User:Urszag/Sandbox

Subpage: User:Urszag/Vowel_devoicing_in_Japanese

=Japanese phonology=

Youngberg 2021 notes:
 * Gives further examples of accented "deficient moras".
 * Claims initial CVJ syllables may show HH pitch (p. 9).

Voiceless coronal affricate
The affricate originated as a contextual variant of the plosive. In the history of Japanese, the coronal plosives were changed into phonetic affricates before the high vowels : as a result, original  developed to. In core vocabulary, and  are in complementary distribution:  never occurs before, whereas  occurs only before.

Therefore, can be analyzed as a conditioned allophone of  before.

In loanwords, however, can occur before other vowels: examples include  ツァイトガイスト;  エリツィン. There are also a few marginal native forms with before a vowel other than, such as otottsan (the standard form of this word is otōsan). Based on dialectal or colloquial forms like these, as well as the phonetic distance between plosive and affricate sounds, argues that the affricate  is its own phoneme, represented by the non-IPA symbol. In contrast, disregards such forms as exceptional, and prefers analyzing  as an allophone of  rather than as a distinct affricate phoneme.

Before the high vowel, original was both affricated and palatalized, developing to the alveolo-palatal affricate [tɕ]. This can be regarded as the palatalized counterpart of : thus, analysts like Hattori treat [tɕ] as a palatalized allophone of (a phonemic affricate  ), whereas analysts like Shibatani treat  as a palatalized allophone of. (A third group of analysts treat as its own phoneme, distinct from both  and, due to the coalescence of original  sequences to , which created a contrast between  and .)

Voiced affricate vs. fricative
Like, the voiced plosive historically underwent affrication before , which caused  to become. Therefore, initially functioned as an allophone of  before. However, the distinction between the voiced affricate and the voiced fricative  is neutralized in Standard Japanese and in most (although not all) regional Japanese dialects. The result of this merger is a phonetically variable sound that can be transcribed phonemically as, though some analyze it as , the voiced counterpart to. The distinction between and, which were originally palatalized allophones of  and , is likewise neutralized in Standard Japanese, resulting in a phonetically variable  sound.

A 2010 corpus study found that in neutralizing varieties, the fricative pronunciations and the affricate pronunciations  could both be found in any position in a word, but the likelihood of an affricate pronunciation was increased in phonetic conditions that allowed for greater time to articulate the consonant: voiced affricates were found to occur on average 60% of the time after, 74% after , and 80% after a pause. In addition, the rate of fricative realizations increased as speech rate increased. In terms of direction, these effects match those found for the use of plosive vs. non-plosive pronunciations of the voiced stops ; however, the overall rate of fricative realizations of (including both  and, in either intervocalic or postnasal position) seems to be higher than the rate of non-plosive realizations of.

As a result of the neutralization, the historical spelling distinction between these sounds has been eliminated from the modern written standard except in cases where a mora is repeated once voiceless and once voiced, or where rendaku occurs in a compound word: つづく[続く], いちづける[位置付ける] from. The use of the historical or morphological spelling in these contexts does not indicate a phonetic distinction: and  in Standard Japanese are variably pronounced with affricates or fricatives according to the contextual tendencies described above, regardless of whether they are underlyingly voiced or derived by rendaku from  and.

Some dialects, e.g. Tosa, retain the distinctions between and  and between  and, while others distinguish only  and  but not  and. Yet others merge all four, e.g. north Tōhoku.

Vowels
is the most difficult vowel to phonetically characterize. Broad phonetic transcriptions often represent it as ; in narrow transcription, its value has been suggested to range over, ,.

There is particular inconsistency in descriptions of its lip position. It is commonly described as unrounded. Akamatsu 1997 lays particular emphasis on its lack of rounding. However, some descriptions state that it is not entirely unrounded, at least on the phonological level. Others state that it is unrounded but not spread or compressed but not protruded (with compression optional in conversational tempos). In contradiction to the preceding descriptions, Nogita and Yamane found the lips to be rounded and protruded, not compressed, during the pronunciation of in a study that recorded the lips of Japanese speakers, and propose that the transcription  is more accurate than  (comparing it to the quality of Norwegian ); they suggest that the perceptual impression of unroundedness could in part be a result of its fronted articulation.

