Jarvik7 Wrote:The paper I linked actually mentioned importance on the first page as an older explanation, but said it was inadequate to fully describe ha/ga.

Just took a look at the first page of the linked article. I bet the author of the paper isn't good at Japanese. Or maybe his or her Japanese grammar is either archaic or too strict.

The first page says 誰かは来ましたか？ and 誰かは来ました。are ungrammatical, but they do sound natural. Perfectly idiomatic. 誰かは来ましたか？ is often used when you think probably no one has come. 誰かは来ました。 is often used when you mean "Sure. We had a visitor. He's not the person you'd expect, though."

A: 俺がいない間誰かきた？
B: うん。確かに誰かは来たよ。(<- Is this ungrammatical? What's the proper wording then?)
A: どーゆー意味だよそれ。誰が来たんだよ！

A: 誰かは来ましたか？ (<- This is ungrammatical??)
B: いや、やっぱり誰もこないねぇ。

If the author has never heard these two sentences or heard them but thinks they're wrong, then his or her Japanese is very poor. It doesn't seem he or she understands subtle differences.

Also, the author seems to try to understand the emphasis/importance in translated sentences. That's wrong. You have to follow your heart, not English translations. He or she says "Why are [they] ungrammatical?... ...I shall attempt to answer all these questions." No thanks. It's just the author doesn't know what he or she is talking about. Or possibly it was like that in 100 years ago.

http://en.wikipedia.org/wiki/Susumu_Kuno

I think his credentials are sound The paper is an excerpt from the first mentioned book.

Jarvik7 Wrote:http://en.wikipedia.org/wiki/Susumu_Kuno

I think his credentials are sound

He'll have trouble understanding subtle differences in Aijin's Japanese. Maybe it was like that when he was young. Maybe he didn't bother to be "updated" in his research filed after moving to US. If his CV on Wikipedia is bona fide, probably my great-grandfather spoke like him. At least the explanation in the article doesn't represent the current usage of は/が. If 誰かは来ましたか is ungrammatical, then the grammar doesn't describe the contemporary Japanese language at all.

The first page says "誰は来ましたか". You seem to have subconsciously inserted the か to make it grammatical.

The second example I'm assuming means ungrammatical in the same context/meaning as the 'a' example (simply "someone came"). It also of course may be new irregular language usage. That simply means that one must add a proviso to that theory, not replace it outright (unless a more universal one is found). In fact the emphasis theory of wa/ga is even older and covers fewer cases!

I'm not gonna tell you what sounds natural or not of course, not being a native I encourage you to read past the foreword though.

Assuming he's still alive, I have no doubt he would find youth-language unfamiliar. He wouldn't say it's ungrammatical though. Linguists know better than to be prescriptive.

Jarvik7 Wrote:The first page says "誰は来ましたか" you subconsciously inserted the ka to make it grammatical.

The second example I'm assuming means ungrammatical in the same context/meaning as the 'a' example ("someone came").

I'm not gonna tell you what sounds natural or not of course, not being a native

Assuming he's still alive, I have no doubt he would find youth-language unfamiliar. He wouldn't say it's ungrammatical though. Linguists know better than to be prescriptive.

If inserting か is cheating, then his second example "誰かが来ました" is also invalid because 誰が来ました。 doesn't sound natural unless it's a question. Neither 誰が来ました。nor 誰は来ました is grammatical. 誰かが来ました。and 誰かは来ました。 are both grammatical and carry mutually different senses, though their translation could be the same sentence "Someone came."

But I didn't read the entire article, so I could be wrong. I'll read it later.

Jarvik7 Wrote:The paper is an excerpt from the first mentioned book.

1973?

magamo Wrote:If inserting か is cheating, then his second example "誰かが来ました" is also invalid because 誰が来ました。 doesn't sound natural unless it's a question.

What I meant is that it doesn't say "誰かは来ました" on the page in the first example, it says "誰は来ました". You subconsciously inserted it when you read it. Are you saying that "誰は来ました" is grammatical as well?

