Why is が NOT a topic marker?

Like many learners, the precise intricacies of は vs が continue to elude me. I’ve seen plenty of explanations of how は is actually used to mark the topic, and how the topic is different than the subject and how direct translations change as a result of it. And while I think it would still take some time for me to puzzle out how to parse more complex usages of は, I generally get the idea.

But everything seems to gloss over が’s function in marking the subject, just trusting that “subject marker” is plenty descriptive enough for English speakers to get how to use it, and instead focusing on when は should be used instead of が. The problem is that が is used in all kinds of ways which don’t map to the English concept of a subject, and sometimes it feels like が is closer to marking the topic of a sentence than it is to marking the subject. For example:

“A cat is wanted [by me]” (:nauseated_face:)
“As for cats, [I] want [one]” (:smiley:)

So then, why is が not just a different kind of topic marker, and why is it instead called the subject marker?

The word subject here is standing in for the Japanese word 主語. The fact that “subject” may have different meanings in English is sort of beside the point. What matters is that が marks the 主語.

は marks the 主題, which also has an accepted meaning within the context of Japanese.

I think this is how it needs to be approached. Yes, “topic” and “subject” are words that have a variety of meanings in English, and English grammar addresses things in its own way, but when we are talking about Japanese, we should think of it as topic主題 and subject主語.

And in that sense, it’s kind of strange to suggest that they would be the same or interchangeable terms.

Let me know if I’m misunderstanding what you’re trying to say or if what I said isn’t clear.


I’m not suggesting that they would be the same (they’re clearly not), but presumably the words “subject” and “topic” were chosen for が and は because they resemble what those words mean in English grammar. Of course the words “topic” and “subject” are arbitrary and can have any definition they need to in the context of Japanese, but the names we use for the particles suggest things about how the particles should be used.

What I’m wondering is why が is called a “subject marker” rather than an alternative type of topic marker (e.g. why は and が aren’t called “type 1 topic marker” and “type 2 topic marker”). Or put another way, what about が makes it more “subject-y” than “topic-y”?


I think it’s just because 主語 is the grammatical subject. We already have a term for that in English, so nothing creative is necessary there.

In 猫がほしい, 猫 is unquestionably the grammatical subject of the sentence. What the grammatical subject of a naturally translated sentence in another language would be just doesn’t really come into play.


What about this example from the Tofugu article on が:


物価が is certainly marking the subject in both cases, but I find it hard to accept that スイスが is also marking the subject in the latter. Unless there’s some way of parsing out the sentence where it actually is?

1 Like

I’m no expert in 国語文法, but to me it doesn’t really change anything if this is analyzed as two 主語 or a 主語 and a 述語 that contains a 主語, or something like that, or even something else entirely, with regard to how we decide to call 主語 and 主題 in English.

I suppose since the English words don’t really matter from my perspective except as placeholders for when we have to use English to discuss things, maybe I’m just the wrong person to discuss it with.


I think if you treat 物価が高い as a single word, an adjective, in the same way 背が高い means “tall” then “Switzerland” can be a subject for the full sentence in way that’s sensible in English.


That’s an interesting way to think about it. I think it will take some time to persuade my brain to actually group the fragments together like that, but that’s quite helpful.

In this case, while I get that the first translation isn’t at all natural in English, it’s really just a matter of expecting a translation to point us in the right direction when it simply doesn’t. Why? Because some things don’t translate, and in the case of English and Japanese, the two languages have such hugely different expressive strategies that it’s not at all uncommon to have to reword and interpret a sentence in order to translate it naturally (i.e. sometimes, you can’t even just stop at matching phrases with equivalents in English; you have to decide on the overall message yourself and write the translation almost as an explanation).

Here, grammatically speaking, 欲しい is an adjective in Japanese. You can’t change that, even if you can translate it using a verb form in English. Like Leebo said, 猫 is the grammatical subject of 欲しい in the Japanese sentence, even if that’s not very natural in English. (Just another example: 好き is an adjective in Japanese; ‘to like’ is a verb in English. The two languages are just structurally very different.)

To be fair, however, Japanese dictionaries don’t list ‘subject particle’ as が’s only function. You’ve also got stuff like ‘marking the target of hope, ability, liking, disliking and so on’.

