Talk:Japanese phonology
{{WikiProject banner shell}}
to this page and add the quality rating to that template instead of this project banner. See WP:PIQA for details.![]() | Linguistics Start‑class | |||||||||
|
{{WikiProject banner shell}}
to this page and add the quality rating to that template instead of this project banner. See WP:PIQA for details.![]() | Linguistics Start‑class | |||||||||
|
{{WikiProject banner shell}}
to this page and add the quality rating to that template instead of this project banner. See WP:PIQA for details.![]() | Japan Start‑class Mid‑importance | ||||||||||||||||
|
Interesting idea I got from a book
It seems you can actually treat Japanese and Sino-Japanese phonology in a different way.
Vowels = a u e o
Y + Vowel = ya yu ye(i) yo
this way you can say that ye is pronounced i and combines with consonants the same way ya yu and yo do
s + ya = sha s + yu = shu s + yo = sho s + ye = shi (because ye is pronounced i)
This stems from the fact that i always palatalizes, while e never does so there is no overlap between the two
so the Japanese syllable table could be written with four columns and the y row will be all filled
the writing system does not accomodate this though -Iopq 07:08, 5 May 2006 (UTC)
>>This stems from the fact that i always palatalizes, while e never does so there is no overlap between the two<<
Not true in actual spoken Japanese in some dialects. There is plenty of overlap between J /e/ and /i/. The fact that [je] could assimilate or change to [ji] or [ii] ought to tell you that. —Preceding unsigned comment added by 133.7.7.240 (talk) 02:27, 12 June 2010 (UTC)
Voicing of H produces B?
Why is B/P the voiced version of H when it makes more sense for it to be the voiced version of M?
Adding dakuten is supposed to change voicing only; movement of the teeth, toungue, and lips stay the same (or so I've learned). This works perfectly fine for K->G, S->Z, T->D (し,ち have different pronunciations, yet じ/ぢ are pronounced the same. Same with す,つ->ず/づ. ち->ぢ and す->ず are fine. I know づ was once 'dzu', but that seems to have gotten lost. Not sure on じ. The tongue doesn't touch the roof of the mouth when pronouncing 'shi' but it does for 'chi' and 'ji'.), and even W->V in katakana (assuming V is pronounced V and not B. [Interestingly, unicode characters exist for ヷ, ヸ, ヹ, and ヺ, though they don't really display properly]). Then there's H->B/P. Lips don't touch together when pronouncing H, but they do when pronouncing B/P. Comparing with the remaining unvoiced consonants N, M, Y, and R, M is the one that has the same mouth movements as B/P.
Another thing is the different pronunciations of ん. ex: /n/ before T/D/N, /ŋ/ before K/G, /m/ before M/B/P. D is voiced T, G is voiced K, but B isn't voiced M, which it might seem to be.
150.210.226.38 (talk) 22:25, 27 April 2009 (UTC)
- This is a historical derivation. For example, in English, /aɪ/ and /aʊ/ are the "long" versions of /ɪ/ and /ʌ/. This is because historically they were *ī *ū versus *i *u in Old English; subsequent sound changes affected long vowels to a greater extent than short vowels. (Compare Old English cū "cow" vs. cuppe "cup".) Japanese /h/ is historically *p. 16th century Portuguese sources record it as [f] (halfway between [p] and [h]), and Chinese words beginning or ending in [p] are spelled out in Japanese with the kana of the /h/ series (Chinese /pan/ → Japanese /han/, Chinese /tep/ → Japanese /tehu/ → post-WWII /tyou/), suggesting at the time that /h/ was still pronounced as a [p]. Modern Japanese /pp/ is the doubled homologue of /h/ (e.g. hatu-ha contracts to happa), while /b/ is the voiced homologue of this historical /p/.
- Oh, and these still are [p] is some "dialects" of Japanese (actually separate but closely related languages). For example, in Yaeyama, "boat" is [puni] instead of /hune/, and "dove" is [patu] instead of /hato/. It is thought that these were historically *pune and *pato. kwami (talk) 23:09, 27 April 2009 (UTC)
There seems to be some sort of morpho-phonemic alternation of J h, b and w. Take for example the morpheme 'hara' (field, 原). In place names, it could be pronounced 'hara', 'bara' and 'wara'. —Preceding unsigned comment added by 133.7.7.240 (talk) 12:16, 8 June 2010 (UTC)
erroneous explanation of vowels
[ed. note: large parts of this discussion that were personal attacks or other nonsense have been deleted. — kwami (talk)]
this discussion has been bowdlerized by kwami to make himself look good. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:14, 13 June 2010 (UTC)
- Your ability to mindread is impressive. -- Hoary (talk) 12:41, 13 June 2010 (UTC)
Your ability to read, such as it is, isn't. —Preceding unsigned comment added by 133.7.7.240 (talk) 13:49, 13 June 2010 (UTC)
The article vowel section states:
Japanese has 5 vowels:
The Japanese vowels are pronounced as monophthongs, unlike in English; except for /u/, they are similar to their Spanish or Italian counterparts.
Except that you can descriptively augment the vowel total with diphthongs like 'ai', 'au' and 'oi'--as well as 'ae'. If English vowel inventories include such diphthongs, why not Japanese? They are central to speaking the language and saying its vocabulary, and people who are learning Japanese would be misinformed to think they need to learn '5 monophthongs'. Totally bogus structuralist 'urban legend' nonsense. —Preceding unsigned comment added by 133.7.7.240 (talk) 12:23, 8 June 2010 (UTC)
- This is the dominant interpretation by linguists. If you've got other sources that says Japanese has phonemic diphthongs, bring them to the table. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 16:48, 8 June 2010 (UTC)
- 133.7.7.240, Japanese vowel sequences may sound to you as diphthongs, but they are not. They are just vowel sequences. One simple demonstration is that in Japanese absolutely any vowel can be followed by absolutely any other vowel, which does not happen in English. Native Japanese speakers don't perceive vowel sequences as units, while on the contrary English native speakers cannot normally separate diphthongs into their vowel components.
- Like Aeusoes says, if you do have sources stating that Japanese has diphthongs (which I strongly doubt), you are welcome add that in the article. — AdiJapan 16:54, 8 June 2010 (UTC)
For example, take a place name like 'Sabae'. The final vowel sequence as pronounced by many Japanese is so clearly a diphthong. They do not say a plus e. They say a sound that brings the two together as a unified sound.
I would say that 'ai', 'au', 'oi', and 'ae' are all diphthongs in spoken Japanese and not just double vowel sequences. But at any rate, [1]
- 2. DIPHTHONGS
Diphthongs refer to a tautosyllabic sequence of two vowels of different qualities. One question that always arises when we discuss this type of vowel is where we can draw a line between `diphthongs' as defined in this way and heterosyllabic vowel sequences, or sequences of any two vowels that occur across a syllable boundary. At least three criteria are used for the definition of diphthongs in general, one being morphological and the other two phonological. The morphological criterion is that two vocalic elements must be within a morpheme rather than across two morphemes in order to form a diphthong. Thus [ai] in the word /ai/ `love' is entitled to form a diphthong, whereas [ai] in the compound noun /ha+isya/ `tooth, doctor; dentist' is not. Of the two phonological criteria, one concerns the sonority of the two vowels in question. Given a sequence of two vowels, V1 and V2, V1 must be at least as sonorous as V2 to form a diphthong. Stated conversely, V1 and V2 belong to different syllables and do not form a diphthong if V2 is more sonorous than V1, e.g. [ia], [oa]. Potential exceptions to this are cases where the first vowel becomes a glide, e.g. [ia] → [ja], as well as cases where the second vowel becomes a schwa, e.g. [ia] → [i?]. The other phonological criterion is related to word accent and specifically applies to Japanese. The accent assigned to V2 by any quantity-sensitive accent rule usually shifts to V1 if the two vowels are within a single syllable, i.e., if they form a diphthong. This contrasts with the case where the accent assigned to V2 remains intact if the two vowels are across a syllable boundary. This interpretation is based on the general observation that accent falls on the nuclear vowel of the syllable, rather than on the mora originally designated as the accent locus by mora-counting accent rules (McCawley 1978, Kubozono 1999a). Although matters may be more complicated in some cases, the three criteria stated above seem to be sufficient when we discuss the two diphthongs [ai] and [au] in Japanese. Generally, both [ai] and [au] satisfy the two phonological requirements as long as the two vowels are within a single morpheme (see Note 5 below for some exceptional cases). Other vowel sequences such as [oi], [ei], [eu] and [ou] can also be interpreted as constituting a diphthong as long as they are tautomorphemic. [iu] and [ui] may be somewhat more ambiguous because their components, [i] and [u], are just as sonorous as each other. These vowel sequences must be tested by accent rules with respect to their syllabic status.
