#language isolate

[victusinveritas]

One of these days I'll make a list of languages that are connected linguistically but really far away on a map. Today isn't that day though.

[thoughtportal]

Japan is an ethnically homogenous nation where everyone speaks Japanese, right? Not exactly. Other groups including the Ainu also have called Japan home, perhaps for longer than the Japanese themselves. Today, the Ainu language is spoken by only a handful of people. One of them, Russian-born linguist Anna Bugaeva, takes Patrick Cox to meet Ainu speakers (and non-speakers) on the island of Hokkaido. Along the way, we learn about the mysteries of Ainu, a “language isolate” unrelated to any other language in the world. Bugaeva says Japanese children aren’t taught about the Ainu because their presence—and language—contradict standard Japanese history.

Music in this episode by Tonality Star, Podington Bear, Circus Marcus and Blue Dot Sessions. Photo of Ainu language activists Maki and Kenji Sekine by Patrick Cox. More on Anna Bugaeva’s research here. Read a transcript of the episode here.

[endangeredlanguagechallenge]

Endangered language challenge #4: Kusunda language

Kusunda (Kusanda, mihaq) is a language spoken by the Kusunda people in Nepal. It was originally thought to have died in 1985. But 2004, 3 speakers were discovered and most of the documentation we have today results from intense work with them. They belong to the ethnic group of Kusunda or Ban Raja, which is translated as “The Kings of the Forest”. It is an exonym, given by outsiders and not themselves, and it points to the fact that many Kusundas live in the forest. They have no problems with the term “Kusunda” (while in some regions of India it is used as an insult) but they refer to themselves as “mihaq” the peoples. Nowadays, there are only 160 ethnic Kusundas left. Of the 3 remaining speakers, none is monolingual, all of them also use Nepali. They all married outside their ethnic group and no child is raised in this language anymore, making Kusunda a moribund language. While most speakers vary widely in their fluency, Gyani Maiyi Sen is the most famous speaker and the most fluent if not only native one, she often gets referred to as the last speaker. (She is depicted in this video.)

The interesting thing about Kusunda is, while earlier researchers thought of it as a Tibeto-Burman variety, it is now evident that it is a language isolate, thus, not related to any other known language. It is probably the last remnant of languages that were spoken before the Tibeto-Burman and Indo-Iranian languages came to that region and it is not related to the Munda or Dravidian languages. The groundwork was laid by David Watters 2005 when he published the workings from the intense working with the three remaining speakers in a Kusunda grammar, which offers grammatical, typological and vocabulary insights. 

Kusunda Phonology

Kusunda hat some interesting features: - It has six vowels that oppose each other in vowel harmonies, making them phonologically only three phonemes. Normally, only /i, ə, u/ or only /e, a, o/ can be together in one word. Theoretically, one set could be replaced by the other set without changing the meaning. - Kusunda is phonologically very distinct to its surroundings: It has uvular and pharyngealized consonants that don’t appear in neighbouring languages, and while retroflex consonants are prevalent in that region they don’t occur in Kusunda. - While English contrasts sounds by the place of articulation and the articulator, in Kusunda, mostly only the active articulator is of importance. So, it wouldn’t matter where the tip of the tongue (apical consonants) produces a sound, but only that it is done by it. Throughout the words, one would find apical consonants that are dental, alveolar, retroflex and palatal without changing the meaning. Even voicing, like the difference between /p/ and /b/, seems to play a minor role, as does aspiration, and many sounds can be stops here and fricatives there.

Kusunda Grammar

Very strikingly and highly unique among the world’s languages is the way Kusunda distinguishes between marked and unmarked structures. Irrealis would be a marked feature in comparison to the unmarked realis, same goes for negativity versus affirmation and other features. Normally, languages would make use of affixes or other means to denote these structures. Kusunda however, deploys a harmonic autosegmental process (often called mutation) that applies to the whole word. An apical consonant would then shift to a laminal consonant, a velar consonant to a uvular one and the first set of vowels would change to the second one, in order to mark the irrealis in contrast to the realis of the negativity in contrast to the affirmative modus. Here is an example: If you say “I go” you would apply the realis because it is present time. The form would be /ts-əg-ən/, but if you want to say “I will go” you have to use the irrealis and thus change the consonants’ articulator: /tʃ-aɢʕ-an/ - as you see, the velar /g/ of the root becomes uvular /ɢʕ/ (among the other changes) and is thus marking the irrealis by mutation.

