What's new

Pitch Accent

TurkeySloth

後輩
8 Jul 2008
96
11
18
While searching various Japanese terms on Wiktionary—which includes an accented transcription, I've noticed the fointricacies below.

1. Long vowels tend to either have a medial downstep (Ryuu [ryúꜜù]) or high tone continuing from the second mora to the following mora, after which the downstep may occur (roomaji [ròómájí]/[ròómáꜜjì])
Related question
Do any words have long vowels that don't vary in pitch? Wiktionary would transcribe them with the same character ([ɯ́ᵝː, ɯ̀ᵝː] as examples; [úú, ùù])

2. There's no downstep after the moraic nasal, which I've only seen as high tone (kanji [kàńjí]). Granted, Wiktionary could just be missing the transcription with the downstep, which would be [kàńꜜjì].
Related question
Is the moraic nasal always high tone? If so, is there ever a downstep after it? If not to main question, how often is it low tome?
 
1)
It totally depends on the word, as I already mentioned in your previous thread.

青龍 (blue dragon): type 1 (the accent is on the first mora) /セ]イリュウ/
清流 (clear stream): type 0 (flat) /セイリュウ=/

青龍刀 (liuyedao/Chinese sword): type 0 /セイリュウトウ=/
清流党 (the name of a party/group): type 0 /セイリュウトウ=/

小山竜 (surname コヤマ + given name リュウ): type1 /コ]ヤマ/ + type1 /リュ]ウ/
小山竜 (the name of a dragon or dinosaur): type 3 /コヤマ]リュウ/
小山流 (the name of a school of tea ceremony, flower arrangement, marshal arts or like that): type 0 /コヤマリュウ=/

竜派 (the name of a school/Ryū school): type 0 /リュウハ=/
流派 (school): type 1 /リュ]ウハ/

竜門 (dragon's gate): type 0 /リュウモン=/
流紋 (current mark): type 0 /リュウモン=/

登竜門 (gateway to success): type 3 /トウリュ]ウモン/
流紋岩 (rhyolite): type 3 /リュウモ]ンガン/
 
Last edited:
Thanks again. Somehow, I'd completely forgotten about that topicuntil after I posted this one. This can be merged into that, if you'd like to do so.
 
I think there's no problem to make a new thread since the previous one is a bit old.

2)
If you mean that the accent is not on ン by "there's no downstep after the moraic nasal", the answer is yes in standard Japanese. The accent can be on ン in some dialects, though.


Do you mean the accent is not on the mora right before ン by "the moraic nasal is always high tone"? If so, it's not correct.

漢字 (kanji) type 0 /カンジ=/
幹事 (secretary/organizer of a meeting) type 1 /カ]ンジ/

喧嘩 (fight/quarrel) type 0 /ケンカ=/
献花 (floral tribute) type 1 /ケ]ンカ/
 
Last edited:
Okay and thanks once more. Wiktionary uses the Tokyo dialect for its transcriptions, which have the accent directly on the moraic nasal with level (high) tone on all subsequent morae. Additionally, I asked with kanji being the only word containing the moraic nasal that I'd looked up. Thus, that question was more of one to see what I can expect from other words/dialects (if any are ever used there).
 
I meant to ask about type 2 and 3 accents before, because type 2 is absent from your list of words with long vowels (presumably absent from such words) and type 3's lack of an explanation.

Are accent types 2 and 3 related to the second and third morae, respectively, under your explanations? For example using extrapolated moraic transcription to illustrate the accents, Oo as [ò̞.ó̞]/[o̞.o̞ꜜ] (type 2) and Mishiima as [mì.shì.ì.má]/[mi.ɕi.i.mäꜜ] (type 3).
 
Last edited:
Are accent types 2 and 3 related to the second and third morae, respectively, under your explanations?
Right, as "]" is inserted after the mora.

小山竜 : type 3 (the accent is on the third mora) /コヤマ]リュウ/

examples of type 2
赤龍 (red dragon): type 2 (the accent is on the second mora) /セキ]リュウ/
お竜 (polite prefix オ + given name リュウ): type 2 /オリュ]ウ/

For example using extrapolated moraic transcription to illustrate the accents, Oo as [ò̞.ó̞]/[o̞.o̞ꜜ] (type 2)
The type differs depending on the position of the long vowel.

