Re: 行取りという考え方について：忘れないうちに

Hi Nat,

I completely agree with your view on the issue. (It's not surprising, if
it's rare).

A Japanese character grid allows for an area to be defined in terms of
> typographic bounds, and to place lines of text in that area according to
> even increments of the underlying grid.


This is very true in Japanese typography, because still today there *is *the
notion of "type body" unlike Western typography after the introduction of
phototypesetting. We say that a line-length (measure) should be specified
with an integral  multiple of the type size used. Also, we often specify a
leading with a factor with respect to the type size used; for example, "12
Q、行間二分四分アキ" (12 Q, 0.75 x type-size leaded). Once these are specified, the
specified grid constrains the typesetting properties unless some or all of
them are changed explicitly for some reason. I understand that there are
cases where you need to make only a few words in a line larger for
emphasis. You can change the type-size of such particular words.  But even
after changing the type size, the leading remains unchanged. And the
leading cannot be changed automatically and implicitly by anyone. The
pre-existence of the underlying grid prohibits it. This is one of the
essential principles of Japanese typesetting.

The *gyo-dori *line-spacing method and the ruling and control of the
typesetting properties of an entire paragraph (or of paragraphs, pages or
whatever) by those defined by their prescribed type-body-based grid are
still existing today, because of the survival of the concept of type body
in Japanese typography.

If it is thought that interculturization is necessary in today's
technological environment where human words written in a wide variety of
languages in the world are used even in a document or a web page, elements
such as the EM-body based typographical grid in CJK typography should be
respected, and we need to find further generalized solutions that do not
force Japanese developers and users to acquiesce to Roman-centric rules.

However, I think that the use of typographic grids have worked well also in
Western typography. I believe many serious typographers in the West agree
with it.

Although this is related to the issue of 'line-height', which I mentioned
briefly at the previous meeting, if I may say, I personally tend to think
that there is a bad habit in the world of web; whenever an attribute gets
specified, some other attributes also tend to be modified implicitly
without any explicit specification or order.

This does not seem to be an issue of "web or print" or "paper or display
screen". But it seems to be an issue related to: *hidden dependencies* and
*transparency*. From typographic viewpoints, it is about *clarity *and
*consistency*.
I remember that in the late 1970s, or earlier, serious Fortran programmers
complained that when a variable name began with i, j, k, l, m, or n, the
variable was implicitly treated as an integer, and they had to write
"IMPLICIT NONE."

Yet, I agree that there are cases where you have to ignore such typographic
grids. Typically, if you insert a large block composed of multiple lines of
mathematical elements (equations, etc.), or a picture, it's natural that
you need to be free from the constraint of the grid for the body text,
though some people may still want and try to return to the grid as soon as
they insert it.

Just my views.

--Taro

2026年5月14日(木) 1:33 Nat McCully <nmccully@adobe.com>:

> As far as the need for 行取り spacing for Japanese, even on the web, I say
> yes absolutely it is needed. The underlying character grid in Japanese
> layout is different from a simple baseline grid for Latin composition, but
> the current CSS specs seem only to be concerned with satisfying the need
> for the latter. A Japanese character grid allows for an area to be defined
> in terms of typographic bounds, and to place lines of text in that area
> according to even increments of the underlying grid. This concept is unique
> to Japanese (and maybe Chinese/Korean), and allows for the white space in
> different text sections to look more harmonious overall. The CSS rhythm
> issue is a variant of this, but seems only to be looking at it from the
> Roman baseline grid point of view.
>
> *Nat McCully*  |  *Principal Scientist*  |  *Adobe*  |  C +1 206 409
> 0624  |  nmccully@adobe.com
>
> ------------------------------
> *From:* MURATA Makoto <eb2m-mrt@asahi-net.or.jp>
> *Sent:* Wednesday, May 13, 2026 2:04
> *To:* Taku Yamaguchi <study.yamahige@gmail.com>
> *Cc:* Yasuo Kida <kida@mac.com>; Japanese Layout Task Force |
> 日本語組版タスクフォース <public-i18n-japanese@w3.org>
> *Subject:* Re: 行取りという考え方について：忘れないうちに
>
>
> *EXTERNAL: Use caution when clicking on links or opening attachments.*
>
>
> 私の知る限り、Florian + Erikaと石井さんとの間で、この方式については大きく意見が分かれています。収束しそうだという話は聞いていません。
>
> 村田 真
>
> 2026年5月13日(水) 17:48 Taku Yamaguchi <study.yamahige@gmail.com>:
>
> 木田さま、みなさま、
> 山口です。
>
> 調べてみました。
>
> 石井さんとエリカさんがeditorやってるCSS Rhythmic Sizingが、10年越しで今日の議論の要件を満たそうとしてる気がします。
> CSS Rhythmic Sizing Module Level 1, W3C Working Draft, 17 February 2026
> https://www.w3.org/TR/css-rhythm-1/

