Hacker News new | comments | show | ask | jobs | submit login
The Logic Behind Japanese Sentence Structure (8020japanese.com)
317 points by tav 102 days ago | hide | past | web | 149 comments | favorite

Hey all, on a topic related to this: here is another way to get some feeling for the different sentence structure.

I recently finished making English subs for a 45 minute Japanese rock concert video from the 1980's.


Here I introduce a concept in subtitling whereby a subtitle template with dashed ("------") blanks appears for an entire English sentence, and the blanks convert to words and phrases as the corresponding concepts appear in the Japanese audio, in that order.

The viewer has a better idea of what is being sung at the moment it is sung, and which words are receiving the emotional emphasis in the song. Also, the revelation of meaning is delayed for the English viewer in the same way. The "kicker" phrase at the end of a verse or a meaning-altering particle (such as an entire sentence negation) isn't prematurely revealed in the translation.

Wow, that's a great way of presenting the lyrics in a different language

I've watched it a bit and it seems very nicely done

"This video contains content from UMG, who has blocked it in your country on copyright grounds." :(

Somewhat ironically, this problem hinges on an excerpt of a studio album song which plays over the ending credits for less than a minute. Not around the actual video album itself. That's what the content ID system latched on to.

That's pretty cool, I've been looking for something like this.

This article does a great job at presenting a gist of the Japanese sentence structure. Nevertheless, it makes me want to point out that it's not the whole story. If you take into account topics such as modality and conjugation, some of the information you add to a verb is placed after the verb and cannot be freely reordered.

Japanese verbs are "greater" than English verbs in the sense that you conjugate/suffixate a verb to express negation, conjunctions, conditional forms etc, making it longer and longer: https://en.wikipedia.org/wiki/Japanese_verb_conjugation

In contrast, English has a relatively simple set of inflections of verbs. Many of those Japanese verb forms and suffixated long verbs are translated into multi-word phrases. Compare:

Anata wa kyou nemuru. (You sleep today.) -- verb is in normal form ("nemuru")

Anata wa kinou nemurenakatta. (You were not able to sleep yesterday.) -- verb is in continuative form ("nemuru"→"nemu") + possibility suffix ("reru"→"re") + negation suffix ("nai"→"naka") + past suffix ("ta"→"tta")

The name for what you're talking about is "morphological typology". Languages occupy a spectrum from analytic (words stay the same) to synthetic (words change). English is usually categorized as analytic, since we only have a few ways to change words: plural -s, past tense -ed, etc., and English has been getting more analytic over time. Other European languages are more synthetic (fusional), like French, German, Spanish, etc. Japanese, Finnish, Hungarian are very synthetic (agglutinative). Chinese (all varieties) is on the opposite end of the spectrum, and it's much more analytic than English.

Interesting. Considering language as a lever and a tool for thought, how do you think the position of a language on the analytic/synthetic spectrum affects its suitability for computer interpretation?

Are there any major Japanese programming languages? It would seem to me that given our relatively primitive compilers, that 'analytic' languages offer simpler mediums for unambiguous programming, whereas 'synthetic' languages may bear greater nuance of expression when we can 'think' programs into existence in the future.

I advise extreme caution when going down this path... it's rife with established lines of pseudoscientific thought. For example, a couple centuries ago you could talk about how English just isn't complex enough to express great ideas the way Latin and Greek are.

Ambiguity is an orthogonal concept to morphological typology, however. There is no extra "nuance of expression" just because you conjugate words more, that's just nonsense.

There's one pretty famous and you may already know it.

> Ruby is a language of careful balance. Its creator, Yukihiro “Matz” Matsumoto, blended parts of his favorite languages (Perl, Smalltalk, Eiffel, Ada, and Lisp) to form a new language that balanced functional programming with imperative programming.

> He has often said that he is “trying to make Ruby natural, not simple,” in a way that mirrors life.

> Building on this, he adds: Ruby is simple in appearance, but is very complex inside, just like our human body.

These will not be considered strictly scientific articles, but appear reasonable..

[http://carlosplusplus.github.io/blog/2013/08/01/ruby-and-the...] - [https://robots.thoughtbot.com/learning-japanese-the-rubyist-...]

>Interesting. Considering language as a lever and a tool for thought, how do you think the position of a language on the analytic/synthetic spectrum affects its suitability for computer interpretation?

Interesting question. I remember reading article many years ago in BYTE magazine, someone, maybe one of the regular columnists who wrote about programming languages, speculating that if a programming language came out of the Orient, it might be different in some interesting ways from existing ones, which mainly came from the West, I suppose, though of course there could have been Oriental contributors to existing ones. For a long time after that I never heard of any such language; then came Ruby.

Well, APL is damned near polysynthetic, if that's your bag. But like Cree or Ojibwe, you won't really "speak" it fully, fluently and with nuance until adolescence.

Ruby comes from Japan

> English is usually categorized as analytic, since we only have a few ways to change words

Words can be stressed in many different ways in an English sentence, and the pattern of stresses in a sentence determines much of the meaning in the surrounding context that isn't communicated by word inflections such as -s and -ed. In programming languages, this would be equivalent to marking up identifiers in different ways (using bolding, italicizing, underlining, etc) instead of using prefix, suffix, and circumfix syntax operators and punctuation.

For Japanese, you learn a handful of rules to conjugate a verb. Then those rules always apply with no exceptions. In English, there are not set rules. Every case is special.

What is the past tense form of 'shake'? 'see'? 'walk'? 'sleep'? 'eat'? 'speak'? 'sit'? 'seek'? 'work'?

There are no rules. You basically have to memorize every word and all the possible ways it can morph.

English possessive: 's or s' or ' or s, depending. Japanese possessive: no

English plural: different for every word. Japanese plural: same as singular, or throw on a -tachi

English past tense: different for every word. Japanese past tense: -mashita for verbs, deshita for adverbs/adjectives

In English, nothing is simple. In Japanese, a multi-word phrase may have more syllables, but at least it will always be the same rule.

You're forgetting that there are only 3 (three) verb conjugations in English, of which two are almost always the same. Only a finite number of verbs have irregular conjugations, so you just learn them along with vocabulary. In Japanese, the number of possible conjugations of all irregular verbs (copula, "suru", "kuru") is probably larger than the number of English irregular verbs that are commonly used. In fact, let's count the number of conjugations of "kuru" in my ICHIRAN [1] database:

    ICHIRAN/DICT> (length (get-kana-forms 1547720))
Are English possessives considered difficult by anyone? Not sure what that demonstrates.

Plurals! Oh, that's my favorite topic that I'm working on right now. -tachi is mostly used with people, so can't be used in most context. For inanimate objects you just say the number of them. And that's where the counters come in... At which point any sane person gives up learning Japanese for good.

Past tense, isn't that the same as conjugations? Also your rules don't really work. "tanoshii" => "tanoshiideshita"? Pretty sure that's not a word. The correct past tense is "tanoshikatta [desu]".

[1] https://github.com/tshatrov/ichiran

We have counters in English too!

Tons of words are uncountable, like water, bread, and so on.

A slice of bread, a loaf of bread, a bread roll (Hey, why did that one come after the 'bread'...)

We even have lots of words that are both countable and uncountable. "I ate some tomato" and "I ate some tomatoes" has quite different meaning.

Overall I think all languages have their foibles, and trying to hold one widely used natural language up as "More regular" or "more difficult" is a pretty fruitless endeavour. Thought it is fun to talk about ;)

> A slice of bread, a loaf of bread, a bread roll (Hey, why did that one come after the 'bread'...)

This doesn't seem that unusual to me, all things considered. "Bread", as a word, is more of a substance-noun than a discrete object-noun.

Moreover, "slice" and "loaf" don't strike me as words which give meaning to the phrases "slice of bread" or "loaf of bread"- in fact, it's the other way around. For instance, "slice" is the primary noun, and "bread" is just meant to distinguish it from other "slices" (e.g. "slice of pizza").

So, when I say "Pass me two slices of pizza", I'm really saying "Pass me two 'slice-of-pizza's", rather than "Pass me 'two-slices' of pizza".

You're somehow comparing 3 verb conjugations in English vs. 186 for "kuru". Well, I hate to break it to you, but there aren't 186 conjugations for "kuru". There are, exactly, 6. 9 if you count the formal/archaic forms[1]. There may be 186 forms you can build with auxiliaries, but then, you'd have to compare to all the variants you can have in english with may, can, shall, etc.

