Skip to main content

The Dictionary Landscape

The catalog lists every dictionary as a table row. This page shows the same corpus as a landscape: when each dictionary appeared, how much of its headword list exists nowhere else, how big it is, and which dictionaries are effectively members of one family.

Era × novelty × size​

Each bubble is one dictionary. Right = later publication; higher = more of its headwords are unique to it (found in no other CDSL dictionary); bigger = more distinct lemmas. Uncheck a language group to declutter; hover any bubble for details.

0%10%20%30%40%50%60%18251850187519001925195019752000publication yearheadwords unique to this dictionary (%)MW — Monier-Williams Sanskrit-English DictionaryPW — Böhtlingk Sanskrit-Wörterbuch in kürzerer FassungPWG — Böhtlingk and Roth Grosses Petersburger WörterbuchAP — Practical Sanskrit-English Dictionary, revised editionMW72 — Monier-Williams Sanskrit-English DictionaryVCP — VacaspatyamLRV — Vaidya Sanskrit-English DictionarySHS — Shabda-Sagara Sanskrit-English DictionaryYAT — Yates Sanskrit-English DictionaryWIL — Wilson Sanskrit-English DictionarySKD — Sabda-kalpadrumaCAE — Cappeller Sanskrit-English DictionaryAP90 — Apte Practical Sanskrit-English DictionaryACC — Aufrecht's Catalogus CatalogorumCCS — Cappeller Sanskrit WörterbuchSCH — Schmidt Nachträge zum Sanskrit-WörterbuchMWE — Monier-Williams English-Sanskrit DictionarySTC — Stchoupak Dictionnaire Sanscrit-FrançaisBOR — Borooah English-Sanskrit DictionaryMD — Macdonell Sanskrit-English DictionaryBUR — Burnouf Dictionnaire Sanscrit-FrançaisBHS — Edgerton Buddhist Hybrid Sanskrit DictionaryBEN — Benfey Sanskrit-English DictionaryPUI — The Purana IndexABCH — Abhidhānacintāmaṇi of HemacandrācāryaAE — Apte Student's English-Sanskrit DictionaryGRA — Grassmann Wörterbuch zum Rig VedaINM — Index to the Names in the MahabharataBOP — Bopp Glossarium SanscritumFRI — Friš Sanskrit Reader VocabularyIEG — Indian Epigraphical GlossaryGST — Goldstücker Sanskrit-English DictionaryARMH — Abhidhānaratnamālā of HalāyudhaPE — Puranic EncyclopediaLAN — Lanman's Sanskrit Reader VocabularyVEI — The Vedic Index of Names and SubjectsMCI — Mahabharata Cultural IndexNMMB — Nāmamālikā of BhojaKRM — KṛdantarūpamālāACPH — Abhidhānacintāmaṇipariśiṣṭa of HemacandrācāryaACSJ — Abhidhānacintāmaṇiśiloñcha of JinadevaPGN — Personal and Geographical Names in the Gupta InscriptionsSNP — Meulenbeld's Sanskrit Names of PlantsMW72LRVMWSHSAPPWGPWVCPPUIBHSACCIEGPGN

Bubble area = distinct lemma count (largest: MW, 194,084). Novelty is a floor — a shared headword is independent attestation in each dictionary, not necessarily a copy.

Data table (43 dictionaries)
CodeDictionaryGroupYearDistinct lemmasUnique %Entries
BHSEdgerton Buddhist Hybrid Sanskrit DictionarySpecialized195317,77757.6%17,839
IEGIndian Epigraphical GlossarySpecialized19667,56857.5%7,932
PGNPersonal and Geographical Names in the Gupta InscriptionsSpecialized197847054.9%485
ACCAufrecht's Catalogus CatalogorumSpecialized196232,58543.3%49,833
PUIThe Purana IndexSpecialized195111,73038.6%17,512
SKDSabda-kalpadrumaSanskrit-Sanskrit188640,81737.1%42,531
KRMKṛdantarūpamālāSpecialized19651,69836.6%2,061
APPractical Sanskrit-English Dictionary, revised editionSanskrit-English195788,86934.7%90,843
MWEMonier-Williams English-Sanskrit DictionaryEnglish-Sanskrit185128,23831.2%32,378
SCHSchmidt Nachträge zum Sanskrit-WörterbuchSanskrit-German192828,45529.6%29,125
NMMBNāmamālikā of BhojaSanskrit-Sanskrit19552,26523%506
BURBurnouf Dictionnaire Sanscrit-FrançaisSanskrit-French186619,14622.7%19,776
INMIndex to the Names in the MahabharataSpecialized19049,45422.7%12,647
PEPuranic EncyclopediaSpecialized19756,62319.7%8,799
MCIMahabharata Cultural IndexSpecialized19932,32518.2%2,643
LRVVaidya Sanskrit-English DictionarySanskrit-English188948,25917.6%53,441
BORBorooah English-Sanskrit DictionaryEnglish-Sanskrit187722,01315.9%24,609
AP90Apte Practical Sanskrit-English DictionarySanskrit-English189034,27715.8%34,882
ACPHAbhidhānacintāmaṇipariśiṣṭa of HemacandrācāryaSanskrit-Sanskrit18961,45314.1%163
MWMonier-Williams Sanskrit-English DictionarySanskrit-English1899194,08412.9%286,525
FRIFriš Sanskrit Reader VocabularySpecialized19567,77512%8,155
VEIThe Vedic Index of Names and SubjectsSpecialized19123,70411.7%3,834
VCPVacaspatyamSanskrit-Sanskrit187348,6369.3%50,135
GSTGoldstücker Sanskrit-English DictionarySanskrit-English18566,7617.9%6,780
AEApte Student's English-Sanskrit DictionaryEnglish-Sanskrit192011,3437.8%11,359
ACSJAbhidhānacintāmaṇiśiloñcha of JinadevaSanskrit-Sanskrit18966347.6%240
MW72Monier-Williams Sanskrit-English DictionarySanskrit-English187251,1594.7%55,390
STCStchoupak Dictionnaire Sanscrit-FrançaisSanskrit-French193223,9864.4%24,574
PWBöhtlingk Sanskrit-Wörterbuch in kürzerer FassungSanskrit-German1879151,3494.4%170,556
GRAGrassmann Wörterbuch zum Rig VedaSanskrit-German187311,1084.1%12,785
ABCHAbhidhānacintāmaṇi of HemacandrācāryaSanskrit-Sanskrit189611,5843.3%1,965
SNPMeulenbeld's Sanskrit Names of PlantsSpecialized19744483.3%453
LANLanman's Sanskrit Reader VocabularySanskrit-English18844,7263%4,944
ARMHAbhidhānaratnamālā of HalāyudhaSanskrit-Sanskrit18616,6733%7,907
YATYates Sanskrit-English DictionarySanskrit-English184644,7202.8%45,206
SHSShabda-Sagara Sanskrit-English DictionarySanskrit-English190046,7331.9%47,326
PWGBöhtlingk and Roth Grosses Petersburger WörterbuchSanskrit-German1855106,0821.9%123,366
MDMacdonell Sanskrit-English DictionarySanskrit-English189320,1031.7%20,749
CAECappeller Sanskrit-English DictionarySanskrit-English189138,4841.6%40,069
BOPBopp Glossarium SanscritumSanskrit-Latin18478,5051.4%8,961
BENBenfey Sanskrit-English DictionarySanskrit-English186617,0350.7%17,310
WILWilson Sanskrit-English DictionarySanskrit-English183243,9380.6%44,577
CCSCappeller Sanskrit WörterbuchSanskrit-German188728,7510.5%30,010

