# Comparative Greek & Latin grammar — sources

The comparative section is **anchored on the site's two reference grammars** and **grounded in
the comparative scholarship**. Every comparandum cites its source; the syntax dimension cites
Whiton by rule number, the other four dimensions cite the scholarly works by page.

## The two anchors (critical foundations)

- **Smyth**, *A Greek Grammar for Colleges* (1920) — the Greek anchor. Live at
  `/greek-grammar/smyth.html`; comparanda deep-link to `#s{N}`. (Perseus-derived TEI, CC-BY-SA.)
- **Woodcock**, *A New Latin Syntax* — the Latin anchor. Live at
  `/cicero-reader/cicero_syntax.html`; comparanda deep-link to `?sec={N}`.
  **Wheelock's Latin** supplies the learner-facing chapter comparator (`#ch-{N}`).

## The comparative scholarship (page-cited)

| Work | Role | Dimensions fed |
|---|---|---|
| **Whiton**, *Parallel Rules of Greek and Latin Syntax* (Ginn & Heath, 1877) — public domain | the **syntax spine**: 81 numbered parallel rules with Greek↔Latin `cf.` cross-references | syntax |
| **Sihler**, *New Comparative Grammar of Greek and Latin* (Oxford, 1995) | the Indo-European background; case-syncretism, the dual, augment, aorist↔perfect, optative→Latin subjunctive | morphology |
| **Pinkster**, *Latin Syntax and Semantics* (Routledge, 1990) | Latin functional grammar; tense/aspect, case-as-valency, mood value, word order, the article, particles | semantics, pragmatics |
| **Mason**, *Greek Terms for Roman Institutions: A Lexicon and Analysis* (Hakkert, 1974) | the Greek↔Roman institutional lexicon | vocabulary |
| **Oniga** (ed.), *Formal Linguistics and the Teaching of Latin* (Cambridge Scholars, 2011) | formal/comparative framing of Latin word order and tense | pragmatics, semantics |

### Sourcing discipline (no fabrication)

- The full Whiton rule inventory was transcribed from the page images (English statements +
  construction tags + `cf.` cross-refs); its OCR-garbled Greek/Latin example text was **not**
  used — clean Greek examples come from the Smyth data, Latin §§ from Woodcock.
- The four image-only works (Sihler, Pinkster, Mason; the Oniga excerpt is born-digital) were
  read page-by-page for specific claims; each curated comparandum carries the page number and a
  verbatim quote. Nothing was bulk-OCR-ingested.
- Smyth §§ and Woodcock §§ are attached only where independently verified against the live data;
  rules with no confident anchor remain Whiton-only rather than carry a guessed §.

### Known limitations

- The Indo-European comparative grammar volumes (3 vols) on disk are image-only with no usable
  text layer; they were not mined — Sihler covers the same ground with a clean modern treatment.
- The on-disk **Oniga** file is a front-matter + Introduction excerpt; its chapter on the
  Greek↔Latin subjunctive (Krasukhin) and the Salvi word-order chapter are not present, so all
  Oniga citations are from the Introduction.
- Two semantics comparanda (`sem-tense-aspect`, `sem-tense-synthetic`) carry an `audit_flag`
  for citation strength; see `AUDIT-FINDINGS.md`.