>
> これって「CSSは行取りの必要性を認める」ということでしょうか。
>
> > 1. Introduction
> > ...
> > Controlling sizes of CSS objects to be multiple of a step unit is
> desired in many cases. This level of the specification focuses on following
> cases.
> > - Adjust heights of block-level boxes to multiples of the specified
> length.
> > - Adjust heights of line boxes to multiples of the specified length.
> > By stacking blocks at multiples, authors can align content across
> columns, pages, scroll-snapped bocks or multiple blocks placed absolutely
> to maintain vertical rhythm.
>
> また、「行取りがWebに必要か？」の議論は、前からあるようです。
>
> 2016-05-12、ウェブページにおけるバーティカル・リズム
> https://hail2u.net/blog/webdesign/vertical-rhythm-on-webpage

>
> 2017-04-02、もっと細かい値を基準にしたVertical Rhythm
> https://yuheiy.hatenablog.com/entry/2017/04/02/210214

>
> CSSでバーティカルリズムを実現しよう！line-height-stepを使ってみる, 石井宏治（グーグル）, 2017年11月14日
> https://html5experts.jp/kojii/24718/

>
> CSS Rhythmic Sizing で Vertical Rhythm, created_at2017-10-09,
> updated_at2024-11-15
> https://blog.jxck.io/entries/2017-10-09/css-rhythmic-sizing.html
>
>
> 2026年5月13日(水) 16:26 Urayama <urara@peace.ocn.ne.jp>:
> >
> > 木田さま、小林先生、皆さま
> >
> > 浦山です。
> > 今日の会議は、けっこう頭のなかがシャッフルされました。
> >
> > 私もそうですが、従来の組版作業では、ページサイズ、版面と余白のサイズ、
> > 文字数と行数を最初に決めます。そして、そこに文字や図版を収めていくために、
> > 見出しや表組や図版まわりの処理を「行取り」で考えると、
> > いろいろと楽でしたし、組み上がりもきれいでした。
> >
> > （ただし、小林先生の言われた「端数はここで調整してね」も、
> > 必要に応じて指示していました。）
> >
> > ところが、デジタルの世界では、ページという概念がなく、
> > しかもさまざまなサイズの文字や行間が自由に設定できるので、
> > とくに「行取り」にこだわる必要はなくなったように思います。
> >
> > なので、JLreq-dで「行取り」に触れなくてもよいのかもしれませんが、
> > 私が思うに、「行取り」にはそれなりのメリットがあるので、
> > たとえば、本日の会議で話の出た、段組みがきれいにそろうとか、
> > 複数の領域を同一の仕上がりにそろえたいという要求に応えられるとか、
> > 「行取り」のメリットをいろいろ述べて、
> > それをエンジニアに選択してもらえるとよいと考えます。
> >
> > 「行取り」を使うと、日本語をきれいに組版・表示できる！というのは、
> > 売りになると思いました。
> >
> > 追伸
> > 今月26日のJLTFミーティングは、旅に出るので欠席させていただきます。
> > よろしくお願いいたします。
> >
> >
> >
> >
> >
>
>
>
> --
> Regards,
> Makoto
>