Verbs are roughly divided into two groups. Class I gets personal prefixes and numeral suffixes: First persons are indicated by /t-/ or /ts-/, second persons by /n-/ and the third persons by /g-/ or others. The singular is indicated by the suffix /-Vn/ (V stands for a vowel) and the plural by the suffix /-da-/ that comes directly after the root and is followed by /-n/. So, you can distinguish “I eat” (t-əm-ən) from “You eat” (n-əm-ən), or “I eat” (t-əm-ən) from “We eat” (t-əm-da-n). The prefixes also resemble the first consonants of the personal pronouns a lot, which are /tsi/ and /tok/ for the 1. P. Sg. & Pl., /nu/ and /nok/ for the 2. P. Sg. & Pl. and /gina/ for the 3. P. which has no distinction between plural or singular. The verbs from Class I are few in number and highly irregular.

Most verbs are in Class II, which differ in a lot of features from Class I verbs. If the verb is transitive, the conjugated forms of “I made, we made, you made, etc.” come after the verb root. So, if you want to put the verb “to buy” /dza/ in the first person singular you add /a-t-n̩/  which simply means “I made” > dza-a-t-n̩ (I made buying). If the verb is intransitive, you simply add the personal pronoun after the root and add a /-n/. From “to enter” /sip/ you would simply come to “I enter” by adding /tsi/ and /-n/ > /sip-tsi-n/. 

That is just a peek into the grammar, there are a lot of very interesting and unique features that Kusunda has to offer and it is definitely worth a further look into the grammar David Watters wrote.

Kusunda Phrases

The Australian National University in Canberra has some very good resources on Kusunda on their website. If you want to hear the sound files to the following phrases or other sentences or even texts, visit this website. 

Sodzaq! - Greetings! (hello or goodbye) Tsi gidzi Gyani Maiya - My name is Gyani Maiya Tsi tugun - I've come (I'm here) Nəti? - Who (is it)? Nu wee? - Are you well? Nətn nitn? - What did you say? Hampena nugun? - Where have you come from? Ni whi hampe? - Where is your house? Tsi whi Kulmor - My house is in Kulmore Garaw əni - It's hot Idang ugun - I'm hungry Pəwhəiran əni - I'm tired Aw təgəi! - Let's go! Hana nyhaan? - Where are you going? Kafera muitsidaq - We'll meet again 

Gyani Maiya’s words about the language loss: Ta gepən mi: əndzi - This language is going to be lost. Nətima gepən ədu - There is none to talk with. Mjaqa tsi toqdu ta qau nətn ba - I am alone and if I died, then there is nothing. Ləmbə ədi ta nəti ləmbə əgəndzi - Though I am ready to teach, is there anyone to learn? Toqdu bela ənin jiodze gepən əgu ta ləmbə əgu ta we: da - It would be better if they learnt since I am very old now.

Kusunda sources

See for further linguistic material the link to Glottolog and check the links under the Wikipedia article or the Links page of the ANU Kusunda Linguistics article.

Ethnologue: Kusunda language Glottolog: Kusunda language https://en.wikipedia.org/wiki/Kusunda_language Kusunda grammar, David Watters (2006) Kusunda Linguistics, information, texts, audio (Australian National University)

[cosmogyros]

feeling really validated bc I was playing the LingQuest language-matching game, and after getting a few right, I was presented with one where I couldn’t identify the language family at ALL. like I literally had no clue, and usually I can tell what family a language is from pretty well, even if I can’t identify the specific language. but guess what language it turned out to be? BASQUE.

[dedalvs]

Do you know of any quantitative studies on whether or not language isolates as such tend toward a certain typology? I've been looking around the internet and can't seem to find anything on this question besides a conference abstract and a reddit post, which is stunning (to me at least).

