@Thora

After asking my Japanese friend in every imaginable way, I concluded that -tai becomes -TAi after any verb, accented or not (except in obscure cases, as you mention, where you have a -tai verb followed by the subject of the verb... and which I frankly don't worry too much about at this point).

Yes, the dictionary says it should be -TAI on unaccented verbs, but I've yet to find someone who says it like that. Everytime I tried to say or to elicit -TAI, my friend kept saying it just didn't sound right. I'm not sure what's happening here. If it had simply been a difference between Tokyoben and NHK accent, I would have expected her to say so, but it just didn't seem right to her even when I suggested that possibility.

I also found that you will always get TAkuNAi. Since you can only have one accent per word, you have to consider these as 2 separate words. Some such words are so strong that the regardless of the verb's accent, they will always surface the same.

As a learner, I'm quite satisfied with the conclusion that native speakers say -TAi virtually all the time, and that this is what sounds natural to them. So I know that if I say -TAi, I will sound right, pretty much 99% of the time. It's just annoying that the dictionary says otherwise... I will ask another friend today.

@SomeCallMeChris -- The problem with the article you're refering to is that the general claim is actually that students will sound weird if you introduce pitch markings to them without any other training. But that's the whole problem -- there is no training and a method for teaching pitch needs to be developed.

AlexandreC Wrote:@SomeCallMeChris -- The problem with the article you're refering to is that the general claim is actually that students will sound weird if you introduce pitch markings to them without any other training. But that's the whole problem -- there is no training and a method for teaching pitch needs to be developed.

Regarding this, one of my linguistic teachers -- who calls items like this suprasegmentals -- argue they can't be taught formally since they require such a developped feeling over organs mostly used inconsciously in our every day life. She mentions songs or shadowing as an introductory way towards awareness, but I really think the only way past that is to train your ear (with lots of input), and to self-control your production once your ear is sound. Having natives correct you might be well for a while, but if you don't hear what you utter or don't know how to correct yourself, it's all for naught.

EratiK Wrote:Regarding this, one of my linguistic teachers -- who calls items like this suprasegmentals -- argue they can't be taught formally since they require such a developped feeling over organs mostly used inconsciously in our every day life. She mentions songs or shadowing as an introductory way towards awareness, but I really think the only way past that is to train your ear (with lots of input), and to self-control your production once your ear is sound. Having natives correct you might be well for a while, but if you don't hear what you utter or don't know how to correct yourself, it's all for naught.

When we speak our native language, we use all of our articulatory muscles unconsciously. Yet, we all have to learn to control them when learning a second language. Are new sounds impossible to teach? Are differences between a given sound in two languages impossible to teach?

We could push that reflection a little bit further -- could you formally teach someone to ride a bike? I'm not sure what formally means in this context, but if someone came to you and said "I can't ride a bike, will you help me?", I'm sure you'd be able to do something, and I'm sure it would be better than saying "just keep imitating the other riders" or "sit on your couch and pedal as if you were on a bike".

AlexandreC Wrote:@SomeCallMeChris -- The problem with the article you're refering to is that the general claim is actually that students will sound weird if you introduce pitch markings to them without any other training. But that's the whole problem -- there is no training and a method for teaching pitch needs to be developed.

I was not interested in the political aims of the article, only the facts that they were using to make their case. I'm not interested in arguments about whether or not pitch accent should be learned or when, only in learning it myself.

Others who feel the same might not realize that there is actually some solid experimental evidence in that article that can be helpful to actually understanding pitch accent, and with that understanding have a better chance of training your ear to hear what is actually important.

@Thora -- another friend confirmed that it's always -TAi regardless of the verb's accent.

Thanks for checking. Well, 111 more pages of Tokyo accent rules to test! You might need some new friends... ;p

Maybe we can ignore the rules about ~tai adjectives as well since Vtai and adjectives like 冷たい have the same conjugation and accent. (tai in Vtai is an adjective.) The trend is younger people adding accents to flat adjs. For eg, 冷たい　and 眠たい　are now both [0]or [3]. Shinmeikai states it's still necessary to distinguish which adjs are originally flat, however, b/c their compound accents don't change (eg aKAI MONO, not aKAi mono.) hmm

AlexandreC Wrote:...I concluded that -tai becomes -TAi after any verb, accented or not (except in obscure cases, as you mention, where you have a -tai verb followed by the subject of the verb... and which I frankly don't worry too much about at this point).

