How I recommend learning grammar

I feel like studying grammar has never really been all that popular. Honestly, I get it. I too remember grammar from my school days. It wasn’t always super exciting. In this post, I’d like to discuss the way I personally study grammar, because it’s pretty different from traditional methods.

The key tool I use to learn grammar in any new language is what I call an example sentence database (ESD). This idea came to me as I was writing the grammar book.

What is a good example sentence?

To me, a good example sentence for a specific grammar point has the following three key properties:

1. The sentence unambiguously highlights a specific usage of that grammar point.
2. There aren’t too many different grammar points competing for attention, or they are all known.
3. The complexity of the sentence is adapted to your current level in the language.

I suggest we look at a specific example to see how these properties can be checked one by one.

First, the question of ambiguity. Many different particles in Japanese have multiple uses. For instance, を is a direct object marker in Japanese. Here’s an example of this:

りんごを食べた。— (I) ate an apple.

That is not its only use, though. For instance, it is also used to denote space through or along which one moves. Here’s an example sentence taken from Dictionary of Basic Japanese Grammar, in the article “o²”.

私は五番街を歩いた。— I walked along Fifth Avenue.

Here, notice that ‘Fifth Avenue’ cannot be the direct object of the verb 歩く, because you don’t walk something or someone, you just walk. As such, it is uniquely different and the sentence makes it obvious it is not the direct object marker を. In other words, it’s unambiguously discussing this other usage of を. Property 1 is satisfied.

Let’s keep this sentence (which DOBJG calls a key sentence for を) and discuss the other properties mentioned above.

Next is looking for all the grammar points in the sentence. What jumps out to me is は, the topic particle, and た, the -た form of verbs. These are really the only two grammar points in that sentence outside of this usage of を. It is likely that you will have seen both of these before you come across this specific usage of を. As such, the sentence also satisfies property 2.

Finally, the complexity of the sentence. It’s a clear XはYをVerb sentence, which is perfectly appropriate here. There isn’t a million subordinate clauses, a ton of other particles, really complicated 敬語 or anything else that might really trip you up. The sentence is pretty simple overall. Property 3 is satisfied.

Since all three properties are fully satisfied, I consider this to be a great example sentence. The way I study grammar is through collecting great example sentences for each new and unique grammar point I come across.

This collection is called my example sentence database (ESD), and it is unique to me. The coolest part is that it comes from my immersion, so it is a bunch of example sentences from my favorite media!

How this works in practice

This is all good and well but how do we go about actually curating a personalized ESD? Usually, one’s first encounter with grammar is through a guide or a textbook. That’s fine when you’re just starting out, because everything is new to you and it’s very hard to find good example sentences in immersion.

What I do is that I use this first grammar resource as a way to decode Japanese sentences. Maybe you just started out and you come across this sentence in your immersion:

りんご、食べたいなぁ。

It’s very likely that your guide or textbook will discuss the -たい auxiliary in Japanese. It’s also likely that it will discuss the sentence-ending particle な (or なあ). From these two, you can start decoding the sentence.

りんご、食べたいなぁ。— Ah, I want to eat an apple…

If you like it and you feel like it satisfies all three properties for either -たい or な, then you can add it to your ESD.

Usually, that is in the form of a text file of some kind. It can be very simple (just a TXT file) or it can be very complicated (fancy markdown or org-mode). Initially, I just used a channel in a Discord server and each time I wrote down the example sentence, the grammar point it was illustrating and my understanding of it.

This is already good enough, but ideally you have some sort of index going on for easy lookups.

Either way, the point is that you have these logged somewhere. Periodically, you can take a look at your ESD and just go through a few of the sentences to refresh your memory. If you want, you can even turn it into your own personalized Anki grammar deck.

In my experience, grammar sticks a lot better if it’s seen in context, and no context is better than your own immersion.

This process is not entirely unlike what most immersion learners already do: Looking up grammar points as they go. That’s fine too, but not every sentence you will come across will be good for remembering that particular grammar point. Maybe there’s a million things going on. Maybe it’s not that clear. Maybe you just don’t like that sentence.

Curating your own personal ESD makes it easier to choose your sentences, kind of like picking which words you mine during immersion.

This is how I tend to learn grammar.

Of course, I find reading grammar textbooks fun, but that’s because I’m a linguistics nerd. If that’s not you, then maybe making a simple ESD might be a good idea.

It doesn’t take a lot more time out of your immersion, and I find it pretty beneficial. It also lets you isolate when exactly you do grammar, because you can just add the example sentence in your ESD and fill out the details (what you understand from it, etc.) later.

Once you get better at the language, you can refine the ESD further, swap the example sentences for other stuff, make it look nice with pictures and everything. But the key requirement is to have the example sentences, an explanation and a way to look them up easily. If you have all that, you’re set to learn grammar through immersion.