1. https://ja.wikipedia.org/wiki/%E3%82%AB%E8%A1%8C%E5%A4%89%E6...


> Past tense, isn't that the same as conjugations? Also your rules don't really work. "tanoshii" => "tanoshiideshita"? Pretty sure that's not a word. The correct past tense is "tanoshikatta [desu]".

'tanishiideshita' is the kind of mistake you make when you're not taught that, in japanese, adjectives conjugate. Sadly, a lot of material glosses over that fact.

Similarly, most material for non-natives like to talk about the -masu form, then describe things as "-masu form without masu" (sigh).

Cumulating "knowledge" from such material, you end up with simplified rules like in GP, which work in some cases, but don't in many others.

Then when you dive more into the language, you either encounter new forms and consider them as such, and are crushed under the sheer number of forms, or have to basically start over, deconstruct what you learned and realize that, in fact, it's all much simpler and structured than what you thought, and what made it all more complex is all the learning material for beginners.

In some ways, it's like maths.

Coming back to the 186 forms for "kuru", I'm sure you only end up with that because of that same learning material "limitations". So you probably end up counting "konai", "konakatta", "konakute", "konakereba", and many other forms as forms of "kuru", when, in fact, they are one form of "kuru" with variants of "nai".

The same material will e.g. also tell you about "-kunai" for the negative form of "-i adjectives", but fail to mention that it's actually "-ku+nai", which explains why you will find forms like "-ku ha nai" or "-ku mo nai". I've never seen those explained in textbooks, but that I'm sure it's not pretty.

No, that's not counting auxiliaries (except "-masu"). The "explosion" comes from the fact that many conjugations can themselves be conjugated (e.g. "konai" can be conjugated as an i-adjective). I don't think katsuyōkei should be counted as conjugations because they're not words by themselves. Like, obviously a godan verb has 5 possible root endings but that doesn't mean it has 5 conjugations.

If we're counting auxiliary verbs, my system can recognize more than 4000 verb/adjective endings.

>Also your rules don't really work. "tanoshii" => "tanoshiideshita"?

I learned this quite early on, but I'm still a little confused for exactly when to conjugate desu instead.

For me, it helps to think of verbal ("-i") adjectives as the same thing as a verb. Or rather, that they're both just predicates about the topic.

tanoshikatta is literally a predicate stating "was-fun", and desu is just a formality afterwards to make it polite.

this differs from the other kind of adjective ("noun adjectives"), which can't conjugate themselves, so you need desu to change instead.

Of course, the even more polite form is "tanoshuu gozaimashita" (which comes from tanoshiku gozaimashita), but even then it seems to me to be the same form as tanoshikatta if you accept that the latter could've derived from "tanoshiku atta".

I'm not a linguist though, so I do not know if the above ideas are correct, but it's the way I understand Japanese verbs.

The main problem beginning Japanese learners often face is that they are taught polite form before plain form. Polite form is a natural extension of plain form, but if you start with that, it's actually quite mind bending to back track to plain form. The secret is to abandon polite form entirely until you are relatively fluent with plain form and then add polite form back in.

For example, "tanoshii" is present/future tense. "tanoshikatta" is past tense. If you want to make it polite, then you just add "desu". Super easy.

While it is grammatically incorrect, it is completely acceptable in normal conversation to do the same with the negation. "tanoshikunai" is the negation. Past tense negation is "tanoshikunakatta" (ye gods, I can't read romaji...). You can do exactly the same thing to make it polite -- just jam "desu" on the end. That's what every child will do. The wrong bit is that "tanoshikunai desu" should really be "tanoshiku arimasen".

For "na" adjectives, it works differently. "suki" is present tense. To make it polite: "suki desu". Past tense is "suki datta". To make it polite "suki deshita". Negation is "suki de wa nai" (seriously, romaji makes me cringe...). Polite negation is "suki de wa arimasen" (though you can very much get away with the mistake of saying "suki de wa nai desu" -- again, every single child speaks this way).

Past tense negation is "suki de wa nakatta". Polite is "suki de wa arimasen deshita" (but again, the easy way is "suki de wa nakatta desu").

So, why is it like this? The reason is that "i" adjectives were originally verbs that had a different set of inflections/conjugations. Very obscure piece of trivia (that most Japanese people don't even know) is that "ohayou gozaimasu" is actually one of those conjugations -- it's actually "(honourific) o hayai de gozaru" in polite form. The "i" ending mixes with "de" to produce the "ou" ending. Anyway, the point is that you have to inflect it because it is literally a verb that is modifying a noun.

"na" adjectives on the other hand are actually adjectives. They are called "na" adjectives because you have to add "na" when modifying the noun. For example, "suki na hito". The "na" is actually a contraction of "ni aru" -- because in Japanese you can only modify nouns with verb phrases.

So this is why there is a difference between the negation of "i" adjectives and "na" adjectives. "ku" is the verb combining form of the old style "i" verbs (like "te" is on modern verbs). So "tanoshikunai" is really "tanoshiku nai" -- you are combining the "tanoshi" verb with the "nai" verb. On the other hand "suki" is actually an adjective, not a verb, so you have to say "suki de wa nai" -- you can't combine them.

Past tense is exactly the same. In "tanoshikunakatta", it's really combining 2 verbs and conjugating the last one (as per the rules" -- "tanoshiku nakatta"). If you want to make it polite, the polite past tense of "nai" is "arimasen deshita" (but you can get away with "nakatta desu" in virtually every situation).

With "na" adjectives -- "suki de wa nakatta", we've conjugated the only verb. Again to make it polite you can say "suki de wa arimasen deshita" (or "suki de wa nakatta desu" if you want to sound like an uneducated bumpkin like me).

Hope this helps! Avoid polite form until you can handle plain form and it's almost all completely logical ;-)

Edit: Fix past tense in the examples of incorrect, but acceptable polite forms.

> The wrong bit is that "tanoshikunai desu" should really be "tanoshiku arimasen".

While it should technically be -ku arimasen, it's actually rarely used, and -kunai desu is more "mainstream".

> Past tense is "suki datta". To make it polite "suki deshita". Negation is "suki de wa nai" (seriously, romaji makes me cringe...). Polite negation is "suki de wa arimasen"

Trivia: all these forms are really variations of "suki de aru". "datta" comes from "de atta", "de ha nai" is really "de nai" with a "ha" for emphasis. "de ha arimasen" is really "de aru", with the "ha" for emphasis, and "aru" conjugated with the "masu" auxiliary at the negative form.

> Very obscure piece of trivia (that most Japanese people don't even know) is that "ohayou gozaimasu" is actually one of those conjugations -- it's actually "(honourific) o hayai de gozaru" in polite form. The "i" ending mixes with "de" to produce the "ou" ending.

Technically speaking, the -i and the de are not combining at all. The -u form (ウ音便) comes from the -ku form (連用形), where the k is removed. Then the preceding sound also changes (like in arigataku -> arigatou ; oishiku -> oishuu, etc.). The typical forms used in keigo are -u gozaimasu and -u zonjimasu (where there is no "de" to combine in the latter ;) )

I think the Japanese learn the ウ音便 in 国語 or 古文, so I don't think it's some obscure trivia that few people know. In fact, you can hear it in e.g. 時代劇 dramas.

Also, the form is pretty common in Kansai dialect (without gozaimasu). In fact, wikipedia claims[1] it comes from there and the gozaimasu was added in Kantou.

1. https://ja.wikipedia.org/wiki/%E9%9F%B3%E4%BE%BF#.E5.BD.A2.E...

Thank you for that. That makes a lot more sense. I admit my source for the combining was 言葉マン on NHK and I may have misunderstood some things :-).

> The main problem beginning Japanese learners often face is that they are taught polite form before plain form.

I often heard that the rationale for doing so is that learners don't sound impolite. It kind of makes sense, but on the other hand, I think it's also part of the "日本語が上手ですね" problem, that is, for the Japanese, the polite form is rather advanced.

Anyways, I do agree that for learners it would all make more sense to start from the basics and learn the polite forms later. Textbooks tend not to, though, sadly.

If I could upvote something multiple times, it'd be this post. The conjugations are well documented and can be intuited with enough exposure, but the insight as to why (い-verbs) and common usages (お早うございます) is something clearly lacking from the literature.

Counters exist in English.

"5 head of cattle" is exactly analogous to "ushi go tou" (牛五頭)

Most English counters have left common usage but the concept exists. Similar but not the same many people enjoy learning all the names for groups of animals I English. "A murder of crows" for example

Oh, I just remembered another fun feature of Japanese verb conjugation.

