Linearity is the convention by which an academic argument moves forward step by step, each idea building on the last, without doubling back, digressing, or asking the reader to reassemble the order. The term entered ELT through Robert Kaplan's contrastive-rhetoric study (1966), which compared paragraph organisation across L1 backgrounds and described English expository writing as linear and deductive in contrast to the inductive, circular, or coordinate patterns observed in other rhetorical traditions. Kaplan's diagrams have been criticised since (Ulla Connor and others reframed the field as intercultural rhetoric to avoid essentialising whole languages), but the underlying observation that English academic prose privileges linear development has held up across decades of EAP corpus work.
A linear text reads as if each sentence answers the question the previous sentence raised. New information enters at the end of one clause and is taken up at the start of the next, the given-new pattern that the theme-rheme system formalises. At paragraph level, the topic sentence projects the structure of what follows and later sentences fill it in without tangents. At essay level, body paragraphs deliver the structure announced in the thesis, in the order the thesis announced.
Most academic writing assessments (IELTS, TOEFL, Cambridge English, EAP exit testing) were calibrated against a linear-expository norm. Markers reward "clear progression throughout," which is the descriptor's name for the perceived effect of linearity at the paragraph and essay level. Learners whose L1 rhetorical tradition tolerates digression, parallelism, or topic looping often produce texts that are grammatically sound and lexically appropriate but score below their language level on coherence, because the reader has to do work the writer should have done.
A text with many connectors can still loop and digress; a text with few connectors can be perfectly linear. Cohesion is the surface signalling, Coherence the perceived logical fit, linearity the directional discipline that produces both. Teaching linking phrases without teaching directional structure is the most common reason intermediate learners stall on writing scores.
The contrastive-rhetoric framing has well-known limits. Culture and rhetorical convention do not map one-to-one, and individual variation within any tradition is large. The pedagogical claim that survives is more modest: the academic-writing genres ELT learners are tested on are conventionally linear, and learners benefit from explicit instruction in that convention regardless of L1. Linearity is taught as a target convention, not an aesthetic universal.