+- kanji koohii FORUM (http://forum.koohii.com)
+-- Forum: Learning Japanese (http://forum.koohii.com/forum-4.html)
+--- Forum: Learning resources (http://forum.koohii.com/forum-9.html)
+--- Thread: Technique to improve listening and speaking (/thread-8742.html)
Technique to improve listening and speaking - magamo - 2012-03-23
SomeCallMeChris Wrote:Now, I don't have a pitch accent dictionary (yet, though it's quickly going up in my priorities, and pitch accents are marked in 大辞林 but nothing else about pronunciations is... I don't think.) so, I don't know if a pitch accent dictionary would also mark which vowels are devoiced. I've only seen devoiced vowels marked in language learning resources for foreigners (I think I might have a romaji dictionary buried somewhere that marks devoiced vowels and pitch accent, or at least I've leafed through one at one point.) I've never seen any marking (or description of) these other allophonic variations that magamo refers to. I'd love to have such a dictionary. Although, if one's ear is well enough trained the sound of an electronic version of an accent dictionary should be sufficient. (Or possibly the EDICT sound files... although those are Tokyo accent not standard, I think?)I don't know if there is a Japanese dictionary or some kind of learning resource that marks allophones and other detailed phonetic information. But most likely there is none.
One of the reasons is that rules for devoicing are so complicated linguists are still trying hard to figure them out. If some genius finds the perfect description, it wouldn't be included in any learning material. For example, recent research on vowel devoicing describes the phenomenon not by "when it occurs" but by "how frequently it happens in a given phonetic/phonological context." We can only tell how many percent a certain vowel gets devoiced in a specific word in a particular sentence. Of course, there must be a set of golden rules that quite accurately describes when it happens and when it doesn't without relying too much on percentage. But this remains out of reach of linguists.
Another reason is that it's the standard among textbooks and dictionaries not to include irrelevant details. This is because, for example, whatever allophone you use, it doesn't change words or phonemes. For instance, う is う no matter how you articulate it as long as it's perceived as う by native speakers (Well, this sentence is painfully soooo captain obvious, but I think you get the drift.). Learning material for foreigners isn't for people who want to achieve native-like accents. Dictionaries and such for native speakers are for people who already speak with native accents. There must be some useful material for (aspiring) voice professionals like actors and news anchors. But probably that's something you need only when you're already really good at pronunciation.
Anyway, this wikipedia article explains simple and very useful rules about vowel devoicing. If I may add a few little things to it, vowel devoicing occurs far more frequently on non-accented moras. As you probably already know, the mora that immediately precedes a pitch drop is felt accented by native speakers in standard Japanese (This is the Japanese pitch accent.) So this accented point is usually voiced while other moras in a word will likely get devoiced if they meet the criteria described in the wikipedia article. Also, pretty much no learning material or casual articles like wikipedia etc. mention this, but there are two types of devoicing: one is pronouncing vowels without vocal cord vibration like a whisper and the other is skipping vowels entirely. There is a degree for the former kind too, so not all devoiced vowels of this kind are equally voiceless. Some may only be partially devoiced.
Ah, and one more thing about the article. It says the speaker's gender plays a role in Japanese pronunciation. I'm no expert on this, but you might want to take it with a gigantic chunk of salt. I have seen such claims a few times for different phonetic features such as how males and females are different in terms of voicing onsets, but I kind of feel like they should be taken cum grano salis en masse.
Either way, the linked wikipedia article has fairly detailed information on other kinds of allophones too. I recommend you read through it from the top to the bottom if you're interested in Japanese phonology. It's not complete, but probably more than enough for a beginner. Note that the article may not make any sense to you if you haven't studied phonetics and phonology in general. If you want a bit more self-explained explanations with less jargon, I searched my own posts about Japanese pronunciation on this forum. I'm pretty sure I missed many related posts, but those I found are mostly written so learners with little knowledge of phonetics can understand, I think. Also, you should be careful because some of them were written ages ago, so there may be factual errors. Like I said, I'm no expert. And if they're confusing, probably that's because of my poor English; I'm practicing English by posting here.