Jarvik7 Wrote:

magamo Wrote:If inserting か is cheating, then his second example "誰かが来ました" is also invalid because 誰が来ました。 doesn't sound natural unless it's a question.

What I meant is that it doesn't say "誰かは来ました" on the page in the first example, it says "誰は来ました". You subconsciously inserted it when you read it. Are you saying that "誰は来ました" is grammatical as well?

誰が来ましたか？
誰かが来ましたか？
誰かは来ましたか？
誰は来ましたか？

All these sentences have different senses. The last one is strange because it sounds like the speaker is asking 誰? and then proceed 来た？ while ignoring the first question, which doesn't make sense. I don't know what the definitional of "ungrammatical" is, but the last one is illogical. The only ways to make it "logical" is replacing は with が so that it doesn't sound like ignoring the 誰？ question or inserting か so that the speaker isn't asking 誰？

If 誰 was あれ (it), then all the four sentences are ok because none of them is illogical or ungrammatical. Using 誰 in the last one just makes it illogical.

Essentially you just said the last one was ungrammatical.

As for the 誰かは来ました, I am unsure if it is an issue stemming from new language usage, or if he meant it was ungrammatical in a context specific sense. In any case simple emphasis cannot explain everything else on that page. It is unfortunate that there is no public-access corpus for Japanese, otherwise one could simply search for the oldest record of "誰かは" to see if it was around when the book was written.

I have another linguistic paper (1981, Senko Maynard) which shows other linguists research into は／が via discourse analysis (new/old). The scan is rather hard to read though since the book I took it from was about 2 ft thick and some got lost in the page gutter. In any case, emphasis is an old-style theory. It simply takes the context of the sentence it was uttered in into account. Discourse analysis takes the entire conversation, the actions/environment of the speakers at the time of speech, and general knowledge into account. Traditional vs modern linguistics.

@Thora: Linguistics syntax research doesn't really move at the speed of light. Thankfully syntax doesn't mutate at that speed either Almost all mutation in post-reform Japanese is lexicon based it seems. (Plus weakening of dialects and the slow death of polite & gendered speech, but thats kind of unrelated)

Like I said, If 誰 was あれ (it), then all the four sentences are ok because none of them is illogical or ungrammatical. Using 誰 in the last one just makes it illogical to the extent that a speaker in his right mind would never say it. I just can't come up with a realistic situation in which a person asks "Who?" and then say "Came?" before getting the answer to the first question. No one would say "Came?" when he doesn't know who he's talking about.

I don't think just because a sentence is illogical means it's ungrammatical.

あれ isn't a question word, which was the point of the example I believe.

There is no sentence in which a question word followed directly by は is correct (aka makes sense), and so it is ungrammatical.

It seems the way you parsed it was to make two sentences. 誰？ (は, ignored) 来ました？

Something can also be called ungrammatical if you posit a rule (ex: が emphasizes what is to the left of it, は emphasizes to the right), and then you create a sentence where what that rule predicts isn't what most natives would understand the sentence as. The sentence itself can be natural, but it just doesn't fit the rule that was posited. It is ungrammatical in the hypothetical grammar that was created by that rule.

Jarvik7 Wrote:あれ isn't a question word, which was the point of the example I believe.

I guess this example the author claims to be ungrammatical illustrates what I'm talking about:

大勢の人はパーティーに来ました。("ni" is misspelled in the article, by the way.)

He says this is ungrammatical. But to me it seems it's just when you say 大勢, you're most likely focusing on the large amount of people. The following is a rare case where the allegedly ungrammatical sentence makes sense:

A: 俺パーティー行かなったんだよね…
B: 大勢の人はパーティーに来ました。あなたは行かなかったんですね。

I don't know if this grammar was irregular when the author wrote the article, but probably lots of young Japanese wouldn't think B is ungrammatical, though 多くの人は or similar wording like ほとんどの人は would be more idiomatic. In fact 少数の人はパーティーに来ました。totally makes sense becasue 少数 can be either important or unimportant. The difference between 少数の人はパーティーに来ました。(A few people actually came to the party.) and 少数の人がパーティーに来ました。(Few people came to the party...) is whether "少数" is focused on, I think.

I think your "makes sense" detection is relying on there being more context than there is. Keep in mind that these are simple artificial examples and thus they have no context other than what is written on the page. Think of it as if someone just walked up to you and said the sentence. You can't say "but it makes sense in this other context" or "it makes sense if you replace だれ with あれ", because that won't explain why it doesn't make sense in the example as given.

Quote:B: 大勢の人はパーティーに来ました。あなたは行かなかったんですね。

This is contrasting usage (many people came, but you didn't), which he also talks about. I think you should read the whole article first. This is neither emphasis nor new/old.

Jarvik7 Wrote:I think your "makes sense" detection is relying on there being more context than there is. Keep in mind that these are simple artificial examples and thus they have no context other than what is written on the page. Think of it as if someone just walked up to you and said the sentence.

Quote:B: 大勢の人はパーティーに来ました。あなたは行かなかったんですね。

This is contrasting usage (many people came, but you didn't), which he also talks about. I think you should read the whole article first.

I see. Actually I found this. It's a very brief summary of Kuno's explanation about は in Japanese. I think you can read it; your knowledge of Japanese is awesome.

It basically says 大勢の人はパーティーに来ました。is ungrammatical unless context says it's ok... I know we shouldn't abuse context when talking about what is grammatical, but at the same time I don't know how you speak a language without context. When I see a sentence, I say it's not natural when I can't come up with a realistic situation in which a native speaker would say it. Of course, ridiculously rare situations shouldn't be taken into account of, but it doesn't seem the example I just gave is that ridiculous.

I guess my definition of unnatural is quite different from the author's definition of "ungrammatical". I'd say 大勢の人はパーティーに来ました。is ok because we say it and don't think unnatural.

I'll definitely read it before making further posts on this.

Thora Wrote:1973?

The Japanese summary I linked says it's from 1973's book 日本文法研究.

Despite being a humanities discipline, linguistics works like science. You form a hypothesis and then try to break it. If your hypothesis doesn't hold in all contexts, then it's incorrect (or limited to certain environments).

In the ungrammatical examples Kuno gave, they didn't work in the contexts and sentences as given, so the earlier theory (emphasis) is proven invalid, or at least not universal. And so Kuno sets out to create a new one that describes all situations.

Your definition of a word and how it's used in the paper being different isn't surprising. Words are used differently in science as compared to normal speech (such as the word theory, which in science is as close to fact as you can get).

Thanks for the JP link. I'll check it out later. I can't wait to add more obscure 専門用語 to my SRS (Did you add anaphoric to yours? It's in the English paper)

-edit-
Took a quick peek at the JP version and his descriptions are a lot more clear and less ambiguous. One can tell which language is his native one.

@Jarvik7
I'm sure you'll find this ebook interesting. It's the source of the summary I linked and is a huge book that explains the contemporary Japanese. Of course it's tough even for a native speaker to read through unless you're used to linguistics, but it cites tons of very recent stuff written in this century. I just skimmed a few pages and found a quote from a 2004 article.

I only read a few pages in 10 minutes, but it seems は/が has a lot more than Kuno's 1973 book says. An explanation very similar to mine is also mentioned in a more elegant and formal style, and neither Kuno's nor mine covers all the cases, which I did mention in the very first post.

About Kuno, it seems he was a famous figure in this area 30 odd years ago, and his theory does seem to be influential; the summary I linked was one of the supplementary explanations in the ebook in spite of the fact that his theory is quite old, which speaks volumes.

Edit:

Jarvik7 Wrote:(Did you add anaphoric to yours? It's in the English paper)

Ha ha. The English paper should have tons of obscure words! And sorry for posting this before finishing it. I just wanted to let you know the ebook.

Edit2: You can find an in-a-nutshell version of the modern explanation in 補説5.2 「は」と「が」の基本. Roughly speaking it says you use は when X in XはY is something both speaker and listener can easily picture. が is trickier.

Edit3: It also says you can use more than one は in a certain situation. I wanna punch the teacher in the face who forced us to avoid multiple はs like a plague.

I also found about a short comment on Kuno's research. It says his theory was based on an idea that was once revolutionary in linguistics especially in the study of the English language, but the attempt to describe the Japanese grammar didn't work well. It also says the current research in Kuno's style is almost exclusive in the study of difference between English and Japanese. I'm not sure if this is true, but if that's the case, it's not surprising his explanation is heavily influenced by English translation and Kuno and his followers' theories are still dominant among teachers/textbooks of the Japanese language in English speaking countries. Here's the link. Search for 生成文法による日本語文法研究 and you can find the comment.

Whoa. I'd just like to say thanks to you guys for this fascinating dialogue. You've dumped a load of smart down my knowledge hole, but now my thinky-thing hurts.

I remember being taught the standard ha=topic, ga=subject explanation at school. It's actually the simplest and best explanation for ha/ga and accurate too. I think a huge source of the confusion and ensuing over analysis of ha/ga comes from the fact that people misunderstand the term "subject". In this rule, "subject" actually refers to the agent of a verb. i.e the person/thing performing the action (or acted on in the case of passive verbs). I think I always unconsciously thought of "subject" in the sense of subject of a conversation, or subject of study, and so was never using ha/ga correctly back in school. I think if you replace the word "subject" with "agent" it all makes perfect sense.
I think very few people actually understand the concept of grammar or how to teach it, and so you see textbooks filled with japanese sentence x = english sentence y explanations, which is bullshit. Not for some airy-fairy language is ambiguous and ever-changing reason, but simply that two different languages are two different beasts. It'd be like trying to describe a square using the parameters that define a circle. I think it's a much better idea to process a foreign language by understanding it's syntax. I think the syntax of Japanese (and perhaps most languages) is actual pretty simple, and all the necessary rules could be described in one page.

In the case of understanding japanese, go from japanese sentence to actual meaning as understood by japanese people. You do this by understanding the function of each word and the syntax of the language. This may well produce some funny sounding sentences when you translate back to your native language, but I think that hardly matters. The goal is acquire japanese language by parsing it the same way japanese people do.

To get back to the ha/ga thing:
I think "ha" actually performs a similar function to a comma (,) in english:

for example:
(in romaaji because people were having display issues ??)

"fuyu ha ~." = "winter, ~." or more elegantly "in winter, ~".

"gakkou ha ~." = "School, ~." or "Regarding school, ~." or "As for school, ~."


"watashi ha hon wo yonda." = "I, read book." = "I read (a) book."
(It is assumed that "I" is the agent, but not by any grammar rules. But rather by common sense)

and "ga" simply defines the agent of the following verb:

"watashi ga hon wo yonda." = "I read a book."
(what the particle ga actually does, is state that the proceeding word "watashi" is the agent performing the following action, the verb "yonda".)

"boku ha fugu ga suki" = "I, fugu (is) like." or more elegantly "I like fugu."
(In this sentence, "fugu" is the agent of the omitted verb "desu" which basically means "is". Also you need to note that "suki/kirai" are not verbs, but rather nouns similar to the english "likes/dislikes" )

"nezumi ga raion ni taberareta." = "mouse was eaten by lion."
("nezumi" is the agent performing the action of being eaten, "taberareta".)

It's really not that complicated folks. This thing about "old info/new info" is an over complication. If you analyze a sentence by it's syntax, logic (not grammar) will tell you if a sentence is wrong or right.

for example:
right:
"fuyu ha neko ga shinda." = "Winter, cat died." = "In winter, (unspecified) cat died."

wrong:
"fuyu ga neko ga shinda" = "winter cat died"
(wrong because the verb "shinda" seemingly has 2 agents both marked by ga. Illogical because there is no way of knowing which died. Was it the cat or winter?)
Perhaps the assumed meaning would be "fuyu (desu) ga neko ga shinda."


What I'm trying to get at through this long winded post, is simply that the syntax is actually fairly simple, and if a sentence is right/wrong is a matter of logic not grammar. I believe magamo said the same thing.

For magamo:
about a/an/the.
It is used to specify wtf you're talking about. If you said "I like car.", the standard response from an english speaker would probably be "huh? what car?" because just car on it's own doesn't refer to any specific car, but rather the abstract concept of a car.
"cars" = cars in general
"a car" = one car (of the many)
"the car" = the specific car (understood from context/mutual understanding)

I wish I could chime in, but I find it very difficult to explain in logical terms the grammar of your own language. I can explain most particles and grammar to Japanese students here, but when they ask about は/が I don't know how to explain it well in English.

But to be honest the differences are not always clear cut. Language is sometimes more artistic than it is scientific, and rigid, logical explanations can't always be counted on. は／が　is a topic that a lot of people studying Japanese seem to overthink sometimes. All the different uses, nuances, exceptions, etc, are difficult to explain, and I think it's easier to let your mind simply pick up on when to use which one. This is a kanji website, so most of you already know firsthand that after a lot of exposure you start to learn which readings to use. Let's say you have experience with 下着、上着, 寝間着, 肌着, 古着, well, when you see 水着 for the first time you KNOW that it's みずぎ and not すいぎ from the pattern of using the 訓読み You don't have to think about it logically, your mind just sees the pattern unconsciously.

It's the same idea with nuances of grammar like this. You do need to understand the basic differences, yes, but don't stress yourself out about all the smaller differences. After a while your brain will pick up on which ones to use in which contexts, just like it does for people who grow up in the language.

＜offtopic＞

nadiatims Wrote:It'd be like trying to describe a square using the parameters that define a circle.

This made me giggle as I just had an exam covering fourier series.
＜/offtopic＞

By the way if anyone is interested I have no clue what you guys are saying because I went to mine more sentences. I would never have thought this topic got so far from a simple explanation.

This post is going to be incredibly long. So please ignore it if you want to. I could be totally wrong anyway.

I read through the English version of the article by Kuno published in 1973, which Jarvik7 kindly gave the link to. I also ran my eye over thorough explanations of は/が found in 現代日本語文法概説, which is based on recent results in linguistics.

Now that I have read through them, it's not surprising to me any more that Kuno's research method is found almost exclusively in the study of difference between English and Japanese these days. I feel sorry for perplexed learners who had to use textbooks that based on his view to learn は/が. Their textbooks are simply wrong. I'm under the impression that their teachers didn't know what they were talking about.

First of all, it seems the consensus among researchers today is that something wasn't quite right about Kuno's view of は/が. According to 現代日本語文法概説, you can read about how Kuno got tripped up in specialist works in academia such as 「助詞の意味と機能について−「は」と「が」を中心に−」 by 柴谷方良 written in 1995. The author is (was?) the dean of the Department of Linguistics at Rice University if you doubt the credibility, though I don't think "who" should make any difference. In 現代日本語文法概説, there is a direct quote in which 柴谷方良 is talking about Kuno's errors.

Anyway, one of the major errors Kuno made was about sentences such as 大勢の人はパーティーに来ました。, which Jarvik7 and I have discussed. This sentence is considered grammatical in recent books including 柴谷方良's work while it is ungrammatical in Kuno's view. What he was saying about the sentence was like "'You are a student' is ungrammatical in English because the literal translation of the verb 'are' in another language X is considered the plural of 'be.' You could say it's grammatical only when context says you can see it as a special singular 'are.' However, 'You are a student' is ungrammatical without context."

In the English article he first explores how native English speakers compose English sentences (He spent the whole 5 pages...) and then tries to explain the Japanese language from the viewpoint of the English language. He defines "themes," which roughly explain how a native English speaker's mind works, and then classifies は according to the English themes. And he claims a Japanese sentence using は is ungrammatical when his version of literal translation into English doesn't much up with the English theme which he thought the translation should have fallen into. But obviously his definition of "ungrammatical" doesn't resonate with native Japanese speakers' intuitions. There are lots of examples he claims to be ungrammatical but native speakers would think natural. So he makes up a special term "contrastive usage" and says it's not ungrammatical when you see it as contrastive は. Of course, this view is considered nonsense by today's researchers.

I could be completely wrong because I'm not used to reading this kind of thing in English. But one thing I am certain is that his work shouldn't be recommended to learners of contemporary Japanese. In page 42 there is an example sentence that is considered a faux pas by the standards of today. It contains a very controversial word that my IME (I'm using ATOK) forbids to write in kanji. I'm guessing choosing the sentence as an example was all right back then. Also, there are a few example sentences that don't sound natural to my ears, though it may be typos; there are a few obvious typos in his illustrative sentences.

Anyway, here are explanations of は/が posted on the forum of Tae Kim's Guide by the author of 現代日本語文法概説. He explains the particles by showing how/why/where popular explanations such as topic vs. subject and new vs. old are wrong. His explanation is based on very recent research in contemporary Japanese. Since the posts are very long and in Japanese, they wouldn't be helpful to beginners. But it might be interesting if you're advanced learners having trouble understanding は and が.

He concludes that the wrong explanations are ok as long as learners can understand proper usage. I'm not sure if he's too optimistic, but it sure doesn't seem like linguistically proper explanations help beginners. It might be better to listen to whichever explanation you think makes sense and forget about it as soon as possible.

Thanks for pointing to that article by 三郎 magamo. I think he basically is saying exactly what I did, all be it with more and better examples. It's great to be vindicated by a native speaker
My post was very long, but what I was meaning to show, is that ha/ga is actually really simple if you don't try to tack on all sorts of assumptions, and just allow the nuance to come from the context or logically thinking about what is actually being said. Ｉ　think a lot of the nuance (about new info/old info etc) becomes very obvious once you start reading a little more fluently, and can read between the lines, but you should also understand that none of this nuance is explicitly implied in the grammar (I prefer the term syntax).

Anyway to reiterate the rules:

"ha" simply defines the topic of the sentence. What it is you're talking about.

"ga" defines the agent of the following action. Nothing more, nothing less.

NOTE: I prefer the term "agent" to "subject", because the word "subject" is inherantly confusing. Interestingly, 三郎 also used the word agent in his post. I use the term "action" instead of "verb", because it is less abstract to me.

When ha/ga is used incorrectly in a sentence, it's not because it violates some grammar rule. It's because when that sentence is parsed for meaning by a native speaker (according to proper syntax), it produces an error of logic.

LTze0 Wrote:＜offtopic＞

nadiatims Wrote:It'd be like trying to describe a square using the parameters that define a circle.

This made me giggle as I just had an exam covering fourier series.
＜/offtopic＞

Oo Sounds like a fun problem to solve heh =D

magamo Wrote:.
Anyway, here are explanations of は/が posted on the forum of Tae Kim's Guide by the author of 現代日本語文法概説. He explains the particles by showing how/why/where popular explanations such as topic vs. subject and new vs. old are wrong. His explanation is based on very recent research in contemporary Japanese. Since the posts are very long and in Japanese, they wouldn't be helpful to beginners. But it might be interesting if you're advanced learners having trouble understanding は and が.

He concludes that the wrong explanations are ok as long as learners can understand proper usage. I'm not sure if he's too optimistic, but it sure doesn't seem like linguistically proper explanations help beginners. It might be better to listen to whichever explanation you think makes sense and forget about it as soon as possible.

To be honest, even natives and more advanced learners of Japanese that I talk to still have no idea when it comes to ha vs ga. I was content with the explanation ADBJG gave me which I still have to re-look up every now and then.

magamo: I think you grossly misread Kuno's article. At no point is he judging grammaticality based on the English translations. The English is just there as an aid to understanding for non-native linguists.

nadiatims Wrote:"ha" simply defines the topic of the sentence. What it is you're talking about.
"ga" defines the agent of the following action. Nothing more, nothing less.

It's not nearly that simple, otherwise linguists wouldn't have been arguing about it for the past few decades.

田中さんが林檎を食べた。
田中さんは林檎を食べた。

In both cases Tanaka is the agent of the action. Both sentences are grammatical and make sense (although they have different nuance).

Re: saburo's forum post: He seems to have no understanding of what Kuno was actually saying. Perhaps he got his information through a third party or only had access to a synopsis. He appears to be attacking 旧情報／新情報 without knowing what they mean in the context of discourse.