a Muslim, Christian, Parsi or Jew.

Read that carefully, because it does two things at once. It asserts that Buddhism, Jainism and Sikhism fall inside the category for the purposes of family law, which those communities have disputed since — Sikh organisations in particular have campaigned for a separate personal law for decades. And it makes the category residual: anyone not on a short list of others is in.

This is the clearest available answer to the question the pillar opened with. A religion usually defines itself and the state registers the result. Here the state had to define it, because there was no body that could.

What the family metaphor buys you

Calling it a family rather than a church is not a way of being polite. It changes what you predict. In a church, disagreement tends to produce separation: a schism, a new denomination, a boundary. In a family, disagreement is normal and does not produce exits, because membership was never conditional on agreement. Two relatives can hold incompatible views about the same ancestor and remain at the same table for fifty years.

That is why the traditions of this pillar borrow from each other so freely and deny it so consistently, why a temple can host a deity whose story belongs to a rival sect, why a village festival can be attended by everyone and administered by one lineage, and why the political question of who speaks for the majority is so hard to settle. There is no chair to sit in. That is the structural fact from which most of the rest of this pillar follows.

Sources and checks

  1. Lorenzen, David N. "Who Invented Hinduism?", Comparative Studies in Society and History 41(4), 1999; expanded in Who Invented Hinduism? Essays on Religion in History, Yoda Press, 2006
  2. Hindu Marriage Act, 1955 Section 2(1), application of the Act; Section 2(2) saves the Act's application to scheduled tribes
  3. Census of India 2011 Religion tables, C-01: Hindu population 966.3 million, 79.80% of total
  4. Fuller, C. J. The Camphor Flame: Popular Hinduism and Society in India. Princeton University Press, rev. ed. 2004
  5. Oberoi, Harjot The Construction of Religious Boundaries: Culture, Identity and Diversity in the Sikh Tradition. University of Chicago Press, 1994

Last checked 30 October 2026 Yellowsea Huang Corrections