What is an allophone?: Two kinds of Japanese J sound
Slightly shorter post in the same thread as above
Allophones and a little bit about Japanese N
The 6 types of Japanese N (You can safely ignore the jargon at the end of this post. Google them if you're curious.)
Japanese N doesn't behave like the English equivalent in a syllable
Thread about the difference between English and Japanese ch's. This also has posts about allophones of h, p, b and some other details on Japanese sounds
One thing I want to emphasize is that those minor details like allophones are reeeeeally minor unless you desperately want perfect pronunciation. As I wrote in the above posts, English also has tons of allophones. But if you're a native English speaker, probably you'd think they're super minor, e.g., how "p"s in "pit" and "spit" are different sounds. Native speakers often don't even realize there are multiple sounds for English "p".
Since we're talking about devoiced vowels, here's a fact you might not know: English also has vowel devoicing. If you're a native speaker, you're also devoicing vowels when speaking English, though in a different way than a native Japanese speaker does in Japanese.
If you want to see if you're devoicing vowels in English, here's how:
1. Touch your throat so you can feel vibration of your vocal cord while talking.
2. To make sure your hand is on the right spot, say T (i.e., tee) and D, and see if you can tell the difference. You should feel the vibration from the start when you say D, but you only feel it for the "ee" part for T.
3. Say "peculiar." You should notice that your vocal cord doesn't vibrate until you hit the u in the middle, i.e., the first vowel (written as "e") is devoiced.
I'm not sure if this works for all English dialects with large numbers of speakers, but it should if you speak American English with a fairly neutral accent. Unfortunately, the rules about when to devoice vowels seem to be different between English and Japanese. But this fact may make devoiced vowels look less exotic to you.
SomeCallMeChris Wrote:As an aside, I'd also love to get ahold of regional accent resources for other regions, but I don't think there are accent dictionaries for other accents (although I haven't actually researched that.)As I said in another post, pitch accent is THE difference. Certainly some phonemes are pronounced differently, but they're uber-minor. They're as minor as devoiced vowels in English. Your average Japanese guy wouldn't know there is a difference other than pitch patterns and grammar/vocabulary. If you google in Japanese, I think you can find some useful websites for regional differences in pitch accent. There might be some dictionaries, I guess.
So, the most important thing is pitch patterns when it comes to regional accents. I'm sure I made a post or two about how pitch patters differ from region to region. But here is the point:
It is believed that all the regional pitch variations come from the same one dialect of classical Japanese. As far as pitch accent goes, every dialect is a reduced version of an older version of the Kansai dialect. So all dialects share a common feature that all words are classified into several groups. Every word in a given group gets the same pitch pattern within a dialect. So, let's say, a dialect X has 4 groups A, B, C, and D. Then all words in A are of the same pitch pattern, which is different from other groups'.
Now take another dialect Y, which also has 4 groups. How are words grouped in this dialect? The answer is that it's pretty much the same as in dialect X. If two words are of the same group in dialect X, then they most likely are of the same group in Y as well. So basically you can use the same 4 labels A, B, C, and D. It's just the pitch pattern assigned to each group is different between dialects X and Y. If you speak dialect X and learn the pitch pattern of a certain word from group A in dialect Y, you can fairly accurately guess the correct pitch patterns of all the rest of the words in group A because most likely they're all the same. It's kind of similar to how two English words that rhyme with each other tend to rhyme in another dialect too.
There are exceptions. For example, very young words which became part of Japanese vocabulary after complete separation of dialects X and Y may fall into the different categories. Some words might have changed their groups only in one dialect, though they're rare. Also, a newer dialect tends to merge two groups into one. So if you learn a dialect with a reduced number of groups first, you might have harder time than those who first learned the Kansai dialect, which is the father of all the regional dialects and has the most complicated pitch accent system.
