How rapidly and radically can a language evolve?

« previous post | next post »

[This is a guest post by Alex Wang, a long-term resident of Shenzhen, China]

I was wondering if there have been any studies on how readily a language can absorb new elements and features.

Yesterday at the Pacific Coffee shop near where I live, by chance I struck up a conversation with a professor who teaches economics at the local Shenzhen University.  He heard me speaking with my younger son in English and, when I went to attend my older son, he struck up a conversation with my younger son.  I suppose he was curious about how my younger son's oral English skills were so “good”, since he has a daughter who is around the same age as my older boy.  It would seem many locals want an English speaking friend for their children so as to have an environment to practice.

Anyway, the point is that, after we started speaking in English, I noticed he spoke English pretty fluently.  I asked if he had studied abroad.  He said no. He watched TV shows and he wrote most of his research papers in English.  After I remarked that’s interesting, he explained that this was quite normal as some English economic terms were easier to remember and more natural.

I have long suspected this, as I have spoken to many professions such as doctors, chemists, etc., and they all say the same thing.

My view is that, by using characters rather than pinyin, the language won't evolve in a natural way. As discoveries in all fields — astronomy, biology, physics, chemistry, culinary items, music types, peoples' names, etc. — are being made in an exponential fashion, the creation of words will become ever increasingly bièniu 别扭 / 彆扭 ("awkward; difficult to deal with") until more and more locals turn toward another language.  Translating research papers, literature, and so forth will fall further and further behind unless perhaps the "machines" become better at it.  Even so, already for popular books like Harry Potter in Chinese, the names of people, spells, and items are awkward for locals. Whereas when I read names such as Hu Jintao or Deng Xiaoping, it's not awkward at all.

You can see that a profound linguistic transformation is already happening in the oral realm.  Many people are just using English words inserted directly into Chinese sentences. The use of acronyms, as discussed in prior Language Log posts, is very widespread.  For example, in the US we say that a picture was photoshopped. Here everyone uses PS, e.g., “zhège zhàopiàn PSle 这个照片PS了" ("this picture was photoshopped") or Ptú P图 ("photoshopped picture").

A reverse example is the word "MA LA" to describe Sichuan food.  Even among expats here, we just use "MA LA", as it's too cumbersome to describe in English while speaking that something is "spicy and numbing".

Many kids now use the actual English names of people in their Chinese oral conversations instead of transcribing them syllabically as if they were written in characters.  It's only a matter of time before it's easier for them to type the English name directly than try to remember the Chinese character version of an English name as more and more kids learn English from pre-school.

Finally, I am reading a book titled Alexander of Macedon by Peter Green.  I couldn’t even begin to understand how such a book could be translated felicitously into Chinese, given that there are so many proper nouns of people and places; it would seem reading all of these names transcribed into characters would be extremely bièniu.

An example. “Artaxerxes used them against Cyrus. Darius clearly thought he had found the magical formula for victory. The plain at Cunaxa, some sixty miles north west of Babylon.…”

Now an English reader not knowing those names still could read it without feeling bièniu.  However, when words are created like “salad” with characters meaning "sand" and "pull" (shālā 沙拉), or the way names of people constructed with Chinese characters that have all sorts of irrelevant meanings, it becomes totally burdensome and odd to read even for native speakers of Chinese.  I know this firsthand, as my wife started reading the Chinese translation of an English encyclopedia with my older son for some mother-son time and even she found it utterly frustrating.

I suppose that's enough rambling for now!  I just feel that an implosion is coming within the next few decades.  Like a giant star collapsing.


  1. Joyce Melton said,

    July 28, 2017 @ 12:21 am

    From historical evidence (the great vowel shift, the creations of Afrikaans, Hawiian Pidgin and Nicaraguan sign language) it takes a generation or two to make such a great transformation in a "language" that one can say it is a new thing. Not sure how to quantify something smaller than that.

    My own life experiences suggests that childhood fads and pop culture memes have a deka-life of about eight years. That is within eight years of noticing a pop-culture or playground phenomenon, 90% of them will be extinct and forgotten. My theory is that it takes about eight years for originators and influencers to age out of their most creative and influential stages and be replaced by a new crop of leaders and innovators.

    It's about that long between puberty and adult responsibilities in many cultures.

  2. Bathrobe said,

    July 28, 2017 @ 1:26 am

    If you check out Wikipedia, you will be surprised how quickly Chinese comes up with equivalents to recent English terminology. Even where an English term is (semi-)borrowed, a Chinese term is quite likely to usurp its position within a few years. One example is 英特网 yīngtè-wǎng 'Internet', which gave way to 互联网 hùlián-wǎng after a few years.

  3. Anonymous Coward said,

    July 28, 2017 @ 3:56 am

    “Artaxerxes used them against Cyrus. Darius clearly thought he had found the magical formula for victory. The plain at Cunaxa, some sixty miles north west of Babylon”

    That's not the best example to illustrate the challenges faced by contemporary Chinese script… 3/5 of the proper names have established, idiosyncratic and extremely well-known Chinese versions that are not going to go away even if the language switches into Pinyin.

