An experimental reading-time method in which participants control the incremental exposure of text (word-by-word or segment-by-segment) by a manual action, and the latency between advances is recorded as a measure of online processing difficulty at each segment while altering normal parafoveal preview and eye‑movement dynamics.

Principle

Longer inter‑segment latencies indicate increased processing cost at or immediately after the displayed segment (e.g., integration, retrieval, disambiguation), but measured times conflate processing at the segment, spillover from previous segments, and task‑specific pacing effects due to the lack of natural eye movements and parafoveal preview.

Demonstration

Illustrative scenario → Situation: Sentence with a temporary ambiguity is presented phrase‑by‑phrase. Recognition: Participant advances through segments by keypresses. Action: Longer pauses occur at the disambiguating phrase. Consequence: Researchers infer the locus of processing difficulty and test syntactic or parsing hypotheses.

Misapplication

Interpreting longer reading times solely as evidence for syntactic complexity without accounting for wrap‑up effects, absence of parafoveal preview, individual pacing strategies, or effect of segment granularity; the error is attributing latency to a single cognitive source.

Consequence

SPR provides controllable, segment‑level timing evidence for theories of sentence processing and integration, but findings must be cross-validated with methods preserving eye movements (e.g., eye‑tracking) to rule out task artifacts.

Reversal

For questions about natural reading where parafoveal processing or regressions are essential, eye‑tracking or gaze‑contingent moving‑window paradigms are preferable; conversely, SPR can be adapted (phrase vs. word presentation) to emphasize different processing scales.

Boundary

Clearly within: word‑by‑word self‑paced sentence reading measuring inter‑key latencies. Boundary case: RSVP (rapid serial visual presentation) where exposure is experiment‑controlled but timing is passive—similar temporal data but different volitional control. Clearly outside: free reading with eye‑tracking enabling natural saccades and parafoveal processing.

Semantic Tension

Control of exposure ↔ Preservation of natural reading: SPR increases experimental control over input timing at the cost of disrupting normal perceptual mechanisms (parafoveal preview, regressions), forcing trade‑offs in inference about natural processing.

Synthesis

Self‑Paced Reading isolates temporal costs tied to specific segments under controlled presentation; it is a powerful tool for testing processing hypotheses but requires awareness and complementary methods to separate genuine cognitive difficulty from task‑induced artifacts.