Vtai + noun isn't obscure. The noun could be subject or object of the verb, or it could be a noun modified by a clause containing the subject and/or object. eg 友達に会いたいとき... ( naKITAI TOki per dict.) Vtai involves 1st and 2nd person, so it's likely to occur in conversation. It's also likely to be followed by function words which apparently can affect accent.

Quote:So I know that if I say -TAi, I will sound right, pretty much 99% of the time.

Unless it's before nouns (and some function words)?

Quote:I also found that you will always get TAkuNAi. Since you can only have one accent per word, you have to consider these as 2 separate words. Some such words are so strong that the regardless of the verb's accent, they will always surface the same.

Meaning negative of Vtai? Not always. We don't always say shiTAkuNAi, for eg. That's more for emphasis, I think.

It could be separated for emphasis, but sometimes the actual pitch curve of an accent phrase can have 2 peaks. One is the pitch accent (one/phrase after adjusting the underlying accents on words in the phrase) and the other is the phrase initial rise (typically on the 2nd mora). The pitch accent can be higher or lower than the initial rise. Any mora in between slope from one to the other in theory, but the pitch sometimes dips. Longer words tend to gradually slope down and rise before a steep pitch accent fall. Also, maybe く is partially unvoiced or something. ie - Learn from listening. :-)

fwiw, Shinmeikai has egs of separated alternate forms of compounds. Factors include strength of bond, emphasis, age of speaker, length of word, etc. One eg is 読むよ'うだ　→　読'む・よ'うだ. Only if main verb is accented.

[b/]Edit: changed "phrase accent" to "phrase initial rise" to avoid confusion.

For folks trying to pay attention to pitch, since almost all adjs have an accent on next-to-last mora, it might make more sense just to make a note of the few unaccented ones. Here's Shinmeikai's list for that purpose:

[3 morae adjectives:]
[0]:
赤い　　 甘い　　 軽い　　 暗い　　 遠い　　 厚い (暑い、熱い[2]) 重い (思い　noun　[2])

[0][2]
薄い　　　 荒い　　 浅い　　 遅い　　 堅い　　 硬い　 きつい　 煙い　　つらい　 眠い　　 丸い　　

[> 3 morae adjectives:]
Some [0] only, some both [0] and next-to-last:

明るい あぶない いかつい おいしい 重たい 悲しい 黄色い 煙たい
冷たい 眠たい 平たい 優しい 易しい よろしい 難しい くすぐったい

edit: added [ ]

Thora Wrote:It could be separated for emphasis, but sometimes the actual sound curve of pitch phrases can have 2 peaks. One is the pitch accent and the other is the so-called "phrase accent" - typically the 2nd mora. The pitch accent can be higher or lower than the phrase accent.

Since there can't be more than one accent per word, I assumed, ad hoc, they'd have to be separate words. TAkuNAi is pretty consistent from what I've been noticing. I'm not sure what the role of phrase accent is.

Thora Wrote:Also, maybe く is partially unvoiced or something. ie - Learn from listening. :-)

I did consider this. But you get -TAkute, too. The speakers I asked also instinctively thought there were 2 accents (ta and na)...

This information might be interesting for anyone relying on pitch markings after learning about the binary LH system (Tokyo Japanese). The LH system would describe an unaccented 3-morae (kana) word as LHH. If accented on the 2nd mora, it'd be LHL.

* The initial L only appears after a pause. In connected speech, the L becomes H.

* Unaccented words whose initial syllable has a double vowel or an ん start with H, not L.* So they're all HHH. Even after a pause. (eg 電話 でんわ　starts with CVN, 交番　こうばん　starts with CVV.) Many kanji compounds fit this category. They start med-high and slope down slightly, whereas LHH rise from L.

* Unaccented LHHH and final-accented LHHH' don't sound the same. Many descriptions say they're identical unless there's a particle to show the pitch drop. (Someone SRSing would still want to include particles.)

* The pitch isn't actually stepped or binary; it's mostly shallow descending waves with the occasional blip.

This graph shows the 2 unaccented word melodies. (It's shown relative to accented which I'm assuming is [1].)


This one surprised me a bit:


*Note: most descriptions of this raised L on unaccented words say it applies to "heavy" syllables" which means double consonants (学校 gakkou) as well as double vowels and nasal N. The slide show these images are from concluded that it's based on duration of voicing (double vowels and nasal N), but not double consonants.

Thora Wrote:* The initial L only appears after a pause. In connected speech, the L becomes H.

* Unaccented words whose initial syllable has a double vowel or an ん start with H, not L.* So they're all HHH.

