Backward characters

« previous post | next post »

Name on a ship that docked in Yancheng (in Jiangsu province) harbor last Thursday:

The reason there are armed public security forces patrolling near the ship is because it was full of smuggled cargo.  The story is reported here:

"Smugglers caught because they got their Chinese characters the wrong way round:  Language blunder gives sugar carriers a bitter lesson after it attracts coastguards’ suspicions" (SCMP, 9/5/17)

The name the smugglers wanted to affix to the ship was Dàqìng 大庆 (the "Oil Capital" of China, located in Heilongjiang Province).  However, after they hastily painted over the old name of the ship with the wrong shade of blue, they reversed the direction of the stencil to apply the new, false name, so that Dàqìng 大庆 came out with the order doubly wrong:  the name read Qìngdà 大 instead of Dàqìng 大庆 and the orientation of the individual characters was inverted.  Interestingly, the transposition of dà 大 doesn't really matter, because it looks the same whether viewed directly or in a mirror.  As it appears on the ship, inverted, 庆 cannot really be read (without mental transposition), but if you look at it in a mirror, it has the correct orientation.

Similar transpositions are described in these posts:

[h.t. Dean Barrett]


  1. Jim Breen said,

    September 7, 2017 @ 8:53 pm

    You see these sorts of things a lot when newspapers have photos with hanzi/kanji in them and the people doing the layout reverse the images to make the balance better, not realising they've screwed up the text. I've also seen advertisements for Kurosawa's "Ran" with the 乱 in mirror image.

    I may have mentioned it here before, but in the 80s a leading stock broker did a big sales pitch in Hong Kong and their specially-printed brochure had 香港 on the front, with the 香 upside down.

  2. Bruce said,

    September 7, 2017 @ 11:24 pm

    Maybe the painters verified their work by sending a selfie to a Chinese speaker

  3. AntC said,

    September 8, 2017 @ 12:39 am

    @Jim You see these sorts of things a lot when newspapers …

    Not just newspapers. A lot of Youtube videos seem to be reverse images, irrespective of what language appears in them. (Presumably to evade copy detection software?)

    So my first question would be: how do we know the photo of the ship is not reversed, rather than the characters on the ship? The SCMP story confirms that it was the reversed characters that drew suspicion, and presumably the characters on the shields of the "security forces" are correct in the photo.

  4. David Morris said,

    September 8, 2017 @ 1:26 am

    I am a total non-speaker and reader of Chinese (though I do handle documents written in Chinese regularly) but something about the first character just looks strange. I just know about radicals as part of Chinese characters. Do these always – or usually – go to the left of the main part?

  5. flow said,

    September 8, 2017 @ 4:13 am

    @David Morris—that reversed character is actually not super strange or typographically impossible; for example, there are 丮卂厾虱 so I think there could plausibly be a character ⿱丶⿹⺄大. These characters are, however, admittedly not very frequent ones, their composition is not very typical, and the ⺄ stroke has a hook at the end that the fake character pictured above is missing. OTOH there are things like 𢎡, 𡆡, 𩡬 that do look broken or incomplete from the standpoint of your average Chinese character, so it's hard to rule out the possibility of an actually existing character similar to the pictured one altogether purely based on calligraphic / typographic considerations.

    Coming to think of it, someone should submit a proposal to the Unicode consortium to include ⿱丶⿹⺄大 in a future version of the encoding, on the grounds that (1) the character has been used in real life, (2) it has achieved a considerable dissemination, (3) has a known reading and meaning.

  6. David Marjanović said,

    September 8, 2017 @ 6:34 am

    presumably the characters on the shields of the "security forces" are correct in the photo.

    They are.

  7. krogerfoot said,

    September 8, 2017 @ 6:47 am

    Chinese and Japanese characters on the starboard side of a vehicle are often reversed in order (though not mirrored), so they are read from the front to the back. E.G., this China Southern Cargo aircraft, with "空 航 方 南 国 中" on the side. I'm not sure if ships do this too, but that might have contributed to the smugglers' confusion. Attention to detail seems not to be their strong suit.

  8. B.Ma said,

    September 8, 2017 @ 6:58 am

    I disagree that the mirrored image is "doubly wrong"; it still reads Dàqìng 大庆 and not Qìngdà 庆大.

    You still read ƎƆͶA⅃UꓭИA as AMBULANCE rather than ECNALUBMA. See this example from Hong Kong for the equivalent in characters:

    It could possibly read Qìngdà 庆大 if it had been painted on the other side of the ship – the reason is given in krogerfoot's comment.

  9. Victor Mair said,

    September 8, 2017 @ 7:46 am

    Kangxi radical 53, 广, as in the qìng 庆 of Dàqìng 大庆, has a dot at the top, distinguishing it from Kangxi radical 27, 厂, and the ⺄ mentioned by flow.

  10. unekdoud said,

    September 8, 2017 @ 7:49 am

    Even if it was painted unmirrored as 庆大 (as would be common in other categories of Chinese signage), I would consider that as being written right-to-left rather than in reverse, so it would still be read Dàqìng.

  11. Victor Mair said,

    September 8, 2017 @ 7:54 am


    I was talking about viewing the characters on the prow of the ship head on, not in a mirror. As for reversing a string of letters or characters by viewing them in a mirror, I'm quite aware of what happens when you do that, and went to great lengths to demonstrate it in the two posts cited at the end of this o.p.

  12. Jonathan said,

    September 8, 2017 @ 10:30 am

    @B.Ma: I'm probably not in the majority, but 'ecnalubma' is a word I use. It means 'an ambulance that has its label painted on backwards'. It's undergone a bit of metathesis, and is sometimes pronounced 'ecnalumba'.

