Kanji of the Year: the tie that binds

« previous post | next post »

We're waiting eagerly for the English Word of the Year for 2011 (to be announced on January 6, 2012) and have already had the Chinese "Morpheme(s) of the year".  Now arrives the Japanese Kanji of the Year:  kizuna 絆 ("bond").

The first two paragraphs of the BBC article reporting the selection of kizuna 絆 ("bond") as the kanji (Chinese character) of the year contain two major mistakes and at least one minor error:

The Japanese word "kizuna", meaning bonds or connections between people, has been chosen as Japan's kanji of 2011.

The kanji, or Chinese pictorial script, for "kizuna" emerged top of a public poll for the character that best summed up the year.

First of all, the initial sentence / paragraph describes kizuna both as a word and a kanji.  Since this is a matter of the Japanese character of the year, it should not simultaneously (in the same sentence) be referred to both as a word and as a kanji.  In the present case, it is the character aspect of 絆 which is at issue; among its readings is kizuna, one of whose meanings is indeed "bond", but it also conveys the less felicitous meanings of "encumbrance; hindrance".  When we consider a fuller range of meanings for kizuna 絆, we find that, more often than not, the implications tend to be on the negative, restrictive side:  "bonds; fetters; a yoke; encumbrance; ties".  I will return to this highly multivalent aspect of kizuna ("bond[s]") below.

Secondly, and more seriously, kanji (the Japanese pronunciation for 汉字 [simplified] / 漢字 [traditional]; Korean hanja; Mandarin hànzì — all three meaning "Chinese character") may not be defined as a "Chinese pictorial script" (they were probably thinking of "pictograph", but wanted to say it in a more elaborate way, only making it sound even more ridiculous).  In what way, pray tell, is 絆 a picture of a "bond, encumbrance, hindrance"?  Nor are kanji / hanja / hanzi "ideographs / ideograms" or "hieroglyphs", those other favorite lay and journalistic denominations for what are more precisely described as morphosyllabograms (however, since that's quite a mouthful, in normal discourse it's probably easiest and safest just to refer to them as "[Chinese] characters", inasmuch as I am also uncomfortable with calling them logographs / logograms, a designation that is popular with many specialists, but the kanji / hanja / hanzi most assuredly do not always represent a word).

The minor error is that a single graph, letter, or character does not constitute a script, which is implied by the beginning of the statement in the second paragraph.

Unlike the monosyllabic Chinese Morpheme(s) of the Year, kòng 控, which has widely variant meanings and usages, the trisyllabic reading of 絆, kizuna, is actually a native Japanese word and may freely be used as such.  However, 絆 also has Sino-Japanese monosyllabic readings, han and ban, in which case, like kòng 控, it must be used in combination with other morphemes:

hanten はんてん / 絆/半纏 ("hanten, a short coat of Japanese style; a workman's happi / livery coat")

bansōkō ばんそうこう /  絆創膏 ("a sticking / adhesive plaster")

kyahan きゃはん /  腳絆 ("leggings; gaiters; puttees")

kihan きはん / 羈絆 ("restraint; fetters; a yoke")

I should note that, aside from the indigenous trisyllabic word kizuna, another polysllabic native word has also been attached to the kanji 絆, namely hodasu 絆す, which has the verbal meaning of  "tie; bind".  A derived noun of this verb is hodashi ほだし / 絆し ("ties; bonds"), an indigenous synonym of kizuna.

Why is kizuna 絆 ("bond") supposedly an appropriate choice for Japanese kanji of the year?  The Japanese people were severely tested by the devastating 8.9 magnitude earthquake that struck the northeast coast on March 11, 2011.  This, in turn, triggered a 30 foot tsunami that washed away entire towns, and this was followed by the meltdown of the Fukushima nuclear power plant.  All of these linked disasters led to tremendous loss of life and enormous property damage.  What was remarkable was the way the entire nation came together and worked hand in hand to restore a sense of normalcy to the afflicted areas.  This has been attributed to kizuna 絆, the bonds that tie the people together in a sense of tight community.

The Kanji of the Year in recent times have often directly reflected the negative aspects of various calamities and catastrophes.  For example, the Kanji of the Year for 1995 (the first year for which a Kanji of the Year was chosen) was shin 震 ("earthquake"), after the Hanshin-Awaji or Kobe earthquake, and in 2004 it was sai 災 ("disaster"), after the typhoons, earthquakes, torrential rains, and torrid heat of that year.  Incidentally, sai 災 ("disaster") came in second this year and shin 震 ("earthquake") came in third.

