In standard dialect:
橋 HAsi
箸 haSI

These two are reversed in Kyoto dialect I think. It gets more interesting when you add particles too.
See http://en.wikipedia.org/wiki/Japanese_pitch_accent

Last edited by uberstuber (2009 August 21, 5:21 pm)

Just listening doesn't work for me, I don't notice it. The only I can finally remember is 一杯 as in "one cup/glass" vs "lots of, full".

I know this is a good question and that pitch accent is important to sound natural, but if we aren't always using the right pitch accent, or forget it, I don't think it is a major major thing.

When speaking Japanese, people are going to know we mean chopsticks or bridge just from the context.

I get misunderstood several times a day due to incorrect pitch accent.

It's always with words I learnt via reading rather than listening. When reading a new word, I tend to assign it an accent of my own making.

The problem is that once my mind has assigned a certain pitch accent to a word it's difficult to shake off, even though I later hear the "correct" accent many times.

The best solution, I suppose, is to always learn the correct accent immediately. Now that WWWJDIC has added pronunciations this has become feasible (~ish). Saiga can be useful too.

Of course it is important if you are aiming to get a perfect native accent, but I think there are much more important things to worry about. It's a lot better to have bad pitch than bad grammar and a small lexicon. I think in this case it's a lot better to just pick it up passively. Active study of it hurts more than it helps in my experience.

I don't know which parts of that were directed at my posts, but..

What I mean to say, in short, is that it is important to learn pitch accent, but you shouldn't study pitch accent. Just try to make your pronunciation match a native without consciously thinking about and memorizing LOW-HIGH-LOW-LOW-HIGH-LOW-MEDIUM-DIAGONAL-UTURN-5mhz.

Last edited by Jarvik7 (2009 August 22, 4:36 am)

Jarvik7 wrote:

One thing to watch out for though is stress accent. English has stress accent, Japanese doesn't. Adding in stress to Japanese is the most obvious marker of a (English-native) gaijin accent.

I've heard some native English speakers doing that and it sounds pretty bad. I don't have to worry about that since my native language is French and I've read that the accent is nearly the same as Japanese for unaccented words (I personally never realized it because to me that's just the natural way to speak).

I guess I just have to try to pay more attention to words with a pitch drop in them too.

mmh.. I have trouble hearing pitch accent.. Does anyone know of a way of "training" it?

I mean.. I think I can here it, but it's so fast, how am I supposed to pay attention to it?

bombpersons wrote:

mmh.. I have trouble hearing pitch accent.. Does anyone know of a way of "training" it?

I mean.. I think I can here it, but it's so fast, how am I supposed to pay attention to it?

Pitch isn't recognized mora-by-mora when you hear words spoken at natural speed. Here's a direct quote from an article on Wikipedia:

Wikipedia wrote:

* When a human hears a fast rhythm, typically faster than 330 milliseconds (ms) per beat, the series of beats is heard as one solid noise. For example, a human can imitate a machine gun sound, but hardly count its beats.
    * Conversely, when a slow rhythm is heard, typically slower than 450 ms per beat, each beat is separately understood. The speed of a slow rhythm can be controlled beat by beat, such as hand clapping in music.

The Japanese "beat" is said to be at approximately 145 ms per mora, so you will hear a word as one solid sound. If someone plays the guitar at an amazingly fast speed, you hear the melody and recognize the music, but can't recognize each note as one musical note. The same goes for Japanese moras. You feel the pitch of a word as one solid sound like a chunk of notes forming a melody. (If you didn't know, "mora" roughly means one spoken kana, so the word 私 (わたし) has three moras: わ, た, and し.)

If you have never heard of isochrony of language, you might want to read a few articles about it such as this http://en.wikipedia.org/wiki/Isochrony I think you can find more detailed information for learners if you google for "mora timing" and "Japanese." I don't know if understanding isochrony would be necessary to grasp the Japanese beat, but getting used to the Japanese beat of moras definitely helps you catch the pitch accent and stuff. If you already listened to tons of natural speech and still don't get the hang of it, learning about isochrony may help you pay attention to it.

