Isolated languages

from Wikipedia, the free encyclopedia

An isolated language is a language that has no genetic relationship to any other language. As far as we know today, the only isolated language in Europe still spoken today is Basque .

This article provides an overview of the world's isolated languages, sorted by continent.

Problems of the term

The apparently simple definition of the term "isolated language" - a language for which no genetic relationship with any other language has been proven - is problematic:

  • Problem of dialects and splits: Even isolated languages ​​can of course have dialects and these can develop into independent languages ​​today or in the future. It is often a matter of definition whether one speaks of "one" language with several dialects or several independent, closely related languages. This not clearly definable difference turns an “isolated language” into a small family of closely related languages.
  • Problem of the extinct sister languages: The Ket (a Yenisei language ) would have to be considered “isolated” today, since no other Yenisei language is spoken apart from Ket. From the previous centuries, however, at least five sister languages, which have since become extinct, are known, which is why the Ket cannot be viewed as isolated, but is the only surviving member of the Yenisese language family. Likewise, z. For example, Sumerian ancient relatives, which are now considered to be isolated, had relatives which - in contrast to Sumerian - were not recorded in writing.

The notion of “isolated languages” according to the above definition is therefore rather unhistorical: in reality, all languages ​​in the world have or had “relatives” who simply could not be determined for various reasons. B. because they became extinct without written fixation or because the research methods are not (yet) sufficient to prove the relationship with other languages. If one assumes that no language was created out of “nothing”, all languages ​​are genetically related to one another . However, this cannot be proven in practice, since any genetic similarities have been completely (sometimes several times) masked by other effects of language development over the millennia. Nevertheless, or precisely because of this, linguistic research on the large hypothetical macro- families , which comprise several “classical” language families and “isolated languages”, is of particular interest in this context . (See also Eurasian , Nostratic , Dene-Caucasian , Australian , Indo-Pacific , Amerindian, and others). With the establishment of each larger genetic language unit, the number of "isolated languages" is reduced.

In the following, most of the languages ​​considered to be “isolated” today are listed alphabetically with their number of speakers by continent and within them. Some are advised of the reasons why they may not be "isolated" after all. The basis is the compilation of Ethnologue , which was supplemented by specialists based on additional knowledge and corrected where necessary.

Not to be confused with the “isolated languages” are the “ unclassified languages ”: these are either extinct and so weakly handed down that classification with the available material is not possible, or they still exist but not yet sufficiently researched to make them one To be able to assign a language family or to claim with great certainty that they are "isolated". Languages ​​can also be unclassifiable if they have the characteristics of two language families. Such languages ​​are called hybrid or mixed languages . In practice, the difference between “isolated” and “unclassified” is not very sharp, so some unclassified languages ​​are also listed here.

Ethnologue as a problematic source

For almost all of these isolated languages, some of which are only known to experts, information about the ethno-linguistic situation, the number of speakers and their geographical distribution can be found in Ethnologue . However, the classification in Ethnologue is not always compatible with the one used here, so that some of the languages ​​referred to here as isolated are assigned to a language family there. Conversely, some languages ​​designated as isolated in Ethnologue are assigned to a language family here, e.g. B. Tinigua , Puelche and others. In addition, some confusion of names in Ethnologue lead to erroneous classifications and assessments. In case of doubt, the current primary sources for the individual language families or area-specific language groups should always be consulted.


Isolated languages ​​of Eurasia

The following languages ​​of Eurasia are considered isolated by the majority of researchers today:

language Number of speakers Occurrence / comment
Ainu nearly † Japan , Sakhalin ; Candidate for the Eurasian macro family, weak candidate for Macro Altai
Basque 800,000 Spain , France ; the extinct Aquitaine may have been a dialect of ancient Basque; probably not related to the old Iberian languages . Possibly. Member of the Dene-Caucasian macro family (JD Bengtson) or the Proto-Eurasian macro family (M. Morvan).
Burushaski 100,000 Pakistan ; perhaps relationship with the Yenisian, candidate for Dene-Caucasian
Elamish Iran : Khusistan ; Relationship with Dravidian possible: "Elamo-Dravidian"
Hattisch Anatolia
Jukagirian 200 Russia , Eastern Siberia ; possibly related to the Uralic languages : "Uralisch-Jukagirisch"
Korean 78,000,000 Korea ; possibly related to Japanese; possibly. along with Japanese to Macro Altaic
Kusunda nearly † Nepal ; from the oldest language class in the Himalayas
Nahali 2,000 India ; alternatively Kalto, Nahal, Nahale, Nihal; from the oldest language class in India
Niwchisch 400 Russia, Eastern Siberia; Candidate for the Eurasian Macro Family
Sumerian southern Mesopotamia ; possibly dene-caucasian

