research-document
Cross-domain explanatory-progression corpus manifest v1.0
Cross-domain explanatory-progression corpus manifest v1.0
Frozen: 2026-08-02
Mission: MS-COMM-2026-0004
Unit: one bounded argumentative or definitional sequence, coded as ordered
functional moves rather than as sentences alone.
Definition
Explanatory progression is the observable order by which a text transforms an audience's starting representation into a target representation by orienting, defining, decomposing, connecting, warranting, exemplifying, qualifying, and synthesizing claims. This definition concerns textual architecture. It does not by itself establish comprehension or causality.
Inclusion criteria (frozen before selection)
- One work from science, mathematics, law, philosophy, and software engineering.
- A stable, citable primary edition; translations must be declared.
- A bounded sequence with an identifiable audience state, target state, and warrant or dependency structure.
- Sufficient genre difference to stress-test the rubric, but a shared explanatory function sufficient for comparison.
- Publicly inspectable text and independent context scholarship where available.
Exclusion criteria
- Fame or reception as a proxy for reader outcomes.
- Whole-discipline representation from a single text.
- Sequences whose effect depends primarily on unavailable audio, visual, or performance evidence.
- Translations without edition provenance.
- Causal claims about comprehension without audience testing.
Selected corpus
| Domain | Work and unit | Edition/provenance | Selection reason | Material limitation |
|---|---|---|---|---|
| Science | Charles Darwin, On the Origin of Species (1859), ch. IV, pp. 80–130 | First edition, Darwin Online, F373 | Mechanism explanation built through analogy, examples, objections, and synthesis | Chapter depends on earlier variation and struggle chapters |
| Mathematics | Euclid, Elements, Book I definitions/common notions and Proposition 1 | Joyce web edition based principally on Heath/Heiberg; translation is slightly modernized | Explicit dependency ordering and a canonical six-part proposition | Ancient authorship/editorial history and translation are uncertain |
| Law | U.S. Supreme Court, Brown v. Board of Education, 347 U.S. 483 (1954) | Official U.S. Reports scan via GovInfo | Compact institutional reasoning under precedent, legitimacy, and remedial constraints | Collective judicial voice; outcome and implementation span Brown I/II |
| Philosophy | René Descartes, Discourse on the Method, Part II | 1853 English translation on Wikisource; French original 1637 | Autobiographical problem framing leads to four procedural rules | Translation can alter rhythm and lexical claims |
| Engineering | Fielding, Nottingham, and Reschke, RFC 9110, §§1 and 3 | IETF Internet Standard, June 2022, current errata page | Layered terminology and normative cross-reference architecture | Consensus specification, not an individual writer or tutorial |
Viable alternatives rejected
- Feynman lectures: strong teaching case, but lecture performance complicates text-only comparison.
- Hilbert's Foundations of Geometry: stronger modern axiomatics, but Euclid provides a more visible proposition microstructure.
- Holmes's The Path of the Law: individual legal voice, but less directly comparable to an opinion's institutional proof burden.
- Plato's dialogues: dialogic discovery is valuable but makes unit and speaker intent harder to code reliably in this first pilot.
- Hoare's 1969 axiomatic-programming paper: closer to mathematical proof than operational specification; RFC 9110 better tests layered retrieval.
Selection bias and boundaries
The corpus is English-readable, heavily Western, male-authored or institutionally voiced, and oriented toward elite written traditions. It is a method pilot, not the diverse writer corpus required by the full program. Non-Western, oral, Indigenous, feminine, vernacular, multimodal, and accessibility-first traditions are missing and are high-priority research debt.