|
 |
"Mispronunciation index" | ||
Tags: Pronunciation | ||
Share with:
| ||
 Language Learning Forum : Specific Languages |  | |
| IronFist Senior Member United States Joined 6439 days ago 663 posts - 941 votes  Speaks: English* Studies: Japanese, Korean | Message 1 of 15 08 December 2012 at 1:55am | IP Logged | ||
|
I propose that something gets created called the mispronunciation index.
This is a measure of how close words are to each other in a language. In practical usage, it means if you mispronounce something, how likely are you to still be understood vs. be misunderstood because your mispronunciation was actually a different word that exists in that language. In other words, if you mispronounce a word, but the word you say is close to your intended word but isn't a real word, there is a better chance you will be understood. If you mispronounce a word, and the word you say is close to your intended word, but it's also another word in the language, there is less of a chance that you will be understood. For example, in English, if you try to say "slip" but accidentally say "sleep," that's another word. You might not be understood (out of context). In English, if you try to say the word "language" and accidentally pronounce it as "län-gu-odge", that's not even close to anything else, and you will still probably be understood. Bad examples but you get the idea. This is then applied against a function containing a variable representing how similar sounds within that language are to each other. Possibilities: 1) high number of words that differ only slightly + many sounds that are very similar = high mispronunciation index (example: Korean) 2) high number of words that differ only slightly + few sounds that are very similar = moderate mispronunciation index (example: Japanese) 3) low number of words that differ only slightly + many sounds that are very similar = moderate mispronunciation index (example: [I don't know but I'm sure there's some language like this]) 4) low number of words that differ only slightly + few sounds that are very similar = low mispronunciation index (example: Spanish) A high value means many words are very similar and you will likely be misunderstood. A low value means many words are different from each other and you will likely still be understood. The scale goes from 1-10. For example, a language like Korean would have a high value because not only are there lots and lots of homonyms, but there are lots of words that are only one letter different from each other and if you get one part wrong you say the wrong word, and it's still another word that exists in the language. Japanese would have a middle value because while there are lots of words that are only one syllable different from each other (such as the story of the foreigner who was talking about her experience in Japan, and she meant to say something like "the grass is the same color here as in America," but she accidentally said "kuso" (sh*t) instead of "kusa" (grass)), but because Japanese sounds are so different from one another, the likelihood of being misunderstood is much lower. You're not going to get hung up on consonants or vowels that sound similar and mix up words that way, like you will in Korean. With one version of each consonant, even if you can't say it correctly, you'll still likely be close, and close is good enough in Japanese. Close is not good enough in Korean. Also, generally speaking the more long words a language has, the less likely they are going to be close to other words, and the more time your brain has to understand what is going on. If you get *most* of a word right, there is a better chance of you being understood. If the word is only one or two syllables you need to say it perfectly to be understood, especially if there are many similar words. I have no idea how such an index would be figured out, but I think it's an awesome idea. I suppose it's something that one can do intuitively after having studied a language for a while. For example, I would put Korean at 9 and Japanese at 4. German is a 3.5 or 4. Spanish is a 3 or 4. Discuss. Edited by IronFist on 08 December 2012 at 1:57am 3 persons have voted this message useful | |||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6599 days ago 9753 posts - 15779 votes 4 sounds Speaks: Russian*, English, FinnishC1, Latin, German, Italian, Spanish, Portuguese Studies: Danish, Romanian, Polish, Belarusian, Ukrainian, Croatian, Slovenian, Catalan, Czech, Galician, Dutch, Swedish | Message 2 of 15 08 December 2012 at 8:17am | IP Logged | ||
|
Interesting. But the perception of similarity depends on your native language.
I believe most Slavic languages qualify for 3), with the hard and soft versions of almost all consonants. But for a native speaker of a Slavic language that's not much of a problem. I think the index should also include the correspondence between the written and spoken text. If you pronounce something poorly enough, it won't be understood even if there's no similar word (also possible if you KNOW what the correct pronunciation is, but just fail to produce the sound;_;). One more factor is whether the language is studied commonly and whether your mistake is common. The more common it is, the more likely you are to be understood. Another problem is of course that it's more likely to be a "legitimate excuse" for someone shying away from a language, not something that would help them to choose one. But nevertheless I'm looking forward to comments from people ranking their target languages :D Interestingly, I like the thought of listing 10 of my languages, but I don't think they're distributed evenly enough. But if you want to rank any language as 1, compare it to toki pona first :P Its pronunciation was designed to be, so to say, user-friendly, so it's probably impossible to beat. 1 person has voted this message useful | |||
| IronFist Senior Member United States Joined 6439 days ago 663 posts - 941 votes Speaks: English* Studies: Japanese, Korean | Message 3 of 15 08 December 2012 at 9:24am | IP Logged | ||
|
Yeah, I was thinking that it would have to be relative to other languages. You can't just say "Korean is hard," because while it is for native English speakers, I'm sure native speakers of other languages might not have as much difficulty with it.
So there would have to be a table with languages along the X and Y axes and you line up native language on the left and target language on the top, and it tells you the mispronunciation index for the target language based on the native language. It's not necessarily going to be an excuse for people to shy away. Some people like a challenge. I think some people on this forum have said they picked a certain language because it's hard. 1 person has voted this message useful | |||
| yong321 Groupie United States yong321.freeshe Joined 5544 days ago 80 posts - 104 votes Studies: Spanish | Message 4 of 15 08 December 2012 at 8:51pm | IP Logged | ||
|
Not sure if this was already studied by linguists. If not, this is an interesting and important research topic. I believe it can be studied mathematically, i.e. write a computer program to analyze pronunciations (using IPA e.g.) of words of a language. Taking into account of word usage frequency, the program determines how close the word's pronunciation is to another word's and returns a number. After checking a few thousand common words, it returns an index.
