On Romanization

Update Nov 2011

Forget it.  There is no policy on Romanization.  You’re on your own.


After about three weeks of blogging, Frozen Garlic announces a new Romanization policy.  As of now, almost all Chinese terms will be rendered in the hanyu pinyin system.  I will make exceptions for a few familiar names and places.  This includes (but is not necessarily limited to):








Lee Teng-hui

Frank Hsieh

James Soong

Lian Chan

Of course, it is not Frozen Garlic’s style to use only a few words when a lot more could be written.  So let me elaborate.

As you might be aware, Chinese characters (unlike French or German words) do not use letters from the Roman alphabet.  Therefore, it is necessary to adopt some conversion method to render them legible in English-language essays.

There is no shortage of systems.  In the academic world, we are most familiar with three systems.  The Wade-Giles system has been around for more than a century and was the academic gold standard until a few decades ago.  It is still used in the humanities, but not in the natural sciences or social sciences.  The Wade-Giles system’s most prominent feature is the use of an apostrophe to represent an aspiration (a puff of air), so that “teng” is pronounced like an English D while “t’eng” is pronounced like an English T.  In common use, many people drop all the apostrophes.  This corrupted Wade-Giles makes it impossible to know how a sound is pronounced.

The Yale system was developed at … wait for it … Yale University.  It has mostly faded away.  It’s too bad, because the Yale system was arguably the one that the average American would come closest to reading with the correct pronunciation.

The dominant standard today is the hanyu pinyin system invented in the PRC.  Many casual readers cringe at this system because they have no idea how to pronounce sounds such as zh, q, and x.  Also, lots of people in both the USA and in Taiwan opposed this system because of its communist origins.  However, almost all language students since the 1980s (including myself) have studied Chinese using the hanyu pinyin system.  (The system was developed in the 1950s with the help of Soviet advisors.  I have been told that all the weird spellings are nearly perfect representations of the Mandarin; you simply have to read them as a Russian would.  I love this story far too much too test it and thus risk having it proven wrong.)

But wait, that’s not all.  There are a few even more obscure systems floating around.  For example, every now and then I would come across people who insisted their English surnames were Perng or Horng, even though I was quite sure there was nothing even remotely similar to an “r” sound in their names.  They were apparently spelling their names according to Gwoyeu Romatzh (Wikipedia is great!), MPS II (a modification of Gwoyeu Romatzh), or a corruption of one of the two.  In those systems, different spellings are used to indicate different tones.  The “r” in Perng and Horng indicates that both are 2nd tone.

Then there is the Tongyong pinyin system.  The Chen administration tried to promote this domestically developed system as an alternative to the (PRC) hanyu pinyin system.  It was supposed to be adapted to the needs of Taiwan.  In other words, its developers thought it could be used for other languages spoken on the island, including Taiwanese, Hakka, and various Aborignal languages.  Speakers and teachers of those languages politely disagreed.  The main effect of the Tongyong system has been to confuse foreigner and cause a lot of road signs to be replaced one more time.  The system has thankfully been dropped, though you still see road signs leading to “Banciao” or “Shyinjuang” and wonder when those places were invented.

So for example, how would we render “presidential candidate Ma Yingjeou”?

Chinese 總統 候選人 馬英九
English president candidate Ma Yingjeou
Wade-Giles Tsung3 t’ung3 Hou4 hsüan3 jen2 Ma3 Ying1 chiu3
Yale Tzǔng tǔng Hoù shwăn rén Mă Yng-jyǒu
Hanyu pinyin Zǒng tǒng Hoù xǚan rén Mă Yīngjiŭ
Gwoyeu Romatzh Tzoong toong How xheuan rern Maa Ing-jeou
Tongyong pinyin Zǒng tǒng Hoù syuăn rén Mă Yīng-jyǒu

The various systems basically agree on some sounds, like President Ma’s surname (which literally means “horse”.  I like to call him by his literally translated name: Handsome Horse Number Nine).  There are other characters, such as the second character in candidate, that no one agrees on.  There are many, many more such examples.