大王 (great king): type 3 /ダイオ]ウ/
偽王 (counterfeit king): type 2 /ギオ]ウ/
(The pronunciation of オウ is the long vowel オー "ō" in both words.)

Mishiima as [mì.shì.ì.má]/[mi.ɕi.i.mäꜜ] (type 3)
I don't understand what "Mishiima" means. If it's a surname or location name, it's type 2 /ミシ]イマ/ or type 0 /ミシイマ=/.

(I edited my previous posts in this thread since I realized I had forgotten to put "=" to the type 0 words.)
 
Doomo arigatoo. It's nice to know I understood "]" correctly in the type 1 words. However, I don't know why I thought the same mark was another mora in the type 3 words, especially since I figured the words without any non-kana characters were type 0.

Mishiima's a surname slightly altered from that of the Tekken series' Mishima family's. Thus, it'd be more classically rendered as Mishīma.
 
@Michealin are you familiar at all with Dogen's youtube channel? I first came across his Japanese comedy videos; he focuses on pitch accent instruction and has a Patreon series dedicated to pronunciation, and has some teaser content on youtube.

This video may be a little remedial but he introduces the different pitch patterns that express in Japanese (頭高、中高、尾高、平板), and provides some examples. As always, I recommend listening to native Japanese to understand how these patterns play out, rather than abstracting Japanese into a ruleset.


Regarding "Mishiima," is this a name that you created? I'm afraid it does not work as a Japanese surname, which are usually made of two kanji. If you take 三("mi" - three)島("shima" - island), you can't insert an extra "i" into "shima" without completely changing the characters, and even then the result would not look like a name, as it would be an assortment of characters to approximate the sound.

Fun fact: according to Wikipedia, the Tekken Character Kazuya Mishima is modeled after the writer Yukio Mishima, who chose it as his pen name when he was 16. There is a city in Shizuoka with the same name and Kanji.
 
Thanks for the help and suggestion. I'll be sure to watch that video and others on the channel when I have time to do so. The other daimiyoo's names I know, aside from Koyama are below. Please assess them. They're all fictional to some extent because of the time gap between the real Sengoku jidai and that of my setting.

Iiefuyaa [iː.eꜜɸɯ.jäː] (original/fictional; might need changed)
Kuutaa [kɯː.täː] (original/fictional)
Ku=ih [kuꜜiç] (original, Ainu-influenced/fictional)
Moro [mo.ɾoꜜ] (original/fictional)
Hiimuraa [çiꜜːmɯ.ɾäː] (Rurooni Kenshin-inspired; probablyhas similar issues to Mishiima) (Wikipedia gives the kanji 緋村 for Himura, that doesn't show up directy on Wiktionary.)
Uuhara [ɯː.hä.ɾäꜜ] (original/fictional)
Yowrá [joʊ.ɾäꜜ] (original, Ainu-influenced/fictional)
 
Last edited:
OK I think I was misunderstanding you as a language student, but it seems that your interest is based around this story you're writing, and you aren't that familiar with how kanji works, correct?

Here's my take on the names you listed. Granted, I'm no expert on how names worked back in sengoku jidai, and a lot has probably changed since then (like in Europe, where the peasants didn't even have surnames). These days, names (particularly surnames) are written with kanji, which may have multiple different readings used specifically for names (called nanori). Because of this and the way that Japanese phonetics works, I would not recommend adding long vowels to make a name unique, if your goal is to be realistic. Also, because of all the different nanori readings, it would be difficult to sight-read the kanji for a name if you're using an unusual reading.

