Babel, page 29
(By the way, please note that compound characters and compound words are very different things. Under 5, we saw that for ‘mother’ is a compound character, composed of a semantic and a phonetic element, but it is not a compound word, consisting as it does of just one syllable: Mā. Here, under 6, we encountered a compound word, (XIÀNGSHÙ) for ‘oak’. It consists of two full characters, both of which are fully pronounced. As it happens, both characters of this compound word are also compound characters, but that is a mere coincidence.)
7. All Chinese languages are written the same way
What’s true. Until 1956, speakers of all Chinese languages used (nearly) the same characters, usually with the same meanings. As a result, two Chinese people who were both monolingual in two different Chinese languages (say, the Beijing dialect of Mandarin and the Cantonese dialect of Yue) would understand most of each other’s writing even though they would not understand each other’s speech. Even within Mandarin, dialects can be so different that people may have trouble recognising some particular word; writing it down will solve the problem. That’s why Mandarin-language films are subtitled in Mandarin: if native speakers of, say, Cantonese do not catch a spoken Mandarin word, they will still usually be able to read it.
All of this is largely still the case today – with the added facilitating factor that everybody has learnt Mandarin at school so that even those who can’t really speak it will still be literate in it. The situation can be compared to what occurs when English speakers with broad and very different accents, say Appalachian and Scouse, run into a comprehension roadblock: writing down what they’re saying will clear up the problem. However, Mandarin and Cantonese are much more different from each other than any two varieties of English.
But then there’s this. There are some grammatical differences between Chinese languages (word order and the use of certain particles) which show up in writing. Also, some Chinese languages, especially Cantonese, have developed special characters for words not used in Mandarin. On the other hand, most Chinese languages are rarely written at all.
Mandarin (and English) subtitles are standard on Chinese films. This is a Taiwanese rom-com called My Egg Boy featuring a chef and his dog.
More importantly, the People’s Republic in 1956 simplified thousands of characters, while in Hong Kong and Taiwan the traditional characters have been retained. Many simplified characters are so different from their traditional shapes that somebody literate in one cannot easily recognise the other.
On balance, however, there’s more truth to this commonly held idea about Chinese writing than to most of the other ones listed in this chapter.
8. Chinese characters are great for punning
What’s true. Mandarin is the ideal language for wordplay, because it’s rich in God’s gift to the punster: homophones.
But then there’s this. The great opportunities for wordplay arise not thanks to the characters, but in spite of them. On paper, most of the Mandarin homophones are easily distinguished. A pun like the one about the duck who orders a beer and tells the bartender ‘to put it on my bill’ would never work in written Mandarin. Two words might sound identical, like bill for ‘beak’ and bill for ‘cheque’, but they would be represented by different characters.
On the other hand, in Mandarin you can occasionally use an incorrect character that is homophonous with the intended meaning, and still get the message across. That would be like writing The Gnu Whirled instead of The New World – computers will be stumped, but most human readers won’t, or not for long anyway. Punning opens up great possibilities to circumvent the Great Firewall (as the People’s Republic’s massive online censorship apparatus is known), which largely relies on computer intelligence to suppress all sorts of utterances; not just political protests, but also simple profanity. The most famous example is CĂONíMĂ, ‘grass mud horse’, which was, until the censors got wise, an alternative way of saying CÀO Nǐ Mā – ‘do your Mom’, in bowdlerised translation. Grass Mud Horse has now even become the title of an online ‘glossary of memes, nicknames and neologisms created by Chinese netizens and encountered in online political discussions’.
9. Mandarin would be better if it ditched characters
What’s true. There’s no denying that learning to read and write characters is much more time-consuming than learning an alphabet, not only for second-language learners, but also for native speakers of Mandarin.
But then there’s this. Even if switching to the Latin alphabet could be proved to be a highly beneficial move, it still wouldn’t happen. And that’s not because Chinese culture is particularly conservative, or some such Orientalist cliché. Rather, it’s because all cultures are conservative when it comes to writing. Even small spelling reforms stir up strong emotions. Big reforms happen only in revolutionary times – think of Turkey under Atatürk. But it didn’t happen in China under Mao (though he toyed with the idea), and it’s not going to happen until the next revolution – maybe not even then.
