Joel of Far Outliers has a post quoting the late University of Hawai‘i professor Donald M. Topping at length on linguists and endangered languages; Topping makes disturbing and important points:

Surprisingly, a major obstacle to the success of the Micronesian linguistics project is one that was unanticipated, and may be fairly assigned to the linguists themselves. That is the problems presented by the “new” orthographies. Mr. Leo Pugram, Coordinator for Curriculum and Instruction in Yap, made the following statement, “When the new orthography was established, it was a time for problems, confusion, and hatred for the new orthography. This still exists today on Yap.”
Obviously, the linguists left their mark: the “new orthography.” The complaint articulated so bluntly by Mr. Pugram was echoed by nearly every other Micronesian educator who attended the Guam conference…
Where then, are the linguists? Have they played a role? Do they now? Each of the three languages in question was described and lexified by nonnative linguists during the 1950s and early sixties. At the time of their work these linguists issued the call of alarm about the precarious status of the languages. Their calls, however, appeared to fall on deaf ears, for there was little response. It took the coming of another generation of young people who were not afforded the opportunity to learn their heritage language at home before the threat of total language loss became real.

Joel has a follow-up post quoting another University of Hawai‘i linguist, Kenneth L. Rehg, on the same subject:

…we set out to promote literacy in the Micronesian languages, but some of our efforts had just the opposite effect. Disputes over orthographies, unrealistic expectations concerning standards, an insufficient understanding of the literacy needs of these communities, and reliance on external funding all hindered progress toward that goal. Consequently, I have come to believe that if the linguistic community is serious about documenting and supporting the threatened languages of the world, we must move such endeavors into the mainstream of our discipline.

Incidentally, Joel recently made his 1,000th post: congratulations!


  1. Michael Farris says

    The difference between what linguists might want in a writing system and what natives are used to/want should never be underestimated. A good practical rule is that if a language has an orthography that’s even remotely usable and/or accepted by native speakers then linguists should forget their phonemic scruples and run with it.
    I like the approach used in the recent Muskogee/Mvskoke/Creek dictionary which gives prominence to the the fairly strange (but established and fairly standardized) conventional orthography and gives a more precise linguist oriented system (basically devised by Mary Haas) afterward.

  2. if a language has an orthography that’s even remotely usable and/or accepted by native speakers then linguists should forget their phonemic scruples and run with it
    I couldn’t agree more.

  3. Out of curiosity, does the Muskogee/Mvskoke/Creek orthography you describe use the English alphabet, or an indigenous one?

  4. michael farris says

    It uses the Roman alphabet but:
    – doesn’t indicate pitch (important, especially in some aspect distinctions) for example kērris represents both /ki:łłés/ (I’m learning [it]) and /kî:łłes/ (I know it) (nb. r is used for /ł/, a lateral fricative like welsh ll)
    – isn’t always consistent in distinguishing short and long vowels, especiall /o/ and /o:/ u always represents short /o/ but o can represent either short /o/ as in cuko /cokó/ ‘house’ or long /o:/ as in cokv /có:ka/ ‘book’.
    – has some counterintuitive letter choices, r as a lateral fricative, v for short /a/ and the values you’d expect for the letters e and i are reversed, that is e = /i/ and i = /e(y)/
    There’s some other stuff, but that gives you an idea. Still, all in all its quite usable and Creek as one of the five civilized tribes has one of the more extensive written legacies among NAMerican languages. If I could change anything about it, it would only be to add some optional circumflexes (falling pitch) for aspect marking. Most falling pitch doesn’t need to be indicated but I think would be a good idea for writers to be able to mark some.

  5. The requirements of linguists (or learners) and those of native speakers are often different. Native speakers do not require more than a hint as to the phonetic realisation of a word (look at Chinese) and may indeed find a close phonemic transcription of their language quite off-putting. If it accurately transcribes one dialect it will alienate the speakers of all the others. Imagine the endless disputes and namecalling that would be caused by any attempt to make the spelling of English more accurate phonologically.

  6. Andrew Dunbar says

    I found it interesting that you’ve chosen to use the Hawaiian spelling “Hawai‘i” rather than the English spelling “Hawaii” (or possibly “Hawai’i”). Usually you speak out against this sort of thing. There is no ‘okina in the English alphabet.
    I wonder if this might be the spelling that the University of Hawaii uses?
    (Less interesting is that you’ve used a punctuation mark rather than the true ʻokina which is a letter – you notice this when selecting with the mouse etc. Unfortunately the true ʻokina is not in most fonts so this is a common workaround.)

