I’ve been making my way through Robert Bringhurst’s The Tree of Meaning: Language, Mind and Ecology, which came out a couple years ago in Canada, but which is now getting an American release from Counterpoint. Bringhurst is probably best known to the readers of this site as the author of The Elements of Typographic Style, though he’s well-known as a poet and translator of Native American languages in Canada. This book is a collection of essays looking broadly at oral, written, and visual language and culture through an ecological lens, a viewpoint not dissimilar to the gatherings of wood s lot, the work of his fellow Canadian Mark Woods.
Much of The Tree of Meaning looks at Native American literature of the Pacific Northwest and how that was gathered by the followers of Franz Boas who gathered and disseminated native stories in songs. I know painfully little about anthropology; perhaps that’s why Bringhurst’s words on the problems inherent in transcribing oral literature seem evocative:

It is true that writing changes literature. It changes it, first of all, by leaving things out. A transcript of an oral poem never captures the fullness of a living performance tradition. And this is where writers become more deviant still. As they take their own dictation, they begin to try to use the resources of writing to patch up the holes and mend the tears they cannot help but make in the fabric of literature as they slip from the oral tradition.

(p. 178 in “The Humanity of Speaking: The Place of the Individual in the Making of Oral Culture”.) The idea of the printed word as being what’s left behind isn’t one we commonly think about, but it instinctively makes sense – perhaps more than ever in a world where a book can be simmered down to a text file constrained to the 127 ASCII characters. Looking to return closer to the original poetry, Bringhurst investigates John R. Swanton‘s typographic transcriptions of Haida myth and finds that document to be more complex than we commonly give it credit for:

The typescript, like the photograph, filters and compresses features of reality. We have to learn to read it – and I don’t just mean we have to learn the language. Learning to read transcriptions of oral literatures is something like learning to read historical photographs. The depth and the color, the sounds and the smells, the coughing and spitting, and a lot of the rest of the nitty-gritty is missing. Through informed imagination, much of that can be restored. And it’s like learning to read music. The point is not just to grasp the grammar and the syntax but to envision, and maybe re-create, a genuine performance.

The typescript looks at first to be plain prose, which is a form designed to minimize the outward individuality of any human voice. Almost no one speaks in genuine prose, but the form is often used – by journalists, linguists, and court reporters alike – for transcriptions. The typographic form we associate with prose makes speakers look like writers.

Swanton went looking for what he couldn’t see, and his typewritten texts are the result. I went looking, in the typescript, for what I couldn’t hear: the oral art, the form and meaning hidden in the flattened landscape of the page.

(p. 188 in the same essay.) He reproduces a portion of the typescript, demonstrating his point:

There’s a lot to think about here, especially now. Each technology in its own way dictates a Procrustean bed, though we’re often not attuned to what the constraints of our technology are. The high-bandwidth networked screen creates possibilities: more can be passed along in the transmission of stories than was before. But constraints dictate form: we know how to read the print book perhaps because that extra information has been sheered away. Learning to read electronically may be complicated.