I understand the concept of pitch accent, morae etc, I simply do not know what the numbers mean.
I remember mentioning it in a thread a while back but I can't seem to find it anywhere.

I'll take the first couple examples from core:

一つ　　 2

一　　 2

一般　　0

これ 　0;1

に　 1

二つ　3

If someone could give me a quick explanation, it would be appreciated. Thanks

The number indicates after which mora the pitch is reduced.

- If the number is 0, the first mora is LOW pitched while the rest is HIGH. There is no reduction in pitch.
- If the number is 1, the first mora is HIGH pitched while the rest is LOW.
- If the number is >1, the first mora is LOW, then the following morae up to and including the nth mora as indicated by the number are HIGH. The rest of the morae after that are LOW pitched.
eg. 一つ [hi.to.tsu] = LHL
に [ni] = H
一般 [i.ppa.n] = LHH

Where there are multiple numbers it means that the word can be read using any of the numbers as the pitch indicator. The number listed first is most common.
eg. これ [ko.re] 0 = L-H or 1 = H-L

Lastly in the case where the number is the last mora of the word, the first mora of the next word would be LOW. (someone correct me on this one, I may be wrong)
eg. 一 [ic.hi] = LH (the first mora of the word that follows 一 would be LOW pitched)


I'm not that great at explaining things so I'll also link you to the relevant wikipedia page:
http://en.wikipedia.org/wiki/Japanese_pi...nary_pitch

Perfect! ありがとうございました

What about, for example, if the accent number is 3 or something? does that mean it's LLHH or HHHL or LLHL?

ktcgx Wrote:What about, for example, if the accent number is 3 or something? does that mean it's LLHH or HHHL or LLHL?

So that's a case where the number is >1. The pitch accent will be LHHL (for a word with four morae).
If it were a word with 5 morae it would be LHHLL.
6 morae: LHHLLL etc.

Things to remember are:
- The accent number is number of mora to count before the pitch goes from high to low.
- First mora is low unless the accent number is 1, in which case the first mora is high.

MindTrick Wrote:

ktcgx Wrote:What about, for example, if the accent number is 3 or something? does that mean it's LLHH or HHHL or LLHL?

So that's a case where the number is >1. The pitch accent will be LHHL (for a word with four morae).
If it were a word with 5 morae it would be LHHLL.
6 morae: LHHLLL etc.

Things to remember are:
- The accent number is number of mora to count before the pitch goes from high to low.
- First mora is low unless the accent number is 1, in which case the first mora is high.

Thank you! Today I'm at my school where the Kyoto sensei makes fun of my not being able to tell the differences between pitches, or hear the difference, for that matter... The other day he was nearly in tears because apparently I was trying to tell him that it looked like lollies would be falling from the sky soon... T_T

You can also rarely have something like a pitch-accent of 4 for a 3 mora word or the like, which is because particles attached to the word follow the pitch of the word.　It doesn't happen much, but if you see a pitch accent larger than the word is long, that's how that works.

Normally pitch drops to the following particle or in the middle of the word (quite often on the last mora) so that following particles are usually L outside unless the word was pitch-0, in which case, of course, they are H.

You know what? Ignore what I wrote, just read this page:
http://learnlangs.com/japanesepitch/index.php/Main_Page

I personally find the number system for pitch accents very difficult to remember.

There's a great Japanese website for looking up the pitch accent of common Japanese words, using lines under and over the syllables to represent the high and low tones:

日本語教育用アクセント辞典

http://accent.u-biq.org/

I also have a couple of physical dictionaries with the pitch patterns, but it's so much easier to look them up using this website.

tom_b Wrote:I personally find the number system for pitch accents very difficult to remember.
There's a great Japanese website for looking up the pitch accent of common Japanese words, using lines under and over the syllables to represent the high and low tones:
日本語教育用アクセント辞典
http://accent.u-biq.org/
I also have a couple of physical dictionaries with the pitch patterns, but it's so much easier to look them up using this website.

It's not that easy without a search bar though...

tom_b Wrote:I personally find the number system for pitch accents very difficult to remember.

I basically wrote down what the numbers mean on a sticky note, and stuck it on the wall next to my computer. After glancing at it over the course of a couple days I memorized it without even trying.

The way I remember them (in case it can help) :

- "0" is a misleading choice. Just take it to mean "infinity" (1/0 !)

- General case ("n") : start Low (*if possible*), then immediately High until mora "n", at which point you drop to Low.

This takes care of
• "0" (= infinity) : start Low (one mora), go to High and never go down (including following particle)
• "1" : start High (no space to start Low since you need to go Low after first mora), then immediately Low till the end. This is the only case where the first mora gets High pitch.

Does this seem to work?

jmignot Wrote:• "0" (= infinity) : start Low (one mora), go to High and never go down (including following particle)
• "1" : start High (no space to start Low since you need to go Low after first mora), then immediately Low till the end. This is the only case where the first mora gets High pitch.

Does this seem to work?

That's right if the word is located in the head of a phrase.
The phenomenon of the first mora being kept lower and the rest raised up occurs per phrase, not per word. As long as a phrase continues, pitch keeps down as it runs into accent kernels, whose location is indicated by the number. Once a phrase shifts to another, pitch can regain.