Might the Chinese be right to stick to their ‘awful writing system’, as the US Chronicle of Higher Education put it? The obvious alternative would be pinyin, the Latin transcription system that was developed in Mao’s day, and which has achieved wide currency, both among students of Mandarin (mostly to find out how characters are pronounced) and among native speakers (mostly to input characters on to telephones and computers). But while pinyin painstakingly indicates the tone of each syllable, as in the famous quadruplets Mā, Má, Mǎ and Mà, it does not differentiate between Mandarin’s many homophones, that is to say, all those items that sound perfectly identical, including the tone. As a result, pinyin would produce many more misunderstandings than the character script.
Or so the reasoning goes. But wait, not so fast: pinyin has something that the character script sorely lacks, namely spaces. What we call homophones in Mandarin are mostly syllables that sound the same, not words. In character script, it’s not immediately obvious whether a character is a word in itself or a part of a longer word. In pinyin, on the other hand, no such ambiguity exists. Earlier on, we saw that XIÀNG could mean ‘oak’, ‘statue, ‘(in the) direction (of)’, ‘elephant’ and ‘neck’. But in fact, Mandarin speakers often do not say XIÀNG for any of these concepts. Just like they frequently say XIÀNGSHÙ for ‘oak’, they will frequently use DÀXIÀNG or ‘big elephant’ for ‘elephant’, JǐNGXIÀNG or ‘neck-neck’ for ‘neck’, DIāOXIÀNG for ‘statue’, where the DIāO part means ‘to engrave’, and FāNGXIÀNG for ‘direction’, where FāNG means ‘place’ or once more ‘direction’.
In pinyinised Mandarin, these words are instantly recognisable as words, whereas in character script they could simply be two words that happen to be sitting next to each other. As a result, pinyin leaves far less scope for ambiguity than might seem at first sight. According to Chinese linguists, quoted by the sinologist William Hannas, no more than around one per cent of Chinese words are homophones. They found seventy monosyllabic words whose different meanings, 164 in all, were likely to create real confusion, as well as thirty-nine problematic polysyllables with eighty-two meanings. Given that pinyin is a highly regular spelling system, words that are doppelgangers in speech (homophones) will also be doppelgangers in writing (homographs).
However, the problem can easily be solved. European languages have homophones too: in English, think of there, their and they’re, rode, road and rowed, here and hear. Mandarin homophones could easily be distinguished in writing by adding a silent letter, like the silent letter that distinguishes morning from mourning. Granted, this prop would make it somewhat harder for children to learn pinyin. Compared to memorising characters, however, it would still be as easy as breaking a Ming vase.
But probably this additional prop wouldn’t even be necessary. Vietnamese, too, has many homophones. Unlike pinyin, Vietnamese doesn’t mark word boundaries, since syllables are usually written separately. Even so, the Vietnamese seem fine with their script.
10. Now you know everything about characters
Not at all true, I’m afraid. Characters are so different from all other writing systems that they raise more questions than I can answer in this chapter. How do you place words written in characters in some kind of order (I’m avoiding the word ‘alphabetical’ here), for instance in a dictionary? (It involves counting strokes.) How do you differentiate between two homophonic characters in speech, without writing them down? (By mentioning a familiar word in which it appears, a bit like saying, ‘weigh as in heavyweight, not as in highway’.) Is it possible to describe a character without writing it down? (The strokes have names, but it’s usually more practical to refer to the two component parts that most characters consist of, as discussed under 5, above.) How is Chinese rendered in Braille? (By writing pinyin in Braille.) Et cetera.
Once mastered, there’s no limit to the creative use of Mandarin: Lego is a bit of a challenge, but baristas will find infinite new outlets for their art.
