What Counts as a Valid Reading? Distributed Interpretive Authority in GCSE English Literature Assessment

High-stakes literary assessment must accommodate defensible interpretation while supporting comparable marking. This article analyses a census of 332 selectable task options from the 2023 and 2024 GCSE English Literature examinations of AQA, OCR, Pearson Edexcel and Eduqas, linking question architecture to official marking rules. Whole-text directed and extract-to-whole-text tasks comprise 63.9% of coded options. All four awarding systems nevertheless require credit for valid responses beyond indicative content. Only 43 of 332 options (13.0%) explicitly invite candidates to accept, reject or qualify a supplied interpretation, with marked differences between awarding bodies. The findings therefore resist a simple open/closed model. I argue that assessment distributes interpretive authority across agenda, stance, evidence and recognition. The resulting condition is one of bounded interpretive plurality: conclusions may remain plural while the procedures through which they become assessable are tightly organised.