I do not. Interesting question! I don’t know if anyone’s done a typological survey specifically and exclusively on language isolates. Personally, I’m not sure it would be very interesting—a bit like doing a typological survey of conlangs in general, without specifying which type. Of course, this betrays my own feeling about what the answer will be: That there is no unifying typological thread for language isolates. I, though, have no study either to back up my claim, or disprove it. Perhaps someone should undertake one! (Or, if anyone knows of one, please share!)

[kephas-rolling]

Language isolates in North America (2/2) : Atlantic coast and central

Coahuilteco (white) : Amongst all the Native American languages that have been recorded, Coahuilteco is one of the poorest documented. The language has been extinct since at least the 18th century. The drastic lack of information concerning Coahuilteco stems from the fact that the Spaniards did not really pay heed to the speakers of the language. For instance, the estimation of the period around which the language went extinct comes from the lack of written mentions around that time. It was never clearly said that the Native Americans went extinct for a particular reason and when this happened. Consequently, most of the available info we garnered today was from studies once the ethnic group and the language ceased to exist. It was once wagered that the whole of south-eastern Texas was teeming with numerous Coahuilteco speakers but is now hypothesised that it was merely a widespread L2. Linguists are running short of details as to the grammar, the number of L1 speakers and if there were dialects. Since it is hardly feasible to classify Cohuilteco into a language family, it is thought of as an isolate.

Zuni (light purple): Zuni (also spelt Zuñi) is still spoken today in New Mexico and Arizona by 9 686 people according to a census conducted between 2006 and 2010. This is a slight increase since the previous census in 2000 which found 9 650 speakers. There are very few monolinguals of Zuni; many Zuni people speak perfectly English. Ethnologue reports that more than 7 000 people (more than 72%) use Zuni at home. The language is used at home, in traditional tribal council meetings, in all religious ceremonies, and occasionally in religious services. Many efforts have been made to try and find a place for Zuni in pre-existing language families but these efforts have been fruitless. Most common points between Zuni and other geographically close languages are mainly about shared vocabulary and phonemes (specifically ejective phonemes) but it is most commonly accepted that these lexical and phonological features are but mere borrowings from surrounding tongues.

Tonkawa (blue) : Tonkawa has been gone since the 20th century, estimations have it at around the 1940s. Tonkawa has been replaced today by English in the Tonkawa tribes but the language is fortunately well recorded, whether it is its phonology or grammar. Tonkawa has a taboo phenomenon remarkably close that of the Kwaio language in the Solomon Islands. Words would fall out of use because of the taboo associated the sonority of a words and that of a deceased member of the tribe. This would cause a need for a new word to express a familiar concept. This is why the vocabulary that Chowell lay down on paper in 1829 was pretty different from the one that Hoijer got a century later. For instance, the words for “beard” or “eyes” had differed in the span of a century because in 1892, 167 Tonkawas were killed in a massacre. 

Atakapa (brown): Atakapa is also an extinct isolate. It was spoken in the Gulf of Mexico region. Although divided in several bands, Atakapan speakers were never numerous to begin with. Regarding dialects, there are conflicting reports when it comes to their number. There might have been two or three. John R. Swanton (1929:122)  indicates though that there were most probably three dialects based on the words and sentences collected from different bands. Nothing is sure. The decline of the Atakapan is mostly due to the growing numbers of settlements from Americans after the Louisiana was sold by the French in the late 18th century. After the Louisiana Purchase, the Atakapan either retreated in the hinterlands or stayed close by. Their numbers dwindled up to the point of reaching 0 in the early 20th century. 

Chitimacha (light orange): This language is officially extinct. However, reality is more complex than that since there have been since in the 1990s great endeavours to revitalise the Chitimacha tongue. There has been a partnership between “[the tribe and] the software company Rosetta Stone on a two-year project to create computer software for learning the language, which today every registered tribal member has a copy of.” Ethnologue makes no mention of the reservation’s school, set in Louisiana, where kids actually exchange in the language. Although not all Chitimacha speak the language, amongst the 1300 members (according to their website), the language is still alive today. Proof: kids learning the language actually come up with slang words. Chitimacha could have never come back from its 60-year-long slumber after the death of its last two speakers: Benjamin Paul and Delphine Ducloux. They were the last speakers and died in the 1930s but not before Morris Swadesh dedicated two years of life to gathering the words and myths that Ben and Delphine had inherited from their ancestors. It was based on these records that the revitilsation effort of Chitimacha is working.

