Discussion: [Article] Japanese Irregular Verbs
From the forum: Sam's Essays
This thread was started by: Glowing Face Man.
Discussion start time: 2009-11-14 16:34:21.
From the forum: Sam's Essays
This thread was started by: Glowing Face Man.
Discussion start time: 2009-11-14 16:34:21.
From: rika, from the old Blog Comments
Date: Nov 18 2008.
Date: Nov 18 2008.
From: komuta, from the old Blog Comments
Date: Apr 27 2009.
Date: Apr 27 2009.
Regarding the naming of the two groups "ichidan" and "godan" : while it is true that in old japanese their were more than two groups of verbs, the number in the naming doesn't have any relation to it. "ichidan" and "godan" respectively means "one row" and "five rows", and was named after the corresponding number of row used in the kana chart by their conjugation. While every ichidan verb form always ends with the same vowel ("i" or "e"), the modern godan verbs use all five vowel of the kana chart in their conjugation, hence the naming.
I agree that Japanese verbs are wonderfully regular; it's one of the many nice simplicities that make Japanese not nearly as hard to learn as people imagine. (See http://www.homejapan.com/2008/02/whats_easy_about_learning_japanese ).
However, I'm going to disagree entirely with one claim, that the copula です (and variants) is vastly different from the English "to be". First, this excerpt:
"In English, the copula is “to be”, and it can express either equality (“The animal IS a cat”), or existence (“There IS a cat”, “I think, therefore I AM”). In Japanese, equality and existence are handled by different words. Existence is itself split into a verb for animate things which exist– いる (iru)– and a verb for inanimate things which exist– ある (aru)."
I'd have to call this misleading. There are two English copulas here, not one. They're contained right in your examples: "is" vs "there is". They're not the same thing; they can't be interchanged freely.
The bigger problem: For what reason would you claim that the copula is a verb in European languages but is not in Japanese? To the extent that this sort of copula (regardless of language) is considered a verb to begin with, there's nothing different about Japanese. Its copula is built on the existential verb ある (indicating existence), but adds で before that to become である (indicating equality). (This way of building one copula from the other even resembles the English pair of copulas somewhat.)
From there, the ある portion of である conjugates exactly the same as the stand-alone verb ある, becoming であります, the formal でござる／でございます (simply replacing ある with the older, more formal ござる), the past tense であった／でありました, and so on. Even the negative construction follows the verb ある slavishly, with the tweak that the added で becomes では: ない becomes ではない, ありません becomes ではありません, なかった becomes ではなかった, and so on.
As perhaps the most-used verb, the copula unsurprisingly gains some special irregular forms, such as だ for である, だった for であった, でした for でありました, じゃない for ではない, and of course です for であります, but again, that doesn't change the verb-like base for construction in any way.
In all of these cases, the forms of both ある and である conjugate like verbs, they function as verbs, and they're used as verbs. They're verb-y through and through, the same as "to be" and "there be" and all the forms thereof, and I submit that there's nothing at all preventing them from being understood with "the tools of English grammar". We can say that ある and "there are" are existential verbs while である and "to be" are copulae, and thus arguably aren't "real" verbs, but my point is that whatever the technical linguistic status of the terms in either language there isn't anything terribly different going in between English and Japanese.
Finally: The regularity of verbs is connected to the use of Chinese characters? I know you're only calling it conjecture, but I'll add that I can't see any support for the idea or its explanations. It all seems predicated on the idea that Chinese writing existed first in Japan, after which the Japanese language later came about, and adapted itself to fit certain aspects of Chinese character usage… which can only leave one scratching his head as to how that worked out on the timeline.
Chinese characters (and writing of any sort) came to Japan rather shockingly late in its history, followed by hiragana centuries later, and the spread of writing throughout the general populace much later still. Is there some evidence that spoken verbs were much more irregular in the past, and underwent a regularization process only after the full writing system was finally developed and promulgated very late in the language's existence?
A final statement:
"In English, a verb, in any conjugated form, is just a homogeneous soup of Roman letters. But in Japanese, verbs have so much more structure, thanks to the kanji-hiragana mingling."
Er, I don't quite follow. Is the above simply noting again that Japanese verbs are more regular than English ones? If so, I don't see any evidence for a role of Chinese characters in that regularity, but otherwise I agree with the observation.
