Museum Signs

The Chinese writing system is incredibly efficient, isn’t it? I mean, it’s a pain to learn and we all forget how to write the odd character from time to time, but you can cram so much into such a small space. It’s not just things like the Analects which, in translation, require lines of English to represent the briefest of the sage’s utterances; even making simple arrangements by SMS/text message seems so much more convenient in Chinese.

Then, as Bryan pointed out, there are signs in your local hospital which protrude unreasonably far, simply so that they can helpfully accommodate the English translation.


Watch out. That thing could take your eye out!

I guess this all first came to my attention whilst wondering around museums and other tourist attractions. The sign below is pretty typical.


To my eyes, the sign is dominated by English. Just a handful of Chinese characters do the job.

Occasionally one would see lengthy pieces explaining the history of a monument in some detail, but the accompanying English text would be laughably brief. Obviously, you just couldn’t fit that much on the signs without the foreign language overwhelming the whole show.

But does Chinese really compress more information into a given space?

Perhaps we can do a little experiment, if you’ll indulge me.

If we take an article, on a general topic, for which a translation is readily available, and place the English and in Chinese texts side by side, we might then adjust the font size of the texts until each occupies approximately the same surface area. If we then step back and gradually approach the texts, one would imagine that one of the texts might become legible before the other. This would be a crude test of which language required the greater space to convey a given amount of information.

So, in the interests of linguistic inquiry, I have borrowed an article from the website of the British Embassy in Beijing, in English and Chinese. And, for the sake of harmony, there’s a second article, from the website of the Chinese Embassy in London, in Chinese and English. Apologies to readers less familiar with simplified Chinese characters. I will set up a similar experiment with traditional (non-simplified) characters, should there be demand.

Before you rush to enlarge the following two images, please read on.

Upon clicking each image:

  • Immediately move away from your screen to a distance from which you *cannot* read the text at all (maybe 3 metres or 10 feet).
  • Move slowly toward the screen until you can *just* begin read a few words or characters.
  • Try to read *both* texts without moving your head any closer to the screen and note how far you can read and with which text you read further.
  • Advance, a little at a time, trying to read both Chinese and English, until you find a comfortable reading distance.
  • Make a mental note of places in the texts that cause you problems.

Click on the first article and retire to a safe distance.

Patrick Stewart

Now the second article:

Mongolian Culture

Please leave a comment with the following information.

  1. The language, English or Chinese, you consider yourself to be normally better at reading.
  2. As you first approached the screen, the language you were able to start reading first. (for each article)
  3. As you moved closer, the language which became comfortable for you to read first. (each article)
  4. Any other information you think relevant (e.g. words or characters you found most difficult to identify).

17 responses to “Museum Signs”

  1. matt says:

    Great experiment! The results surprised me a little…

    i. I read English better than I do Chinese (though I have been reading lots of Chinese lately).

    ii. I was able to start reading the English first.

    iii. I was able to comfortably read the Chinese before I was able to comfortably read the English (this is what was surprising to me, in light of ii).

    iv. Numerals were the first bits of text I was able to make out at all or to read comfortably – and I was able to read them in the Chinese side of the text before the English (though I guess that might only be a font issue?).

  2. matt says:

    (oh, everything above is equally true for both articles)

  3. Claw says:

    Matt does allude to a good point about fonts. You use a serif font for the English but sans serif for the Chinese. Serifs have an effect on distinguishing letterforms/strokes at farther distances (this is why road signs are always sans serif) so you should try to use the same serif type in your samples to make it a fairer comparison.

  4. Air Phloo says:

    1. I read English better than Chinese

    2. Some of the less complicated Chinese character became legible first (古,文化,中), even before any English words.

    3. I could comfortably read every English word before I could even make it more complicated characters like 蒙 ,愿望,歌舞

    Same for both articles.

  5. Mark S. says:

    A somewhat related post: Do Chinese characters save paper?

    Chinese typography is subject to what I call the weakest link (or densest Hanzi). If strokes run together (a very different thing than ligatures in alphabetic fonts), then the character is too small and should be enlarged — which means enlarging all of the others too. Because simplified characters comprise only a limited set, it generally makes no difference if the text is in simplified characters (because an extended text is essentially never going to be only in Hanzi that were “simplified” to have fewer strokes) or traditional ones.

    Or at least that’s the way it ought to be. In practice, especially in China, lots of texts (esp. in books) have horrible typography. Perhaps this is improving now?

  6. konw says:

    if you take a look at the sighs at roads, titles, neon light sighs at hotels gates, shops, on top of hospital buildings,etc and compare the ones in chinese and ones in english you will easily find the answer:

    very few chinese letters would make sense(exp, PHARMACY and 药店, or sometimes just one character, a very big neon light of 药 protruding out in the air above the shop gate as you can see in the streets of many hongkong movies, :))
    the same goes for many traffice sighs(tho many of them were pics without texts), notices(exp: 非请勿入), etc

    although , i believe this depends largely on the way you’re writing them, like the phrase 非请勿入, you can make it really long if you like(sth like 未被邀请,请勿入内/如未受到邀请,请不要擅自入内/etc). you can really squeeze many of the chinese sentences into really short if you want…

  7. konw says:

    sorry for the typo,

  8. ETH says:

    i. Better at reading English

    ii. First article: English first. Second article: Chinese first.

    iii. First article: English first. Second article: Chinese first.

    iv. The second article never became comfortable to read. Close enough to read was uncomfortably close to my screen. Characters with the most strokes were hardest to read. Because of that, I would guess that traditional characters will be somewhat harder to read.

