The Wall Street Journal reported today on the findings of a 2008 linguistic expedition to Arunachal Pradesh, the most northeastern state of India. In a mountainous region already populated with a plethora of other spoken tongues, researches claim to have linguistic evidence of yet a wholly new language called Koro. If true, it’s an astonishing discovery — and bittersweet. Like so many other languages of the world, Koro is likely endangered; the younger members of this speech community are learning more standard languages in school (Hindi and English) and are not speaking Koro in their villages. The Journal article states there are about 800 living speakers of Koro.
Koro has been identified as a member of the Tibeto-Burman subgroup of the Sino-Tibetan language family. It does not exist in a written form, but presumably the linguists who have made the recordings of Koro speakers have transcribed these. The linguistic details of Koro are to appear in an upcoming volume of the journal Indian Linguistics. There will also be an online dictionary of the language available at some point. If you go to the WSJ article (the link above) there are several audio recordings of basic sentences in Koro — it’s interesting to hear them. The recordings and research were conducted as part of National Geographic’s Enduring Voices project.
It has been estimated that within this century as many as half the 6,909 known human languages will become extinct; many of them exist in isolated, small speech communities. The topics of language birth and language death will be the subject of a future post.
Word Travels 
Thanks for this, Ann. I looked at the WSJ story you linked to, and found it very interesting and well-presented. One thing it made me wonder about is what constitutes a language. The article says “the local Koro speakers themselves didn’t consider theirs a separate language, even though it is as distinct from those spoken by other villagers as English is from Russian,” and I wondered why not…
My second question is the statement by Gregory Anderson, who runs the project, that languages like Koro “construe reality in very different ways. They uniquely code knowledge of the natural world in ways that cannot be translated into a major language.” Why can’t they be? What does that sentence mean? If a Koro speaker sees a mountain lake or a rainbow or bare trees against a snowy field, isn’t his experience the same as ours would be?
Hi BBear,
There’s an aphorism among linguists that a language is just a dialect with an army. 🙂 Some truth in that. As you read in the WSJ article, that region of India has about 120 known languages, many spoken by small communities. I’m not an expert on Sino-Tibetan or Sino-Burmese, but hopefully the data collected from this speech community, once published, can be assessed by other linguists and contrasted to other known languages. The scholars reporting on Koro have already done their own analysis and based on it, concluded that this is a separate language. Perhaps there will be issues with their claims as data is made available, but perhaps not.
People self-reporting on their language can be tricky — there are cases of speech communities claiming to -not- understand the speakers in the next village, even though by linguistic criteria, there should be extensive mutual intelligibility. In that instance, there was social/political animosity and an emotional desire to distance themselves from the others, perhaps cognitively they could actually understand a lot of the other ‘dialect’, but refused to do so in some sense. There is a famous dialect continuum in Alsace where village by village the spoken dialect takes on more German or French attributes, depending on the proximity to the two countries. So your question of ‘what is a language’ is a good one, and without a simple answer.
It seems Gregory Anderson was referring to a strong version of the Sapir-Whorf Hypothesis, which conjectures that the categories of thought are determined by the grammatical structures and semantic partitions of a given language. Personally, I believe it’s a more malleable relationship, and we can learn other languages and we can all understand and reason about mathematics and discover things about the natural world. But, it is certainly an eye-opener for anyone who looks at the grammars of Amerindian languages or Asian languages, for example, after only ever previously being exposed to IndoEuropean, especially western European languages. Some of the deepest metaphors about time can be different, e.g., in English we think of time as ‘ahead of us’, we ‘look forward’ to things, and so on – you can think of many more of these. I -believe- it’s Navajo (maybe Apache) that has a different set of metaphors in which the speaker is facing toward the past and is not looking ahead at anything in the future, flying blind so to speak. So the temporal expressions get packaged differently in regards to speaker, prepositions, timeline. However, does this mean a Navajo and an English speaker cannot map their own systems to the other one, and both walk around in the same world? I don’t think so. Cognitive anthropologists at Stanford conducted famous ‘color wheel’ studies, where they showed that if a language lexicalizes some portion of reality, the speakers of that language have better recall when presented with that lexicalized data, and worse recall if they don’t. One case tested Mayan Indians in the Yucatan; apparently the Mayan languages don’t have separate words for ‘green’ and ‘blue’. When subjects were presented with specific objective swatches along the color spectrum (via a machine called the color wheel) they were not as good at recognizing specific swatches of blue and green that they were shown earlier. Speakers whose languages had separate words for these colors had better recall for these types of swatches. Probably speech communities with many words for snow are able to distinguish corresponding distinctions in the environment than communities that have only one word for frozen water. But vocabularies can expand or contract in languages, so these effects may be marginal rather than central. That’s my bias anyway. 🙂