An audience is the person or group a piece of writing addresses — the reader the writer has in mind when making decisions about structure, vocabulary, tone, and level of detail.

Audience isn’t an abstraction. Karen Schriver distinguished three methods of audience analysis: classification-based (assigning demographics), intuition-based (imagining a reader), and feedback-based (testing with real readers). Her research demonstrated that writers systematically overestimate how well readers understand their text — a gap that classification and intuition can’t close [@schriver1997]. Only feedback-based methods, where writers observe actual readers using the document, reliably reveal comprehension failures.

Linda Flower made a related distinction between writer-based prose — text organized around the writer’s own process of discovery — and reader-based prose — text reorganized for the reader’s needs. The revision from writer-based to reader-based is one of the core moves in producing usable technical writing [@flowerhayes1981].

The vault’s plain language specification defines a default audience: “An adult reader with general literacy, no specialized background in the topic, and a practical reason for reading.” When writing for a different audience, the specification requires an explicit “Assumed audience” section.

  • genre — genres are shaped by their audience’s expectations and recurring needs
  • readability — one measure of whether a text is appropriate for its audience
  • revision — the process of transforming writer-based prose into reader-based prose
  • protocol-aided audience analysis — Schriver’s most rigorous method: observing readers thinking aloud while using a document
  • accessibility — accessible writing extends audience analysis to include readers with diverse abilities
  • translingual writing — challenges assumptions about whose language defines the audience
  • plain language writing — the vault’s operational standard for audience-appropriate writing