Unclassified living languages ​​of Eurasia

These languages ​​are listed in Ethnologue as " unclassified " because they have not yet been researched or not sufficiently researched to be able to classify them genetically. It is rather unlikely that these are "isolated" languages.

language Number of speakers Occurrence / comment
Andh 80,000 India ; only as a second language, the speaker's mother tongue is Marathi , probably Indo-Aryan
Chak 25,000 Myanmar ; probably Sino-Tibetan, branch unknown
Malakhel 2,000 Afghanistan ; probably identical to Ormuri , then Iranian
Mukha-dora 17,000 India; today only second language, poss. Mother tongue is Telugu, speakers also use Adivasi Oriya; probably Dravidian
Warduji 5,000 Afghanistan ; Possibly a dialect of Persian, very likely Iranian (Pamir language?)
Waxianghua 300,000 China ; very likely Sinitic branch of Sinotibetan

Unclassified extinct languages ​​of Eurasia with legible scripts

The extinct old European and oriental languages ​​in this list could not yet be assigned to any language group, as they are only very sparsely preserved and documented. Whether these are isolated languages ​​cannot be conclusively clarified ( pre-Indo-European substrate ).

language Occurrence / comment
Eteocetan Crete ; Language of some Iron Age inscriptions
Eteocyprian Cyprus, Crete; Language of some Iron Age inscriptions in the Cypriot syllabar
Etruscan Italy ; a relationship with Lemnian and other Aegean languages as well as with Rhaetian as so-called Tyrsenic languages were u. A. postulated by Helmut Rix , but has not been proven.
Gutäisch Mesopotamia; only proper names known, cuneiform , from the Zāgros Mountains , possibly related to the Lulubian
Iberian Iberian Peninsula; Connections to Basque rather unlikely; clearly distinguishable from Tartessian ; not to be confused with Ibero-Celtic
Kassitisch Mesopotamia; only proper names and a tablet known, cuneiform, from the Zagros Mountains
Lulubian Mesopotamia; only proper names known, cuneiform, from the Zagros Mountains, possibly related to the Gutaean
North Pikenish Italy; in contrast to the Italian South Piken, it is certainly not the Indo-European language
Pictish Scotland ; is sometimes assigned to the British branch of the island Celtic languages
Sican Sicily; in contrast to the neighboring languages Elymian and Siculian, which are mostly regarded as Italian, probably non-Indo-European
Tartessian or
South Lusitan
Relationship with the Iberian unlikely, certainly not Ibero-Celtic

Unclassified extinct languages ​​of Eurasia in as yet undeciphered scripts

The languages ​​in this list cannot be assigned to any language group, as their scripts could not yet be deciphered. So they are actually hypothetical languages.

language Occurrence / comment
Harappan Pakistan , India ; Language of the Indus script that has not yet been deciphered; mostly regarded as related to Dravidian, then of course not in isolation
Minoan Crete; Language of the texts in the not yet fully deciphered Linear A script and some other not deciphered Cretan texts, e.g. B. the disc of Phaistos; possible relationship with Eteocyprian and Eteocetan
Kyprominoic Cyprus; Language of the not yet deciphered Bronze Age inscriptions of Cyprus, especially from Enkomi ; possible relationship with Minoan and Eteocyprian

Languages ​​of Eurasia, often incorrectly classified as isolated

language Number of speakers Occurrence / comment

Hurrian ( Syria , Eastern Anatolia) and Urartian (Eastern Anatolia, Armenia , northwest Iran) together form the (isolated) Hurrian-Urartian language family ; not classified as isolated due to mutual relationship; Relationship to the North Caucasian languages ​​postulated but not recognized
Japanese 126,000,000 Japan; not isolated, as the widely differing Ryūkyū languages ​​of Okinawa are mostly evaluated as independent languages; possibly related to Korean, possibly member of Macro-Altaic; Candidate for Eurasian and Nostratic
Ket 200 Russia, Siberia ; Yenisan language, related to the now-extinct Yenisan languages Yugh , Kott , Pumpokol , Assan and Arin ; the Yenisan languages ​​are a candidate for the Dene-Caucasian ; more recent hypotheses of the genetic relationship to the Na-Dené languages or to Burushaski


The Austronesian and the so-called Papua languages belong to the Indo-Pacific area . While the Austronesian languages ​​form a clearly identifiable genetic unit - there is little doubt whether a particular language is Austronesian or not - the genetic situation of the Papuan languages ​​is very complex. Some of the non-Austronesian and non-Australian languages ​​of New Guinea and some of the neighboring islands (the so-called Papua languages) are isolated in the sense that they cannot be assigned to any “family” (even if only one with two members) . There are also some little-researched, so far unclassified Papuan languages ​​that could well belong to one of the ten or so Papuan language families.

