This study examines what happens when questions are shortened for machine-assisted use: what compression preserves, what it removes, and who remains answerable for the result. It treats summarization as an editorial practice rather than a neutral act, and separates usefulness, source fidelity, and authority to represent. The argument proceeds in stages. Chapters 1–6 establish terms, purposes, a taxonomy of losses, and a research contract distinguishing evidence types. Chapters 7–12 turn to information, context, relevance, chains of summaries, and reversibility. Chapters 13–18 consider collective judgment, groupings, weighting, and agenda setting. Chapters 19–24 examine support, evaluation, preference, and the evaluation of evaluators. Chapters 25–31 develop provenance, documentation, consent, translation, and correction. Chapters 32–36 compare deliberative arrangements and the design of editorial power. Chapters 37–47 move toward benchmarks, failure cases, costs, prospective studies, and an accountable compression protocol. Chapter 48 closes with a research ledger separating established deductions from open commitments.
Readers seeking the core proposal may begin with Chapters 1–6 and 47–48; the remaining chapters supply the distinctions those parts depend on. Every example in this study is constructed, and every empirical evaluation is a proposed design rather than a reported result. The cited sources are used within stated reading limits. Where the analysis points beyond its evidence, the last chapter records the gap as open.