« previous post | next post »

This will be a mini-disquisition on fish terminology, focusing on one particular species.

Reader hanmeng, after seeing a reference to bàyú 鲅鱼 (a kind of fish — discussion below) in the opening scene of the 32nd episode of " Méndì" 门第 ("family status; pedigree; ancestry; lineage; families related by marriage equal in social status" — title of a popular TV drama series), googled to find what the equivalent word is in English, and was directed to Baidu (a search engine for Chinese-language websites), where they render it as "Japanesespanishmack—erel".

hanmeng comments:

Obviously an error; how could it be called "Japanese Spanish" anything?

[VHM: When I queried hanmeng what he meant by this, he replied, "I meant offhand I had thought any given item should either be Japanese or Spanish but not 'Japanese Spanish'. Although in a different context I am sure I could believe in, for instance, a Japanese Spanish professor.]

In fact not only is there such a fish as the "Japanese Spanish Mackerel", the misprint (with or without the missing space) is found several other places. Apparently it started from this paper (see the entry in the bibliography under Kishida, T., K. Ueda and K. Takao at the bottom left column of p. 94.)

[VHM: I don't see anything wrong with this entry, nor with the previous one, both of which make reference to "Japanese Spanish mackerel", unless hanmeng considers that itself to be a mistake — but see below for more on this name.]

I'd have thought the Japanese would know better, but there's at least one Japanese page that apparently also uses baidu.

[VHM: hanmeng's claim here seems to be valid, since this page (about 2/3rds of the way down) copies " Japanesespanishmack-erel" in the middle of a line, where the hyphen is not necessary.]

Since the dash in "Japanesespanishmack—erel" in the Baidu article occurs at the end of a line, it is obvious that the Baidu editors meant that to be "Japanesespanishmackerel" (in his message to me, hanmeng referred to that as a "ghost-word"). Unpacking "Japanesespanishmackerel", whoever is responsible for that monstrosity must have been intending to write "Japanese Spanish mackerel", but as hanmeng intimates, would that make any sense?

Let's start with the usual bilingual dictionary entry, which is:

bàyú 鲅鱼 ("Spanish mackerel" [Google translate, Baidu Fanyi, Bing Translator, ABC Comprehensive Chinese-English Dictionary, Oxford Chinese Dictionary]).

Citing a number of correspondents who are expert Chinese and Japanese lexicographers or language teachers, and one who is the world's authority on Japanese fish markets, we will hone in on "Japanese Spanish mackerel" to determine whether it is a "ghost-word" or not, and ultimately — if possible — to decide what is the most accurate translation for bàyú 鲅鱼. I began by asking them why the Japanese write the name of the fish in katakana (sawara サワラ), rather than in kanji (鰆), and if they had any comments ready to hand about this fish and terminology relating to it.

From Ted Bestor, author of Tsukiji: The Fish Market at the Center of the World (UCal Press, Amazon):

I think this is the standard form, but fish names are incredibly variegated, not just in Japan but throughout the world — partly because fishing is highly seasonal and the same fish may be caught in many different locations at different times, fish names proliferate.

I am not certain, but I imagine that most Japanese consumers would not be able to read this kanji. A popular motif in sushi restaurants are teacups and towels that have 50-60 fish kanji on them and little hiragana (furigana) tags. Most people in my experience only recognize about 1/3 of them.

So, in a fish monger's stall, I would expect to see sawara with a label using hiragana, or maybe the kanji plus hiragana. As for katakana, it would be scientific usage. That is, in the biological sciences in Japan, katakana is used for scientific names of species, subspecies, etc.

From Nathan Hopson:

Spanish mackerel is サワラ (鰆) in Japanese. As the kanji suggests, it's used as a spring seasonal word (季語) in poetry, both alone and in compounds like 鰆東風 (さわらごち), a (spring) wind that blows during the サワラ fishing season. There are several etymological hypotheses for the derivation of サワラ, but 狭腹 (さはら [VHM: "narrow belly"]) seems most likely to me.

Alternative kanji include:

(Sorry, small graphics pasted from source.)

Oh, and they're delicious. One of my favorite lunches at the university cafeterias in Japan was a set with sawara braised (?) in miso.