Whereas Japanese kizuna 絆 is a noun, the most prominent meanings of Mandarin bàn 绊 are verbal:  "trip, stumble, hinder, impede, encumber, entangle"; the original signification of Sinitic bàn 绊 was "to restrain a domesticated animal by tying a rope around its leg(s)", i.e., "hobble; fetter", from which was derived the more general meaning of "restrain; bind; control", curiously recalling the primary meaning of this year's Chinese morpheme(s) of the year — kòng 控 ("control").  Mandarin bàn 绊, however, may also function nounally in expressions such as jībàn 羁绊 ("fetters").  Pronounced pàn, 绊 may be considered as equivalent to pàn 襻 ("loop for fastening a button or something shaped like a button loop and used for similar purposes; fasten with a rope; tie").

The ambivalence of kizuna 絆 is explained in philological terms by Jim Unger:

The gloss kizuna < kiduna < ki-tuna in the nice, chummy sense of 'bond' is derived.  I guess in Chinese the word represented by the character originally was a noun meaning 'hobble'.  (The same character is used to write hodasi 'fetter'.)  One can see how the nice connotation developed from the original denotation, but I guess this is a good case illustrating the foolishness of believing that every character has an unalienable, eternal "core meaning."

Wondering whether the Sinitic bàn 绊 morpheme shows up in Korean and Vietnamese vocabulary as it does so prolifically and productively in Japanese, I asked Bill Hannas how it manifests itself in these languages, to which he replied:

A large dictionary of Sino-Korean does list PAN () with that character and approximate meaning but the term doesn't appear in a (regular) hangul dictionary of the language, which means it was not indigenized.

The correct Vietnamese spelling is BẠN with a dot under the vowel….  The term does not appear as a word (with that meaning or anything close to it) in modern Vietnamese.  It shows up in a modern Vietnamese-Chinese bilingual dictionary but only as one of the characters that have that sound, not as a lexical entry per se.

To which Bob Ramsey adds:

Of course the Sinitic morpheme associated with the character 絆 is part of the Korean lexicon, but only in compounds.  (It's read pan or ban, depending on the romanization system you use—Koreans don't distinguish voiced from voiceless; and this particular obstruent is voiceless and slightly aspirated in initial position, but voiced intervocalically.)  But of course kizuna is a native Japanese word, and the character is only used here to indicate a kun reading.  (You may know this, but in the Three Kingdoms period people on the Korean peninsula also used this unwieldy device, called hun by them, to write native words.  But then, Chinese character readings were completely standardized by the powerful monarch King Kyongdok in the Unified Silla period, and kun (or hun) readings largely disappeared from use thereafter.  Not to put too fine a point on it, but it seems pretty clear that the early Japanese learned this and most other ways of writing from people from the Korean peninsula, Paekcheans probably, but Sillans might also have played a role in the transmission of scribal methods.)

But to get back to the topic, if I understand what you're asking, I'm not at all sure Koreans have a really close equivalent to kizuna.  A favorite Korean word that's somewhat similar, though, is inyon (因緣); that's of course a Buddhist concept that came to Korea via Chinese, but Koreans have taken this idea of karma or fate to heart and think of it as a tie that transcends other worldly things, including death.  That may not be the same thing exactly as kizuna, but it definitely is an idea about ties that is deeply embedded in Korean culture.  There may be some other, better equivalent here, but I can't think of anything else right now that's so close to Koreans' ideas of what binds people together.

After having completed this post, ringing in my mind are these lines from an old hymn used by Thornton Wilder in "Our Town".  The hymn, lines of which are sung three times in the play, has many verses, of which I quote only the first here:

Blest be the tie that binds
Our hearts in Christian love;
The fellowship of kindred minds
Is like to that above.

I quote these lines with the proviso that the bonds ones feels to one's fellow man and woman need not necessarily be of the Christian sort.

[A tip of the hat to Carmen Lee and thanks to Cecilia Segawa Seigle, Nathan Hopson, Miki Morita, and Fred Dickinson]



  1. yeajung said,

    December 26, 2011 @ 8:33 am

    Brief comment as a Korean speaker: It is true that 絆 is not part of any commonly used word in the Korean language. In passing, about the suggestion for a Korean equivalent of kizuna, I'd like to suggest the term "연대(yeondae; 連帶)", which means 'solidarity' or 'banding together'. The word "inyun", though indeed popular, is used mostly for personal relationships between two or a few people.