Natchez (green): Native to Mississippi , Natchez belongs to the long list of extinct language in North America. Natchez went extinct with the death of Watt Sam in 1944 and his cousin Nancy Raven in 1957. The Natchez people dissolved into the neighbouring the communities after the Natchez chiefdom was destroyed in the 1730s by the French. The Natchez people assimilated into other tribes which had the impact of creating bilingual speakers which eventually had the sad consequence of relegating the Natchez language to a distant memory since the language was no longer needed. The early 1900s saw linguists John R. Swanton (in 1907) and Mary R. Haas (in the 1930s) work with these two speakers. One of the only voice recordings of Natchez on wax cylinders of Watt Sam and is kept at the University of Michigan. One of the most remarkable features of Natchez lies in its phonology; particularly the voicing contrast that is applied to its sonorants (namely /m̥ m/, /n̥ n/, /l̥ l/ and /w̥ w/) and not to its obstruents as it is most common. Linguists have tried to link Natchez to the Muskogean languages but the proposal has yet to convince a majority of linguists. 

Tunica (red): Tunica is in effect a dead language, it was spoken in Mississippi, Louisiana, and Arkansas. However, there have been initiatives that aimed at breathing life back into it. Fortunately, as the Tunica tribe website indicates, thanks to these efforts, there are now around 60 speakers ranging from beginner to intermediate level. “ In 2010, the tribe partnered with the Tulane University Linguistics Program, Department of Anthropology, to start the Tunica Language Project. In an effort to re-awaken the Tunica language, plans were put in place to work with the tribal community to bring the language back.” Tunica had become dormant since the 1950s after the death of Sesostrie Youchigant but not before he worked with linguist Mary Haas to describe what he remembered of the language, which he had learned as a child.    

Yuchi (purple): Yuchi’s fire is burning ever so dimly. The language had, in 2012, 5 L1 speakers, who were at least 80 years old at the time, according to a video on their YouTube channel. As of 2017, it seems that there only three of them remaining as one passed away in 2013 and another one in 2016. Despite dwindling numbers there is still some hope as there has been since 1994 a Yuchi Language Project, notably with language classes or Annual Yuchi Knowledge Bowl. The language was native to the Tennessee and later Georgia regions before being forcibly moved to Oklahoma, following the passage of the Indian Removal Act in 1830. An event that was also called the Trail of Tears. It is claimed that Yuchi is distantly related to the Siouan family but nothing is sure. 

Timucua (orange): Timucua has disappeared since the turn of the 19th century. The tribe had thin ranks by 1700, around 1000 of them and the conflicts with and enslavement by English colonists and other warring tribes had completed their eradication by the time the USA acquired Florida. A Spanish member of clergy, Father Pareja, recorded in his 31 years of service 9 to 10 different dialects. 

Language isolates in North America ( 1 / 2 ) : Pacific Coast

[mapsontheweb]

Map showing the most isolated languages

[kephas-rolling]

Language isolates in North America ( 1 / 2 ) : Pacific Coast

Haida (dark blue) : Haida is native to Queen Charlotte Islands (Canada) and Prince of Wales Island (USA). Haida is categorised by UNESCO as a “critically endangered” language with only 24 speakers in 2014. It experienced a sharp decline from the 15,000 speakers that were wiped out by epidemics brought by the European explorers a little after the mid-18th century. Efforts have been made to revitalise the moribund language, such as classes at the University of Alaska Southeast or an iPhone application based on a "bilingual dictionary and phrase collection." Anthropologist Franz Boas proposed a kinship between Haida and the Tlingit language in 1894 and linguist Edward Sapir considered Haida as a member of Na-Dené language family in 1915. Contemporary linguists are of the mind that there is no tangible evidence of such claims and therefore, Haida is an isolate.    

Kutenai (brown) : Kutenai is the language of the Ktunaxa people (who are spread over Montana, Idaho and British Columbia); there are approximately 1,510 of them, and amongst them, only 31 can speak the language. That is barely above the 2% mark. To remedy the decaying state of Kutenai, the Confederated Salish and Kootenai Tribes of the Flathead Nation have founded the Salish Kootenai College. This college gives the opportunity ti earn a certificate in Native American studies by attending classes about the history, culture and language (pronunciation and grammar) of the Salish and Ktunaxa people. One of the most in-depth works carried out on the language was written by Lawrence Morgan in 1991. He wrote A Description of the Kutenai Language in 1991 as his PhD thesis. 

Takelma (pink) : The last speaker of Takelma died in 1934. Her name was Frances Johnson. She talked with Edward Sapir in 1906 when he was writing his thesis The Takelma Language of Southwestern Oregon. Frances Johnson (also known as Gwísgwashãn) last documented participation in efforts to document Takelma was in 1933, with anthropologist John Peabody Harrington. The latter took up to 1,200 pages of field notes while discussing with Johnson. Takelma was thought for a time to be a member of the Kalapuyan languages before this hypothesis was debunked by Tarpent & Kendall in 1995. Sapir proposed Takelma’s membership to the controversial Penutian family. 

Siuslaw family (green) : To talk about the Siuslaw language family can be somewhat problematic in so far as there are both arguments in favour and against the divide of the two Siuslaw entities into two dialects or two languages.It appears that the differences between the two are on a phonological and lexical level (Page 234). Is it enough to categorise two dialects into languages? Regardless of the controversy, The language(s) was/were spoken in south Oregon where the last speaker died in the 1970s. Sapir included Siuslaw in his aforementioned Penutian hypothesis.

Klamath (yellow) : Native to south Oregon, Klamath can be a controversial inclusion in the category of North American language isolates because it is often included in Sapir’s proposed language family of the Penutian languages. American linguist and indigenous American language specialist Lyle Campbell recreated the evolutions and equivalences of consonantal mutation between several languages (Klamath, Maidu, Wintu, Patwin, Yokuts, Miwok andCostanoan) from a hypothetical common ancestor as evidence of the existence of the Penutian family. There are nowadays no more native speakers of Klamath; the last one of them died at age 92 in 1993. There were however 6 L2 speakers in 2007.

Karuk (dark red) : Spoken in northwestern California, Karuk is an endangered language, with barely 12 L1 speakers but also 30 L2 speakers thanks to efforts by the Karuk people and American linguist William Bright, who “was made an honorary member of the Karuk tribe”. “Advocates for Indigenous California Language introduced the master-apprentice program to the tribe (linguist Hinton is one of the main trainers). It's an immersion method in which a fluent indigenous language speaker is paired with a beginning speaker, and they spend up to 40 hours a week together for three years, doing everyday things while communicating only in the indigenous language. About 20 Karuk teams have completed the program and more are signing up”. Sapir classified it as a member of the Hokan family but Lyle Camplbell (2000) estimates that evidence is far too lacking to back up this hypothesis.

Chimariko (blue) : The language went extinct around the 1950s, when its last speaker, Martha Ziegler passed away. Chimariko is hypothesised to have had three dialects spread around Northern California. Once again, Edward Sapir tried to include an indigenous American language in a super language family but linguists tend to view as an isolate despite the presence of cognates and phonological traits common to other neighbouring languages. It is argued that these similarities merely exist because of intense cultural and commercial exchanges between the communities of Native Americans at the time, giving opportunities to some lexical and phonological traits to spread around an area in spite of genetic linguistic differences. 

Yana (light green) : This central Californian language went extinct in 1916, when the 57-year old Ishi died of Tuberculosis. Dubbed the “last wild Indian” of the USA, he lived outside modern culture and therefore had not developed an immune system to the diseases that Europeans had brought with them as they pushed the borders evermore westward. He was the last speaker of Yahi dialect, a sub-branch of one of the four recorded Yanan dialects. Sapir had classified Yana as belonging to the Northern Hokan sub-branch (which included the above Chimariko and Karuk) but Yana remains in the minds of linguists today an isolate.

