When people's names are listed with both sides aligned (either in horizontal or vertical writing), for example, in the ending scroll of a movie, there is a seemingly complicated rule (to me).

When the length of the name (family plus given) is not three characters, the positioning is such that the characters are distributed evenly, without particular segmentation between the family and the given name parts:

X* = family name, Y* = given name

X          Y
X   Y   Y   Y
X   X   Y   Y
X   X   X   Y
X  Y  Y  Y  Y
X  X  Y  Y  Y
X  X  X  Y  Y
X  X  X  X  Y

However, when the length is three, it seems as if a space character is inserted between the family and the given parts and the positioning is done counting in the space character, like this:

X       Y   Y
X   X       Y

Considering the rules for the other cases, I would rather expect:

X     Y     Y
X     X     Y

Why is it like as is? Is there a rule behind that makes the length-3 cases non-exceptional, or is there a reason why length-3 is exceptional, or, is it just due to inconsistency/stupidity of whomever started it?


Or to see it from the other side, it is natural that a space be inserted in between the family and the given names (as Dave points out in the comment). Then, why are the cases other than length-3 (except the length-2 case, which can be interpreted in either way) not having a space? I would rather expect:

X     Y  Y  Y
X  X     Y  Y
X  X  X     Y

and so on.

Edit These are links to some pictures that describe the situation (with some variations):

  • Couldn't it be that length-3 (and therefore only one "sure" kanji on each side) makes it too difficult to instinctively guess whether the middle one belongs to the last or first name? (I'm sure in most cases, it can be guessed, but I'd imagine it can take some effort and time, which is not a good idea for rolling credits)
    – Dave
    Apr 11, 2012 at 0:51
  • @Dave If that is the case, why doesn't it happen with other lengths? And this is not just in movies. The same thing happens, for example, in a printed form of a signature-collection on a political issue.
    – user458
    Apr 11, 2012 at 1:03
  • Moved my comments to an answer: feel free to move or delete your comment.
    – Dave
    Apr 11, 2012 at 3:30

3 Answers 3


I believe the rule you are looking for is called 人名字取り. 5字取り is the most commonly used for the closing credits in movies. What this means, is everything is aligned based on 5 characters.

Here is an image for reference (link).

enter image description here

I found this other reference which has a few more patterns (posted below). If I'm understanding the question correctly, 5字取り or 7字取り pattern B seems similar to what sawa is explaining. However, since there are many different patterns, it might be impossible to find an exact match.

enter image description here

  • 2
    Wow, you found detailed information. I have never seen such information on that. It is interesting. That is one case. But there are also cases where length-4 names are evenly distributed regardless of its composition, and I see that one more frequently.
    – user458
    Apr 12, 2012 at 1:17
  • 2
    @sawa: I found a link, I believe the case you are talking about is either 5字取り or 7字取り pattern B.
    – Jesse Good
    Apr 12, 2012 at 1:42
  • 1
    This is great! Kudos for finding this. However, I would point out that the layout presented is very different from what @sawa originally mentioned: in this image, last and first names are clearly separated by spacing in all cases...
    – Dave
    Apr 12, 2012 at 14:42
  • @Dave: Yes, this is only one example, the other link I posted in the comments has a few more examples.
    – Jesse Good
    Apr 12, 2012 at 20:28
  • @Jesse: I might be missing something, but I still do not see anything looking like what Sawa described in the page you linked. If there is an example matching his description, could you post it instead of the one currently up there? (fwiw, the illustration above is what you would get with basic kerning rules).
    – Dave
    Apr 15, 2012 at 3:51

Making my previous comment an answer, since it turns out to be not as obvious a suggestion as I thought

I think we can all agree that the particular spacing makes it easier to read the 3-kanji length names, so the question is why this is done only for 3 kanji.

My personal guess is that this is because 3-kanji last+first names combinations are particularly hard to "parse", compared to other lengths:

  • 2 kanji is trivial

  • 4 kanji is statistically very likely to be 2+2 (considering how much more common 2-kanji names are than 1- or 3-kanji).

  • 5 (and more) kanji means that at least the first 2 kanji belong to the last name, making it much easier to decide "on the fly" whether the 3rd belongs to the last or first name.

Remains the case of 3 kanji, where only one kanji on each side belongs for certain to last/first name. Knowing whether the middle kanji belongs to the last or first name could prove difficult without some deeper analysis (and since we are talking about rolling credits, the ability to be read easily would be quite essential).

This approach seems a decent compromise between aesthetics and legibility. It seems strange that it's been adopted as a standard in all Japanese movie credits (I never realised it to be the rule, but definitely saw it often). Maybe it has its root in some previous older media (e.g. lists of authors on a book?).

  • Actually, I don't agree with 3 kanji being difficult to guess. In Japanese 3-length names, 2+1 is very likely. For Chinese, 1+2 is very likely. The likelihood of 2+1 over 1+2 in Japanese is more than the likelihood of 2+2 over 3+1. (Probably 1+3 is not much popular).
    – user458
    Apr 11, 2012 at 22:59

What if the rule is "Do not insert a gap that is larger or equal to one fullwidth space (sometimes used to separate names), unless it breaks up the family name and the given name"?

Names with three characters put a larger space before/after the one-character name, in order to make the spacing between the characters of the two-character name smaller than one fullwidth space, so that it can be read as a single name more easily.

Names of four characters or longer are long enough within the typeset area that the spacing does not need to be adjusted - the characters are close enough together for the name to be read naturally.

I suppose that if you can find an example of a credits screen where the typeset column for names is so wide that even four-character names have greater than one fullwidth space between the characters, it would shoot a hole in this theory.

(I also agree with Dave's assessment that "length 3 names are the hardest to parse".)

  • I think you are grasping something similar to what I had imagined before, but I cannot fully understand what you are expressing.
    – user458
    Apr 11, 2012 at 21:56
  • @sawa I'll edit to make it more clear if you like, but I'm not sure it's worth it; I have found a book that has a 3-character name spaced (vertically) as 早  川   浩. 早 and 川 are more than one space apart, so my argument doesn't hold water. It is typeset differently from the 4-character names, but still the 川 is not central - enough to make it clear how to parse the name, but not enough to support my argument. I think Dave is closer than I am to the truth here - for 3-character names it helps to have a visual cue as to the correct parsing.
    – Hyperworm
    Apr 11, 2012 at 22:24
  • @sawa I would also ask (just to be sure) how confident you are that the patterns X X X Y and X X X X Y don't get special spacing, to rule out that it's something special with one-character names.
    – Hyperworm
    Apr 11, 2012 at 22:26

You must log in to answer this question.