  2. Yuji said,

    December 26, 2011 @ 8:41 am

    As a Japanese born and raised in Japan in his thirties, I say the word "kizuna" is mostly taken positive among not-so-old Japanese, often used to describe affectionate ties among friends and lovers in novels and animés. In colloquial usage, I didn't remember anybody using the word "kizuna" in negative way. Just my 5 yen.

  3. Mary Bull said,

    December 26, 2011 @ 10:07 am

    Formal request: Since the pale pink font is so hard on my old, cataract-ridden eyes, is there any possible way to use something with more contrast — possibly dark blue — to highlight and set apart quotes in the posts at LL?

    I would be immensely grateful if it could be done.

    This long and extremely interesting post has tired my eyes so much that they have begun to ache, and I could not read more than half of this fascinating text.

  4. J.W. Brewer said,

    December 26, 2011 @ 11:06 am

    Just on the hymn text quoted by Wilder: I would have thought the stock phrase was "ties that bind" rather than "tie that binds" but it turns out interestingly enough that per the google n-gram viewer the statistical dominance of the former is merely a phenomenon of the last six or seven decades. (The trend lines cross in favor of "ties" circa 1925 but the divergence doesn't get wide until after 1955 or so.) OTOH, plural "bonds of matrimony" has been notably more common that singular "bond of matrimony" continually since 1800, although the latter has persisted as the minority variant. In general "bond" and related words seem susceptible of both positive and negative meanings in English, much as "kizuna" does in Japanese, although maybe there are some additional details (i.e. English verb "to bond" is generally positive whereas "to bind" can go either way, even though both can sort of be glossed "to form a bond").

  5. Mr Fnortner said,

    December 26, 2011 @ 11:28 am

    @Mary Bull, You may be able to control the colors yourself. In IE try: Tools > Internet Options > General > Appearance > Colors > Use Windows colors, or set your own. In Firefox try: Tools > Options > Fonts & Colors > Colors > Link Colors. Other browsers act in a similar way. I've had the best success letting the browser set the colors. Good luck.

  6. Nathan said,

    December 26, 2011 @ 11:30 am

    @Mary Bull: I've never seen text in any shade of pink at LL, using at least four different browsers over the years. It's something in your own settings. In fact, the quotations are dark blue by default.

  7. Brett said,

    December 26, 2011 @ 11:39 am

    @Nathan: In this particular post, the quotes were a pale red-orange hue (Mary Bull's pink, I assume) when I first viewed the post. However, they have no reverted to the more standard blue, even though my browser has not changed in any way.

  8. Eric P Smith said,

    December 26, 2011 @ 12:24 pm

    On “the tie that binds”, the paradoxical relation between the positive and negative aspects of binding is a recurring theme in Christianity. My favorite exploration of the theme is in the well-known hymn by the Scottish theologian George Matheson, Make me a captive, Lord, and then I shall be free. You can see the four verses at http://www.hymnary.org/hymn/CH4/534.

    (Note that in verse 2, the imagery is of a pocket-watch: the word wind is from the verb wind and rhymes with ‘find’.)

  9. Bertil Wennergren said,

    December 26, 2011 @ 12:56 pm

    "Koreans don't distinguish voiced from voiceless"

    That is most definitely not true.

    [(myl) Indeed. Korean is famous for its three-way voicing contrast: the alternatives are generally called other things, like "lenis/fortis/aspirated", or "plain/tense/aspirated", or "lax/forced/aspirated", but (among other characteristics) they do differ in voice onset time.

    I'm not sure what Bob Ramsey meant -- maybe that none of the alternatives are exactly like English voiced and voiceless stops?]

  10. Peter said,

    December 26, 2011 @ 2:37 pm


    According to Wikipedia, the oral stops are distinguished phonemically into 3 categories: unvoiced aspirated, unvoiced unaspirated, and "faucalized voice". The voiced stops are allophones of the unvoiced unaspirated ones. Presumably the claim that "Koreans don't distinguish voiced from voiceless" should be taken to mean that the voiced-unvoiced distinction is not phonemic (aspiration is though).