    Cyrus → Jūlǔshì
    Darius → Dàliúshì
    Babylon → Bābǐlún

    Xerxes is conventionally Xuēxīsī, so Artaxerxes is Āěrtǎ-xuēxīsī by analogy, prefixed with the automatic transcription for Arta-.

    This leaves only one proper name for which a translator needs to scratch her head a little bit and which can be profitably replaced with an unassimilated representation.

  4. Anonymous Coward said,

    July 28, 2017 @ 4:03 am

    For an individual that Anglophones call /sairəs/, Russians call kir and Kurds call Kûruş, it's half-expected that the Chinese will not be bewildered with an idiosyncratic name written in characters like 居鲁士.

  5. David Marjanović said,

    July 28, 2017 @ 4:29 am

    I was wondering if there have been any studies on how readily a language can absorb new elements and features.

    These are two different questions:
    – How quickly can a language take up a bunch of loanwords or create new words of its own?
    – Do Chinese characters have an effect on this speed?

    …and the headline led me to expect something entirely different, like "how quickly can grammar or pronunciation change" – much as in the first comment.

    英特网 yīngtè-wǎng 'Internet'

    Some 15 years ago I was taught another version of the same, some kind of yinda-wǎng (I have long forgotten the first two tones and characters).

  6. Victor Mair said,

    July 28, 2017 @ 6:06 am

    "If you check out Wikipedia…."

    Wikipedia is BIG! Please give a link or two.

  7. Victor Mair said,

    July 28, 2017 @ 7:14 am

    @Joyce Melton

    I like what you say about "a deka-life of about eight years", but I'm not sure that I understand exactly what it means, though I sort of grasp the general import.

    Some links in my quest to understand it better:



    How did it come to have the meaning attributed to it in The Urban Dictionary? An anabolic steroid?

  8. languagehat said,

    July 28, 2017 @ 8:29 am

    When I saw the post title, I thought there was going to be a reference to Robert Sheckley's wonderful story "Shall We Have a Little Talk?" (a representative of Earth's evil capitalists goes to an alien planet to negotiate a purchase of land, but the language changes every day, in both grammar and vocabulary; at one point, he shouts "Stop agglutinating!"), but there wasn't one, so now there is.

  9. Sean Richardson said,

    July 28, 2017 @ 8:43 am

    Regarding deka-life, that one was a puzzler to me too until I read "90% of them will be extinct and forgotten", which made it clear that deka-life must be on the paradigm of half-life … but that only gets confusing again. The way half-life is defined as the time at which only half of the original radiation remains, would not the word for when only 1/10 remains be deci-life?

    But, learning English in the 1970s, I heard decimate in contexts that meant almost completely destroyed, i.e., ~1/10 remaining, and was surprised and reached for a dictionary when I read accounts of battles where it was used to mean ~ 1/10 destroyed. Apparently since then that shift in meaning has only continued, to the point where dictionaries give the latter meaning only as historical now. ( ;

    So, blame contronymity? Could deka-life have been settled upon to avoid using deci?

    Which raises the question, is English in particular not the best language to borrow from when it come to numerical comparators, or is this kind of potential confusion a feature common across languages?

  10. Dan Blum said,

    July 28, 2017 @ 9:22 am

    That's actually "Stop agglutinating, you devious dog!"

    Linguistic discussion of the story at Tenser; said the Tensor.

  11. languagehat said,

    July 28, 2017 @ 9:27 am

    Thanks! (Although now you've reminded me how much I miss Tenser, said the Tensor.)

  12. Guy_H said,

    July 28, 2017 @ 9:36 am

    Its a bit tangential to the OP, but the Chinese translations for Harry Potter are honestly awful. It relies a lot on transliteration rather than conveying the clever meanings that JK Rowling uses. For example, Slytherin is 斯莱特林 (si lai te lin) in the simplified version and 史萊哲林 (shi lai zhe lin) in the traditional version. Hufflepuff is 赫奇帕奇(he qi pa qi) in simplified and 赫夫帕夫 (he fu pa fu). It doesn't capture anything of the original feeling. I think the spells are more successful, because the translators often ignored the original sound and focused on meaning (e.g. expecto patronum! is 呼神護衛!), although it still sounds clunky.

  13. Bathrobe said,

    July 28, 2017 @ 10:01 am

    Wikipedia is BIG! Please give a link or two

    Basically, if you look at Wikipedia articles on recent subjects, you'll generally find that there is a corresponding Chinese-language article using Chinese terminology. For instance, 'string theory' (which I realise isn't terribly recent but is nevertheless a relatively new term) has a Chinese article at 弦理论. I notice this because a language like Mongolian lacks both a Wikipedia article and a standard term for 'string theory'.

    On the whole, Chinese does pretty well coming up with terms for relatively recent concepts. Japanese does pretty well, too, often by borrowing and katakana-ising the English term.

  14. Victor Mair said,

    July 28, 2017 @ 10:47 am

    "Japanese does pretty well, too, often by borrowing and katakana-ising the English term."

