How to transcribe the name of the ruler of the PRC
« previous post | next post »
This is a follow-up to "How to pronounce the name of the ruler of the PRC" (10/26/25). Surprised by the amount of dissension over how to pronounce his name and how to represent the pronunciation in romanization, I decided to try another approach. I asked all of the students (undergrads and grads) in my Fiction and Drama and in my Language, Script, and Society in China classes to write down the best way that could think of to transcribe Xi Jinping's in roman or Cyrillic letters — other than the official Hanyu Pinyin version, Xi Jinping.
Only two of the students were linguistics majors, about a dozen were East Asian Languages and Civilizations majors. The remainder were drawn from a wide variety of disciplines and fields (humanities, sciences, and social sciences) across the university. About 90% had a Chinese background (ranging in ability from minimal acquaintance to full fluency). There were a couple of students from Taiwan, a few from Cantonese and other topolect areas, one had a Korean background, and two or three had no prior exposure to any East Asian languages.
About half of them had some familiarity with Pinyin, but I told them not to use Pinyin or be influenced by Pinyin in drawing up their proposals.
All of the students have native or near-native competence in English
Here are the results of the survey of those who attended class that day:
1. Shee gee pee-n
2. Shi Ginping
3. She-tsin-ping
4. Shee jihn-ping
5. she jeen peen
6. she jean ping
7. She gin ping
8. She Jean pin
9. hi dgiŋ Phiŋ
10. Tsi Jen Ping
11. se-zing-ping
12. shi chin ping
13. See jing Ping
14. sh-e ji-ping (hold your tongue in the sh position and move it up a bit)
15. See Jin Ping
16. see jin ping
17. She-Jin-ping (she as in her)
18. I tell my Hispanic friends "Si", but my American friends say "C". I say the jing ping pair rhyme with zing.
19. cee jing ping
20. htsee jin ping
21. shee zhing ping
22. She Qin Ping
23. she-jing-ping (as in ping pong)
24. Щи джин пинь
25. Си Дзиньпин (Dungan)
26. Си Цзиньпи́н (official Russian transcription)
N.B.: Within a week or so, I will make a new post on Dungan and explain how it is that we have a Dungan speaker at Penn this year.
P.S.: Yesterday, I heard a television newscaster repeatedly pronounce "Xi" as "Gee".
Selected readings
- "How to pronounce the name of the ruler of the PRC" (10/26/25)
- "How to say 'Xi Jinping' en français" (4/16/18) — very short, but it makes me laugh every time I hear it. Must watch / listen. "Sissy Ping".
- "Sino-Russian Transcription and Transliteration" (9/17/08)
- "Putin in Russian, Mandarin, and English" (1/21/18)
Dave J. said,
November 4, 2025 @ 9:02 pm
Did nobody suggest Hsi Tsin-p’ing, or would the Wade-Giles transcription be cheating?
Victor Mair said,
November 4, 2025 @ 9:26 pm
They never heard of Wade-Giles.
Bei Dawei said,
November 5, 2025 @ 1:45 am
I vaguely remember some TV person pronouncing it as "Eleven Jinping" (mistaking "Xi" for a roman numeral)
Andreas Johansson said,
November 5, 2025 @ 2:46 am
From the perspective of getting furriners to pronounce things approximately correctly, it might have been a good idea to merge x j q and sh zh ch as sh j ch, and write the empty rhyme as something else than i. You'd have to distinguish more systematically between u and ü.
So, by that logic I guess Shi Jinping.
Jarek Weckwerth said,
November 5, 2025 @ 3:18 am
What we learn from this experiment is that (1) English speakers have trouble representing the name in writing, (2) there is considerable variability and (3) these students didn't consider the task from a pragmatic perspective of what a good transcription is. We already knew (1) and (2).
This is like asking someone who can walk to design footwork for tennis. Or asking someone who has a disease to design a test for it. The old story of "is an elephant a good zoo manager".
If we wanted to find out what a good transcription would be*, then we would have to design a bunch of transcriptions, ask many people to read them, and then analyse their pronunciations.
And from the previous post we know, for example, that a good transcription should use for the second consonant since it's more unambiguous.
(*) Unless we are looking for people who will be making transcriptions. Then this was a good test if we know what a good transcription is beforehand.