* Unaccented LHHH and final-accented LHHH' don't sound the same. Many descriptions say they're identical unless there's a particle to show the pitch drop. (Someone SRSing would still want to include particles.)

* The pitch isn't actually stepped or binary; it's mostly shallow descending waves with the occasional blip.

Interesting points to keep in mind; I remember reading (and testing) the LH becoming HH in certain phonological contexts. I need to test it again. Luckily, it's going to be predictable and doesn't affect the underlying pitch rules.

I created this pdf verb table for pitch. If anyone is interesting in commenting or pointing out errors, I'd be most obliged. I had a native Japanese teacher confirm all these, but a few differ from the dictionary (though very few).

Earlier in the thread, I referenced a FAQ page on pitch I had written at the HTLAL forum.

In order to update the information, I've decided to move it to another location.

http://learnlangs.com/japanesepitch

Much updating still needs to be done, but you are welcome to comment on it or make suggestions.

Romuald Huszcza, Maho Ikushima
Gramatyka japońska
ISBN 83-86483-69-5

Some verb forms.
page 161

ò è à – accented vowels o, e, a
xu stands for -ku, -gu, -mu, -bu, -nu, -su, -tu, -u, -ru (dictionary form)
m – unaccented mora (except for ...mài and masu forms, m is just the phoneme m)
M – accented mora
-ðeba (two last morae, conditional form)
-ðeyo (two last morae = imperative form + emphatic partical よ)
(cV) consonant verb, (vV) vowel verb
... a sequence of morae in a given verb
(Mmxu) etc, verbs like 入るhairu, 帰るkaeru, 参るmairu – accented mora is moved backwards in some forms, one mora to the left


both accented and unaccented verbs
...màsu
...màsita
...masèn
...masèndesita
...masyòu


unaccented verbs
m...xu
m...nai
m...ta
m...te
...òu
...ðèba
...xudaròu
...ròyo (vV); .... ðèyo (cV)
m...aseru
m...tai
...mài
...xuyòuda


accented verbs
...Mxu (Mmxu)
...Mnai
M...ta
M...te
...òu
...Mðeba (Mmðeba)
...Mxudarou (Mmxudarou)
...Mroyo (vV); ...Mðeyo (cV) (Mmðeyo)
...asèru
...tài
...mài
...Mxuyouda (Mmxuyouda)

Thanks for the detailed info Buonaparte, but don't you think my pdf is a little more learner friendly?

AlexandreC Wrote:Thanks for the detailed info Buonaparte, but don't you think my pdf is a little more learner friendly?

Canadians are famous for their learner friendly stuff all over the cyberspace.

The more resources the better, you can always check one against the other. The good thing about the table I posted above is that it is from a source by pofessionals in the field. That's what my granny used when she started to learn Japanese some time ago and she says it was good enough for her. If you don't believe ME, ask her.

With 八重歯 like that, you're granny will do well in Japan.

nadiatims Wrote:With 八重歯 like that, you're granny will do well in Japan.

And the concept of high and low will be that much simpler to grasp.

I am sorry to stay that I am sticking to my pronunciation (accent) dictionary, and communication textbook. They just seem to be much easier to work with. HNK and Sanseido both have dictionaries for were designed for native Japanese speakers, but more straight forward in indicating the correct accent.

NHK日本語発音アクセント辞典 新版
新明解日本語アクセント辞典

コミュニケーションのための日本語発音レッスン

NHK日本語発音アクセント辞典 CD-ROM版 - 動詞 verbs:

I prefer the notation used by 大辞林：

http://www.sanseido-publ.co.jp/publ/dict...in_ac.html

I have the top chart hanging on the wall above my monitor.

CJ

cjon256 Wrote:I prefer the notation used by 大辞林：

http://www.sanseido-publ.co.jp/publ/dict...in_ac.html

I have the top chart hanging on the wall above my monitor.

CJ

I fail to see the usefullness of the top chart.

A noun is either [HL...] or [LH(...)H*L....] , ie. starts low and stays high until an accented mora, unless the accented mora is on the first mora. A few words suffice to sum up the content of the entire chart.

Off-line version of this site: http://tisc.isc.u-toyama.ac.jp/pronunciation/

some missing audio vocab added (some phrases are still missing), .m3u playlists added.
http://www.mediafire.com/?7cl24q5cq57b2b6

@AlexandreC

The chart merely illustrates the notation used by 大辞林. Pretty much every entry is annotated with these codes. I've seen it used elsewhere too, including earlier in this thread.