    That's a completely different strategy from the Chinese approach of first hanzi-izing the sounds of new terms (yīngtèwǎng 英特网 ["Internet"]) and then sometimes (but by no means always) translating them (hùliánwǎng 互联网).

    Katakana-izing is closer (but by no means identical) to the approach advocated by Alex Wang. And then there's the whole problem of proper nouns with which Alex was concerned. Seldom would it be suitable, or even possible, to translate names like Donald Trump, Barack Obama, Jakob Ludwig Felix Mendelssohn Bartholdy, or Johann Sebastian Bach.

  15. Alyssa said,

    July 28, 2017 @ 10:55 am


    That's a shame, there's so much charm in the names Rowling uses. I know these sorts of things are subjective, but in my opinion it's far more important to get across the "slithering" in Slytherin and the "huff and puff" in Hufflepuff rather than trying to replicate the sounds.

  16. Joyce Melton said,

    July 28, 2017 @ 11:29 am

    Is deka-life a useful term? How long might it take to enter the lexicon? As far as I know, I invented the term in the middle of typing that sentence last night.

    Why didn't I use deci-life instead since deci- is the established prefix for one tenth? I thought of that, but decided I liked deka-life better, partly that I just liked the sound better. Also deci- is the SI standard prefix for the measurement of physical qualities. Deka- goes back to the original Greek word which felt more appropriate. I also wanted to avoid the fringe of the decimate controversy.

    Getting back to Victor's original post, I do rather expect Chinese to change more than other important world languages in the next few decades because China is changing more than other nations.

  17. Bathrobe said,

    July 28, 2017 @ 11:48 am

    As discoveries in all fields — astronomy, biology, physics, chemistry, culinary items, music types, peoples' names, etc. — are being made in an exponential fashion, the creation of words will become ever increasingly bièniu 别扭 / 彆扭 ("awkward; difficult to deal with") until more and more locals turn toward another language.

    The point being made was that Chinese word creation will find it difficult to keep pace with new discoveries. I disagree with this. Chinese has a remarkable tendency and ability to use native morphemes to render foreign terminology and does keep up remarkably well with new discoveries.

    The other problem, transliteration, is a real one. Transliterating foreign names into Chinese characters is a painful exercise. It isolates Chinese speakers from foreign names behind a wall of strange syllabifications and even stranger characters (although there is a lot of regularity in the characters used). It is also quite deadening in translating literature (as in Harry Potter).

    My point about Japanese was simply that Japanese is also successful in keeping up with new terminology although it uses a different approach from Chinese — transliteration into native syllabic form. I was not claiming that the two systems were equivalent.

    As to whether Chinese vocabulary is 'unnatural' in discussing concepts borrowed from the West, I'm not qualified to comment. However, I don't think that transliterating English words into katakana (as in Japanese) or the wholesale borrowing of English words into Vietnamese or Russian words into Mongolian is necessarily a better approach. The only result is a huge swathe of vocabulary that doesn't make sense unless you know word building in the foreign source.

    There is always a tension between purism and heavy borrowing. Purism allows the creation of vocabulary that is relatively accessible to speakers of the language. Wholesale borrowing tends to make the language impenetrable to ordinary speakers. On the other hand, for people dealing in fields that rely heavily on specialist knowledge in a foreign language, it is probably more comfortable to use vocabulary from that foreign language than 'translate' it word by word into their own language. Puristic vocabulary is likely to be seen by people in the field as simply a clumsy rendition of the English, which is the real 'language of thought'.

  18. LAL said,

    July 28, 2017 @ 11:48 am

    @Victor : Re Deca-How did it come to have the meaning attributed to it in The Urban Dictionary? An anabolic steroid?

    Decadron (Deca) is a brand name for dexamethasone a steroid widely
    used in medicine.

  19. JK said,

    July 28, 2017 @ 3:07 pm

    I can't remember the last time I read 英特网 in print, it seems like everyone just calls it 网络 nowadays.

    I went to check what Baidu has to say and accidentally typed "yintewang" and 因特网 came up:因特网/114119

  20. Jerry Friedman said,

    July 28, 2017 @ 4:59 pm

    To mention a writer who gives translators difficulties similar to Rowling's, Tolkien insisted that his place names, personal names, and words hobbit and orc had to be left as he wrote them, with English "local colour". In a letter to his publisher on a Dutch translation of The Lord of the Rings, he wrote, "I am no linguist, but I do know something about nomenclature, and have specially studied it, and I am actually very angry indeed." To judge by the names in Wikipedia articles on LotR in various languages, his insistence accomplished very little. I wonder how he thought his names should be handled in Chinese.

  21. Bathrobe said,

    July 28, 2017 @ 6:15 pm

    Incidentally, the idea of just "letting them use English" is a fair-seeming but insidious and treacherous proposal. It opens up a can of worms that is potentially even worse than the problem it is trying to solve. It is one of the most simple-minded, wrong-headed suggestions ever made, usually by people who haven't thought the consequences through.