  11. KWillets said,

    December 26, 2011 @ 4:43 pm

    Koreans don't distinguish voiced from voiceless; and this particular obstruent is voiceless and slightly aspirated in initial position, but voiced intervocalically.

    He's referring specifically to the different sounds of ㅂ between word-initial and intervocalic positions. Kim-Renaud's Korean An Essential Grammar describes it almost the same way. However it is clear that Koreans do distinguish the sounds, since they make the distinction consistently in speaking.

    I guess he's saying that Hangul doesn't allocate two characters to the distinction.

  12. Andy said,

    December 26, 2011 @ 6:59 pm

    It is great to see here all sides of the same kanji coin.
    As the heavily promoted 絆 (especially by the Japanese government) as the ties that bind were also the きずな fetters that kept people some from revealing and others from pursuing the facts about the real situation at Fukushima.

  13. Mr Fnortner said,

    December 26, 2011 @ 7:33 pm

    Isn't distinguishing between the character and the word in Chinese not much different from distinguishing between the spelling and the word in English? For example, would we promote the "spelling of the year," as in t-a-k-e or s-t-a-n-d as distinguished from one of the multiple meanings of the words take or stand? To most people I suspect the map IS the territory, and therefore the character IS the word.

  14. Victor Mair said,

    December 26, 2011 @ 8:01 pm

    @Mr Fnortner

    Characters are usually NOT words, since the average length of a word in Mandarin is roughly two characters / syllables, yet again, sometimes characters ARE used to write whole words that are free, not bound.

  15. Victor Mair said,

    December 26, 2011 @ 8:20 pm

    From Bob Ramsey:


    I don't mean to get involved in a lengthy discussion of this particular point; there are too many more interesting things to explore in this thread. But maybe I should try to explain my off-hand remark.

    In his comment, Bertil turns the discussion to voice onset time, and sure, the aspirates certainly represent a long lag in voice onset time. (All those values related to voice onset time (VOT), as well as plain/tense/aspirated, have been carefully measured in Korean since at least the sixties–as he probably knows.) But "voicing" addresses a particular range of values on the voice onset time scale. When one says a consonant is "voiced", it means that the vocal cords are already vibrating at the time the obstruent is articulated and released (which means that in the case of voiced consonants the VOT is a minus value). Now notice that in the case of Korean, no obstruents have such VOT values in initial position. That is, they are not voiced there. Korean does have a [b] that occurs in the middle of a word, but that consonant is just another realization of the lax [p] that is found at the beginning of words. Both are allophones of /p/. They are not distinct from each other. The difference is not phonemic.

    That Korean does not have a voicing distinction in the obstruents has a real-world consequence in decisions about how Korean is romanized. For example, when the Korean surname 박 (朴) is romanized, most people write it "Park". That's because the spelling is close to the way English speakers have heard it. We hear an initial "p". And it really is phonetically a [p], unvoiced with slight aspiration (but not aspiration DISTINCTIVE in Korean). On the other hand, when Koreans say, for example, "Mr. Park", to American ears what they're saying often sounds like Mr. "Bark", with a [b]. Why? It's because Koreans pronounce "Mister" with a vowel sound at the end, and so the lax initial consonant at the beginning of the name automatically becomes voiced between those vowel sounds. What's important to remember is that Koreans can't actually HEAR the difference between the two sounds; they think they've pronounced the name exactly the same way in both cases. Within the Korean phonological system, the two sounds [p] and [b] are just different allophones, one unvoiced and one voiced, of the very same phoneme. Thus, the sound [p] is not distinguished from [b] in Korean words, and so when Koreans borrow an English word with an initial [p], they give that consonant the heavy aspiration (long VOT lag) of their own consonant /ㅍ/. They don't render it with a /ㅂ/, because they equate that consonant with our [b], even though /ㅂ/ has what is objectively speaking a phonetic value of [p] closer to that of our /p/ in initial position.

    Peter writes, "the claim that 'Koreans don't distinguish voiced from voiceless' should be taken to mean that the voiced-unvoiced distinction is not phonemic (aspiration is though)". That's right; he understood what I meant. –But he shouldn't say the difference is a "distinction". By definition, a "distinction" means the difference is "phonemic".


    In response to my request to post Bob's comments in this thread, he added:


    You're more than welcome to use those remarks as you see fit. My initial reluctance to give them was because I didn't want people to get sidetracked by that topic.