  • Iiefuyaa - sounds closer to a first name (I know a guy named 尚史, Naofumi, for example). There are surnames Iie and containing Fuya
  • Kuutaa - There are names with kuta, and apparently one kuuta, but none with both long vowels
  • Ku=ih - I can't speak to Ainu names, but in Japanese the only consonant that can end a syllable is N. Names containing kuih~ exist, but the h is actually at the head of a mora, and is always followed by a vowel.
  • Moro - This name does exist: Moro
  • Hiimuraa - Like you mentioned, Himura is a real name, and lengthening the vowels has the same problem as I explained before. In the anime, they used an unusual kanji for the first character, meaning "red/crimson," probably referring to his hair color. Swapping a common kanji for an unusual one with the same possible reading for a fictional character is a common approach.
  • Uuhara - No, but there is Uehara
  • Yowrá - There seem to be some entries for Youra

I recommend digging a little for more info on names used during that period. If you're making up fictional characters, I would suggest basing them on actual people from that area and use similar names, and not modify the spellings listed without knowing the alternate characters they would use:

 
Mishiima's a surname slightly altered from that of the Tekken series' Mishima family's. Thus, it'd be more classically rendered as Mishīma.
Long vowel marks like ō, ū or ī are not "classical". These are just a way to represent long vowels in romaji (cf. Hepburn romanization, Kunrei-shiki romanization, Nihon-shiki romanization).

You also need to know that romaji is not a way to show the pronunciation accurately. It's just a way to transliterate hiragana/katakana into Latin alphabet. Thus, whether it's written as Kyōto, Kyouto, Kyooto Kyohto or customarily Kyoto, there is no problem as far as it can clearly represent きょうと.

Mishiima can be something like 実椎間, 三四今 or 御子井間, i.e., "mi + shii + ma", "mi + shi + ima" or "mi + shi + i + ma", respectively, which are completely different from 三島, as nice gaijin-san mentioned.

However, these names are never written as Mishīma because "shii" is しい, not しー. For example, the province 紀伊 is Kii. Kī is the romaji for キー, i.e., the loan word for "key". Kī is not used for 紀伊 even though the pronunciation is usually the same as キー.

Thanks for the help and suggestion. I'll be sure to watch that video and others on the channel when I have time to do so. The other daimiyoo's names I know, aside from Koyama are below. Please assess them. They're all fictional to some extent because of the time gap between the real Sengoku jidai and that of my setting.

Iiefuyaa [iː.eꜜɸɯ.jäː] (original/fictional; might need changed)
Kuutaa [kɯː.täː] (original/fictional)
Ku=ih [kuꜜiç] (original, Ainu-influenced/fictional)
Moro [mo.ɾoꜜ] (original/fictional)
Hiimuraa [çiꜜːmɯ.ɾäː] (Rurooni Kenshin-inspired; probablyhas similar issues to Mishiima) (Wikipedia gives the kanji 緋村 for Himura, that doesn't show up directy on Wiktionary.)
Uuhara [ɯː.hä.ɾäꜜ] (original/fictional)
Yowrá [joʊ.ɾäꜜ] (original, Ainu-influenced/fictional)
Those names don't sound natural to my ears except Moro. Japanese words don't end with "yaa" "taa" or "raa". Mishiima is rather odd, too.
 
I see and was aware that vowels with macrons aren't classical. I think I conflated the term with also for some reason. However, it now seems that ī is rarely used in native words. My substitution of ī for ii probably relates to Japanese Name Gender Finder's wildly inconsistant indacation of long vowels. For example, under O, you'll find Ohba, Ohka/Ooka (presumably the same name), Otani (should be Ootani, for example), and Otomo (should be Ootomo, for example). Mind you, I couldn't find the surnames Ohba and Ohka/Ooka anywhere else. Additionally, I expected the Ainu-inspired names to sound strange to you because Hokkaidoo Ainu's the only extant Ainu language, but it's nearly extinct and only spoken on Hokkaidoo. I decided to give the Ainu people a bit more recognition in my setting. Thus, the Yowra clan controls Hokkaidoo and the Tsugaru Strait while the Ku=ih clan controls most of northeastern Honshuu.