    The Latin alphabet has 26 letters, which seems simple enough but conceals real problems of spelling and pronunciation. 26 letters is not sufficient to write even English. Thanks to its history, English spelling is a train wreck. How are Chinese speakers, brought up on the regular system of pinyin, supposed to deal with English words like Graham? Or Urquhart? Or Campbell? Is foisting these kinds of spelling on Chinese readers really going to help them?

    But it gets worse. Those 26 letters are used for languages all around the world, and all of those languages have their own conventions for reading those letters. English speakers are notorious for their ability to mangle the pronunciation of languages other than English. How are Chinese speakers supposed to deal with this? How should they be expected to read Einstein ('ainstain' or 'ainshtain') or Zuckerberg ('tsukerberg' or 'zakerberg')? Many languages employ diacritics, which English speakers feel free to ignore. How are Chinese speakers supposed to deal with Dvorak (or is that Dvořák?) or Ho Chi Minh (or is that Hồ Chí Minh?).

    The simple proposal to use English words is a slippery slope. It leaves Chinese wide open to importing foreign vocabulary in the original spelling. If you think the Chinese devotion to Chinese characters is a strange religion, the fetish for 'spelling' in languages that use the Latin alphabet is voodoo run wild. Languages that borrow other people's alphabets are often fatally tempted to use the spellings of the source. While some languages have taken steps to force the spelling of borrowed terms to conform to local norms (e.g., German, Turkish), this tends to break down over time. Eventually people start importing vocabulary in the original spelling because it is 'more authentic'. So while German has tried to Germanise the spelling of borrowed words, as in Frisör, the tendency is to use the original (and more prestigious) French spelling Friseur. Now, borrowings from English are generally spelt the English way in German. Borrowings from Russian are spelt the Russian way in Mongolian. The result is to introduce new irregularities into the spelling system. People who worship English (or Russian) may love this, but it wreaks havoc with spelling over time. Since the whole reason for adopting English words is that Chinese characters are too complex, difficult, and anachronistic, what is the point of replacing them with something that is possibly even worse?

  22. Alex said,

    July 28, 2017 @ 6:46 pm

    Its the script that I have an issue with.

    If Jūlǔshì, Dàliúsh, Bābǐlún were used in the book my mind would do the same thing as it did with the English names. It would glance over each of them as one unit and due to the capitalization move onto to the rest of the sentence.

    Spacing and capitalization make a big difference for me.

    Perhaps the native reader of Chinese is used to seamlessly subconsciously disregarding the most used meanings of the characters when reading proper nouns and their mind switches to using the characters as phonetic based sounds.

    Perhaps some native readers of Chinese can comment.

  23. flow said,

    July 28, 2017 @ 7:33 pm

    @Bathrobe Yes, I fully agree. What the Chinese writing has too many of—individual signs at various graph(em)ical levels—English (and other languages using the Latin alphabet) has too few of. There's a stubborn resistance not only against adopting new letters, but also against reproducing those variations there are. That it can in principle be done is shown by the letters C/G, I/J, U/V/W, which are not (all) classical; unfortunately, some hopeful additions like Ðð, Þþ, Ŋŋ, Øø are not getting the attention they deserve, IMO.

    This paucity of the Latin alphabet and the internal tensions it produces can be held responsible for the unfortunate choices that were made when Pinyin was designed (still in and by itself not such a bad (basis for an) orthography; Wade-Giles is considerably worse); hence we get the rather idiosyncratic initials of qun, zhun, xun, the unhappy extraneous -g in hang, the digraphs sh-, zh-, ch- (whose -h- is entirely different from the lone initial h-) and the confusing role of -i in xi, ji, qi / shi, zhi, chi / si, zi, ci, ri. (Earlier plans that used some new letters were not upheld in later versions). Moreover, a more recent victim has surfaced that parallels the downfall of æsh, eð, þorn (and ƿynn) in English, namely nü, lü, which are often rendered as nv, lv nowadays, pressing a letter with otherwise rather predictable consonantal interpretation [v, f] into service as a vowel [ü]. For the lack of a nail.

    We may lament the difficulties of Chinese writing day in day out but the sobering fact is that even neighboring European cultures find it often too difficult to get that clumsy äöüßøæł of the next country on the map written out correctly, and it would seem that even the great China with its strong domestic electronics sector finds it too difficult to supply their citizens with keyboards that have a ü key, Pinyin be damned.

  24. languagehat said,

    July 28, 2017 @ 7:41 pm

    the digraphs sh-, zh-, ch-

    I've never understood the animus people have against digraphs; the purist "one symbol per sound" attitude has produced some horrific alphabets (half the languages in Africa got ruined by "scientific" alphabets a few generations back). Digraphs are an excellent and parsimonious solution to many a problem of representation.

  25. flow said,

    July 29, 2017 @ 4:17 am

    @languagehat—Yes. Digraphs are a parsimonious solution that I actually do like. I also like diacritics.

