Concept

Tertiary orality

Tertiary orality is the condition of language produced without a composing subject.

Walter Ong described a primary orality that preceded writing, and a literacy that displaced it. He added a secondary orality, carried by radio and television, which reproduced the intimacy of speech while remaining anchored in a literate order — its scripts were written, its speakers composed.

A third condition has now arrived, and it is not a variation on the second. Large language models emit utterances bearing every mark of composition: register, cadence, argumentative shape, the appearance of having been meant. Behind them stands no one who meant anything. The utterance has a form without having an origin.

This matters because the whole apparatus of literate trust runs on the assumption that a text implies an author, and that an author can be held to it. Citation, attribution, plagiarism, authority, liability, the byline itself — each is a device for connecting words to a person who can be asked about them. Tertiary orality severs that connection while leaving every one of its outward signs intact.

The consequence is that our procedures for evaluating a claim now run on a signal that has been decoupled from what it once indicated. We read fluency as evidence of thought, coherence as evidence of care, and grammatical confidence as evidence of a mind that has checked something. Those inferences were reliable for as long as producing fluent text was expensive. They are now free.

Tertiary orality is my own term. The argument is set out at length in a paper currently under review; this page exists to fix the definition and the date. The Gutenberg Parenthesis, which frames it, belongs to Lars Ole Sauerberg and Thomas Pettitt.