  7. I’m afraid all it reflects is my lazy inclination to copy-and-paste; if I’d typed it out, I would have spelled it Hawaii, unless I went to the trouble of checking the university site… *checks site*… yup, they do include the mark, so if I were feeling picky I might have included it anyway. But I would have used a simple apostrophe.

  8. OK, this is very confusing… Andrew Dunbar, the first “Hawai’i” in your comment should have a left single quotation mark in it, as in Joel’s post? And Hat, did you change the spelling in your post after this exchange? Because they’re all apostrophes now.

  9. The one in my post looks like a single open quote to me. I haven’t changed anything. But embrace your confusion! Chaos is good!

  10. Andrew Dunbar says

    the first “Hawai’i” in your comment should have a left single quotation mark in it, as in Joel’s post
    I’m using a different computer now and on both of them I see a left single open quote. Cutting and pasting from one page into a “find” box in the other matches. Maybe on your setup the fonts are causing differences?

  11. Linguists can’t win here. For every group who dislike linguists because they burdened a community with a new orthography there’s a group who complains that young people don’t speak the language probably because the linguists didn’t write it down correctly. I have had both these accusations levelled at me by different people in the same community.

  12. Sorry, pressed “post” too quickly. Michael, it’s not that simple. There may already be more than one orthography to choose from (in Arnhem Land there was the “Government orthography” and the “Missionary orthography”).
    We can go the way of practical orthgraphy first, then a phonemic guide of some sort, but this lengthens and complicates the entries, which can be an issue in producing learner’s materials.
    Underspecified orthographies create headaches for revitalisation materials too.

  13. The official spelling adopted by the University of Hawai‘i system is with the ‘okina, but only for the place name, not for English derivatives like Hawaiian or Hawaiiana. It’s a bit of a problem for handling search terms online, of course, so some people use the uglier keyboard `, while others resort to the less intrusive apostrophe, and others leave it out altogether. I don’t think the ‘okina is very common in business usage.
    There is a special Hawaiian font set used by a lot of Hawaiian language sites that remaps the ÿ into an ‘, but that really looks awful (Hawaiÿi) if you don’t have those fonts available for your browser. The same font set remaps the umlaut vowels as macron vowels.

  14. I like your approach to this matter. I would love to know more about it.

  15. John Emerson says

    God forgive me, Hannah, but I suspect you of veing a spambot.

  16. I agree — that’s why I deleted the URL from her effort. I’m not sure why I didn’t delete the whole comment as usual; maybe I was seduced by the unusual elegance, politeness, and just plain grammaticality of the comment.

  17. “A good practical rule is that if a language has an orthography that’s even remotely usable and/or accepted by native speakers then linguists should forget their phonemic scruples and run with it.”
    Simply not true for Chinese. John DeFrancis spent his entire career showing how the ortography accepted and use by native speakers in China is detrimental to widespread literacy, and there is legitimate research showing that changing to full romanisation, repugnant as it might be to those proud of their system, has tremendous benefits.

  18. Maybe on your setup the fonts are causing differences?
    This is really peculiar. I’m getting different results in different browsers. And it’s not a question of the font – I’m actually getting different characters in Konqueror and Mozilla. (It’s not Konqueror universally replacing U+2018 with U+0027, either – it shows up fine in Joel’s post, and in his comment above.)
    ‘ U+0027 APOSTROPHE
    (Your ʻokina (U+02BB) came through fine, although it doesn’t display in the font.)

  19. The majority of Languagehat readers, who aren’t interested in character set issues, should probably skip this comment.
    Joel’s post is in UTF-8. It includes left and right single and double quotation marks (‘ ’ “ ”). Hat’s site is in ISO 8859-1 (Latin 1). Latin 1 doesn’t contain characters for any of these quotation marks. In order to use them in ISO 8859-1 HTML, they ought to be represented by HTML entities. That happens automatically when I paste them into this comment text box.
    However, the quotation marks from Joel’s post in Hat’s post above are not represented by HTML entities. Nor can they be represented by their proper Unicode characters, which don’t map to anything in ISO 8859-1. Instead, for some reason, they appear as codepoints 91, 92, 93 and 94. These are either unassigned or control characters (what with the different versions of standards, it’s a little hazy). In any case, they’re not supposed to be used like this – this page is not valid ISO 8859-1. I don’t know why this substitution has occurred, but apparently it’s common enough that the browsers turn them into quote marks when they encounter them. Unfortunately, while Mozilla (and probably IE) map them to the original left and right forms, Konqueror uses the neutral vertical forms, ‘ and ” (which are, after all, present in Latin 1), thus turning Andrew’s comment above into nonsense.

  20. michael farris says

    Claire, yeah, there’s about a thousand factors to be taken into consideration. But I stand by my statement, a sub-par orthography from a linguist’s point of view that is widely accepted by native users is better than a linguistically perfect one that will alienate native users (in the link in question the decision to always mark glottal stops with q seems more or less insane and I’m not surprised the orthography wasn’t accepted). And yes, I realize it can be hard to actually figure out what native users will find acceptable (and that people might tell the linguist one thing and actually do something else).
    When there’s no widely accepted orthography and/or competing orthographies then the linguist needs to proceed with caution instead of rushing ahead with a system that makes it easy for the linguist. In terms of trying to help native users create written traditions for a vital language, the linguist’s concerns about phonemic accuracy should come last. In preservation/revitalization efforts, again the linguist’s concerns are not paramount and parallel practical / linguist orthographies may be the way to go.

  21. michael farris says

    “the ortography accepted and use by native speakers in China is detrimental to widespread literacy … changing to full romanisation, repugnant as it might be to those proud of their system, has tremendous benefits.”
    I kind of agree in the abstract, but …
    – as long as most Chinese users prefer hanzi then that’s what they’ve got and non users of the language (like me, and I presume you) don’t really have any say
    – romanization is a process not an event, Vietnamese romanization took generations and was a grass roots movement with a long period of mixed usage (some magazines and books in nom and some in quoc ngu)
    – the best way to promote script reform (or literacy in the first place) is through _authors_ that is train and produce romanized authors and let them produce and promote romanized materials (and expect it to take a while)
    – I’ve noticed a news item or two from the pinyin site about legal restrictions to romanized chinese words/names/signs if there wasn’t some attraction there then there wouldn’t be legal restrictions.

  22. Michael, I think we’re basically in agreement. I was reacting to the absoluteness of your first post. Perhaps we are also looking at this from different angles, from the point of view of producers of materials versus consumers of them?
    For example, someone confronted with a word that’s not in the dictionary doesn’t necessarily know if it’s just a variant spelling or a word that’s not there. If there are more than a few of them it quickly turns people off looking things up. In my experience fluent speakers with reasonable literacy don’t care that much about spelling in texts, but then again most of the people I work with are either not fluent speakers or not very literate.
    I’m not particularly prescriptivist in general (my fieldwork motto is “whatever works”) but “spelling doesn’t matter” as a general rule is not one I’ve found to be particularly effective, particularly in lexicography.

  23. I have some ideas about where a “predictive science of language vitality” might go. The thing I’ve noticed watching my parent’s fist language go off to extinction is how they characterize it as “things don’t happen in Plautdietsch anymore”. As soon a day-to-day life can only be discussed with other speakers of the same language by relying on their knowledge of some shared second language, there is real endangerment. I suspect that’s what is happening in Polynesia.
    As for writing standards – I think that where there is a script people will identify with, and it isn’t harder to master than a second language, it’s probably better to go with trying to standardize it than creating a new linguistically better system. But I think the principal issue is identification rather than entrenched users. People prefer a writing system that they can identify as their own. Part of the problem with linguistically sound writing is the notion that some linguist thought it up.
    China should have a standardized syllabic writing system indigenous to it rather than characters. (In fact, it already has half of such a system in the fanqie scheme used in 19th century dictionaries.) The Cree should use a standardized form of syllabics rather than a Roman scheme foisted on them by linguists. There are usually ways to make existing but inconsistent schemes at least work somewhat better without abandoning them entirely.

  24. pick up the cards When handling the cards in a handheld game, here are a few craps dice game Christ its a miracle I was even born! (mutters under his breath .

  25. Internet Archive links to Topping’s paper and to Rehg’s followup. Both papers are excellent and non-technical (at least by LH standards).

Speak Your Mind