Tables / History, Culture & Language
Lexical distance — how similar are the main languages?
How similar are the main languages of A and B? How many of their everyday basic words look and sound alike?
- How to read a cell
- We take the most widely spoken language of each country and compare 40 everyday words — I, you, two, water, fire, eye, hand, name and so on — as they sound in each language. LDND counts how many sounds must change to turn one word into the other, and scales the result so that two unrelated languages come out at about 100: 0 means the same words, Czech and Slovak score about 30, German and Dutch about 55, English and German about 75. It measures shared basic vocabulary, not how easily people understand each other, and the main language of a country is a simplification — many countries speak several. “No data” means the source has no figure for that pair — the note in the cell says why; “not applicable” means the question does not arise for that pair.
- Reliability
- medium — standard ASJP method (LDND, Wichmann et al. 2010) on 40 basic words per language, re-implemented from Wichmann's reference program; lists with fewer than 28 of the 40 words are not used. ASJP's word lists are short and come from different transcribers: two lists of the SAME language typically score 20–45 against each other, so the value is the median over all pairings of ASJP's lists for the two languages and differences under about 10 points mean little. LDND compares sounds of basic words, not grammar, spelling or mutual understanding (Spanish–Portuguese 68 scores further apart than Spanish–Italian 57, although Spanish and Portuguese speakers understand each other more easily). Main language = the most widely spoken (native + second language at A2+) in LINGACON's speaker data (LSI v1.48); when the runner-up is less than 10 points behind, the official language (L03) wins, otherwise the one with more native speakers; for the 21 states LINGACON does not yet cover, the official language from L03, the one the constitution names first or as national language. Debatable choices: Belgium — French (Dutch has more native speakers); Switzerland — German (official; Swiss German is the everyday spoken form); India — Hindi; Paraguay — Spanish (Guaraní nearly as widespread, slightly more native speakers); Philippines — Filipino (Tagalog); Kazakhstan — Kazakh over Russian (near tie); Belarus — Russian; Luxembourg — French; Qatar and UAE — English (largest by speakers, not official; Arabic is); DR Congo and Cameroon — French over Lingala / Cameroonian Pidgin (near tie, official one chosen); Uganda — English; Nigeria — Nigerian Pidgin; Djibouti — Somali; Malta — Maltese and Andorra — Catalan (near ties); Hong Kong — Cantonese; Mauritius, Fiji, Sierra Leone — English (Mauritian Creole, Fijian, Krio are everyday languages without LINGACON data); several Caribbean states — their creole where LINGACON shows it as most widely spoken (Jamaica, Saint Lucia, Dominica, Saint Vincent). Arabic-speaking states are compared through their spoken dialects, not Modern Standard Arabic
- Answered
- 100 % of 37,056 directed cells have an answer; the rest say why not
- Second source
- 6 % of the cells (marked ≈) rest on a secondary source or an estimate because the main source has no figure — the note in each such cell names it; verify before relying on it
- Refresh
- when ASJP publishes a new database version (v20 2022, v21 2025) or LINGACON releases a new LSI that changes a country's main language
- Credit
- Lexical distances computed by CountryPair from the ASJP Database v21 (Wichmann et al. 2025), CC BY 4.0 — licence
- LINGACON Project — licence
- Source: Wikidata (CC0), via QLever; de-facto additions by CountryPair — licence
One country against all others
Loading…
Hover a line for its note, or switch the notes on.
The table
Rows are country A, columns country B; the two-letter codes are the usual country codes (AM = Armenia, DE = Germany…). Hover a cell to see the full value and its note. A dash means “no” or “nothing recorded”; a striped cell means either “no data” (the source has no figure for this pair — the note says why) or “not applicable” (the question makes no sense for this pair, for example a sea route for two landlocked countries). “All regions” on both sides shows the whole table.