Glossary of Place-Name and Script Terms: Endonym, Exonym, Romanisation
An endonym is the name a place has in its own language, an exonym the name others give it; this glossary explains these and the other terms that appear in the country, language and alphabet tables, from romanisation and transliteration to ISO 3166, ISO 639 and ISO 15924.
Names, scripts and codes
Every table on this site rests on three questions. What is a place or a language called by the people who live there or speak it, and what do others call it? In which script is that name written, and how is it rendered in Latin letters? Which international code identifies it? The terms below are the vocabulary for those three questions, grouped in the same order.
Two words carry the first question. Deutschland is the endonym of the country, the name in its own language; Germany, Allemagne and Alemania are exonyms, the names other languages use. The United Nations Group of Experts on Geographical Names defines both terms for places, and the same pair is used here for languages too.1
Terms for names of places and languages
5 terms.
| Term | Meaning |
|---|---|
| Endonym | The name of a place in an official or well-established language of the area where it lies: Deutschland, Wien, Hrvatska.1 |
| Exonym | A name used in one language for a place outside the area where that language is spoken, when it differs from the local name: English Germany and Vienna, Spanish Alemania, French Allemagne.1 |
| Autonym, xenonym | Synonyms of endonym and exonym, used less often. They also serve for what a people or a language calls itself (autonym) and what others call it (xenonym).1 |
| Native name | On this site, the endonym of a country or the name a language uses for itself, set in its own script: Deutschland for Germany, brezhoneg for Breton, euskara for Basque. |
| Short and official form | Country pages give two forms where they differ: the short name in everyday use (France) and the full official name of the state (République française). |
Terms for scripts and writing systems
8 terms. A script is not the same as a language: Serbian is written in Cyrillic and in Latin, and the Latin alphabet serves hundreds of languages.
| Term | Meaning |
|---|---|
| Script | A set of graphic characters used to write one or more languages, such as Latin, Cyrillic, Arabic or Devanagari. One language can use several scripts, and one script serves many languages.2 |
| Alphabet | A script with letters for both consonants and vowels, such as Latin, Greek, Cyrillic, Armenian or Georgian.3 |
| Abjad | A script whose basic letters stand for consonants only; vowels are left to the reader or marked with optional signs. Arabic and Hebrew are written this way.3 |
| Abugida | A script in which each consonant letter carries an inherent vowel, changed by attached vowel signs and removed by a mark called the virama. Devanagari, Bengali, Tamil and the other Brahmic scripts are abugidas.4 |
| Syllabary | A script whose signs stand for syllables or morae. Japanese hiragana and katakana are the best-known examples.5 |
| Right to left | Arabic, Hebrew, Syriac and Thaana run from right to left. Every name in these scripts is marked on the page so that browsers set it in the right direction. |
| Long s | The letter ſ, used inside words in older German and English printing. German entries show it next to the modern spelling, as in Deutſch beside Deutsch. |
| Zawgyi | A widely used Burmese font that put Myanmar characters on code points in a non-standard way, so the same word could be stored in several ways. Myanmar officially moved to Unicode on 1 October 2019. Burmese on this site is in standard Unicode.6 |
Romanisation, transliteration and transcription
4 terms.
| Term | Meaning |
|---|---|
| Romanisation | Writing a name from another script in Latin letters. Every non-Latin name on this site is followed by a romanisation in brackets or in its own column.7 |
| Transliteration | A romanisation that maps the characters of the source script one to one, so that the original spelling can be restored from it. Diacritics such as ā, š or ṭ mark letters that Latin lacks.7 |
| Transcription | A romanisation that follows pronunciation instead of spelling. It is easier to read aloud but cannot always be turned back into the original script.7 |
| IPA | The International Phonetic Alphabet, a set of symbols with one symbol per speech sound. The IPA chart lists them. |
Why the codes matter
Names change, codes are meant to stay put. The ISO 3166 code DE stands for Germany whatever language the page is in, and ISO 639 does the same for languages. Codes do change when the world does, though: when a country splits, merges or reorganises its regions, the ISO 3166 maintenance agency adds, withdraws or replaces codes, and withdrawn country codes move to ISO 3166-3.8
Language tags combine the codes. A tag starts with the ISO 639 code of the language and can add the ISO 15924 code of the script and a region, so sr-Cyrl marks Serbian in Cyrillic and zh-Hant Chinese in traditional characters.10 Every native-script name on the site carries such a tag, which lets browsers choose a suitable font and screen readers the right pronunciation.
Country, language and script codes
10 terms.
| Term | Meaning |
|---|---|
| ISO 3166-1 | Country codes in three forms: alpha-2 (DE), alpha-3 (DEU) and a three-digit numeric code (276) that matches the UN Statistics Division list. Alpha-2 codes are the basis of most country-code internet domains.8 |
| ISO 3166-2 | Codes for the principal subdivisions of every country in ISO 3166-1, such as states, provinces or regions. Each code starts with the alpha-2 code of the country, as in DE-BY for Bavaria.8 |
| ISO 3166-3 | Codes for country names removed from ISO 3166-1 since its first edition in 1974, such as those of states that no longer exist.8 |
| ISO 639-1 | Two-letter language codes: de, fr, ar, eu.9 |
| ISO 639-2 | Three-letter language codes. For 22 languages there are two codes, a bibliographic (B) and a terminology (T) code: ger and deu for German, baq and eus for Basque.9 |
| ISO 639-3 | Three-letter codes intended to cover all individual languages, maintained by SIL International.9 |
| ISO 639-5 | Three-letter codes for language families and groups, such as gem for the Germanic languages.9 |
| ISO 15924 | Four-letter codes for scripts, each with a numeric code as well: Latn (215), Cyrl (220), Arab (160), Deva (315). The Unicode Consortium is the registration authority.2 |
| Language tag (BCP 47) | The code that tells software which language a piece of text is in, built from subtags in a fixed order: language, then script, region and variant, as in sr-Cyrl, zh-Hant or de-CH. It is defined in RFC 5646 and RFC 4647.10 |
| Code point | The number Unicode gives each character, written on this site in the form U+0915 (the Devanagari letter ka). Unicode has 1,114,112 code points, from 0 to 10FFFF in hexadecimal.11 |
Questions about the terms
What is the difference between an endonym and an exonym?
An endonym is the local name of a place in a language of that place (Wien); an exonym is the different name another language uses for it (Vienna).
Is transliteration the same as transcription?
No. Transliteration converts letter by letter and can be reversed; transcription follows the sound and usually cannot be turned back into the original spelling.
Why does German have two ISO 639-2 codes?
ISO 639-2 gives 22 languages a bibliographic and a terminology code. For German these are ger and deu; ISO 639-1 has the single code de.
- 1United Nations Group of Experts on Geographical Names, definitions of endonym and exonym, as cited in Wikipedia, “Endonym and exonym”, checked 28 October 2026. ↑
- 2ISO 15924, Codes for the representation of names of scripts; Wikipedia, “ISO 15924”, checked 28 October 2026. ↑
- 3Wikipedia, “Abjad” (consonant-only scripts; alphabets with letters for consonants and vowels), checked 28 October 2026. ↑
- 4Wikipedia, “Abugida”, checked 28 October 2026. ↑
- 5Wikipedia, “Syllabary”, checked 28 October 2026. ↑
- 6Wikipedia, “Zawgyi font” (non-standard use of the Myanmar block; U-Day, 1 October 2019), checked 28 October 2026. ↑
- 7Wikipedia, “Romanization” (transliteration as one-to-one mapping of characters; transcription of the spoken word; reversibility), checked 28 October 2026. ↑
- 8ISO 3166, Codes for the representation of names of countries and their subdivisions, Parts 1–3; Wikipedia, “ISO 3166”, checked 28 October 2026. ↑
- 9ISO 639 (consolidated as ISO 639:2023); ISO 639-2 Registration Authority, Library of Congress; Wikipedia, “ISO 639”, checked 28 October 2026. ↑
- 10IETF BCP 47: RFC 5646, Tags for Identifying Languages, and RFC 4647, Matching of Language Tags. ↑
- 11Wikipedia, “Code point” (1,114,112 code points, 0 to 10FFFF hex), checked 28 October 2026. ↑