Japanese is well known as a language which allows only CV syllables and has no consonant clusters. And yet, due to a phonological rule that high vowels are deleted the following words are pronounced so as to sound to English ears as if they clusters.
Yasuko [jasko] pikaso (Picasso) [pkaso]
kutusu (shoe) [kutsu] sikaru (scold) [shkaru]
karosita (killed) [karoshta] kusai (smelly) [ksai]
HOW do I explain this in terms of canonical and non-canonical pronunciation with English anologies? I have searched for a week on the Internet for clear definitions of these terms and have come up empty and beat from all the reading I have skimmed over.
Please assist if you can, I would greatly appreciate it.
thanks in advance.
Explain canonical and non-canonical pronunciation(Japanese)
Moderators: Dimitris, maneki neko2, Lorikeet, Enrico Palazzo, superpeach, cecil2, Mr. Kalgukshi2
I am quite unqualified to respond to your question, having done two semesters of Japanese over 30 years ago.
However somebody should at least register that you post has been read, so:
Not all syllables are CV - you can have a syllable ending in [n] as in /shinkansen/.
And is there not one allowable initial consonant cluster, namely, [ts]? Is there not, for example, a word /shi-tsu/?
I doubt whether "high vowels are deleted" is a general rule of Japanese pronunciation. I think there will be environmental parameters restricting deletion to the presence of certain consonants. Thus, I agree that there is deletion in /yas-ko/, /karos-ta/ and /s-karu/, [sk] and [st]/[sht] being 'easily pronounceable'; but I disagree that the of /pikaso/ is elided, since it leaves a phonetically 'unnatural' cluster; and am dubious about the second of /kutusu/, since (see above) there is potentially another word /kutsu/ which must be differentiated in speech.
As to your search for canonical and noncanonical English analogies I can not help you, although I fancy there are none, as English doesn't use elision, but reduction to schwa for unstressed vowels.
Harzer
However somebody should at least register that you post has been read, so:
Not all syllables are CV - you can have a syllable ending in [n] as in /shinkansen/.
And is there not one allowable initial consonant cluster, namely, [ts]? Is there not, for example, a word /shi-tsu/?
I doubt whether "high vowels are deleted" is a general rule of Japanese pronunciation. I think there will be environmental parameters restricting deletion to the presence of certain consonants. Thus, I agree that there is deletion in /yas-ko/, /karos-ta/ and /s-karu/, [sk] and [st]/[sht] being 'easily pronounceable'; but I disagree that the of /pikaso/ is elided, since it leaves a phonetically 'unnatural' cluster; and am dubious about the second of /kutusu/, since (see above) there is potentially another word /kutsu/ which must be differentiated in speech.
As to your search for canonical and noncanonical English analogies I can not help you, although I fancy there are none, as English doesn't use elision, but reduction to schwa for unstressed vowels.
Harzer
-
Seiichi MYOGA
- Posts: 64
- Joined: Thu May 06, 2004 6:13 am
vowel devoicing
CV syllable (or mora) is a fundamental unit but a V may be deleted, resulting in a consonant cluster. This phenomenon that gets you interested is something that is generally called "vowel devoicing."
Roughly speaking, Japanese close vowels /i/ and /u/, as you have pointed out, tend to be devoiced if they appear in between voiceless consonants.
1. Yasuko [jasko]
2. s(h)ikaru (scold) [shkaru]
("Si" is represented in Hebon-shiki (Hepburn-System) as "shi." You would be required to use "shi" but not "si" in your passport.)
3. koros(h)ita (killed) [koroshta]
4. Pikaso (Picasso) [pkaso]
5. kutsu (shoe) [ktsu]
6. kusai (smelly) [ksai]
In terms of phonetics, vowel devoicing is subcategorized into two types. The first group is something where the close vowel is completely deleted (or if not, pronounced very shortly). This is generally good for /si, ci, hi, su, cu, hu/. The second group is something in which the close vowel is pronounced as voiceless. As for /ki, pi, ku, pu, sju, cju/, /i/ and /u/ are typically there but as a voiceless vowel, meaning that even if you put your index finger on your Adam's apple, you cannot feel any vibrations from the vocal cords.
So strictly speaking, the idea of being a consonant cluster may apply to (1) to (3). (As far as I am concerned, however, I always completely drop /u/ in (6).)
This is what we usually mean when we talk about vowel devoicing. However, some speakers may also devoice Japanese open vowels /a/ and /o/, as shown below.
7. kakashi (scarecrow) [kkashi]
8. katana (sword) [ktana]
9. kokoro (mind) [kkoro]
10. hokori (dust) [hkori]
You might be able to get useful information about the latter case, out of Vance's An Introduction to Japanese Phonology.
There are many factors that enter into vowel devoicing. You need, for instance, to consider variations among dialects. What is called Standard Japanese is largely based on the Tokyo dialect, where vowel devoicing is something usual. Broadly speaking, speakers of Eastern Japan, which includes Tokyo, tend to devoice the close vowels whereas speakers of Western Japan (especially those who speak the Kyoto or Osaka dialect) far less often devoice the same close vowels.
Seiichi MYOGA
Roughly speaking, Japanese close vowels /i/ and /u/, as you have pointed out, tend to be devoiced if they appear in between voiceless consonants.
1. Yasuko [jasko]
2. s(h)ikaru (scold) [shkaru]
("Si" is represented in Hebon-shiki (Hepburn-System) as "shi." You would be required to use "shi" but not "si" in your passport.)
3. koros(h)ita (killed) [koroshta]
4. Pikaso (Picasso) [pkaso]
5. kutsu (shoe) [ktsu]
6. kusai (smelly) [ksai]
In terms of phonetics, vowel devoicing is subcategorized into two types. The first group is something where the close vowel is completely deleted (or if not, pronounced very shortly). This is generally good for /si, ci, hi, su, cu, hu/. The second group is something in which the close vowel is pronounced as voiceless. As for /ki, pi, ku, pu, sju, cju/, /i/ and /u/ are typically there but as a voiceless vowel, meaning that even if you put your index finger on your Adam's apple, you cannot feel any vibrations from the vocal cords.
So strictly speaking, the idea of being a consonant cluster may apply to (1) to (3). (As far as I am concerned, however, I always completely drop /u/ in (6).)
This is what we usually mean when we talk about vowel devoicing. However, some speakers may also devoice Japanese open vowels /a/ and /o/, as shown below.
7. kakashi (scarecrow) [kkashi]
8. katana (sword) [ktana]
9. kokoro (mind) [kkoro]
10. hokori (dust) [hkori]
You might be able to get useful information about the latter case, out of Vance's An Introduction to Japanese Phonology.
There are many factors that enter into vowel devoicing. You need, for instance, to consider variations among dialects. What is called Standard Japanese is largely based on the Tokyo dialect, where vowel devoicing is something usual. Broadly speaking, speakers of Eastern Japan, which includes Tokyo, tend to devoice the close vowels whereas speakers of Western Japan (especially those who speak the Kyoto or Osaka dialect) far less often devoice the same close vowels.
Seiichi MYOGA