One useful tip is that sometimes differences are regular. For example, the accent position of many groups in standard Japanese are often different only by one mora from that of the Kansai dialect. So, if you speak standard Japanese, then by shifting the pitch drop point forward by one mora, your Japanese suddenly gets a Kansai feel to it. Not all groups are like this, so you should fix pitch patterns for some groups by listening to several sample words. Words beloning to merged groups should be relearned too if you want something more than a fake accent. But I think this technique can help, especially when you want to learn a minor dialect whose pitch patterns are not well documented.
Wow, you've read this loooong post this much? Then here's some things that I guess you haven't thought about but may interest you because you're interested in accent. The first thing is the extremely common but wrong assumption that native speakers are better at discriminating different sounds within a phoneme. This is not the case. At all. Actually it's the opposite. Nonnative speakers are better at perceiving differences within a phoneme. This is a well-known fact called the perceptual magnet effect in linguistics.
For example, it is (in)famous that the Japanese can't tell the difference between l and r. This is true in a sense. But Japanese speakers do have an acute within-phoneme distinction ability. And research has shown that it's even better than native English speakers. So, your average Japanese guy can hear a slight difference between very precisely articulated r and slightly "off" r even when native English speakers can't even hear the difference. What those nonnative speakers typically can't tell is where the boundary between l and r is. The Japanese hear r's and l's more precisely than native English speakers do, but that doesn't help if they don't know from where a sound is perceived as r (or l) by native English speakers.
This magnet effect appears to be universal. It's not only Japanese guys or only between l and r. So, in a sense, you already have a better listening ability than native speakers when it comes to sounds your native language doesn't have.
What you need to learn is the boundaries. You're not supposed to "sharpen" your perception. You're supposed to draw a new mental circle that represents a phoneme. What native speakers are really awesome at is recognizing whether a sound falls into a specific category or not. This ability, you need. Some people say adults lose the ability to hear a foreign sound. They don't. They just don't know the categories used in a foreign language for obvious reasons.
The other thing you should consider is that we all have an acoustic version of color-blindness. Who cares if p's in "pin" and "spin" are different? Well, to speakers of some languages, this difference is too clear to fail to notice. But native English speakers typically find it difficult to hear the difference, though they always use the two sounds correctly themselves. If you feel cheated because these are allophones, which, by definition, are harder for a native speaker to notice the difference of, the perceptual magnet effect within categories has been demonstrated in many researches for various phonemes too.
So, it's like native speakers of two different languages have different kinds of color-blindness. You can't see the difference between red and green, and I can't tell if it's blue or yellow. This means that it can be pretty much pointless to perfect your pronunciation just because you hear a difference.
Let's say, I ask you to make a copy of a painting as accurately as possible. And you perfect the shade of blue and yellow while using red and green at random thinking they're the same color. Now you finished painting. It looks like an exact copy to you. Surely it is when it comes to the balance between blue and yellow, though I wouldn't even notice if you used only blue. But your seemingly perfect copy is awful and extremely poor to me because you used red when it should be green. To me, it's more than what the deuce. And I say I'll show you how to paint. And the result is a copy that is horrible as ***** to your eye when I don't know where I got wrong.
Anyway, what I'm trying to tell you by this poor analogy is that you might want to focus on the kind of difference Japanese people immediately notice, not the kind of difference you clearly hear. Pitch accent IS what the Japanese notice. Focusing on vowel qualities and such is like perfecting the greenness of a green apple when you're judged by red-green color-blinds. A red apple works just fine.
Technique to improve listening and speaking - vileru - 2012-03-23
@magamo
You're delusional if you think you have "poor English." Honestly, you're writing is better than most American university graduates. You could even be published if you were so inclined.
Anyway, thanks again for another excellent post. Please continue to help us.
P.S. I will use this opportunity to also shout out to Katsuo for sharing deep knowledge of Japanese resources and yudantaiteki for unbelievably informative grammar explanations. I have never skipped over any of the posts you three have made because I always learn something useful whenever I read your posts. At any rate, I apologize for being emotional, but I wanted to express my gratitude.