It is notably fronter than cardinal and can be described as central or near-back. It may be especially fronted after and palatalized consonants, and possibly also after.

suggests it can be narrowly transcribed as after  and  elsewhere.

Long vowels and double vowels
All vowels display a length contrast: short vowels are phonemically distinct from long vowels:
 * {| cellpadding="5"


 * 小母さん
 * お婆さん
 * 怪訝
 * 軽減
 * 蛭
 * ヒール
 * 都会
 * 倒壊
 * 区
 * 空
 * }
 * 蛭
 * ヒール
 * 都会
 * 倒壊
 * 区
 * 空
 * }
 * 倒壊
 * 区
 * 空
 * }
 * 空
 * }
 * }

Long vowels are pronounced with around 2.5 or 3 times the phonetic duration of short vowels, but are considered to be two moras long at the phonological level. In normal speech, a "double vowel", that is, a sequence of two identical short vowels (for example, across morpheme boundaries), is pronounced the same way as a long vowel. However, in slow or formal speech, a sequence of two identical short vowels may be distinguished from an intrinsically long vowel:


 * {| cellpadding="5"


 * 砂糖屋
 * 里親
 * }
 * 里親
 * }
 * }

In the above transcription, represents a hiatus between vowels; sources differ on how they transcribe and describe the phonetic realization of hiatus in Japanese. says it can be "a pause or a light glottal stop", and adopts the transcription. states that there is no complete glottal closure and questions whether there is any actual glottal narrowing at all. Vance describes it as vowel rearticulation (a drop in intensity) and transcribes it as or.

In addition, a double vowel may bear pitch accent on either the first or second element, whereas an intrinsically long vowel can be accented only on its first mora. The distinction between double vowels and long vowels may be phonologically analyzed in various ways. One analysis interprets long vowels as ending in a special segment that adds a mora to the preceding vowel sound (a chroneme). Another analysis interprets long vowels as sequences of the same vowel phoneme twice, with double vowels distinguished by the presence of a "zero consonant" or empty onset between the vowels. A third approach also interprets long vowels as sequences of the same vowel phoneme twice, but treats the difference between and  as a matter of syllabification, with the long vowel consisting of the phonemes  pronounced in one syllable, and the double vowel consisting of the same two phonemes split across two syllables.

For the most part, double vowels (as opposed to long vowels) are found only across a morpheme boundary. However, exceptions may exist: for example, some speakers may pronounce the word 炎 with a double vowel, even though this word is arguably monomorphemic in modern Japanese.

Vowel sequences and (quasi)diphthongs
Any pair of non-identical vowel phonemes can occur in sequence within a word or phrase (although [only a subset of vowel sequences] can be found within a morpheme in native or Sino-Japanese vocabulary).

Japanese is often described as having no diphthongs. There is consensus that sequences such as  do not constitute unitary vowel phonemes in Japanese (unlike English ). However, some analysts use the term diphthong or quasi-diphthong to refer to a sequence of two different vowel phonemes pronounced in the same syllable, as opposed to a sequence of two vowel phonemes separated by a syllable boundary. (The existence of the syllable as a unit of Japanese prosody is itself controversial; argues against the syllable as a unit of Japanese phonology and thus concludes that no vowel sequences ought to be analyzed as diphthongs.)

It has been argued that diphthongs, like long vowels, cannot bear a pitch accent on the second mora. By this criterion, the three vowel sequences, , usually behave as diphthongs (as well as  to some extent, although it is not stably distinguished from the long vowel ); other vowel sequences do not constitute diphthongs, as they can freely bear accent on the second mora. The distinct accentual behavior of vowel sequences ending in compared to other vowel sequences can be seen in the following contexts:
 * A generalization can be made that when the dictionary form of a Japanese verb has an accent, it almost always falls on the second-to-last syllable. The form has accent on the third-to-last mora, which is consistent with this generalization if  is a diphthong here. (The accentuation of some other verbs, such as, indicates that even vowel sequences that could validly constitute a diphthong or long vowel may potentially be pronounced with hiatus in this context.)

