வளம்: மொழி — OPEN ACCESS REFERENCES FOR LINGUISTICS SUB-DOMAINS

One primary reference + one best alternative per topic

Verified: July 2026

═══════════════════════════════

GRAPH (Writing Systems, Scripts, Character Encoding)

═══════════════════════════════

PRIMARY:

Name: Unicode — The World Standard for Text and Emoji

URL: https://home.unicode.org/

Type: Standards body / Character encoding reference

Access: Fully open, free

Maintainer: Unicode Consortium

Content: Unicode Standard (latest version), emoji specs, character charts, normalization forms, bidirectional algorithm, line breaking, script properties

Key Feature: Defines universal text representation for all scripts including Tamil (U+0B80–U+0BFF), Devanagari, IPA extensions

Citation: "The Unicode Standard" (cite specific version)

License: Unicode Data Files license (open, royalty-free)

ALTERNATIVE BEST:

Name: ScriptSource (by Unicode Consortium)

URL: https://scriptsource.org/

Type: Collaborative reference for scripts, keyboards, fonts

Access: Fully open

Content: Script usage data, keyboard mappings, font resources, locale data for world scripts

Advantage: Practical implementation details beyond the standard itself

═══════════════════════════════

PHONE (Phonetics, Phonology, Phoneme Inventories)

═══════════════════════════════

PRIMARY:

Name: PHOIBLE 2.0 (Phonetics Information Base and Lexicon)

URL: https://phoible.org/

Type: Database of phonological inventories

Access: Fully open, CC BY-SA 3.0

Maintainer: Steven Moran & Daniel McCloy (CLLD project)

Content: 3,020 phonological inventories across 2,186 languages; 3,183 segment types; distinctive feature data based on Hayes' Introductory Phonology + Moisik/Esling extensions

Data Format: CLDF (Cross-Linguistic Data Formats), downloadable CSV, GitHub repository (github.com/phoible/dev)

Coverage: Global, linked to Glottolog and ISO 639-3

Citation: Moran, Steven & McCloy, Daniel (eds.) 2019. PHOIBLE 2.0.

ALTERNATIVE BEST:

Name: International Phonetic Association (IPA)

URL: https://www.internationalphoneticassociation.org/

Type: Professional association / IPA chart reference

Access: IPA chart freely available; full handbook paid

Content: Official IPA chart, sound recordings, IPA Unicode guide

Advantage: Authoritative standard for phonetic transcription itself

Limitation: Not a searchable database like PHOIBLE

═══════════════════════════════

MORPH (Morphology, Word Formation, Etymology)

═══════════════════════════════

PRIMARY:

Name: Wiktionary

URL: https://www.wiktionary.org/

Type: Collaborative multilingual dictionary

Access: Fully open, CC BY-SA 4.0

Content: Definitions, etymologies, morphological breakdowns, pronunciations (IPA), translations across 1000+ languages

Tamil Pages: Extensive Tamil entries with morphological information

Key Feature: Largest free multilingual lexical resource; covers derivational and inflectional morphology cross-linguistically

Limitation: Variable quality (crowdsourced); verify against primary sources for research use

ALTERNATIVE BEST:

Name: Dictionaria (CLLD)

URL: https://dictionaria.clld.org/

Type: Open-access journal publishing dictionaries worldwide

Access: Fully open

Maintainer: Martin Haspelmath, Barbara Stiebels (chief editors); CLLD project / Max Planck Institute

Content: Peer-reviewed dictionaries from diverse languages; standardized format, searchable, downloadable

Advantage: Academic quality (peer-reviewed), unlike crowdsourced Wiktionary; follows CLDF standards

═══════════════════════════════

SEMANTICS (Lexical Semantics, Meaning Relations)

═══════════════════════════════

PRIMARY:

Name: WordNet (Princeton University)

URL: https://wordnet.princeton.edu/

Type: Lexical database of English semantic relations

Access: Free for research and commercial use (citation required)

Maintainer: Princeton University (George Miller et al.)

Content: Nouns, verbs, adjectives, adverbs grouped into synsets (cognitive synonyms); relations include hypernymy, hyponymy, meronymy, holonymy, antonymy, entailment

Size: ~155,000 words, ~117,000 synsets

API: NLTK (Python), spaCy, TFDS, CRAN wordnet package

Citation: Miller, G.A. (1995). WordNet: A Lexical Database for English. Communications of the ACM, 38(11).

