The world’s languages, grouped by ancestry
This reference organises the world’s languages into their major language families — groups descended from a shared ancestral language. For each family it shows representative member languages, an approximate first-language speaker total, and the region where the family is concentrated. It is a quick way to see how the roughly 7,000 living languages cluster into a few dozen related groups.
How it works
Linguists assign languages to families using the comparative method: they look for regular, systematic sound correspondences and shared inherited grammar across languages, which reveal descent from a common parent rather than mere contact or borrowing. The result is a family tree — Proto-Indo-European, for example, split into branches that became the Germanic, Romance, Slavic, Indo-Iranian and other languages. A handful of languages, like Basque, resist any grouping and are called isolates. Speaker totals here are rounded L1 estimates, so they show relative scale rather than precise headcounts.
The major families at a glance
| Family | Est. L1 speakers | Geographic core | Representative languages |
|---|---|---|---|
| Indo-European | ~3.2 billion | Europe, South & Central Asia | English, Spanish, Hindi, Russian, Portuguese |
| Sino-Tibetan | ~1.4 billion | East & Southeast Asia | Mandarin, Wu, Tibetan, Burmese |
| Niger-Congo | ~700 million | Sub-Saharan Africa | Swahili, Yoruba, Igbo, Zulu |
| Afro-Asiatic | ~500 million | North Africa, Middle East | Arabic, Amharic, Hausa, Hebrew |
| Austronesian | ~350 million | Southeast Asia, Pacific | Malay, Javanese, Tagalog, Māori |
| Dravidian | ~225 million | South India, Sri Lanka | Tamil, Telugu, Kannada, Malayalam |
| Turkic | ~200 million | Central Asia, Turkey | Turkish, Uzbek, Kazakh, Azerbaijani |
| Tai-Kadai | ~100 million | Southeast Asia | Thai, Lao, Zhuang |
| Japonic | ~125 million | Japan, Ryukyu Islands | Japanese, Ryukyuan |
| Koreanic | ~80 million | Korean Peninsula | Korean |
Figures are rough L1 (first-language) estimates and vary between sources such as Ethnologue and Glottolog.
Isolates and unclassified languages
A language isolate has no demonstrable relatives — it stands alone as a family of one. Well-known examples:
- Basque (Euskara): spoken in the western Pyrenees across Spain and France; unrelated to any surrounding Indo-European languages and predates the Roman period.
- Korean: sometimes grouped tentatively with Japonic as a macro-family, but the connection is debated; most classifications treat it as an isolate or as its own family.
- Sumerian: an extinct isolate from ancient Mesopotamia.
Isolates remind us that language families are histories, not geographic clusters — they show which communities shared a common ancestor, not merely which groups live near each other today.
Tips
- Search a specific language such as
TamilorHausato find which family it belongs to. - The same family can span continents — Austronesian reaches from Madagascar to Hawaii and Easter Island, while Indo-European covers from Ireland to Bangladesh.
- L1 (first-language) speaker counts are shown; widely learned languages like English, French, or Arabic have far larger total speaker populations when second-language users are counted.
- Language boundaries within families reflect historical splits, not political borders, which is why Hindi and Urdu are mutually intelligible but politically treated as separate languages.