  9. ETH says:

    iv. should be:
    The second article never became comfortable to read in English (but was ok in Chinese).

  10. Limao Luo says:

    It’s true that Hanzi isn’t necessarily more efficient at storing information in a given space than English, but this tends to be more in the context of prose; signs, on the other hand are composed of short sentences, phrases, and/or words, in which case Hanzi would easily trump English. In the case of large signs, the use of descriptive phrasing is generally unnecessary(i.e. comprehension and space efficiency is more important), and store names can usually be compressed into to 5 or 6 characters (unlike English).

  11. Lin says:

    1. I read Chinese better than English.

    2. Both Chinese ones are much easily for me to read.(English seem smaller comparing to Chinses, especially the second article)

    3. Both Chinese ones are more confortable for me read first.

    4. I found the first artcle is more equal.

  12. Sima says:

    @Mark S.
    Thanks for the link. I’m sure I must have read your post at some point and should have linked to it in the first place. I very much like the straightforward approach.

    One of your commentators said:

    Its bizzare that your measurements showed English books to be more compact since its common knowledge that the reverse is generally true.

    This kind of comment kind of is inspiring. Where would we be without common knowledge?

    I think your “weakest link” argument is true up to a point, but you don’t need to completely decipher a character (or word) to ‘read’ it. That is to say, if some of the strokes run together, the character may still be legible. An accomplished reader of Chinese no more needs to see every stroke than an accomplished reader of English needs to see every letter. Obviously there are cases when every stroke or letter might need to be recognised but, especially in the context of a text, these are really relatively few.

    If there is a higher number of complex characters (in which the strokes might run together at smaller size or greater distance), then the chances of ‘recognising’ a character from a combination of outline and context is likely to be reduced. Rather than simply a “weakest link”, I suspect that there is a limit to lack of clarity the reader can endure across a text.

    For this reason, I think that the choice of either simplified or non-simplified characters might well prove important for the legibility of a text at a given font size.

    Incidentally, I think the degree of clarity required is also dependent on one’s familiarity with a language and with the concepts likely to be expressed, and this can be seen in other settings. I know that when I listen to Chinese radio in a taxi, I require the volume to be much higher than the driver does (or than I would require it to be for an English broadcast). My ability to piece together elements lost to traffic noise is much lower in Chinese than in English. Likewise, I can decipher pretty appalling handwriting in English, much more so than many of my Chinese friends, but they are generally able to read Chinese handwriting that I simply can’t unlock.

  13. Sima says:

    @Claw, Matt
    Good point about the font. I really should have been more careful.

    Can anyone see any other potential shortcomings ?

  14. Sima says:

    Thanks to those who have left their ‘test results’

    I guess I was pushing it a little with the font size and should have perhaps settled for fewer paragraphs on the page.

    It interests me that for you, at least, “equal” legibility would be somewhere between the two pairs of texts.

  15. Mark S. says:

    Yeah, commenters like that are part of the reason I seldom get involved in comments, even on my own site. If I were to respond just to all of the wonderful comments I receive, I’d feel so guilty about the rest that I’d feel obliged to answer everything — and I’m not so sure that would be in the best interests of civil discourse. So I generally prefer not to open up that can of worms.

    Anyway, I agree that not necessarily every stroke must be distinct for Hanzi to be read. I’m looking at things from a somewhat different approach, though. At some point there’s a line between real reading and context-based guessing. And, oh, could I ever go on about how little attention in general is paid to the difference between knowing and guessing in the case of Hanzi and the question of literacy. But I digress.

    Readers might be able to compensate more or less for texts with strokes all scrunched together, just as readers of texts in the Roman alphabet could compensate for, say, letters to small that the counters and eyes disappear (e.g., the spaces inside a’s and e’s). They shouldn’t ever have to, though, IMO.

    It’s interesting that this happens within a culture in which many people fervently embrace the idea that every stroke is important, because, regardless of what they might practice, they believe that there is but ONE TRUE WAY for stroke order and that those who do not follow it are but ignorant peasants whose foolish behavior in handwriting not only betrays a lack of culture but may put Chinese civilization itself at risk. (Is there a Mandarin equivalent for “Miss Thistlebottom” (i.e., the English teacher who warned us never to split an infinitive, etc.)?) And that’s packaged with the common notion that Chinese characters are inherently beautiful, graceful, and expressive, especially when compared with mere Roman letters. Yet the same culture can be so damn sloppy when it comes to texts, with strokes running together into a big inexpressive lump and an overall typography perhaps best described as fugly.

    All of this is just tangential to the points of your interesting post, though.

  16. Sima says:

    @Mark S.

    At some point there’s a line between real reading and context-based guessing.

    This is a very interesting point. I think context-based guessing (anticipation?) is an essential part of literacy. We all do it as we read and, I think, the better we are at it, the better we read. That said, with a printed text, it should never be essential to guess.

    As for “the one true way”, actually, I think stroke order is very important. Once one starts to write more quickly, the form of the character is lost if the stroke order is not ‘right’ and it becomes very difficult to read. Though it’s not true to say that there is only one way to write a character, there are only a limited number of ways in which the character can be fluently written and remain legible.

    Somehow, Miss Thistlebottom seems to me less likely to emerge in China than in the US or UK. Does anyone know of such a figure?

  17. Sima says:

    One friend (a very capable translator, native Chinese speaker) who had a look at this test, found that the Chinese became legible first, for both texts, but comfortable reading distance came for English before Chinese.

    Just out of interest, does anyone set their web browser to display Chinese and English at different font sizes?

Leave a Reply