ALTERNATIVE BEST:

Name: Open English WordNet (Global WordNet Association)

URL: https://en-word.net/

Type: Open-source fork of Princeton WordNet

Access: Fully open, GitHub-hosted (github.com/globalwordnet/english-wordnet)

Content: Same synset structure as Princeton WordNet plus ongoing community improvements; GWN-LMF format; includes Open English Namenet (named entities) since 2025

Advantage: Actively maintained, community-driven improvements, part of Global WordNet Grid (wordnets in 100+ languages)

═══════════════════════════════

SYNTAX (Grammar, Sentence Structure, Dependency Parsing)

═══════════════════════════════

PRIMARY:

Name: Universal Dependencies (UD)

URL: https://universaldependencies.org/

Type: Framework for cross-linguistic grammar annotation

Access: Fully open, CC BY-SA 4.0 (treebanks vary slightly)

Maintainer: Community effort (600+ contributors)

Content: 200+ treebanks in 150+ languages; POS tags, morphological features, syntactic dependency relations; annotation guidelines, relation inventory

Data Format: CoNLL-U format, downloadable from GitHub

Key Feature: Consistent annotation across genetically diverse languages; includes Tamil (UD_Tamil-TTB), Hindi, Japanese, Arabic

Query: Online search at https://universaldependencies.org/tools.html

Citation: Nivre et al. (2020). Universal Dependencies v2.

ALTERNATIVE BEST:

Name: WALS Online (World Atlas of Language Structures)

URL: https://wals.info/

Type: Typological database of grammatical features

Access: Fully open (CLLD)

Maintainer: Matthew Dryer & Martin Haspelmath

Content: 130+ structural features (word order, case marking, negation, tense/aspect, etc.) for 2,679 languages

Advantage: Broader typological coverage than UD (more languages, more features); not parsed trees but categorical features

Limitation: Coded feature values, not actual treebanks

═══════════════════════════════

PRAGMATICS (Frame Semantics, Discourse, Usage)

═══════════════════════════════

PRIMARY:

Name: Berkeley FrameNet

URL: https://framenet.icsi.berkeley.edu/

Type: Lexical resource based on Frame Semantics

Access: Fully open (registration required for full data download)

Maintainer: International Computer Science Institute (ICSI), UC Berkeley

Founder: Charles J. Fillmore

Content: ~1,200 semantic frames; ~13,000+ lexical units; 200,000+ manually annotated example sentences with frame elements (semantic roles) and syntactic realization

Coverage: English (plus Spanish, Japanese, German FrameNets linked)

Data Format: Full-text XML, FrameSQL, Lucene index

Citation: Fillmore, C.J., Johnson, C.R., Petruck, M.R.L. (2003). Background to Framenet. International Journal of Lexicography, 16(3).

ALTERNATIVE BEST:

Name: Global FrameNet / FrameNet Multi-lingual

URL: https://framenet.icsi.berkeley.edu/fndrupal/multilingualFramnets

Type: Multilingual FrameNet network

Access: Open (varies by language-specific site)

Content: Frame-semantic annotation in Spanish, Japanese, German, Swedish, Danish, Chinese, French, Brazilian Portuguese

Advantage: Cross-linguistic frame comparison; useful for contrastive pragmatics and multilingual NLP

═══════════════════════════════

CORPUS (Text Corpora, Language Archives)

═══════════════════════════════

PRIMARY:

Name: Internet Archive

URL: https://archive.org/

Type: Digital library / text + media archive

Access: Fully open (most content public domain or CC-licensed)

Content: Millions of digitized texts, historical newspapers, audio recordings, linguistic field notes, grammars, language documentation materials

Tamil: Significant Tamil text collection (books, periodicals)

Key Feature: Universal scope; historical depth (19th century onward)

Limitation: Not linguistically annotated (raw text/images)

ALTERNATIVE BEST:

Name: Language Description Heritage (LDH) Library

URL: https://ldh.clld.org/

Type: Curated digital library of language descriptions

Access: Fully open (CLLD project)

Maintainer: Robert Forkel (Max Planck Institute); Zenodo community

Content: Scanned reference grammars, dictionaries, linguistic descriptions from rare/out-of-print sources; linked to Glottolog language identifiers

Advantage: Linguistically curated (vs. general-purpose Internet Archive); every item linked to specific Glottolog language/family; Zenodo DOIs for citation

═══════════════════════════════

COMPUTATIONAL (Cross-Linguistic Data, Typology)

═══════════════════════════════

PRIMARY:

Name: CLLD — Cross-Linguistic Linked Data

URL: https://clld.org/

Type: Platform hosting interconnected linguistic databases

Access: Fully open (all datasets free)

Maintainer: Robert Forkel et al., Max Planck Institute for Evolutionary Anthropology, Leipzig

Hosted Datasets:

Glottolog — https://glottolog.org (language catalog, all families)

WALS Online — https://wals.info (typological features, 2,679 langs)

APiCS — https://apics-online.info (pidgin/creole structures)

WOLD — https://wold.clld.org (loanword database)

PHOIBLE — https://phoible.org (phonological inventories)

Concepticon — https://concepticon.clld.org (concept list linking)

Dictionaria — https://dictionaria.clld.org (open-access dictionaries)

LDH — https://ldh.clld.org (language descriptions)

Tsammalex — https://tsammalex.clld.org (plants & animals lexicon)

Data Format: CLDF (Cross-Linguistic Data Format); CSV downloads; GitHub repos with versioned releases