Can you conjugate the (regular) verb "へる" (heru)? There happen to be two verbs that are pronounced "heru" but with completely different conjugations! Do you remember which one is which? In fact any verb that ends with -eru or -aru can be potentially conjugated like an ichidan verb or a godan verb. For each such verb you have to remember its conjugation class, lest you conjugate it completely wrong. So these aren't so regular after all.

Also a related topic is intransitive/transitive verb pairs. These are also heavily irregular, and you must use the correct one in a sentence or it will be ungrammatical.

I think this is overstating the regularity of Japanese a bit. For example, what is the past tense of "kuru"? What is the present tense of "datta"?

Yes. It is overstating Japanese regularity a very little bit.

And understating the irregularity of English an enormous amount.

My Japanese's pretty rusty, so corrections welcome.

IMO you're not "overstating Japanese [verb][0] regularity a very little bit", but severely overstating it. The ichi-dan / go-dan verbs are indeed quite regular, and though there are irregular verbs like the grandparent post's example, IIRC there aren't actually that many (three?). However, that's not the only "axis" one could conjugate Japanese verbs. The complicated stuff are the transitive / intransitive forms, and the compound verbs. While some rules apply, they are more like the "I before E, except after C" English spelling rule.

It would be akin to saying that there are only three verb tenses in English: past, present, and future; simple, right? While it's technically not wrong, it's glossing over the progressive, perfect, and subjunctive forms (maybe there are more, don't remember), and their combinations.

[0] Hopefully from context it's clear we're only talking about grammar pertaining to Japanese verbs, and not Japanese in general. Japanese grammar has plenty of quirks, like how some colors are adjectives (red: akai, blue: aoi) but others are nouns (green: midori, purple: murasaki), or how loan words are usually written in katakana, but tobacco is not, etc.

The thing with English is people in general can understand most things even if you screw up tenses or tones. You put a few words together and people will mostly get you. In Japanese you have to remember extra things to add so you identify different parts of a sentence. Same with gender in romance languages. They just add more complexity. In a way Korean and Chinese are a little easier in terms of rules.

You should definitely see Turkish. it's belong to same family with Japanese. But verbs are even bigger. There are more variants as suffixes to the verbs. Sometimes a whole English sentence can be translated into one Turkish word. Have a look. "okutamadıklarımızdansınız" just one word. It means you are one of those who we can not make them read.

If you meant "family" in linguistic sense, then, no. Japanese is of Japonic languages family, and Turkish is of Turkic. The proposed Altaic family (which was a parent to both, in its expanded "Macro-Altaic" version) was discredited, and Japonic is generally classified as an isolate language family.

"can not make" or "cannot make"?

> makes me want to point out that it's not the whole story

Surely that's the whole point of the "8020" in the URL?

Personally I thought it was excellent, compact, understandable introduction. Sure it leaves things out, but if it didn't it would be a textbook!

The conjugation of japanese verbs are non negotiable - it has to be there for you to make any sense, and I don't think OP was claiming that you can move(?) how a verb is conjucated - just that the subject and object order doesn't matter (which is correct). The tense of the verb or conjucation of the verb has nothing to do with that.

Why do you use nemuru as the root? Mostly I would hear and say neru. So I could not sleep would become nerarenakatta.

寝る(neru) and 眠る(nemuru) are different verbs. Different kanji in writing, own conjugations.

Neru tends to indicate more intentional sleep (arranging to go to sleep: deciding to retire to bed, to lie down etc), whereas nemuru is spontaneous sleep (falling asleep --- and not necessarily lying down).

"Nerarenakatta" has the possible interpretation of something like, "I couldn't sleep (because of no opportunity to get away from activity and lie down)" whereas "nemurenakatta" is "I couldn't fall asleep" (insomnia).

I think if you stayed up all night studying and so because of that you couldn't sleep, that's when you might best use "nerarenakatta". I couldn't sleep (because I needed to do something else with the time, not due to failure to fall asleep).

I learned Japanese very much the same way I learn programming languages and found it to be very easy to learn spoken Japanese.

As far as languages go, Japanese is structured a lot like a programming language. If you learn five or six "bunpo" or grammar rules, you can go a very long ways. Then, to improve, just add rules to your mastery.

When I first learn any programming language I start with basics: variable binding/assignment, types, conditionals, looping, etc. Japanese fits very nicely into the same learning method.

Does a language have if/then? is it 'if (<expression>) { expression }'? Or 'if <expression> then <expression> end if'? Is there an 'unless' form? What about 'else'?

For Japanese, it's <expression> naraba <expression>. That's it. Unless? <expression> nakeriba <expression>.

How about while? <expression> nagara <expression>

For people who can learn the gist of a programming language in a week, you could learn the gist of Japanese in a week or two. That doesn't mean you would be fluent. You'd still need to learn thousands of vocabulary words. But the basic mechanics can be mastered in days or weeks. More mechanics can be layered as needed.

"For Japanese, it's <expression> naraba <expression>. That's it. Unless? <expression> nakeriba <expression>."

There are actually a number of ways to say "if" in japanese, and the one you mention can only be used in certain contexts.

People might get your gist if you use the conditional tense for everything, but you'll be wrong a lot. The -tara/nara grammar is at least as commonly used, if not more so.

I bring this up to illustrate only that the "programming language" metaphor doesn't go very far. Japanese, like any human language, is loaded with weird, illogical exceptions.

What? You mean there's a ternary operator?

I bring this up to illustrate that the metaphor isn't perfect, but it has some legs. I meant that BNF diagrams would be easier to create and understand for Japanese than for English.

You are right that there are more forms for saying 'if'.

Then again, I've seen programmers who spell 'if' 'f-o-r'.

No, I mean that there are at least a half a dozen different ways to express "if", and they all have specific semantic uses. It is way, way more complicated than syntax.

Agreed. This [0] and the subsequent lessons go onto explain that if clauses change depending on whether something is a fact, whether it is an invitation, whether it is volitional, etc.

I've been studying Japanese for years and I am at a basic intermediate level. I still find grammar hard to read, and even harder to produce naturally. I may just be particularly thick but I think saying "you can master Japanese grammar in a matter of weeks" is pretty misleading.

[0] http://www.learn-japanese-adventure.com/japanese-conditional...

Yeah, anyone who suggests that is either exaggerating, or doesn't know as much as they think they do.

Japanese grammar is simpler than English in some ways, harder in others. Overall though, it's really, really hard.

To be fair, all things considered, if you have to learn either language from scratch from a native language that has no link[1] whatsoever to it, I think English is harder than Japanese.

1. by which I mean, in the case of English, for example, if your native language is not french, german, etc.

In fact, I think English is harder to learn for Japanese native speakers than Japanese for English native speakers.

Yeah, I can't really speculate. It seems to me that the #1 biggest obstacle for Japanese speakers of English is getting past the Japanese education system, and actually speaking. Beyond that, I don't know.

But as a general principle, I believe that learning any new language is really, really hard. The people who claim to master languages in a few weeks are usually full of it (and by "usually", I mean "essentially always".)

I don't think this metaphor has legs. "naraba" can mean other things, and there are lots of cases where you'd say "if" in English but "naraba" wouldn't work.

I certainly agree that Japanese grammar is simpler than English, and relatively easy to learn, but I don't think this case (either "naraba", or the "if/then" case generally) illustrates that.

Japanese actually has a confusing array of conditional constructions

  VERB (conjugated) + nara (or the ridiculously formal "naraba")
  VERB (conjugated) + some noun, such as baai (case)
  VERB (plain form) + to
  VERB (izenkei form) + ba
  VERB (past form) + ra
And there are complex forms such as:

  VERB (plain form) + to shitara
  VERB (plain form) + to sureba
And you can add moshi or moshimo in the beginning to increase the level of supposition.

Each has a slightly different use and meaning of course, but it's still confusing enough for me even now.

5 or 6 grammar points can go a long way? If it's that few then I guess they must be the formal forms. So you'll need at least another 5 or 6 to understand the casual forms that you would use with friends or when you're the sempai in the situation. Want to talk to kids or young adults? Better learn another 5 or 6 forms of manga styled grammar slang. And if you interact with sales clerks, listen to train announcements or want to hear other official announcements, they're going to speak keigo, so there's another "5 or 6" forms to learn.