Isolated and unclassified Papuan languages

language Number of speakers Occurrence / comment
Abinomn 300 Indonesia: West New Guinea; maybe to Yoke-Warembori
Burmeso 250 Indonesia: West New Guinea; maybe to Ost-Vogelkopf
Doso ? 700 Papua New Guinea
Karkar-Yuri 1,100 Papua New Guinea; maybe to Pauwasi
Kembra nearly † Indonesia: West New Guinea
Kibiri 1,100 Papua New Guinea; maybe to Trans New Guinea, Kiwai group
Lepki 500 Indonesia: West New Guinea; perhaps related to Murkim
Murkim 300 Indonesia: West New Guinea; perhaps related to Lepkin
Odiai or Busa 250 Papua New Guinea
Yale 600 Papua New Guinea


According to new findings of one of the best experts in these languages, RMW Dixon (see reference), the genetic unity of the Australian languages cannot be proven because the phase of possible uniformity is far too far back (at least 20,000 years). In particular, the genetic unit of the largest group of Australian languages ​​to date, the Pama Nyunga languages, is being abandoned by Dixon. According to this model, there are various smaller genetic and areal units, many languages ​​remain "isolated". Only the "isolated languages" that are still alive are listed here, including the most spoken languages ​​in Australia. In total, only 40,000 people of the Aboriginal people speak one of the 80 or so native languages; there were at least 350 languages ​​before the colonization.

language Number of speakers comment
Kala Lagaw Ya or West Torres 1,000 Queensland: Western Torres Strait; Papuan language with Australian substrate
Kuku-Yalandji 240 Queensland; the dialects Yalandji, Djangun and Muluridji are sometimes considered as separate languages, then a small language family
Tiwi 1,800 Northern Territory; possibly small language family
Whyungu 500 Northern Territory
Western Desert Language 6,500 Western Desert; the dialects Martu Wangka (700 speakers), Karkutja (300 speakers), Pintupi-Luritja (1,000), Ngaatjatjarr (1,200), Pitjantjatjara (2,500) and Yankuntjatjarra (200) are sometimes considered to be separate languages; then the Western Desert Language Group is a small language family

In addition, a number of extinct or near-extinct languages ​​of Australia must be considered isolated or unclassified. See the external link “The Classification of the Australian Languages”.


Isolated languages ​​in Africa

According to Joseph Greenberg's now undisputed achievement of an overall classification of African languages into the four great phyla Afro-Asian , Niger-Congo , Nilo-Saharan and Khoisan (last version published in 1963), no African language was initially "isolated" because they all belonged to one of the four phyla.

Some Africanists are convinced, however, that Khoisan and Nilo-Saharan are not language families in the narrower sense, but only groups of languages ​​that are typologically similar and form an areal linguistic union . (Details on this topic in the article African languages .) So while some experts consider certain languages ​​to be part of the Phyla Nilosaharan or Khoisan, if no close relatives are recognizable they are considered by others as "isolated" or "unclassified". In addition, there are a number of unclassified and hybrid languages ​​(mixed languages) in Africa.

Isolated languages

The following languages ​​are now considered to be isolated, in the past they were added to the Khoisan languages ​​(so still in Ethnologue):

language Number of speakers Occurrence / comment
Hadza 800 Tanzania
Sandawe 40,000 Tanzania
Kwadi Angola
ǂHoa 200 Botswana ; possibly a South Khoisan language

Unclassified African languages

The following languages ​​probably belong to one of the great African phyla, a reliable assignment has not yet been possible:

language Number of speakers Occurrence / comment
Exercise nearly † Cameroon ; Probably Niger-Congo language, bantoid
Centúúm 200 Nigeria
Kujarge 1,000 Chad
Lufu 3200 Nigeria; maybe Niger-Congo language, jukunoid
Luo Cameroon; not to be confused with the Nilosaharan Luo
Meroitic Sudan ; the language of the Nubian kingdom of Meroe ; possibly Nilosaharan
Mpre Ghana ; probably Niger-Congo language, maybe Kwa branch; only described once in 1931
Pré 200 Ivory Coast ; probably Niger-Congo language; to the Volta-Congo branch rather than the Mande branch

