PWK — Sanskrit-Wörterbuch in kürzerer Fassung (1879–1889)

Chapter authored per Decision 29 Tier A. Position 3 in the atlas ordering — the missing link between PWG's kosha-rich apparatus and MW's hedge-collapsed apparatus.

Trust Block

• Evidence: CDSL v02 pwk.txt, src/data/cross-dict.json, and linked dictionary-profile/audit notes.
• Limitations: narrative atlas profile; counts summarize committed/generated artifacts and do not replace source edition inspection.
• Validation: checked by npm run build; chapter consistency is tracked in _consistency_audit.
• Owner repo: csl-atlas.
• Next use: use this dictionary profile as context, then open source-linked records or compare the lemma in Reader Lookup.

1. Overview

Otto von Böhtlingk's own shorter counterpart to PWG, published St. Petersburg 1879–1889 in seven slim Lieferungen / volumes. Where PWG (1855–1875) ran to seven folio volumes and 570,817 source citations, PWK condenses the same lexical material into a smaller seven-part reference by dropping nearly all of PWG's indigenous-kosha apparatus: the top PWG sigla ŚKDR. (20,109×), MED. (7,176×), and H. an. (6,619×) all collapse to zero in PWK. The result is a dictionary with more headwords than PWG (170,556 vs 123,366 — Böhtlingk added new material in 14 years of intervening scholarship) but fewer citations per headword (0.51 vs PWG's 4.63 — an order-of-magnitude reduction). PWK is therefore the missing link in the European-Sanskrit-lexicography lineage: the editorial moves that MW 1899 would complete (collapse named-kosha sigla, compact the citation apparatus) are first attempted by PWG's own editor, here.

2. Profile table (5 primary types × profile)

Per PAPER.md §5, restricted to the types PWK's data supports:

Citation profile spread: 7.7 pts — between PWG's 0.4 (uniform) and MW's 11.3 (most selective). PWK is the compact-Petersburg midpoint of the type-citation differentiation effect: dropping from PWG's near-uniform 98 % citation rate to ~40 % across most types, with adjectives slightly favoured (45.2 %) and noun-m slightly disfavoured (37.6 %).

Block-presence by name (cross-dict-common subset):

3. Citation density and apparatus

0.51 <ls> per record — sparser than MW (1.09) and 9× sparser than PWG (4.63). Top 12 sigla (analysis/LS_HEDGE_CHECK.md):

Compare to PWG's top 12: PWK's top sigla are completely different from PWG's. PWG's top three were all indigenous Sanskrit lexicons (ŚKDR, MED, H. an.); PWK's top three are an indological catalogue (GAL.), a manuscript catalogue (OPP. CAT. 1), and a Tanjore manuscript inventory (BURNELL, T.) — none indigenous-kosha. Böhtlingk's editorial choice for PWK was to drop the named-kosha apparatus in favour of textual and cataloguing references that had emerged in the 14 years between PWG vol. 7 and PWK vol. 1.

This is the central PWK finding: PWK already collapsed PWG's indigenous-lexicon apparatus before MW did (Phase 3.10 finding). MW's <ls>L.</ls> is not the first compression — it is a re-introduction of a compressed indigenous-lexicon marker that PWK had effectively eliminated.

The arc: PWG kept the named-kosha apparatus (high cost, high fidelity); PWK dropped it (the editor's own retraction); MW re-introduced an aggregate marker that says "the evidence is kosha but I don't say which" (a third compromise).

4. Hedge analysis — PWK uses typographic * prefix

PWK has 0 tagged <ls>L.</ls> instances but does use a typographic asterisk * as an entry-initial mark (the digitisation preserves these inline in the headword field, not as tagged <ls>). The asterisk semantics in PWK is not the same as Cappeller's 1891 asterisk: PWK uses * to mark words that occur only in lexicons or grammarians — the same general meaning — but applied less systematically than Cappeller and never tagged. PWK therefore sits between the typographic discipline of CAE 1891 and the no-hedge discipline of PWG: it has the concept of a hedge but neither systematic typographic application nor any tagged implementation.

5. Lineage statement

PWK occupies the abridgement node in the European-Sanskrit-lexicography lineage. It is derivative-by-design: Böhtlingk's stated intention was to produce a handier working reference for scholars who could not afford or carry the seven-volume PWG. The abridgement strategy — keep PWG's lemma set, drop most of PWG's citation apparatus, expand selectively where new scholarship demanded — is what MW 1899 would later adapt for English-language users. PWK predates MW 1899 by 10–20 years and demonstrates that the editorial moves MW would make were already on Böhtlingk's mind a decade earlier.

Full lineage in PWK/DICT_PROFILE.md.

6. Cross-references — divergence/convergence with adjacent chapters

7. Decisions log

• 5-type profile (not 8+1) per same reason as PWG: PWK lacks <info>, <bot>, <bio>, <lang> so encyclopedic / IE / verbal-lemma primary types from MW's refactored typology cannot be detected.
• No <ls>L.</ls> but typographic * exists — PWK uses asterisk less systematically than Cappeller (CAE); the digitisation preserves these in the headword field, not as tagged <ls>.
• The "missing link" finding is the chapter's central narrative: PWK drops PWG's kosha apparatus before MW; MW's <ls>L.</ls> is a re-introduction in compressed form.
• Effect-size threshold (D19): all numerical claims on this page exceed |Δ| ≥ 5 pt threshold.
• D21 three-stage hedge lineage: PWK sits between PWG (named-kosha tradition) and the lineage proper — uses typographic * for kosha-only entries but does not scale or tag the convention.

8. Data dictionary + reproducibility manifest

• DATA_DICTIONARY.md — full tag inventory
• Source file: csl-orig/v02/pw/pw.txt
• Block-detector script: figures/scripts/export_data.py
• Cross-dict aggregate JSON: src/data/cross-dict.json (includes PWK/PW block matrix data)
• License: CC-BY-SA-4.0

See also (tools)

• Cross-dictionary comparison
• Lineage Sankey — PWG → PWK → MW kosha-collapse arc
• PWK#115 — docs-pass tracking issue

Source: CDSL pwk.txt 2026-05-23 · MWS docs-pass commit reflects audit pipeline as of 2026-05-27 · CC-BY-SA-4.0