- 133.7.7.240 (ever consider registering?), although your personal attacks and "wiki-cursing" almost caused me to dismiss your post as so much hot air, I think the sources you cite and your basic point that five monophthongal vowels are not the end of the story are sound. Although the article already has more phonological detail than a general reader would care about, the sources you cite are enough to justify modifying the paragraph on vowels to acknowledge the existence of diphthongs in Japanese, though less common than in English. --Meyer (talk) 00:47, 11 June 2010 (UTC)
- Simple sonority wouldn't do it, as then there would be no distinction between diphthongs and sequences, whereas there's an obvious difference between [iu], [ju], and [iw]. I'll have to read the refs for accent rules. I have, however, frequently read that Japanese vowel sequences commonly reduce to diphthongs in conversational speech; that would give Japanese phonetic diphthongs but not phonemic ones.
- (I've often wondered if the moraic nature of Japanese were due to the writing system and poetic tradition rather than to anything inherent in Japanese.) — kwami (talk) 00:56, 11 June 2010 (UTC)
- Wait, V1 must be more sonorous than V2? Nonsense. The author apparently doesn't know the difference between rising and falling diphthongs. I wonder if that's because there are no diphthongs in his language? — kwami (talk) 01:03, 11 June 2010 (UTC)
- Being dismissive doesn't help advance the discussion, either, kwami. The author at least knows enough to get published in two linguistic works. If you are saying that the author's ideas are so flawed that the sources should be excluded, you should cite your own souces supporting that conclusion. Otherwise, I think at least a note acknowledging that some scholars believe Japanese has diphthongs is justified. --Meyer (talk) 02:28, 11 June 2010 (UTC)
- I'm not being dismissive, I'm pointing out apparent errors. Once the errors are removed, is there enough left to support the thesis? There very well may be, but I can't tell that until I dismiss apparent strawmen. For example, I just read an argument that Japanese must have syllables because the antepenultimate-mora accent loan-word rule needs to be modified to account for syllables. However, in all examples given but one, the accent corresponds to the accent in the source language, and this potential contributing factor is never considered. (Also, the "rule" does not appear to be very productive to begin with, as lots and lots of loans violate it.) I'd like to see some statistics that divergences form the source-language accent are statistically unlikely if syllables are not taken into account. (The section on compounding is much more convincing, at least given his poor examples from borrowing.) There's also the assumption that if there are two vowels in a syllable, then they form a diphthong. I could make a similar argument with consonants, that if a stop+fricative occurs in a syllable onset or coda, they form an affricate, but I'd be wrong: Polish contrasts the two in that position. Is an analogous phenomenon possible with vowels? I don't know: Diphthongs are notoriously difficult to account for. Anyway, if the argument holds, it's not just for /au/ and /ai/, but also /oi/, /ei/ (= /e:/), /ou/ (= /o:/), /ae/, /ue/, /ui/, /iu/, /ao/, /uo/, etc.--quite a number of diphthongs! Anyway, I've only gotten a start on the refs, both so far by the same author. — kwami (talk) 04:12, 11 June 2010 (UTC)
- I still haven't gone thru all the refs, but I thought I'd start checking on my own. First, AFAICT, VC "syllables" (Vn and V-geminate) are always accented on the V. But that could be a restriction on stressing (downstepping) on consonants, so I will disregard those as potentially irrelevant for demonstrating that Japanese has syllables. Second, AFAICT, all Sino-Japanese lexemes are accented, if at all, on the 1st mora. That is, ROku, never roKU. [Except for numerals and a few similar words which were still CVCV when a accent-shift to the 2nd mora took place] So the fact that these are never accented on the 2nd mora of a "diphthong" is again irrelevant. That's a big chunk of Japanese vocab. (A similar complicating factor may be at play in English etc. loans, I don't know.) Third, there are relatively few vowel sequences in native Japanese words, and since these all arose either thru mimesis or consonant elision, a large number are bimorphemic, and many which seem to be monomorphemic today were historically bimorphemic, so some counter-examples could be dismissed on that argument. Fourth, the putative sequence */ou/ is in nearly all cases /o:/; citations of */ou/ are in most cases perhaps an artifact of the writing system.
- That said, there are some patterns which support the thesis. I checked vowel-initial words in Sanseidoh's Sinmeikai Kokugo Jiten, 4th ed. (1989). Looking at native words w sequences not divided by morpheme boundaries, as stipulated by K., /i:/ and /a:/ are each accented on first mora in 2 lexemes (/a:/ is mimetic), /e:/ in 1 (mimetic), and accented /u:/ is not found. /o:/ is the interesting one here: there are multiple examples of /o:/ accented on the second morpheme, often alternating with the first, but in oo'doka as the only possibility listed. I suppose you could argue that these are all disyllabic, but then we get into circular reasoning about justifying syllables. (There's a similar case spelled /ou/, which in this case I believe truly is /ou/.) If we expand to SJ and English loans, there are many more examples, all accented on the first.
- Looking at /ue, uo, ui, ie, io, ae, ao, oa/ (many in loans, but a reasonable number native), there is no huge pattern. A tendency to 1st mora, but many examples on the second. /ei, oi/ are harder to tell, given an uncertainty on which forms are monomorphemic. (/ei/ is heavily dominated by SJ loans.)
- That brings us to /ai, au/, which are universally accented on the first mora. /ai/ is dominated by SJ and to a lesser extent English. All native words starting w /ai/ are compounds of a single element, which is stressed on the first mora. For /au/, there is a single token, which is an English loan (/auto/, from "out"), as there is no SJ /au/.
- So is the apparent nonsyllabicity of /ao/ and /ae/ and syllabicity of /ai/ and /au/ a reflection of native vs. loans? Native compounds may be the more convincing route to go. I'm not clear on how we can tell a priori whether VV is mono- or di-syllabic in Japanese. — kwami (talk) 06:46, 11 June 2010 (UTC)
- My god, I just realized all refs are to the same guy! His examples are quite poor: they don't take the most elementary complicating factors into account. For example, /ae, ao/ are relatively common in native Japanese, and may be accented on either mora, while /ai, au/ are almost exclusively from Chinese (/ai/ and /o:/ from historic /au/) or English, and are accented on the first mora. Long /o:/ (apart from those written "ou" from Chinese) is largely native, and may be accented on either mora; /a:, e:, i:/ are largely loans, and accented on the first mora. Don't we have any peer review of this stuff? — kwami (talk) 07:09, 11 June 2010 (UTC)
- Martin (2006) says that an accent on the second mora makes it a separate syllable. Does that mean that ha'e is one syllable, as K. transcribes it above, but that hae' is two? But M. (A Reference Grammar of Japanese) is not a rigorous book.
- Iwasaki (2002) Japanese says s.t. about "the" three diphthongs, ai, oi, ui, but I can't get access. Tsujimura (1996) quotes Vance (1987) as saying that 'any possible vowel sequence may be a diphthong'. If that's true, does 'diphthong' have any real meaning in Japanese?
- Sato (2006) Interaction between Phonetic Features and Accent-Placement in Japanese Family Names says, "Two-mora native Japanese words that form a descending diphthong usually have initial accent, with the accent being placed on the dominant mora, such as "koi" [...] and "kai". (I read this as meaning they usually are accented, and if so the accent is on the first mora.) No-one claims Japanese has rising diphthongs. But there are plenty of two-mora words such as hae which are accented on the 2nd mora. Also, verbs like taosu are accented on the 2nd, and K. maintains that verbs if accented are accented on the antepenultimate syllable. So it would seem that /ai/ and /oi/ are syllables, but that /ae/ and /ao/ are not, despite Tsujimura reporting that they are, or at least can be.
- In Kyoto, the second mora of a "syllable" can be accented, in words like koi and tombi.[2] If Kyoto has syllables the same way that Tokyo has, and Yoshida & Zamma use the same terms for both, then it does not follow that two vowels in the same "syllable" form a diphthong. — kwami (talk) 07:21, 11 June 2010 (UTC)
Much of the confusion about monophthongs vs. diphthongs in Japanese stems from the idea that older forms of Japanese didn't have diphthongs, and the writing system reflects this. Also there is the inherited concept of 'mora' often being taken to mean that because a sound is treated moraically it is treated syllabically. However, it has never really been made clear what it means to treat something 'moraically' where as a syllable is phonetically realistic and a fairly workable language universal in concrete sense.