    As for digraphs, things get difficult when they get used in non-minimal examples and across languages. For example, is it 'thres-hold' or 'thresh-old'? Not earth-shattering, sure. How do you read pho? Vietnamese took over from the French the habit to write a refined ph- for initial [f]; in romanizations of other EA languages, ph is most of the time -p.h- (a final -p and and initial h-) or else an initial aspirated p: ph-. You can't tell apart those two where boundaries meet. As for Vietnamese, their ph- might even have a historical antecedent, but the -h- in Pinyin sh-, ch- and so on are purely a European habit in an orthography otherwise proudly independent and innovative system (what with q-, x-). There, the digraphs actually work well because there's no final -s, -c, -z in Pinyin. OTOH sr- cr- zr- would be conventionally less acceptable, but certainly systematically more telling digraphs, bringing more phonological depth. -ng works not so well because -ng- is ambiguous as to the syllable boundary (-n.g- versus -ng.-). Again, not earth-shattering, and a phonologically maximally systematic system will often neither be logically possible (b/c there are conflicting analogies in different phon. sub-systems, such as single-element and triple-element finals in MSC) anyway (and always think of the non-uniqueness of phonological solutions, see Y.R. Chao).

    I said I do like diacritics, but I also think the dots on Pinyin ü are, well, unhappy. Their designers didn't like them and hid them away in all syllables that don't crucially need them (hence nü, lü versus qu, ju). Where the dots are needed they somewhat collide with the diacritics that indicate tone. -ü has been replaced with -v in some practical situations, so now we have nv, lv, qu, ju, chu, zhu, which is not inherently simpler. With a dedicated letter for a dedicated sound, that same series could be, say, ny, ly, qy, jy, chu, zhu (or insert another form for y). MSC [y] can be shown to have strong affinities to both [i] and [u], so a digraph solution -iu- offers itself: nü, lü becomes niu, liu, PY xiong becomes xiung, PY xun becomes xiun (this necessitates to write e.g. niou, liou for PY niu, liu, of course).

    The point where I stop to think diacritics are a good solution is when I see Vietnamese orthography. It can be argued that it has been shown to work, though, so maybe this is pure prejudice.

  26. Anonymous Coward said,

    July 29, 2017 @ 4:23 am

    The Vietnamese ph- is due to the value [pʰ] at the creation of the script, and not to some strange devotion to French, which gladly used the simple letter f for words of native (Vulgar Latin or Germanic) origin.

  27. Victor Mair said,

    July 29, 2017 @ 11:02 am

    Here we go again (and again and again and again), with some folks arguing at length and ad nauseam against a red herring. I haven't seen anyone advocating the replacement of Chinese characters by Romanization as a writing system. What I do see are observers describing what is actually happening in China: the gradual emergence of digraphia, and the increasing use of English, both in bits and pieces in the oral and written realms, but also as integral texts, passages, and messages in writing.

    These things are taking place before our eyes in China and among Chinese people, yet some nonnative individuals take it upon themselves to inveigh against these natural developments in language usage, I know not why.

  28. Bathrobe said,

    July 29, 2017 @ 11:57 am

    I discussed digraphia (specifically the writing of foreign names) in the first part. The second part was related to the "slippery slope" that would occur if Chinese were to be written in Latin letters.

    I am not a nonnative individual taking it upon myself to inveigh against the development of digraphia, which doesn't concern me greatly, if at all. I was writing about Alex Wang's suggestion that Chinese will need to use foreign words because Chinese characters won't be able to keep up (and he appears to approve of this situation), and his suggestion that Chinese should write foreign names in English.

  29. Victor Mair said,

    July 29, 2017 @ 12:15 pm

    "he appears to approve of this situation"

    Which situation? That "Chinese characters won't be able to keep up"?

  30. ~flow said,

    July 29, 2017 @ 5:53 pm

    @Anonymous Coward—I wasn't aware that Vietnamese ph was originally meant to be an aspirated voiceless bilabial stop, thanks for filling me in! I know very little about Vietnamese spelling and do not speak the language myself, but (apart from those many diacritics) I hear commendable things, like it tries to bridge differences between dialects by using differences in spellings that are not reflected in all dialects. A superficially similar thing seems to have happened in China when the great rhyme books were written, which likewise would appear to reflect a language that was not spoken by any single individual at the time, but rather represented a synthesis of multiple forms of speech. During the 20th centuries some attempts were made to do something similar for a Chinese romanization, like keeping distinctions apart in writing that had merged in speech.

    It's interesting to see that a rather young Latin-based orthography like that of Vietnamese already has acquired traces of linguistic evolution.

  31. Bathrobe said,

    July 29, 2017 @ 5:58 pm

    That Chinese characters won't be able to keep up and locals will turn toward another language. Alex stated in a comment that he has an issue with the script. His "ramblings" were clearly against Chinese characters and in favour of alternatives. I have nothing against his thesis but do feel that he has a partisan viewpoint.

