Reason(s) for the spelling ty-/sy-

[Gavril]

Hello,

In some Japanese textbooks (and in some scholarly texts on Japanese), there is a convention of writing ty-, sy- and si where standard Romaji would have ch- and sh(i). For example,

tyō/tyoo instead of chō 町

syō/syoo instead of shō 正

hanasi instead of hanashi 話


I know that the standard Japanese pronunciation of these consonants (the ones written ch and sh in standard Romaji) is not exactly the same as the standard English pronunciation of "sh" as in shake, or "ch" as in child, but I don't know of any phonetic environment in which these consonants are pronounced [tj]/[sj], as the spellings ty-/sy-would suggest.

Are there any contexts in which the Japanese consonants normally written ch-/sh- are pronounced [tj]/[sj] (respectively), or is there another reason for using the spelling ty-/sy-to represent these sounds?

Thank you for any assistance

 

[ms291052]

This is the difference between transcribing phone versus transcribing phonemes. When you write "cho", "sho" or "shi" you're trying to write down the sounds as they are actually produced by your mouth. When you write "tyo", "syo" or "si", however, you're not writing sounds, but you're writing down the syllables as they occur in the mind of a Japanese speaker.

That is to say in the mind of a Japanese speaker たちつてと are just /ta/ /ti/ /tu/ /te/ /to/, even though they wind up being pronounced as [ta] [chi] [tsu] [te] [to].
(Traditionally in Linguistics we use slashes to denote phonemes and brackets to denote phones.)

 

[Gavril]

This is the difference between transcribing phone versus transcribing phonemes. When you write "cho", "sho" or "shi" you're trying to write down the sounds as they are actually produced by your mouth. When you write "tyo", "syo" or "si", however, you're not writing sounds, but you're writing down the syllables as they occur in the mind of a Japanese speaker.

That is to say in the mind of a Japanese speaker たちつてと are just /ta/ /ti/ /tu/ /te/ /to/, even though they wind up being pronounced as [ta] [chi] [tsu] [te] [to].

I can see how one might conclude this in cases like the following:

desu yō > desshō
hanasu / hanashimasu
etc.

In these cases, there seems to be an alternation (though I don't know if "alternation" is the most accurate term here) between [S] > [ʃ] within individuals' speech, and you could claim on this basis that a speaker is "thinking" of alveolar consonants even when he pronounces alveopalatal consonants. (Am I correct to think that this is the reasoning linguists follow when they write dessyoo, hanasimasu and so on?)

But, in cases of originally Chinese words like shō, chō and so on (or in non-Sino-Japanese words like shiru “know”), where is the non-palatal form that the palatal form would alternate with? If there is no such form in modern-day Japanese, on what basis is it claimed that a speaker is thinking of syō, tyō and siru when he pronounces these forms as shō, chō and shiru?

Incidentally, do you know how much consensus there is (among people who teach and research the language) on the reality of ty-, sy- etc. in Japanese?

Thanks again.

 

[Flaminius]

I can see how one might conclude this in cases like the following:

desu yō > desshō
hanasu / hanashimasu
etc.

In these cases, there seems to be an alternation (though I don't know if "alternation" is the most accurate term here) between [s] > [ʃ] within individuals' speech, and you could claim on this basis that a speaker is "thinking" of alveolar consonants even when he pronounces alveopalatal consonants. (Am I correct to think that this is the reasoning linguists follow when they write dessyoo, hanasimasu and so on?)

At least in the modern standard variety, there is no /si/. I will show it in two ways. First, if /si/ existed, it should be immediately perceived as [si] in actual utterances. However, [si] does not exist in the actual speech. Second, if presented with [si], a Japanese speaker would recognise it as a different pronunciation. E.g., In natural conversations I have fooled a few people into believing that I am saying /suimasen/ while I actually said to them [simasen].

I am not going into details but [s] > [ʃ] before /i/ is palatalisation induced by the extremely front articulation of the Japanese I. The consonant /y/ has the same effect on the preceding consonant. Thus the changes you noticed in your previous post are more accurately described as below (suffixes are attached to the verb stem, not to the dictionary form):
des + yō > desyō > deshō
hanas + imasu > hanasimasu > hanashimasu
In fact, all consonants, when followed by /i/, undergo palatalisation. I am very cautious to testify for myself but I think I was completely unaware that /Ci/ is any different from /Ca/ (C stands for a consonant) in terms of the pronunciation until I studied English at school.

There are a few interesting points but I will take them up later. It's very difficult to understand why people feel the need to replicate the Japanese syllabic phonological rules in the Latin transcription.

 

[Flaminius]

Incidentally, do you know how much consensus there is (among people who teach and research the language) on the reality of ty-, sy- etc. in Japanese?

When it comes to phonology, there is no doubt that all /ch/ and /sh/ derived from ty- and sy-. In other words, they are all result of palatalisation. Sino-Japanese vocabulary involves a few hypotheses I do not fully understand. For instance, the Chinese 商 was roughly shang when the Japanese first came into contact with the word. Since the phoneme /ŋ/ did not exist, they perceived it as shau. The historical kana-representation of 商 is しやう. The actual pronunciation was either shiyau or shau. The most possible theory is that the former developed into the latter and the latter into the modern shō.

 

[Ajura]

When it comes to phonology, there is no doubt that all /ch/ and /sh/ derived from ty- and sy-. In other words, they are all result of palatalisation. Sino-Japanese vocabulary involves a few hypotheses I do not fully understand. For instance, the Chinese 商 was roughly shang when the Japanese first came into contact with the word. Since the phoneme /ŋ/ did not exist, they perceived it as shau. The historical kana-representation of 商 is しやう. The actual pronunciation was either shiyau or shau. The most possible theory is that the former developed into the latter and the latter into the modern shō.

Sorry for necroposting, I had not posted for a long time.


To explain this short ch, sh and j were phonemes that are loan phonemes naturalized at least partially to Japanese due to chinese contact and when s, t, d and z are palatalized they became sh, ch and j like the word Tefu"Butterfly" became Chou, there are Sino-Japanese words with Ch that have S forms like Cha(Tea) being Sa in Sado(Tea Ceremony) I think at one point in time [s] in japanese was pronounced as [ts], I think the Old Japanese tried to mimic the palatal consonants of Chinese and other languages by using sj-, tj- and d(z)j pronounced as Sj,tSj and dZj at first and later becoming S,tS and dZ causing sh and ch and j to at least partially naturalize in the language but not before /e/.