3 ms·
An interesting assertion, but the article is both vacuous and confusing. It doesn’t link to the underlying study, combines what one tends to think of as kanji (
by vsnf 7y ago
An interesting assertion, but the article is both vacuous and confusing. It doesn’t link to the underlying study, combines what one tends to think of as kanji (single characters) with more complex juku-kanji (multiple characters) with straight up whole dictionary words. Some of those words are presented with English definitions to the side, but most aren’t. Some of the kanji are presented with pronunciations, but most aren’t. There’s no support for the thesis statement.
I have no particular reason to disbelieve the headline, it sounds reasonable enough, but this page in specific is doing nothing to persuade me of its main point.
- sova 7y agoA Japanese Logographic Frequency List (2000) Chikamatsu et. al https://researchmap.jp/YOKOYAMA_Shoichi/%E8%B3%87%E6%96%99%E5%85%AC%E9%96%8B/?action=multidatabase_action_main_filedownload&download_flag=1&upload_id=6391&metadata_id=18550 https://researchmap.jp/YOKOYAMA_Shoichi/%E8%B3%87%E6%96%99%E...
- fenomas 7y agoThanks for the link! It looks like the article author used the kanji from the research as a starting point, but then randomly added in words endings or compound characters, perhaps as their own study aid.
- wodenokoto 7y agoGreat find, digging up what appears to be the base of TFA’s list. Some critique of using the linked list: That is based on a newspaper corpus from 93, which explains why day/sun is at the top, which they won’t be for list created from Wikipedia or television subtitles, both sources that are arguably closer to “in the wild”
- fiblye 7y agoSeems like someone just pulled entries from their study list and turned it into an article. Loads of incredibly basic words aren’t there, but some random uncommon things are. And absolute loads of duplicates and variations of the same thing.