Hybrid African languages ​​(mixed languages)

The following languages ​​have characteristics from different language families:

language Number of speakers Occurrence / comment
Imeraguen 500 Mauritania ; Features of Afro-Asian (Semitic: Hassaniya) and Niger-Congo (Mande: Soninke)
Laal 750 Chad ; Characteristics of Afro-Asian (Chadian) and Niger-Congo (Adamawa); unknown substrate
Shabo 450 Ethiopia ; Features of Afro-Asian (specifically Omotic) and Nilo-Saharan
Weyto Ethiopia; possibly Nilo-Saharan (East Sudan?) or Afro-Asian (Kushitic?); probably hybrid

Sometimes African languages ​​incorrectly classified as isolated

language Number of speakers Occurrence / comment
Kara (Fer) 4,800 Central African Republic ; Nilo-Saharan , Central Sudanese, Bongo Bagirmi group; at Ethnologue wrongly stated as "unclassified"
Ongota (Birale) nearly † Ethiopia; after Harold Fleming a separate branch of Afro-Asian; at least Afro-Asian


"Only the living languages" are listed here. A complete overview of the current state of knowledge - including the extinct, but somehow handed down languages ​​- is given on the web links "Classification of North, Meso and South American Languages" and the article Language families of the world . If Joseph Greenberg were right in classifying American languages into just three genetic units - Na-Dene, Eskimo-Aleut, and Amerind - there would be no isolated American languages. The present presentation is based on the now generally recognized genetic groups of America, whereby attempts are made again and again to demonstrate larger units in which the "isolated" languages ​​listed here can then also be absorbed. According to the majority opinion of researchers today, America (especially South America) is the continent with the most - sometimes quite large - isolated languages.

North America

Isolated languages ​​in North America

Note: The groups Hoka languages ​​and Penuti languages ​​are not considered here as genetic units. (See L. Campbell, American Indian Languages ​​(1997) and various other current sources.)

language Number of speakers Occurrence / comment
Haida 50 Canada; often counted as Na-Dene
Karuk (Karok) nearly † United States
Keres 8,000 UNITED STATES; possibly family of 2 languages ​​with 7 varieties
Klamath-Modoc nearly † UNITED STATES; possibly related to the Sahapti languages ​​and the Molala
Kutenai nearly † Canada
Washo nearly † United States
Yuchi nearly † United States
Zuñi 10,000 United States

Central America

language Number of speakers Occurrence / comment
Huave 19,000 Mexico ; four dialect groups that could also be counted as languages
Seri 700 Mexico
Purépecha 120,000 Mexico
Tol 350 Honduras
Lenca nearly † Honduras , San Salvador ; possibly two languages: Honduras and San Salvador Lenca

South America

Isolated languages ​​in South America
language Number of speakers Occurrence / comment
Agavotaguerra Brazil ; unclassified; maybe Arawak language
Aikaná (Tubarao) 100 Brazil; unclassified; maybe Arawak language
Amikoana nearly † Brazil
Andoque 600 Colombia
Camsá 4,000 Colombia
Candoshi Shapra 3,000 Peru
Carabayo 150 Colombia
Chiquitano 6,000 Bolivia; according to Kaufmann, a macro-Gé language
Cofan 2,500 Ecuador, Colombia; Loans from the Chibcha languages
Guarao (Warao) 18,000 Venezuela , Guyana
Himarimã nearly † Brazil
Irantxes 200 Brazil; maybe Arawak language
Itonama nearly † Bolivia
Karahawyana nearly † Brazil
Leco nearly † Bolivia
Araucanian languages
( Mapudungun , Huilliche )
260,000 Chile, Argentina
Movima 1,500 Bolivia
Paez 80,000 Colombia
Puinave 2,000 Colombia
Taushiro probable † Peru
Ticuna 41,000 Brazil, Colombia, Peru
Trumaí 100 Brazil
5,500 Bolivia; Tsimané and Mosetén form a family of two.
Kinship with the Chon family postulated by Kaufmann.
Urarina 3,000 Peru
Uru-pa-in 200 Brazil
Waorani (Sabela) 1,700 Ecuador
Yámana probable † Chile
Yari 800 Colombia; maybe to Carib or Huitoto
Yaruro (Jaruro, Pumé, Llaruro, Yuapin) 3,400 Venezuela ; perhaps related to the Esmeralda †
Yuracare 3,000 Bolivia
Yuwana (Joti, Waruwaru, Chicano) 300 Venezuela


Web links