How is readability measured in Japanese?

Question

First of all, I apologize if this question is better suited for the meta. It's a little theoretical, but it's also very much about the Japanese language, so I figured it was fine here.

I was looking at the wiki page for readability the other day, which is full of formulas and methodologies for quantifying readability in English, and it got me wondering what this might look like in Japanese. Unfortunately, the equivalent Japanese wiki page (可読性) provides no such convenient formulas, and it appears most of its sources are about English anyway.

A cursory Google search for 可読性 or 可読性の測定法 turned up mostly design or programming related results, things about code readability, etc. I'm not sure if there's a better term for this in Japanese that I don't know, or if I am barking up the wrong tree in some other sense.

I know that there are grade levels assigned to a large number of kanji and that these are taken into account when producing a lot of written work, but what are other ways in which the readability of Japanese text is measured?

Edit: I've had much better luck Googling for 読みやすさ, which seems to be the more commonly used term. Not sure I have a complete picture yet though, so going to leave the question open for now.

Answer 1

It's unfortunate that wikipedia uses 可読性, because the relevant terms here are 読みやすさ and リーダビリティ.

In terms of how readability is actually measured in Japanese, as with English, it looks like there have been a variety of suggested solutions.

This blog post references a paper titled Derivation of a Readability Formula of Japanese Texts, which was published in English as A Computer Readability Formula of Japanese Texts for Machine Scoring. As detailed in the abstract, their formula aims to compute readability from a combination of four "surface-level" (see: no parsing required) things:

1. The average number of characters per sentence
2. For each type of characters (Roman alphabets, kanzis, hiraganas, katakanas), relative frequencies of runs (maximal strings) that consists only of that type of characters
3. The average number of characters per each type of runs
4. 読点 (comma) to 句点 (period) ratio

Numbers three and four were particularly interesting to me as they look at different types of characters, which is obviously not metric we can use for measurement in English.

In terms of more recent work, Study on Japanese Text Readability and "Easy Japanese" gives a nice overview, and consequently also cites several other papers which propose more concrete methods for measuring readability. I think it's a pretty good place to start for a more in depth look at readability in Japanese.

I'm not going to try and reproduce every formula I've seen mentioned, but the suggested metrics were mostly:

• Measurements of the distributions of character types (what % were hiragana, kanji, etc)
• Measurements of the distribution of word origins (what % were 漢語、外来語、和語, etc.)
• Qualifications of grammar complexity, such as clause or predicate counts
• Average sentence length

Lastly, this presentation from 長岡技術科学大学's readability research center had some interesting graphs, and this is a fun tool that will do some readability analysis on Japanese text for you.