Kusunda, a language like no other?

May 25, 2012 by

Kusunda is a dying isolate language. Gyani Maiya Sen, a 75-year-old woman from western Nepal, is its last known speaker. There are some 100-160 people in the Kusunda tribe, and some of them know a few words of the language, but nobody else speaks the language with any degree of fluency. Its isolate status means that Kusunda is not related to any language of the world, though there were several attempts to link the language to an established language family. B. K. Rana (2002) maintains that Kusunda is a Tibeto-Burman language as it has been traditionally classified. Others have linked it to Munda (Watters 2005); Yeniseian (Gurov 1989); Burushaski and Caucasian (Reinhard and Toba 1970; this would be a variant of Gurov’s proposal if Sino-Caucasian is accepted); the Nihali isolate in central India (Fleming 1996, Whitehouse 1997); and again with Nihali, as part of the Indo-Pacific hypothesis (Whitehouse et al. 2004). However, Kusunda does not exhibit much affinity to any languages spoken in the area.

This prompted some journalists to claim that Kusunda “looks nothing like any other language”. But is Kusunda really outside the realm of what has been so far identified as “universal grammar”? Not really. Like all other languages, it distinguishes consonants and vowels; makes use of nouns and verbs; and constructs larger lingusitic forms out of smaller ones (words out of morphemes, sentences out of words). Here are five specific ways in which Kusunda fits into the cross-linguistic typology.

  1. Vowel inventory. Kusunda has three vowel phonemes: /e/, /a/, and /o/. Each phoneme has two realizations (higher and lower). This vowel inventory qualifies as small. Ninety-three languages in the 564-language WALS sample have small vowel phoneme inventories (2-4 vowels, mostly 3), including Akkhaz, Malagasy, and Yupik.
  2. Case alignment. Kusunda has a nominative-accusative case system, much like Latin, Russian, or even Old English. This means that in Kusunda—unlike in many neighboring languages, which are ergative-absolutive—the subjects of transitive and intransitive are marked the same way (nominative), while objects are marked differently (accusative).
  3. Locative cases. In addition to nominative (for subjects), accusative (for objects), and genitive (for possessors), Kusunda has three locative cases: -lage ‘for’, -əna ‘from’, -ga, -gə ‘at, in’, as well as a commitative case (‘together with’).
  4. Tense-aspect and negation affixes. Kusunda marks tense/aspect distinctions by suffixes on the verb, as do 668 languages in the 1132-language WALS sample. Options employed by other languages include tense/aspect prefixes, tense/aspect tones, mixed types, or not marking tense/aspect distinctions at all. Negation is likewise marked by an affix (more specifically, a suffix) in Kusunda. Other languages that mark negation by affixes including Nepali, Lezgian, Xhosa, Sakha (Yakut), Uzbek, and Ubykh.
  5. Word order. Kusunda has the SOV order, the most commonly found order cross-linguistically: 565 languages in the 1377-language WALS sample have this order, including Basque, Japanese, Hindi, and Enets. Moreover, like most SOV languages, Kusunda has postpositions rather than prepositions: of the 408 SOV languages for which the order of adposition and noun phrase is known (see WALS page), 374 languages (92%) have postpositions.

Thus, though Kusunda may be quite different from and unrelated to its neighbors, it is hardly “unlike any other language”, as journalistic reports proclaim it to be.

Subscribe For Updates

We would love to have you back on Languages Of The World in the future. If you would like to receive updates of our newest posts, feel free to do so using any of your favorite methods below:

  • CC

    I could be wrong, but I’d also imagine that a good amount of borrowing has happened between Kusunda and its neighbors, which leads to the interesting problem of “it looks similar, but is not similar in the right way.”

    • This a great point. I am not sure how much lexical borrowing happened, but as far as I could tell, not much grammatical borrowing did. It’s an interesting issue to investigate further, though…

  • I left this comment at the io9 site you linked to:

    Language isolates are not known to be related to any other language, but that doesn’t mean they are unlike any other language. You probably can’t trace the family relationship between yourself and some random person in Africa (unless you were born in Africa!), but we all belong to the same species and are fundamentally very much alike. Similarly, all languages probably have a common origin (though the evidence isn’t there) and have a great many general features in common.The isolate status of Korean is not accepted by all scientists either. Some believe it is related to Japanese; others, that it is related to a whole family of languages from Turkish through Mongolian as well as Japanese. It’s all about how you evaluate the evidence, whether you think it’s convincing evidence of a common origin or not.

    • That’s a good way to put it, John!

      I’ve written on the Altaic hypothesis and Korean elsewhere on this blog…

  • BH

    Is it really fair to blame a journalist for not being able to distinguish between typological profile and genetic inheritance

    • If a journalists undertakes writing about language, is it really unfair to expect them to know something of the subject matter they write on?

  • Ivan Derzhanski

    Actually, Watters 2006 (_Notes on Kusunda Grammar_) says that the contrast between higher and lower vowels is phonemic in some words, though in others they are in free variation.  Kind of like Bulgarian, which has six vowel phonemes but a lot of neutralisation (esp. in some dialects).  Doesn’t seem very exotic to me.

    • Of course, you don’t find something that’s commonplace in Bulgarian to be exotic! But it doesn’t mean that it’s not exotic from the crosslinguistic perspective.

      • Ivan Derzhanski

        That was tongue in cheek, of course: the similarity I noted is superficial.  In fact Watters doesn’t make it quite clear (and, it seems, never quite found out) just how neutralisation of height works in Kusunda.

  • RuTHaL

    So for no. 3 it’s the Ablative case from Latin which was used to indicate direction.