- BTW, I'm sympathetic to the idea that traditional wisdom in Japanese phonology may be more reflective of the writing system than of the language itself. I've seen some weird phonological claims about Chinese based on the traditional onset-rhyme dictionary format. (On the other hand, similar criticisms have been made of a phonemic analysis being an artifact of using an alphabet.) Syllables aren't universal, and although Japanese may have 'normal' syllables, I have yet to see a coherent explanation as to what constitutes a diphthong in Japanese, or when a "long" vowel is monosyllabic and when it's disyllabic, whereas for English there is plenty of reference to justify A being a diphthong and B not being one, even if some of them are a bit fuzzy. — kwami (talk) 10:02, 11 June 2010 (UTC)
Some people define diphthongs as a unique co-articulation of two vowels. Others argue it is a vowel-glide-different vowel. Some argue for both. Still yet others have argued that, for example, the vowels of American English are typically 'diphthongs' because they finish with some sort of glide away from the vowel. I have yet to see a coherent argument here as to what a diphthong is for the purpose of this argument. Talk about circular. All vowels are monopthongs in Japanese, therefore Japanese has no diphthongs (apparently based on everyone citing Shibatani and Vance or simply plagiarizing a Japanese language learning site somewhere that says this). The argument that a monopthong is impossible because you can show acoustically that no vowel is a constant state. And you can show in articulation there is movement throughout. Please by all means Kwami show me a language that doesn't have syllables. What is and what is not a syllable is the issue. In spoken Japanese, long oo could be phonetically represented as something more like ou. No one here has made it clear as to why sequences like 'au' as in 'kau' shouldn't be considered a diphthong in a one-syllable word in Japanese. It is circular to say it isn't a diphthong because Japanese doesn't have diphthongs. And that is the argument being made here. —Preceding unsigned comment added by 133.7.7.240 (talk) 10:55, 11 June 2010 (UTC)
- or simply plagiarizing a Japanese language learning site somewhere that says this If you are claiming that part of this article is plagiarized, please specify the source. I'll delete it. Or somebody else will. Indeed, you can. If on the other hand nothing appears to be plagiarized, please don't suggest that something is. -- Hoary (talk) 09:56, 12 June 2010 (UTC)
And I'm sure you are going to counter with the languages that contain consonant sequences without medial vowels, but all that does is problematize the concept of syllable. All these languages have vowels, and all of them use vowels medially in some words. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:06, 11 June 2010 (UTC)
>> Such as? Apart from the diphthongs you are trying to establish, which would be a circular argument?
"like their Spanish counterparts" is intended for readers who may not know what a "monophthong" is.<<
Isn't that kind of dumb even for wikilinguistics? What if they don't know Italian or Spanish. And you never did address the issue that this article is employing a phonetic notion of diphthong while at the same time dismissing any notion of diphthong in Japanese as phonetic. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:10, 11 June 2010 (UTC)
>>You say minimal pairs. Do you have examples of minimal pairs of /ai/ vs /a.i/? — kwami (talk)<<
Using that narrow of criteria, show the phonemic status of a diphthong in English. In 1987 Vance's work on Japanese phonology claimed that in 'actual speech' some vowel sequences are achieved as diphthongs. That was however in a discussion of phonology, not phonetics. I would bet most here don't have a clear distinction between the two anyway. However, if so many vowel sequences treated moraically are achieved in actual speech syllabically, we must abandon the idea that two mora equals two syllables, which could lead us to abandon the idea that two mora means no diphthongs. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:29, 11 June 2010 (UTC)
Secondly, Kubozono knows his stuff.
Thirdly:
- "like their Spanish counterparts" is intended for readers who may not know what a "monophthong" is.
- Isn't that kind of dumb even for wikilinguistics? What if they don't know Italian or Spanish.
If you want to say that linguistics in WP isn't much good, I'd agree. Is this dumb by the standards of WP? No, I don't think so. Of course many readers will be unfamiliar with these languages. But many readers will be familiar with them. So you explain something (perhaps only halfway) in one way, and you also explain it (perhaps better) in another way; what's wrong with that? -- Hoary (talk) 11:42, 11 June 2010 (UTC)
Let's take a word like 'au'--to meet. You say the two vowels are a sequence of monophthong [a] and [u]. Well, no one says a.u. They say 'au', quite like the Enlish 'ow'. If you said a.u everyone would look at you like you were crazy. There is no two-syllable word a.u. Moreover, no one has really made a coherent argument as to what a mora is here. If there are two mora, what the heck are they? Two isochronous units of time? What the heck does that mean? —Preceding unsigned comment added by 133.7.7.240 (talk) 11:47, 11 June 2010 (UTC)
I haven't got a clue as to who Kubozono is or if he knows his stuff, but I was challenged here to find a citable source to support the idea of diphthongs in Japanese (the usual suspects doubted it). I think one issue is that monophthong was originally meant to be an ideal type in order to explain a contrast with a much different sort of vowel--such as a 'natural' diphthong with a lot of mouth movement. But it got into the abstract realm of phonemic theorizing and turned into yet another stupid monster in the room of linguistics. A PHONEME! —Preceding unsigned comment added by 133.7.7.240 (talk) 11:50, 11 June 2010 (UTC)
- What's stupid about phonemic analysis? -- Hoary (talk) 13:05, 11 June 2010 (UTC)
>>But many readers will be familiar with them. So you explain something (perhaps only halfway) in one way, and you also explain it (perhaps better) in another way; what's wrong with that?<<
What if I explained the sounds in Italian or Spanish by saying they are like the ones in Japanese? Would that prove useful too? I think statements like this ought to be avoided because they would require more clarification than the information they provide just to make them appear useful. —Preceding unsigned comment added by 133.7.7.240 (talk) 12:03, 11 June 2010 (UTC)
- Less so, I think, but merely because a smaller percentage of readers of English are familiar with Japanese than are familiar with Spanish or Italian. -- Hoary (talk) 13:05, 11 June 2010 (UTC)
>>So is the apparent nonsyllabicity of /ao/ and /ae/ and syllabicity of /ai/ and /au/ a reflection of native vs. loans? <<
No, it isn't because (1) Sino roots have so nativized they are called Sino-Japanese for a reason and (2), for example, the sequences [ai] and [au] are quite common in native Japanese. Take the verb 'au' , to meet. One of its stem forms is 'ai', it's dictionary form is 'au'. —Preceding unsigned comment added by 133.7.7.240 (talk) 12:07, 11 June 2010 (UTC)
As stupid as this criteria is, I can show 'phonetic' diphthongs creating what Kwami would call a phonemic contrasts: 'nai' (there is not), 'nae' (seedling). To say na-i or na-e is to say nonsense non-words. —Preceding unsigned comment added by 133.7.7.240 (talk) 12:23, 11 June 2010 (UTC)
- The same verb au has forms with other vowel sequences besides ai: aeru, aou. Are you claiming that the vowel sequences become diphthongs in some cases but not in others? Is there a rule for that?
- Your statement that "To say na-i or na-e is to say nonsense non-words" is exaggerated at best, but most certainly it's just wrong. That's exactly how natives speak, with monophthongs in hiatus. Any two vowels can meet in sequence in Japanese, probably in every possible pitch pattern, and there is no compelling evidence that they become diphthongs in some cases and not in others. Of course the mouth changes shape even during a monophthong, of course there is a continuous transition even in a hiatus, but none of these means there are diphthongs in Japanese. — AdiJapan 13:05, 11 June 2010 (UTC)
- Kubozono explicitly excluded words like au "to meet" as bimorphemic. There are very few cases of native monomorphemes with au or ai. If you take K's argument that a diphthong can only be accented on its nucleus, then ae and ao cannot be diphongs in Tokyo, and in Kyoto neither can the others.
- Yes, SJ is integrated into Japanese phonology. However, the phonotactics of SJ differs from that of native vocabulary, for example in being predominantly bi-moraic and accented on the first mora if at all. (Gaijin loans differ yet again.) Therefore I don't think it's unreasonable to ask whether a phonotactic pattern is related to SJ vocab. English has special phonotactics for Greco-Latin vocab. We wouldn't generalize that to all of English.
- oo, though written <ou>, is not pronounced [ou]. It's simply [o:]. That's pretty elementary.
- Au "to meet" clearly is pronounced as two vowels in hiatus in careful pronunciation. Don't you know the basics of the language?
- Of course, in practice, every vowel shifts in articulation. The question is of the ideal vowel, as perceived by a native speaker: do they have a single target? In Japanese, they do. In English, many of them (/e:, o:, i:, u:/) do not. If you're going to use that argument for diphthongs, you might as well argue that /p, t, k/ are the same consonant, because they're all pronounced the same, and that it's only the surrounding vowels which differ.
