Oral archive searchable by voice
A studio exploration, with no client behind it. An elder passing on a remedy or a story leaves no text to index, only a recording nobody knows how to search. This concept searches the sound itself: you ask the question out loud, the archive answers with excerpts.
This project is an internal studio exploration. It was not commissioned by a client and has not been deployed.
Context
A great deal of knowledge is passed on by speaking, in languages barely written or not written at all. The recordings exist, often hours of them, carefully kept, but they sit idle: with no text, no archival tool knows how to find a given passage.
Challenge
Finding one sentence in hours of speech normally requires a transcription. When a language has no settled spelling, that transcription does not exist, and waiting for it to appear amounts to doing nothing.
Our approach
Search rests on acoustic resemblance, not on the written word. You speak, the archive returns the passages that sound like the request, and you move by ear from one excerpt to the next. Writing stays available for those who want to read, but it is no longer the compulsory way in.
Result
An interface prototype and an indexing principle, left at the exploration stage: neither commissioned nor deployed.
What this concept sets out to prove
Question an assumption rarely discussed, that all search goes through writing, and look at who is left outside when it is kept.