[VHM: 鰆 consists of 鱼 ("fish", the radical / semantophore) and 春 ("spring", the phonophore, but in this case also conveying secondary semantic significance)]

From Hiroko Sherry:

The kanji consisting of fish +spring, is the standard. 国語審議会(?)(maybe 動物学会/植物学会)[VHM: these are the National Language Council, the Zoological Society, and the Botanical Soceity] has decided to use katakana for animal/plants names. (I understand that they adopted this system over 50 years ago.) I am just guessing here, but probably because kanji names have variation, just like your list shows. Besides, if they are written in katakana, they stand out in the kanji+hiragana sequence. Some Japanese hate this system because katakana destroys the images/feelings which kanji names resonate. However, I think this way of visual presentation probably helps us to get the information from the sentences faster. Not good for literature, but good for science, I guess.

From Richard Warmington:

Katakana are used for other purposes than just writing loanwords. For example (from Wikipedia):

Technical and scientific terms, such as the names of animal and plant species and minerals, are also commonly written in katakana.[5] Homo sapiens (ホモ・サピエンス Homo sapiensu?), as a species, is written ヒト (hito), rather than its kanji 人.

The Japanese Wikipedia article for "rabbit" is titled ウサギ (katakana), and searches for "usagi" written in hiragana or kanji are redirected to the article with its title written in katakana:


I would add that the use of katakana for species names is not limited to technical writing. It commonly appears in everyday usage.

e.g. ペットといえばウサギ♪なんでそんなにカワイイの♪
"Speaking of pets, why are rabbits so cute?"

(This quote from a blog also illustrates the use of katakana for emphasis (similar to the way italics are used in English) – in this case, of the word "kawaii" (cute) – and for the rendering of a loanword, "petto" (pet) (also similar to the way italics are used in English to render loanwords.))

Apart from the issue of katakana usage, the kanji that is usually indicated in Japanese dictionaries for this fish (鰆) is not included among the common use (常用) kanji, and that would be another reason not to write the name of the fish in kanji, I presume.

From Jim Breen:

The kanji used is 鰆, which has the "kun" reading of さわら (sawara).

As commonly occurs with the names of plants, animals, etc. this is often written in katakana, hence the サワラ.

My dictionary entry: link, Wikipedia: link.

A very popular fish in Japan – common in the Inland Sea and considered very tasty.

Invariably called さわら. The generic term for mackerel is さば (saba), but this is not used for the (Japanese) Spanish mackerel as it's usually applied to the "chub" mackerel (Scomber japonicus).

You can't predict whether the kanji, katakana or hiragana written forms of these critters will be used. In this case, it seems to be usually in hiragana, which in the Google n-grams leads サワラ by about 2:1. The kanji form is used quite often too. The only other use I can see of the kanji is in the name of the "wahoo" (kamasusawara) which is nothing like as common. It has the kanji forms of 叺鰆 and 梭子魚鰆 but is invariably written カマスサワラ.

You sent some other kanji used for mackerel. The only ones I know are: 鯖 – saba さば
and 鯵/鰺 – aji – あじ horse mackerel. Also very common and popular.

From Michael Carr:

Kana are used because 鰆 is not in the jōyō kanji [ "regular-use Chinese characters"]. I’ve only seen it used in sushiya [寿司屋 "sushi shops"].

Here are the low-hanging fruit links (from Wikipedia/Wiktionary):

Japanese Spanish mackerel =
サワラ =–
"stone-head fish"!

Not in Wiktionary:

The Chinese dragon names article mentions lots of fish.

I wrote a paper in grad school about the Erya chapter on fish names, which was a trial of how to proceed with my dissertation on the botanical name chapters, but it wasn't publishable.

See also Michael Carr, "Tiao-Fish through Chinese Dictionaries", Sino-Platonic Papers, 40 (September, 1993), 68 pages.

Chinese character variants for bà 鲅 are given here.

From this article and other sources, we know that it really is called Japanese Spanish mackerel, that its scientific name is Scomberomorus niphonius, and that it is also known as the Japanese seerfish (sometimes written as searfish.

According to Wikipedia, there's a whole biological tribe of Spanish mackerels, including the Atlantic Spanish mackerel, the Monterrey Spanish mackerel, the narrow-barred Spanish mackerel, the West African Spanish mackerel, and the Serra Spanish mackerel, and still many more species that don't have "Spanish" in the name, even though they're members of the "Spanish mackerel" tribe.

