Tag Archives: Chinese

The Chinese Characters Dictation Competition Is a Test Few Could Pass – WSJ.com

The Chinese Characters Dictation Competition Is a Test Few Could Pass – WSJ.com.

Linda and I watched this programme when we were in Shanghai as a bunch of (junior?) middle school students were tortured with various characters.

The particular word for “toad” given on the page is not in my big dictionary. That gives 癞哈蟆 (làiháma; with 癞虾蟆 given as a variant), but when I type the characters in, I get 癞蛤蟆, which, ignoring the final character, is the expected word in the programme.

This very much looks like a matter of contrary orthography so that the word which is given in the programme is no more correct than any of the others. The first character, 癞, means “leprosy” while 蛤 means “clam”, and 虾 (here pronounced há) appears to be a bound morpheme in this case, otherwise being xiā, the word for “shrimp”, which is much more familiar to me.

It’s a pity that I don’t have an etymological dictionary of Chinese because I’m also wondering whether this word has been borrowed from another language.

Mind how you pronounce your pinyin

That’s not ‘jaguar’ with an ‘sh’.

Not so long ago when I was in the Blue Bar one evening, Nick mentioned a student called /ˈʃægjʊə/1 and lamented how inappropriate his choice of name seemed to be. I was a little curious about the name and asked Nick to spell it because I’d begun to suspect it wasn’t /ˈʃægjʊə/ at all, but, in fact, 傻瓜 (shǎguā) “idiot”.

The pronunciation of Chinese cannot be successfully inferred from pinyin using the pronunciation of letters in English as a guide. It is worth knowing pinyin before coming to China so that even if the natives pretend that they can’t understand what you’re saying (or deliberately refuse to make any effort to  understand), you aren’t seriously mangling the language by using grossly wrong sound values for pinyin letters.


1. I’ve listed most of the fonts on my machine which include the IPA extensions block. If nothing appears, go to SIL and download the Doulos or Charis font.

Some brief thoughts on pinyin

The muddy world of transcriptions.
One of the things I forgot to mention about Joe Bennett’s book was the lack of his use of pinyin (拼音) for the few Chinese words he included and his assumption that because it’s not a transparent system of transcription, it’s another instance of oriental obscurantism. It’s true that you can’t know the pronunciation of Chinese from pinyin without first knowing what sound each letter represents and, in the case of some vowels, the context, but it’s also true that you can’t know how to pronounce any language from its writing system, even those in which spelling reflects phonemic-level pronunciation with a high degree of accuracy. English, as we all know, is fairly hopeless because the spelling and pronunciation of the standard language parted company long ago.
With pinyin, as with any language, you have to learn what sound each letter represents. As I said, for some vowels, the context is important because, for example, the i of 四 sì “four”, 士 shì “scholar” and 戏 xì “play” is different in each case. The first two are often described as a prolongation of the consonant, the former sounding like a high central vowel, and the latter like a rhotacised vowel; the third is the only one which is a high front vowel.
It struck me that pinyin is the sort of system of transcription which a structuralist linguist might devise within the constraint that only the letters of the Latin alphabet, as used in English, should be employed to do the job. From an English perspective, although some of the values of the letters are utterly opaque, others such as t and d seem fortuitously cunning because English t is inherently aspirated, while d, in an initial position, is half voiced so that they do sound, roughly, like Mandarin /th/ (t) and /t/ (d) respectively. On the other hand, it doesn’t work with c /tsh/ and z /ts/ because it appears that /s/ doesn’t just suppress the aspiration in /sC/ clusters in English, but also in /Cs/ clusters. While English speakers may hear a contrast between /ts/ and /dz/ as the way in which we pronounce c and z, that’s not how it works in Mandarin. In fact, I find that the pronunciation of c sounds more like /tšh/ (i.e., an aspirated English ch) than the aspirated version of the affricate /ts/.
Once you are familiar with pinyin, you know what you’re doing, but you have to surmount that initial hurdle so that you don’t keep telling everyone, as one of my erstwhile colleagues once did, that she came from “Zizilan” (to be pronounced as in English) because the x’s of xīnxīlán (新西兰 “New Zealand”) are, roughly speaking, palatal fricatives, and not /z/ as we pronounce word-initial x in English.

If you ask nicely

I might tell you the answer.

