Hacker News Reader

Comment by prvc | Hacker News Reader

One thing that has puzzled me about complaints about English speakers' pronunciation of transliterated Chinese words (the only topic of TFA), is why the speaker is blamed rather than the transliteration. If the official transliteration into the Roman alphabet leads to less accuracy in pronunciation than some other (for example see the second vs third columns in the table in TFA), then would it not make better sense for the aggrieved name-holders to adopt a more accurate transliteration instead? Could it be that the transliteration rules which are now standard are not optimally suited to this purpose?

View on HN

Pinyin is pretty close. At least it's predictable. Wade-Giles is way off -well, I guess it was based off Cantonese?

In any case, my pet peeve is people expecting speakers of English to accommodate foreign sounds, but not extending the same ask to foreign speakers who often mangle English name sounds just as much.

I say just let them say it however they know how and let it be. I don't mind if they can't pronounce my name well.

But back to pinyin, that was intended as an alphabetic alternative for native Chinese speakers to use instead of logographs/characters. So it's audience wasn't foreign language speakers and thus will have shortcomings for them. This was back when they were tackling illiteracy and pinyin was seen as both potential replacement of characters or at least an aid in learning a simplified version of the original characters.

Pinyin is predictable if you memorize every single syllable that exists, it is not logical on a letter-by-letter basis.

For example, "liu" should be pronounced "liou" and "shui" should be pronounced "shuei", but the "o" and "e" are omitted in Pinyin for brevity. Similarly, the "u" in lu and mu is different from the "u" in "ju" and "qu"; the latter is actually "jü" and "qü" as in "nü" and "lü", again the diareses are omitted purely for the convenience of the writer. The "a" in "yan" and "ya"/"yang" is completely different, etc.

I like Pinyin as a reasonably concise input method, but it's not great for romanizing names.

> liu" should be pronounced "liou" and "shui" should be pronounced "shuei"

This is nuanced: in some cases it’s omission for brevity [1] and in other cases it’s how the phonemes sound 100 years ago [2]. And if you pronounce them fast enough there isn’t much difference.

The omitted diaeresis is certainly for brevity, although using v is more common to me.

[1]: https://www.zhihu.com/question/26010099 [2]: https://www.zhihu.com/question/313646560

Probably you speak more Chinese than I do, but as I understand it, you don't have to memorize the syllables, just the 22 (?) onsets and 17 (?) vowel+codas, same as in bopomofo, just more sneakily. Plus nü and lü, I guess. But that still means memorizing 41 pronunciations instead of the 374 you suggest. Is that right, or is the situation far more complicated than it appears?

I've never learned bopomofo in earnest, and for pinyin I just memorized everything that felt like an exception to me, even though it might have some internal logic. For example, I guess that the "a" in "yan" is different from "lan" because "yan" is its own coda, whereas "lan" is "l" + "an"?

Hmm, I don't know.

> In any case, my pet peeve is people expecting speakers of English to accommodate foreign sounds, but not extending the same ask to foreign speakers who often mangle English name sounds just as much.

Sorry... what? Us non native speaker live in a world where English is taught in almost any school on the planet and where you have a peer pressure to know and speak decently English. This on top of one of the most irregular languages in the world wrt pronunciation of dictionary words, let alone names.

PS good luck getting right Greenwich, for example.

As a native English speaker, for the most part I typically know what someone means when they say "gl-OW-kester square" or "Green-witch mean time", so don't worry about it.

I have a Russian friend who has specifically asked me to go out of my way to correct their mispronunciations, but for the most part if someone says an English word in a weird I don't even blink.

I'd say part of being an effective English communicator is not just about speaking English, but also about being able to understand foreign English accents (and I include different dialects of native English in that as well).

English is also unusually flexible, something I noticed abroad. In English you can do ridiculous wrong things and still create something that a skilled listener can easily understand. Subject and verb all you know? Eh, I can guess from context.

Try putting the subject and verb in the wrong order in Japanese or forget your counter particle for a particular kind of counting and you just get blank stares.