How to read it. The 19th-century Sanskrit–English/German core (MW, PW, PWG, AP90) sits low: these dictionaries share most of their headword lists — a genealogy, not plagiarism, since later lexicographers built on earlier ones. The high outliers earn their place in the corpus by coverage no one else has: BHS (Buddhist Hybrid Sanskrit, 57.6% unique), the specialized indexes (ACC proper names, 43.3%), and the Sanskrit–Sanskrit kośas (SKD 37.1%). A dictionary low on this chart is best reached through MW; a dictionary high on it answers questions MW cannot.

Trust block
  • Evidence: src/data/atlas-extract.json (vendored per-dictionary rows from the csl-atlas OBS-R headword-collapse artifacts, headword_multiplicity.csv / headword_collapse.json, 2026-07 regeneration: 1,496,157 entries → 410,259 distinct lemmas corpus-wide) joined with the generated catalog for year and language group. n = 43 dictionaries (PD has no digital source → no atlas row; PWKVN has no catalog row).
  • Limitations: "unique %" is a floor for novelty — a shared headword is independent attestation in each dictionary, not proof of copying; entry counts include homonym-split records, so per-dictionary splitting policy leaks into the lemma counts.
  • Validation: totals cross-checked against the csl-atlas headword_collapse.json summary; the join is render-time only (nothing recomputed here).
  • Owner repo: csl-atlas (data) / this repo (rendering).
  • Next use: pick a high-novelty dictionary you have never opened and read its featured guide.

The headword family tree​

The same headword lists, clustered: dictionaries that join near the right share most of their headwords; late joiners on the left carry vocabularies nothing else has. This is the csl-atlas L0 lemma-overlap cladogram, rendered read-only.

0.40.30.20.10join height (Jaccard distance between headword sets — smaller = more shared)PGNSNPACSJKRMACPHACCGSTPDIEGBHSMCIPUIINMPEPWKVNSCHGRAVEISKDAPAP90MDCAECCSMW72STCMWPWPWGLRVVCPYATSHSWILFRILANABCHARMHBURBENBOP

Chip color = catalog language group (see the landscape chart above). Hover a code to trace its path to the root. Tight pairs (SHS–WIL, YAT, CAE–CCS, PW–PWG) join at low distance — heavy headword-list inheritance; the singleton indexes and koshas on the outer rim share few headwords with anything.

The clusters recover the known genealogy from the data alone: PW–PWG (Böhtlingk's two Petersburg lexica) and MW hanging just above them; the SHS–WIL–YAT chain (Wilson's list recycled by his successors); CAE–CCS (the two Cappellers); AP–AP90 (the two Aptes); PWKVN–SCH (the two supplements to PW). The specialized indexes (PGN, SNP, KRM, ACC) join last — their headword lists are mostly proper names no general dictionary carries.

Trust block
  • Evidence: src/data/cladogram.json — the vendored csl-atlas sanhw1_cladogram.newick (UPGMA over pairwise Jaccard distances between per-dictionary headword sets, built by lex_l2_cladogram.py). n = 41 dictionaries (the kośa-format and gated dictionaries are outside the sanhw1 snapshot).
  • Limitations: headword overlap only — it measures shared word lists, not shared definitions, and UPGMA assumes clock-like divergence the corpus does not literally have. It is one signal of the csl-atlas genealogy work, not the canonical family tree.
  • Validation: rendering parses the committed Newick verbatim; known pairs (PW–PWG, AP–AP90, CAE–CCS) join at the lowest heights, as expected.
  • Owner repo: csl-atlas (data) / this repo (rendering).
  • Next use: compare with the human-curated origins timeline — the cladogram is computed from headwords alone, the timeline from documented history.

See also​