How to Record Your Temple's Oral History

The knowledge that has no shelf

In most old temples there is at least one person who knows things that were never written down. The priest who can tell you which stones in the compound wall came from the earlier shrine. The woman two lanes away who remembers the year the tank went dry and the utsava murti, the processional image, was carried to a neighbour's house for safekeeping. The family that has always tied the first garland, and cannot quite say why, only that it has always been theirs to tie.

This is oral history — testimony rather than folklore in the loose sense. A named person, on a known date, saying what they saw or what they were told and by whom. It sits alongside the sthala purana, the temple's own account of its site and origin, which usually reaches back into legend. The sthala purana explains why the place is holy. Oral history tells you what happened here in the last eighty years, and who was standing there when it did.

Hinduism has never been shy about knowledge that travels by mouth. The Gita says of its own teaching that it came down through parampara, a succession of people handing it on, and that the royal sages knew it that way — and in the same breath that the chain lapsed with the passage of time and had to be spoken again (4.2-4.3). That is the honest shape of oral transmission: not a thing that survives by itself, but a thing that survives when someone takes the trouble. The Mahabharata reaches us framed as a recitation by the bard Ugrashrava to sages in the Naimisha forest; the Ramayana as a song learned by two boys and performed aloud. The written text is the sediment of speech, not its replacement.

What is lost, precisely

When a temple's oral memory goes, three different things go with it.

The first is narrative: the founding story, the flood that moved the image, the dispute that split one festival into two. The second is reasoning. Why the shrine faces an odd direction. Why the procession halts at a particular corner where nothing now stands. Why this town keeps a festival on a different tithi — the lunar day a festival is actually fixed to — from the next district. Practices survive longer than their explanations, and a practice whose reason is forgotten becomes vulnerable — the first thing to be dropped for convenience, or "corrected" by a well-meaning visitor with a printed manual and no local knowledge.

The third loss is the plainest: names. Who paid for the roof in a hard year. Who cooked for four decades. Who walked eleven kilometres each morning to open the door. None of them appear in any record, and within two generations they are nowhere at all.

Doing it well

Start with who to ask, and resist the instinct to ask only the priest. He is essential, and he is one source. So are the cook, the person who sweeps, the drummer, the flower seller at the gate, the oldest women of the surrounding houses, the trustee's rival, and the man who left for the city decades ago and returns each year — migration makes people notice change that residents stop seeing.

Ask concrete questions and let them open outward:

• Tell me about the first time you came here.
• Walk me around and tell me what each thing is.
• What is done differently now from when you were young?
• Who taught you that? Who else would know?
• Was there a year when things went badly here?

Get consent before you record and again when you finish, in the person's own language, in plain words: this will be public, it will be searchable, your name will be on it unless you would rather it wasn't, and you can ask me to remove it. Some things will be said after the recorder is off. Those things stay off.

Record audio even if you also take notes; memory of a conversation is worse than everyone believes. Transcribe in the language spoken, and keep any translation separate and marked as yours. Leave in the hesitations and the corrections — do not iron speech into an essay. Then attach provenance to every file: name, date, place, language, who asked the questions, and the distinction that matters most, whether the speaker saw it or was told it. "I was standing there" and "my grandmother used to say" are two different kinds of statement, and the difference should survive in the record.

The honest difficulties

Versions conflict. Two families will each tell you their ancestor brought the deity here. Record both, name both, and do not adjudicate. A page that says "there are three accounts of this, and here they are" is more honest and more useful than a smooth single story.

Stories also grow. Distances lengthen, crowds swell, dates drift, and marvels attach themselves to founders. Do not strip this out. A miracle account is accurate evidence of what a community holds about itself, even when it is not an event log. Keep it, and keep the register visible.

Who speaks is never neutral. The trustee's version travels easily; the sweeper's does not. Some people were, and in places still are, kept outside the very doors you are documenting, and their memory of the temple is a memory of exclusion — that is part of the history, not an interruption of it. Women often say different things when a husband or son is not in the room; interview separately and unhurriedly, wherever people are comfortable.

Traditions differ genuinely, and a recorder is not a referee. The same shrine's origin may be told with a Shaiva emphasis in one telling and a Vaishnava one in another. Festival dates and observances vary by region, lineage and family. Advaita, Vishishtadvaita and Dvaita hold different positions on the relation between the self and God, and a devotee's account will carry the grain of their school. Note the difference plainly and move on.

Finally, hold the line between a story and a historical claim. That a community tells you a Chola king stood on this spot is excellent evidence about the community. It is not, by itself, evidence about the king. Both belong in the record, with the seam left visible. And a recording is never a title deed: oral history has been used to press claims over land, entry rights and precedence, and anyone collecting it should refuse that use.

Why this belongs online, and what that takes

An atlas — this site keeps one, a few thousand entries deep — can tell you that a temple exists, its deity, its town. It cannot tell you why the family in the third street brings the first mango of the season, or what the older image looked like. That knowledge is held by people who are getting older in villages their grandchildren visit twice a year, and it does not survive by being cherished. It survives by being written down, attributed, and put somewhere the grandchildren will actually look.

But online is not a place; it is a set of promises about upkeep, and they are worth making explicitly.

• Keep the master. Record and archive in an uncompressed or lossless format, and keep a second copy somewhere that is not the phone you recorded on. The recording will outlive the device, the app, and probably the service.
• Publish the transcript as plain text beside the audio. Text is what search finds and what still opens in twenty years. Audio alone is invisible to anyone who does not already know it exists.
• Put the provenance on the page, not in a private spreadsheet. Speaker, date, place, language, interviewer, and whether the speaker saw it or was told it. A recording without those fields is an anecdote.
• Say who may reuse it, and on what terms. Decide this with the speaker at the time, in the same conversation as consent, and write it down.
• Write the takedown policy before you publish anything. When a speaker or their family asks for something to come down, it comes down promptly, and the policy that says so should be visible on the page from the first day.
• Link it both ways. An account should hang off the temple's atlas entry, and the atlas entry should point back at the voices, so someone who arrives looking for a place finds the people, and someone who arrives for a person can find the place.

If there is a temple your family goes to, the next time you are there, find the oldest person sitting in the corridor. If there isn't, the same is true of anywhere your family has been attached to for fifty years. Ask them one question. Write down what they say, and their name.