(And if someone were to ask "So, why are Japanese verbs to regular?", I think I'd have to respond with "The real question is, why are English verbs so irregular? How did *that* happen?" : )
Best –
However, I'm going to disagree entirely with one claim, that the copula です (and variants) is vastly different from the English "to be". First, this excerpt:
"In English, the copula is “to be”, and it can express either equality (“The animal IS a cat”), or existence (“There IS a cat”, “I think, therefore I AM”). In Japanese, equality and existence are handled by different words. Existence is itself split into a verb for animate things which exist– いる (iru)– and a verb for inanimate things which exist– ある (aru)."
I'd have to call this misleading. There are two English copulas here, not one. They're contained right in your examples: "is" vs "there is". They're not the same thing; they can't be interchanged freely.
The bigger problem: For what reason would you claim that the copula is a verb in European languages but is not in Japanese? To the extent that this sort of copula (regardless of language) is considered a verb to begin with, there's nothing different about Japanese. Its copula is built on the existential verb ある (indicating existence), but adds で before that to become である (indicating equality). (This way of building one copula from the other even resembles the English pair of copulas somewhat.)
From there, the ある portion of である conjugates exactly the same as the stand-alone verb ある, becoming であります, the formal でござる／でございます (simply replacing ある with the older, more formal ござる), the past tense であった／でありました, and so on. Even the negative construction follows the verb ある slavishly, with the tweak that the added で becomes では: ない becomes ではない, ありません becomes ではありません, なかった becomes ではなかった, and so on.
As perhaps the most-used verb, the copula unsurprisingly gains some special irregular forms, such as だ for である, だった for であった, でした for でありました, じゃない for ではない, and of course です for であります, but again, that doesn't change the verb-like base for construction in any way.
In all of these cases, the forms of both ある and である conjugate like verbs, they function as verbs, and they're used as verbs. They're verb-y through and through, the same as "to be" and "there be" and all the forms thereof, and I submit that there's nothing at all preventing them from being understood with "the tools of English grammar". We can say that ある and "there are" are existential verbs while である and "to be" are copulae, and thus arguably aren't "real" verbs, but my point is that whatever the technical linguistic status of the terms in either language there isn't anything terribly different going in between English and Japanese.
Finally: The regularity of verbs is connected to the use of Chinese characters? I know you're only calling it conjecture, but I'll add that I can't see any support for the idea or its explanations. It all seems predicated on the idea that Chinese writing existed first in Japan, after which the Japanese language later came about, and adapted itself to fit certain aspects of Chinese character usage… which can only leave one scratching his head as to how that worked out on the timeline.
Chinese characters (and writing of any sort) came to Japan rather shockingly late in its history, followed by hiragana centuries later, and the spread of writing throughout the general populace much later still. Is there some evidence that spoken verbs were much more irregular in the past, and underwent a regularization process only after the full writing system was finally developed and promulgated very late in the language's existence?
A final statement:
"In English, a verb, in any conjugated form, is just a homogeneous soup of Roman letters. But in Japanese, verbs have so much more structure, thanks to the kanji-hiragana mingling."
Er, I don't quite follow. Is the above simply noting again that Japanese verbs are more regular than English ones? If so, I don't see any evidence for a role of Chinese characters in that regularity, but otherwise I agree with the observation.
(And if someone were to ask "So, why are Japanese verbs to regular?", I think I'd have to respond with "The real question is, why are English verbs so irregular? How did *that* happen?" : )
Best –
Ugh, I thought I'd fixed these forums to support Unicode. Forgive the mangling of kan(a/ji), the forum is still in "alpha" phase :)
What I meant about Chinese characters having a regularizing effect (and you're right, it's just wild speculation) is that kanji stratifies a conjugated verb into two parts, the kanji and the okurigana. There's no such stratification for English verbs. The consequences of giving the kanji part of the verb an irregular reading are drastic: you introduce a whole new reading to that kanji. So there's a big incentive not to do that. Note that the two highly irregular verbs, suru and kuru, are often written without kanji, which kind of supports the hypothesis. (I suppose you could also introduce irregularity in the okurigana part of the verb, though, so that's kind of a hole in my theory)
Good points all 'round.. as someone pointed out somewhere else, English irregularity probably owes a lot to England being everybody's punching bag for a couple thousand years ;)
What I meant about Chinese characters having a regularizing effect (and you're right, it's just wild speculation) is that kanji stratifies a conjugated verb into two parts, the kanji and the okurigana. There's no such stratification for English verbs. The consequences of giving the kanji part of the verb an irregular reading are drastic: you introduce a whole new reading to that kanji. So there's a big incentive not to do that. Note that the two highly irregular verbs, suru and kuru, are often written without kanji, which kind of supports the hypothesis. (I suppose you could also introduce irregularity in the okurigana part of the verb, though, so that's kind of a hole in my theory)
Good points all 'round.. as someone pointed out somewhere else, English irregularity probably owes a lot to England being everybody's punching bag for a couple thousand years ;)
as someone pointed out somewhere else, English irregularity probably owes a lot to England being everybody's punching bag for a couple thousand years ;)
Huh? That's precisely the opposite of the direction of correlation we find in practice. Languages which see a lot of use by large communities of speakers with different first languages, like English in your example, tend to become morphologically and syntactically [i]simpler[/i], basically since large populations of second-language learners round off a lot of the complicated corners. Languages which are used exclusively in tightly-knit niche communities aren't subject to this force and tend by contrast to develop and retain irregularities. I don't know the optimal citation for this but a quick Google yields [url=http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0008559]this recent paper of Lupyan and Dale[/url], and a [http://blogs.ubc.ca/evolsds/me-tarzan-simple-language-morphology-as-a-feature-of-large-cross-cultural-language-communities/]blog post[/url] discussing it.