Of course, that’s what you use if you are trying to learn the language.  For the casual English reader who doesn’t much of a chance of grasping tones (it took me 18 months before I really started hearing tones), all the numbers and accent marks are automatically dropped.  So if you are reading a newspaper and they translate this phrase using hanyu pinyin, you will simply see “zongtong houxuanren Ma Yingjiu.”

What do Taiwanese children use to learn Mandarin?  It seems like that should be the commonly used Romanization standard.  In the PRC, kids learn hanyu pinyin, so everyone knows the standard.  Great idea, except they don’t use a letter-based system here in Taiwan.  Instead, kids learn a character-based system, called the Mandarin Phonetic Symbol system, commonly known as the Bopomofo system (named after the first four sounds).  So where the PRC uses “b,” “p,” “m,” and “f” Taiwan uses ㄅ, ㄆ, ㄇ, and ㄈ.  (This is also the system I use to type Chinese characters on my computer.)

Since no one in Taiwan grew up using a letter-based Romanization system and the government has not consistently insisted on a universal standard, people tend to spell their names however they like.  Sometimes, they obtain an official-looking table that tells them how to spell their name, but often they just sound it out and write whatever they think looks good.  It is chaotic, but a name is also a deeply personal thing, and it’s not such a bad thing for people to have their own personal spellings that they feel attached to.  Americans certainly don’t feel the need to be consistent in their spellings of names such as Carrie, Cary, Cari, Keri, Kerry, or my cousin whose spelling is so strange that I can’t remember it.  Less forgivably, traffic signs have likewise been a mishmash of different systems.  I once saw three signs at the same intersection calling the road “Heping Rd,” “Herping Rd,” and “Hoping Rd.”  This left me “hoping” that “he” was not really a “her.”

Oh, and one more thing.  People don’t always use the Mandarin version of their name.  For example, I once interviewed the chair of the Taiwan Indpendence Party, Xu Shikai 許世楷.  After I called him Chairman Xu a couple of times, he stopped me and said, “I am not Chairman Xu.  I am Chairman Ko.  Ko Seh-kai.”  He used his Taiwanese pronunciation, because that was how he thought of himself.  Bi-khim Hsiao 蕭美琴 also does this.

So that’s why people in Taiwan can’t spell anything consistently and why all the English-language newspapers spell the same things differently.  As for me, I’m giving up on trying to remember how everyone spells their name.  I know lots of people, especially Taiwan nationalists, dislike having the PRC hanyu pinyin system.  I dislike imposing it upon them.  However, in only three weeks, I’ve found myself spelling the same thing in multiple ways and wasting time looking up how I did it last time.  Since almost no one reads my blog anyway, I’ll do it the way that’s most convenient for me.

I’ll also always provide the Chinese characters so that people who can read them know who I’m talking about.

UPDDATE (Nov. 3, 2010)

I have decided to grandfather in all the conventional spellings of the mayoral candidates.  Most of these people have spent some time in the legislature, so I’m using the spelling from the Legislative Yuan’s English website.  This has the benefit of being official.  It has a few drawbacks, too.  They capitalize the first letter of the third character, which Koreans usually do but Taiwanese usually do not.  Also, they aren’t necessarily consistent.  For example, the first person I looked up was 沈智慧,who was spelled Shen Chin-hwei in the 6th term but Shen Chih-hui in the 5th term.  From this, I conclude that she doesn’t care enough about the English spelling to make sure that they get it right.  Similarly, the Taipei Times is calling the DPP candidate for Taichung Mayor 蘇嘉全 Su Jia-chyuan while the LY website calls him Su Chia-chuan.

Because it is difficult to remember some of these unconventional spellings, I’m going to call them by English names (William Lai, Eric Chu, Jason Hu) a lot.   Maybe the only person who I can’t complain about too much is Hsu Tain-tsair.  I have no idea where he came up with that (horridly confusing) spelling, but he lived in the US for many years, so I imagine he actually identifies with that spelling.