Also, there are plenty of other myths around, including the following: ‘Every character represents one syllable.’ (There are hundreds of exceptions, though the Chinese government doesn’t accept most of them.) ‘New characters are no longer being created.’ (They are, both officially and ad hoc.) And then there’s this one: ‘Japanese too is written in Chinese characters’.
Is it? That deserves a chapter in its own right.
* Or perhaps they would. Certainly Mandarin, if sinologist David Moser is to be believed. Do yourself a favour and read his very funny article ‘Why Chinese Is So Damn Hard’: http://bit.ly/MoserMandarin.
* SēN is one of many Chinese characters that can be said to have a meaning without being – in the modern language anyway – an independent word. Some examples of this same phenomenon can be found in English: the were in werewolf means ‘man’, the ly in quickly is derived from a word meaning ‘body’ and ceive in receive and other verbs can be said to mean, or have meant, ‘seize’. For those who like their linguistics well-spiced with jargon, the name for these bits is ‘bound morphemes’.
* Technically, this list of somewhere between 201 and 214 items contains radicals. A radical is not exactly the same thing as a semantic component, but for present purposes, it’s a good-enough approximation.
2b
Japanese (revisited)
A writing system lacking in system
If London’s King’s Cross station can have a platform 9¾ (and no longer just in fiction), surely a book can have a chapter 2b? My reason for inserting one here is because, before moving on to the world’s most widely spoken language, I’d like to revisit a smaller giant, Japanese. One of the things that make the language exceptional is a system that, if not magical or fictitious, is certainly extravagant and harder to learn than any spell, curse or charm. I am talking here about Japanese writing. The reason I didn’t discuss it in chapter 13 is that it’s based on the Chinese character script, which in itself is quite a challenge, as we’ve just seen.
‘Based on the Chinese character script’ should not be interpreted as ‘nearly identical to it’, for Japanese writing has many more tangles, snarls, coils and knots than Chinese – so much so that it’s widely considered the most complex writing system currently in use. So let’s walk straight into the seemingly impenetrable wall of Japanese writing to see if we can magically get to the other side.
Kanji and how to pronounce them
The earliest texts in Japanese were written entirely in the Chinese character script, which was introduced to Japan in the fifth or sixth century by Korean scholars. Unlike the Vietnamese and the Koreans, who no longer use the characters, the Japanese never replaced them, but instead added plug-ins. No writing system based on Chinese characters is going to be simple. But since Japanese and Mandarin are fundamentally different in both structure and basic vocabulary, the character script was not particularly suitable for Japanese to begin with. As a result, its adoption for writing Japanese had profound complications.
So what happened when the Japanese decided to use Chinese characters – or KANJI (‘Han letters’), as they call them?* For one thing, the clues to their pronunciation were lost. As we saw in the previous chapter, most characters consist of a semantic and a phonetic component, giving the reader clues to their meaning and pronunciation. The semantic part holds good in Japanese too, but the phonetic one doesn’t. After all, these characters are now pressed into service to represent Japanese, not Chinese, words, and there’s no reason why words sounding similar in one language should also do so in another. To return to the classical example: if the character for ‘mother’ visually refers to ‘horse’, that’s because the words sound similar in Mandarin, but in English they don’t – nor in Japanese. Therefore, children and foreign students have an even harder time in Japanese memorising the relationship between the visual shape and the correct pronunciation than they do in Chinese. In an effort to make writing easier, several post-war governments have published lists of ‘regular-use characters’, thereby standardising the script and limiting the total number. Even so, there are currently as many as 2,136. In practice, at least another thousand are still in use.
So does the student of Japanese have to learn the correct pronunciation of 2,136 characters? If only. Many of the characters have more than one ‘reading’ – that is, more than one meaning with a pronunciation of its own. Usually, one of these is authentically Japanese. For instance, can be pronounced as /te/, giving us the Japanese word for ‘hand’. This is the native Japanese reading. But in the compound (literally ‘touch hand’, meaning ‘start’), the second character is pronounced /shu/ rather than /te/. This is based on a Chinese pronunciation of many centuries ago, when the word was borrowed. The first half of the word, , is sounded as /chaku/, based on a long-outdated Chinese pronunciation, /chak/. But again, this character can also represent a native Japanese word, as exemplified by the compound (‘kimono’, literally ‘a thing to put on’), where it’s pronounced as /ki/.