    More central to your discussion maybe is the suggestion that Korean 연대(連帶)might be the Korean word closest to kizuna. That's an interesting idea. But the way I imagined the meaning of kizuna (and I'm of course not saying that I necessarily understand all the nuances a native speaker takes from it) was that it was a word with more emotion and personal sentiment attached to it. The literal meaning of kizuna, after all, is one of cords and bonds, and it's often used together with words like 'love' and 'affection'. It's true, as Yeajung says, that 연대(連帶) is used in the sense of group solidarity, while inyon(因緣) is more often used with a couple, a husband and wife, say, or a relationship with a family member, or a close friend. But it's that more personal connection that I thought made that word 인연(因緣) closer to the emotional nuances of kizuna. 연대(連帶) (which is a Korean word that may even come from the Sino-Japanese word rentai) is used when talking about belonging to, and working with, a group, and my impression is that it implies very little in the way of personal emotion. –But of course I'm not a native user of either Japanese or Korean, so I could certainly be off base.

  16. Henning Makholm said,

    December 26, 2011 @ 8:47 pm

    The only thing I know with certainty about kanji/hanja/hanzi is that there is no possible general English description of their nature and communicative role that a journalist can safely use without risking rebuke from specialists who know better.

    They don't encode words. They don't encode meanings. They don't encode sounds. They are not ideograms. They are not logograms or morphograms. They are not hieroglyphs, plctograms, or any of the preceding options with -graph or -glyph instead of -gram. They are definitely not letters. Apparently they are not monosyllabograms either, unless the Japanese somehow manage to pronounce "kizuna" with but a single syllable.

    I find it difficult to fault a journalist who just gives up and picks a wrong description at random, when there are no safely right descriptions to pick.

  17. Victor Mair said,

    December 26, 2011 @ 9:03 pm

    From Bob Ramsey:


    Just noticed Yuji's comment about the meaning of kizuna. That positive nuance of the word describing "affectionate ties among friends and lovers" is pretty much in line with what I was thinking of as the meaning, and what I was trying say here.


  18. marisu said,

    December 26, 2011 @ 10:26 pm

    To quote from Yomiuri News, the reasons for choosing 絆were「 東日本大震災や台風12号など相次いだ災害で再認識された家族や仲間、地域とのつながりの大切さや、サッカー女子ワールドカップ(W杯)で優勝した「なでしこジャパン」のチームワークの良さなどを理由に挙げた人が多かった。」(http://www.yomiuri.co.jp/national/news/20111212-OYT1T00581.htm) This reflects the usage I'm most familiar with: the close bonds between friends and families and within communities. @Andy, that is a usage one would not see; you are more likely to find negative connotations in a Buddhist context. The 『広辞苑』for example gives the following in its definition:「妻子といふものが、…生死(しようじ)に流転(るてん)する―なるが故に」。

  19. Mark F. said,

    December 26, 2011 @ 10:40 pm

    Henning – what's wrong with "Chinese character"?

  20. KWillets said,

    December 27, 2011 @ 1:00 am

    I agree with Bob's comments, particularly if they mean that Busan should be called Pusan again, and Daegu Taegu. I think somebody in the romanization office got drunk and voiced all the initial consonants by accident.

  21. Kenny said,

    December 27, 2011 @ 1:44 am

    I think the only inaccuracy the BBC article includes is that a kanji=a word, but it seems to me that it's mostly the fault of the people who selected the kanji of the year.

    In their announcement, they present a dictionary entry for the character 絆 and several reasons the character is appropriate; however, their reasons involve only the reading kizuna. Nothing in the reasons can be read hodashi (as far as I can tell), and there are no compounds that involve 絆. It really does come across as if the word kizuna had been selected as a word of the year, and the kanji had been dragged along for the ride. Also, although the kanji and supposedly the reading kizuna can mean negative forms of bonds or ties (such as fetters), every single example is the positive sense of bonds between people (I'm not a native speaker, but I've never seen or heard a negative use of kizuna).

    I don't find it unreasonable to say that kanji are pictorial. Of course they aren't all pictures of exactly what they mean, but several of them are stylized pictures of their referents (mountain 山, fire 火, sun 日, water 水, river 川), and many of them are composed of more strictly pictorial characters for things that are related to their meaning, though some are composed of other kanji for their sound rather than their meaning. Many Kanji ,such as forest (森), are still basically pictorial. The writer was clumsy for defining kanji in the broad "script" sense in a context that did not use the word that way.