Citation: Each dataset has its own cite recommendation; stable versioned releases on Zenodo with DOIs

ALTERNATIVE BEST:

Name: Glottolog (standalone)

URL: https://glottolog.org/

Type: Comprehensive catalog of world languages

Access: Fully open, CC BY 4.0

Maintainer: Harald Hammarström, Martin Haspelmath, Robert Forkel

Content: Classification of all known languages (living/extinct), dialects, geographic coordinates, bibliographic references for each language, ISO 639-3 mapping

Advantage: De facto standard language identifier system in linguistics; Glottocodes used as persistent IDs across CLLD and beyond

GitHub: https://github.com/glottolog/glottolog

═══════════════════════════════

SEMIOTICS (Sign Theory, Symbol Systems, Meaning-Making)

═══════════════════════════════

PRIMARY:

Name: Semiotics Encyclopedia Online (Semioticon)

URL: https://semioticon.com/ (or https://semioticon.com/seo/)

Type: Peer-oriented reference encyclopedia

Access: Fully open

Content: Key concepts, theories, and figures in semiotics; covers Peircean, Saussurean, Eco, Barthes, Greimas, biosemiotics, cybersemiotics, visual semiotics

Maintainer: Semiotic Society of America / community

Format: Searchable articles, alphabetical browse

Advantage: Dedicated exclusively to semiotics (not a general encyclopedia with occasional semiotics entries)

ALTERNATIVE BEST:

Name: Daniel Chandler's Semiotics Encyclopedia / Semiotics for Beginners

URL: https://visual-memory.co.uk/daniel/Documents/S4B/ (Beginners); library portals host the 300-entry encyclopedia version

Type: Glossary + introductory encyclopedia

Access: Fully open (university-hosted)

Author: Daniel Chandler (University of Wales, Aberystwyth)

Content: 300 scholarly entries on signs, symbols, codes, paradigmatic/syntagmatic relations, denotation/connotation, key thinkers (Saussure, Peirce, Barthes, Eco, Jakobson)

Advantage: More pedagogically structured than Semioticon; excellent for students entering the field

═══════════════════════════════

NLP (Natural Language Processing, Computational Linguistics Tools)

═══════════════════════════════

PRIMARY:

Name: NLTK — Natural Language Toolkit

URL: https://www.nltk.org/

Type: Open-source Python library + educational platform

Access: Fully open (Apache 2.0 license)

Maintainer: NLTK Team (Steven Bird, Ewan Klein, Edward Loper)

Content: 50+ corpora and lexical resources (incl. WordNet, Penn Treebank sample, Brown Corpus, Gutenberg Corpus); tokenization, stemming, lemmatization, POS tagging, chunking, parsing, named-entity recognition, semantic reasoning, text classification

Book: "Natural Language Processing with Python" (free online at https://www.nltk.org/book/) — full O'Reilly book

GitHub: https://github.com/nltk/nltk

Python: pip install nltk (requires Python 3.10+)

Citation: Bird, S., Klein, E., Loper, E. (2009). Natural Language Processing with Python. O'Reilly Media.

ALTERNATIVE BEST:

Name: spaCy

URL: https://spacy.io/

Type: Industrial-strength open-source NLP library

Access: Fully open (MIT license)

Maintainer: Explosion AI (Matthew Honnibal et al.)

Content: Tokenization, POS tagging, dependency parsing, NER, lemmatization, sentence segmentation, text classification, word vectors, transformer integration

Languages: 60+ trained pipelines incl. Tamil (ta), Hindi (hi), Chinese, Japanese, Arabic, German, French

Advantage: Production-grade speed; neural models; seamless transformer integration; better for deployment than NLTK

GitHub: https://github.com/explosion/spaCy

Limitation: Less pedagogical than NLTK (fewer built-in corpora/lessons)

═══════════════════════════════

SUMMARY TABLE

═══════════════════════════════

Graph — Primary: Unicode — Alt: ScriptSource — Access: Open / Open

Phone — Primary: PHOIBLE 2.0 — Alt: IPA — Access: Open / Partial

Morph — Primary: Wiktionary — Alt: Dictionaria (CLLD) — Access: Open / Open

Semantics — Primary: Princeton WordNet — Alt: Open English WordNet — Access: Open / Open

Syntax — Primary: Universal Deps — Alt: WALS Online — Access: Open / Open

Pragmatics — Primary: Berkeley FrameNet — Alt: Global FrameNet — Access: Open / Open

Corpus — Primary: Internet Archive — Alt: LDH Library — Access: Open / Open

Computational — Primary: CLLD — Alt: Glottolog — Access: Open / Open

Semiotics — Primary: Semioticon — Alt: Chandler Semiotics — Access: Open / Open

NLP — Primary: NLTK — Alt: spaCy — Access: Open / Open

ALL PRIMARY REFERENCES ARE FREE AND OPEN ACCESS.