Just to test if the names need changed completely, altered spellings and/or pronunciations for the Japanese names other than Moro are below.
Iiefuya [iꜜː.e.ɸɯ.jä]
Kuuta [kɯꜜː.tä]
Hiimura [çiꜜːmɯ.ɾä]
Uuhara [ɯꜜː.hä.ɾä]
 
I see and was aware that vowels with macrons aren't classical. I think I conflated the term with also for some reason. However, it now seems that ī is rarely used in native words. My substitution of ī for ii probably relates to Japanese Name Gender Finder's wildly inconsistant indacation of long vowels. For example, under O, you'll find Ohba, Ohka/Ooka (presumably the same name), Otani (should be Ootani, for example), and Otomo (should be Ootomo, for example). Mind you, I couldn't find the surnames Ohba and Ohka/Ooka anywhere else. Additionally, I expected the Ainu-inspired names to sound strange to you because Hokkaidoo Ainu's the only extant Ainu language, but it's nearly extinct and only spoken on Hokkaidoo. I decided to give the Ainu people a bit more recognition in my setting. Thus, the Yowra clan controls Hokkaidoo and the Tsugaru Strait while the Ku=ih clan controls most of northeastern Honshuu.

Just to test if the names need changed completely, altered spellings and/or pronunciations for the Japanese names other than Moro are below.
Iiefuya [iꜜː.e.ɸɯ.jä]
Kuuta [kɯꜜː.tä]
Hiimura [çiꜜːmɯ.ɾä]
Uuhara [ɯꜜː.hä.ɾä]
I accidentally forgot Arikawa [ä.ɾiꜜka.wa] on the original list and, therefore, this one also. Does that name pass muster?
 
Otani (should be Ootani
The short vowel Otani also exists (e.g., 小谷, 尾谷).

I couldn't find the surnames Ohba and Ohka/Ooka anywhere else.


Additionally, I expected the Ainu-inspired names to sound strange to you because Hokkaidoo Ainu's the only extant Ainu language, but it's nearly extinct and only spoken on Hokkaidoo.
Do you know a popular manga/anime Golden Kamuy? Ainu people and their culture are one of the keys of the story, so lots of Aainu names are used there like アシㇼパ Asirpa (it's pronounced as アシリパ Ashiripa in the anime). It doesn't sound like common Japanese names, of course, but it's not unnatural at the same time, unlike yours.

You need to know that we already know many location names that were derived from Ainu language such like Noboribetsu. This is from ヌプルペッ nupur-pet, thus, the pronunciation changes to Japanese-like ones just like loan words. I believe the same phenomenon will occur also on the surnames, unless they only use Ainu language.

Just to test if the names need changed completely, altered spellings and/or pronunciations for the Japanese names other than Moro are below.
Iiefuya [iꜜː.e.ɸɯ.jä]
Kuuta [kɯꜜː.tä]
Hiimura [çiꜜːmɯ.ɾä]
Uuhara [ɯꜜː.hä.ɾä]
Iiefuya is still rather unnatural.
Nice gaijin-san already mentioned Kuuta.
Hiimura could be barely possible.
Uuhara is not changed from your previous post, which I already pointed out that it's unnatural.

I accidentally forgot Arikawa [ä.ɾiꜜka.wa] on the original list and, therefore, this one also.
Arikawa already exists.

I'm not going to offend you, but it seems to me that the problem in your coined names is that you made them just from the viewpoint of the pronunciations without knowing Japanese grammar, vocabulary or how the words/names are made in Japanese language.
 
Firstly, nice gaijin-san, thanks for all of those very helpful links. I'll look into them more later.

Toritoribe-san, thank you from double-checking me on Wikipedia. I, certainly, should've done that before posting. Additionally, I should've specified the I couldn't find the names anywhere that offers kanji reliably for its names, which consists of Jisho and Wiktionary at the moment. I, simply, swapped Uuhara's accent to the only place it could possibly be, if the name had any chance to work. Since you're still asying it's unnatural, it looks like that name as well as Iiefuya will be changed. If I keep the name Hiimura, it'll be altered back to Himura, its stated basis. And, you're correct that I threw together the completely original names from elements I like without regard for grammar, which was daft on my part because I'm perfectly aware of why Japanese names are how they appear. Chalk it up to me wanting the names to sound too unique for my and their own good. For example, Koyama Ryuu, my character, has a perfectly normal Japanese name, as do mum (Akemi) and dad (Toshiro). They aren't the first Koyama family I've used either because that honor belongs to the family of a now-disused neko named Koyama Yaeko, whose family members I've forgotten the names of.
 