If you're interested in learning to read Japanese, you should give it a shot. It's intimidating but pretty do-able. Set yourself up with flash cards (I suggest AnkiDroid) and you can be pretty good at reading hiragana and katakana within a week. Then, study kanji flash cards for an hour a day and you'll know all the common kanji before you know it. Try to memorize 10-25 kanji per day and you can learn all the jōyō kanji within a few months. If you spend a few minutes telling yourself a story about each radical and kanji as you study you'll have no trouble memorizing them.

What AnkiDroid cards do you use?

I had an ex-colleague (a dev) who also knew Japanese well. He was doing Japanese <-> English translation part-time in the same company. Later quit software to do translation full time. When I asked him how difficult it was for him to learn Japanese, he said that Japanese grammar was very similar to Marathi grammar (which, I'm guessing, may be similar to Hindi grammar). Wonder if any one else has observed the same points as him, even for the similarity of Japanese to some other language.

I've only glanced at japanese before but after reading this article i had a similar notion.

It's as if you populate the parameters to a function by placing them in specific registers (using specific particles) and then execute the function (verb) once you're done.

I imagine this idea breaks down horribly once multiple verbs are introduced to a single sentence...

I like that you can add 'sen' at the end of a sentence to make it negative, or 'ka' to make it a question.

"Ka" at the end turns a sentence into a question, yes. You get a yes/no question, or you can swap in an interrogative pronoun if you like. This is ignoring pragmatics.

"Sen"... no. The closest is that present-tense polite (teineigo) forms of verbs change from -masu to -masen to make them negative. However, that's basically the only case that rule works. You also often end up changing the rest of the sentence, just like in English or other languages. For example, in English, indefinite pronouns will generally change when you negate, as the negative of "somebody is here" is "nobody is here", not "somebody is not here". And of course, being a natural language, it's full of exceptions (like how "everyone" doesn't follow the pattern of "everything", "every time", etc.)

One minor quibble with the author's example sentences: They use "watashi wa hito desu" to mean "I am a person."

I'm not a native Japanese speaker, but I'm pretty sure that "hito" is only used to refer to other people, never to oneself (source: the excellent "Nihongo Notes" series by the Mizutanis).

Maybe watashi ha ningen desu 私は人間です (I am a human) would be a better example that still illustrates the grammar pattern.

So I am sitting in Ebisu in Tokyo right now. I spend about 3-4 months a year here and have for about 8 years. My understanding of spoken Japanese is pretty decent. However it is by pure memorization over time. This just sorted a whole bunch of things out in my head as to the why. Very good stuff. Thank you.

Oh cool. I'm in Koenji.

If you would like more of this kind of thing, I highly recommend Tae Kim's Guide to Japanese Grammar. It's a lot like OP's article, but much more thorough, organized by topic, and provides auxillary resources so you can dig in as much as you please.

I have been here for 4 years, the sites down at the moment but I'm hoping it will help me!

I love how precise and detailed this article is written. If only more documentation were like this.

It was always my assumption that grammar is the most important thing to learn about a language. Vocabulary accumulates almost automatically over time, with practise (and a dictionary). Interesting to see how that holds in this case.

Just to play the Negative Nelly role a bit, while the main thesis is on point, the example sentences I find to be somewhat unrealistic and misleading.

I have ideas for how to improve things and wouldn't mind brainstorming things together with the author of that's something they're interested in.

Most important for what? For speaking about linguistics, maybe. But if you are using the language to communicate, there is no most important single thing to "learn", since using the language to communicate means that you are using the language in a way that resembles a skill, and the language grows in your brain as an implicit system – you acquire it bit-by-bit, but the internal system doing the acquiring doesn't care what counts as "grammar" and what counts as "lexicon". You can teach yourself rules, but that doesn't mean that you are able to use them spontaneously.

TL;DR: knowledge about language is different from skill of communicating in language.

> It was always my assumption that grammar is the most important thing to learn about a language

I think it depends on what your goals are. If you just want to communicate, vocabulary is the most important thing to get started, since even with poor grammar people can infer what you mean (gestures help too!). I've seen too many people here in Japan who have crammed English grammar all their school years but are paralyzed when trying to communicate because they're just focused on getting the grammar right.

I think you should learn a handful of grammar rules right up front.

Obviously, * Desu/deshita/de arimasen/de arimasen deshita * Masu/masen/mashita/masen deshita

Then add a few more * if * while * want to <verb> * passive * honorific * etc.

You can learn one a week or one a month. There aren't very many to learn. But having just four or five up your sleeve (in addition to verb conjugation) can really improve your comfort.

Absolutely! You're not going to get anywhere in any language with only either grammar or vocabulary.

Anyone actually recommend the book from which the article is taken? I also tried to read the wa v. ga blog post on the site to get a further sense of the author's approach, but the server returns an out of memory error (from a blog post?!).

I've been in Tokyo now 18 months, took private lessons twice costing about $2,000, and feel I learned 10 words. That's $200/word. I joke with people I stopped taking lessons because learning Kanji would bankrupt me. Japanese just doesn't stick in my older and very Western brain. It doesn't help that my office does business in English and one can get by in Tokyo with minimal Japanese and a lot of pointing and gesturing. The glacial progress becomes discouraging.

I tried Rosetta Stone. It takes the same phrasebook approach as the first textbook I was given, Nihongo Fun & Easy, which was neither. The textbook at least had short sidebar discussions of grammar and somewhat useful phrases. I had no idea where I'd get to use the phrase "The children are swimming," that Rosetta offers.

The 8020 article was the first discussion of particles that actually made sense. When I'd asked teachers about particles before the answer was usually something like "Don't worry about that yet, just memorize the phrases." If the remainder of the book is in the same vein I'd pay twice the asking price. I flipped through parts of Nihongo Fun & Easy after reading this article and it suddenly made much more sense. I wasn't staring at a list of phrases I was supposed to memorize and slowly reverse engineer the language, but could deconstruct the basic sentences.

It's much easier for me to learn construction, and use the break down of other sentences to construct my own, even if the rules fail sometimes and lead me to construct sentences no native speaker would utter. That's the other 80% of language idiosyncrasies that takes time.

I don't expect to be fluent in Japanese any time soon, however moving past "sumimasen kore onegeihshimasu" while pointing at a menu item would be awesome.

I really recommend the Japanese For Busy People books, I learned tons from them quickly. They start off very basic (obviously), but they explain the grammar in a really good way, and progress into more advanced topics as you go along. Make sure to get the kana versions, not romaji!

For learning Kanji, the most efficient option I've found is Remembering the Kanji by James Heisig. It's a bit of a long-term investment, in that it takes a while for them to really pay off (you don't learn the readings/sounds until book two, the first one is completely focused on the meaning and writing of the characters), but in the long run I think it's a much better option than for example the books trying to teach you the characters by showing you their similarity to the things the represent.

I strongly recommend a method of language learning called TPR (Total Physical Response).[1][2][3] It is by far the fastest and easiest way I know of learning vocabulary and grammar.

In a nutshell, with TPR the teacher gives the student a command in the target language, demonstrates the action the command is asking for, repeats the command and finally the student copies the action.

For example, if English was the target language, the teacher could say "sit down", then sit down themselves, then once again say "sit down" and the student would copy the action by sitting down.

This can then be repeated for "stand up", for "pick up the fork", or any arbitrarily complex and sophisticated command.

As you've no doubt noticed, the commands are given in grammatically correct sentences, in context. Grammar is not explicitly taught, however. It is implicitly taught and implicitly learned.

What makes this method work really well is that when you learn words and grammar, you're not doing it with just your mind and maybe some visual cues, you're using your body and doing so in a specific physical context (the place where you're learning), associating what you're learning with parts of that place. It's somewhat analogous to using a memory palace to learn, only without any extra effort of constructing the palace or imaging placing things you want to learn there. With TPR you actually physically interact with the things you learn in that space.

Another great thing is that a TPR teacher need not have any special training in education or really even in the method. TPR takes maybe a minute or two to explain to anyone, so you can recruit helpers from any friends or acquaintances you have who know the target language and are willing to help, though if you want consistent lessons and dedication you'll probably want a professional tutor or teacher anyway.

TPR focuses on learning to understand, in emulation of the first step of a child's language learning process. Children first learn to understand, then to speak, then to read, and finally to write. TPR helps with the first part.

When I taught my tutor this method, he told me that I was by far the fastest of his students to pick up vocabulary, and he wound up switching completely to teaching with this method. It was really effective for me, and I highly recommend it.

TPR has its limits, and it can't be used for all aspects of language learning, but it's fantastic for getting your language learning bootstrapped really quickly.

