Of mynas and miners, bells and whistles

« previous post | next post »

Over at Spicks & Specks, Greg Pringle has a virtuoso post on "The Bell Miner:  How orthography and ornithology catalysed a new folk etymology" (8/9/15).  It's about an Australian honeyeating bird — Manorina melanophrys — that used to be called the Bellbird, but was renamed Bell Miner through association with the South Asian bird called in Hindi the mainā मैना (" starling").

As Greg explains:

For a long time there was no single way of spelling this word. People wrote it as they felt fit, resulting in a variety of 19th-century spellings like “minah”, “minor”, “minar”, and “miner”. The principle behind these spellings was that of representing the pronunciation roughly as it was heard.

If these spellings appear peculiar today, it's because number of them reflect a sound change that took place in southern British English during the 18th century. This involved the disappearance of /r/ from any environment in which it wasn't followed by a vowel. Sounds spelt ‘er’ , ‘ur’ , ‘ar’ , ‘or’, and ‘ir’ in the middle of words turned into long vowels, and sounds written ‘-er’, ‘-ar’, and ‘-or’ at the end of words usually turned into the vague vowel known as schwa. Varieties of English that adopted this change in pronunciation are now called “non-rhotic”; those that preserve the /r/ are called “rhotic”.

This sound change led to a reinterpretation of the function of ‘r’ in spelling. In the middle of words it was seen as a sign of lengthening, and at the end of words it was interpreted as schwa. This is why many of the 19th-century spelling pronunciations of “mynah” end with an unpronounced ‘r’.

Interestingly, this ornithological name kept the spelling “miner”, despite the fact that both “minah” and “mynah” were current during the 19th century and were used for both the miners and the bellbird. The spelling “miner” would have encountered little resistance in Australia, which was a fortress of non-rhoticism. Until well into the 20th century, Australians continued to use ‘r’ to represent impressionistic colloquial pronunciations, without necessarily involving any actual /r/ sound. When authors wanted to represent the way people actually spoke, it was accepted practice to write “to go” as “ter go” or “talked and laughed” as "torked an' larft" [5]. This can still be found up to the present time.

But there are inherent problems with this “pronunciation-based” spelling. The first is that, while it may have looked fine to Englishmen or Australians, it makes less sense to speakers of other varieties of English where the letter ‘r’ is actually pronounced. Such “rhotic” varieties are spoken in places like Scotland, Ireland, regional areas of England, and much of the United States and Canada. For such speakers, there is an audible and meaningful difference in pronunciation between “myna” /ˈmaɪnə/ and “miner” /ˈmaɪnər/. The first could only be applied to the bird, the second, in ordinary pronunciation, would normally be used only in words like “minor” and “miner”.

The problem is not confined to rhotic speakers. Although it is generally understood that alphabetic systems are based on the equation of letters to sounds (often observed in the breach in English spelling), spellings have a propensity to become fixed, and in their fixity come to be regarded as invariant manifestations of words rather than direct representations of sound. Even in non-rhotic English, the old fluidity of spelling by which Hindi मैना mainā was spelt “minah”, "minor”, “minar”, or “miner”, eventually settled on “mynah”, and finally “myna”. In the world of fixed spellings, “miner” is tightly linked to the concept of “mining”, “minor” is tightly linked to the concept of “lesser” or “younger”, and “myna” or “mynah” is tightly linked to a type of bird. Once spellings are fixed, the older practice of using “miner” to represent the pronunciation /ˈmaɪnə/ in an ad hoc fashion is easily overlooked or forgotten. This is what has happened to the spelling “miner” for these honeyeaters. It is now difficult to see the word as related to anything other than mining, opening the way for “miner” to be interpreted in a completely different manner from what was originally intended.

To be fair, of course, the connection with the old meaning is not yet completely dead. There is still some awareness of the roots of “miner”, and some Internet sites even give “Bell Mynah” as an alternative to Bell Miner. But in general the spelling “miner” serves to distract contemporary English speakers from the intended meaning.

Greg's argumentation is backed up by OED (mynan.; minern.3) and Hobson-Jobson, which suggest that miner was originally a variant of myna(h), referring to the S/SE Asian bird, appearing that way in Raffles' History of Java (1817). (We've talked about similar non-rhotic pronunciation spellings on Language Log many times, e.g. Burma, Myanmar, and Eeyore.) Before the name miner was applied to what was formerly called the Bellbird, it had already been attached to another Australian honeyeater, Manorina melanocephala.

I'll end this post on the same note as Greg does his, by pointing out how the Chinese have translated the "miner" of "Bell Miner" as kuàngniǎo 矿鸟 ("mining bird" or "mineral bird"), whereas the myna has been known in China since the Tang period (618-907) as the bāgē(r) 八哥(儿).

[Thanks to Ben Zimmer]


  1. Dave said,

    August 16, 2015 @ 11:46 am

    Language change and Australian birds; what a wonderful post!

    The bell miner isn't the only miner (bird) in Oz. There's also the noisy miner, the yellow throated miner, and (rare) the black eared miner; all are honeyeaters. I suppose the word "miner" only got adopted once, though? (There's also the common myna, a quite different bird, and not native, still with a more clearly birdy spelling.)

  2. Dave said,

    August 16, 2015 @ 12:01 pm

    Let this be a lesson to you, kids: read *before* commenting! That way, you don't fill space with what's already covered in the article. Whoops!

  3. Margaret Dean said,

    August 16, 2015 @ 3:02 pm

    Reminds me of a clueless American kid (me) putting a lot of incorrect "r" sounds into the names of Kipling's animals in The Jungle Book, because the pronunciation guide in the book was written with non-rhotic speakers in mind.

  4. Sky Onosson said,

    August 16, 2015 @ 6:50 pm

    As far as I'm aware, the vowel in the first syllable of Hindi 'mainā' is pronounced like the e in 'met', i.e. it's not a diphthong as in English 'mynah'. Does anyone know if that was historically different in Hindi at the time of borrowing into English, or did English speakers adopt a 'spelling pronunciation' based on the transliteration?

  5. Matt McIrvin said,

    August 16, 2015 @ 7:39 pm

    J. K. Rowling's dialogue for Hagrid in the Harry Potter books involves a lot of eye dialect with "er" endings in it, and I have to consciously remind myself of what that means. And of course there's the literary interjection "er" itself, which many Americans don't even interpret as "uh".

  6. Neil Dolinger said,

    August 16, 2015 @ 9:22 pm

    Of course, Hagrid is supposedly Scottish, so he would be pronouncing his r's.

  7. Neil Dolinger said,

    August 16, 2015 @ 9:26 pm

    I'm wondering about the origin of 八哥. It sounds like a borrowing from another language.

  8. Bathrobe said,

    August 16, 2015 @ 10:11 pm

    It's supposedly from the markings on the wings of Acridotheres cristatellus (Crested Myna), which look like the character 八 (eight) when in flight. See here, for instance.

  9. Margaret Dean said,

    August 16, 2015 @ 10:55 pm

    @Neil Dolinger, IIRC Rowling says that Hagrid's accent is meant to be West Country rather than Scottish.

  10. Victor Mair said,

    August 16, 2015 @ 11:23 pm

    There are a lot of websites that say the bāgē 八哥 was originally called qúyù 鸲鹆, but because of the need for taboo avoidance of the sound of the name of the Southern Tang emperor, Lǐ Yù 李煜 (937-978), a new name had to be chosen, and that is supposedly why they picked "eight brother" (allegedly because of the markings on the wings, as Bathrobe has pointed out).


  11. Moonfriend said,

    August 17, 2015 @ 7:58 am

    The thing with using 'er' to mean 'uh' is that there is no other way with an Australian accent, short of writing something like /ö/, to represent that sound. If it's intended for a local audience, it doesn't matter because everyone knows how to say it. That's why, to imitate a North American accent, people will often write 'lerv' for 'love'. In cases of hypercorrection, people will even pronounce 'lerv' rhotically.

  12. carlageek said,

    August 17, 2015 @ 1:41 pm

    @Sky Onosson – the pronunciation of the vowel ऐ in मैना varies from region to region. There are variants where it is pronounced more like AmE "eh", but there are also variants where it comes very close to the sound represented in the spelling "Mynah". And it is indeed often analyzed as a diphthong. Here is one source that treats both ऐ and औ as diphthongs, and notes that their pronunciation varies regionally.


  13. Neil Dolinger said,

    August 17, 2015 @ 2:40 pm

    @ Margaret Dean, you are correct, as I discovered after I wrote my comment, THEN checked my facts!

  14. Sky Onosson said,

    August 17, 2015 @ 3:32 pm

    @carlageek Thank you! I was going entirely off long-ago Hindi lessons.

  15. Jerry Friedman said,

    August 17, 2015 @ 4:00 pm

    Word of the day: Bell Miners live on bell lerps.

    For folk etymology in bird names, there are widowbirds and whydahs, both in Africa. Both have (in the most familiar species) long black tails like the plumes widows used to wear. The OED says, "Etymology: Name of a town in Dahomey, West Africa. Whidah bird is an alteration of WIDOW-BIRD n., q.v., due to association with this as one of the habitats of these birds."

    (Whydah, Dahomey, is now Ouidah, Benin.)

    The OED pronounces the "y" as in "widow", but when I saw one of these birds in Kenya, our guide pronounced it as in "why".

  16. Brett said,

    August 17, 2015 @ 8:59 pm

    @Moonfriend: How then is "uh" pronounced?

  17. Moonfriend said,

    August 18, 2015 @ 7:37 am

    @Brett: "uh" would be something like /aː/, as in the final sound in "bar".

  18. Piyush said,

    August 19, 2015 @ 5:35 pm

    @Sky Onosson, carlagreek: I am a native speaker of Hindi, and I have never come across the pronunciation of ऐ that Sky Onosson alludes to. My own pronunciation matches the one described in the wikibooks article, and there are certainly some variations but none of the ones I have heard come close to the shortness of the vowel sound in "met".

  19. Piyush said,

    August 19, 2015 @ 5:37 pm

    @Sky Onosson, carlagreek: I am a native speaker of Hindi, and I have never come across the pronunciation of ऐ that Sky Onosson alludes to. My own pronunciation matches the one described in the wikibooks article, and there are certainly some regional variations but none of the ones I have heard come close to the shortness of the vowel sound in "met".

  20. Sky Onosson said,

    August 19, 2015 @ 5:41 pm

    @Piyuysh: thank you, very glad to be corrected! Just to be clear though, I wasn't referencing the vowel in 'met' for it's shortness, but for its tongue position and monophthongal quality.

  21. carlageek said,

    August 21, 2015 @ 10:11 am

    Thanks for the input, Piyush. The Wikibooks article matches what I was taught by a Hindi teacher (who was also a native speaker) the range of pronunciations, and I've seen the diphthong analysis in textbooks myself, so I didn't have reason to question it.

    I wonder if what's going on is that the diphthong pronunciations were once more attested but have now fallen out of use in nearly all dialects. Here is a link to the front matter of Platts, a leading Hindi/Urdu-English dictionary of the late 19th century, that also analyzes ऐ as "ai" and says it can be pronounced "like our 'i' in 'wise', or the 'ai' in German 'Kaiser'." (Page vi)


    So I raise this not to discount Piyush's experience as a native speaker, which is clearly superior as to how the language is pronounced today – but to try to get to the bottom of what is going on to connect मैना to "mynah," especially in the 19th century when the English spelling seems to have crystallized.

  22. Piyush said,

    August 21, 2015 @ 9:53 pm

    @carlageek: The diphthong pronunciation of ऐ that I have heard is /əi/ (roughly अइ). I cannot think of any English words that have this sound. Wikipedia suggests the sound of "i" in pile, but for me that is more like /ai/ (roughly, आइ), and hence completely irreconcilable with the diphthong pronunciation of ऐ. The same goes for Platt's suggestion (i in "wise", which too I pronounce as /ai/, and wiktionary seems to agree).

    The /əi/ pronunciation is common in at least a few Hindi continuum languages, some South Indian languages, and even Sanskrit. In contrast the /ai/ pronunciation suggested in Platts reminds me of how stereotypical British characters in Bollywood movies speak (or try to speak) Hindi.

    I can think of two reasons why Platts includes "wise" as a pronunciation guide for ऐ and why mynah is spelled the way it is: The first, which I think is far less likely, is that the British English pronunciation of the vowel in "wise" has changed in the last 100 or so years. The second, which I would put my bets on, is that there really are no common English words, certainly none that I can think of in the dialect I speak, which have the /əi/ sound, so /ai/ was used as the closest approximation.

RSS feed for comments on this post