Some other evidence has been put forward to support the analysis of as diphthongs:
 * they are sometimes set to a single note in music:

There is some conflicting evidence in regard to the status of and. Certain verbs have dictionary forms that end in or  (where /C/ is some consonant): according to the generalization discussed above, accent on the third-to-last mora is unexpected if  and  are in separate syllables here, so it can be interpreted as a sign that these sequences can constitute diphthongs. However, Vance argues that these are better interpreted as containing  in hiatus with exceptional accent placement, rather than diphthongs. It is possible that verbs with accent on the first mora of /ae/ or /oe/ acquired this accentuation pattern as a result of having variant pronunciations with /ai/ or /oi/; although currently nonstandard, such pronunciations may have been historically prevalent.

Another sequence that shows some conflicting evidence of being a diphthong is.
 * Evidence of hiatus:
 * When the particle の is placed after the dictionary form of an unaccented verb, it causes the preceding syllable to be accented. This accent falls on in, indicating that  in this context is separated by a syllable boundary.
 * Accent is placed on in the compounds  "River Donau",  "people of Lindau"
 * However, in the accentuation of foreign city names, can behave like a diphthong.

Some prior descriptions based on criteria other than unaccentability of the second mora have reached incompatible conclusions about which vowel sequences can constitute diphthongs. For example, some prior literature has described sequences such as, , , as diphthongs when they occur within a morpheme.

Assuming the pitch-accent criterion is adopted, the presence or absence of a morpheme boundary does not consistently correspond to whether a vowel sequence is a diphthong or not. In some contexts, a VV sequence that could form a valid diphthong is separated by a syllable break that corresponds to a morpheme boundary, as in 'well with a pulley', composed of  'wheel, car' and  'well'. However, in other cases, it seems to be possible for a VV sequence to be pronounced in one syllable even across a morpheme boundary. For example, 歯医者 is morphologically a compound of 歯 and 医者 (itself composed of the morphemes 医 and 者); despite the morpheme boundary between and  in this word, they seem to be pronounced in one syllable as a diphthong, making it homophonous with 敗者. Likewise, the morpheme used as a suffix to form the dictionary form (or affirmative nonpast-tense form) of an i-adjective is almost never pronounced as a separate syllable; instead, it combines with a preceding stem-final  to form the long vowel, or with a preceding stem-final ,  or  to form a diphthong.

When followed by, there is a tendency for , , to be split across syllables rather than pronounced as diphthongs. Hiatus seems to be usual when the vowels are separated by a morpheme boundary (as they always are in the context of Sino-Japanese words). Kubozono argues that these sequences are usually syllabified as, , even when they occur within a single morpheme in words of foreign origin.

In general, it is not so rare for a heterosyllabic sequence of two non-identical vowels to occur within a morpheme.

In contexts where the accentual criterion cannot be applied (as when neither vowel is accented) it can be difficult to distinguish between a diphthong and a vowel sequence:

Another environment where it may sometimes be difficult to determine whether, , constitute diphthongs is when they are preceded or followed by another vowel.


 * 多い /oꜜRi/ 'numerous' (also pronounced [ooꜜi])
 * 遠い /toRi/ 'far'
 * 可愛い /kawaiꜜi/
 * Vance cites /pureiboHiꜜša/ as an example of unambiguous /oH.i/.

Aside from the dispute about their prosodic status, whether or how diphthongs differ phonetically from vowel sequences is another point of discussion.
 * Youngberg 2021 states that diphthongs, like long vowels, cannot normally be pronounced with a glottal stop or vowel rearticulation between their two moras, whereas this may optionally occur between two vowels that belong to separate syllables.
 * In contrast, states that a pause or glottal stop may intervene between the two vowels of a putative diphthong in slow, careful speech. Labrune also suggests that there is normally not a gradual transition between the qualities of the first and second vowel; instead, the two vowel qualities are pronounced distinctly.