How a ‘chakubutsu’ is almost a kimono
Two entirely different pronunciations per character, one native, the other imported. Pretty bad, huh? But it gets worse. Some characters have not one native reading but two, and a few have even more. More importantly, many were borrowed two or three times over, in different periods and from different regions of China, resulting in as many different pronunciations. Not all of the 2,136 ‘regular-use’ characters can be pronounced in many ways, but a great number of them have two readings that are in everyday use and one or two more that occur in specialised jargons only, for instance in Buddhist religious writings. The character , which means ‘swim’, has a native pronunciation rendered in Latin script as OYO, but a Chinese-derived pronunciation EI. It appears in ‘to swim’ (, OYOGU) and in ‘swimming style’ ( EIHō). It’s rather as if English were to spell the words ‘swimming’ and ‘natation’ the same way. Some characters go one better by amassing a lot of readings. The record is held by the notorious , which has dozens, including nine in indigenous Japanese words alone and many more in borrowings from the Chinese, with a panoply of meanings ranging from ‘give birth’ to ‘raw silk’ and ‘student’.
What all this means is that reading Japanese involves a continual decision-making process: pronunciation depends on context. The characters of the word ‘kimono’, , could also be sounded as /chakubutsu/, but that doesn’t convey any meaning; the reader has to pronounce it as /kimono/, which of course does. The English language too has a few dozen words where the correct pronunciation must be deduced from context. These so-called homographs include sewer (rhyming with either lower or viewer), sow (rhyming with either cow or low), the well-known read (rhyming with either lead or lead – sorry, make that bead and bed) and, in honour of this chapter’s protagonist, sake (rhyming with make or Iraqi). But in English, your average text has very few such potential pitfalls; in Japanese, there are alternative readings for the majority of characters.
Or am I making this sound harder than it really is? After all, if is pronounced as /te/ whenever it’s a separate word but as /shu/ in the compound /shuchaku/, the reader is best advised to focus on whole words rather than on separate signs. That’s more or less what we do in English as well: we don’t know how to pronounce the letters CHA until we have seen the word they’re part of, be it CHARACTER, CHAPTER, CHAMPAGNE, CHAOS, CHAFE, CHAISE, CHA-CHA, CHALK or GOTCHA (or even CHANUKKA or CHALYBEATE). True enough – except that Japanese doesn’t mark word boundaries, as English and most other languages do: there are no spaces. This means that two characters that sit right next to each other may or may not belong to the same word. Any skilled Japanese reader can nonetheless tell which ones do and which ones don’t, but it requires paying close attention to context. Reading Japanese is like reading English sentences with lots of SEWERS, READS, sows and SAKES in it.
Keeping the endings happy
While the 2,000-plus characters are the hardest part to master, the intricacies don’t end there. In Chinese, words have no grammatical endings, so there are no characters to write them. Japanese on the other hand has lots of endings, and writers noticed very early on that ignoring them would make their texts nearly unintelligible. What to do?
Their first stab at a solution was by using characters that sounded like the endings, regardless of their meaning. To better understand what this was like in practice, imagine we were to do the same thing in English. Our language has some grammatical endings too, such as -ing, so if we had by some accident of history adopted kanji writing, we’d have felt the same need as the Japanese. So how would we write our words ending in -ing, say buying? ‘Buy’ would be , because that’s what this character means. (The Mandarin pronunciation is /mǎi/, but no matter.) The -ing bit is slightly problematic, as there is no character pronounced /ing/. But borrowing a foreign script always involves a degree of compromise, so we’ll just make do with one that is pronounced /ying/: (and never mind what it means in Mandarin). Hence, would be the correct spelling for ‘buying’.