[1] - https://en.wikipedia.org/wiki/Total_physical_response

[2] - http://tola.maf.org/collect/missionb/index/assoc/HASH01da.di...

[3] - http://tola.maf.org/collect/missionb/index/assoc/HASH0162.di...

This article resembles two different ways of designing protocols. In tcp [0], information are encoded in position, e.g. the first 16 bits are for source port and the next 16 bits are for dst ports. While in, say, FIX [1], information are encoding be delimiters and position doesn't matter.

[0] https://en.wikipedia.org/wiki/Transmission_Control_Protocol

[1] https://en.wikipedia.org/wiki/Financial_Information_eXchange

Using particles in this way almost sounds like using flags to specify parameters when calling a function. This allows them to be placed in any order. A Powershell-like example:

  Construct-Sentence -subject Taro -object Noriko -verb to_see -time Past  
  > "Taro saw Noriko."

  Construct-Sentence -object Noriko -time Past -verb to_see -subject Taro 
  > "Taro saw Noriko."
The original sentence is "Tarō wa Noriko wo mimashita." And "masu" appears to be the root verb "to see".

  Construct-Sentence -wa Taro -wo Noriko -verb masu -time Past
  > "Tarō wa Noriko wo mimashita."
Something more Bash-like:

  csent wa:Taro wo:Noriko masu -past_affirmative
  # "Tarō wa Noriko wo mimashita."
Meanwhile, subject-object-verb (SVO) and similar patterns depend on the order of inputs:

  Construct-Sentence Taro Norkio to_see Past
  > "Taro saw Noriko"

  Construct-Sentence Norkio Taro to_see Past
  > "Noriko saw Taro"
This allows for invalid outputs:

  Construct-Sentence Taro to_see Norkio Past
  > "Taro Noriko'ed see"

I've made the analogy of function calls before, it's a pretty cool way to think about it. But I think your example is slightly off. The verb is the head of the phrase, and would thus be the function (or command) itself:

    miru --topic=Taro --object=Noriko # Taro sees Noriko
This uses "miru", the plain non-past form of "to see". Conjugating the verb is a little trickier to translate, but I think it would be analogous to higher-order functions, functions that modify the verb to create a new verb. Using a Python-like syntax:

    formal(past(miru))(topic="Taro", object="Noriko")
    # Taro wa Noriko wo mimashita
I'm trying to think of a way to include relative clauses in this analogy, but it's a little harder. A relative clause is of the form "[verb-phrase] [noun]", e.g. "doresu wo kiru Noriko" (Noriko who wears a dress). Maybe we could use positional arguments for this:

    formal(past(miru))(topic="Taro", object=kiru(object="doresu", "Noriko"))
    # Taro wa doresu wo kiru Noriko wo mimashita ??
EDIT: or maybe currying would be a nice way to solve this. So each verb would be a function that takes arguments, and returns a new function that takes a noun to apply that verb to:

    formal(past(miru))(topic="Taro", object=kiru(object="doresu")("Noriko"))
    # Taro wa doresu wo kiru Noriko wo mimashita

Could this situation be described by nesting sentences?

  "Taro saw (Noriko wears a dress)."
To extend your example:

  formal(past(miru))(topic="Taro", object="Noriko")
  "Taro saw Noriko."

  formal(present(kiru))(topic="Noriko", object="doresu") 
  "Noriko wears a dress."

  formal(past(miru))(topic="Taro", object=(formal(present(kiru))(topic="Noriko", object="doresu")) )

  "Taro saw (Noriko wears a dress)."
  "Taro saw Noriko wear a dress."

  "Taro wa (doresu wo kiru Noriko wa(?)) wo mimashita"
  "Taro wa doresu wo kiru Noriko wo mimashita"
Does that make sense?

*Please help correct any errors, as I have never formally studied the Japanese language.

It wouldn't type-check, since `object` is expected to be a noun and not a verb. :)

It's a difficulty inherent in describing one unfamiliar language with another unfamiliar language. :)

This could work in Powershell. The function would have to be written to expect nested recursive calls, to remove the omitted particle.

  Construct-Sentence -verb miru -Formal -Past -Topic Taro -Object (
  Construct-Sentence -verb kiru -Formal -Present -Topic Noriko -Object doresu

  "Taro saw (Noriko wears a dress)."
  "Taro saw Noriko wear a dress."

  "Taro wa (doresu wo kiru Noriko wa(?)) wo mimashita"
  "Taro wa doresu wo kiru Noriko wo mimashita"

Filipino is really similar in terms of the use of particles/markers! For example, to say: "The cat is eating the fish", we say: "Kumakain ng isda ang pusa". The verb (is eating, kumakain) always comes first. The subject (cat, pusa) is identified by the "ang" marker, while the object (fish, isda) is identified by the "ng" marker. We could also say "Kumakain ang pusa ng isda", although that's rarely used.

The "-um-" affix in "kumakain" makes the verb active ("is eating"). If we instead used the "-in-" affix (as in "kinakain"), it would make the verb passive ("is being eaten by"). So we could alternatively say: "Kinakain ng pusa ang isda" to mean: "The fish is being eaten by the cat".

I quite like the headlining diagram. It's a simplified view that shows the schematic approach of the languages - Japanese relies on case particles rather than ordering (unlike English). Of course, there's a lot of complexity that goes on under the hood when you start to figure out the appropriate verb conjugations to use (which aren't shown in the figure).

Small side comment, if anyone's learning Japanese and wants to ask or answer questions about it, you're welcome to join a little Discord chat group (including native speakers and advanced learners) at https://discord.gg/6sjr3UY

Yeah it's "logical", except in casual language the rules are broken all the time. A lot of things can follow the verb.

The article doesn't mention subclauses at all, but it's where things become hairy. There's particle "ga" which is similar to "wa" except it works as a subject of subclause, except sometimes it means "but". There are dozens of ways to incorporate subclauses into main sentence, using different particles. It's very common for the entire sentence to be a subclause ([something] no/n desu).

Interestingly, many Indic [2] languages follow similar verb-centric grammars. In the canonical Vyakarana tradition (of Panini), sentences are seen as revolving around the verb [1].

The "noun-cases" or कारक (karaka) are generally equivalent to the "particles" in Japanese. The genitive (eqv. の) is not a karaka, since it has no relation to the verb. Of course, since there is technically no syntactic difference between adjectives and nouns in Sanskrit, the semantics of the genitive in particular can be very undeterministic. This is not the case in others though.

I wish there were more studies on how Indic traditions affected East/SE Asia [3]. Sadly, most academics/people here don't believe there exists a world outside N. America & W.Europe (often no India either!).

[1] There is a competing tradition of semantics called "Nyaya" where sentences are seen to be Noun-centric. These discourses are generally not easily accessible.

[2] Dividing the languages based on presence/absence of noun inflections would appear not to have much discriminative power to claim anything about historical origins. Historical Linguistics, I believe, is mostly a politicized pseudoscience.

[3] This documentary highlights the kind of things I mean.


It is also fascinating to look at the Thai/Khmer scripts and realize these are related to current day Telugu/Kannada scripts.

Since you brought up Sanskrit and topic of discussion is sentence structure I will provide one more data point. Because Sanskrit is a highly inflected language there is a lot of flexibility in sentence structure in Sanskrit. In fact one can put the words in pretty much any order in a Sanskrit sentence. This flexibility comes at a very high price though - there are tens of ways in which a noun can transform and theoretically thousands of ways in which a verb can transform. The good thing though is that Panini and later grammarians gave us rules to go by so it is not as bad as it sounds.

> In fact one can put the words in pretty much any order in a Sanskrit sentence.

But will all of those variants still be considered the "same" sentence? I'm asking because it's popular to make the same claim about Hungarian, but it's not really true. You can switch things around a lot and still get fully grammatical sentences that all relate to the same event. But due to Hungarian's topic/focus structure, the actual meanings expressed by the variants are so different that Hungarian speakers wouldn't consider them "the same sentence expressed a bit differently". Some examples: https://en.wikipedia.org/wiki/Hungarian_grammar#Emphasis

In contrast, I believe Latin is really liberal in its sentence structure, especially in poetry.

Yeah, and reading Latin is complicated by the fact that Latin authors like to put the adjective and its noun on oppposite ends of a phrase and then stuff other things in between. So, instead of "the yellow car in my driveway that belongs to my brother" they'll say "the yellow in my driveway that belongs to my brother car". Because an adjective must match its noun in number, gender and case, the authors think of the noun and adjective as holding the phrase together. And this is only the beginning: in poetry there are even fewer patterns because the words are generally ordered to fit the meter: the only thing you can really count on is that certain particles must be the second word in a sentence (important because the manuscripts didn't have punctuation) and that a preposition always precedes its object.