  32. Bathrobe said,

    July 29, 2017 @ 8:12 pm

    There is a lot in Alex's post that needs to be unpacked. Some of the claims / theses / suggestions that he makes are as follows:

    1. Some English terminology in technical fields is easier to remember and more natural than Chinese.

    This is possibly true but uncovering the reasons is not so simple. a) It could be because all cutting-edge work in these fields is being done in English. As a result, all 'conversation' in the field is in English. Translating into Chinese adds an additional layer and is felt to be 'unnatural' by practitioners. By an accident of history, English just happens to be the locus for this work. If 'cutting-edge' work were being done in Confucianism then Chinese would be the language of choice, but Confucianism is, for better or for worse, not cutting edge. b) English innately lends itself better to expressing concepts in such fields. This would be hard to prove or disprove. Linguistically any language is theoretically capable of expressing all the nuances required for developing new theories. But Chinese word-building is possibly inherently clunky, inhibiting thought, while English encourages it. I think that this is one of the imponderables of language. c) Chinese characters place too high a cognitive burden on thought, whereas Latin letters are minimalist and easy to manipulate.

  33. Bathrobe said,

    July 29, 2017 @ 8:12 pm

    2. Chinese word-building (or the process of creating vocabulary to express newly rising concepts) is too slow, which will force people to turn to foreign languages.

    This is also possibly true. But the situation that Alex is describing is not "digraphia"; it is the wholesale import of foreign vocabulary written in a non-native script. Alex sees the roots of this problem in Chinese characters, but 'word-building' and 'Chinese characters' need to be kept conceptually separate. Chinese calls 'string theory' (which, as I said, is not terribly new) 弦理論 / 弦理论 or 弦論 / 弦论 using the morpheme 弦 xián. 弦 refers to the string of a bow or a musical instrument. Japanese also uses 弦理論 using the same (borrowed) morphemes. Vietnamese uses lý thuyết dây, where lý thuyết is the Vietnamese term for theory (in Chinese characters 理說) and dây is a general term for cord, string, vine, etc. Vietnamese no longer uses Chinese characters but still has a huge amount of Chinese-based vocabulary. Lý thuyết is one such word. The Vietnamese term for 'string theory' is an example of native word-formation within Vietnamese that happens to use Chinese vocabulary. The problem is thus not one of "Chinese characters" as such; it is whether native word-formation can keep coming up with equivalents to new foreign (English) developments in terminology.

    Chinese word-formation is highly dependent on the use of native morphemes. Japanese tends to borrow foreign words. This is a cultural issue, not simply a linguistic one. Until now, Chinese has performed remarkably well using native linguistic resources to express foreign concepts, and there appears to be an strong cultural preference for using native morphemes instead of importing foreign words. What Alex is suggesting is that foreign words will increasingly take the place of native coinages.

  34. Bathrobe said,

    July 29, 2017 @ 8:15 pm

    3. Alex notes that translating names in Harry Potter is clumsy in Chinese. This is not just an issue of Chinese characters; it relates to conventions for rendering foreign names in Chinese. Mainland China has adopted a system that mimics katakana, spelling out each individual letter. Taiwan uses a more impressionistic system that renders how English names sound in Chinese. An example that has come up here often is the rendition of 'Trump' as Tèlǎnpǔ in Mainland China and Chuānpǔ in Taiwan. Chuānpǔ actually sounds closer to English but doesn't reflect the English spelling. The same difference is found in rendering the names of characters in Harry Potter. For instance, Draco Malfoy is Délākē Mǎ'ěrfú in the Mainland translation and Zhuǎigē Mǎfèn in the Taiwanese. The Mainland name means very little, but the Taiwanese name happens to mean something like 'arrogant-big-brother horse-part'. The poor rendition in China is related to over-rigid conventions for rendering foreign names.

    The problem of rendering foreign sounds in Chinese is partly related to decisions made in the 20th century about the use of phonetic scripts. The system known informally as bo po mo fo was a good method for rendering Chinese sounds according to Chinese phonology and had the advantage of blending in reasonably well with Chinese characters. Taiwan still uses bo po mo fo, and it is actually used at places in Harry Potter. But Mainland China decided to adopt pinyin and phase out bo po mo fo. The reasons are fairly transparent: pinyin is more international (being used all over the world) whereas bo po mo fo can only be used for Chinese. For practical reasons, it is simpler to teach children one system of phonetic representation than two. It is interesting to speculate whether keeping bo po mo fo might not have brought about its use in a role similar to katakana in Japanese, representing foreign words in terms of Chinese phonology without resorting to either clunky Chinese-character renditions or the direct use of English spelling.

  35. Joyce Melton said,

    July 30, 2017 @ 12:58 am

    Anonymous Coward beat me to the explanation about Vietnamese ph digraph. A thing to remember is, it was the Portuguese who created the Romanized script for Vietnamese, not the French; Portuguese monks to be more exact.

    This may be the origin of using nh for the sound Spanish spells with ñ; Portuguese uses that digraph for that sound (or a similar one, they aren't exactly the same).

  36. ahkow said,

    July 30, 2017 @ 1:03 am

    Dr. Mair wrote, "I haven't seen anyone advocating the replacement of Chinese characters by Romanization as a writing system."