Finally, I wish to report a strange phenomenon that I noticed once when checking the frequency of the characters in a fairly large set. I forget whether the magnitude of the set was in the 10,000 or 20,000 range. In any event, there were far more characters in the set than anyone could possibly remember or know. What was striking to me was the fact that an overwhelmingly disproportionate number of characters having extremely low frequency, and often a large and cumbersome quantity of strokes, were names of fish! Having done all the research necessary for this post, I have come to the conclusion that characters for the names of fish are often arcane, arbitrary, and recognized by very few people. What matters is the spoken pronunciation of the names, since the people who most often deal with fish (fishermen, fishmongers, housewives, and so forth) are apt to have low levels of hanzi / kanji literacy. Furthermore, fish terms vary greatly by region, even within the same language (e.g., Japanese, Mandarin), and, as one of the experts cited above pointed out, even by season.

[Hearty thanks to everyone quoted in this post]


  1. leoboiko said,

    November 13, 2013 @ 2:00 pm

    Precisely because of this complexity and non-utilitarianism, characters for fish seem to be popular in Japanese quizzes, tests, trivia and other forms of kanji virtuosism. For example, searching for "魚 漢字" in Google Play Android Apps returns at least a dozen mobile apps designed specifically to exercise fish kanji (like 1, 2, 3, 4).

  2. Steve said,

    November 13, 2013 @ 2:53 pm

    At the risk of asking a stupid question, am I correct in assuming that the Japanese word "baiyu" is not made up of characters that mean anything like "S

  3. Steve said,

    November 13, 2013 @ 3:16 pm

    There are few things as embarassing as deciding NOT to submit a comment, and then accidentally submitting an incomprehensible and fragmented version of it.

    Anyway [awkward cough] the dumb question, which I had thought better of asking, was: Do the characters / units of speech that comprise "Bayu" have individual meanings that are anything like Japanese + Spanish + Mackerel, or, as I suspect, is it simply the name of a fish that, in English, is called a Japanese Spanish Mackerel?

  4. John said,

    November 13, 2013 @ 3:17 pm

    Um, so Chinese wiki calls it 蓝点马鲛, 日本馬加鰆, 正馬加, 土魠, but not 鲅鱼. What gives?

    How would you say 鰆 in Mandarin? Chun1?

  5. John said,

    November 13, 2013 @ 3:23 pm

    Steve: 鱼 (yú) means fish. Several, but not all fish names in Chinese are (randomcharacter)鱼.

    The random character on its own has the same meaning, i.e. 鲅 and 鲅鱼, in writing, are interchangeable. However, if you just say 鲅 in speech, nobody will understand what you are talking about, hence the need to append a 鱼, i.e. "the fish called 'ba'"

    The English name of anything is rarely related to the Chinese name, except when it is a direct sound transliteration (or by convention, such as obscure element names)

  6. Rubrick said,

    November 13, 2013 @ 6:04 pm

    I'm a little confused about the part regarding katakana being used for scientific names of species. The implication seems to be that this is a case where katakana is used for something other than foreign words, but Linnean names are (by fiat) Latin. Isn't that a pretty textbook use case for katakana?

  7. Ken Brown said,

    November 13, 2013 @ 6:05 pm

    Systematic names of fish, or any other organisms, are governed by some very prescriptive international conventions,and in Latin alphabet the binomial names are printed in italics. I doubt if it would be possible to write them in kanji (or Chinese characters). Katakana might be needed. (Though I'd guessJapanese and Chinese scholarly writing mostly uses Latin for them)

  8. Ethan said,

    November 13, 2013 @ 7:00 pm

    @Rubrick: The Linnean name is apparently Scomberomorus niphonius, which is nothing like sawara/サワラ/さわら/鰆. I think the only question is why, having decided not to use the kanji character, a native speaker would choose katakana over hiragana.

  9. Leonardo Boiko said,

    November 13, 2013 @ 7:17 pm

    @Ethan: I can't look it up at the moment but I've heard katakana is the convention adopted to substitute the use of kanji for animal and plant names, which, as exemplified by this post, is particularly cumbersome. As for why not hiragana, I'd guess it's because in adult Japanese text hiragana tends to be used for inflections, particles and function words, while kanji and katakana more commonly take up the content words. Since the script is unspaced, this specialization of character shapes helps in finding words. Compare:


    Also for this reason, hiragana-only children's texts will add spaces to help in delimiting words. (It's interesting that they don't add spaces in the same places that we do when writing romanized Japanese.)

  10. Matt said,

    November 13, 2013 @ 8:10 pm

    I'd have thought the Japanese would know better, but there's at least one Japanese page that apparently also uses baidu.

    That Japanese page is located inside a Chinese encyclopedia, though, and the translation clearly wasn't done by a native speaker. (I'm not even convinced it was done by a human. Surely anyone in the CJK area would balk at the use of 人口 to refer to a population of fish.)

  11. Jim Breen said,

    November 13, 2013 @ 11:13 pm

    I confess to being a little sceptical of Nathan Hopson's "As the kanji suggests, it's used as a spring seasonal word (季語) in poetry…" I don't think the presence of 春 as part of the 鰆 kanji means anything apart from providing the ON reading (シュン).

  12. RussianFisherMan76 said,

    November 14, 2013 @ 7:54 am

    I think it's might be helpful (found in fishbase):

    Synonyms for latin name:
    Scomberomorus niphonius (Cuvier, 1832)
    Sawara niphonia (Cuvier, 1832)

    Common names from fishbase:

    China: Bà yú
    Japan: Sawara

    Publication with bayu mentioned:
    Yamada, U., S. Shirai, T. Irie, M. Tokimura, S. Deng, Y. Zheng, C. Li, Y.U. Kim and Y.S. Kim, 1995. Names and illustrations of fishes from the East China Sea and the Yellow Sea. Overseas Fishery Cooperation Foundation, Tokyo, Japan. 288 p.

    Publication with sawara:
    Masuda, H., K. Amaoka, C. Araga, T. Uyeno and T. Yoshino, 1984

    Page of one of the departement of Russian Academie of Science with common name SAWARA:

  13. Victor Mair said,

    November 14, 2013 @ 10:06 am

    The Chinese dragon names article referred to by Michael Carr may be found here:


  14. leoboiko said,

    November 14, 2013 @ 10:24 am

    @Jim Breen: As Prof. Mair noted, in many cases (though not always) the sound element of a Chinese character may also carry "secondary semantic significance". Consider the 包 hou "wrap" component: Besides being a perfect phonetic, 包 clearly suggests meaning in characters like 抱 "hug", 泡 "bubble" or 胞 "placenta, sac".

    Since we don't have detailed records of what the character creators were thinking, and since it's easy to create semantic associations a posteriori, I think it's often hard to distinguish whether a component was originally intended as only a pronounciation hint, or as a pronounciation hint with a side-help of mnemonic association.

    At any rate, I think that even a posteriori associations may be culturally important. For example, some say the character for "east", 東, is a phonetic borrowing that originally meant a basket or bundled package (related to 束—compare the oracle forms). But all Japanese I've asked about it think of it as "the sun behind a tree"; even if this analysis is historically incorrect, it's synchronically the dominant model. In the same way, sawara are definitely a "spring fish" in seasonal poetry, so even if the 春 element was originally intended strictly as a sound-hint, I bet contemporary people will understand the character 鰆 as "spring fish". (I wonder—could it be possible that the sawara fish acquired this role in poetry because of the Chinese character, or at least influenced by it?…)

  15. Nathan Hopson said,

    November 14, 2013 @ 10:54 am

    @Jim Breen:

    Point taken. I may have been a bit sloppy about the cause and effect there, but sawara is a spring fish with spring in the kanji, and the reference source I was looking at suggested a connection.

  16. Matt said,

    November 14, 2013 @ 11:47 am

    … even if the 春 element was originally intended strictly as a sound-hint, I bet contemporary people will understand the character 鰆 as "spring fish".

    The 日本国語大辞典 has the following citation:

    *名語記〔1275〕八「魚の名のさはら如何。鰆とかけり。はるのひいでくれば春の魚とつくれり」 ("What about 'sawara', the name of a [type of] fish? It is written 鰆. The construction [meaning] 'fish of spring' is used because [the fish itself] comes out on the first day of spring,

    … So people have been understanding the character that way for at least 700 years.

    Some sources claim that 鰆 was first applied to the fish in question in Japan (either reinvented independently or repurposed, but either way with no connection to whatever fish 鰆 referred to in China/Korea). I have no idea if this is true, but if so it would strongly imply that the parts were chosen for purely semantic reasons (because that's how Japan rolls when it comes to "national characters" 国字).

    In other words, it might be the case that the 春 in "鰆-as-in-Spanish Japanese mackerel", the Japanese usage, can legitimately be interpreted as purely semantic even if the 春 was purely phonetic when the character 鰆 was first invented in China!

  17. richard said,

    November 14, 2013 @ 2:30 pm

    I would say that the biological use of katakana for species etc. names is perhaps better thought of a specific example of a more general academic practice: using katakana for abstract nouns as opposed to specific instances. For example, in my field (ethnomusicology), if we talk about a koto we use 琴, but if we talk about "the koto" or use koto to gloss all Asian long tube zithers (qin, zheng, tranh, geomun'go, gayageum, etc.), then we use katakana: コト. I think of it as signalling the difference in philosophy between beauty and Beauty.

  18. Ken Brown said,

    November 14, 2013 @ 5:41 pm

    Apparently sawara can be called Spanish mackerel in English. So /Sawara niphona/ might indeed be Japanese Spanish mackerel.

  19. Victor Mair said,

    November 15, 2013 @ 12:43 am

    From Michael Carr:

    This is definitely a kigo for spring (the 鰆 = 春 "spring" + 魚 "fish" character pun / folk etymology is too good to overlook), specifically chūshun (zhòngchūn) 仲春 "middle of spring". While not found in the English WP Kigo list, it is in the Japanese interwiki list, under Shokubutsu 食物 "Foods" for Chūshun.



  20. Victor Mair said,

    November 15, 2013 @ 12:56 am

    From Miki Morita:

    It is interesting to see how to give fish a common name or translate them from/into foreign languages is an ongoing issue (http://www.fish-isj.jp/iin/standname/index.html). I also found an article about names of fish and seafood recorded in "engishiki", a Heian-period book on laws and customs (attached file). I cannot go into the details, but pick up some interesting examples in the article.

    あゆ 鮎 年魚
    In Engishiki, the fish that we call あゆ (sweetfish [Plecoglossus altivelis]) is written as both 鮎(considered to be a later addition) and 年魚 (older and more common at that time). However, the original meanings of 鮎 in Chinese is catfish, and 年魚 is salmon. In Engishiki, namazu(catfish) is written as 鯰, and salmon as 鮭.

    This author does not refer to the origin of ”あゆ” itself, but I found its references in Nihon shoki and Shosoin monjo in a website. But I am not sure if this one is reliable or not.

    What happened in the process of selecting kanji for ayu?
    I am not sure if I understood what he says in his article, but I was amused to know a little about all these mysterious combinations between Chinese characters and fish names.

  21. Marcos said,

    November 15, 2013 @ 3:01 am

    To be fair to hanmeng, "Japanese Spanish mackerel" looked made up to me as a native English speaker, for whatever reason. I was quite surprised to discover it was a real thing. Maybe I'm just not used to seeing two nationalities used to describe one non-human noun?

  22. Daniel said,

    November 15, 2013 @ 2:28 pm

    And presumably if this species of fish were to be cultivated in Canada, it would be "Canadian Japanese Spanish mackerel" (…or in Spain: "Spanish Japanese Spanish mackerel"…).

  23. Daniel said,

    November 15, 2013 @ 2:51 pm

    Apparently this species occurs off the coast of Australia – so presumably "Australian Japanese Spanish mackerel" is a real thing. Google, however, knows only about the existence of "Australian Spanish mackerel."

  24. Daniel said,

    November 15, 2013 @ 3:12 pm

    One Tokugawa-period treatise lists sawara, but written with the Chinese characters for majiaoyu (a related species of mackerel) rather than the single character "fish+spring". It mentions that this species is most numerous from the fifth to the tenth month, so it is unclear why it should have become a seasonal word for spring, unless that is when it tastes best.

    (Apologies for the vagueness – I tried to submit a different version complete with details, kanji, etc. but something prevented it from posting.)

  25. Daniel said,

    November 15, 2013 @ 3:14 pm


  26. William Steed said,

    November 17, 2013 @ 8:27 pm

    With regards to the surprising number of characters for types of fish, one of my Chinese lecturers (an Anglo guy, but teaching Chinese) talked about the huge number of words for types of horses in the 词海. I suspect that for 文言 at least, there would have been a proliferation of rarely used but readily understandable ad-hoc characters for horses turning up. The same proliferation could be used for fish.

  27. Alon Lischinsky said,

    December 7, 2013 @ 7:46 am


    Linnean names are (by fiat) Latin

    That is a common misconception. While terms are often Latinized for use in binomial nomenclature, there is no requirement to do so, and exceptions are plentiful. Ipomoea, already assigned by the old man himself as a genus name for the morning glory, is pure Greek; Naja naja, the spectacled cobra, has a fittingly Sanskrit name, and a genus of beetles is named Ytu after the Guarani word for ‘waterfall’.

    I don't know any binomial names fully derived from Japanese, but it's far from impossible.

RSS feed for comments on this post