Pretty much everything you said about Latin holds for Sanskrit too. Sanskrit has additional complication 'Sandhi' - euphonic combination [1]. There are many rules which are used to combine words depending on the ending sound of the first word and the beginning sound of the second. And you can keep on combining words as long as the rules are applicable. So you can get very long words without a break. Newbies often struggle with this.

[1] https://en.wikipedia.org/wiki/Sandhi

Yes, in most cases the sentence will still mean the same unless it's a very complicated sentence with multiple nouns and adjectives where there may be some ambiguity as to which adjective goes with which noun (and that too only if declensions happen to be in the same case). Having said that, the most common sentence structure used is still subject-object-verb.

It is for the most part (see answer above).

That's the story that's told in introductory texts, but the genitive generally forbids such permutations (depending on the flavor of usage in the geographical area).

There are actually only 7 kinds of सुप्s (x 3 numbers); there are about 10 तिङ्s, but only about 3-4 are in general usage. Then there are other sentence transformations outside these.

The relational modifiers "stick" to the roots, so moving words around the sentence is quite easy - this is particularly useful to poets who need to satisfy certain mathematical rules in their compositions. Much of this is also true for languages like Kannada/Hindi. There was also a paper about a KR scheme inspired by something of this kind.

Sanskrit is generally easy to pick up with the right teachers, in an immersive environment; the grammar-technique, esp. in a "sterile" environment, is rather terrible IMO.

>The genitive (eqv. の) is not a karaka, since it has no relation to the verb.

Unless I'm misunderstanding your point, Hindi's sambandh kaarak (ka-ke-ki-ra-re-ri-na-ne-ni) is the genitive case. watashi no hon <-> meri pustak, etc

It is indeed the सम्बन्ध; it's not a कारक technically. The canonical defn. of कारक is that which relates a noun to the verb. सम्बन्ध relates relations between nouns (you can't have sentences with only a verb and a noun in genitive).

:shrug: It's part of the list of kaarak and is taught as such.

Shrug all you want, but it most certainly is not and neither will it ever be. The definitions have been laid out in the Ashtadhyayi ~2000 years back.

It's not my concern if certain folk have no education in Panini, and are arrogant enough to ignore his work.

I'm wondering if the main factor is one's ability to learn a new language - itself affected by many factors such as age, for example?

I've been married to a native Japanese for going on 19 years now.

I have tried to learn the language. I have lived in Japan for 6 years, hoping that full immersion would help. I even embarked in the Kumon Japanese course whilst in Japan, from beginner level to more advanced. I have piles and piles of the work books cluttering my home.

I ended up being able to read katakana, hiragana, and learned some 250 Kanji.

What I didn't end up managing was being able to have decent conversation in Japanese. Sure, I could ask for a beer, directions, talk about the weather, but that was about it. I had reached some plateau and could go no further.

In the end I gave up. It was basically something I couldn't do. I tried many different ways of learning, found none which could not prevent my sheer frustration at not being able to take the knowledge in.

Are some people simply 'wired' to learn language more than others? Is there an age limit, for example? Was it my low tolerance for frustration? Was it my perfectionist tendencies? Probably a 'yes' to most of those.

But I stopped after more than a decade of trying.

The article says:

> What this means is that the sentences, “This is a car”, and, “This is the car”, would both be, 「これは車です」. There is no differentiation.

This is not always true. The latter could be 「これが車です」. The は and が particles are very similar but are still different. Fully grasping this small difference is one of the biggest problems Japanese learners encounter when studying grammar.

Closer to the beginning, the article also mentions:

> The topic of a Japanese sentence is very similar to what other languages refer to as the subject. The subject of a sentence is the person or thing that does the action described by the main verb in the sentence. These are, in fact, slightly different concepts, but for now, we will treat them as being the same so as to keep things simple.

It turns out that は marks the topic and が marks the subject. I feel that many times the confusion between は and が in Japanese learners happens because the learning material tries to make this simplification in the beginning. When it's time to learn が, it's hard to retrain the brain.

This isn't quite true.

は has two distinct usages. One is to mark the topic of the sentence. The other is to mark the subject of the sentence with a contrastive connotation.

It also has a use for emphasis. And a few others uses.

And what does 'contrastive connotation' mean?

Anyone that would benefit from this style of learning (rapid, focused on structure and rules) may actually be hurt by the rush to cover many topics without treating any precisely. I'm by no means an expert but here are some issues in just the first section.

* example that glosses over the difference between a topic and a subject is frustrating because, in fact, the similarity is fairly superficial.

* there is no "a", "an", or "the" in Japanese, however to specify "this is the car" (implying that it is in answer to some question about which car) one would say これが車です。Using the が particle instead of は.

I'm always on the lookout for useful resources. So far, Tae Kim's guide [1] has been the best I've found. Kim doesn't assume much about the reader's pre-existing knowledge yet he is able to remain succinct.

[1] http://www.guidetojapanese.org/learn/grammar

Remember those English diagramming classes everyone hated? I'm not very familiar with the education system in Japan, but I doubt they have diagramming classes. In Japanese, the diagramming is built into the language. You tag the subject, the direct object, the indirect object, etc. Everything gets markup.

Which part of the sentence is the direct object? Uh... the part with the direct object tag hanging off it? Correct!

Have you ever heard a programming language described as "designed for teaching"? Japanese is a language designed to be as simple as possible to learn.

Coming from English, the idea that a natural language could actually be designed was a shock to me. I thought they just evolved sloppily and haphazardly. Well, Japanese is proof that it doesn't have to be that way. Clear rules and not too many of them. No exceptions. Rigidly consistent. It's like a language created in a lab that never got dirtied up by real world usage. Except, oh wait, it's a real language used by millions of people every day.

I never had the pleasure of taking them myself, butI'm pretty sure Japanese grammar classes are not so fun as you imagine.

Putting aside that the norm in Japanese school is rote memorization and pedantic attention to details (e.g. you'd memorize dates of historical events), the Japanese grammar that is taught in class is traditional Japanese grammar. It's pretty streamlined compared to the abomination that is medieval Latin grammar (which remains the basis for English grammar taught in class), but it still targets Classical Japanese and uses rather obscure terms where clear diagrams would suffice.

I'll have to ask, but I think that instead of diagrams, Japanese students mainly need to memorize the difference between Izenkei, Mizenkei, Renyoukei, Rentaikei and Shuushikei, even though the last one is irrelevant for modern Japanese.

And of course Japanese wasn't "designed" any more than English was. There are dialects with widely varying grammar and vocabulary, and there some aspects which are very hard to learn even if you ignore the writing system (e.g. the proper use of Wa vs Ga, the proper use of emphatic sentence endings like no/n'da or yo). Inflection is way more regular than English, and syntax is pretty streamlined, which is a boon.

Wa/Ga aren't that hard. Mostly just subject vs topic. Just like objective and subjective isn't that hard in English. Of course, most Americans can't master objective/subjective.

The trick for wa/ga is often just to remember that subjects are often implied in Japanese and missing from the actual text/speech. So, "Nekko ga suki desu" because the "watakushi wa" was implied.

I think you're probably falling into a common trap: There are plenty of exceptions, you just don't know them yet.

For sure, Japanese verbs are more regular than English verbs. But most materials written about Japanese are aimed squarely at beginners. They naturally skip over any foibles for pedagogical reasons. There isn't all that much written (in English) about tricky sentences, syntactic ambiguity and similar mistakes.

There's a theory that humans have a certain "complexity budget" for languages. People will add things to simple languages, and forget things from complex languages, to arrive at some sort of equilibrium. Language complexity is a bit of a touchy subject though.

By comparison, Japanese might have simpler syntax and phonology than English, but higher orthographic and morphological complexity (and English is already orthographically complex).

Sanskrit is very consistent in its rules too.


We had it in school in grades 8-10 or so and I really enjoyed learning it. The grammar rules are fairly simple and there are not many of them, IIRC. Words can be made up out of smaller words by joining them together by simple rules.

Another interesting thing is that many words have multiple meanings, a fact that poet and prose writes in the Sanskrit literature leveraged heavily. So a passage of text can have multiple layers of meaning.

It is almost like a programming language where you compose bigger structures such as functions and classes out of the atomic elements of the language.