Chinese Pinyin LY spelling,English name
郝龍斌 Hao Longbin Hau Lung-bin
蘇貞昌 Su Zhenchang Su Tseng-chang
朱立倫 Zhu Lilun Chu Li-luanEric Chu
蔡英文 Cai Yingwen Tsai Ing-wen
胡志強 Hu Zhiqiang Jason Hu
蘇嘉全 Su Jiaquan Su Chia-chuan
郭添財 Guo Tiancai Kuo Tien-tsai
賴清德 Lai Qingde Lai Ching-te,William Lai
許添財 Xu Tiancai Hsu Tain-tsair
黃昭順 Huang Zhaoshun Huang Chao-shun
陳菊 Chen Ju Chen Chu
楊秋興 Yang Qiuxing Yang Chiu-hsing

Everyone else will still be referred to in pinyin.

11 Responses to “On Romanization”

  1. David on Formosa Says:

    I think you have a great blog here. The issue of romanisation is always a difficult one to deal with in Taiwan. Unfortunately a practical linguistic matter has been made into a political football and this causes a lot of confusion. I’d just like to make a few suggestions.

    In this article you make reference to Su Zhenchang, You Xikun and Cai Yingwen. This is not how any of these people choose to spell their names in romanised form. Most people know them as Su Tseng-chang, Yu Shyi-kun and Tsai Ing-wen. I know you have explained above that it is a lot of trouble to get the correct spelling, but I think for people’s names it is important to respect their personal choice. For example, if your name was Jon who would you feel if somebody insisted on spelling it John? Also a Taiwanese person whose father and grandfather spelt their name Tsai should not be expected to change it to Cai.

    Another important point is that most people in Taiwan spell their name with a hyphen between the first and second syllable. For example, Ma Ying-jeou not Ma Yingjeou.

    I know it seems pedantic, but there are good practical reasons for doing this. One is to make things easily searchable on the web. A Google search for “Tsai Ing-wen” and “Cai Yingwen” will return very different results. Another point is that it reflects a difference between Taiwan and China. When I see Chou Hsi-wei I know it is referring to a Taiwanese politician, but Zhou Xiwei gives me the impression of referring to a Chinese Communist official.

    It is not that hard to check the spellings using Google, Wikipedia or government websites. You seem quite meticulous about your analysis of opinion polls and election results. It would be nice if you could show the same amount of care with people’s names.

  2. frozengarlic Says:

    David, thanks for the complements.

    In an ideal world, I would certainly do as you say and look everyone up. Unfortunately, there are significant costs in time in doing so. I tried it for the first three weeks and decided that it simply required too much time.

    It is, as you say, fairly easy to look up someone like former Premier Su to see how he prefers to see his name spelled. However, many of the people I reference are not so easy to look up. Of the people running for the Taichung City Council this year, how many have widely accepted English spellings for their names? Where do I draw the line between well-known, standardized spellings such as Tsai Ing-wen, and less well-known names such as Ma Zhaoling, a very local politician referenced in an article about Zhonghe City factions? (Damnit, I wish I knew how to put in those neat links. It would have been useful there.)

    Even if I were to try to respect individual choices, it is not the case that English spellings are standardized. For years, one newspaper referred to Premier Hsiao Wan-chang, while another called him Vincent Siew. During the 1990s, as hard as it is to believe now, we never really new whether the future president’s name was Chen Shui-pien or Chen Shui-bian. Not many people know this, but he even has an English name, Walter. Should I call him Walter Chen?

    Should I arbitrarily follow the Taipei Times? They slavishly follow the Tongyong system (one of the few things I hate even more than Pinyin) on lots of naming conventions. Also, their website has become really slow, so it takes lots of time to look things up.

    The broader point is one about identity. For most people in Taiwan, politicians included, the English name is not a core part of their identity. I am very careful to get Zhou Xiwei’s (Chou Hsi-wei’s?) Chinese name correct, as that is most definitely part of his core identity. (In fact, he changed the characters a few years ago from 周錫偉 to 周錫瑋.) I just don’t have any reason to believe that he cares as much about his English name. (The appropriate analog to your John/Jon argument is insisting on writing his name with the former characters instead of the latter characters rather than the choice of Chou/Zhou; one is a primary identity, the other is much more peripheral.)

    Conceptualization is one thing; operationalization is a completely different matter. In this case, I have basically thrown up my hands in frustration and succumbed to the pinyin convention, something I swore I would never do a decade ago.

  3. Tim Maddog Says:

    First of all, those aren’t “English names” — they’re spellings used in English contexts. “Mark Ma” could be thought of as an English name.

    Secondly, if using Tongyong spellings of local place names (as the Taipei Times does) is “slavinsh” then so is your use of Hanyu, right? Even if you use Hanyu, the “Pinyin convention” for names of people in Taiwan is to use a hyphen between the two syllables of their given name. Your reply to David doesn’t explain why Ma Ying-jeou doesn’t get that hyphen or why you don’t write it as “Yingjiu” or how you can remember that spelling but not “Tsai Ing-wen.”

    It’s incredibly easy to check spellings if you don’t know them via everyday exposure. Here’s an example which shows my preferred method.

    I use Google to do a site search of the Taipei Times — an English-language web site which usually puts Mandarin names in parentheses after the spelling which is used in English contexts — and search for the Hanzi in quotation marks.

    For Tsai Ing-wen, this is what I’d put in the Google search box:
    site:taipeitimes.com “蔡英文”

    You’ll sometimes get variations among the search results, but doing a second search for one or two will give you a good idea of which one to use.

    Another way to find the correct (common) spelling is to find the Mandarin name via Wikipedia. For Tsai Ing-wen, you’d go to this page:

    Once you’re there, look for “English” in the Languages menu in the left sidebar. There isn’t always an English version of the page (which is why I didn’t suggest this first), but if there is it should give you the answer.

    Either method should take less than a minute — not a “significant” amount of time. I’m confident that you can do it!

  4. Alton Says:

    I second the requests of my colleagues that you use the spellings of personal names that are preferred by the persons wearing them. For prominent figures such as Tsai Ing-wen these are easy enough to find. You can hardly miss them, in fact, as you are surely already encountering these spellings in your reading.

    It only took me a few seconds to go here–

    Official DPP site:

    –and see the chair is Tsai Ing-wen.

    You just use the spelling the person prefers, the one that’s already out there, just as you would write ‘Barack’ and not ‘Barak.’

    Granted, variant spellings make life a little sloppy and complex instead of neat and orderly and linear. But as you recognise, that’s life in a free society. It’s as you say about the third-party candidate running now in Kaoshiung: the world isn’t fair.

    I do enjoy the blog. I like the approach, the tone, the analysis. Many thanks.

  5. Alton Says:

    I wanted to add my thanks for an informative and balanced overview of the romanisation options in Taiwan.

    I share your fondness for the now-obscure Yale system. It’s rarely seen now, but it does what it was designed to do: yield accurate phrasebook pronunciations from American speakers of English who are not trained in Chinese. Linguists have other priorities, of course. And its advantages were always going to have geographical limits. German or French speakers, for example, would not find it so helpful in yielding the right pronunciation.

    (The Yale system for Cantonese is still going strong, though, I see. Boolah boolah.)

    Tongyong Pinyin was designed not only to help romanise a variety of island languages. It was also designed to help tourists who used English (not just native speakers, but Japanese and Koreans and anyone who uses English in international travel) to get recognisable pronunciations even though they have never been trained in the use of the system. The goal was to bring some of the virtues of Yale and other familiar approaches into a Pinyin system that retains the advantages of that approach as well.

    Tongyong does a pretty good job realising what it was designed to do. Visitors with no Chinese training will get more accurate pronunciations when reading ‘Banciao’ rather than ‘Banqiao’, ‘Sindian’ rather than ‘Xindian’, and ‘Gaosyong’ rather than either ‘Gaoxiong’ (Hanyu) or ‘Kaohsiung’ (corrupted Wade-Giles).

    Students of Chinese complain about Tongyong because they used another system in class. Fine. They don’t like it because they learned another system in class. It just goes to show that students of Chinese are creatures of habit like everyone else.

    Business people jetting between Taipei, Hong Kong and Singapore prefer not to deal with two romanisation systems while they are also dealing with Traditional and Simplified Chinese characters. Understood.

    But it’s always struck me as grandly irrelevant to dismiss Tongyong as a system for being ‘political.’

    Hanyu Pinyin is political. Politics have everything to do with its acceptance as the UN-accepted standard, everything to do with why any system invented in Taiwan is dead on arrival as an international standard, and why it was adopted in Taipei when Tongyong was the standard everywhere else in Taiwan. Politics have everything to do with Hanyu Pinyin was the system taught to you, too, and why it is the system you now feel most comfortable using. The name ‘Hanyu’ itself is political.

    If Hanyu can be a useful system regardless of who invented it or pushed it, then Tongyong can be the same. The question of political motive falls out of the picture. What’s left are needs, options for meeting those needs, and choices.

    Speaking of which, it’s worth mentioning that in Taiwan the system is officially called the ‘New Phonetic System.’ Taiwan drops the ‘Hanyu’ stuff.

    I don’t knock the decision to use it. Sometimes having a decision, any decision, is the most important thing. One likes standards that really have a fighting chance to be standards. Still, it’s OK to try to invent something that better meets your needs. In free societies, as you know, treating settled policy as open to improvement is always allowed.

  6. Carlos Says:

    I would also prefer seeing widely-used Romanizations where they exist, because using something different slows down the reader and draws attention away from the content. It probably takes a conscious effort on your part, and to top it off you’re reducing the number of hits you could get off web searches. As for drawing the lines between well-known and not… you’re more familiar with these names than most of us, so if you don’t know their Romanized name then chances are we don’t either (so you can go ahead and use hanyu pinyin).

    I’m American with Spanish and Taiwanese parents, and Spanish is probably my first language (my Mandarin and especially Taiwanese are barely existent, sadly). Maybe that causes me to dislike the argument that a Romanization system is better just because it’s easier for English speakers to understand. If you’re Eastern European (or Spanish from 500 years ago), the Hanyu Pinyin pronunciation of the letter ‘c’ is totally normal. In Valencian, ‘x’ is said like English ‘sh.’ (Valencian is my dad’s dialect… it’s to Catalan as Hoklo/Taiwanese is to Fujianese, and under Spain’s own generalissimo all dialect speakers were treated just like Taiwanese speakers were treated – lots of parallels, actually). Another side effect of speaking Spanish is that Portuguese-derived Peh-oe-ji makes a lot of sense to me, especially the b/p/ph and g/k/kh series. I wish it were healthier.

    Sorry, not sure where I was going with that paragraph. Just jumping into the bigger debate I guess. Love the blog!

  7. frozengarlic Says:

    I like the Spanish comparison. For the past few world cups, I have been pronouncing “Xavi” as if it were pinyin. Then I thought maybe it should be pronounced like the X in Mexico.

    I might revisit this whole romanization question sometime after the current election cycle ends, but for now, as you might guess, it is not my top priority. I’m interested in elections, not romanization. So everyone should feel free to give your input, but I’ve basically said my piece for now.

  8. Carlos Says:

    A few hundred years ago (including when Mexico was founded) Spanish and Catalan were more similar, and ‘x’ was pronounced like English ‘sh’. In Spanish it changed to become more like English ‘h’ (Spanish ‘j’) so Xavier became Javier. But everyone in Spain says and spells Xavi’s name in Catalan, so you were saying it right.

    Actually, Spanish speakers can’t really pronounce that sound so it ends up sounding more like “Chavi”.

  9. Carlos Says:

    Thanks for being flexible and acceding to our wishes. It really did make your blog easier to read during the run-up to the mayoral elections.

    Have there been any attempts by news organizations in Taiwan to come up with a standard?

  10. Tetsuo Says:

    I just wanted to say that if you ever find yourself referring to the Taipei Times or China Post for romanizations of anything, you’re doing it wrong. The Taipei Times, at least, can’t even keep romanizations straight in the same day’s paper, let alone across time.

    • frozengarlic Says:

      That’s the problem in a nutshell. Even they can’t always keep these things straight. BTW, I’m using the legislative yuan english website as my source for names, not newspapers. Alas, even that has different spellings in different legislative terms.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: