An exonym or xenonym is an external name for a geographical place, a group of people, an individual person, or a language or dialect. It is a common name used only outside the place, group, or linguistic community in question. An endonym or autonym is an internal name for a geographical place, a group of people, or a language or dialect. It is a common name used only inside the place, group, or linguistic community in question; it is their name for themselves, self-designated, their homeland, or their language.
Marcel Aurousseau, an Australian geographer, first used the term exonym in his work The Rendering of Geographical Names (1957). The term endonym was devised subsequently as an antonym for the term exonym.
Exonyms exist not only for historical-geographical reasons but also in consideration of difficulties when pronouncing foreign words.
All four of the terms (exonym, endonym, autonym and xenonym) are from the Greek root word ónoma (), 'name', from Proto-Indo-European *h?nómn?. The prefixes are from the Greek éndon (), 'within'; autós (), 'self'; éx? (), 'out'; and xénos (), 'foreign'.
As pertains to geographical features, the United Nations Group of Experts on Geographical Names defines:
For example, India, China, Egypt, and Germany are the English-language exonyms corresponding to the endonyms ? (Bh?rat), (Zh?ngguó), ? (Masr), and Deutschland, respectively. Chinese and German are exonyms in English for the languages that are endonymously known as "" ("Zh?ngwén") and "Deutsch", respectively.
Exonyms may derive from different roots, as in the case of Germany for Deutschland, or they may be cognate words which have diverged in pronunciation or orthography, or they may be fully or partially translated (a calque) from the native language. For example, London (originally Latin Londinium) is known by the cognate exonyms Londres in Catalan, Filipino, French, Galician, Portuguese, and Spanish; Londino (?) in Greek; Londen in Dutch; Londra in Italian, Maltese, Romanian, Sardinian and Turkish; Londër in Albanian; Londýn in Czech and Slovak; Londyn in Polish; Lundúnir in Icelandic; Lontoo in Finnish; Landan in Persian. An example of a translated exonym is the French name Pays-Bas for the Netherlands, Nederland in Dutch, all of which mean "Low Countries".
Exonyms can also be divided into native and borrowed, i.e. from a third language. For example, Slovene uses the native exonyms Dunaj (Vienna) and Benetke (Venice), and the borrowed exonyms Kijev (Kiev) and Vilna (Vilnius), from Russian. A substantial proportion of English exonyms for places in continental Europe are borrowed (or adapted) from French; for example: Navarre (Spanish: Navarra/Nafarroa), Belgrade (Serbian: Beograd), Cologne (German: Köln), Munich (German: München), Prague (Czech: Praha), Rome (Italian: Roma), Naples (Italian: Napoli), and Florence (Italian: Firenze).
According to James A. Matisoff, who introduced the term "autonym" into linguistics, "Human nature being what it is, exonyms are liable to be pejorative rather than complimentary, especially where there is a real or fancied difference in cultural level between the ingroup and the outgroup." For example, Matisoff notes Khang "an opprobrious term indicating mixed race or parentage" is the Palaung name for Jingpo people and the Jingpo name for Chin people; both the Jingpo and Burmese use the Chinese word yeren (literally "wild men") "savage; rustic people" as the name for Lisu people.
Exonyms develop for places of significance for speakers of the language of the exonym. Consequently, many European capitals have English exonyms, e.g. Athens (/Athína), Belgrade (?/Beograd), Bucharest (Romanian: Bucure?ti), Brussels (Bruxelles, Brussel), Copenhagen (Danish: København), Lisbon (Portuguese: Lisboa), Moscow (Russian: /Moskva), Prague (Praha), Rome (Roma), Vienna (Austrian German: Wien), and Warsaw (Polish: Warszawa), while for instance historically less prominent capitals Ljubljana, Reykjavik, and Zagreb do not (but do have exonyms in languages spoken nearby e.g. German: Laibach and Agram, though "Agram" is old fashioned and not used any more). Madrid, Berlin, Oslo, and Amsterdam, with identical names in most major European languages, are exceptions. Some European capitals might be considered partial exceptions in that whilst the spelling is the same across languages, the pronunciation can differ; thus Paris in English sees the 's' vocalised, whilst in Swedish Stockholm is pronounced with a more emphasised glottal stop which is missing in English. For places considered to be of lesser significance, attempts to reproduce local names have been made in English since the time of the Crusades. Livorno, to take an instance, was Leghorn because it was an Italian port essential to English merchants and, by the 18th century, to the British Navy; not far away, Rapallo, a minor port on the same sea, never received an exonym.