What I meant about Chinese characters having a regularizing effect (and you're right, it's just wild speculation)
One dogma current among most linguists is the [i]primacy of speech[/i] (over writing), which holds (loosely) that since we all learn to speak before we learn to write, at the cognitive level language is fundamentally speech and writing is merely a sort of superstructure. Therefore one wouldn't expect to see deep influence like this in the other direction, from writing on spoken forms. If you ask me, nothing in cognition is this absolute, and in modern times now that at last many places have essentially universal literacy we'll probably increasingly see counterexamples to this; even so, I'd say it's right about the usual direction of influence.
Accordingly I don't think your conjecture's likely. The Japanese are hardly averse to kanji having lots of readings. Although I'm not good enough at the language to cite examples, IIRC many kanji even have multiple [i]kun-yomi[/i] which are used in derivationally related words from the same root, except that one includes one more mora than another. It's cross-linguistically quite a common pattern of irregularity that the stem of some word gains an augment in some but not all inflected forms. If Japanese were to come to possess verbs like this, they might just never write the augment and fold it into the reading of the kanji in the forms that have it. Then the written forms would still appear entirely regular. Or they could leave the kana reading invariant, and just have an extra okurigana syllable in certain forms, without breaking any of the principles of the system. (The latter is I reckon more likely; it's actually pretty similar to the difference between ichidan and godan, if you imagine [i]r[/i] to be the ichidan augment.)
Do bear in mind that there are lots of morphological ways for a verb to be "irregular". A suffixal alternation like the above is one. Another is outright suppletion: if Japanese had suppletion they'd probably just use different kanji for the different suppletive stems. Another one, which you seem especially to have in mind considering your English example, is a stem internal modification (like the vowel alternations in English or other IE languages known as [i]ablaut[/i]): it's less easy to think of a slick representation of this in written Japanese, yes. And there are many more.
Hi,
I am French and just like you I have been repelled by languages.
I spent 8 years learning German and don't remember anything at all.
I spent 5 years learning English and at the end of Uni, I could not ask the time in a real situation...
Then I went overseas out of interest. 1 year of studies.
Because of need, and out of frustration for not being able to express myself properly I learned English. I was nearly fluent in 1 month and in 3 months I can say I was owning the language.
Later I flew to Asia. And as of today I spent close to 3 years in China. I speak Chinese almost fluently... And I recently decided to own Chinese characters! Initially my attitude towards it made it impossible - I thought it was impossible.
Now that I know more than 500 characters I have come to a similar realisation as you:
- Chinese characters are a lot smarter than our alphabet. They just make so much more sense. Learning curve sure is steeper, but the way they are constructed not only tell the meaning but also the sound. After the first 500 characters are learned, many other can be deducted.
After I have completely owned Chinese (speaking [70% owneed], reading[40% owned], writing[40%owned]), I want to learn Japanese!
I give myself 1 more year in China to completely master the language!
That comment is a bit too long and does not make much sense at the end, I mostly tell my experience... But it's just that I found your opinion to be coinciding with mine... Which was interesting!
冯涛 ；）
I am French and just like you I have been repelled by languages.
I spent 8 years learning German and don't remember anything at all.
I spent 5 years learning English and at the end of Uni, I could not ask the time in a real situation...
Then I went overseas out of interest. 1 year of studies.
Because of need, and out of frustration for not being able to express myself properly I learned English. I was nearly fluent in 1 month and in 3 months I can say I was owning the language.
Later I flew to Asia. And as of today I spent close to 3 years in China. I speak Chinese almost fluently... And I recently decided to own Chinese characters! Initially my attitude towards it made it impossible - I thought it was impossible.
