drivelength

Drivelength is a proposed metric used to quantify the share of non-substantive or trivial content in a text or spoken discourse. It measures how much language adds little in terms of information, argument, or value relative to total length.

Coined from the words drivel and length, the term appears in discussions of linguistic quality, editorial tooling, and content moderation. It is meant to complement readability and semantic-density measures by focusing on redundancy and filler rather than syntax alone.

A drivelength score can be computed by identifying segments that function as filler, hedges, repetition, or generic statements, then dividing their total length by the text’s overall length. Approaches include lexical inventories, machine-learning classifiers, and metrics for semantic density.

Context matters: some hedges or repetition serve rhetorical or stylistic purposes, and low drivelength does not guarantee quality. Conversely, concise prose can still convey little meaning. The metric is best used as a guide to editing, not an absolute rule.

Applications include editing support, quality benchmarking for user-generated content, and evaluation of summarization or translation systems. Limitations include subjectivity, domain dependence, and the risk of penalizing legitimate nuance. Clear criteria and human oversight improve reliability.