> the idea that a natural language could actually be designed was a shock to me

Perhaps the Japanese grammar was designed by a small group of individuals, maybe even one individual, a long time ago, and promoted officially by the Japanese Emperor of the time. Many writing systems have been designed from scratch in the past. Many spoken languages have risen from regional dialect to imperial language, and at least one language, Israeli, was brought to life after being dead as a spoken language for centuries. No-one's ever found a definitive link between Japanese and any other language. Perhaps at some time, the Japanese grammar was designed from scratch, perhaps using a mixture of vocabulary from many other nearby languages existing at that time, and promulgated to become the language of all Japan.

The grammar perhaps, but then you get to the writing system :)

Meanwhile over in Korean, the writing system was designed from scratch to be very simple (I'm led to believe - I don't know any Korean)

I really like the Korean writing system. It's very cool. Take the 1-4 sounds that make up a syllable, arrange them into the syllable template, and you have a single character that represents that syllable. It's pretty slick, space-efficient, phonetic, and full of yummy goodness.

Then, just to mess with people, optionally throw in thousands of Chinese characters.

I have to enter a caveat here though. Spoken Japanese is a little bit different, and in some cases, arguments will follow the verb. It's pretty rare, but I did hear things like "Dou sureba ii ore" or "nani yatteru omae"?

This is VERY rough and informal though, and not Japan being very polite, it's not something that I'd hear everyday. Maybe on TV or from really close friends, and even then I'm not sure if everyone would say that.

But it just goes to show that natural languages are very complex creatures, and even the tidiest rules have exceptions sometimes.

I wouldn't describe that pattern as incredibly rough or even particularly informal. It's common to at the subject after the sentence when speaking quickly. You use it when you think the listener may have to quite picked up what you were saying etc.

Something I like about the whole "verb at the end of the sentence" thing is that you can totally flip over the meaning of what you're saying, right at the end. In English, you can achieve the same effect with awkward forms (like "not" at the end of the sentence), but in Japanese, it's just the natural form.

Try to imagine the kind of snarks you could do if you could put things like "I don't reckon" on hold until the end of the sentence.

Sadly (ironically?), that tends not to be the kind of language subtlety/humor the Japanese go for.

> Try to imagine the kind of snarks you could do if you could put things like "I don't reckon" on hold until the end of the sentence. Sadly (ironically?), that tends not to be the kind of language subtlety/humor the Japanese go for.

If that's true, I think it might be more that such kind of humor is seen as too basic, or childish, and not particularly subtle for adults. Akin to simple puns or "ghost jokes" and the like in English (What do ghosts like for dessert? I Scream!) that kids — or foreigners — may enjoy, but native speakers don't really count as witty.

You can still see what you're talking about in some of their comedy skits, though: https://www.youtube.com/watch?v=4nW4jhqPbd0

As much as people claim to despise puns, the major English authors show very little reluctance to use them: they are everywhere in Shakespeare, for example, and you can also find them in more recent authors like TS Eliot (for example, in Ash Wednesday, he puns on dissembled/disassembled)

I was just thinking about negation by adding the suffix. We have it in Bangla too. In fact, we share the same word, 'nai', for 'is not there'.

I think many Indian languages have it too.

And in fact the Japanese negation at the end - negation suffix ("nai"→"naka") - from whym's comment in this thread - sounds like the Marathi one - same sound (naka or nako).

Beginner question: In casual, spoken Japanese, I've been taught that I can drop the particles (including pronouns). Hence "watashi wa tabemasu" can be colloquially shortened to "tabemasu".

Thanks to this article, I've come to understand particles much better and why they're important, but does it change in casual spoken Japanese? Are some particles okay to drop whereas others are kept? Thanks in advance.

> I can drop the particles (including pronouns)

Sure. "piza wo taberu?" can be "piza taberu?" and that's fine, even normal.

> (including pronouns)

This is different and I think there are a few misconceptions bundled up in this assumption.

1. Particles don't exist on their own; they're permanently linked to the word that precedes them; if you've ever studied a Romance language, you can think of them as a way of declining nouns.

So "watashi wa" is the nominative, "watashi wo" is the accusative.

2. Pronoun dropping is done in the sense that the pronoun is not essential to the sentence and can be inferred from context.

For example, if you and your friend are eating and you ask "motto taberu [gonna eat more?]," nobody's going to be confused about whether the subject of that sentence is "watashi wa" or "anata wa."

It happens in English, too, but people overthink it a lot when presented with it consciously in Japanese.

"Dropping the pronoun" is a very English way of thinking about things. In English sentences have to have subjects. Consider the following paragraph:

John went to the shop. John bought a cake. John took it to Bill's house. John and Bill ate the cake.

It's not grammatically wrong, but it's clunky. We just use pronouns to sound 'normal' in English:

John went to the shop. He bought a cake. He took it to Bill's house. They ate the cake.

In Japanese instead of changing a noun (John) to a pronoun (he) you just don't say the subject if it hasn't changed:

John went to the shop, bought a cake and took it to Bill's house. They ate the cake.

Which, as you can see, is something that works in English at times too.

Thanks. Sorry for the confusion; I have a decent understanding of when pronouns can be dropped, but my question was moreso about when particles can be dropped. For instance, from the linked article, this sentence:

Tarō wa Noriko wo toshokan de mimashita.

In casual spoken Japanese, can any of these particles (wa, wo, de) be dropped?

None in that case. I don't know of any general rules for when particles can be dropped, but sitting here thinking of examples, it's usually done for simple, direct sentences where the particle would be tying together two words whose relationship is already obvious:

> eiga miru?

> ano hon, suki?

> ame futte kita.

> okaasan iru?

And so on. But I don't think that's a rule you can work backwards from, it just describes all the cases that occur to me.

generally, no

Well, to be clear, particles and pronouns are very different things. (In fact, Japanese pronouns basically just act like nouns, from a grammatical perspective.)

Japanese is a so-called "pro-drop" language, which means it's normal to omit pronouns completely when they can be inferred. "Watashi wa tabemasu" would only be used for specific emphasis, as in: "I specifically (as opposed to somebody else you were just talking about) am eating." In any other context, it sounds unnatural, and you would just say "tabemasu" even in formal situations.

Separately, you can omit particles in colloquial speech when they're obvious. So for instance, in a polite setting you might ask "ashita wa, nani wo shimasu ka?" (meaning "what are you doing tomorrow?") But in a casual environment, you could say "ashita, nani suru?" and still be understood perfectly.

That's a common question... and unfortunately the answer is that "watashi wa tabemasu" and "tabemasu" are absolutely different from each other and you can't just go around substituting one for the other. You also can't really translate the difference between the two of them, but in context, a good translation might be:

Watashi wa tabamasu. -> As for me, I'm eating.

Tabemasu. -> I'm eating.

It bears repeating: this depends on context. But you can see how "watashi wa" sticks out like a sore thumb, and you can instantly recognize novice Japanese second-language speakers because they say "watashi wa" all the time. And you'll later learn that "watashi" is sometimes an inappropriate way to refer to yourself...

There's another article on the site [1] that explains this in detail. It seems that you can drop the "watashi wa" only when the context of the statement or answer is obvious from previous parts of the conversation.

[1]: https://8020japanese.com/wa-vs-ga/#comparison

As others have commented here, Japanese relies heavily on inference. Subjects and topics are very commonly dropped.

I shouldn't say that the language relies on inference. Rather, speakers of Japanese rely on inference. It's like chaining functions in some languages.


By the time you get to the washing in the conversation, the car is hard to find in the sentence, and the driver never does appear. But they are both there if you go looking.

In general, you can drop everything except the jitsugo (verb or adjective most usually) if the other bits can be inferred.

eg: Taberu (eat) or Hayai (fast)

It sounds like cavemen talk if you translate it literally, but it is perfectly valid japanese

Syntactically we might call "hayai" a verb, which would then be translated as "is fast" rather than just "fast". Semantically, it is an adjective.

A interesting property of Japanese is that a sentence is also a subordinate clause. For example

Tarou wa Noriko wo toshokan de mimashita. (Tarou saw Noriko at the library.)

Tarou wa Noriko wo mimashita. (Tarou saw Noriko.)

Tarou wa Noriko wo mimashita toshokan (The library where Tarou saw Noriko)

Generally "<sentence> <noun>" means "the <noun> such that <noun> <particle> <sentence> is true for some choice of <particle>".

Your third example isn't valid. It needs a bit of tweak.