- I'm not objecting to the concept of Japanese having diphthongs. But if we're going to add that to the text, we shouldn't say "Japanese has diphthongs, but nobody knows which they are". So far I haven't seen anything that would enable us to say which they are, or at least no consensus in the lit as to which they are. Come on, if we're all too doofusy to know what to do, you'll have to explain it to us in little words. — kwami (talk) 19:45, 11 June 2010 (UTC)
>>What's stupid about phonemic analysis? -- Hoary (talk) 13:05, 11 June 2010 (UTC)<<
It's circular to say this or that phonemic contrasts create lexical contrasts (carry 'functional load') and then use lexical contrasts to show what phonemes are. It's obvious, for a start, that minimal pairs are fairly rare in the lexicon of a language. Moreover, we can show counterexamples--words that are pronounced the same but have entirely different meanings. Also, no one has ever demonstrated in speech processing or articulation research that phonemes exists. If anything, however you interpret the various results of these sorts of research, the one thing you can say is that no one has clearly demonstrated a phoneme. Also, most people can't even get close to agreeing what a phoneme is. It's a speech sound, it's a phone, no it's a sound category, it's a sound category existing outside human cognition (a Platonic cave projection), it's a sound category existing inside human cognition. Other attempts to use things like complementary distribution fail on contradictions, counter-examples and gaps in dealing with real phonology and phonetics. The phoneme is a nice little fiction to discuss a language in print in a somewhat more consistent manner than native writing systems allow. That is about all I can say for it. —Preceding unsigned comment added by 133.7.7.240 (talk) 02:35, 12 June 2010 (UTC)
- But you're assuming a "real phonology". So far as phonology is real, what should its units be? Has anyone posited a unit that has been, or that promises to be, more useful than the phoneme? You have a point about the uncertainty of the nature of a phoneme, but I think you overdo it, and anyway the uncertainty is not unusual to phonemes -- the last I heard, linguists didn't even agree on what a syllable is. -- Hoary (talk) 09:56, 12 June 2010 (UTC)
>>Kubozono explicitly excluded words like au "to meet" as bimorphemic.<<
Show me where because 'au' to meet is not bimorphemic and it is not Sino-Japanese. —Preceding unsigned comment added by 133.7.7.240 (talk) 02:37, 12 June 2010 (UTC)
>>If you take K's argument that a diphthong can only be accented on its nucleus, then ae and ao cannot be diphongs in Tokyo, and in Kyoto neither can the others.<<
What syllable with a nuclear vowel ISN'T accented on its nucleus? The question is whether or not the two vowels separate enough in order to justify saying there are one or two syllables in the sequence, a discussion further complicated by the concept of 'mora' which modern linguistics has adapted as a timing or rhythm unit, so giving it other content seems questionable. —Preceding unsigned comment added by 133.7.7.240 (talk) 02:55, 12 June 2010 (UTC)
>>Au "to meet" clearly is pronounced as two vowels in hiatus in careful pronunciation. Don't you know the basics of the language?<<
There is no hiatus. In spoken Japanese 'au' is almost a perfect rhyme with English <ow> as in how, now, cow. You are one nucleus short of being a full diphthong. —Preceding unsigned comment added by 133.7.7.240 (talk) 02:58, 12 June 2010 (UTC)
>>Of course, in practice, every vowel shifts in articulation. The question is of the ideal vowel, as perceived by a native speaker: do they have a single target? In Japanese, they do. In English, many of them (/e:, o:, i:, u:/) do not. If you're going to use that argument for diphthongs, you might as well argue that /p, t, k/ are the same consonant, because they're all pronounced the same, and that it's only the surrounding vowels which differ.<<
The point was Daniel Jones, I believe it was, originated the concept of 'monophthong' as an ideal type in phonetic description. He never meant it to be some sort of absolute category. Anyone who has studied the acoustics and articulations of vowels can tell you there is nothing that is a constant state about them. Formants emerge as something of a constant in acoustic analysis but these are not the same from beginning to end either. In terms of articulation about the only thing you can say is some vowels are more dynamic and involve more movement to create the tongue-mouth contours than others. —Preceding unsigned comment added by 133.7.7.240 (talk) 03:05, 12 June 2010 (UTC)
>>In Japanese, they do. In English, many of them (/e:, o:, i:, u:/) do not. If you're going to use that argument for diphthongs, you might as well argue that /p, t, k/ are the same consonant, because they're all pronounced the same, and that it's only the surrounding vowels which differ.<<
The arguments about [p], [t] and [k] make no sense whatsoever. With consonants you typically have points and manners of articulation to refer to that help you to pin down what the overall articulatory target/gesture might be. —Preceding unsigned comment added by 133.7.7.240 (talk) 03:09, 12 June 2010 (UTC)
I'm a native ja speaker and if asked to pronounce "au" (meet) slowly or clearly for non-native ja speakers, I would say the verb "a (pause) u". Oda Mari (talk) 06:31, 12 June 2010 (UTC)
- "There is no hiatus. In spoken Japanese 'au' is almost a perfect rhyme with English <ow> as in how, now, cow." --- That, I'm guessing, is your own opinion, probably formed by your own listening of natives. Then of course you wouldn't mind if I did the same sort of original research. Which I did. I asked a native (speaker of the Tokyo dialect) to pronounce two sentences in the most natural way. (He didn't know what I was looking for, usually I ask him this sort of questions when I'm learning the pitch accent of new words.) Here are the two sample sentences:
- 東京で会う。 Toukyou de au.
- 何時に会う? Nan ji ni au?
- In the first sentence you could argue that phonetically [u] is weaker than [a], because of the downstep as well as the nature of the two vowels, so a case could be made for a diphthong. But in the second sentence it was without any doubt a clear hiatus; there is no way you could say it was a diphthong. Is it possible that a diphthong becomes a hiatus just with a change in intonation? I doubt it is. My conclusion is that /au/ is a hiatus.
- I may not know much about phonetics, but I have a few advantages: in my native language (Romanian) we have both [a.u] and [aw], totally distinct; I've been speaking Japanese for some ten years and I've been surrounded by natives all this time; and finally I'm much less influenced by English than you seem to be.
- This, of course, is not an argument in itself (it's "original research" as I said), but might get you thinking that what you hear is not necessarily what others hear, nor what natives perceive. — AdiJapan 08:21, 12 June 2010 (UTC)
- BTW, I assume that K. would not count au "to meet" as a diphthong because it's bimorphemic. The question is whether a monomorphemic word like kai or koi has a diphthong, at least in Tokyo. Perhaps if our friend supplied us with a reference that actually listed the diphthongs of Japanese, we'd have more to go on. — kwami (talk) 08:24, 12 June 2010 (UTC)
A list of possible diphthongs in Japanese. This one looks pretty good: http://www.fonetiks.org/sou3ja.html
Even a conservative account of Japanese like Kindaichi allows for two diphthongs in spoken standard (Tokyo dialect) Japanese. http://books.google.com/books?id=tMNXDsBv2fgC&pg=PA103&lpg=PA103&dq=how+many+vowels+in+Japanese&source=bl&ots=9ggLO1dFzT&sig=DehliVC5B7Mng_dyajYaVHuiChM&hl=en&ei=0awUTKadOYnGcPr77foL&sa=X&oi=book_result&ct=result&resnum=9&ved=0CDYQ6AEwCDhQ#v=onepage&q=diphthong&f=false —Preceding unsigned comment added by 133.7.7.240 (talk) 10:06, 13 June 2010 (UTC)
- The 1st ref you gave is just a bunch of vowel sequences. You didn't even bother to check if they might cross compound boundaries, and many of them do. There's nothing on that site to suggest any of them are diphthongs.
- The 2nd ref you gave said that the diphthongs are ya, yu, yo, and wa on one page, and discussed the likelihood of half of the alleged 8 vowels of Old Japanese being diphthongs on the other, not a new idea and irrelevant to the present discussion. Nowhere does it mention diphthongs like ai, au, oi in Modern Japanese.
- Come on, a simple description of which diphthongs are found in Japanese, preferably including how it was determined they are diphthongs. If you can't provide that, there's not a lot we can do with diphthongs in the article. — kwami (talk) 10:23, 13 June 2010 (UTC)
More evidence that any phonological treatment of spoken Japanese needs to consider the possibility of diphthongs:
http://ado.lib.kagoshima-u.ac.jp/handle/10232/855
First language prosodic timing is a priority issue for TESOL teachers and phonologists alike (Celce-Murcia et al 1996; Cutler et al 1994). Vowel sequences are important because they reveal several crucial aspects of this timing in a microcosm. However, cross-linguistic comparisons need to be reassessed according to whether the focus is on micro-phonetic phenomena such as the transition between two adjacent vowels, or broader phonological indices such as vowel length or accent. Conclusions hinge on a few central observations. Phonologists tend to think of phonetic nucleus as being characterized by phonological accent. However, the comparisons of Japanese and English vowel sequences found in the literature and tested by experimental studies (Gore 2003) show that phonetic nucleus and phonological accent do not necessarily go hand in hand. It seems possible for a phonological diphthong to influence accent placement in Japanese, while being phonetically indistinguishable from a sequence which has no effect on accent; and it also seems possible for a sequence not normally considered a phonological diphthong to show more diphthongal phonetic characteristics than one which does constitute a phonological diphthong.