  13. Victor Mair said,

    September 8, 2017 @ 11:22 am

    @David Morris

    Your instinct / intuition is correct. Radicals like 53 广 and 27 厂 go on the left side.

  14. David Morris said,

    September 8, 2017 @ 5:50 pm

    flow: thanks.

    Victor: but, as flow has explained, radicals *can* go on the right.

  15. Guy said,

    September 8, 2017 @ 8:36 pm

    @Jim Breen

    Also it seems unlikely that the three people in the photo would be carrying their shields in their right hands and holding the batons in their left hands.

  16. Jim Breen said,

    September 8, 2017 @ 8:42 pm

    @Guy. I never suggested it was a reversed photo. I could see that from the hanzi on the shields.

  17. Guy said,

    September 8, 2017 @ 9:24 pm


    Sorry, I meant to respond to AntC. Careless error on my part.

  18. Victor Mair said,

    September 8, 2017 @ 11:17 pm

    @David Morris

    Of course, a few radicals can appear on the right, but the element flow wrote about (⺄) is not a radical. Your initial instincts were right.

  19. AntC said,

    September 9, 2017 @ 4:55 am

    @Jonathan … a bit of metathesis, and is sometimes pronounced 'ecnalumba'.

    A little more metathesis, and it'll get pronounced 'encalumba'. Which is Spanish for "it hardens"? Or Zulu for "in vain". (Both from Google translate.)

  20. flow said,

    September 10, 2017 @ 6:28 am

    I'm afraid I have to make clear that whether ⺄ and 广 and 厂 are radicals or not is quite non-substantial to the question whether or not the can or can not go on the right or left of a character, and also whether the character shown in the picture is possible or not. The term 'radical' is a western invention, and ultimately a misconception. It does have some justification, but to state that 广 'is' a radical while ⺄ 'is' not—is just saying that one Mr. Mei Yingzuo included the former but not the latter in his list of 214 graphical components (including some variants) that he found would make for a good classification system for his dictionary, published in 1615. It is nothing inherent or principled about 广 and ⺄ that makes one of them appear in that list and the other not.

  21. Victor Mair said,

    September 10, 2017 @ 10:27 am

    "I'm afraid I have to make clear…".

    Why are you afraid? What are you afraid of?

    Call them bùshǒu 部首 (lit., "section headers"), radicals, or what have you, the concept, existence, and use of semantic classifiers for Chinese characters have been around since the beginning of the 2nd c. AD (Shuōwén Jiězì 說文解字 ["Explaining Graphs and Analyzing Characters"]). It is sheer nonsense, or errant pedantry at best, to declare that they are meaningless or imply that they were invented by Westerners.

    The system of 214 Kangxi radicals has been in use for more than four centuries. Monuments of Sinitic lexicography such as the Dai Kan-Wa Jiten 大漢和辞典 ("The Great Chinese–Japanese Dictionary") of Morohashi Tetsuji are arranged according to this system of semantic classification. When I was a graduate student, my classmates and I memorized the shapes, meanings, pronunciations, and numbers of the 214 Kangxi radicals, and we relied on them to do our Sinological spadework.

    丮, 卂, 厾, and 虱 are not listed under a supposed ⺄ semantophore. Furthermore, as I pointed out earlier in the comments, the mirror inversion of 广 is not ⺄. Even those who are not literate in Chinese characters can see that at a glance.

    There's a lot more that I'm prepared to say, and many LLog posts to reference, if what I've written here is insufficient to bring a halt to this vain scholasticism.

  22. flow said,

    September 11, 2017 @ 4:22 am

    So know I'm really afraid I have to write in once more. First, when I wrote I was afraid, I was afraid that people could be tired by a lengthy comment, and I apologize for that.

    My comment was triggered by this exchange:

    > David Morris said, "Victor: but, as flow has explained, radicals *can* go on the right.

    > Victor Mair said, "Of course, a few radicals can appear on the right, but the element flow wrote about (⺄) is not a radical."

    When I wrote "The term 'radical' is a western invention, and ultimately a misconception." I should have made it clear that the misconception is on the one hand in the term itself, when understood as, say, "root element" or "fundamental part of characters", and on the other hand in the thinking that "all Chinese characters are made up of 214 fundamental parts (and their variants), and such an analysis is exhaustive and sufficiently explanatory".

    I did *not* mean to say that "the 214 Kangxi radicals are a useless and / or misguided way to build a dictionary or to get orientation as a learner of the script". Quite the contrary, most of the 214 radicals are absolutely essential for writing Chinese, and with all their shortcomings, their introduction was a huge step forward as far as Chinese dictionaries are concerned.

    You point out out that "丮, 卂, 厾, and 虱 are not listed under a supposed ⺄ semantophore. Furthermore, as I pointed out earlier in the comments, the mirror inversion of 广 is not ⺄. Even those who are not literate in Chinese characters can see that at a glance".

    I do not think I suggested that 丮, 卂, 厾, or 虱 contain ⺄ as a semantophore. They do contain ⺄ as graphical element, though.

    And you're right, neither 广 nor 厂 are exact mirror images of ⺄; apart from the dot in 广 there is also a difference in stroke details: 厂 ends in a downwards, pointed stroke, ⺄ has an upwards hook at the end. The character painted on that ship's hull has a pointed downwards stroke that goes to the right, which is, I believe, not possible in Chinese characters; examples like 大 and show how that right-going stroke is 'flattened out' when it has no hook. But these are finer details, akin to serifs in the Latin alphabet. I meant to present 厾 as an actual character that is "almost 庆 mirrored" (except 去 is not 大 and 广 is not *exactly* like ⺄ flipped, but you get the point).

RSS feed for comments on this post

Leave a Comment