Tarou ga Noriko wo mita toshokan

The particle "ga" and "wa" both introduce a topic. But in a phrase to explain a noun, we use "ga" exclusively. Your main point still holds, in a sense that "Tarou ga Noriko wo mita" is a valid sentence. But to be precise, "mita" in those two sentences are different conjugated forms; it just happens that two conjugated forms are the same in the verb "miru" (to see).

> The particle "ga" and "wa" both introduce a topic

Forgive me for saying this, since you seem to be a native speaker, but don't you mean that they both introduce the subject, not topic (using 'topic' as a linguistic term)?

"Wa" would be the topicalising subject marker, denoting known information:

Tanaka wa nihon ni itta.

Tanaka went to Japan. -> As for Tanaka, he went to Japan. Tanaka = known information (i.e. Tanaka is familiar to the listener)

"Ga", while also a subject marker could denote/introduce new information:

Tanaka ga nihon ni itta.

Tanaka went to Japan. -> e.g. It was Tanaka who went to Japan.

Tanaka = new information (e.g. the listener is did not not Tanaka was the one going to Japan.)

(Note: I realise there are other constructions for my interpretation of the ga-sentence)

My knowledge of Japanese grammar is in Japanese, so I'm not certain about the English term of 主語, to be honest. We use the same term to describe 'subject' in English grammar. I used 'topic' just because the original article used it.

Your explanation of 'ga'/'wa' is spot on as far as I can understand as a layman of native speaker with standard Japanese grammar education in Japan but no advanced linguistic degree.

I'd say that, because 'wa' emphasizes the introduced subject as the center of interest, it isn't used in the subordinate clause.

Tanaka ga nihon ni itta hi. (The day Tanaka went to Japan) ; ok - the interest is on 'hi'

Tanaka wa nihon ni itta hi. ; invalid

Thanks for your reply. I believe 主語 covers both subject and topic. Since an English sentence such as "John loves Mary." can be understood as e.g. "It is John (not James) who loves Mary." or "John loves Mary (not Lisa)", it might have several formal representations in Japanese via e.g. the use of wa/ga.

Also, see user gizmo686's excellent explanation for one approach below.

Not a native speaker, but have studied Japanese linguisticly (as well as as a second language).

Wa is a bit of a complicated topic. The prevailing thinking is (roughly) that it has two distinctive meanings: topic marking, and contrastive. As a topic marker, wa does not introduce the subject (although in many cases, there is a null anaphora referring to the topic).

In anycase, the common linguistic explanation for shiro's correction is that the subject of subordinate clauses resists topicalization.

(Is this where people start flaunting their phd:s, professor titles? j/k ;-) academic here as well - I do not hold a phd)

Well, I realise the topic + contrast bit but is it really treated as a null anaphora, rather than acting as both topic and subject marker in my example...? My examples referred to information structure more than anything.

Yes, I realise you can have sentences like "Ashita wa Tanaka ga..."/"Sou wa hana ga nagai." - I've even seen a discussion on double topics (some old, theoretical text by Yasuo Kitahara IIRC, probably more known for 'Mondai-na Nihongo'). I also realise that in some contexts where it seems to denote a subject its noun is only a topic ("watashi wa unagi desu").

Logically, it would indeed be quite difficult for a subordinate clause to contain the/a topic.

Anyway, I'm curious if you happen to have further explanations (or articles)!

(Unrelated note: why is it that Japanese of all things make us crawl out from under our rocks...? :-))

No PHD here either, just undergrad followed by some hobbiest reading (of scholarly sources) on Japanese linguistics.

To be clear, the comment about null anaphora was more of a throwaway comment anticipating the objection that sometimes the topicalizing wa does mark the subject. While I have seen this explanation presented, and it is my prefered explanation, I would not necessarily call it pervasive. Now, for the explanation itself (unfourtantly, I am on vacation, so cannot check any of my references).

Japanese is a clear example of a pro-drop language, so using pronoun dropping (aka, null anaphora) as an explanation requires less justification than it would in English, where we only see it in specific contexts. Additionally, we see the topicalizing "wa" in various contexts, not all of which can be understood as subjects, so a unified explanation that can account for all of them would be preferred.

For example, consider the sentence

1) Mary-ga ringo-o tabeta

We can topicalize Mary with the following derivation:

2) Mary-wa Mary-ga ringo-o tabeta

3) Mary-wa anohito-ga ringo-o tabeta (Pro-form substitution)

4) Mary-wa ringo-o tabeta (Deletion)

Simmilarly, we can topicalaize ringo with

2) Ringo-wa Mary-ga ringo-o tabeta

3) Ringo-wa Mary-ga are-o tabeta (Pro-form substitution)

4) Ringo-wa Mary-ga tabeta (Deletion)

We also have the following sentence (kudamono = fruit)

Kudamono-wa Mary-ga ringo-o tabeta

Admittedly, I struggle to think of a context where the speaker would not drop Mary due to context, but that should not be relevent here, and I am sure that there exists better examples.

Notice that, under the null anaphora explanation, all three of these examples could be explained in the same way. If we were to explain the first example as wa being a subject marker, then we would need to explain the second example as wa being an object marker, and the third example as wa being just a topic marker.

I have seen an alternative explanation that describes topicalization in Japanese as a transformation rule. I have mostly seen this by researchers who view Japanese non-configurationally, who argue that a rule such as [ga/o] -> [wa] in a non configurational language is directly analogous to a movement rule in a configurational analysis. Even under this approach, you still need to account for sentences where the topic has no co-referential place in the rest of the sentence.

Further, even under this alternative explanation, I would still not call wa a topic marker. Rather, I would say that when the listener reconstructs the deep-structure, he uses pragmatics to infer what syntactic role the topic plays. Indeed, If you consider a sentences such as ringo wa tabeta and Mary wa tabeta you can see that there is no syntactic way to identify where the topic falls in the deep structure.

Thank you so much for this writeup - it does ring a bell!

Don't worry about references. This is more than enough to get me re-started, dig through my old books/articles and find new ones.

> Further, even under this alternative explanation, I would still not call wa a topic marker.

-> "subject marker"? ;-)

Enjoy your vacation (and maybe pursue a phd)!

1. You can't topicalise the subject of a subordinate clause. 2. Only plain form is valid for verbs in a subordinate clause. (maybe the particle "kara" could be seen as an exception here)

"anata ga kukkii wo taberu" (you eat the cookie)

"anata ga kukkii wo taberu no wo mita yo" (I saw you eat the cookie)

"Anata" is now the subject of a subordinate clause.

I think it can be topicalized (in the sense of becoming the grammatical topic, marked by ha) like this:

"Anata ha, kukkii wo taberu no wo mita yo". (As for you, I saw (you) eat the cookie.)

The subordinate clause now has an unspecified/implicit subject, and that identifies with the topic that was introduced.

I think an extraposition of "anata" (into a non-topical position) is possible like this, where subordinate clause minus its subject is now the topic:

"kukkii wo taberu no wo mita ha, anata da yo". (the one I saw eating the cookie is you)

I feel like this is being shilled too much. I see it everywhere on facebook, reddit japan topics and general.

Of course, generally speaking I am learning Japanese.

Learning Japanese, it's been hard to realize that the most important part of a sentence is always at the end.

Japanese structure is very reminiscent of Forth.

so ? How

Subject, object, etc -- all of the "arguments" -- come before the verb.

Hey, it was a joke - though maybe not a good one, Forth-syntax-wise. Just something I quickly came up with. I actually know Forth a bit. Had played around with it some earlier.

Oh, I see. Sorry. Dense very I was being. :-)

Anyway, good joke or not, I'm glad I made it, because it made me look at your profile and then your blog. Checking out a few of the articles, like Why Lisp? and Lisping at JPL, looks interesting ...


For anyone interested in learning the logical rules that dictate the more confusing parts of Japanese grammar (which were, as others have pointed out, dramatically over-simplified in this article), this is a decent starting point: https://www.tofugu.com/japanese/kobun-reading-introduction/

I've personally been happy with the following books as well: Bungo Manual: Selected Reference Materials for Students of Classical Japanese Classical Japanese Reader and Essential Dictionary Classical Japanese: A Grammar

Many of the confusing rules we have today (i-adjectives vs na-adjectives, different verb conjugation classes, etc.) are subsets of much larger rule-sets from early in the language's written history. Bound particles are probably the most confusing of these rules, which were too convoluted to survive over time, but still inform common usage patterns today.

The Japanese language (particularly its written form) is very young, so it's actually feasible to gain an in-depth understanding of the language's entire history without spending a decade on a doctorate.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact