Today’s video is my longest so far, clocking in at nearly 20 minutes. Possibly also my most controversial, though honestly I don’t see how anyone can seriously try to refute what I am saying.
I think a long video was in order because I am tackling one of the really key issues in Japanese structure and one of the core misconceptions that spreads its tentacles over many areas of Japanese…
…and makes what is in fact clear and simple into an absurd tangle in the minds of most students.
I am dealing here with a core problem of Western “Japanese grammar” and showing how simple and understandable real Japanese grammar is.
It affects many areas of Japanese, but by way of example, I decided to focus on desire/emotion expressions because they throw up this problem very extensively.
And this kills two birds with one stone by allowing me to introduce these structures at the same time (and clarify them for those who already know them).
A piece of semi-trivia here is that I deliberately chose the crepes example because it is the one used by Tae Kim-sensei in his “proof” that there is no grammatical subject in Japanese. I have the utmost respect for Tai Kim-sensei. And as I have said before, he makes this assertion not because he is illogical but because he is much more logical than the usual explicator of conventional Eihongo Japanese.
If you accept the premise of Eihongo grammar then it logically leads to conclusions that eat away the entire structure, ending up – quite logically – with denying the grammatical subject altogether.
So Tae Kim-sensei’s logic is impeccable, but unfortunately a false premise can only lead to a false conclusion. The textbooks continue to talk about the existence of a grammatical subject while saying things incompatible with it, because they fudge the logic.
If you are interested in this “controversy” I wrote an article about it here.
I have great respect for Tae Kim-sensei and more rather than less because he relentlessly thought out the logic of his position where everyone else shied off and fudged it.
I had hesitated over linking to Tae Kim-sensei’s article because it really can tie one’s logic-circuits in knots about how Japanese works.
However, I think it is worth doing. Please don’t read it until you are comfortable with the logic of this video. Because if you aren’t, it will make the confusion worse than ever.
If you have absorbed the video, however, I think you will gasp and roll on the floor in paroxysms of amazement (well, perhaps not actually on the floor, but I wanted to show off that I can spell paroxysms).
Seriously, I think you will be astonished that a mind as fine as Tae Kim-sensei’s can produce quite such an all-fired mess of the whole structure of Japanese.
And I say this with no disrespect at all. It is precisely because his mind is so fine that he is worth referring to at all. This is what the inside-out illogical premises of Eihongo grammar will do to the very best.
So be glad you got the Magic Talisman from some odd-looking android!
First, I really enjoy your videos and your rigorous logic with respect to Japanese grammar. But I can’t help feeling that your train went off the rails a bit in the second part of this video. I cannot conceptualize that the unspoken が in 食べたい or 怖い is 私. First, when would you ever say 食べたい as a sentence without the thing you wanted to eat made explicit?
a: ケーキはどう?
b: 食べたい!
I think quite obviously the unspoken subject of the sentence above is ケーキ.
I’ve never heard a Japanese person say 食べたい as meaning simply “I’m hungry”. It always means I want to eat (something), even if that something is unspoken, and it might just refer to the upcoming meal.
Similarly, if you were to say, 怖い! the unspoken subject is not yourself, but the situation, or the howling wind, or a random clown. (“I’m scary”? No, it means “It’s scary”).
But apart from that, I find your approach to analyzing Japanese grammar helpful and refreshing. I’m only commenting because you’re usually so consistent in your logic, but I feel like you had a lapse this one time.
Thank you so much for commenting. And thank you for keeping a close watch on my logic. I like to see people following the arguments closely and it’s good to let me know if I make a slip.
In this case though I can say pretty definitely that I haven’t.
You write:
You are of course correct about the interchange you put forward. But in fact people do say 食べたい meaning “I want to eat”. Have you ever seen the アンパンマン anime? In this very popular show which is very much centered on eating and nourishment, characters very often say:
お腹が空いた!早く食べたい!
So much so that it could almost be considered a catchphrase of the show. In some cases it is true that there is a particular food they have in mind, but very often not. A very common scenario is that a character or group of characters are stranded somewhere and fainting from hunger, and they will use this phrase. They have no particular food in mind and are highly surprised (if they haven’t met him before) when Anpanman pulls off a piece of his (made-of-anpan) head and feeds them.
Just as in Japanese “I want to eat” is less common than “I’m hungry” (or in Japanese “tummy became empty”) but it is possible to say “I want to eat” in both languages and sometimes it is said.
