AllSet Learning Pinyin Chart: now with Gwoyeu Romatzyh!

Yesterday we released version 1.6 of the AllSet Learning Pinyin iPad app. We’ve been getting lots of good feedback on the app (thank you everyone, for the support!), and this latest release is just a small taste of some new functionality coming to this app.

The major thing we added this time that all users can enjoy is the “play all 4 tones in a row” button. It works really well in conjunction with the audio overlay window (not sure what to call that thing semi-transparent rounded-corner box that pops up when you adjust volume on an iPad or play audio in this app). So not only are you hearing the tones, but you’re also seeing the pinyin text in big letters right in front of your face as it plays. The key point is that because the text is big, the tone marks are also clearly visible. (This can be a problem with some software.)

Aside from that, we also added four new romanizations to the chart as addons (click the links below to learn more about each):

  1. Yale Romanization
  2. Gwoyeu Romatzyh
  3. Tongyong Pinyin
  4. MPS2 Romanization

The latter two are of interest mainly to Taiwan-focused sinologists. The first one is of interest to sinologists that like to poke around in musty old texts (the same types that are interested in Wade-Giles). The second one, however, is rather special.

Zhao Yuanren

Gwoyeu Romatzyh was invented by Chao Yuen Ren (赵元任), as legendary a linguistic badass as any that has ever existed. I won’t dwell on him in this article, but one of his accomplishments is inventing his own romanization method (Gwoyeu Romatzyh) which uses alternate spellings to indicate tones rather than tone marks or numbers. The idea was that tones should be an integral part of each Chinese syllable, not merely something tacked onto the end as an afterthought, and that binding tone to the spelling of each syllable is a way to enforce that.

Unfortunately, Gwoyeu Romatzyh (AKA “GR”) is a bit confusing. You can’t have regular alternate spelling conventions without running into conflicts, which forces a certain amount of irregularity, and well… it gets a little messy. It was definitely an interesting experiment, nevertheless.

While I would never considera using GR for any practical purpose, I do find that having GR on the AllSet Learning Pinyin chart breathes new life into the system for me, specifically as I play through the four tones of various syllables and watch the text update accordingly. Patterns start to emerge. Check out the following video, where I first play the syllable “shang” in pinyin (all 4 tones), then switch over to GR and repeat it, and then go through a whole slew of syllables in all 4 tones.

If you have an iPad, please be sure to check out version 1.6 of the AllSet Learning Pinyin app. Note that GR is available as an addon in the “Addons” section.


John Pasden

John is a Shanghai-based linguist and entrepreneur, founder of AllSet Learning.


  1. Thanks John, I wish I had this app when I started learning Chinese! Good re-enforcement of the tones from the beginning.

  2. In physics/math we call this “multiple representations“. Interesting seeing it applied to pinyin as an active learning strategy. I remember flipping through my Wade-Giles Chinese dictionary for a different perspective on how to pronounce Chinese, but I was always a bit hesitant to use what I inferred from that because I was afraid that the originators of W-G might have used old/dialect-accented Mandarin as a basis for their romanization. In any case, still a fun and useful strategy for language geeks.

    • I wouldn’t say we’re implementing these as a “learning strategy.” It’s more of a reference.

      If you’re already interested in Wade-Giles or GR, then it’s great to have an app like this to clarify everything visually. But the app certainly doesn’t push you into the alternatives.

  3. Neato – will check it out if I ever do get an iPad.

    re: multiple representations — I had a cache of second-hand textbooks in Yale romanization at the outset, and I think it was a big help for my pronunciation, at least as far as getting me to a more-or-less good-enough point by bypassing pinyin “x” and “q” and telling me that Pinyin “zhi” was just “jr.” W-G is actually a great system when it’s used properly, and it represents some relationships between sounds that Pinyin doesn’t…but I don’t know if it would’ve helped me any at the outset. The thing with WG “yu/yü” equalling Pinyin “you/yu” would’ve messed me up bad, I suspect.

  4. Peter Nelson Says: July 27, 2012 at 11:40 am

    @MicahS All of the romanization systems are basically isomorphic (in the sense that you could automatically translate syllables written in one system into another, without any context). They all reflect more or less the same understanding of initials and finals as well. As to the particular letter sequences assigned to the initials and finals in the different systems, I would not attach a great deal of importance to them. They’re all standing in for the same pronunciations anyway, and if you want to know what those pronunciations are, I’d recommend looking at the IPA values for the initials and finals or (I know it sounds crazy!) listening to Chinese people speak.

    • Peter,

      No, I don’t think you can say that “all of the romanization systems are basically isomorphic (in the sense that you could automatically translate syllables written in one system into another, without any context).” There are some significant differences in how j/q/x are represented vs. zh/ch/sh (in some systems contrasting pairs are conflated into one representation, since they exist in complementary distribution, and there’s no actual conflict). Same goes for yu vs. yü and some of the different “-i” sounds.

  5. Fascinating! Thanks for introducing another interesting romanization of Chinese to me, John. Like other commenters above, I wish that I had an iPad and some of these great apps that are out nowadays back when I first started learning Chinese.