Are the effects of changes in speech rate on F2 transition slopes in languages with and without unitary diphthongs explicable in terms of general phonetic and physiological principles? Five American English diphthongs and four Japanese vowel sequences were recorded, spoken at three different tempos by eight and seven speakers, respectively. LPC spectra were computed at 10-ms intervals, and the resulting F2 contours were analyzed by a computer program designed to objectively define and measure steady-state and transitional portions of the vowel. Contrary to the claims of Gay [J. Acoust. Soc. Am. 44, 1570–1573 (1968)], speech rate was found to significantly affect two measures of transition slope in English, with slope decreasing as speech rate increased. Individual diphthong populations in both languages displayed high correlations between a transition, slope and its F2 range, indicating that rate adjustment is not simply a matter of target undershoot at faster speeds. Japanese showed less rate-dependent variability, suggesting that temporal reorganization for different speech rates is affected by language-specific structures. [Work supported by NIH.]
http://cat.inist.fr/?aModele=afficheN&cpsidt=12403493
http://www.lit.kobe-u.ac.jp/linguistics/KPL/4_2004/KPL_2004_katrin.pdf
http://books.google.com/books?hl=en&lr=&id=QXEgn1PCQsAC&oi=fnd&pg=PA147&dq=diphthongs+japanese&ots=7mED5FOPu9&sig=kkvLRMjTdEKpUrzzFT5cEdz9AxU#v=onepage&q=diphthong&f=false —Preceding unsigned comment added by 133.7.7.240 (talk) 10:56, 13 June 2010 (UTC)
>> Nowhere does it mention diphthongs like ai, au, oi in Modern Japanese.<<
Which is why I described it as a CONSERVATIVE account!
This page here, http://www.fonetiks.org/sou3ja.html , specifically refers to them all as diphthongs. I am not arguing that here, but you had better come up with some sort of criteria as to what is and what is not to be considered a diphthong. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:05, 13 June 2010 (UTC)
The article also states:
>>In most phonological analyses, all vowels are treated as occurring with the time frame of one mora. Phonetically long vowels, then, are treated as a sequence of two identical vowels. For example, ojiisan is /oziisaɴ/ not /oziːsaɴ/.<<
With hiatus or without? One use of the concept of 'mora' simply states that a long vowel is two mora and a short vowel one. The problem is no one can agree as to what a mora is in the real evidence of speech. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:22, 13 June 2010 (UTC)
- Yes, the mora is problematic, but you're getting into WP:OR at this point.
- A diphthong is a vowel. A sequence is not. So the question of whether ae is a diphthong depends on whether it's one vowel or two. There may not be an easy answer to that question, but what I would expect would be s.t. like the reasoning behind finding that /tʃ/ is an affricate in English, but /ts/ is not. That at least is one approach to the question, which bypasses the problem of what a syllable is. But there's also a diphthong as a vocalic nucleus+coda. Either way, it's going to be a phonological decision, not just a phonetic one.
- BTW, there's a lot of imprecise use of the word 'diphthong' in Japanese. For example, in one source I just saw, nyu contains a diphthong, but yu does not. And there is clearly a diff tween what various refs accept as a diphthong even of the /ae, ai/ kind. So no, I don't need to define what a diphthong is in Japanese, your sources do. That's the whole point of presenting evidence in an argument: whoever presents the argument needs to provide the evidence. So far you've demonstrated that there is a lot of talk about "diphthongs" in Japanese. But there are plenty of sources which state that Japanese, like Russian and Hungarian, has no diphthongs. If we're to cover this adequately, we're going to need to know (a) what the diphthongs are, (b) why there is disagreement over there existence, and (c) how well accepted they are in the field. — kwami (talk) 11:32, 13 June 2010 (UTC)
>>Either way, it's going to be a phonological decision, not just a phonetic one. <<
That itself is not clear. There has been an awful lot of phonetic deciding done here in these discussions.
>>So far you've demonstrated that there is a lot of talk about "diphthongs" in Japanese.<<
This article cites very few sources of a scholarly nature on Japanese. I just cited a whole slew where diphthongs are discussed. A cited an old classic, Kindaichi, which at least admits of two diphthongs. You are the embodiment of a bad faith discussant Kwami. You say you are going to need to know what the diphthongs are. Take an analogy. What are the phonemes of English--counts vary by almost a dozen. Back to Japanese, a conservative account would admit of a handful, a more flexible approach would expand the number considerably and enunciate a principle as to why this or that particular sequence should not be treated as a diphthong. So far no one here has done that--you only dismiss the idea of diphthongs in a circular argument that has lacked evidence totally. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:49, 13 June 2010 (UTC)
- Well let's not worry about the frightful Kwami. Go ahead, Mr/Ms IP, tell us what you want to say. -- Hoary (talk) 12:57, 13 June 2010 (UTC)
I just did in how many words Hoary? But you can't have a discussion with people who alter the record to suit themselves. At any rate, any description of Japanese vowels that doesn't include the possibility of diphthongs does not reflect how many years of discussion in actual scholarly treatment of Japanese phonology and phonetics. I don't care what Kwami says, he has been so demonstratively wrong on most points so far.
At one extreme, just about any vowel sequence could be treated as a potential diphthong in Japanese. Then we might use auditory perception to determine whether or not hiatus rules them out. Even if we can't provide phonemic accounts of said dipthongs, that does not rule out inclusion in any discussion of the phonology of Japanese--you might start with some of the sources this article has in its bibliography but fails to consult or specifically cite. —Preceding unsigned comment added by 133.7.7.240 (talk) 13:07, 13 June 2010 (UTC)
- No, no: you go ahead. You can either edit the article directly, or you can write a draft in your own user area (at least after you've taken the trouble to get your own username). -- Hoary (talk) 13:11, 13 June 2010 (UTC)
I have a user name. I just don't use it until an area has been cleared of bridge trolls. Some articles you can never get them out no matter what (e.g. 'phonology', 'phoneme', etc.). Anyone can edit, but how do you stop the re-edits and reversions to suit the bridge trolls? It's a wiki-dilemma.
Here is kwami being 'civil' in his own way, to a published scholar (can you feel the envy?): >>Wait, V1 must be more sonorous than V2? Nonsense. The author apparently doesn't know the difference between rising and falling diphthongs. I wonder if that's because there are no diphthongs in his language? — kwami (talk) 01:03, 11 June 2010 (UTC)<< —Preceding unsigned comment added by 133.7.7.240 (talk)
Before I was so rudely interrupted yesterday, I was digging into sources on Japanese phonology and phonetics, particularly diphthongs. But wait, first the so-called syllabic N. If this is realized as a nasalized vowel (or a nasalization and lengthening of the preceding vowel), what does that do to statements like 'all Japanese vowels are pure vowels or monopthongs'? Now about the diphthongs. It has to be remembered that phonological and phonetic accounts of languages exist for different reasons. In the case of Japanese, most typically they exist for these purposes: 1. Japanese description for native literacy and language awareness, 2. JSL and JFL learning (to make the language more understandable, accessible and learnable, it is hoped), 3. academic, somewhat technical descriptions for phonetics and phonology. This article seems to draw on all three trends in a somewhat mixed up fashion. Finally for now, even if we can not demonstrate a strict 'phonemic' status for any Japanese diphthongs, that does not mean they do not participate in phonological and morpho-phonological processes in Japanese. Therefore, I will undertake to writing up an account of what are the possible diphthongs of Japanese in some fairly standard sources and amend this article appropriately. Of course the explanation will include the qualifications and controversies (e.g. why some would claim Japanese has no diphthongs, while others would claim it has no phonemic dipthongs, while others would say it has at least three, while others would claim a long list of them). This will include the source Vance, which is a fairly conservative mostly phonemic account of Tokyo dialect Japanese. It is cited here in the bibliography, but apparently no one has bothered to consult it on diphthongs, among other things133.7.7.240 (talk) 08:11, 14 June 2010 (UTC)
- Now that sounds reasonable.
- As for nasal vowels, that's a different thing than a diphthong. You can have "pure" nasal vowels as well as nasal diphthongs. As for the case in Japanese, it's allophony. We have allophonic nasal vowels in English too, but we don't therefore say that "English has nasal vowels". — kwami (talk) 09:20, 14 June 2010 (UTC)
The question though is: are these nasalized, vowelized forms of syllabic /N/ monophthongs? Also, it's an interesting sort of allophony because it strains 'phonetic relatedness', crossing categories from a consonant into a vowel. A different way of dealing with the issue, though, is to say the syllabic /N/ results in a lengthening and nasalization of the preceding vowel. But this also might depend on how you analyze syllables in Japanese, with many accounts being syllable-avoiding (or at least they wouldn't like the idea that the syllabic /N/ is actually a syllable coda, even if it adds to mora counts).133.7.7.240 (talk) 02:41, 15 June 2010 (UTC)
The article needs a re-think here, but before I get into a tug-of-war over who has the right to make statements about Japanese's vowels, let me look at this issue here:
The article states: >>Vowels have a phonemic length distinction (short vs. long). Compare contrasting pairs of words like ojisan /ozisaɴ/ "uncle" vs. ojiisan /oziisaɴ/ "grandfather", or tsuki /tuki/ "moon" vs. tsūki /tuuki/ "airflow". In most phonological analyses, all vowels are treated as occurring with the time frame of one mora. Phonetically long vowels, then, are treated as a sequence of two identical vowels. For example, ojiisan is /oziisaɴ/ not /oziːsaɴ/.<<
So is the article claiming that 'vowel length' is its own phoneme? Some accounts simply say a short vowel, relatively speaking, is 'one beat', and a long vowel, 'two beats'. In other words, in terms of linguistic timing (and it is hoped somewhat isochronous in actual research), that means a short vowel is in a one-mora syllable while a long vowel is in a two-mora syllable. WHY EXCLUDE SUCH EXPLANATIONS? Moreover, it should be pointed out that this quoted bit above then says "Phonetically the long vowels, then, are treated as a sequence of two identical vowels", but what are they phonologically and phonemically (since so much has been made of the need to exclude diphthongs on theoretical phonemic basis)? 133.7.7.240 (talk) 12:20, 16 June 2010 (UTC)
Deletion
Would anyone like to remove this section? Well that's too bad, you can't. -- Hoary (talk) 12:41, 13 June 2010 (UTC)
Archiving
This talk page is long and promises to get longer. MiszaBot's instructions ask me to get your agreement before asking it to handle all this automatically. So, would it be OK to get MiszaBot to archive? -- Hoary (talk) 10:13, 12 June 2010 (UTC)
- We've had this talk page 4 years, and yet this last "discussion" takes up half the page. I doubt we're likely to get things like that often, so I doubt there's any need for automated archiving. We could archive manually after this is over, and then let it sit for another four years. I don't object exactly, just don't see the point. — kwami (talk) 11:25, 12 June 2010 (UTC)
- I've archived half the content of this page. -- Hoary (talk) 14:40, 12 June 2010 (UTC)
Citation style
The entire article is suspect and most likely represents the views of very few people. The bibliography is of some length, but it's not really clear how this material has been used to compose most of the article. For example, the previous waste-of-time discussion diphthongs, when the article itself cites a work that discusses diphthongs in Japanese while the article denies their existence. More wiki-garbage in linguistics. —Preceding unsigned comment added by 133.7.7.240 (talk) 11:58, 13 June 2010 (UTC)
- So go ahead and improve it. -- Hoary (talk) 12:36, 13 June 2010 (UTC)
I can't even post to the freaking edit/discussion page without being deleted, such as the stunt Kwami pulled here. Who has the time to spend hours and even days fighting every deletion and reversion of the bridge trolls of wiki? —Preceding unsigned comment added by 133.7.7.240 (talk) 13:09, 13 June 2010 (UTC)
- A stunt? Please identify the offending "diff". (You can find it here.) And sign your posts: you do this by hitting the twiddle key "~" four times in a row. (Even a humorless ignoramus dogmatic doofus -- did I forget anything? -- such as myself can manage to do this.) -- Hoary (talk) 13:16, 13 June 2010 (UTC)
- Please, Hoary, no feeding the trolls. You're our link to sanity here. — kwami (talk) 13:17, 13 June 2010 (UTC)
No, I won't sign my posts. They auto-sign, something even a moronic wiki like this can do. —Preceding unsigned comment added by 133.7.7.240 (talk) 13:20, 13 June 2010 (UTC)
Did you guys want to discuss why this article isn't in conformance with the usual citation style or did you just want to continue to act like egomaniacal doofuses? —Preceding unsigned comment added by 133.7.7.240 (talk) 13:21, 13 June 2010 (UTC)
- Thanks for the choice; I'll plump for the latter. Meanwhile, if you would like to improve the citation style, go ahead. -- Hoary (talk) 13:38, 13 June 2010 (UTC)
>>A stunt? Please identify the offending "diff"<<
From memory I can see that Kwami deleted part of the discussion about the Tokyo informant for the Romanian who knows for sure what a Japanese diphthong is because he/she has less exposure to English than I do. But you two seem to have a lot of time on your hands so you go figure it out. —Preceding unsigned comment added by 133.7.7.240 (talk) 13:23, 13 June 2010 (UTC)
- Nope. If you want to complain, you decide whether or not you can be bothered to do the spadework. If you can, it's likely that the complaint will be investigated. If you can't, it won't be. -- Hoary (talk) 13:38, 13 June 2010 (UTC)
>>Nope. If you want to complain, you decide whether or not you can be bothered to do the spadework. If you can, it's likely that the complaint will be investigated. If you can't, it won't be. -- Hoary (talk) 13:38, 13 June 2010 (UTC)<<
Certain people think they are above the rules and working policies of wiki, such as what went on here yesterday, which included the deletion of others' content in order to alter the outcome of an argument. However, getting back to the purpose of this particular discussion, it really is a major fault of the article that so many good sources have been tacked on to a bibliography, but that these sources have not been integrated into the actual discussion of the article on Japanese phonology. That was my point in starting this section, the vandalistic actions of certain self-appointed wiki experts notwithstanding. Answer the charge or allow others to improve this woeful article.133.7.7.240 (talk) 08:19, 14 June 2010 (UTC)
- We're still waiting for you to provide sources for improvements. You have yet to provide us with so much as an account of which diphthongs Japanese has. — kwami (talk) 09:16, 14 June 2010 (UTC)
- I'm less worried about the bibliography than you are, but I can see that some people may get the wrong impression that the article has been improved by reference to, or even that it reflects the insights offered in, those works. I've therefore commented out every item that isn't cited. (As the markup isn't conspicuous and might otherwise go unnoticed, I've added strings of "+" characters to draw attention to it.) ¶ I allow you to improve this woeful article, if you do so by citing respected phonology texts. (But I don't have to allow you -- the article is not "protected" or even "semi-protected".) -- Hoary (talk) 11:06, 14 June 2010 (UTC)
>>We're still waiting for you to provide sources for improvements. You have yet to provide us with so much as an account of which diphthongs Japanese has. — kwami (talk) 09:16, 14 June 2010 (UTC)<<
First, I provided you a page that lists all possible diphthongs of Japanese (or at least all possible vowel combinations) and it even includes sound files so you can listen as to whether or not they might meet some criteria for determining what is and what is not a diphthong (better than a report of someone in Tokyo being asked to read a sentence aloud if only because we can all hear it). I have also provided you with the citation of Kindaichi, an older, traditional 'phonemic' approach to the phonology of Japanese, which compasses two or three diphthongs (I can't remember right now). Moreover, as I have pointed out, the other diphthongs in a phonological discussion are listed in Vance, which this article had in its bibliography but didn't obviously make use of. I plan to use Kindaichi, Vance and at least one more, but having moved offices recently, those books are apparently in storage somewhere.133.7.7.240 (talk) 02:49, 15 June 2010 (UTC)
- "all possible diphthongs of Japanese (or at least all possible vowel combinations)" -- these are not even close to being the same thing. Also, I challenge anyone to tell from sound files of English /dz/, /dʒ/, /ts/ and /tʃ/ (say, adze, cage, catsup, catch) which are the affricates.
- Kindaichi proposes four diphthongs in Japanese, actually: /ja/, /ju/, /jo/, and /wa/. That seems a bit off-topic.
- But I sympathize with the move. I can't locate half of my refs either. (And I don't have Vance.) — kwami (talk) 06:27, 15 June 2010 (UTC)
>>"all possible diphthongs of Japanese (or at least all possible vowel combinations)" -- these are not even close to being the same thing<<
Well it is a site with phonetics. To be generous to it, it takes the most expansive view of diphthong possible. Also, it's easier to hear vowel sounds like this in sound files than it is fricative, affricate and sibilant consonants. I do have to wonder how anyone anywhere could claim Japanese 'oi' isn't a diphthong once you hear it shouted. 133.7.7.240 (talk) 14:02, 15 June 2010 (UTC)
- Interjections and mimesis are not good data for arguing the general phonology of a language. If we did that, we'd have to say that English is a click language that allows fricatives as nuclei. — kwami (talk) 21:23, 15 June 2010 (UTC)
Except 'oi' is like the equivalent of 'Hey there' in Japanese and warrants entries in dictionaries. So you can continue to dissent all you want, but I think readers will benefit that this article is finally going to get a bit of improvement. Am looking for Kindaichi, Vance and the CUP publication 'Languages of Japan' (I forget the author but he is a top academic linguist). 133.7.7.240 (talk) 03:21, 16 June 2010 (UTC)
- Huh? "Hey" is also an interjection. Lots of interjections get into dictionaries, but that doesn't mean that that they follow normal phonology. "Tsk" and "tchik" made it into the OED, but that doesn't make English a click language.
- The CUP book is Sibatani. Little help there: he never mentions diphthongs. He simply has 5 vowels, long and short. Long vowels are not single segments, as tone drop occurs in the middle of them, but neither are they equivalent to vowel sequences involving grammatical endings or compounds. So it would appear their are monosyllabic vowel sequences and disyllabic vowel sequences. That is, /a/ vs /a:/ vs /aa/, where /:/ is a segmental phoneme, or /a./ vs /aa./ vs /a.a/. [My OR here:] the oo in ookii and the ae in hae would therefore need to be disyllabic, since the tone can rise in the middle of them, and therefore hae cannot be a diphthong. ai would normally seem to be monosyllabic, but (extrapolating from Sibatani) a monosyllabic sequence, not a diphthong, not a segment.
- Yes, I'm aware that diphthongs are often defined simply as two vowels in the same syllable, but Sibatani shows things are not that simple for a moraic language. There is another common definition of diphthongs: two vowels in the same segment (a vocalic contour segment). (After all, contours consonants (affricates) are defined segmentally, not syllabically.) That might be a better test for a moraic language like Japanese. — kwami (talk) 06:44, 16 June 2010 (UTC)
I'm still looking for my copy of Shibatani. He has 5 vowel sounds, but how many vowel phonemes in his inventory? In other research, it's been noted formant changes across the long vowels, so calling them a monophthong is doubtful (even if this goes back to an articulatory phonetic distinction between 'pure' and 'complex' vowels). Both the terms 'diphthong' and 'segment' are problematic. For example, we don't speak speech in segments; linguists analyze speech into segments. Still, I'm not sure what use I can make of Shibatani or how it squares with the explanation in this article. Much of the problem is there seems to be very little explicit agreement as to how to use terms like sound sequence, segment, syllable, diphthong, etc. I'm not even sure what it means to say a language is 'moraic' since you can apply a term like 'mora' to any number of languages, including English (if you can discern an isochronous unit of timing, or at least a linguistically relative unit of timing). Clearly this article explanation of the vowels could benefit from taking in a few more accounts and explaining the differences. There are so many controversies in linguistics its misleading and not useful to be dogmatic about a given language133.7.7.240 (talk) 12:36, 16 June 2010 (UTC)
- I agree with you there. Many of these terms are arguable; in fact, that what a lot of phonologists spend much of their time doing. Although we should include the controversies over Japanese in this article, it's too much to include arguments over whether 'phoneme' is a valid concept: leave that to the phoneme article.
- Sibatani posits 5 vowel phonemes: /a e i o u/, and discusses that Kindaiti would have a 6th, /:/. He hedges his bets a bit, admitting that long vowels cause difficulties and are not easy to analyse. He said somewhere (now I can't find it) something about the difficulties of trying to get a perfect phonemic analysis, and doubting that it's worthwhile.
- You're really hung up over monophthongs! Yes, of course, any vowel will have variation in its articulation. The question is whether such variation is meaningful. When you ask someone to pronounce a monophthong clearly, they attempt to make it as steady as possible. With a diphthong, it's always a contour. With a monophthong, the more carefully it's enunciated, the less contour it displays; with a diphthong, the more carefully it's enunciated, the more contour it displays, reaching a farther target. With a monophthong, the contour is variable in direction, depending on surrounding consonants (and especially adjacent vowels!); with a diphthong, the contour may be variable to a degree, but these are variations on a single direction: say, low-back to high-front, even if the start and end points, or the curve of the path between them, vary. — kwami (talk) 20:32, 16 June 2010 (UTC)
>>since the tone can rise in the middle of them, and therefore hae cannot be a diphthong. ai would normally seem to be monosyllabic, <<
Could you clarify what you mean here? Just re-state the part about what 'tone' does and what this means as to what is and what is not possible. Thanks. 133.7.7.240 (talk) 12:54, 16 June 2010 (UTC)
- I think Kwami is talking about Japanese pitch accent. Hae/fly, hai/lung aitai/want to meet and kai/buying are all LH accent and the stress is on the second vowel. Oda Mari (talk) 15:38, 16 June 2010 (UTC)
- Right. One of the things I keep reading over and over in accounts that posit diphthongs is that we know they're diphthongs because the 'accent' (which in Tokyo dialect is a downstep) can only occur on the first element. But in hae etc. are accented on the 2nd element, so by that argument cannot be monosyllabic.
- Oda Mari, my dictionary says hai 'lung' has no accent. If it has high pitch on the i, presumably that's the same rise in pitch that we get with ha 'leaf', which also has no accent (unlike ne 'root'). A diphthongist would say that it's a simple rise in pitch across the diphthong, just as across the vowel of ha. You can see it isn't 'accented' on that syllable in haiga 'lung(NOM)', where the high tone continues onto the subject particle ga. (Assuming you have the same stress patterns as my dictionary!)
- As for aitai and kai 'buying', AFAICT all of our 'syllabic' sources would say these are not monosyllabic, because they are not monomorphemic. They all mention that you can get 'accent' on the 2nd element of a vowel sequence when it's a separate morpheme, such as the final u or i of a verb.
- BTW, Sibatani, who our anon. protester recommends as an expert, uses this restriction in accent placement in words like kai 'shell' to argue that accent placement is determined by syllables rather than by moras in Tokyo Japanese. However, he does not conclude from that that Japanese has diphthongs: he posits 5 vowels plus length, and says that "The basic vowel phonemes [of Tokyo] are quite straightforward." — kwami (talk) 19:56, 16 June 2010 (UTC)
>>Right. One of the things I keep reading over and over in accounts that posit diphthongs is that we know they're diphthongs because the 'accent' (which in Tokyo dialect is a downstep) can only occur on the first element. But in hae etc. are accented on the 2nd element, so by that argument cannot be monosyllabic.<<
1. Some sources on 'syllable' and 'diphthong' (not just in Japanese) do not exclude differences in accent going either way. The only requirement is that there is a difference between the two vowel elements that form the diphthong.
2. I am not recommending Shibatani, I'm just trying to find my copy of his book. I had long suspected he was too abstract and antiquated in his handling of phonology and phonemics, and that these probably weren't his strong points anyway (probably a background in socio-linguistics and ethnography, given the expertise on Ainu).
3. I wouldn't make too much of my anonymity Kwami because no one here thinks that is your real name. Let's just say for now my IP address is my pen name and that tells more than your pen name since yours doesn't show an IP address even.
4. My points about vowels, diphthongs and syllables in Japanese comes down to this: this article presents a somewhat erratic but technical view of vowels, diphthongs, mora, etc. but is oblivious to the fact that there are various accounts of these concepts and these concepts as applied to Japanese. So what is not clear is how such an idiosyncratic view was derived from the given sources (but it looks like two plus the synthesis of Kwami and several others, citations lacking). What is erroneous about the presentation of vowels etc. here is the omission of other descriptions, accounts and explanations. 133.7.7.240 (talk) 02:10, 17 June 2010 (UTC)
>>You're really hung up over monophthongs! Yes, of course, any vowel will have variation in its articulation. The question is whether such variation is meaningful.<<
Right, I don't like the term 'monophthong' very much at all, because I doubt it exists in any spoken language (just as cardinal vowels don't). Because phonemics, however limited and paradox-prone, doesn't need such descriptions. As I pointed out before, I suspect it was a phonetician who came up with the idea, in order to contrast in articulatory terms somewhat simple vs. somewhat complex vowels (and on to dipthongs, triphthongs, from there). I'm not sure whether the question of such variation depends on 'meaningful'. You seem to ignore that criterion in your follow-up explanation anyway. At any rate, if Shibatani, one of the published experts on Japanese, says the vowel length does create difficulties for phonemic accounts, isn't that worth mentioning?133.7.7.240 (talk) 02:17, 17 June 2010 (UTC)
>>Huh? "Hey" is also an interjection. Lots of interjections get into dictionaries, but that doesn't mean that that they follow normal phonology. "Tsk" and "tchik" made it into the OED, but that doesn't make English a click language.<<
But notice the BIG difference here. Tsk, psst, shhh, clicks, these might be possible in English but hardly typical sequences of sounds, and not ones important to creating the phonic substance of the lexicon. On the other hand, Japanese may have relatively few 'vowel types' or 'vocalic values', but as a spoken language it is filled with vowels sounds. And it also has many vowel possible patterns of vowel combinations, some of which might be diphthongs in a phonological sense. Even if one or two sources (unclearly cited) tell us not. 133.7.7.240 (talk) 02:21, 17 June 2010 (UTC)
>>I agree with you there. Many of these terms are arguable; in fact, that what a lot of phonologists spend much of their time doing. Although we should include the controversies over Japanese in this article, it's too much to include arguments over whether 'phoneme' is a valid concept: leave that to the phoneme article.<<
Not quite because some of the controversy comes from the difficulty in applying concepts that were got through analyzing, for example, English, to a relatively unrelated language (isolate), for example, Japanese (and Japan is a land of many academic linguists). Also, the way this article deploys such terms as syllable, monophthong, diphthong, tone, etc. you get the feeling there is a nuanced intelligence behind these decisions, but for an encyclopedia article, that can be obscurantist, not explanatory. There might even be a whiff of reverse Nihonjinron about. By the way, the phoneme article sucks big time. That is all they could do there, argue, and the ones with the last reverts and entries prevailed (much to the detriment of the article). 133.7.7.240 (talk) 02:28, 17 June 2010 (UTC)
>>As for aitai and kai 'buying', AFAICT all of our 'syllabic' sources would say these are not monosyllabic, because they are not monomorphemic. They all mention that you can get 'accent' on the 2nd element of a vowel sequence when it's a separate morpheme, such as the final u or i of a verb.<<
Well, in some languages, even morpheme boundaries are not clear (English for one). But I'm following the argument here both for monophthongs and two-vowel sequences (that could or could not be diphthongs, which themselves may or may not exist). For example, the monopthongs. At least in a simplistic phonetic sense for the sake of argument (what native speakers say in controlled speech, what we hear in the acoustic data, what me might perceive--so even this is not so simple), we could say about monophthongs that we might a a one-mora type, a two-mora sequence of the same vowel with not break in morpheme, and a two-mora sequence that crosses morpheme boundaries (such as from the end of content morpheme hiatus onto a grammatical morpheme, most notable 'o' or 'wo').
Now about the diphthong examples, let's assume 'kai' is two-morpheme because of the highly inflectional nature of Japanese verbs. But what about the word 'kai' as in shell? There are actually all sorts of 'kai' morphemes in Japanese, most of which are not bi-morphemically inflected. 133.7.7.240 (talk) 02:41, 17 June 2010 (UTC)
Also, it seems to me we are making morpho-phonological arguments against diphthongs in Japanese, and not clear-cut phonological or phonetic ones. 133.7.7.240 (talk) 02:43, 17 June 2010 (UTC)
>>Right. One of the things I keep reading over and over in accounts that posit diphthongs is that we know they're diphthongs because the 'accent' (which in Tokyo dialect is a downstep) can only occur on the first element.<<
Over the past few days, I've seen pitch, tone and accent used as something different but also as something almost the same. Could you clarify what you mean by the terms? Traditional accounts say there is limited (and dialectal) use of contrasting pitch in spoken Japanese. Some say this is a type of contrastive tone. You seemed to have used 'tone' before in a different sense. Accent could mean a rise in pitch and/or an increase in stress. I've even seen acoustic accounts of 'long' vowels in Japanese that shows they are stressed (which is not to say that isochrony in Japanese makes the same use of stress as English does, although most linguists are going to weasel words like 'stress-based', 'syllable-based' and 'mora-based' to avoid the over-committment of terms like 'stress-timed', etc.). 133.7.7.240 (talk) 02:52, 17 June 2010 (UTC)
Moraic obstruent
This is an inadequate treatment of the phenomenon, to say the least--actually it's rather quaint in its adherence to archi-phonemes, etc.
Without discussing the glottal aspects of this/these sounds, you can't really make much sense of them. —Preceding unsigned comment added by 133.7.7.240 (talk) 12:08, 13 June 2010 (UTC)
- So go ahead and improve it. -- Hoary (talk) 12:35, 13 June 2010 (UTC)
By what criteria would any improvements be judged? Someone from Romania talked to someone in Tokyo and got them to read a sentence out loud? —Preceding unsigned comment added by 133.7.7.240 (talk) 13:10, 13 June 2010 (UTC)
- Well, life is a bitch. So what are you going to do about it? If you really believe that editing Wikipedia is a hopeless enterprise, it's hard for me to understand why you're spending your time here. Briefly, the sensible options would seem to be (i) writing well and persuasively, and (ii) spending your time on some other website (or indeed "real life") instead. -- Hoary (talk) 13:21, 13 June 2010 (UTC)
The immediate above comment has absolutely nothing whatsoever to do with the moraic obstruent and the inadequate explanation of it in the article. After addressing diphthongs, I will return to this, because it participates in some very interesting aspects of Japanese language and deserves better, more detailed sourced explanation. 133.7.7.240 (talk) 08:21, 14 June 2010 (UTC)
- I look forward to your improved explanation. -- Hoary (talk) 10:50, 14 June 2010 (UTC)
As an aside, from Yuko Yoshida, "Licensing to let", in Living on the edge (Kaye & Ploch eds, 2003), p 456ff,
- Aim of paper: "that /N/, the so-called "moraic nasal" [...] should not be treated as a consonant [...] but instead involves a nuclear position"
- "In many cases, "bimoraic" Sino-words comprise a diphthong or a long vowel, with an optional consonant preceding it, and the second vowel of those vowel sequences is avoided for accent assignment."
- [Note: she calls them both 'diphthongs' and 'sequences']
- When accented, 145 of 177 accented "Sino-words" (kanji) are stressed on mora 1. Said to reflect that in Chinese this "is the head of the word-domain. This headship manifests itself in Japanese in terms of pitch accent. [...] the headship on N1 [mora 1] is inherited from the original language Ancient Chinese, and thus N2 does not have to fulfil the conditions [...] to be the head of that domain unless N2 receives an accent via a specific process such as the case seen in numbers".
Transcribes /VN/ as [Vɯ̃] which, given that there's no IPA symbol for Japanese /u/, sounds reasonable, though there may be allophonic lip closure. If we take this to be a nasal vowel, then depending on which sequences are "diphthongs", we might have nasal diphthongs in J.
Note: French nasal vowels borrowed as /VN/, in contrast to French /Vn/, which is borrowed as /Vnu/. — kwami (talk) 11:07, 14 June 2010 (UTC)
Removals of comments
There've been several removals of comments hereabouts, some even citing vandalism.
I hereby waive any right (?) to be treated "civilly". Please don't remove any comment because it's incidentally impolite to or about me. The requirement of civility is rather a bore, and I think that we can assume that anybody at Talk:Japanese phonology (unlike, say, those at Talk:Ashlee Simpson) will have a mental age of 14+ and will thus be unfazed by the odd bit of criticism. -- Hoary (talk) 14:09, 13 June 2010 (UTC)
- I'm sure you can handle it. But even your good-humored comebacks, when there gets to be pages of them, are inappropriate: Talk pages are for improving the article, not for trolling or playing games. Personal attacks and other such nonsense should be deleted, so that readers coming here see a page discussing the article, not ourselves. — kwami (talk) 14:30, 13 June 2010 (UTC)
- Incivility and personal attacks create a hostile editing environment for all concerned, and should be avoided no matter who the specific target is Please do not comment on other contributors, focus instead on the editorial content of the article. Generally, it's inappropriate to remove the comments of others on any talk page other than your own user talk page, per WP:NPA Removal of text. Only truly egregious attacks should be removed, and it would be wise to seek assistance from an admin instead of removing them yourself. Dreadstar ☥ 19:21, 13 June 2010 (UTC)
Bibliography
As noted above, I've just now commented out every item in the bibliography that isn't explicitly cited from within the article. -- Hoary (talk) 11:08, 14 June 2010 (UTC)
- I've taken the commented out items and put them in a separate "further reading" section. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:40, 14 June 2010 (UTC)
What is wiki style on this? Further Reading or Suggested Reading? Some styles would have a works cited list followed by a more expansive bibliography. It's these sorts of things that scholars expect and benefit from. I'll defer to the experts on wiki as to what the wiki style is.133.7.7.240 (talk) 02:44, 15 June 2010 (UTC)
- "WP:MOS" probably explains. I'm no expert, but I'd use "Further reading" if only because "Suggested reading" seems to suggest value judgements. (Well of course either way there are value judgements involved, but one can choose whether or not to remind the reader of this.) -- Hoary (talk) 07:39, 15 June 2010 (UTC)
Genesis of this article
It would appear to be to quite an extent a 'lifting' of the World Lingo page on 'Japanese phonology', which helps account for the idiosyncracies of the bibliography (e.g., works listed not cited, over-reliance on a few sources, such as Akamatsu, etc.). Or at least it has been lifted from the same source that WorldLingo lifted theirs from. On the downside, that is not good because wiki pages typically come up first or very high in search results, but the article does not appear to be an originally written encyclopedia article. The good side is that the wiki article does seem to have undergone some development and revision, at least as far as description of the consonants is concerned. However, I could also hold out the other possibility--that the copyrighted World Lingo article has been lifted from the wiki article. At any rate, I actually like its more straightforward treatment of consonants over the one in the wiki article.
>>http://www.worldlingo.com/ma/enwiki/en/Japanese_phonology
133.7.7.240 (talk) 06:22, 16 June 2010 (UTC)
- Since you provided the link, I deleted the material you lifted from them.
- WorldLingo just plagiarizes other sites. It's obvious in this case, because there was development of this article before it got to the where it is in the WorldLingo mirror; also, some things which I wrote myself appear in the WorldLingo version. — kwami (talk) 06:41, 16 June 2010 (UTC)
- It's a mirror of Wikipedia, read the note at the bottom of the page: "The original article is from Wikipedia. To view the original article please click here." It's a mirror copy of an older version of our article. Dreadstar ☥ 07:12, 16 June 2010 (UTC)
I see that note now. Before all I saw was the copyright. But the mystery is, how did an article that cites so much in its bibliography come to cite so little in its explication (looks like mostly Akamatsu and Itoh used). 133.7.7.240 (talk) 11:58, 16 June 2010 (UTC)
I don't know. Maybe the person or people (not me) who added them all did so in part as a to-do list: "I've heard of these; must get around to reading them some time." -- Hoary (talk) 14:07, 16 June 2010 (UTC)