  22. Kenny said,

    December 27, 2011 @ 1:44 am

    For those who can't read the linked page, here's my paraphrase/translation of the explanation. Kizuna 絆 is the always given as bond and no other word was translated as bond.

    "From experiences of large-scale disasters (such as the earthquate and the typhoon in Japan, the earthquake in New Zealand, and the floods in Thailand), people realize anew their bonds with the irreplaceable people close to them, such as their family and friends.

    "The little connections between people are becoming bonds between humans on a global scale that overcomes national boundries, not just communities like provinces and societies.

    "Through social media, which began with social networking services, new bonds between people were born and bonds with old friends deepened.

    "Also, in international society their were many movements for democracy.

    "All Japan was moved at and given courage by the bonds of the Nadeshiko Japan team, which claimed the championship in the World Cup."

  23. Laura Miller said,

    December 27, 2011 @ 11:27 am

    Victor was correct to list the various readings for the Kanji of the Year, an event sponsored by the Japanese Kanji Proficiency Society as part of their mission to promote appreciation and mastery of kanji. On their official website they list four readings for the character (han, ban, kizuna, and hodasu) and they also select two “meanings.” The event is not intended to celebrate a “word” but rather awareness of kanji.

  24. Victor Mair said,

    December 27, 2011 @ 12:21 pm

    From Sung Shin Kim:

    The closest Korean word I can think of is 结束 (결속, kyŏl sok). "Kyŏl sok ŭl tajita (결속을 다지다)", "hardening kyŏl sok" is, for example, a common expression.

  25. Rurousha said,

    December 27, 2011 @ 6:44 pm

    "This, in turn, triggered a 30 foot tsunami … " If only the tsunami were only 30 feet! It was 30 meters; as a matter of fact, it reached 40 meters in certain areas.

  26. Kenny said,

    December 27, 2011 @ 6:57 pm

    @Laura Miller
    I agree with you. Victor's discussion of the various readings was proper, and I think entirely accurate (as I said I have no exposure to the other uses of the word kizuna or of the character, but the dictionaries I can access at the moment agree with him). I just think it's sort of silly to claim to celebrate an entire kanji or to claim to promote awareness of kanji without giving some sort of importance to the other readings. I suppose we can do an equivalent thing with a word of the year by de-emphasizing different meanings or senses or concentrating only on one sense, but it seems more extreme to emphasize only certain senses of a certain reading of a kanji and call it celebrating the kanji.

    In this situation journalists are understandably confused. Even the Kanji Proficiency Society treated the kanji of the year as a word of the year.

  27. Laura Miller said,

    December 27, 2011 @ 10:15 pm

    I love the idea of celebrating a kanji for itself, and not any particular meaning or reading it may have, and hope I’m not that silly to think so. Many calligraphy associations and schools in Japan also believe that the form of a character is more worthy than pinning a specific meaning on it. When I see the priest at Kiyomizu-dera performing the kanji, it is not a “word” that is being expressed. In any case, the intent of the creators of the Kanji of the Year was not to focus on one meaning. In some ways the event reminds me of Korea’s Hangul Day, also a celebration of a writing system.

  28. Chad Nilep said,

    December 28, 2011 @ 1:41 am

    "I am also uncomfortable with calling them logographs / logograms, a designation that is popular with many specialists, but the kanji / hanja / hanzi most assuredly do not always represent a word".

    At least in my understanding, "logogram" need not mean "a character that represents a word" but rather "a character that represents a morpheme or similar semantic unit". Etymology is not destiny, but unless I'm mistaken logos can mean either "idea" or "word".

    "Morphosyllabogram" is also problematic, at least for Japanese kanji, since it seems to suggest that each character maps to one syllable as well as one morpheme. This is often, but not necessarily the case.

    The first two examples that come to mind are 学ぶ (to learn) and 風邪 (a cold, a respiratory infection). In 学ぶ (manabu) the hiragana is bu, so the kanji must map to the first two syllables. Both 風 (the wind) and 風邪 are pronounced 'kaze', so the second character contributes meaning but does not indicate pronunciation. Then of course there are odd cases like 河豚 ('fugu'; blowfish), where the characters don't seem to relate to the pronunciation in any way I can see.