Now that I know more than 500 characters I have come to a similar realisation as you:
- Chinese characters are a lot smarter than our alphabet. They just make so much more sense. Learning curve sure is steeper, but the way they are constructed not only tell the meaning but also the sound. After the first 500 characters are learned, many other can be deducted.
After I have completely owned Chinese (speaking [70% owneed], reading[40% owned], writing[40%owned]), I want to learn Japanese!
I give myself 1 more year in China to completely master the language!
That comment is a bit too long and does not make much sense at the end, I mostly tell my experience... But it's just that I found your opinion to be coinciding with mine... Which was interesting!
冯涛 ；）
I was googling and I came across your article. I know why Japanese is so regular.
Japanese is what is called an agglutinative language. Agglutinative languages are known for primarily expressing grammatical features by using suffixes and prefixes, e.g. the Japanese word ikō can be glossed as ik-ō, go-VOLITIONAL, "shall we go?".
One interesting property of agglutinative languages is that they all tend to be very regular. Turkish, for example, has only 1 irregular verb.
Thus, Japanese's regularity is because of how the grammar works; it's not due to a property unique to the Japanese, writing or non-writing. (Also see the post from Alex Fink on how speech has primacy in languages)
Japanese is what is called an agglutinative language. Agglutinative languages are known for primarily expressing grammatical features by using suffixes and prefixes, e.g. the Japanese word ikō can be glossed as ik-ō, go-VOLITIONAL, "shall we go?".
One interesting property of agglutinative languages is that they all tend to be very regular. Turkish, for example, has only 1 irregular verb.
Thus, Japanese's regularity is because of how the grammar works; it's not due to a property unique to the Japanese, writing or non-writing. (Also see the post from Alex Fink on how speech has primacy in languages)
Hummm... that sounds plausible, but isn't German agglutinative? (And even English, to an extent... if you don't believe me, ask the antidisestablishmentarianists!) I guess one could say that agglutinativeness is not a binary property but rather, a spectrum. Is it true that Japanese is more agglutinative than German? It's hard to say for sure-- the lack of a space in Japanese makes it ambiguous when you have two words and when you have one... Anyway, definitely an interesting idea and you might be onto something. Isn't Mandarin also highly agglutinative? And sure enough, Mandarin is quite regular
No, there are two kinds of agglutination, one for word formation, the other for indicating grammatical features. A language is considered agglutinative when the latter type of agglutination is
English and Mandarin use some agglutination for deriving words, but use very little in their grammar (the main exceptions being -ed and -s in English). They're what are called isolating languages, each word tends to contain few components (the technical definition being that they a low morpheme-per-word ratio).
Languages like German and Latin use more agglutinative features in their grammar, but of a different type than agglutinative languages, e.g., the word "bonus" in Latin (meaning "good) has -us at the end, indicating the nominative cause, masculinity, and singular number. In an agglutinative language, all of that would be indicated by separate affixes. Languages like German thus fall into their own class, fusional languages.
Languages like Japanese and Finnish use agglutination extensively, most words are formed out of multiple components (high morpheme-per-word ratio). As noted above, they also express each grammatical feature using a different affix, instead of combining them together.
Polysynthetic languages like Classical Ainu and Cherokee take agglutination to an extreme, where what would be an entire sentence in another language is one word in a polysynthetic language.
You're right about it being a continuum; the division between fusional and agglutinative languages is a bit blurred -- Japanese is less agglutinative than Finnish, for example.
English and Mandarin use some agglutination for deriving words, but use very little in their grammar (the main exceptions being -ed and -s in English). They're what are called isolating languages, each word tends to contain few components (the technical definition being that they a low morpheme-per-word ratio).
Languages like German and Latin use more agglutinative features in their grammar, but of a different type than agglutinative languages, e.g., the word "bonus" in Latin (meaning "good) has -us at the end, indicating the nominative cause, masculinity, and singular number. In an agglutinative language, all of that would be indicated by separate affixes. Languages like German thus fall into their own class, fusional languages.
Languages like Japanese and Finnish use agglutination extensively, most words are formed out of multiple components (high morpheme-per-word ratio). As noted above, they also express each grammatical feature using a different affix, instead of combining them together.
Polysynthetic languages like Classical Ainu and Cherokee take agglutination to an extreme, where what would be an entire sentence in another language is one word in a polysynthetic language.
You're right about it being a continuum; the division between fusional and agglutinative languages is a bit blurred -- Japanese is less agglutinative than Finnish, for example.