Read the How to use section below if you're a bit confused on how to use grammar dictionaries with Yomichan.
Youtube Demo (turn the captions on)
| Website | Yomichan dict name | Info | Deconjugation support |
|---|---|---|---|
| Nihongo no sensei | 毎日のんびり日本語教師 | Although the site was made primarily for Chinese students studying Japanese, you might still find some value in this dictionary as a filler (and for deconjugation); although you might want to download a few of shoui's chinese dicts just for this specific dict because its 意味 section is in chinese, its 解説 section is still in Japanese though |
Y |
| E de wakaru | 絵でわかる日本語 | Really good example sentences because most of the time they rephrase sentences so you can better understand them (my second most preferred grammar dict next to nihongo kyoushi) | Y |
| Nihongo Kyoshi | JLPT文法解説まとめ | Probably the best overall of the grammar dicts, although I feel that for most of its entries, it doesn't go too deep into nuances but is really great for quick lookups. I usually cross-reference the definitions and explanations with other monolingual dicts if I feel this dict doesn't cover it that much. Some of the entries might have missing examples sentences, but whatever, just click the links | Y |
| Donna Toki | どんなときどう使う 日本語表現文型辞典 | Written both in English and Japanese. Most of its Japanese definitions are written as 類義語 so they're great for quick lookups but it also has explanations and further notes at the end which might go deeper into the nuances of the grammar point | Y |
| DoJG | 日本語文法辞典(全集) | Written in English so this is great for beginners. The first dict I made. Its entries aren't complete since I made this dict from the itazuraneko anki deck (as opposed to web scraping, which i used on the other dicts). I manually removed some entries which I thought were too easy and might simply clog your yomichan search results; Incomplete readings; no external links | N |
If some of the dicts fail to import, check that you're using at least Yomichan version 22.4.4.0.
- Depends of your needs but I have all of them installed. Do note that these dictionaries are much smaller in size than full-fledged dictionaries like 大辞林, they barely affect the performance of your device if at all. In fact, I have a pathetic 6+ year old android phone with Kiwi + TTU + Yomichan with around 10 monolingual dictionaries and all of these grammar dicts and it runs just fine.
- If you're a beginner, I'd still recommend that you install at least two grammar dictionaries for parsing purposes. If you can't understand what the dictionary is saying you can always try asking through discord, language exchange apps or DeepL. (In fact I recommend that you install a lot of monolingual dicts even if you're a beginner just for parsing).
Note: No special settings or installation procedures required, simply Install the dictionaries like you would install any other Yomichan dictionary
- Go through a basic grammar guide first, even with these dicts in hand.
Unless you've had some basic grammar study and a bit of immersion experience, some of the dictionaries might simply confuse you so do that first. - I recommend going through Tae Kim first, then Cure Dolly.
- For Cure Dolly in particular I recommend that you at least go through the first 20-30 or so videos just so that when you come across a grammar point, you aren't just blindly memorizing 'rules', instead you realize that most of the time, these 'grammar points' can make sense when you look at their components.
-
Skim through a list of grammar points to get a feel for what they look like. (i.e. go to one of the sites I listed above and try hovering over those grammar points, then try hovering over the grammar points in the context of the sentence, similar to what I did on my youtube demo). Grinding a grammar deck also helps.
-
Look up the grammar points you come across in your immersion and create cards as necessary. The more you immerse and look stuff up, the more you'll start to get a feel of what structures look like a grammar point, tbh, you'd probably surprised that what looks like some random string of hiragana scribbles actually turns out to be a grammar point.
-
When you common across patterns without an adequate explanation inside the grammar dicts/monolingual dicts, just right click whatever it is that yomichan parsed and do a google search (something like
<grammar_pt> 文法解説). I use a plugin called Context Menu Search to make such searches easier. -
Keep in mind that the main reason that I made these dictionaries is because I dislike the 10 seconds it takes to google a grammar point, which you aren't even sure is a grammar point. Also when googling
<grammar_pt> 文法解説, they almost always point to the same 2-4 sites anyway, so why not turn those sites into a yomichan dictionary. With that said, do not expect to find a lot of info about grammar inside these dictionaries. Most immersion learners would probably agree anyway that these grammar points aren't even grammar most of the time, instead, they're usually common phrases, expression strategies or even vocabulary. -
You might want to check this out if you want to keep dictionary definitions concise by using yomichan's collapsed dictionaries feature + some CSS.
-
Do note however that these still do not account for every conjugation possible because most of these sites list the conjugations relative to
Ichidan Verbs (v1)andsuru-verbs (vs). For example: https://nihongonosensei.net/?p=18790(～されるまま／されるがままに), the yomichan entries that can be automatically made from these areされるままandされるがままに.
However, take a look the example sentences. You'll see 言われるがままに, unfortunately, they only listed～されるまま／されるがままにas the entries and not われるがままに, which makes sense because otherwise you'll have to list every conjugation for thegodan verbs (v5), which is a huge pain in the ass and probably why they didn't do it and also why I won't (just takes too much time for marginal benefit). IMO, immersion would bridge this gap anyway. (but if someone can think of a solution to this, feel free to edit and reupload any of the dicts). -
For grammar points structured as
なになに～なになに, e.g. 〜か〜ないかのうちに, it's usually best to break them apart into their own entry (e.g. one entry for か another for ないかのうちに). It's obvious though that か is just too vague and common to be an entry so i only madeないかのうちにas an entry. Feel free to send a PR though if you think I've made a mistake with some of the entries (probably did), but do note that having too many vague entries can make yomichan hard to navigate (e.g.なになに～なに, scanning for ない would give you a LOT of entries, most of which you probably already know anyway). -
These dictionaries aren't perfect and yomichan's deconjugation isn't perfect either (particularly for very colloquial expressions), you could try deconjugating(or converting a godan conjugation to an ichidan/suru-verb conjugation) something that looks like a grammar point before looking it up on yomichan (it works sometimes). If you're a bit more advanced, you could try adding regex replacements rules inside yomichan or even editing the deinflection json file.
-
In case some people might want to scrape other sites to create another dict, I also uploaded the scripts I used just to give you guys an idea. It's a really half-assed script though and I bet you guys can come up with a better one.
-
Btw, when I first tried to scrape Nihongo Sensei and Nihongo Kyoushi, at first I only intended to create a dict containing only the grammar points for yomichan parsing and its only content would be the link, but then I went further for some reason. If you're a bit short on time, just keep in mind that the content isn't even the most important part of grammar dictionaries, it's the parsing part (because tbh, you'd be surprised what things count as a grammar point) that's most important. After yomichan has parsed whatever shit that is, you could simply right-click and google the grammar point, or if you have the link then click the link. That'd be the most bare-bones but still very powerful grammar dict.
Remember to share them though, even if they're just barebones :D
- https://jn1et.com/jlpt/
- https://nihon5-bunka.net/japanese-grammars/
- https://japanese-bank.com/jlpt-grammar-all/
- The 4th entry of a dictionary entry represents its
part of speech(p.o.s.)(e.g. vs, vk, v1, v5, adj-i, empty=noun). It's responsible for yomichan's deconjugation stuff.
- (Send me a message first if you plan to edit any of the dicts just to ensure that we aren't doing something redundant)
- Complete the part of speech info (classify according to these:
vs, vk, v1, v5, vz, adj-i), afaik, only these 6 part of speech markers work with yomichan (according to the yomichan deinflection rules or the schema), if it's a noun or already conjugated, leave it blank - There are some entries that would be worth duplicating and deconjugating. For example, let's say there's a grammar point called
をちゅうしんにする, if you add its p.o.s., i.e.vs(suru-verb), yomichan would catchをちゅうしんにして, をちゅうしんにしました, etc.
However, for example the grammar point ないこともない, if you add its p.o.s.,adj-i, it'd catchないこともなくて, ないこともなく, etc.but it won't catchないこともありません, in order for yomichan to catch that, you'll have to create a fake entry, ないこともある. - You might want to check this and the other grammar dicts for examples on how to format entries.
- Send a PR I guess
For all dictionaries: I used web scraping for all of the dicts except DOJG so there might be a few missing entries and parsing mistakes. For some of the entries, I used a script to generate the reading when the reading info isn't available, so look out for reading mistakes.
Send me a message or open an issue if you find any corrections, missing entries, etc. Might take a while to update though cause I'm lazy.
Incomplete readings(there might be some entries where the dictionary entry is also the reading), incomplete entries (particularly for 基本), no deconjugation/p.o.s. info
Some entries have the example sentences cut off, some entries might have ads at the end; also, some entries have wildly differing HTML structures (e.g. some use tables, some have more than one 例文 sections), etc., might have a few missing entries from N4, N5 that i didn't bother to add
Might have a few entries missing, incomplete kanjification (all of donna toki's entries are in hiragana, I converted some of them into kanji but not all)
Might have a few entries missing?, a few reading mistakes maybe, a few ads at the end of some entries maybe