This is about the nuance conveyed by が, which is an exclusive one. The first sentence is just a comment about Switzerland. The second emphasises the fact that Switzerland (and not any other place) has a high cost of living. If you think about it, even in English, something being the subject of a particular verb/adjective does have similar consequences: ‘John is using the dryer’ likely implies that no one else is using it – or no one else can, at any rate – even if we must admit that this inference is also partly due to the use of ‘the’ before ‘dryer’, which implies there’s only one dryer available.

This is a possible approach, yes, but given how Japanese tends to work, with sentences being parsed as you read them, with modifications to the functions of phrases being made as particles appear (e.g.「これはすごい」is just ‘this is amazing’, but「これはすごいと言った人」is ‘the person who said this is amazing’), well, you could also treat one of those subjects as a ‘sub-subject’. Honestly, I haven’t seen many double-が sentences, so I don’t really think this is something to worry about, but you really can just take it that the statement「物価が高い」applies to「スイス」, and that this is logical because スイス comes first in the sentence, and hence likely influences what comes immediately after it, particularly since it can’t be the subject of anything else. In other words, the sentence ultimately looks something like this:


I don’t really know how else to explain it besides saying that you will get a feel for this if you keep observing how the two particles are used in Japanese, and in fact, more often than not, if you insist that it’s a subject particle and then attempt to create an explanation based on that axiom, you’ll find that the end result is a better fit for Japanese grammar than if you had gone with, say, a ‘topic particle’ explanation. There are other things you’ll see, like how most particles have a shorter range of influence within a sentence than は and も, or how most case particles (i.e. が、を、で、に and other particles marking a grammatical function) tend to play their role under the influence of the nearest verb/adjective, but that’s not immediately relevant to this discussion.

Maybe a final thing to think about is this: a very common and simplistic – but nonetheless helpful! – explanation given by Japanese teachers to beginners is that が places emphasis on what comes before it, whereas は places emphasis on what comes after. In that light, the Switzerland example feels more like this:

For further reading, here’s something I translated that based on a book from a Japanese grammarian discussing the five main differences between は and が. (My translation was quite literal, so I’m sorry if it doesn’t read very well…):

PS: I’ve just realised that I should have translated 主格 as ‘[the word in] the nominative case’ for difference (3), and not as ‘the subject’. Oops. Now I see why the authors chose to use 主格 (nominative case) there instead of 主語 (subject). Oh well… sorry about that.


You’re trying to cram Japanese into English grammar.

And that makes sense. English grammar is what you’re used to. Your brain has this idea of how a language is supposed to work - which is English (or some grammatically similar language - I have no idea where you’re from, but let’s go with English) - and Japanese is confusing because it doesn’t quite work that way.

Here’s the thing about translations: they’re not a 1 on 1 mapping of “the same words/concepts but in a different language”. Translations are the closest approximation of the same ideas expressed using whatever the other language lets you do. Trying to get the grammar to line up neatly when doing that is doomed to fail eventually, and here you see one example. If you try to line up the meanings and grammar neatly and cram Japanese into English, what you get is indeed “a cat is wanted”, because that’s basically what the Japanese word 欲しい means, in general terms. But as you also point out, that’s a very weird sentence.

The reason for that is fairly simple, even if it’s confusing for a good while: Japanese is not English. It’s not even anything like English. Some concepts map nicely. Plenty don’t map as nicely.

Japanese is a lot less confusing when you stop trying to twist it into English. But that is unfortunately a lot easier said than done.


This seems straightforward to me: because in simple sentences like 太郎くんがりんごを食べる the が is marking the subject of the verb in a similar way that を is marking the object. Yes, there are some sentence patterns where what が is doing doesn’t seem to line up with what in English looks like an object, but overall if you want a one-word description of what が is marking then “subject” looks pretty good. If you’re into linguistics you can call it the nominative particle, which is a fancy way of saying the same thing.

One thing that makes が not very topic-y, incidentally, is that unlike は you can find it in relative clauses, which are not places which set topics.

When you do start to look at the not-so-neat cases like 猫が欲しい you have basically two choices[*]. You can say “this is a subject because it is が-marked”, and correspondingly broaden your definition of “subject”. Or you can say “が does more things than merely mark subjects”, and broaden your list of what が does. But even in the latter case it’s still reasonable to say it’s a ‘subject marker’ because that’s short, descriptive and covers the majority of what it’s doing.

[*] There are multiple schools of thought on how to explain Japanese grammar, including ones derived from a Japanese grammarian tradition, ones from Western linguistics, and ones from Japanese-as-second-language instruction. None of these are necessarily any better or worse than the others, but they can differ on definitions of terms, which is a bit confusing if you switch between them.


Right, I understand all that. I think that sample sentence was probably not a great choice because as you and others have mentioned, ほしい is an adjective. I wanted to provide an example sentence to explain where が is confusing, but to be honest this question is really more about how English is used to describe Japanese (and what insights can be gotten about Japanese from that) than it is about Japanese itself, so trying to provide that example was not particularly useful.

Interesting! So in this example (hopefully these sentences aren’t crazy, even if this isn’t the most natural phrasing):

[A: [B: かもがいる]いけが]すき (I like [A: the pond where [B: there are ducks]])
[A: かもは][B: いるいけが]すき ([A: as for ducks], I like [B: the pond where they are]—wait, weren’t we talking about ducks?)

が is just stating the subject of the かもがいる clause matter-of-fact-ly, whereas if it were acting like a topic marker (like in the は version), it would suggest that the sentence was talking about the ducks themselves (which it clearly is not). Is that a reasonable interpretation?

I would interpret the first as saying “I like the pond where there are ducks”, while the second would be “the ducks like the pond they are in”. There are technically different possible interpretations since in the first sentence the topic (は) is implied/unstated and in the second the subject (outer が) is implied/unstated, but I think the meanings I gave are most likely.

I recommend reading this blog post to better understand the differences between は and が.


I see, that makes sense.

So then in the は version, it is talking about the ducks themselves, and that’s the issue!

Your sentence 2 is somewhere between weird and ungrammatical, because the は cannot be inside the relative clause, which leaves it being just いる – and that makes no sense. It can’t be ‘I like’ because the explicit は-marked topic is there and takes precedence over “no explicit topic, probably speaker by default”. So it’s more like “ducks, …wait, what kind of pond?”.

Sentence 1 is fine. In theory かもが could be a subject clause for the main sentence, but we know it isn’t because that would leave the relative clause as the nonsensical just-いる again.

You can construct sentences where it’s genuinely an ambiguous parse whether a が-marked clause is inside or outside a relative clause, and は doesn’t have this ambiguity.


Why is that? It doesn’t look like は is in the relative clause to me. Putting aside the brackets that aren’t actually part of the sentence, wouldn’t it be essentially かもは(じぶんが)いるいけがすき. Probably better with a leading この・その・あの since it would be weird to say this about ducks in general, and maybe a みたい or whatever after the すき since we can’t know what the ducks are thinking. But the core structure seems fine to me. Am I missing something?


Yes, it’s exactly because the は isn’t and can’t be in the relative clause that it seems weird to me. I just felt like a relative clause that’s only いる is missing too much to make sense or be anything anybody would say – I think it needs an explicit subject like 自分が or 他のかもが if you’re actually trying to say “ducks like the pond where ducks are”.

This might be the equivalent of an English sentence like “The horse raced past the barn fell” which is confusing on first read but can technically be parsed, or it might be an “OK in the right context” sentence, or it might be “never OK”. You’d need to ask a native speaker. To me いる alone doesn’t work as a relative clause, but that might just be me.


I wonder…
Isn’t that the が that in old Japanese had the meaning of modern の? (Like in 我が国).

I would then parse スイスが物価が高い as:


That is, what is high is not merely the cost of things, but the cost of things of Switzerland.


I did consider that possibility, and yes, が did have such a meaning in Old Japanese, but I’m not sure if it’s still a possible interpretation in modern Japanese. It’s implied, perhaps. At the very least, context tells us that the rest of the statement must apply to Switzerland.

I’m not certain that the version with いる alone is unparsable, but it did strike me as a little strange. Adding 自分が does make the sentence feel more natural.