I've decided to use the names below. However, Wiktionary lacks their accent patterms. What are their accent patterns? Additionally, I'd love to stop using Wiktionary for Japanese because of these deficiencies. What's another site that could provide the info I need? I'm not against using the names of clans from our Sengoku jidai, but, if I use any, I'd rather use three or fewer because of the hundereds of years between that and the one happening in my setting. In fact, I, originally, had a Tachibana clan in the setting and may reinstitute it.

Arikawa
Himura
Kuihira
Kuuta
Mishima
Moro
 
Last edited:
Arikawa: type 2
Himura: type 1
Kuihira: type 2
Kuuta: type 1 or 0
Mishima: type 0
Moro: type 1 or 0

What's another site that could provide the info I need?
The following site has pitch accent information, but only for very common surnames.

I'm not against using the names of clans from our Sengoku jidai, but, if I use any, I'd rather use three or fewer because of the hundereds of years between that and the one happening in my setting.
I think given names would reflect more strongly the difference between the periods rather than surnames.
 
Found this chart that may make it easier to visualize. Match up the number of mora and try speaking the names with different patterns.

With a lot of exposure to spoken Japanese it becomes pretty immediately apparent whether the pitch accent is right or not, but this aspect of Japanese is not often taught because while it's crucial for accent reduction, it's not crucial for being understood, which is the main goal of most language learners. There are only a small handful of homonyms that can be differentiated by their pitch accents.

If you're writing this story in English, selecting names that are possible/realistic in Japanese and getting their romaji transcriptions correct is enough; knowing the correct pitch accent is only useful when saying the character names aloud, and even then it's forgivable to fudge the pitch if you are saying their name in the context of a sentence in English. When I'm speaking English with English speakers, I will generally say "karate" with the American accent and stress pattern; although it's less "correct" than suddenly switching to my "Japanese" voice and saying it properly, it's better understood by the listener, and saves me from looking too pedantic.

Japanese-Accent-Patterns.jpg
 
Thanks for the wonderful help, guys.

Toritoribe-san, Am I correct in thinking you're saying I should give peasants full names as well and ditch the idea of jidaigeki naming customs? Additionally, that site's lovely, even though a computer translates 東 as "east" rather than any of the names it could be—for example. And, that accent information helped a bunch.

Nice gaijin-san, I may not be able to read the image's text. But, I, certainly, understand what it's illustrating. Honestly, I'm glad you posted something in kana/kanji rather than roomaji because of the exposure it gives me.
 
Am I correct in thinking you're saying I should give peasants full names as well and ditch the idea of jidaigeki naming customs?
I'm talking about given names of daimyo/samurai and their wives. Surnames of daimyo/samurai are not so different from, or almost the same as the ones of modern people, but given names are often different. Indeed Toshiro or Akemi was used also in Sengoku period, but it's uncommon for Daimyo and their wives. Those names sound more likely modern ones.
 
I'm talking about given names of daimyo/samurai and their wives. Surnames of daimyo/samurai are not so different from, or almost the same as the ones of modern people, but given names are often different. Indeed Toshiro or Akemi was used also in Sengoku period, but it's uncommon for Daimyo and their wives. Those names sound more likely modern ones.
Okay. Are you saying I should use period-correct given names, like Ginchiyo and Nobunaga, or modern ones, like Toshiro and Akemi?
 
There were several female lords of a castle in sengoku period, but they were not daimyo class. They used usual female names like おつやの方(艶) O-tsuya-no-kata(Tsuya), 氏姫 Uji-hime or 誾千代 Ginchiyo, as you wrote, as the lord of their clan. There was no female daimyo as far as I know, but I think it would be no problem with "Ryū + an honorific title". In the first place, you keep the name at least for 5 years, right? Is it OK to abandon your idea now?
 
Back
Top Bottom