I think that English, for its many faults, is at least extremely error tolerant in terms of it being possible to say something that's comprehensible enough dozens of ways and often with words practically in random order. It's probably the most amazing thing about English to me, although I only know a few other languages they are much more strict.

> I have a Russian friend who has specifically asked me to go out of my way to correct their mispronunciations, but for the most part if someone says an English word in a weird I don't even blink.

Yes, English has so many weird native dialects that weird foreign pronunciations shouldn’t throw you off either.

The Finns seems obtuse in their insistence on using Finnish phonology when speaking English, which is easy to make light of on the surface, but when you think about it it’s not actually a bad system. It’s consistent, easy to understand and it lets them focus on gaining fluency over memorising pointless irregular pronunciations.

I think you have to understand that "international English" is not the same as English. We have the same thing in France where lots of people come to study from other countries where they speak French, and their French is a little different. I'm not going to waste anyone's time on correcting someone because they said "le table" and not "la table". Same thing with the pronunciation. I can understand them, they can understand me, and that's what matters. That's the "price" of having a huge influence in the world.

>But back to pinyin, that was intended as an alphabetic alternative for native Chinese speakers to use instead of pictogram/characters. So it's audience wasn't foreign language speakers and thus will have shortcomings for them.

Yes that makes sense, and is reasonable. But the expectation for all foreigners to be familiar with it is not.

But I'd say the expectation that professionals who have to pronounce Chinese names pretty frequently, like those who do announcements at an airport, or newscasters, could spend the time to familiarize themselves with it IS reasonable.

I agree, it's crazy how often the news talk about China nowadays and almost nobody pronounces the names correctly.

It’s not a big deal. Watch some news in Chinese and you’ll hear similar butchering.

It’s so exaggerated US cities are given totally different names in Chinese -there is little resemblance to the North American name of the city.

That's true, although in Chinese these are official transliterations as Hanzi.

I think if someone says Milan in English instead of Milano it's fine because the city has an English name as well.

Right, often what they do is they try to find "close" syllables. Close is the wrong term as some are not close at all. Vancouver and Philadelphia for example.

It's like taking Milano and cutting it up into Mil + Ano and saying, hmm, in English the closest syllables are approximately Mill + Anus, we'll call it Millanus, oh, but that's got a word we should avoid so let's call it MillLand.

The point is, it's not wrong or bad for (any) people to get foreign words wrong when they pronounce them.

You fail to understand most of the world use international English where nobody really cares about correct pronunciation, but about understanding each other and only small minority of English users are nowadays native English speakers, one could say it's questionable nowadays what is the correct pronunciation (intended original use changed), so it would be quite foolish to insist on correct pronunciation in international enviroment.

I don't think same can be said about English native speakers who usually don't speak even second language and can't go their way to at least pronounce few names correctly, while rest of the world is learning whole other language, not just pronunciation of bunch of names.

Agreed with your pinyin commnets.

Pinyin was not designed as a system for making it easier for English speakers to pronounce Chinese, but as a way for native Chinese speakers to write their own language. During the early years of the revolution in China, there were proposals of various levels of seriousness to ditch characters entirely in favor of a Latin script like pinyin, and this actually happened in Vietnam.

Wade-Giles, an earlier system of transliterating Chinese, was designed more with English speakers in mind, but suffered from a number of problems of its own, particularly requiring apostrophes all over the place.

And in Japan you can still see two competing systems duking it out, with the Japanese-for-Japanese Kunrei system ("Huzi") officially anointed as standard but in practice the Japanese-for-Americans Hepburn ("Fuji") being far more common.

>Pinyin was not designed as a system for making it easier for English speakers to pronounce Chinese, but as a way for native Chinese speakers to write their own language.

In this case, it makes even less sense to blame English speakers for incorrect pronunciations resulting from this system.

Kunrei Shiki is really annoying. Nobody knows how to pronounce Mitutoyo correctly because someone in Japan got scared of the supposed ambiguity caused by "tsu".

I recently saw some publicity video by Mitutoyo America on YouTube that pronounced it [mitutɔjou]. All the machinists I watch on YouTube pronounce it [mɪtsutɔjou] though.