Benjamin E. Orsatti said,
November 5, 2025 @ 7:47 am
Complete outsider's opinion: How did we ever get "x" for /(h)s(h)/ to begin with? Isn't the goal of transcribing for foreigners to get them to approximate the native sounds using their own phonologies? In that case, English already has "sh" for /sh/, and even the Wade-Giles "hsi" at least suggests the presence of a "tensed" /s/, but "x", especially for us Hellenephiles, really does a number.
Andreas Johansson said,
November 5, 2025 @ 9:55 am
Using X for /ʃ/ has a long history, going back to the Middle Ages – it was even sometimes used in Middle English, I'm told. Apart from pinyin, the chief modern example is Portuguese.
Rodger C said,
November 5, 2025 @ 10:27 am
And Catalan.
Jarek Weckwerth said,
November 5, 2025 @ 10:42 am
@Benjamin Orsatti The goal was not too transcribe for foreigners but make a Romanisation. And even if it was, English is not the only foreigner language ;)
Benjamin E. Orsatti said,
November 5, 2025 @ 3:49 pm
Jarek,
Ah, let me rebut: If the goal is ROMANisation, using the ROMAN alphabet, why not do it with letters and pronunciations approximating Latin?
Ben
cliff arroyo said,
November 5, 2025 @ 4:31 pm
"And Catalan."
And Maltese as in xemx (sun, pronounced shemsh) or m'hux (he isn't, pronounced moosh)
"why not do it with letters and pronunciations approximating Latin?"
Did Latin have a sh sound? What about ch- final -ng?
Personally I always like hs- more than x- but x- doesn't bother me
Benjamin E. Orsatti said,
November 5, 2025 @ 5:08 pm
Cliff,
“Did Latin have a sh sound? What about ch- final -ng?”
Sure it did, “sci / sce”; so then you would have, “Scī Dzin Peń”…
Oh. Yes, I see. That’s not really any better, is it?
Hangul it is, then (!).
Tom Dawkes said,
November 6, 2025 @ 7:58 am
@Benjamin E. Orsatti
Later Latin — in some areas — did develop a sh sound. But in Pinyin 'sh' and 'x' are distinct sounds, so it's not really appropriate to conflate them. Neither Greek nor Latin had an 's' ~ 'sh' distinction, so that we have in the New Testament Jesus < Joshua and Simon < from Shim'on. And modern Greek has a sibilant which to English ears is neither 's' nor 'sh', and similarly for Dutch and Finnish.
In any case orthographies from Latin are very much constrained, so that foreign names will often be mispronounced. Even BBC presenters more often than not pronounce German names with -er (Merkel, Merz) as though rhyming with 'work' , even though English actually has a very similar vowel as in 'care'
Roscoe said,
November 6, 2025 @ 11:32 am
@Bei Dawei: My GPS used to pronounce the alternative name for Manhattan’s Lenox Avenue as “Malcom the Tenth Boulevard.”
David Marjanović said,
November 6, 2025 @ 2:01 pm
Like King Jong the Second?
That's rather specifically Italian you're talking about. Until around the end of the western empire, c was [k] without exceptions, and after that changed, it must still have taken a while for [s̠tɕ] to simplify to [ɕ~ʃ].
In Sardinian, none of this ever happened. Only [kj] became an affricate.
The Venetic language, by what little evidence remains, had a separate /ʃ/ (from earlier /sj/ and /sw/), and used a separate letter for it. Latin didn't, so it didn't.
And so did Latin; one hint at this is the fact that it survives as such in those kinds of Spanish that distinguish s from z/c.
Jarek Weckwerth said,
November 7, 2025 @ 5:15 am
@ Benjamin Orsatti If the goal is ROMANisation, using the ROMAN alphabet, why not do it with letters and pronunciations approximating Latin?
As you may have noticed ;) this is not done in any language that uses the "Latin" script today. For a very simple reason: The script was designed for Latin and does not match the phonemic systems of the other languages. Hence digraphs, trigraphs and diacritics; and language-specific renditions of "similar" sounds.
The specific problem here is that the solutions adopted in Pinyin do not match the solutions adopted in English. But they don't match those in e.g. Spanish or Czech either, and those don't match each other, because that's how things have gone historically.
That's why we need IPA-like solutions for serious linguistics.
BTW the awareness among the speakers of any language of the grapheme-to-phoneme relations in other languages is always very fragmentary. Are English speakers any better at guessing e.g. Italian G2P than they are at Pinyin?
Benjamin E. Orsatti said,
November 7, 2025 @ 8:39 am
Jarek,
In my understanding, Italian is one of those languages which, among Swedish, Korean, etc., believes its orthography to be 100% phonetically predictable.
B.
Jarek Weckwerth said,
November 7, 2025 @ 4:24 pm
@ Benjamin believes its orthography to be 100% phonetically predictable — I'm afraid you'll find that this is untrue for most languages with any kind of longer history of writing. Even in Italian and other languages that non-linguists call "phonetic", you can't predict the spelling from pronunciation with 100% accuracy. The same /k/ is spelled differently in caro and che; the same /ʃ/ is spelled differently in scerno and scialo, etc.
The fact that native speakers aren't really aware of this just shows that they aren't capable of sufficient reflection wrt their knowledge of Italian orthography and phonology (which, BTW, is typical of native speakers).
But we are discussing the opposite here, i.e. whether you can derive pronunciation from the spelling. Here, Italian etc. are much much better, and certainly way better than English. However, this state of affairs is language-specific: The rules for grapheme-to-phoneme conversion are very good, but different in each language. That was my point above.
In other words:
(1) Why do we have X for /ɕ/? — Because that's the convention in Pinyin.
(2) Why is that the convention in Pinyin? — Because a grapheme was needed for /ɕ/ and someone decided it would be OK to use it.
(3) Why does it not use some other convention (English, Latin, whatever)? — Because each language has its own conventions.
(4) Why does each language have its own conventions? — Because that's the traditionally accepted solution to the problem of the Latin script not being designed for other languages.
Michael Watts said,
November 8, 2025 @ 9:15 am
You appear to be making some argument that doesn't match what you're saying. Caro and che are exactly what you would predict for the Italian spelling of the pronunciations /karo/ and /ke/. Whether a phoneme is always represented the same way in all contexts within the orthography is not at all the same question as whether its representation can be accurately predicted.
Jarek Weckwerth said,
November 8, 2025 @ 10:35 am
@Michael Watts: Well, possibly true*, I'm making a weak claim ("even in Italian … not 100%") in response to Benjamin's strong claim which sort of caught me off guard — I was thinking more in terms of G2P than "predicting" the spelling from pronunciation.
But even in Italian you seem to have enough examples of established borrowings to be able to easily make the claim that you can't predict the spelling from the pronunciation. /kaki/, /kajak/, /kilo/, /ketʃup/, etc. (Mind, I don't speak Italian.)
And you can only "predict" the usual spellings if you know the grapheme-to-phoneme rules beforehand. I'm not fully convinced that that counts as prediction.
(*) And yes, on reflection I do agree that it's two separate questions ;) But still.
Philip Anderson said,
November 8, 2025 @ 2:28 pm
@Jarek Weckworth
I have no issue with counting the ability to reliably predict spelling from pronunciation or vice versa, from knowledge of the rules as “prediction”. But you are right that it can’t be done from knowing the rules in a different language. English is notorious for being unpredictable full stop.
Some letters have more or less the same pronunciation in a range of languages, or at least speakers can’t hear the difference. But some letters and digraphs have very different pronunciations in different languages, like C and CH (in English, German and Italian for instance). So it is no surprise that letters like this, including X and Q were “available” for non-standard uses in Pinyin.
Jonathan Smith said,
November 9, 2025 @ 12:08 am
The typical story about why "j" "q" "x" for this series is that these consonant letters were left over. Maybe it's true or at least truthy. But what is true and more interesting is that this series was in earlier Romanization schemes written (largely or entirely) the same way as the "zh" "ch" "sh" series with which it's in complementary distribution — an eminently reasonable solution. (Or you could write it into the [also complementary] "z" "c" "s" series to save a letter, or into the [also complementary] "k" "g" "h" series to be funky like that.) But the earlier systems were devised by Westerners or so inspired. Pinyin, on the other hand, simply re-represented 1-to-1 the consonants of the early 20th century Zhuyin system, which in turn has partial and murky but culturally pure origins in the pseudo-Seal-Script phonetic symbols devised by Zhang Binglin. So these sounds were kinda destined to have unique representations, ease of interpretation by readers of English etc. be damned… which is fine.
Xiaofeng He said,
November 10, 2025 @ 1:30 am
The pronunciation /si/ for 習 (x vs s, q vs ts, etc.) at existed in Yuan dynasty, as in Zhongyuan Rhyme 《中原音韻》, which still existed as Yunbai 韻白 or Zhongzhou Rhyme 中州韻 in Peking (Beijing) opera 京劇 from Qing dynasty till now. Wade-Giles and the Tongyong Pinyin in Taiwan (i.e. cin for 秦 qin and syuan for 宣 xuan) is close to that way. From all the good/'standard' Mandarin speakers I have met, and from the modern Mandarin TV plays and movies, I believe it is now still the de facto standard. The Pinyin standard is somehow simplified and more "northern", where there is not distinguishment between the sharp/round sounds, i.e. the Mongol or Manchu people in Yuan and Qing dynasty.
Andreas Johansson said,
November 10, 2025 @ 2:12 am
In my understanding, Italian is one of those languages which, among Swedish, Korean, etc., believes its orthography to be 100% phonetically predictable.
Swedish is far less phonetically predictable than Italian, with many sounds having multiple spellings, which are at best etymologically predictable. (And that's before going into un- or partially assimilated loans.) No-one with modest grasp of Swedish orthography can believe it's 100% phonetically predictable.
I don't think the pinyin uses of X and J are particularly exotic. As for Q, the explanation I heard somewhere is that it's based on the vaguely similar-looking Cyrillic Ч, which stands for [tʃ] and similar affricates.
ajay said,
November 10, 2025 @ 5:16 am
My GPS used to pronounce the alternative name for Manhattan’s Lenox Avenue as “Malcom the Tenth Boulevard.”
This reminds me of the kid in my history class at school who made a number of very strange assertions about the course of the Reformation that only made sense when we realised he thought Martin Luther was the same person as Martin Luther King.
Jarek Weckwerth said,
November 10, 2025 @ 7:23 am
@Philip Anderson: I have no issue with counting the ability to reliably predict spelling from pronunciation or vice versa, from knowledge of the rules as “prediction”.
I dunno. Personally, I would think "convert" or perhaps "derive" are far better terms. "Convert" in particular is what is typically used (grapheme-to-phoneme conversion etc.).
There's absolutely nothing that allows you to predict, absent prior knowledge of the writing system, that the letter S is how [s] should be represented in writing. There is some iconicity in O, and arguably you can derive a bit of Hangul from its overall principles, but "prediction" strikes me as a suboptimal term.
The underlying principle of an ideal alphabetic writing system is that each contrastive unit of sound gets a unique graphical representation. That is not the case in the normal orthographies of most languages. A single phonological merger or split after the writing system is designed breaks this principle.
Chris Button said,
November 10, 2025 @ 9:10 pm
I doubt longstanding F1 fans would protest much with "x" as a sibilant.
The "x" in the name of Belgian driver Jackie Ickx technically represents /ks/, but the /k/ part is redundant after the "k" so the "x" becomes a de facto "s".
And the "x" in Xuxa, the name of Brazilian driver Ayrton Senna's famous (ex)girlfriend, represents /ʃ/ or "sh".
Matt Anderson said,
November 11, 2025 @ 5:18 pm
I've long thought that pinyin "q" came from Albanian (where it writes the similar /c/ or /tɕ/), but at this point I no longer remember if I was taught that, if I read it somewhere, or if I just made it up myself after learning what Albanian /q/ represents. But, coincidence or not, that's at least one similar sound written with the same letter.
Matt Anderson said,
November 11, 2025 @ 5:24 pm
Corrections to the above:
I've long thought that pinyin "q" was inspired by Albanian…
…after learning what Albanian "q" represents…
Tom Dawkes said,
November 13, 2025 @ 6:19 am
On the pronunciation/orthography (mis)match in Italian. I recall granddaughter (fully bilingual English-Italian) writing "squola" for "scuola" when she was around 6.