In earlier times, the name of the first tribe or village encountered became the exonym for the whole people beyond. Thus the Romans used the tribal names Graecus (Greek) and Germanus, the Russians used the village name of Chechen, medieval Europeans took the tribal name Tatar as emblematic for the whole Mongolic confederation (and then confused it with Tartarus, a word for Hell, to produce Tartar), and the Magyar invaders were equated with the 500-years-earlier Hunnish invaders in the same territory, and were called Hungarians.
The Germanic invaders of the Roman Empire applied the word "Walha" to foreigners they encountered and this evolved in West Germanic languages as a generic name for all non-Germanic speakers; thence, the names Wallachia, Vlachs, Wallonia, Walloons, Cornwall, Wales, Wallasey, Welche in Alsace-Lorraine, and even the Polish name for Italy, W?ochy.
During the late 20th century the use of exonyms often became controversial. Groups often prefer that outsiders avoid exonyms where they have come to be used in a pejorative way: for example, Romani people often prefer that term to exonyms such as Gypsy (from Egypt), and the French term bohémien, bohème (from Bohemia). People may also avoid exonyms for reasons of historical sensitivity, as in the case of German names for Polish and Czech places that at one time had been ethnically or politically German (e.g. Danzig/Gda?sk and Karlsbad/Karlovy Vary), and Russian names for Russian locations which were subsequently renamed (e.g. Kiev/Kyiv).
In recent years, geographers have sought to reduce the use of exonyms to avoid this kind of problem. For example, it is now common for Spanish speakers to refer to the Turkish capital as Ankara rather than use the Spanish exonym Angora. According to the United Nations Statistics Division, "Time has, however, shown that initial ambitious attempts to rapidly decrease the number of exonyms were over-optimistic and not possible to realise in the intended way. The reason would appear to be that many exonyms have become common words in a language and can be seen as part of the language's cultural heritage."
In some situations the use of exonyms can be preferred. For instance, for multilingual cities such as Brussels, which is known for its linguistic tensions between Dutch- and French-speakers, a neutral name may be preferred so as to not offend anyone. Thus an exonym such as Brussels in English could be used instead of favoring either one of the local names (Brussel in Dutch/Flemish and Bruxelles in French).
Other difficulties with endonyms have to do with pronunciation, spelling and word category. The endonym may include sounds and spellings that are highly unfamiliar to speakers of other languages, making appropriate usage difficult if not impossible for an outsider. Over the years, the endonym may have undergone phonetic changes, either in the original language or the borrowing language, thus changing an endonym into an exonym, as in the case of Paris, where the s was formerly pronounced in French. Another example is the endonym for the German city of Cologne, where the Latin original of Colonia has evolved into Köln in German, while the Italian and Spanish exonym Colonia closely reflects the Latin original. In some cases no standardized spelling is available either because the language itself is unwritten (even unanalyzed) or because there are competing non-standard spellings. Use of a misspelled endonym is perhaps more problematic than the respectful use of an existing exonym. Finally, an endonym may be a plural noun and may not naturally extend itself to adjectival usage in another language, like English, which has a propensity to use the adjectives for describing culture and language. The attempt to use the endonym thus has a bizarre-sounding result.
The names for a country and a people are often different terms, which is a complication for an outsider, with a noticeable examples being people from the Netherlands being called the Dutch by native English speakers and in Italian the use of tedesco for German and Germania for Germany.
As modern technology removes many of the barriers between peoples, it is increasingly becoming the case that younger people may be more familiar with an endonym than with its official exonym. For example, many Italian cities are now more famous for their football teams and Torino and Napoli are becoming more common than Turin and Naples.
Sometimes the government of a country tries to endorse the use of an endonym instead of traditional exonyms outside the country:
Following the declaration in 1979 of Hanyu Pinyin spelling as the standard romanisation of Chinese, many Chinese endonyms have successfully replaced English exonyms, especially city and most province names in mainland China, e.g. Beijing ( B?ij?ng), Guangdong ( Gu?ngd?ng) (province), Qingdao ( Q?ngd?o), although older English exonyms are sometimes used in certain contexts - e.g. Peking (duck, opera, etc.), Canton, etc. (In some cases the traditional English exonym is based on a local Chinese dialect instead of Mandarin, in the case of Xiamen, where the name Amoy is closer to the Hokkien pronunciation.) In the case of Beijing, the adoption of the exonym by media outlets quickly gave rise to a hyperforeignized pronunciation, with the result that many English speakers actualize the j in Beijing as [?].