Also knowing that 水 and 花 are both LH (as per the notation I was contrasting with) is not as useful as knowing that the former is a [0] and the latter is a [2] (although perhaps the other notation captures this somehow?).

水 http://dic.yahoo.co.jp/dsearch?enc=UTF-8...ss&stype=0

花 http://dic.yahoo.co.jp/dsearch?enc=UTF-8...&pagenum=1


CJ

cjon256 Wrote:@AlexandreC

The chart merely illustrates the notation used by 大辞林. Pretty much every entry is annotated with these codes. I've seen it used elsewhere too, including earlier in this thread.

Also knowing that 水 and 花 are both LH (as per the notation I was contrasting with) is not as useful as knowing that the former is a [0] and the latter is a [2] (although perhaps the other notation captures this somehow?).

水 http://dic.yahoo.co.jp/dsearch?enc=UTF-8...ss&stype=0

花 http://dic.yahoo.co.jp/dsearch?enc=UTF-8...&pagenum=1


CJ

Some people contrast LH / LH or else LH / LH* with the same effect.

The reality is that a low percentage of nouns have pitch (I read 30%). Although nouns are the first examples given for pitch, they are the least problematic. On the other hand, a large percentage of verbs and adjectives have pitch, and the pitch changes depending on the various forms, so they are the most problematic.

Romuald Huszcza, Maho Ikushima
Gramatyka japońska
ISBN 83-86483-69-5


N = 名詞 めいし NOUNS
to be learnt by heart
N composed of n morae – n + 1 possible pitch distributions

N + particles
1. N + は　が　に　を　へ　で　と　も　や
から　しか　だけ
no change

2. N + の
2a. N with the last mora accented change to unaccented: 男otokò -> 男のotokono
2b. the rest – no change

3. The majority of two- and three-mora particles (except for から、しか、だけ)
より, さえ, でも, のみ, まで, かしら, すら, とも, だって, など, なり, こそ, では
3a. unaccented N – the first mora of the particle is accented
3b. accented N – no change

4. だけ
4a. either according to 1. or
4b. all N become unaccented
(both are correct)


Surnames (last names) + titles
1. + さんsan, 様　さまsama
no change
2. + 君　くん
2a. accented surnames – no change
2b. unaccented surnames – no change or the last mora of the surname is accented, both are correct
3. + 先生　せんせい sensèi　博士　はくしhàkusi　大使　たいしtàisi
3a. unaccented surnames – the title is accented
3b. accented surnames – no change, but there is a tendency to accent the title (particularly by the younger generation)

N = 名詞 めいし, AN = 形容動詞 けいようどうし + the copula
the copula であるde aru （だ、です、だった, etc）
forms of the copula with -màsu, -masèn, etc, are accented

1. unaccented N, AN – the first mora of the copula is accented (except for だ)
2. accented N, AN – no change

buonaparte Wrote:Romuald Huszcza, Maho Ikushima
Gramatyka japońska
ISBN 83-86483-69-5


N = 名詞 めいし NOUNS
to be learnt by heart
N composed of n morae – n + 1 possible pitch distributions

N + particles
1. N + は　が　に　を　へ　で　と　も　や
から　しか　だけ
no change

2. N + の
2a. N with the last mora accented change to unaccented: 男otokò -> 男のotokono
2b. the rest – no change

3. The majority of two- and three-mora particles (except for から、しか、だけ)
より, さえ, でも, のみ, まで, かしら, すら, とも, だって, など, なり, こそ, では
3a. unaccented N – the first mora of the particle is accented
3b. accented N – no change

4. だけ
4a. either according to 1. or
4b. all N become unaccented
(both are correct)


Surnames (last names) + titles
1. + さんsan, 様　さまsama
no change
2. + 君　くん
2a. accented surnames – no change
2b. unaccented surnames – no change or the last mora of the surname is accented, both are correct
3. + 先生　せんせい sensèi　博士　はくしhàkusi　大使　たいしtàisi
3a. unaccented surnames – the title is accented
3b. accented surnames – no change, but there is a tendency to accent the title (particularly by the younger generation)

N = 名詞 めいし, AN = 形容動詞 けいようどうし + the copula
the copula であるde aru （だ、です、だった, etc）
forms of the copula with -màsu, -masèn, etc, are accented

1. unaccented N, AN – the first mora of the copula is accented (except for だ)
2. accented N, AN – no change

ooh, great info! MORE! MORE!