  6. John, I’m very surprised to see that you have incorporated GR into the app! But you may be surprised to know that there is at least one person (me) out there who listens to your Chinesepod podcasts and visualizes the words in GR. If you (or Dilu) says 食欲, in my mind’s eye I visualize it as “shyryuh”. And I have written a script that converts podcast transcripts from pinyin to GR. In Wenlin, I have imported a GR version of CC-CEDICT as a “supplementary dictionary”. Etcetera. 🙂 And knowing GR also means that I can read YR Chao’s books.

    • Wow! I must say, I am a little surprised. How did you get into GR? Do you feel that it really has benefits?

      • When I began studying Chinese in 1997, I subscribed to the Kenyon College Chinese listserv. I was a little out of place there as a complete beginner amongst all the professors, but the discussions were pretty interesting. A substantial minority of the academics on the list were from a cohort of students who had graduated from a handful of schools in the US which taught Mandarin via GR, and it was their influence that inspired me to get into GR. Some of the professors would write part of their posts in GR. Back then, sending hanzi in an email was sometimes problematic, and even pinyin tone marks could be a hassle. In those circumstances, GR provided a way to write Chinese that relied on only ASCII characters.

        I made friends off-list with another learner (non-academic) on the list who, like me, became a GR enthusiast, and who later wrote most of the Wikipedia article on GR (a scrupulously neutral-POV article, which has been “identified as one of the best articles produced by the Wikipedia community”). There may still be one or two colleges in the US that use GR for romanization. In 2000, the Princeton Chinese Primer series was published in both GR and pinyin versions.

        Does GR have benefits? That’s something I have considered from time to time over the past 15 years. I will try to sum up my thoughts here.
        1) GR really is a bit complicated and takes some effort to learn
        2) It truly is a pinyin world out there, and almost anyone wanting to learn Chinese with GR has to be a determined eccentric 🙂 … like the folks who still listen to vinyl using a valve amp, perhaps
        3) GR “gives an individuality to the physiognomy of words” [YR Chao] and it does help one to remember tones, but:
        a) You need to have a lot of exposure to GR-romanized Chinese
        b) Exposure to pinyin can cause interference, so you might need to be self-taught, like me, and you would need to have various software tools to convert pinyin to GR, etc.
        c) If you aren’t a good speller in English, with a strong visual memory of the shape of words, then GR may be less useful (or even harmful) to you

        To quote one research finding ~
        One study conducted at the University of Oregon in 1991–1993, compared the results of using Pinyin and GR in teaching elementary level Chinese to two matched groups of students, and concluded that “GR did not lead to significantly greater accuracy in tonal production.”

        However, in correspondence with me last year, the author of that study wrote:
        “In a two-decade retrospection, I think my conclusions should probably be re-interpreted as: (1) GR is potentially an effective means of phonological acquisition, but NOT for; (2) very fluent mature adolescents of American English! “

        My feeling is that it depends on the individual, and those who are good spellers, for example, are more likely to benefit from GR than the average student. It probably helps, too, if the learner has a certain flexibility that comes from having previously learned other languages that are written in Roman script.

        Anyway, it’s all very academic, so to speak, in my case, because in practice, I have had to become “bilingual” in pinyin and GR rather than immersed totally in GR. But regardless of its efficacy, I simply prefer GR. As someone who can read pinyin and Gwoyeu Romatzyh with equal ease, I would say that GR feels like the way Chinese was meant to be romanized. It feels more natural and “wholesome”, if that makes sense. The tones are built into the spelling of words, rather than appended in the form of diacritics. To me, pinyin feels a bit like HTML source code (the tone marks are like tags), whereas GR feels like WYSIWYG text rendered on the webpage, just as the writer intended it to look.

      • Thanks for sharing!

        It seems to me that one of the benefits of GR is that it forces the student to really pay attention to every syllable and every tone. You can do this with pinyin, but the majority of learners will not. A good teacher will make sure that beginners REALLY learn the syllables of Chinese, though, no matter whether they’re learning with pinyin, or GR, or zhuyin.

  7. That’s right. One of the scripts I wrote for myself is for looking up CC-CEDICT via GR spelling — e.g. enter “shyryuh” to find 食欲. It forces me to “include” the tones and prevents me from just entering enter a toneless pseudo-Pinyin spelling like “shiyu”.

    I also modified the Javascript for one of those online pinyin-to-hanzi IME’s, to create a GR-to-hanntzyh IME. This IME makes impossible to be lazy and write in toneless Pinyin. For Sinosplice readers to play with, I have just uploaded it here:

    It seems to work ok in Chrome now, although I seem to recall having to use IE in the past.

    To start with, try typing “shyryuh” to get 食欲.

    Another point is that a lot of one’s knowledge of pronunciation and tones can come independently of any romanization system. That is, when you know the word and what it looks like in hanzi, and you hear it (and then repeat it aloud) again and again (from a podcast and/or its hanzi transcript, for example).

Leave a Reply