    Any road, we come to the same end: I also tend to call hanzi / kanji / hanza "(Chinese) characters".

  29. Jongseong Park said,

    December 28, 2011 @ 3:39 pm

    Then of course there are odd cases like 河豚 ('fugu'; blowfish), where the characters don't seem to relate to the pronunciation in any way I can see.

    Wow, I learn something new every day. The Korean for blowfish is bogeo 복어, which seems likely to be related to fugu. Now, eo 어 is Sino-Korean for 'fish' and bok 복 fits the more stringent phonotactics of Sino-Korean syllables, so I had assumed until now that the whole word bogeo was a Sino-Korean term. But no, a quick look at the dictionary reveals that while eo is indeed the Sino-Korean element, there is no Chinese character corresponding to bok. This is what happens when you no longer use Chinese characters in writing—I probably have made countless other incorrect assumptions about Sino-Korean etymologies because I don't know the correct hanja for Korean words.

    So bok is a native (non-Sino-Korean) term, and the Japanese fugu must be cognate to it, but the Japanese decided to write the Chinese term 河豚 as the kanji for the fish while using the kun reading for it.

  30. Victor Mair said,

    December 28, 2011 @ 6:07 pm

    @Chad Nilep

    On logos.

    From Don Ringe:


    My reading dictionary gives 'word; talking; statement, assertion, promise, resolution, condition, command; condition, conversation, expression, power of speech, repute; language; thought, reason; opinion; …'–you get the idea: in different contexts _lógos_ can have quite a range of different meanings. The basic meaning really is 'word', though; it doesn't seem stretchable all the way to 'idea', and 'morpheme' is a concept the Greeks just didn't have (not even the grammarians).

    I think 'logograph' and 'logogram' *are* intended to mean 'sign for a whole word'. If I understand what I've been told, that was probably true in Ancient Chinese but is certainly no longer true. The system can't be called a syllabary, though, because each character indicates both a syllable and its meaning(s), not just the sounds of the syllable. I think your suggestion that character = morpheme must be about right.


    From Joe Farrell:


    > > LOGOS means "word," right?

    Yes, but it covers a wider range of meanings than "word" does.

    > > If we say that it also means "speech", what does that imply?

    It could mean a speech or an account of something. I think it could also mean the capacity for speech, including whatever that might imply, such as the capacity to think and reason. For instance, the Greeks would have said that men had logos, but animals not, meaning the ability both to speak and to reason.

    > > Can it possibly mean "morpheme", "idea"? (I don't think so.)

    No, I don't think so. "idea" is a Greek word.

    > >
    > > I'm struggling with the meaning of "logograph / logogram" — which I don't think are appropriate for Chinese characters, since they imply that the basic unit of writing is a word, and that is certainly not the case for Chinese, where the average length of a word is approximately two syllables / characters.

    I see. I would guess that "logogram" is a fairly uninspired calque for "word-picture" or some such. But it may have been better thought through than that, since Chinese characters are "marks" that convey "sense." So are letters, though, so neither word does much to distinguish Chinese characters from western alphabets.

    For what it's worth, "character" is a Greek word, as well, meaning something like "stamp" or "type" (in the sense of a mark, not a category). And the Greek word for letter of the alphabet is stoicheion, which is also the word for "element" (as in physics), although collectively words and all of the knowledge and skills needed to use them properly are known as "grammata," which is just the plural of gramma, "a mark or drawing."


  31. Victor Mair said,

    December 30, 2011 @ 10:53 pm

    From Eric Henry:

    …there is a Han Viet word "bán" (絆), which means stocks or fetters. It can also be used as a verb: "to impede, to prevent forward motion." In Chinese, you can use the resultative zhù with this word: 絆住.

  32. JQ said,

    December 30, 2011 @ 11:34 pm


    Wiktionary lists 鰒 as an alternative kanji for fugu, which sounds like a good fit for the bok in bogeo, but would have expected the Japanese reading to be fuku instead.

    I think somebody in the romanization office got drunk and voiced all the initial consonants by accident.

    Romanization != English

  33. JQ said,

    December 30, 2011 @ 11:37 pm

    OK, I should have clicked on before submitting, which says the Korean is indeed bok and fuku is the ONyomi while fugu is a KUNyomi. Apparently, it can also be pronounced awabi which means 鲍 (abalone)

RSS feed for comments on this post

Leave a Comment