[Lingtyp] Traditional view of language and grammar in indigenous societies

[randylapolla]

Hi Adam,

> Thanks for providing this very interesting description of your experience working with the non-literate elderly Qiang historian. Did the (presumably younger) Qiang speakers you worked with later also insist on talking about the content of narratives/histories rather than the code? From what you described, it sounds like there was a mismatch between what you wanted to learn about (the language) and what the historian wanted to teach you (the history of his people). If he felt that his pedagogical goal was to teach you the histories themselves, then it's not surprising that he rejected the hypothetical sentences you constructed.
>
> What I'm less comfortable with is the suggestion that this individual teacher's preference for the content of the histories rather than the linguistic code itself tells us anything deep about the distinction between oral and literate cultures. (But I am open to being persuaded that I am wrong about this.)

The younger Qiang people have all learned Chinese (and come to think in Chinese categories), and in the process learn about Chinese structures, but this can also be a problem. Our training can be a blinder to understanding the language. In the past at least, before the spread of typology, Indian scholars focused on paradigms and participles, but not tones (not describe them), even if they found the latter in the language. In China it was the opposite: they learned Chinese grammar, which has no paradigms or participles, but does have tones, and so they looked for tones in all languages, but didn't find participles or paradigms even when they existed in the language because they were unfamiliar to them. Back in the 1990’s I gave a talk on middle-voice marking in Tibeto-Burman languages at the Academy of Social Sciences Institute for Ethnic Studies. They had never analysed any of the languages as having such marking because that category does not appear in Chinese and so they never heard of it. They thought I had just made up the concept of middle voice. I said, “No, the Greeks knew about it a very long time ago.” Another example was when in 2005 I gave a talk at the Central University for Minority Studies on the importance of having some typological knowledge when describing languages, and afterward a woman came up and said to me she worked on Uyghur, and was having trouble finding a 补语（’complement'，but not what we normally call a complement) in the language. I asked her how she was defining the term, and she said “As in Chinese”. I said “In Chinese, it is defined as something that comes after the verb, but Uyghur is a verb final language, so there is nothing following the verb. If Uyghur manifests such a category, it will not be the same as in Chinese, but you can look for functional equivalents.”

> My experience working with the speakers of Tupari, a Tupian language spoken in the Brazilian state of Rondonia, has taught me that even non-literate speakers can have quite impressive metalinguistic awareness about the structure of their language. The Tupari pronominal inventory doesn't encode a formal-informal distinction, but the language has lots of number-sensitive verbal suppletion and agreement; and traditional speakers always make sure to talk to/about their in-laws as duals rather than singulars, with many morphological consequences seen in the suppletive verbal roots and in the agreement suffixes. The reason I mention this is because many speakers I have worked with, including speakers who aren't literate in Tupari or in Portuguese, know how to talk about this aspect of their language: they explain that one must speak about one's in-laws as if each individual in-law were two people, not just one. There's even a specific term for this kind of respectful speech (kiarowak ara). I'm not claiming that non-literate Tupari speakers have metalinguistic awareness about all of their language's grammatical properties, but in this area at least they can and do talk about structure.

Having awareness of the “proper” (conventionalised) way of speaking one’s language is not the same thing as what I was talking about. The old man I worked with was well-known as an excellent speaker, and he often presided over weddings and other get-togethers where he made speeches, but unless they make a special effort to understand the structure, such speakers won’t understand them. I even knew a Chinese teacher at Stanford University who was a famous writer (in Chinese), but when asked any question about structure, he would just say “That is the way we say it.” He could not analyse the structures. But some people do put special effort in trying to understand their language. The Panini example is one where they were trying to understand older texts, and so outlined the patterns found. The same was true for Chinese linguistics, where the study of phonology was about trying to understand older texts that didn’t seem to make sense if the modern pronunciations and meanings were used.
But I agree with your point about education being a key factor in how we think about language. The idea that there are ungrammatical sentences is entirely a pedagogical notion, and should not be used in language description. There is no such thing as ungrammatical sentences in natural linguistic data. And asking speakers if you can say some sentence means you are testing their ability to create a context in which the sentence makes sense rather than really testing grammaticality.

All the best,
Randy

> On 4 Jul 2025, at 8:21 AM, Adam Singerman via Lingtyp <lingtyp at listserv.linguistlist.org> wrote:
>
> Dear Randy,
>
> Thanks for providing this very interesting description of your experience working with the non-literate elderly Qiang historian. Did the (presumably younger) Qiang speakers you worked with later also insist on talking about the content of narratives/histories rather than the code? From what you described, it sounds like there was a mismatch between what you wanted to learn about (the language) and what the historian wanted to teach you (the history of his people). If he felt that his pedagogical goal was to teach you the histories themselves, then it's not surprising that he rejected the hypothetical sentences you constructed.
>
> What I'm less comfortable with is the suggestion that this individual teacher's preference for the content of the histories rather than the linguistic code itself tells us anything deep about the distinction between oral and literate cultures. (But I am open to being persuaded that I am wrong about this.)
>
> My experience working with the speakers of Tupari, a Tupian language spoken in the Brazilian state of Rondonia, has taught me that even non-literate speakers can have quite impressive metalinguistic awareness about the structure of their language. The Tupari pronominal inventory doesn't encode a formal-informal distinction, but the language has lots of number-sensitive verbal suppletion and agreement; and traditional speakers always make sure to talk to/about their in-laws as duals rather than singulars, with many morphological consequences seen in the suppletive verbal roots and in the agreement suffixes. The reason I mention this is because many speakers I have worked with, including speakers who aren't literate in Tupari or in Portuguese, know how to talk about this aspect of their language: they explain that one must speak about one's in-laws as if each individual in-law were two people, not just one. There's even a specific term for this kind of respectful speech (kiarowak ara). I'm not claiming that non-literate Tupari speakers have metalinguistic awareness about all of their language's grammatical properties, but in this area at least they can and do talk about structure.
>
> I would like to bring up one other potential counterexample to the idea that a society without writing is automatically going to be one in which "speakers are not familiar with the concept of analyzing it as a thing in the world separate from its daily use" (quoting Chris Donley). Didn't Panini (https://w.wiki/EP5f) compose and teach his treatises on Sanskrit grammar orally, centuries before the language came to be written down? Perhaps the important distinction isn't oral versus literate cultures but rather the question of whether there are schools or educational institutions in which language is taught as a subject in its own right. We're used to schools emphasizing written language, but as the case of Panini shows, language can be taught without writing, too.
>
> If I'm wrong about Panini, Sanskrit, etc, please let me know — I'm just regurgitating what I'd read in the literature on non-Western grammatical traditions.
>
> All the best from Syracuse,
> Adam
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20250706/881d3852/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1424 bytes
Desc: not available
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20250706/881d3852/attachment.bin>