But more importantly than this, with intransitive verbs in general the implicit が-marked subject has to be oneself. Take:
東京に行きたい
or
走りたい
In these cases what would you suggest the subject to be if not oneself?
And of course the same pattern is found in the potential. In:
本が読める
The が-marked subject is, of course, the book. But when the same word is used intransitively:
読める
There is no question that the が-marked subject is whomever we are declaring to be literate (by default oneself).
Thank you again for your close attention and for taking the time to comment. I really appreciate it.
Thanks very much for your reply. Food for thought. I’ll digest it slowly.
食べない?
【冗談】
And thank you again೭੧(❛▿❛✿)੭೨
PS – on a personal note I understand the need to think this over. It was a point that puzzled me for a while too. Having realized where about the が in たい clauses I was rather stumped by 東京に行きたい at first. Then it dawned on me that this is a pattern affecting all intransitives in a variety of those cases where Eihongo grammar inverts the が. And has to do so, since the positioning of the が in the cases that Western grammar gets wrong actually depends on their tranisitivity, if you think about it. In fact it makes the error slightly more forgivable, in that it is correct in certain cases. Or at least makes the confusion more understandable (though I have a feeling it would have happened anyway).
There is more to this story, that I am still digesting myself, and since it isn’t vital for learners it isn’t a priority, but… when I say “intransitive” here I really mean it (imposing a Western term which I think is relevant and necessary here at least until we find something better).
Transitivity and intransitivity are not in fact fully aligned with 他動詞 and 自動詞, which makes translating them as transitive and intransitive deceptive. But here we are speaking of actual transitivity. However… I think there is something connected with the Japanese conceptualization of what Western grammar calls transitivity that would make this “shift” of the が more “transparent”.
I don’t know if I can discover what it is. I have found “missing links” before.
If I get enough money on Patreon I will definitely invest in a deerstalker.
However, as I say, I don’t think the answer will be of huge practical help to learners. The current formulation covers all the bases as far as I can see.
I’m already persuaded of the basic logic. But (like a previous commenter) I am unsure about whether the second half does justice to the logic.
Why not translate 〜たい as “calls-for-eating” and say that the logical subject is sometimes “the situation” or “the occasion”? That seems more straightforward. Isn’t it also more intuitive to suggest 怖い as an elliptical sentence means “[This situation] is-scary,” parallel to 寒い (“[This situation] is-cold”) and 危ない (“[This situation]” is-dangerous”)?
Then again, I’ve never heard a person say “私が食べたい” or “私が怖い” 。。。While the first might mean “I [subject] call-for-eating [predicate]” the second just seems weird. Does it make sense to say 私が怖い (apart from needing emphasis to answer a question like, “Is there anyone here who is scared?”).
Excellent question. It is really good to think around these subjects.
Why not translate たい as “calls-for-eating”?
Essentially because I think it does violence to the meaning. たい does in fact express personal desire.
Even though in transitive sentences it is applied to what in English would be the “object” of desire, desire is not absent. In fact it is the salient point. The usual English version is not wrong about that. Its error lies in trying to bring the way in which desire is expressed in Japanese into line with the way it is expressed in English.
Ironically, saying that ○○たい means “calls for ○○” is bringing Japanese into line with English from the other side – that is to say, throwing away the other half of the equation. The assumption here is that since it is applied to the “object” of desire it can’t (at least essentially) be about desire. It must be about simple “appropriateness” or “calling for”. But it is about desire.
Apart from the fact that we know this from experience of how the word is used, we also know it from the use of たがる. If たい simply meant “calls for” rather than expressing subjective desire, there would be no objection to using it with other people.
With 怖い and 食べたい I agree that the full 私が isn’t used. They are usually somewhat exclamatory, but still, I believe, full sentences.
In
早く食べたい
(unless we are prepared to sacrifice the primary desire-meaning of たい) I think we have to see the subject as “I”.
I have certainly heard あたし怖い and I interpret it to mean that the subject is あたし as it clearly is in
東京に行きたい
(unless, again, we are prepared to throw the personal desire meaning of たい completely away) However it could be read as short for
あたし (は状態が) 怖い
In general though I would say that there are too many cases where the emotion is clearly predicated of the speaker for that to have much bearing on the overall principle.
I do agree that the changing of the “direction” of the が feels a little unsatisfactory. On the other hand Japanese, as well as being very logical, is also very adept at directing things to where context dictates.
I still have an idea that there is a “missing piece” here – and I suspect it has to do with the conceptualization of “transitivity”. However I also strongly suspect that when and if we find the “missing piece”, while it will put our hearts at ease it won’t make much practical difference to the model as a guide to understanding and usage.
I have been thinking about this discussion on and off since it happened. Both contributors, while accepting the general logic of the overall argument, are a little uncomfortable with the idea that the が particle shifts its focus according to whether the sentence is (what would in English and other European languages) be considered “transitive” or not. And I confess that this bothers me a little too. In a way we are spoiled by Japanese because of her diamond-like logic. We don’t expect her to be throwing up random irregularities as if she were a European language.
And I’m not at all sure she is. I’ve got a feeling there is something we’re missing. Not that she isn’t doing what she appears to be doing. I am sure she is. While one can try to argue around it in the たい form I don’t see how we can possibly deny that it is happening in the potential.
歩ける means “(I) can walk”. It can’t possibly be interpreted to imply “∅ does possible for (me) to walk”, while 本が読める does mean “the book does possible for ∅ to read”.
The problem is more, I think, that we need to look at it in the right way to see its actual logic, and I am not 100% certain what that way is. It won’t, I think, make much difference in practice but it will make it all feel neater.
I don’t think I’m quite there yet but let me share my unfinished thinking so far. I am pretty sure it revolves around the idea of “transitivity”. And one part – an important part, I think – of the rule is that Japanese does not like transitivity in the case of “subjective” phenomena.
Adjectives of desire are clearly subjective. The potential is also “subjective” in the sense that it refers to an individual’s ability (it isn’t referring to 可能性 – the objective possibility or impossibility of a thing). Thus the たい-form, the potential and adjectives like 嬉しい and verbs like 分かる all belong to a particular class – words that describe our subjective or personal condition, either in relation to something else (what in English would be a grammatical object) or not.
But that something else never is a grammatical object in “good” Japanese.
It is true that some people make a grammatical object of it. People do in fact say things like 本を読める but if you look at online discussion boards in Japanese you will see that a lot of people consider the usage incorrect or at least sub-standard. Some people (mainly a certain type of young person in Tokyo, I understand) even say コーヒーを好き which I don’t think even they consider to be “correct”. All of this I suspect is heavily influenced by English and by the Western ego-centered world-view.
(The practical point for the learner to realize here is that even when Japanese people do ego-centrify these expressions they always do so by changing the particles, not by misrepresenting them the way Eihongo grammar does)
So to summarize, I think that in “good” Japanese the rule is that subjectivities cannot be transitive. It doesn’t matter too much whether the subject is what in English would be the object or what in English really would be the subject. The が marks (that is to say, the grammatical subject is) the fount of the subjective condition, whether that fount is external or internal.
As I say these are my current sketches toward a solution rather than a finished thesis. What do we think?
お人形さんへ
I really appreciate your thoughts on this and trying to reconcile apparent inconsistencies. In a way, I feel like we are approaching this study of Japanese grammar the way physicists approach the physical world.
In physics, there is the assumption that ultimately everything must work according to the same basic principles. But physicists have a difficult time finding one set of laws that explain both astrophysics and sub-atomic physics. Similarly, this quest to explain why が is the subject particle always (unless it’s “but”) no matter the form of the sentence, and to insist that spoken or unspoken (“spooky action at a distance”), it is always present, makes it difficult to find one unifying theory.
I’m tempted to believe that in this case, it’s because no matter one’s ideas about Japanese being “logical” it is in fact a language shaped by humans who are imperfect and did not set about to create a perfectly logical language. Grammar tries to explain the language, but does not define it. Although Japanese certainly exhibits more regularity in its patterns than does English, ultimately I think absolute logic may be too much to expect of any human language with so many centuries of history.
I found this page interesting reading:
https://theguywholovesjapanese.wordpress.com/2016/05/31/japanese-doesnt-need-a-subject-because-the-elephants-nose-is-long/
And of course, I’m more confused than ever. Because if even Japanese linguists can’t settle this stuff among themselves…
But then, I’m getting pretty far away from effective language learning. I find these videos and discussions fascinating food for thought (早く食べたい), and at least the videos about Japanese using helper verbs rather than conjugations makes complete sense to me and I think really does help my language learning.
As to が, I’m still a bit up in the air, though I think I’m leaning a little more towards Tae Kim’s explanation, but I’m keeping an open mind. Guess I should reread Jay Rubin, and all the comments on Tae Kim’s post…
Anyway, thanks for keeping this conversation alive. Maybe there really is a theory of everything.
The analogy of grammar to physics is very apt I think. Apt in more ways than people are accustomed to imagining.
Both are attempts to describe a pre-existing reality. Neither, in my view, can lay claim to “truth”.
The very concept “truth” has no place in physics or any other empirical discipline. The concept of “truth” is an invalid import from metaphysics. (I am not arguing for or against the concept of truth here – I am simply saying that it has no relation to empirical disciplines and even if it exists, cannot be determined by them.)
The most an empirical discipline can do is produce theories that account satisfactorily for all the observed phenomena. The question of whether these theories are “true” or not is entirely beyond the remit of empiricism.
The same applies to grammar, which is essentially the same as physics in being an empirical attempt to describe a complex pre-existing set of phenomena.
We can model grammar in various ways, and the only test of the validity of a model is its consistency with all the phenomena.
That is the only test of its validity (or quasi-“truth”). But not the only test of its utility. Utility is a different question altogether.
Utility is my primary concern. Utility requires validity, but not vice versa. So my criteria for a model are
A) validity
B) utility
i.e. does it provide a helpful model for the learner that will clarify and simplify her actual use and understanding of the language?
I have read the article you link to and believe I can adequately counter all the arguments against the existence of a subject (most of them I already have).
This does not mean that my model is “true” and the other one “false”.
Provided a model can account for all the relevant phenomena it is valid. My argument for my own models is not that they are “more valid” but that they are more useful.
Specifically, more useful for the non-native learner.
Japanese academic linguists are not primarily concerned with utility for the foreign user* (neither should they be). So really there is no “argument” here. They may be “right” insofar as their own attempt to free Japanese grammar from foreign influences is concerned.
The concept of the zero pronoun, for example, is quite patently a way of making Japanese clearer to people who already have the concept of a pronoun as part of their apparatus for formulating meaning.
I don’t think Japanese people themselves have any need of (and from a “Japanese purity” perspective should very likely avoid) the concept of the zero pronoun.
However, it does clarify Japanese very effectively for foreign learners, and, unlike Eihongo grammar, does no violence to the Japanese structure. Indeed it is a wholly valid (but not exclusively valid) way of describing what is happening.
____
* To put it in plainer terms: Japanese academic linguists are constrained by no pedagogical considerations and can make their formulations of Japanese grammar as complex and convoluted as they like without risking the smallest damage to the ability of Japanese people to speak Japanese
I am now satisfied that I have the answer to this question. It is a little long to explain, but you can find it here (it has to do with the self-move and other-move concepts in Japanese)
However there is a shorter answer that I am ashamed of not mentioning earlier. I put it as a follow-up to the longer post:
This is a rider to the last post. I think the self-move/other-move explanation is important, but some people will still have the question:
But what do the words actually mean? Does こわい mean “scary” or “scared”? Does たい mean “want-inducing” or “wanting”? Does (え-row)る mean “possible” or “able”?
The answer to this is very simple. They mean both.
Now I know some people don’t feel happy with that, but the only reason for that is a really excessive English-prejudice because it isn’t even the case that this sort of thing doesn’t happen in English. It does. It just doesn’t happen in those particular cases.
For example:
I am happy
Personal state of happiness
A happy occasion
Occasion that induces happiness
I am suspicious
Personal state of suspicion
Her story is suspicious
Suspicion-inducing story
There are various other words like this in English and in most cases (including the two above) they are similarly flexible in Japanese.
The only difference is that Japanese extends the same flexibility over a wider range.
And that gives the answer without even going into self-move and other-move concepts!
Though it kind of takes the sport out of it, doesn’t it?
NOTE:
There is even a direct equivalent to こわい in English.
I am fearful that the ship will sink
personal state of fear
There is a fearful monster out there
Fear-inducing monster
When you note in the video the difference in mindset between Ego-centric English/European and animistic Japanese, that was all you needed to do for me. Everything else fell into place.
It seems to me, those who are having a hard time conceptualizing the unspoken が in 食べたい are still looking at Japanese through a Western lens. And you point this out brilliantly in the Patreon post linked above and in the follow-up short answer, also above.
I suppose the explanation will serve those who still see Japanese through Western lenses.
But I believe the idea is, first and foremost, to see Japanese through a Japanese lens, ね。
ありがとうございます。
Yes that’s absolutely right. Later I did a much more detailed video on the whole “puzzle” of why 食べたい etc. appear to “flip polarity”. This affects a large range of Japanese, so it matters to get a grasp of it. I called it “the final problem” since for many learners it remains a problem even after (and in fact because) they have learned about the logical structure of Japanese, and because it involves one last “paradigm shift”.