A Recording from 1987 Is Now Telling a Living Singer He Is Wrong

In a room in Dharamsala there is a shelf of cassette copies, digitised twice since, of performances recorded between 1984 and 1991 by a collecting project that was one of the more admirable things the exile community did in that period. Old men and a few old women, most of them born in Tibet, singing and reciting things they had learned by ear from people who had learned them by ear. It was salvage work and it was urgent and the people who did it were right.

Last winter I watched a singer in his fifties perform a section of an epic at a gathering, and afterwards a young man told him, politely and with real respect, that two of his lines were wrong. He had the recording on his phone. He played it. The lines were different.

The singer said: yes. That is Tashi's version. I am not Tashi.

He was not being defensive. He explained afterward, without much heat, that he had learned from a man who had learned in Kham, that the recording was of a man who had learned in Amdo, that the passage in question is where singers have always put their own region's names, and that the two versions had coexisted for as long as anybody could establish. Then he said the thing I have been chewing on since: that in 1987 nobody could have played him a recording, and so nobody could have told him he was wrong.

The archive saved the material and it is beginning to kill the practice, and the mechanism is not neglect. It is authority.

Consider what a recording is, in a tradition that had no fixed text. Before, the answer to "which version is correct" was that the question did not arise in that form; there were lineages, and a lineage was a chain of people, and you could say who you learned from, which was a claim about transmission rather than about accuracy. The performance was the work. Variation was not error — it was how a form covers a plateau the size of Western Europe with no roads.

A recording converts one performance into an object that outlasts the performer, and an object that outlasts everybody becomes, by default, the reference. Nobody decided this. Nobody at that collecting project would have wanted it. But a young man with a phone has a fixed thing and a variable thing in front of him, and the fixed one wins every time, because that is what fixed things do.

The other side of this has better standing than I do, and I have argued it in rooms where that was plain to everybody including me. So let me put it properly.

Without those cassettes there would now be nothing. That is not rhetorical. Of the singers recorded in that decade, all are dead. Several transmitted to no one, because their sons were in a different country doing different work and the whole apparatus of apprenticeship — a boy attached to a man for years, in a place, with time — did not survive the crossing. The material on that shelf exists in the world only because somebody carried a tape recorder up a hill in 1986. Against that, my worry about authority is a luxury complaint made possible entirely by the thing I am complaining about.

And the deeper version of the objection: a tradition without any reference point does not stay various, it dissolves. Variation is only meaningful against something. Ten singers with ten versions is a living form; ten singers with ten versions and no memory of what the others sing is ten people each doing something alone.

I concede that. What I want to argue is narrower and, I think, still true: the problem is not that the recordings exist, it is what we have failed to record alongside them.

Every one of those cassettes is catalogued by title, singer, date and place. Not one is catalogued by lineage. Nobody wrote down who the singer learned from, or where that man had learned, or which passages he considered fixed and which he considered his own. The collectors were preserving a text, because they came out of a manuscript culture and a text is what they knew how to save. The information that would have made the recording readable as one version rather than as the version was in the room, in the singer, free, and nobody asked, and now cannot.

So the young man with the phone is not doing anything unreasonable. He has been handed an object stripped of the only metadata that would tell him what he is holding.

I have a recorder now. It is a small stupid thing and I am not a scholar and I am aware of how this looks. I have been going to see four singers, two here and two in Nepal, and I ask them the same three questions before I ask them to sing: who did you learn this from, where did he learn it, and which lines in it are yours. The third question is the one nobody has ever asked them, and two of the four found it funny, and one of them thought about it for a long time and then went through the passage line by line for forty minutes, naming which were his.

That forty minutes is the most valuable thing I have recorded. It is also, I notice, exactly the sort of material that does not go into an archive well, because it is a man talking about himself rather than performing the tradition, and every instinct anybody has about preservation says the singing is the artefact and the talking is the preamble.

The singer from the gathering died in June. I had asked him the three questions in February. His answer to the third one included the two lines the young man had played at him from a phone, which he had made himself, when he was about thirty, on a road, and had kept.