Technique to improve listening and speaking - SomeCallMeChris - 2012-03-23
Thank you, magamo, I'll be sure to follow all those links tomorrow and read the ones I haven't read before.
I absolutely understand your color-blindness allegory (not just becomes I have a mild variety of RG colorblindness... ) but also because of your example of 'pit' and 'spit'. I pronounced each one several times... and I can -feel- the difference in air flow on my lips. They surely must not be the exact same sound. And yet, I hear them as the exact same sound. That's a very strange feeling (and I don't suppose that there's any purpose to sharpening my hearing on it unless I start trying to perfectly mimic other English accents, and even then it would only matter if those regional speakers could hear the difference between their P and other people's P... )
Well, that's certainly given me a lot to read and to think about, thank you very much for your informative response.
Technique to improve listening and speaking - Irixmark - 2012-03-23
Yes, magamo, thanks so much. I'm learning much more here than in the Japanese classes in university... pitch accent wasn't even touched on at all there. What I remember though was that my first Japanese teacher was from a small village in southern Kyushu, and my second teacher from somewhere else... I think 長野県... and they sounded very different to me. But the first was a very short guy (for a Japanese) who had to overcompensate a lot, so asking him about his 田舎 accent was not a good idea if you cared about your grades.
The corePLUS deck has a tag "homophones" -- are those truly 100% homophones, i.e. with identical pitch accent, if there is such a thing? Or are they words with different pitch accents?
One idea might be to try to train your ear specifically on the most common homophones, ideally of course with a native speaker giving feedback (... who should on top of that speak standard Japanese).
Technique to improve listening and speaking - Realism - 2012-03-27
AlexandreC Wrote:No jokeRealism Wrote:I rarely speak Japanese to anybody because I don't know any Japanese people where I live, and I can speak Japanese just fine. A couple of months ago, I spoke to a Japanese native on Skype and we could talk normal without any problem, she didn't have to slow down or dumb down her speech or anything.Is this a joke? The idea that someone would speak a foreign language, let alone Japanese, to perfection ("no accent") without hardly ever practising it ("I rarely speak Japanese to anybody"; "I just listen to Japanese audio") is only conceivable if you are also claiming to be an extremely gifted language learner. Actually, this would go beyond any claim I've ever heard from the gifted polyglots I know or heard of.
[...]
I just listen to Japanese audio pretty much all day long and that's it really. I read Japanese websites as well, this is all I do. I can speak without any problem, no accent either.
You've never met any Europeans who could speak English almost fluently??
I mean, how often do they have to speak English in their everyday life?
Technique to improve listening and speaking - AlexandreC - 2012-03-27
Realism Wrote:Let's take the random example of a Swede or a Dutch person. Their language is related to English, they start learning it in school as a child, English music is heard on the radio all the time, English literature and movies is easily accessible and they most likely use English daily online. So the average adult might have been studying or using it for 10 or 15 years.AlexandreC Wrote:No jokeRealism Wrote:I rarely speak Japanese to anybody because I don't know any Japanese people where I live, and I can speak Japanese just fine. A couple of months ago, I spoke to a Japanese native on Skype and we could talk normal without any problem, she didn't have to slow down or dumb down her speech or anything.Is this a joke? The idea that someone would speak a foreign language, let alone Japanese, to perfection ("no accent") without hardly ever practising it ("I rarely speak Japanese to anybody"; "I just listen to Japanese audio") is only conceivable if you are also claiming to be an extremely gifted language learner. Actually, this would go beyond any claim I've ever heard from the gifted polyglots I know or heard of.
[...]
I just listen to Japanese audio pretty much all day long and that's it really. I read Japanese websites as well, this is all I do. I can speak without any problem, no accent either.
You've never met any Europeans who could speak English almost fluently??
I mean, how often do they have to speak English in their everyday life?
Now, Japanese, which would typically take 3 to 4 times longer than English, is another story.
You see, you said you spoke Japanese with no accent just from listening to Japanese. That sounds like a joke.