> If the official transliteration into the Roman alphabet leads to less accuracy in pronunciation than some other (for example see the second vs third columns in the table in TFA), then would it not make better sense to adopt a more accurate transliteration instead?

Can't be done. The Roman alphabet is a pretty good alphabet for Latin, but it isn't for Chinese, which has different sounds than are in Latin (or English).

For example, stop consonants in English have a voiced-unvoiced distinction: /p b t d k g/

Latin is similar: /p b t d k g kw gw/

Chinese is totally different. Instead of a voiced-unvoiced distinction, it has aspirated-unaspirated: /p ph t th k kh/ (with the h denoting an aspirated stop).

Is there a resource (maybe with sound samples) that expands on the voiced/unvoiced vs aspirated/unaspirated distinction? It sounds fascinating, but I'm struggling to conceptualize it just with the text here.

The Wikipedia articles on it might be approachable for a layperson with no particular background. I studied some linguistics in university and it's a hobby interest so I can't really judge that anymore. The articles have plenty of sound examples. https://en.wikipedia.org/wiki/Aspirated_consonant https://en.wikipedia.org/wiki/Voice_(phonetics)

English actually has all the relevant sounds in most dialects so I will try to explain in it reference to that. Most English dialects have environments where consonants are reliably aspirated, or not, depending on their voicing and surrounding environment. Unvoiced stops (p/t/k) in isolation "pat" or "cap" are moderately aspirated. Voiced stops (b/d/g) are not like "dog" or "bad". (Leaning on this pattern as a partial hack using a feature common in a lot of European languages, is why p and b are what they are in Pinyin, and many other Chinese romanizations.)

In principle, if your dialect does not have a length contrast for /æ/ after a voiced consonant (many do though including mine, making the vowel longer in that environment, so you'll have to try to say them with equal vowel length) then "pap" and "bab" (to rhyme with cap and cat) when whispered (no vocal chord movement, so all equally unvoiced) should be differentiated only by aspiration similar to Mandarin's p/b difference. Though aspiration is Mandarin is a bit more forceful.

As an aside, most people can hear the difference between those words when whispered, hearing a voiced or unvoiced consonant, based solely on the aspiration hint. Phonetic interpretation is a strange and subtle thing. And voicing in Mandarin? It plays the inferior role to aspiration's main starring role, a role reversal of the English arrangement.

> Unvoiced stops (p/t/k) in isolation "pat" or "cap" are moderately aspirated. Voiced stops (b/d/g) are not like "dog" or "bad". (Leaning on this pattern as a partial hack using a feature common in a lot of European languages, is why p and b are what they are in Pinyin, and many other Chinese romanizations.)

Some languages, such as Hindi, have both the voiced/unvoiced and aspirated/unaspirated distinction, so they have 4 consonants (/t d th dh/) where English only has two (/t d/).

Hindi also has retroflex equivalents of those 4 consonants, and there are minimal pairs between all 8 of them. A typical English speaker would struggle to tell the difference between them.

For these sorts of reasons, a Romanisation of names in different languages that (1) is easy to pronounce for English speakers, and (2) sounds like the word in the original language, is going to be a tall order.

Compare the t in 'stellar' vs the t in 'tell'.

What's the difference between the t in 'stellar' vs the t in 'tell'? I seem to pronounce them identically, dictionary pronunciation guides (https://www.oxfordlearnersdictionaries.com/definition/englis... and https://www.oxfordlearnersdictionaries.com/definition/englis...) do not make a distinction..

[edit] ah, one (!) of the two "tell" audio recordings in that dictionary seems to pronounce it in a different way that would illustrate the aspiration, but that seems different to how I've usually heard 'tell' being pronounced.

Both recordings of the "tell" word exhibit aspiration. I guess it is perhaps more noticeable to me because I'm quite well-versed in phonetics.

A common way of illustrating aspiration is: hold a sheet of paper in front of your mouth (or your hand), as you say the words. "tell" will give you a noticeable puff of air as it leaves the mouth, as opposed to "stellar", which won't.

