TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Ethics & Provenance · AGE 19—21 · POSITIONAL · HELD
Consent, community and the generated monument
Living heritage is decided with the people who keep it, and consent comes before circulation.
The dilemma
A temple in a Karnataka town, still in worship, its outer mandapa altered in the 1970s with concrete and enamel paint. For a studio project you generate a "restored" view stripped back to an imagined twelfth-century state, and it is beautiful. The temple committee has not seen it. The priest's family has repainted that mandapa for three generations. A heritage group would like to use your image in a fundraising post. Do you let them?
The choices
Let it go out — it is your image and it argues for conservation. Hold it until the people who use the temple have seen it and said what they think of being shown their building without the parts they made. Or change what you make: generate nothing "restored" until you have been on site and asked what the building's keepers consider its heritage to be — which may include the 1970s mandapa.
The consequence
The image that goes out becomes the reference. Fundraisers, journalists and the next student will restore toward it, and the committee will be asked why their temple does not look like the picture. Holding the image costs a conversation, and the image may come back changed — the keepers may want the paint kept, the concrete gone, or neither. The third choice costs the most time and produces the only version with any standing: a proposal made with the people whose heritage it is, inside a tradition India's own conservation charter describes as living — master builders, sthapatis and raj mistris who "continue to build and care for buildings following traditions of their ancestors."
The case
A fourth-year elective on a town's heritage precinct. One team generates views of the bazaar street "as it was" — no signboards, no cables, no modern shopfronts — and the local paper runs them as how the bazaar should look. The shopkeepers whose fronts were erased were not asked; several had rebuilt those fronts with their own money after a fire. The tutor's question at jury was not about the images' quality. It was: who is in the picture, who was removed from it, and who decided?
Our position
Living heritage is decided with the people who keep it — not by a model, and not by the architect alone. Custodians, owners, trusts, the state's archaeology departments, scholars and the wider public all hold real claims on a place; the Lab's position is not that the users' voice is the only one, or always right, but that the architect may not be the one who decides by default. A generated reconstruction of a place in use is circulated only after the people who use it have seen it and spoken, and it is captioned as a hypothesis with its author named. Consent comes before circulation.
Why we hold it
The Nara Document (1994) says judgements of value and authenticity "may differ from culture to culture, and even within the same culture," and must be made "within the cultural contexts to which they belong." UNESCO's own definition of intangible cultural heritage says it "can only be heritage when it is recognized as such by the communities, groups or individuals that create, maintain and transmit it." India's INTACH charter (2004) recognises the living tradition of builders and the concept of jeernodharanam — renewal as a continuing practice — as part of conservation itself. A model has no community and no context. It has a distribution over photographs, most of them of monuments with no people in them. So a generated "restoration" is a claim about a contested past that nobody has made, and an image nobody made is the easiest thing in the world to pass around. The architect's job is to put a person back behind it: a name, a caption, a conversation.
The strongest objection
"Ask the community" assumes there is one, and it assumes that whoever answers is the community. The body recognised as the temple's voice may be a dominant-caste trust, an all-male committee, the property-owning faction or the state department — and it may exclude the women who sweep the mandapa, the tenants against its wall, the lower-caste devotees, the migrants who use it, the dissenting descendants. Giving "the community" a veto can legitimise that hierarchy instead of protecting the heritage, and it can collide with statutory conservation duties and with legitimate scholarship and public interest. Consent from one is not consent from all; waiting for all is a veto on any proposal; and meanwhile the image tools are public and cheap — if the careful student does not make the reconstruction, a tourism board or a political campaign will, with less care. A position that hands the judgement to "the people who keep it" has to say which people, and it has not.
What would make us revise it
Evidence from real projects that consent-first practice prevents good conservation rather than improving it — proposals that died in consultation while the buildings fell. A settled Indian procedure for community consent on heritage imagery, at which point this card stops being a position and becomes the procedure. Or a workable method for hearing the voices a committee excludes — at which point the position is rewritten around that method, which is the revision the objection above is asking for. Reviewed every edition against what heritage bodies say and what the Lab sees in its own work.
Prove it
Take one generated reconstruction of a place in use — yours or a published one. Write the caption it should have carried: author, method, what is documented, what is conjectured, who was consulted. Then list the people and bodies with a stake in that place and mark which of them saw the image before it circulated.
Take it to crit
For any generated reconstruction of a place people still use, ask the student three questions in order: who was asked, who is shown, and who decides what counts as the heritage here? If the answers are "nobody, nobody, me," the project has not started.
How it works
FOUR PERMISSIONS on this map separates legal, contractual, institutional and ethical permission. This card is the ethical question at its hardest, because many people are affected and the harm is to memory, not to property. The Venice Charter's line at conjecture — RESTORED OR INVENTED — protects the record; Nara and the INTACH charter widen the question to whose judgement counts. India's protection frame — the Archaeological Survey of India for centrally protected monuments, state departments of archaeology for others, and a majority of architectural heritage that INTACH's charter notes "still remains unidentified, unclassified, and unprotected" — decides who may touch a building. None of it decides who may picture one, which is why this question sits in ethics and provenance rather than in law. SAYING YOU USED IT is where the caption gets written. The objection above is real: the Lab has no formula for which voice in a divided community prevails, nor for hearing the voices a committee leaves out — only the rule that the architect may not be the one who decides by default, and the practice of asking more than one body before the image moves.
What this idea builds on
What this idea opens up
Sources
Open this idea on the map · The complete map · Logika · RBDS AI Lab, India · revised every edition.