That eliminates the factor of the human speaker's native language. To add that factor, the program may assign a difficulty number to each vowel and consonant relative to the speaker's native language. Alternative to this "objective" method, what you described here, the "subjective" human assignment of the indexes, works too. Since your "mispronunciation index" is exclusively about how probable a mispronounced word is interpreted as another *existing* word, the name of this index is slightly misleading. But an accurate name would be a big mouthful (to me) unless someone thinks of a better term. It seems according to your words, a language with many homonyms will help push the index high. But in this case the language naturally has so many homonyms; there's no mispronunciation involved. The meaning of a sentence is determined by the context. So I don't think homonymy of a language should play a role. In fact, we may need to consciously exclude that factor when assigning an index to that language so it won't be artificially rated too high. 1 person has voted this message useful | |||
| IronFist Senior Member United States Joined 6439 days ago 663 posts - 941 votes Speaks: English* Studies: Japanese, Korean | Message 5 of 15 08 December 2012 at 10:18pm | IP Logged | ||
|
The program that compares words should only use commonly used words, however.
Many homonyms makes the index go higher, but that's only half the equation. The other half is how close the sounds are. For example, if Korean didn't have so many homonyms, it wouldn't be quite as hard, even though it's still very difficult to hear/pronounce words correctly. It's like how even though Japanese has a lot of homonyms, its pronunciation is much simpler than that of Korean, so the only difficulty comes from maybe not knowing which word with the same pronunciation was used. The added difficulty of not even being sure what word was pronounced isn't there. In other words, if I hear "hashi" in Japanese, that can mean "bridge," it can mean "end," or it can mean "chopsticks." The only confusion comes from possibly not understanding which word was meant. I am sure that the word I heard was "hashi." Listening and repeating new words in Japanese is easy. In Korean, if I hear anything, I'm never 100% sure what I just heard. Not only is there the issue of homonyms, but there are also "almost-homonyms" (words that sound the same to me and you but are actually different to Korean ears). But even if you do hear it clearly, or see it written, there are still many words it could be. Listening and repeating new words in Korean is extremely difficult. A hypothetical third category would be a language with many "almost-homonyms" but few actual homonyms. In this case, once you learn to clearly hear the words, the confusion goes away. This is what Korean would be if it didn't have tons of homonyms. Finally, just for completeness, there are languages with easily distinguished phonemes and few homonyms/almost-homonyms. It's easy to hear words, and there's no confusion as to what the words mean. Edited by IronFist on 08 December 2012 at 10:22pm 1 person has voted this message useful | |||
| shk00design Triglot Senior Member Canada Joined 4446 days ago 747 posts - 1123 votes Speaks: Cantonese*, English, Mandarin Studies: French | Message 6 of 15 14 December 2012 at 8:30pm | IP Logged | ||
|
Talk about languages that have many words with very similar pronunciations Chinese
Mandarin it is all over the place. Have to assume Vietnamese the same. Suppose if you ask somebody how well did he "sleep". The word in Chinese is 睡 and pronounce shuì. If you mispronounce the word as 水 shuǐ you are asking the person how is your "water". In the pronunciation guide you can see the difference between the accent mark over the "i" although the characters are not the same. The English language we have subtle difference between Australian, British & N. American accents. On the radio you hear words like "at all" such as "It is not like this at all." in England the "t" is pronounced in front of the "all". You hear something like "at-tall" but in N. America you tend to hear "at-all". Another word I've noticed is the word "error". In N. America you tend to hear something like "e-er" with the "r" in the middle silent. Edited by shk00design on 14 December 2012 at 8:31pm 1 person has voted this message useful | |||
| vonPeterhof Tetraglot Senior Member Russian FederationRegistered users can see my Skype Name Joined 4774 days ago 715 posts - 1527 votes Speaks: Russian*, EnglishC2, Japanese, German Studies: Kazakh, Korean, Norwegian, Turkish | Message 7 of 15 15 December 2012 at 8:21am | IP Logged | ||
Edited by vonPeterhof on 15 December 2012 at 8:21am 2 persons have voted this message useful | |||
| atama warui Triglot Senior Member Japan Joined 4703 days ago 594 posts - 985 votes Speaks: German*, English, Japanese | Message 8 of 15 17 December 2012 at 9:23am | IP Logged | ||
|
Pitch accent is 100% irrelevant for Japanese, as different regions use different pitch anyways and people understand each other by context.
While there are a lot of ways to interpret こうかい, there's always only 1 that fits the context. Now if you say 草 or クソ this is a whole different problem. Those are not homonyms by any stretch of imagination, and even tho they might be "one letter off" for Westerners, they're one mora off for Japanese people. I totally mixed up words in my conversations, like そのあと電話に乗る instead そのあと電車に乗る because sometimes stuff slips out like that, but it never bothered native speakers. They knew that I meant I'll take the train, not "ride the phone". This problem is overrated and belongs to the realm of "if I mispronounce it, it might come out as may i eat your dog OMGOMGOMG". Seriously. I can tell you that this doesn't exist at ALL in German, either, we will always understand you or not get what you want to say at all, but we'll never misunderstand the meaning because you mispronounced something "slightly" like in above example. And by the way, that girl who said "the shit here has the same color" was right. And for the message as such, it doesn't even make a difference. grass, shit, what she tried to say was transmitted correctly, no matter how it sounded. You may make people smile when messing up like that, but you won't start WW3, so this kind of activity is... killing time. Better spent by sitting down to learn :p Edited by atama warui on 17 December 2012 at 9:24am 2 persons have voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