Maybe these recordings are helpful to you: https://en.wikipedia.org/wiki/Aspirated_consonant#Allophonic

Thanks, that was helpful, I seem to get the difference now - just the "tell" example wasn't working as I (apparently unlike most English speakers) seem to pronounce that word with an unaspirated t.

Are you a native English speaker? It's common for native speakers to be pretty deaf to the differences between allophones. Try recording yourself in Audacity saying "Tell the stellar tart to start stewing a tall stall too," then highlight the two "tell" sounds and listen to them repeatedly. Also the two "tart" sounds, the two "too"/"tew" sounds, and the two "tall" sounds. Maybe you'll be able to hear the puff of white noise that's missing in "stellar", "start", "stew", and "stall". That's the aspiration.

This is so hard to explain in text, given the variation of English accents and that aspiration is not meaningful in English. For my regional speech, the strongest difference would be something like the difference between the /k/ sounds in “kale” and “scale”. Nothing to do with my feelings for the vegetable, but it gets an extra scornful bit of turbulence.

I'm always puzzled by certain British imperial transliteration schemes that seem to do their best to set you up for failure and embarrassment.

Apparently "Burma" is supposed to be pronounced in a non-rhotic fashion, but would it have killed the non-rhotic speakers to write "Bama"? Then the rhotic speakers who form the majority of English speakers would have at least had a fighting chance to get it right.

Another thing is their scattering of seemingly pointless "h"s in their transliteration of Devanagari, which together with "s"s leads many speakers to incorrectly assume a fricative: https://www.youtube.com/watch?v=8l8x36fjvjk

OK, I'm Indian and I will stipulate that transliterations of Devanagari are inconsistent, but that guy you linked to is just being a crude racist asshole. (0:39 "and it's so appropriate because she's Indian." Fuck you, Paul Henry!)

In the specific case, her last name would be pronounced "deekshit", with both "d" and "t" being soft like in Spanish. Both those consonant sounds do not exist in standard American/British English.

What is more exasperating to me is that we are not even consistent in English spellings of the same names. For instance, "Sameer" and "Samir" are both pronounced the same but the first is closer to the actual pronunciation.

As far as pronunciation goes, I have made peace with foreigners incorrectly pronouncing sounds that don't exist in their languages, but it bugs me when they can't be arsed to get the syllable emphasis right or when they use the wrong vowel sound even though the correct vowel sound is familiar to them.

Example of #1: A friend of mine is named "Manohar". The syllable emphasis and spelling is: "Muh-NOH-huhr". But his name gets pronounced as "Man-oh-HAAR" in the US. And it's not because he's pronouncing it wrong. Imagine if he pronounced every "Michael" as "Mikhail". People would be pissed.

Example of #2: I can understand them not pronouncing the soft "t" and "d" in "tandoori", but not so much having the first syllable rhyme with "can" instead of "run". Or "Punjab" becoming "Poon-jab" instead of "Puhn-jaab".

Also, #PSA, Hindi does not have the "zh" sound in it, so please stop pronouncing "Raj" as "Razh". Thank you. :)

I'm studying Hindi as a second language, and the inconsistency of transliterations drives me crazy, I'd rather nobody transliterated and just used devanagari if it's not going to be IAST.

Worse, Google has updated the (abc -> हिंदी) keyboard on Android so that in many cases it actually refuses the standard transliterations in favour of popular ones, I assume. And so if you want पीता you have to type 'peeta'; 'piita' won't get it, even if you click the caret for more options.

It also reinforces things like हूं - which I can understand if you're writing by hand, but why wouldn't a digital keyboard input the 'correct' / 'proper' / whatever हूँ?

(To write it there I used the direct input method rather than transliteration one, with which afaik it's not possible to write it.)

In fairness to butcherers of Manohar's name though, you really do need to be told/have learnt how the transliteration (IAST or roughly) is supposed to work. Otherwise of course you're going to read it in the language you know.

The trouble is, once you do know, you're left wondering if something is transliterated 'correctly' (in standard as opposed to popular fashion) or English-phonetically... Particularly with vowels: 'e' for example is either 'ay' as in 'pay' or 'e' as in 'pee' accordingly. (The latter 'should' be 'ī' or 'ii'.)

