Practicing Reading Cursive Japanese with Miwo

Most people involved in Japanese studies with access to a smartphone or tablet will be aware of the kuzushiji (cursive character) recognition app, Miwo, which launched on August 30, 2021. It is not my intention to review Miwo in this article, but rather to explore how I have been using the app within my own studies. Indeed, perhaps one of the most promising aspects of the app is its potential use for kuzushiji education in both formal (classroom) and informal (personal study) settings.

Video showing how to use Miwo and its features.

Although it is possible to use flash card and other educational apps to supplement one’s study of kuzushiji, the key to learning to read cursive Japanese is engaging with and reading historical materials. There are several ways to go about doing this. One could pick up a copy of Kobayashi Masahiro’s Toite oboeru! Kuzushiji・Komonjo doriru 解いて覚える!くずし字・古文書ドリル or a similar text and read facsimile versions of original sources alongside transcriptions, but perhaps a more common method and one that doesn’t limit the student to a specific set of texts that have already been transcribed is to transcribe materials from one’s own collections or collections available online. I find this particularly useful since it allows me not only to practice reading historical documents, but also to prepare documents for publication. It can, however, be quite a monotonous task, and if there is a character that one can’t decipher it can be extremely difficult to work out what it is even with the help of the multiple tools available on the market.

Miwo can potentially help with both of these issues. Miwo allows users to take (or upload) photographs of pre-modern Japanese texts which the app will then automatically transcribe through optical character recognition (OCR) technology. As with all OCR software the app does not produce guaranteed results, but one can use Miwo in order to gain some ideas (or confirm ideas) about unidentifiable characters. As noted, it isn’t always accurate, but I have found that it often offers ideas that help me to think about characters in new ways. For example, it may suggest a wrong character but help me to identify certain radicals present in the character or similar looking characters which when combined with the context of the passage may lead me to a eureka moment.

Example of a transcription made by Miwo.

More relevant to the theme of this article, however, is the ability to use Miwo to add a bit more variety into one’s studies. I enjoy using Miwo to transcribe a text and then going back through and fixing the transcription. I’ve found that this helps me to think about texts in different ways. When I transcribe a text from scratch I tend to focus on individual characters, which occasionally leads me to overlook other aspects of the text. Correcting a transcription created by OCR, on the other hand, encourages me to focus on the meaning of the passage at large. This use of Miwo could easily be incorporated into the classroom as an activity that would help students not only think about the shape of individual characters, but also the overall meaning of a passage. It might form a lengthier activity in a single class and could be coupled with reflection on the relationship between digital tools and Japanese literature, or if used with short passages it could be used as a warmup activity. I recommend that this method is used on transcriptions with lots of errors, rather than transcriptions that are mostly accurate, since it is important to be actively engaging with the text rather than just passively reading. In any case, I’m not offering a particularly novel or groundbreaking idea here, but if you are looking for a little more variety in your studies or new activities in the classroom I highly recommend using Miwo as a study tool.