I also heard there were several methods to improve your intonation and other pitch related aspects of your second language(s). If a learning method can improve your pitch accent, it probably improves your listening skills the same way. It seems shadowing is one of the most popular methods when it comes to the melody of language. I heard professional translators also used the technique to improve their skills, so it might be worth giving a try.

On a side note, I think why the AJATT guy recommends you learn kanji and listen to fun stuff regardless of whether you understand it before ever learning written language grammar is because that way you can improve your listening skills of intonation, pitch, and phonemes, and avoid developing a foreign accent. If you read text before you can hear pitch and stuff, you hear Japanese with your own thick gaijin accent in your head, which definitely internalize the foreign accent in your mind. So when you hear a simple sentence spoken by a native speaker, you may not be able to understand it because you expect a totally different sound you deeply ingrained in your mind by yourself by reading native materials in a gaijin accent.

I know how difficult to get rid of the weird accent in your mind developed by reading text in a foreign accent. We are FORCED to read and speak English in katakana at school by English teachers who can't speak English at all. The textbooks were written in the Latin alphabet, but we read it in katakana, i.e., we heard our own voice speaking English using Japanese sounds and intonation. It becomes almost impossible to lose your foreign accent in your head if you ingrain it too deeply...

If your listening is still very weak, I can't recommend you read anything in Japanese. If you think listening to butchered gaijin-ish Japanese by other foreigners hurts your listening skills and accent, you may not want to listen to your own bizarre Japanese spoken in your head while reading books for the very same reason.

Last edited by magamo (2009 August 22, 10:06 am)

bombpersons wrote:

Ha, "Butchered gaijin accent" For some reason I think the voice in my head sounds much better than my real voice. Maybe it's because of all the listening.

When I'm watching something, I'm gonna try and listen to the pitch carefully and see if I can notice it more.

Also, I think I might give shadowing a try.. Is there an easy way to do it? Do you just pick a random show and try and copy the speech? Or maybe using something like subs2srs might be useful?

We've been butchering English by japaneesu engrishu, so I don't know how well shadowing works to improve speaking skills, but it seems a lot of English learners find shadowing effective in improving listening skills. I guess it would work the same way for Japanese learners, too.

Anyway, these two methods worked well for me:

1. Use a PC and slow down the audio material to the extent that you can definitely hear every detail and perfectly follow it while shadowing.
2. Shadow the same material at natural speed plenty of times so that you half-remember every line and can "sync" with the speaker, i.e., try to speak a millisecond ahead of the speaker with the exact same intonation.

The first method helps you realize subtle pitch/tone/whatever of audio materials you missed when you heard them at natural speed. The second method gives you immediate feedback when you shadow a native speaker because you're 100% synchronizing with the audio; if you're following the speaker with a millisecond delay, it'll be harder to be corrected by the audio while shadowing. It seems to me that trying to speak a millisecond ahead of a native speaker while mimicking intonation is more effective than trying to catch up with the audio when it comes to listening skills. This is kind of a rougher shadowing than the typical immediate following method, so I think it'd be better to do it in conjunction with the slowing down method to polish your pitch and stuff and fix rougher points even if you think you can properly shadow an audio at natural speed. If you shadow to reduce your accent, the usual immediate following method might be better, but it seems letting the audio follow you is better to improve listening skills for some reason.

As for the materials to shadow, I think you should use natural native stuff with enough context such as dialogues in movies, anime, etc. The melody of language like pitch and intonation does have meanings like words and phrases, but you can't look in a dictionary because there is no such thing as an intonation dictionary.

I don't think it would work very well to shadow an isolated example sentence taken from a textbook if you don't understand its nuance given by prosody. It's like trying to understand a word taken in isolation by pronouncing it repeatedly without using a dictionary. I think it's kind of obvious you need context to understand the meaning of a pitch patter. So I don't think a phrase book listing isolated sentences will greatly improve learners' prosody if it has a CD. Native speakers' voices don't help much if it doesn't teach the meanings of the pitch patterns used in its example sentences. It's as if showing a word without giving its definition or any context. What you can learn is its spelling and nothing more, i.e., you may become able to mimic a pitch pattern but will never know its meaning and when to use it.