Matisoff wrote, "A group's autonym is often egocentric, equating the name of the people with 'mankind in general,' or the name of the language with 'human speech'." For example, various Native American autonyms are sometimes explained to English readers as having literal translations of "original people" or "normal people", with implicit contrast to other first nations as not original or not normal. Exonyms often describe others as "foreign-speaking", "non-speaking" or "nonsense-speaking". The classic example is the Slavic term for the Germans, Nemtsi, possibly deriving from a plural of nemy ("mute"): standard etymology has it that the Slavic peoples referred to their Germanic neighbors as "mutes" because their language was unintelligible. The term survives to this day in the Russian nemtsy (), Bulgarian nemtsi (), Ukrainian nimtsi (), Polish Niemcy, Czech N?mci, Slovak, Slovenian and Serbo-Croatian Nijemci/Nemci (?/), Montenegrin Njemci (), as well as in the Hungarian Német and Romanian Nem?i (both adopted from the Slavic), and even in the Turkish Nemçe and Arabic al-Nimsa (). The Turkish was adapted from the Slavic, and the Arabic from the Turkish, the words in both cases referring specifically to Austria.
One of the more prominent theories regarding the origin of the term "Slav" suggests that it comes from the Slavic root slovo (hence "Slovenia," "Slovakia"), meaning "word" or "speech". In this context, the Slavs are describing Germanic people as "mutes"—in contrast to themselves, "the speaking ones".
Another example of such development is the exonym "Sioux", an abbreviated form of Nadouessioux, derived most likely from a Proto-Algonquian term, *-a·towe·, "foreign-speaking".
In Basque, the term erdara/erdera is used for speakers of any language different from Basque (usually Spanish or French).
While the Irish and Scottish Gaelic words for England and its people are Sasana/Sasann and Sasanach/Sasannach ("Saxons"), the word for the English language is Béarla/Beurla, which derives ultimately from a word meaning "lips". In Old Irish, this word was applied to any foreign language, but by the medieval period it had come to be used exclusively for the English language.
Exonyms and endonyms must not be confused with the results of geographical renaming as in the case of Saint Petersburg, which became Petrograd () in 1914, Leningrad () in 1924, and Saint Petersburg (- Sankt-Peterbúrg) again in 1991. In this case, although St Petersburg has a German etymology, it was never a German exonym for the city between 1914 and 1991, just as Nieuw Amsterdam, the Dutch name of New York City until 1664, is not its Dutch exonym.
Old place names that have become outdated after renaming may afterwards still be used as historicisms. For example, even today one would talk about the Siege of Leningrad, not the Siege of St. Petersburg, because at that time (1941-1944) the city was called Leningrad. Likewise, one would say that Immanuel Kant was born in Königsberg in 1724, not in Kaliningrad (), as it has been called since 1946.
Although the pronunciation for several names of Chinese cities such as Beijing and Nanjing has not changed for quite some while in Mandarin Chinese (although the prestige dialect shifted from Nanjing dialect to Beijing dialect during the 19th century), they were called Peking and Nanking in English due to the older Chinese postal romanization convention, based largely on the Nanjing dialect, which was used for transcribing Chinese place names before Pinyin, based largely on the Beijing dialect became the official romanization method for Mandarin in the 1970s. Since the Mandarin pronunciation does not perfectly map to an English phoneme, English speakers using either romanization will not pronounce the names correctly if standard English pronunciation is used. (For example, in Mandarin the first phoneme in Beijing is an unaspirated voiceless bilabial stop, which is not the English b, the voiced bilabial stop, but an allophone of p that yet does not occur at the beginning of stressed syllables, where instead the aspirated variant is used. Thus an English speaker pronouncing either "Bei" or "Pe" according to English pronunciation rules will mispronounce the city's name in Mandarin.) Nonetheless, many older English speakers still refer to the cities by their older English names and even today they are often used in naming things associated with the cities like Peking opera, Peking duck, and Peking University to give them a more antiquated or more elegant feel. Like for Saint Petersburg, the historical event called the Nanking Massacre (1937) uses the city's older name because that was the name of the city at the time of occurrence. Likewise, many Korean cities like Busan and Incheon (formerly Pusan and Inch?n respectively) also underwent changes in spelling due to changes in romanization, even though the Korean pronunciations have largely stayed the same.
The name Madras, now Chennai, may be a special case. When the city was first settled by Englishmen, in the early 17th century, both names were in use. Possibly they referred to different villages which were fused into the new settlement. In any case, Madras became the exonym, while more recently, Chennai became the endonym.
Likewise, Istanbul (?stanbul in Turkish) is still called Constantinople (?) in Greek, although the name was changed in Turkish to disassociate the city from its Greek past between 1923 and 1930 (the name Istanbul itself derives from a Medieval Greek phrase). Prior to Constantinople, the city was known in Greek as Byzantion (Greek: , Latin: Byzantium), named after its mythical founder, Byzas.