掲示板 Forums - FAQ: Why does this sentence have が (or は)? Shouldn't it be the other one?

This is one of the most commonly asked questions by learners on renshuu when looking at example sentences.

The reality is that because Japanese is such a contextual language, why は or が is used is often unclear when looking at a single sentence, and not a sentence within a larger dialogue. Because of this, the answer is usually "it could be either, depending on what came before it."

Let's take a look at a simple example:

カレーはおいしいよ。(Curry is delicious.)
Now, this sentence could use は, but it could also use が. It's hard to tell when seeing it by itself. Let's look at two dialogues:

A: Bさん、いつもカレーを食(た)べていますね

B: カレーはおいしいよ。

A: B, you're always eating curry.

B: (Because) Curry is good.

In this setup, B is using the は to take the conversation and switch the topic from themself over to curry, and the conversation would presumably continue with talk about curry (the topic).

A: このレストランは、何(なに)がおいしい？

B: カレーがおいしいよ。

A: What's good (delicious) at this restaurant)?

B: (The) curry is good.

Here, a previous topic has already been established (the restaurant), and A is asking about what might be worth having. が is often used to state information that was (directly or otherwise) requested by the other person in the dialogue. This is why you see it come up so often in answers to questions.

This example is not meant to cover all or even most situations, but hopefully it helps show how there is often not a "correct" particle choice between は and が when viewing a sentence in isolation.

I'd like to note that while this "zero-pronoun" model as mentioned by Anonymous123 was popularized by Cure Dolly for many learners, it originally came from Jay Rubin's Making Sense of Japanese (who was likely simplifying the research of the 70s-90s for learners). What's important is that this comes from a model linguists came up with to explain subjects and objects missing in the sentence; in reality whether or not speakers think this way at all is debatable. Many learners also seem to think that it only applies to が, when actually it also applies to other particles like を.

Personally, I think the model works great for linguistic analysis but is a crutch for learners, especially those from languages like English that want to have a subject in everything. It's ongoing research and debate in Japanese linguistics on what "subject" actually refers to, while the topic is a lot easier to identify (Handbook of Japanese Syntax, Shibatani et al., 2017, chapter 3). Instead, most prefer to say that が marks the nominative, which then depending on the model may have plenty of functions like marking the subject, but not always and not exclusively by (this Japanese StackExchange answer goes over some of it: What exactly is the subject?).

I think that instead learners should just learn the functions of は and が by themselves and only leave this model (and other models) in the back of the head. In fact, learning how they function themselves will be a lot more helpful when one gets to other ways of marking things that seem topic/subject-like such as に(は), って, について, or even no particle at all, and suddenly have to grapple with how they differ from は and が. Of course, if Cure Dolly's way of explaining it helps you make it click anyways then all the better!

My preferred resource for these two particles (at least one that isn't completely academic) on this is Imabi's:

The Subject Marker が

The Particle は I: The Topic/Contrast Marker

The Object Marker が

Note that they do use the zero-pronoun model, but it's in a way that's a lot more rigorous, as a mean of analysis instead of as a way to translate to English. They also say that が can mark objects which disagrees with Jay Rubin's model and other models too. If you prefer that が always marks a subject, this may be more preferable: Double Subject Constructions.

It works the same in that model, カレーはおいしい would be analyzed as カレーは(∅が)おいしい where ∅ = カレー. Jay Rubin's explanation of this is that は only marks the topic and nothing else, so we have to suppose that this invisible が exists to keep the idea that "every sentence has a subject" and "the subject is always marked by and only by が".

These two statements are pretty debated to say the least. Some would say that Japanese has no concept of subject at all (e.g. Mikami), some would say that it exists and only が marks the subject (e.g. Jay Rubin), others would say that but it can be marked by more than just が (e.g. Shibatani). It gets even more complicated when we talk about sentences that contains both a は and が, or one that contains two がs, etc. and there's even more disagreements.

For subjects, I personally prefer the third of those for myself (and for whether が marks objects, generally I say yes, but I can be persuaded on this one lol), but with the caveat that the subject is a difficult thing in Japanese, and instead we should focus more on it's topic-comment structure. Something は-marked is the topic, the rest is a comment on the topic, and sometimes the subject is obvious and sometimes it's not.

For any beginners reading this, none of this actually matters! Pick any tutorial or guide that works for you and go with it! Theory helps in the short run, but in the long run the main thing that matters is reading, listening, getting input so that your brain simply just intuits the right thing from what it has seen and heard before. And to that end, check out the book club on Discord!

A correction: nominalization refers to turning verbs, adjectives, etc. into nouns. が marks the nominative case. Many Indo-European languages have this concept of case too. No other particles marks the nominative other than が; for example を is the accusative, に is the dative, etc. は is not considered a case-marking particle but an adverbial one.

And for your second point, this depends on the model. Depending on the model, nominative-marked phrases could be: always subjects, subject or object, or something else entirely. And again, depending on the model, non-nominative-marked phrases could be: never the subject or sometimes the subject.

(Also before this gets out of hand, I wanna say that I'm only an amateur in linguistics, don't take my word as gospel, and please correct any errors I might make )

Thanks, that was sloppy of me. Glad you understood what I meant.

I don’t really speak Japanese or GB theory well enough to argue the other point, so let’s just note that it can get complicated.

I read somewhere (I forgot where) that an easy way to choose between は and が for our sentence is to determine which part of the sentence we want to highlight.

• Use は if the main message comes after は. For example:

私は寿司を食べた。 (I ate sushi.)

Here, we want to emphasize that the act of eating sushi is more important than who did it. This is why the part with は can often be omitted as it is less important. --> 寿司を食べた。

• Use が if the main message comes before が. For example:

私が寿司を食べた。 (I ate sushi.)

This is the opposite of the previous example with は. In the sentence with が, we want the listener to focus on the doer (私, I) rather than the action (eating sushi). In other words, the act of eating sushi is less important than who did it.

Here are some more examples:

猫は木に登っています。 (The cat is climbing up the tree.) — The act of climbing up the tree is the main information.

猫が木に登っています。 (The cat is climbing up the tree.) — The cat itself is the main information.

漢字は難しいです。（Kanji is difficult.) — The difficulty of kanji is the primary focus.

漢字が難しいです。（Kanji is difficult.) — The subject "kanji" itself is being highlighted, focusing on it being the thing that is difficult.

Of course, this method doesn’t work for all contexts and situations, as there are other uses of は such as for contrast, or が that is used with certain words/phrases like 好き、 できる、　分かる、 etc. But it has helped me a lot in deciding which particle to use, and hopefully, it will also be helpful for anyone who read this post as well.