Washo (light red) : Washo belongs to the category of Native American languages that are still kept alive by a small but driven community. The case of Washo is even more unique in relation to other isolates listed above in that in the pre-colonial times, it is wagered that there might have been no more than 1,500 speakers. Depending on the sources and dates, the number of estimated speakers varies between 10 (by 1998), 20 (by 2008, according to Victor Golla in California Indian Languages [2011]) or “no more than 50″ (by 2011). If all these numbers are true, it is because of the efforts of revitalisation that the Washo people have engaged in, such as the creation of an English-Washo dictionary or the setting up classes to teach Washo through the Cultural Resource Department for the Washoe Tribe. 

Salinan (turquoise) : Salinan is also one of these languages that invite controversy regarding its family ties to other languages. Although there aren’t definitive and convincing evidence of such a classification, Salinan has been put in the Hokan family, following Edward Sapir’s 1929 completed Native American languages classification for the Encyclopædia Britannica. Salinan has two fairly well-documented dialects, Antoniaño and Migueleño. Linguists have written quite a fair amount with speakers of both dialects, as well as collected corpora of important words and phrases. Salinan was spoken by the Salinan people around the central coast of California. Its last speaker disappeared in 1958.

Esselen (gold) : Esselen is unfortunately one of the worst attested Native American languages. It may have gone extinct somewhere during the 19th century. So little is known that most information that linguists have are actually second-hand. It was passed onto linguists by non-Esselen speakers who had heard earlier in their life people speak Esselen. Linguists and anthropologists like Kroeber and Harrington could only collect words from Native Americans who had heard and remembered certain words. Other than that, there are very few written traces, apart from bilingual Spanish-Esselen catechism.  

Seri (dark purple) : Seri is an isolate language spoken in the Mexican State of Sorano. UNESCO has listed Seri as a vulnerable language. The number of Seri speakers has not been pinned down precisely but it is estimated that in 2000, there were around 716 speakers, 900 in 2007 but 760 by 2010. Numbers are not really accurate since three different sources give different statistics but it is crystal-clear that Seri is indeed in danger. Edwar Sapir, who apparently did not like dangling languages, put it in his hypothetical family of Yuman. However, there again, there is a lack of conclusive evidence to credibly range it in this grouping. Therefore, it is still considered an isolate. Remarkably, Seri is a North-American language isolate with one of the smallest phonological inventories. 

Purépecha (grey) : This language is still spoken today by the Purépecha people of Michoacán, in southern Mexico. This isolate faces a paradoxical situation, according to Villavicencio Zarza 2006 (page 53); while the numbers of speakers have been steadily growing over the past few decades (going from 58,000 in 1960 to 96,000 in 1990, 120,000 in 2000 and 125,000 by 2010), this took place in a backdrop of increasing numbers of bilingual speakers (at the expense of native speakers whose total has been lowered by 50% since the 1970s) and increasing numbers of people in the region. This resulted in the lowering of the percentage of Native speakers overall. Because of this paradox, Purépecha is considered an endangered language. Although linguists like Campbell (1987) put Purépecha in the Chibchan language family, it is speculated that the language is an isolate. 

Cuitlatec (dark green) : It is believed that Cuitlatec saw its last speaker, Juan Can, die in the 1960s. He, liked his ancestors, resided in the Mexican State of Guererro. Like Prépecha, Cuitlatec possesses a small sound inventory, especially regarding the consonants. There has been debate to classify the language as belonging to the Chibchan or Uto-Aztecan family but no substantial evidence has been put forward to back up this claim. 

Huave (pink) : Huave is spoken in only four villages in the Isthmus of Tehuantepec in Oaxaca. The language was still used by around 18,000 people in 2010. While Huave is endangered because it is mostly confined to the social life of one of the four villages, it is reported that a new interest in Huave was born out of the Huavean teenagers’ will to hide their conversation from adults. By speaking Huave, which has fallen out of use for many adults, teenagers are able to protect the contents of their conversation from undesired hearers. Youngsters seek and appreciate the privacy that Huave gives them. Huave is experiencing a resurgence thanks to texts but also thanks to local initiatives such as Radio Ikoots, which broadcasts to the local communities in Huave.