It’s ironic that in my final term with the programme I’m teaching parts of Book 4 that I’ve never taught before. As far as I’ve ever got is early in Unit 24 and then exams usually turn up. Not this year. We’re actually going to have completed most of the material in Book 4. Sort of. (I can’t explain “sort of” right now.)

Anyway, we got to the third conditional today (which is what they call unfulfilled conditions these days). These are sentences like “if my students had paid attention, they would’ve got better marks in the final exam last term”. In other words, the condition is about something which never happened in the past. There was the inevitable question in the book about how you make these sentences in your language, which got me wondering just how Chinese tackles these sentences.

My initial hypothesis was that Chinese might use some irrealis particle (perhaps a little like ἄν) in Classical Greek. But from the examples in Ross and MA (2006:278ff.), it appears that Chinese makes no formal distinction between real and unreal conditions. It even has six words for “if”, which are used in different registers:

Conjunction Usage
要是 yàoshi Formal and informal speech and writing.
如果 rúguǒ Formal and informal speech and writing.
假如 jiǎrú More formal speech or writing.
假使 jiáshǐ Formal written Chinese.
倘若 tǎngruò Formal written Chinese.
倘使 tángshǐ Formal written Chinese.

The conjunction occurs before or after the subject of its clause. As for the nature of the condition, that would appear to be determined by the context. For instance, one of the examples in the book is

rúguǒ nǐ shì wǒ, nǐ yě bù huì tóngyì tāde kànfa de.
If I were you, you wouldn’t agree with his viewpoint either.

It’s not impossible to say “If I am you” in English (e.g. in a wacky Star Trek/Stargate time-travel, meet-yourself episode), but “if I were you” is more plausible, I think, for this sort of utterance. But there are no extraneous particles in the Chinese to indicate that this is an improbable condition. Similarly, the following sentence is also translated as an improbable condition from context:

jiáshǐ rénrén dōu qí zìxíngchē huò zuò gōnggòng qìchē, huánjì wūrǎnde wèntí jiù róngyì jiějué le.
If everyone rode a bicycle or took a bus, the pollution problem would be easy to solve.

The only example of a sentence translated as an unfulfilled condition in the past in Ross and Ma is

rúguǒ bù shì nǐ bāngzhù tā de huà, tā shì bù huì chénggōng de.
If you hadn’t helped him, he wouldn’t have succeeded.

I don’t know whether this can also mean “If you don’t help him, he won’t succeed”.

Po-Ching and Rimmington (2006) don’t really have anything useful to say about conditionals in Chinese as far as I can find.

Ironically (that word again), TY Ancient Greek has also reached conditionals. Open conditions (which would be the so-called types zero and one in a modern grammar of English) are usually expressed by the indicative in the protasis (conditional) and the apodosis (main). Present unfulfilled conditions are similar to type two in English with the imperfect indicative in both clauses, but the verb takes ἄν in the apodosis. The aorist indicative is used in past unfulfilled conditions (type three in English) and, in the same way, the verb in the apodosis takes ἄν. But the imperfect may also represent the continuous form of past unfulfilled conditions.

Future remote conditions are formed by the optative in the protasis and the optative + ἄν in the apodosis. For this type of condition and present unfulfilled conditions, English only really has the second conditional. Morwood (2001:183ff.) has

εἰ ταῦτα ἔλεγες, καλῶς ἂν ἔλεγες. (Present unfulfilled)
If you were saying these things, you would be talking sense.
εἰ ταῦτα λέγοις, καλῶς ἂν λέγοις. (Future remote)
If you were to say (or, “If you said etc.”) these things, you would talk sense.

Of course, the distinction between continuous and non-continuous forms to translate these two types isn’t really valid because certain verbs in English are inherently perfective (e.g. verbs of knowing and perceiving). Other translations seem to disguise the situation so that Weir Smith (1920:526) resorts to “should…would”, which is, if I remember rightly, how Reading Greek did it.

εἰ ταῦτα ποιοίης, καλῶς ἂν ποοίης.
If you should do these things, you would do well.

But whether this is really any different to “If you were to do etc.” or “If you did etc.”, I don’t know. The use of “should” here doesn’t feel quite right either, having a slightly dated feel to it.

I was also curious to know how all this was handled in Old English, although all I have available to me is Sweet’s Anglo-Saxon Primer. The only information about conditional sentences is found in the section on the subjunctive. The subjunctive is used in hypothetical conditions such as gif mannes heafod tobrocen sie “If a man’s head is broken” or þas flotmen þe cwicne bindaþ, butan þu mid fleame þinum feore gebeorge “These pirates will bind you alive, unless you save your life by flight”.

When the condition is unreal, both clauses have their verbs in the subjunctive and, apparently, the preterite refers to present time. Thus, me leofre wære þæt ic on gefeohte feolle, wiþ þæm þe min folc moste hiera eardes brucen “I would rather fall in fight, provided that my people might possess their country”.

A verb in the subjunctive may express a condition without gif “if” by preceding the subject. I don’t know whether this is the source of Modern English conditional constructions such as “Had you asked, I would’ve helped you”.

Well, you didn’t ask, but I may have helped you anyway.

Po-Ching, Yip and Don Rimmington (2006). Chinese. An Essential Grammar. 2nd edition. Routledge: Oxford.
Davis, Norman (1953). Sweet’s Anglo-Saxon Primer. Ninth edition. Clarendon Press: Oxford.
Ross, Claudia and Jing-heng Sheng Ma (2006). Modern Mandarin Chinese Grammar. A practical guide. Routledge: Oxford.
Weir Smith, Herbert (1920). A Greek Grammar for Colleges. American Book Company: New York. (Available from Textkit.)

When I was in the shower

I got thinking about stress-timing, again.

Mark Liberman has been discussing stress timing and syllable timing following some comment made about the speech pattern of some American politician of Italian decent (Stress timing? Not so much; see the article for older links to the story). Investigating this further, Liberman seems to come to much the same conclusion that I vaguely recall Elizabeth Couper-Kuhlen1 came to in her second (?) book: stress timing in English is a rather approximate notion at best.

The idea is that languages tend to fall somewhere a long a continuum between stress-timed (stresses at equal intervals) and syllable-timed (syllables at equal intervals). I don’t know whether there’s any correlation between the mode of stress assignment and the nature of the language. For example, are syllable-timed languages more likely to have quantity sensitive stress assignment or quantity insensitive stress assignment; or is there no observable tendency here. English, which has quantity sensitive stress assignment, is generally regarded as the epitome of a stress-timed language. Apart from the usual sorts of stress patterns, I’m not sure whether there’s much going on in the phonology of English where LL = H2 means anything. In other words, on the surface such a pattern, which might be regarded as a typical marker of a syllable-timed language, is insignificant in English.3

As I’ve probably said here before, a syllable is perceived as stressed in English because of a number of features. One of those features is relative prominence. One of the reasons why we perceive a syllable is stressed may be because some adjacent syllable has a lesser degree of stress (right down to being unstressed). The former is relatively more prominent than the latter. Adjacency is also important because a stressed syllable doesn’t become more prominent when a whole string of unstressed syllables follow. One is enough.

But although the notion that stressed syllables come at roughly equal intervals seems all well and good, I have thought in the past that even unstressed syllables have duration. When you think about it, it kind of implies that in two strings of syllables such as áxáxáx and áxxxáxxáxxxx, the unstressed syllables of the latter are spoken faster than those of the former to maintain that whole notion of equal intervals between the stressed syllables. Speakers of Chinese may have problems distinguishing (unstressed4) shwa from ø, but shwa isn’t without duration.

Another matter to consider is the structure of English words and how stress is a assigned.5 I don’t know whether it’s possible to find a non-derived word in English where the syllable bearing primary stress is preceded by four syllables to the start of the word, and the initial syllable is stressed as well (i.e., àxxx[á]x or àxxx[áx]x). Thus, although English (and other so-called stress-timed languages) may not quite satisfy the standard definition of stress timing when we study the issue more closely, stressed syllables are going to occur relatively regularly in an utterance.

Actually, I think it’d be more interesting to study this phenomenon in Russian, which has morphologically determined stress. As far as I’m aware, Russian lacks secondary stress which, being phonemic in English, must contribute to how speakers perceive rhythm in English. Yet Russian is, unless my information is antiquated, usually listed under the heading Stress-timed Languages.


  1. WTF?! A search via the Google toolbar for her in Firefox got me redirected to Baidu, although a search via the Google website produced the expected results. Ditto a search for Harry van der Hulst.
  2. L = light syllable; H = heavy syllable.
  3. “Currently not significant” may be nearer the mark. Old English poetry points to a mid point between the two camps. One aspect of verse is LL or LH = H; yet certain parts of the verse allowed for an indeterminate number of unstressed syllables. This sort of pattern isn’t subsequently found in English until the 16th century when Sidney and others tried to write syllabic verses after the pattern of Greek and Latin poetry. My PhD supervisor, Chris McCully, wrote an article about this which, I think, was published in the Journal of Linguistics back in the mid 90s. Nonetheless, English verse has shown no true interest in LL = H in a very, very long time.
  4. In fact, Chinese does have shwa, but it’s always stressed. In English, shwa + sonorant is always reduced to a syllabic consonant, hence Chinese /ən/ sounds very different from it’s English counterpart (e.g. the second syllable of “sudden”).
  5. In nouns: stress the penultimate syllable if it’s heavy, otherwise the antepenultimate; in verbs, the final syllable if it’s heavy, otherwise the penultimate. I think Harry van der Hulst proposed that the language then has rhythmical left-to-right secondary stress. But without really long words, it’s hard to confirm this. From the history of stress assignment in English, I think van der Hulst is right. I believe that this pattern of secondary stress has its origins in the assignment of primary stress in Old English.

Tomorrow today

Mingtian jintian.

Another story from The Independent: Schools import China’s teachers for lessons in ‘language of tomorrow’.

Sir Cyril Taylor, chairman of the Specialist Schools and Academies Trust, said it should be seen as the key language for future generations to learn – replacing European languages.

Although Chinese is becoming increasingly important, the argument seems to be “China – 1.3 billion people – surging economy – therefore language is important”. At the moment, the language has a practical importance, but because it’s principally limited to one country, it seems to lack that one thing which makes a language important beyond its native seat – an international dimension. I think I’ve said before that after English, the next most useful language on the planet is Spanish; after that, probably French or Portuguese.

Chinese is likely to remain a minority sport in the outside world in spite of what Sir Cyril would like to believe. It’ll gain adherents, but for practical reasons, when you’re in the UK, the European languages are a more immediate concern than one eight hours away across the Earth.


And if not Chinese, how about Latin? (Id quod circumiret, circumveniat – Latin makes a comeback) It’s not the first time that Latin has been undergoing a revival in schools. Well, either that or classics in general.

Mr Mount is also adamant that learning to write in Latin is not simply ars gratia artis (6). He says there is a real quid pro quo (7) in having a Latin qualification on your curriculum vitae (8), because after all that time spent learning to distinguish a nominative from a genitive, “you’ll never get an apostrophe in the wrong place again”.

(Mr Mount is a Torygraph hack who has written some book on Latin. The numbers are for footnotes in the original article.) An interesting statement because unlike the relationship between Latin and English that once pertained, the grammar of the former being the basis for describing the grammar of the latter, things have got a little muddled. The apostrophe may be a marker of the genitive in written English, but such punctuation was never used in Latin. Therefore, quite how knowing mensa from mensae and mensarum or focus from foci and focorum is going to help with an apostrophe in English, I don’t really know.

The Tenth Don’t

A cultural interpretation.

Just recently I posted a picture of ten don’ts from Machang Lu. The tenth was the somewhat baffling 不说服务忌语 (bù shuō fúwù jì yǔ) “Don’t say service language”. As I was having a shower this evening, I suddenly realised what the tenth don’t might actually refer to. In Tongzhou, the scrap merchants would prowl the streets once in a while and dispense their cries via a megaphone. In Changzhou, they bang enamel pots with spoons. Here, they call out as they lug around their lumpen handcarts. Their cries might be “service language”, hence the injunction.

It makes sense because they can be annoying if you don’t want to be importuned.

[06.08.14. Youdao translates this as “Don’t say service JiYu” and won’t even attempt to render 忌语 into English. Is this another instance of some sort of resultative clause in Chinese? Or is it a separate clause? For example, the phrase might mean “Don’t offer services, refrain from speaking.”]

Irony bonus.

The eighth don’t is against graffiti and posting bills. What is there on the sign but un graffito: 办证刻印 (bàn zhèng kèyìn), which, I assume, is an offer of the noble art of forging stamps for official documents. These signs, with various phone numbers, are all over the place around here.

A film by any other name

This makes so much sense.

I was going through the films I’ve bought over the past three months, and wondered what the Chinese title of The Incredibles is. I know the first two characters say “superman” or “super person/people”. It says chao1 ren2 te4 gong1 dui4 which means “Super people Special Attack Team”. <span class = “sarcasm”>It’s the title the film should’ve had.</span>

Meanwhile, here’s a notice from the Phoenix Hotel in Changzhou.

Security Scattering Sketch Map

1. Please don’t worry if a fire is occurring our hotel have owned superior facilities to ensure you transmitted safely.

2. Please follow the direction rounte to the information corridor and the safeguards will take you out to the safe belts.

3. Please do not use in case of fire alarm.

4. Point profess your excellency.

In the case of the third point, it’s not clear what you shouldn’t use. You’ve been warned. And “rounte” isn’t a typo on my part.

Old linguists never die. They just can’t remember the word for it

Twenty-five Years On

Although the circumstances of my life might change, I still remain interested in languages and linguistics. I have a greater aptitude for languages than most of the foreigners I’ve met here in China. I don’t kid myself that I’ll ever be fluent in Chinese. A lot of people come here thinking they’ll learn the language, but clearly have no idea what it takes to learn one.

I tend to learn bits of Chinese when I need to be able to communicate some specific piece of information. Of course, I promptly forget the word the moment I no longer have any use for it. I like to know what the characters say because so much written information goes begging in this country because the writing system is unfamiliar.

I didn’t turn up in China completely ignorant of Chinese. I knew the facts that everyone knows about Mandarin Chinese – it’s a Sino-Tibetan language; it has the largest number of speakers in the world; it’s a tone language; it’s allegedly monosyllabic; it’s written with characters; characters have radicals and stroke counts. And that was it.

What have I learnt about it since then? Well, it’s not really monosyllabic. Most words are disyllabic because the language has a very limited syllable structure  –  (C)V(N) – which means lots of potential ambiguity. The meanings of these disyllabic words can either be transparent, partly transparent, or utterly opaque. In the last instance, this means that Chinese words have spelling. weiji (危机) means “crisis”. The character wei means “danger”, but ji can mean “machine”, “aircraft”, “opportunity” etc. In other words, weiji happens to be spelt with these characters. Dongxi (东西) means “thing”, but is written with the characters for “east” and “west”. Zuoyou (左右) can mean both “vacillate” and ”control, influence”, and is written with the characters meaning “left” (zuo) and “right” (you).

The Chinese consonant system is heavy on affricates. (This is only true of standard Mandarin; most regional varieties have simplified the consonant inventory to such an extent that pinyin is wholly inaccurate.) Curiously enough, the basic set of stop consonants – labial, alveolar, and velar – are phonetically similar to English, although phonemically the languages are different. Mandarin Chinese opts for a contrast between plain and aspirated non-continuant obstruents.

Chinese has [l], but no [r], although it has a consonant that is acoustically similar, and rhotic vowels. (Japanese goes the other way with an r-like consonant, but no [l].)

The vowel system is a bit weird with centralised and back unrounded vowels, and a rhotic vowel. The realisation of the high front vowel is determined by the preceding consonant. It’s centralised after [s]; rhotacised after retroflex consonants; but otherwise [i].

Chinese does have inflectional and derivational morphology. The inflectional system is minimal, but it’s there. Nouns have a possessive form and verbs tend to take temporal and aspectual particles. In fact, it might even be possible that there are a few more nominal inflections, even if they aren’t analysed in that way. Adjectives are actually stative verbs. And prepositions are verbs of a sort as well. Chinese word classes tend to be a little nebulous.

Syntax is SVO with some SOV-like elements. Locative phrases come first in the clause. There is a ba-construction (把) which places the direct object at the start of the sentence. There’s not much co-ordination, or even subordination. The existential verb is you (有) “is/are; have”. Chinese uses measure words (a little like “piece” in “a piece of paper”), although their range of use is often restricted and there’s the all-purpose measure word ge (个). But just to be quirky, the Chinese word for “two” when you count two things is liang (e.g. liang ge piao “two tickets” 两个票), but er (二) “two” when counting.

18.06.13. Edited the HTML, added Chinese characters, and added tags. I also added some extra information and broke up one of the longer paragraphs.