    Wang is predicting a wholesale replacement. Although he does not advocate for it, based on preceding paragraphs he seems sympathetic to an outcome like this. "I just feel that an implosion is coming within the next few decades. Like a giant star collapsing." A giant star collapsing / "implosion" is a very sudden and thorough change — these phrases don't seem to suggest "the gradual emergence of digraphia", to quote Dr. Mair.

    The following strikes me as descriptively true, but I'm not convinced that it will lead to digraphia. Bad translations ideally give way to better translations. It's certainly possible that some of these might use digraphia, but it's also possible that other translations might exclusively use Chinese characters.
    "Even so, already for popular books like Harry Potter in Chinese, the names of people, spells, and items are awkward for locals."

    This passage was also interesting.
    "Finally, I am reading a book titled Alexander of Macedon by Peter Green. I couldn’t even begin to understand how such a book could be translated felicitously into Chinese, given that there are so many proper nouns of people and places; it would seem reading all of these names transcribed into characters would be extremely bièniu."

    I imagined a 5th century Chinese scholar who just heard about the book called the "Diamond Sutra". It's come all the way from this place they call "India", but some holy man called Kumārajīva has just translated it into Chinese!
    "I couldn’t even begin to understand how such a book could be translated felicitously into Chinese, given that there are so many proper nouns of people and places; it would seem reading all of these names transcribed into characters would be extremely bièniu."

  37. Victor Mair said,

    July 30, 2017 @ 7:43 am

    Desultory observations prompted by the current discussions


    In exhaustively / extensively critiquing the theses and preferences of others, one would do well to carefully and critically examine one's own.


    One of my Mother's favorite expressions was: "If the shoe fits, wear it."


    The "slippery slope" that has been invoked against the borrowing of English words in Chinese writing sounds like a bogeyman, especially in light of the fact that it is already happening on a large scale. If one opposes this "slippery slope", one should at least:

    a. spell out its supposed deleterious consequences if left unchecked

    b. describe what one proposes to bring a halt to it

    If requested, I can supply numerous references to earlier LLog posts on the use of Romanization and the borrowing of English in Chinese.


    The issues raised in this post and in the comments appended to it should be considered in light of the following two posts as well:

    "The naturalness of emerging digraphia" (7/28/17)

    "Learning to write Chinese characters" (7/29/17)

    See also the many links to earlier LLog posts on writing Chinese provided in the second part of the latter post. There will be many other posts on writing Chinese and language change in China during the coming weeks, months, and years.

  38. Jay said,

    July 30, 2017 @ 7:46 pm

    A dekametre is 10 metres, so deka- isn't the appropriate prefix here. Deci- for 1/10 metre is more appropriate, so decilife, or perhaps better tenth-life in analogy to half-life.

  39. Jay said,

    July 30, 2017 @ 7:55 pm

    Since Chinese doesn't have [v] (if I'm not mistaken), and so Pinyin doesn't use I don't see anything wrong with using for "ü" if it's more convenient to type. We've gotten used to Pinyin's idiosyncratic use of and .

  40. Jay said,

    July 30, 2017 @ 7:56 pm

    Well, that didn't work.

    Since Chinese doesn't have [v] (if I'm not mistaken), and so Pinyin doesn't use "v", I don't see anything wrong with using "v" for "ü" if it's more convenient to type. We've gotten used to Pinyin's idiosyncratic use of "q" and "x".

  41. Bathrobe said,

    July 30, 2017 @ 8:22 pm

    @ Victor Mair

    I don't think your vision of digraphia leads to a slippery slope. But Alex'sproposal is something else. I'm sure you imagine you belong to the same party, but from what I can see you are a Menshevik while he is a Bolshevik.

  42. Alex said,

    July 30, 2017 @ 9:18 pm

    I am not for replacing Chinese with English.

    I know and have stated on old posts that there are languages that have a closer fit/less exceptions to their script. The momentum for pinyin is too great, similar to English as a global language. Whereas the pure use of Chinese characters is in decline.

    I am for script reform as that will increase literacy, increase knowledge due to time saved can be used for reading etc, increase creativity as often children change the composition to use words they know how to handwrite. Nursing a broken bird's wing might change as they cant write wing or stories about 麒麟 might change to a story about a cow. Imagine writing that in a 9 mm by 9mm square. I am for fewer scolding and hitting. Protect the culture by increasing readability of history/lower barrier for entry.

    Here is an example
    sushi 寿司 shòusī

    let us think about the child, he hears the word, knows the word but doesn't know how to write or even begin to look up. So the story changes we went to eat pizza last night oh wait

    For both words the pinyin are easy to remember for kids.

    I am wondering why the school systems here choose to start with pinyin here in first grade before teaching how to write characters?