いまの１ ひと０ (a person who has just passed me by)
.い い と
. ま の ひ と ま の、ひ

いまの１ ねこ１
。い い ね
。 まのね まの、 こ
。 こ

On the other hand, a word can be emphasized. In that case, accent kernels are highlighted, in other words, moras before the accent kernel are also kept lower like this.
. か
.(あ) た た い
. あ

viharati Wrote:That's right if the word is located in the head of a phrase.
The phenomenon of the first mora being kept lower and the rest raised up occurs per phrase, not per word. As long as a phrase continues, pitch keeps down as it runs into accent kernels, whose location is indicated by the number. Once a phrase shifts to another, pitch can regain.

いまの１ ひと０ (a person who has just passed me by)
.い い と
. ま の ひ と ま の、ひ

いまの１ ねこ１
。い い ね
。 まのね まの、 こ
。 こ

On the other hand, a word can be emphasized. In that case, accent kernels are highlighted, in other words, moras before the accent kernel are also kept lower like this.
. か
.(あ) た た い
. あ

Thanks for correcting me.
Now I am a bit frustrated because, while I understand that there is more to pitch rules than applying them on a by-word basis, I was not able to fully understand your explanation.
Perhaps I need first to know the meaning of "accent kernel"…
Where can I find more detailed material on this subject?

Quote:"accent kernel"

Maybe, I should have said "accented mora" instead.

Quote:Where can I find more detailed material on this subject?

If it's OK to be written in Japanese.
http://www5a.biglobe.ne.jp/~accent/accent.htm
The point is
i) "Accent" stands for fall of pitch in the next mora, which is unique to each word.
Accent marker 0 represents the word has no accented morae, 1 represents the first mora is accented. As a result, 0-marked words are pronunced in level pitch (not physical but linguistic sense) in sequence of a phrase while 1-marked words have a down-step in the second mora and the rest are level (or lower than the part before the accented mora) until it runs into another "accented mora" in sequence of a phrase.

ii) "Prominence" stands for an obligatory intonation, which occurs per phrase.
When a phrase starts with a 0-marked word, pitch is raised up in the second mora of the word and so on. (In this case the first mora can either be lowered or level than the last mora of the previous word.)
With a 1-marked word, pitch is raised up in the first mora then it falls down in the next.

In other words, when rise of pitch occurs, it's perceived as change of phrases (in order to enunciate words etc.)


iii) These are features of New Tokyo dialect, which is the virtual standard.

viharati Wrote:If it's OK to be written in Japanese.
http://www5a.biglobe.ne.jp/~accent/accent.htm
The point is
i) "Accent" stands for fall of pitch in the next mora, which is unique to each word.
Accent marker 0 represents the word has no accented morae, 1 represents the first mora is accented. As a result, 0-marked words are pronunced in level pitch (not physical but linguistic sense) in sequence of a phrase while 1-marked words have a down-step in the second mora and the rest are level (or lower than the part before the accented mora) until it runs into another "accented mora" in sequence of a phrase.

ii) "Prominence" stands for an obligatory intonation, which occurs per phrase.
When a phrase starts with a 0-marked word, pitch is raised up in the second mora of the word and so on. (In this case the first mora can either be lowered or level than the last mora of the previous word.)
With a 1-marked word, pitch is raised up in the first mora then it falls down in the next.

In other words, when rise of pitch occurs, it's perceived as change of phrases (in order to enunciate words etc.)


iii) These are features of New Tokyo dialect, which is the virtual standard.

Many thanks for this enlightening explanation. Now I see that I missed an important part of that story.

So, if I understand correctly, the only fixed point is the pitch drop occurring after the accented mora (indicated by "n") — or the lack of such drop in unaccented ("0") words.
[at this point the number convention appears to make sense!]
What occurs *before* the accented mora depends on the position of the word in the phrase.
According to your examples:
- The pich rise on the second mora of accentless words occurs only if they are located at the beginning of a phrase (after period or comma). Otherwise the pitch remains at its previous level.
- The same applies to the pitch rise before the first mora of "1"-type words.
- Further changes occur if special emphasis is placed on a word.

A couple points are still unclear to me (sorry but reading the reference in Japanese using Rikaichan would cost me my entire Christmas vacation…)
-How about words of type "2", "3", etc. ? Do those words necessarily have both the pitch rise after the first mora and the drop after the accented one, or is it also somehow context dependent?
- From your description, there can be 3 different pitch levels in one phrase. But if I assume a longer sentence with particular combinations of words, and I apply the same rules, it seems that the number of levels can be even larger. For instance a succession of "1"-type words (if this is ever possible in a real sentence) would lead to a pitch going down by successive steps.
Is this correct?

Quote:-How about words of type "2", "3", etc. ? Do those words necessarily have both the pitch rise after the first mora and the drop after the accented one, or is it also somehow context dependent?

たのしい3 クリスマス3
. の し ｜ り す
. た い ｜ く ま す

. の し
. た い く り す
. ま す
(More often than not, it's probablly pronounced separately, though)
In the beginning, the second mora is raised up and it keeps that level until it reaches to the accented mora（し）, then dropps and keeps that level until it reaches to another accented mora (the third mora of くりすます).

Quote:From your description, there can be 3 different pitch levels in one phrase. But if I assume a longer sentence with particular combinations of words, and I apply the same rules, it seems that the number of levels can be even larger. For instance a succession of "1"-type words (if this is ever possible in a real sentence) would lead to a pitch going down by successive steps.
Is this correct?

Yes. But of course too many down-steps in one phrase are virtually meaningless, it winds up to be a murmur.

This sounds easy to understand, if not to apply in actual speech !
It will take me a lot of practice.
Thanks again.