You have my condolences for Google's nonsense. I have not written anything in Hindi for the past couple of decades, I think, despite growing up in Delhi and speaking the language fluently. All my work communication and conversation with friends is in English. The good thing about Hindi is that if you know the script, it's pronounced exactly as it's written. There are no strange inconsistencies like the monstrosity that is English.

> In fairness to butcherers of Manohar's name though, you really do need to be told/have learnt how the transliteration (IAST or roughly) is supposed to work. Otherwise of course you're going to read it in the language you know.

I have absolutely no ill-will for people who pronounce any language's names solely after reading them. But if you work with someone and they introduce themselves to you, it's just common courtesy to at least make a best effort attempt to get the syllable emphasis and the vowels right. If you _hear_ it and still get it hopelessly wrong, I would be less charitable.

> In the specific case, her last name would be pronounced "deekshit", with both "d" and "t" being soft like in Spanish. Both those consonant sounds do not exist in standard American/British English.

So he did get the fricative right after all? The sounds certainly exist, unfortunately there is no unambiguous way of denoting them without resorting to the now archaic "ð" and "þ".

Could you perhaps shed some light on what a seemingly silent "h" after a consonant really indicates? Take names like "Delhi" and "Bhopal".

> Example of #1: A friend of mine is named "Manohar". The syllable emphasis and spelling is: "Muh-NOH-huhr". But his name gets pronounced as "Man-oh-HAAR" in the US. And it's not because he's pronouncing it wrong. Imagine if he pronounced every "Michael" as "Mikhail". People would be pissed.

I agree that errors in emphasis are far more jarring than errors in sounds, but using the native cognate in the language being spoken of a foreign name in that category is actually very common and often encouraged by the bearer of the name. Perhaps because it has already been mangled so many times since its Hebrew origin through Koine to all the modern languages that use it.

On thing that has always puzzled me is despite of what a dumpster fire Devanagari transliteration to Latin is, it seems to still have a massive buy-in with native speakers online. In places like social media or the Indian channels on the company Slack, it seems more common than actual Devanagari. I see the same pattern with Malayalam being written in Latin rather than its own script. If there's no need to accommodate people who don't know the script, I fail to see the point. I've asked but people usually get very defensive about it. Are they ashamed of having different scripts from the globally dominant one, Latin? I would rather consider it a point of pride.

> So he did get the fricative right after all?

Did not sound like that to me. He pronounced "d" and "t" as ड and ट when it should be द and त. I think the first consonant is closer to how you might pronounce the article "the".

> Could you perhaps shed some light on what a seemingly silent "h" after a consonant really indicates? Take names like "Delhi" and "Bhopal".

Oh boy. OK, "Delhi" is the Anglicised name of the city name, which in Hindi is "Dilli", so it's actually pronounced "Del-hi", hard "D", faint "h". Though it could be clipped to sound just like "deli" depending on whom you're speaking to.

But in cases where it's a transliteration, _usually_ it means that it's an aspirated version of the consonant in question. "Bhopal" is one such instance. The Devanagari script has, for some consonants, multiple versions: soft, aspirated soft, hard, aspirated hard. (I'm not a linguist; there may be better terms for these.) Let's take "d". There's the soft "d" pronounced as Spanish-speaking people pronounce "d". There's a hard "d" pronounced as in "door". Then there's "dh" and "ddh", which are the aspirated forms of the respective letters.

(Google "pronounce Bhopal" and switch to the "Indian English" pronunciation option. I also checked the top 5 or so youtube videos and ALL of them are wrong; they have an American pronouncing it. WTF?)

But as I mentioned, our transliteration is not standardised and it's indeed a dumpster fire, so sometimes, the consonant is "ddh", but it's written as "dh" in a word, which does not help things at all.

Want more complications? In South India, "th" is often used instead of "t" in names where there's a soft "t". But it's pronounced exactly the same. [1]

Oh, and Devanagari is a simpler script. Other Indian languages like Malayalam have even more sounds that Westerners have zero hope of pronouncing correctly. Heck, Indians from other states struggle with it.

As for names, you'll find multiple spellings for names that are pronounced the same. I have friends named "Aarti" and they are spelt "Aarti", "Arathi", "Aarthi", and "Arati". A royal mess.

And this is before some idiotic astrological/numerological superstition makes them add unnecessary letters to their names. (Not common, but does happen.)

> In places like social media or the Indian channels on the company Slack, it seems more common than actual Devanagari.

I can only speculate that there are two reasons for this:

1) PC keyboards in India are English QWERTY and typing regional language text on them is a pain.

2) The medium of instruction in middle class education is English, so many of us grow up speaking English well. And in some cases like mine, think in English too. As we progress in our careers, communication in the workplace is in English, and writing fluency in native languages decreases. It's just easier to type it out in English. I consume pretty much all my Internet content in English (though I can speak 3 Indian languages), and it bugs the hell out of me too because I have to translate from the weird English transliteration to the native language in my head first. And that's before people use SMS-like contractions for words.

[1] Speculative answers here: https://www.reddit.com/r/india/comments/5uwcwi/people_from_s...

This will only make sense to you if you read Hindi: https://qr.ae/pG5zMP

> "Delhi" is the Anglicised name of the city name, which in Hindi is "Dilli", so it's actually pronounced "Del-hi", hard "D", faint "h". Though it could be clipped to sound just like "deli" depending on whom you're speaking to.

To be fair to the English orthography, the English 'Delhi' doesn't come from the Hindi दिल्ली, but rather an anglicization of the Persian دهلی when the East India Company arrived, retained in modern Urdu as دہلی or 'dehli'.

> Apparently "Burma" is supposed to be pronounced in a non-rhotic fashion, but would it have killed the non-rhotic speakers to write "Bama"? Then the rhotic speakers who form the majority of English speakers would have at least had a fighting chance to get it right.

To be fair, Bama wouldn't be the right pronunciation. The easiest way for an RP speaker to indicate what the right pronunciation should be is to write "Burma". An American might try Buhma.

And your complain about "rhotic speakers" being "the majority of English speakers" could only have been taken into account of the British transliterator had access to a time machine to determine the future.

At the time, a person in the British empire wouldn't have been considering the opinions of an American, because America and Britain had separate spheres of influence and a generally amicable but not integrated existence. To them, RP was the language of the elite and that was all that mattered.

But if they were to consider the situation in America, they would have seen that in America as in Britain, non-rhotic accents were on the rise and were somewhat prestigious. If they were going to consider the future situation of the 21st century globe and how someone might pronounce the word in an American-centred word, it would have been entirely reasonable for them to have supposed it would have been one with a non-rhotic accent.

Another matter you might consider is that a very strict spelling would not necessarily help matters anyway. The J in Beijing is pronounced more-or-less like an English j, but that doesn't stop people from mutilating it into the s of measure. Or the person who responds to you, who observes that "Punjab" should be pronounced roughly like the English spelling should indicate, but which is often pronounced with the u of push instead - both because of a widespread assumption that all words in foreign languages are pronounced like Italian, Spanish or French.

And the final matter, which might hopefully set your blood at some ease, is to note that when a word is borrowed into a foreign language, it is subject to different pressures. You are not obliged to pronounce "Burma" or "Myanmar" the way any particular natives of Burma or Myanmar do. The English words are English words, intended to help English speakers identify a place in a generally non-offensive manner. There will always be variation between the source pronunciation and the destination pronunciation, especially once a few hundred years have intervened, and that's fine. The question is can we all agree that these are suitable words for referring to that country?

English is one language with multiple dialects, so words that enter the language from a certain dialect will reflect that history to a greater or lesser degree. Even when there isn't an "r" in the spelling, Americans still manage to put an r into the pronunciation, as with Goethe or Gödel. So to blame the British for unetymological r is just false. Americans do it because they recognise the unity of the language, and they're communicating rather than upholding any particular notion of perfection. It is unreasonable to say that the only reason for the British to do the same is that their life depends on it.

I wondered this too when learning Pinyin. I believe the reason is that while Pinyin's X and Q may sound like "sh" and "ch" to English speakers, they are actually different sounds (and contrast with the "sh" and "ch" sounds that also exist in Mandarin). That said, I still wonder why -ui was chosen for "wei", "shi" is pronounced "shr", "-ian" is pronounced "yen", etc...

MPS-II seems much closer to intuitive pronunciation for English speakers, although it never caught on: https://en.wikipedia.org/wiki/Mandarin_Phonetic_Symbols_II

> I believe the reason is that while Pinyin's X and Q may sound like "sh" and "ch" to English speakers, they are actually different sounds (and contrast with the "sh" and "ch" sounds that also exist in Mandarin).

Neither Mandarin sh nor x exist in English. X is sharper and is a little more like shy-, whereas sh is darker and a little like shr-. In reality they both strike English speakers as being quite like sh, because we use the space differently.

I've heard various native English speakers who are fluent in Mandarin saying e.g. Xi as Si. I don't know if there's a dialectal difference in English or in Mandarin or both that accounts for it.

> That said, I still wonder why -ui was chosen for "wei"

It looks better, if you have a certain aesthetic.

> "shi" is pronounced "shr"

As far as I know Mandarin speakers just consider them the same. Pinyin is designed for native speakers.

> "-ian" is pronounced "yen"

The e in Pinyin isn't really the English e sound. It's more like the American short u in fun or the NZ short i in pin. It's not obvious what letter to use to represent that, but e is not completely unreasonable. When they do have an e-sound according to the internal logic of Mandarin, it's a variant of an a-sound.

> That said, I still wonder why -ui was chosen for "wei", "shi" is pronounced "shr"

Shi is pronounced Shr? That sounds more like your typical pirate Beijing hua. Which can get quite confusing when you get out of Beijing to let's say to Shanghai business trip and you can't figure if something cost 4 or 10 kuai at street vendor (my personal anecdote, I think they actually do it on purpose since normally 4 would have very different sound).

A Chinese teacher once told me that pinyin is based off Russian

That's more wrong than correct. There's only one minor part of the design where we can see a relation: Latin q resembles Cyrillic ч.

I find the title misleading, as the link is only covering Chinese. But the same problem exists with other languages that are already using the "Roman alphabet". So, with that in mind, what makes you think the transliteration should be optimized for English? :)

Yeah, I was kind of blueballed by the page. Maybe the author expected pull requests for other languages?

In some place, the transliterated rules aren't up to the name bearer. There's official rules and when you apply for international documents, the government decides how your name looks like in romanized form.

Do you also blame French system when you mispronounced Guillaume Dubois? Like, French should use a better system to transliterate into Roman alphabet for English pronunciation?

Pinyin is not a transliteration into English, it's a way to write Chinese using the latin alphabet. Learning how to pronounce Pinyin is no difference than learning how to pronounce French or German text.

So when you say "why the speaker is blamed rather than the transliteration" it's like saying "why the English speaker is blamed rather than the French spelling" about an English speaker having problems pronouncing a French word.

The goal of pinyin was not to make Mandarin words pronounceable to English speakers, but to be a consistent means of representing the sounds of Mandarin to Mandarin speakers.

Of course.

But the "official" (as you say) transliterations of world languages enjoy wide usage not because they're necessarily the best anyone could come up with --- rather, they just represented the right local maxima at the right time in history.

Displacing those romanization systems with better ones (assuming those exist) would be an uphill battle lasting decades, and I wager there aren't many people who would find that struggle worth the effort.

Particularly when the dominant language that uses roman letters is lacking a consistent alphabetic encoding system itself. Before you start replacing pinyin, you might want to take a long, critical look at how English is written.

There are plenty of other transliteration systems that might lead to more accurate pronunciations for English speakers in particular (as opposed to say, native Russian speakers); see Wade-Giles or Gwoyeu Romatzyh. The factors that led to the current de-facto standard are more related to 20th century history rather than optimizing for the particular use case of reading aloud by non-natives.

Maybe it's transliterated for a more Latin-like crowd where vowels have exactly one pronunciation, like in languages like Spanish, where reading the second columns as they'd naturally think to, would likely be 95% OK

Reply on news.ycombinator.com