    As for using English personal names I am for that. Regardless of my personal opinion, that is the trend as I am beginning to see it in books and subtitling in tv shows. Game of thrones is popular here. It is easier to match Sansa with Tyrion saying on TV. Mixing English names into the subtitling at least with the people ive asked seem to increase readability and speed. It is important because China wants to be global so when meeting business people or saying where you are going its actually becoming more natural for locals than the phonetic use of Chinese characters. It allows locals to communicate with global counterparts. We have had staff do engagements in several English speaking countries and at night when they go out together often they both like the same actor or would love to discuss a famous historical person like a famous general or scientist but the discussion doesn't happen. When people talk about what famous street they visited or what famous castle etc etc many now use English and I have started asking since its summer and many of my wechat friends are traveling abroad. I ask them by the way why did use English names for places and street names when for some places wechat location it has the Chinese etc. I guess we all know the most common answer.

    Its adapt or illiteracy will increase as more people switch to voice. I'm beginning to see people now have trouble selecting the right character for infrequently used characters since wechat introduced voice messages.

  43. Alex said,

    July 30, 2017 @ 11:40 pm

    By the way, I am curious for the pro written Chinese character group, how many of you had to teach your children Chinese characters to fluency, if you did, how long ago was it?


  44. Alex said,

    July 31, 2017 @ 8:14 am


    perhaps a better analogy is who is Yeonsangun of Joseon and Choe Manri

    what's funny is I can do nothing and time/technology will take its course.

    What is the slippery slope? usually that implies something will start upon a negative path. I thought languages and script were meant to evolve?

    Did something terrible happen when English changed throughout the centuries?

    Did something terrible happen when traditional Chinese changed to simplified?

    Or when we went from brush to ballpoint pen to write Chinese characters?

    Will I be happy, sure as in my mind it means the country and the people are progressing and less kids are suffering. I only wish it happened so my kids didn't have to suffer but perhaps my grandkids wont need to suffer as much.

    My first email it concerned GDP. I still believe it. I believe the change will bring about increased GDP and thus better the lives of hundreds of millions of people.

  45. cliff arroyo said,

    July 31, 2017 @ 11:40 am

    "The momentum for pinyin is too great, similar to English as a global language. Whereas the pure use of Chinese characters is in decline."

    One problem I perceive is that Chinese characters have tremendous emotional capital for users.
    English has tremendous emotional capital for learners.
    Pinyin has…. very little to know emotional capital so far. Victor has mentioned that people who claim to have loved his late wife can't be bothered to read her beautifully written memoirs becasue they're in pinyin (a system they supossedly know but are unused to reading in).

    Vietnamese speakers do have an emotional attachment to their writing system (more than they ever had for Chinese characters AFAICT) and apparently acquired that over decades (or centuries) of bottom up usage.

    The way to further pinyin is to figure out how to give its usage an emotional weight that it so far lacks.

  46. Victor Mair said,

    July 31, 2017 @ 12:37 pm

    For those who say they disapprove of Romanization and English borrowings for Chinese, yet simultaneously say they don't care about digraphia, well, they should care, because half of digraphia is Romanization, including English words directly borrowed into Chinese. It's called shuāngwénzhì 双文制.

    In China, the Romanization they use is Hanyu Pinyin (HP), which is an official system of representing the sounds of Modern Standard Mandarin (MSM) that we've talked about often enough on Language Log for regular readers to be familiar with it. We've also mentioned more than once that the PRC has an official orthography for writing integral MSM sentences and texts in HP.

    As for the dreaded "slippery slope", I'm still not clear what's so scary about it, of what it actually consists, and how to stop it if it really exists and is such a bad thing.

  47. Philip L. said,

    August 1, 2017 @ 4:54 pm

    Being fairly unfamiliar with simplified characters, until I read the comments to this post I did not know that the simplied form of 網 is 网. I love it! The graph actually looks like a net, especially a tennis court net! I looked up the simplified form for 球 and was disappointed that it was not O.

  48. Eidolon said,

    August 9, 2017 @ 7:16 pm

    Digraphia, in the sense of people in China combining the use of Chinese characters, pinyin, and possibly some English, especially in the class of acronyms, is both observable and, perhaps, inevitable. But what isn't inevitable, or even likely, is a large-scale *oral* language shift towards English, either due to the difficulty of translating English terms into Standard Mandarin or due to the difficulty of transliterating English names into Standard Mandarin. The former isn't much of a problem, and the latter, while a problem, is simply not important enough to warrant language shift. To whom does it matter that the Chinese transliteration of Harry Potter's name is terrible? I would argue, not other Chinese.

    English being the prestige international language of the modern world, and a favorite second language of upwardly mobile East Asian students, is a given. But I see no hurry in countries like Japan and South Korea, which are much further down the path of "Westernization," to shift towards speaking English, so I have no reason to believe it'll happen in China, either. Large-scale oral language changes are usually a product of government policy, and with respect to the PRC, the only such policy is wide-scale Mandarinization – and that, just like digraphia, is an observable phenomenon in modern China.

  49. Victor Mair said,

    August 10, 2017 @ 8:43 am

    Digraphia is happening / emerging in China today. We have given mountains of evidence for it on Language Log, there is plenty more out there in the real world to observe, and you can rest assured that we shall be reporting on it here in the coming months and years.

    Someone's opinion that digraphia cannot or should not happen in China cannot negate what is actually happening.

RSS feed for comments on this post