Japanese Phonetics Episode 33—LIVE (Updated with native recordings!)
Added 2017-12-18 14:43:56 +0000 UTCJune 4, 2021 Update: In the previous version of this lesson, I talked about how the initial い in 行くか often gets partially dropped in rapid speech, and attributed this to devoicing. However, after doing additional research, I've come to the conclusion that this is more of a form of slurring / slang than devoicing, and thus removed this section from the lesson. To clarify, I no longer believe that devoicing occurs at the beginning of words, though word initial い certainly does sometimes dropped in the form of slurring or slang（食べていく？ or, 'Want to eat something before you go?' often becomes 食べてく？, for instance). In terms of pure devoicing, I believe the following to be the best way to describe things:
い and う are devoiced when between voiceless consonants, or when they are preceded by a voiceless consonant and followed by a pause (source below).
I have updated the lesson 33 follow-up to reflect the above revisions, and as part of this, removed a small section in which I said the つ in 失礼 may be partially be devoiced in very rapid speech.
Nihon Onsei Gakkai (1976:748) and Tanaka and Kubozono (1999:8) cited by Vance (2008:206—207). Japan Foundation (2009:52—57).
Bibliography
Japanese Phonetics Index Page
The thirty-third episode of Japanese Phonetics, now updated with native recordings, is live! In this lesson we look at the extremely important linguistic phenomenon known as devoicing. This concept was a game changer for me, and I hope it will be for you as well. Apologies for the delay—this lesson ended up being by far the longest so far at over twenty minutes, so production was a bit more time consuming than usual.
As always, if you have any questions or concerns with regards to this lesson, please don't hesitate to leave a comment below. Thank you very much for your continued support, and good luck with your Japanese studies!
Best from Beppu,
Dōgen
Comments
It's good to know the rule behind something like this. I'm glad you're catching it!
Dogen
2025-08-19 09:20:27 +0000 UTCThank you very much for the useful lesson! I'm gonna start paying attention to this. The first time I noticed devoicing was while watching Naruto's anime :D (in the Italian dub Sasuke has the U voiced, while in the Japanese dub it's Sas'ke). Now I know there's a rule behind that!
Nicolas Rota
2025-08-10 18:52:28 +0000 UTCHi Jakob! I think you could be onto something here, but at the same time I can't comment with any authority on this matter as 1) Devoicing at the end of words followed by a pause seems to exhibit a high amount of volatility when it comes to the rule, and 2) In the past I've jumped the gun and said that a learner's observations are correct only to come across a counter point later. That said, Vance, in the Sounds of Japanese (P211, see bibliography in post above), notes that 'When す' [and presumably other う sounds?] is the last syllable in some other kind of word [that is to say, a word other than a ます form verb] before a pause, the vowel doesn't devoice as consistently, and notes that in 話す the す is likely to be voiced, as is the case with 烏 (からす）, or crow. So again, it is true that う doesn't devoice as consistently at the end of certain words, but I can't make a definitive statement about the final vowel of dictionary form verbs. Cheers!
Dogen
2024-09-26 01:09:17 +0000 UTCHi - at 23:40, you say that each of the "u"s in 尽くす could technically speaking be devoiced. However, based on my experience, it seems like the final vowel of a dictionary form verb is never devoiced (when not part of a larger phrase - unlike 行くか, where it is part of a larger phrase) - is this not accurate? I'm thinking for example of 行く, 貸す, 探す, 出す, etc. (In particular, I'm wondering if this is a general exception to the rule that "u" and "i" can be devoiced if sandwiched between an unvoiced consonant and a pause.) On that same note - at 28:58, the graphic marks the final "u" in "hikiokosu" as being devoiced, but it sounds to me like you're fully pronouncing it?
Jakob
2024-09-19 22:31:48 +0000 UTCWoah, I had noticed for so long people pronouncing "Sukiya" (the gyudon chain) as S-kiya, and I always wondered why. Now I know. Thanks Dōgen.
Polarwaves
2024-09-09 11:20:20 +0000 UTCThank you for the feedback. I'm glad this helped!
Dogen
2024-08-26 00:54:58 +0000 UTCThis is amazing! I've always heard something different about these sounds, but couldn't put it together, so therefore couldn't pronounce it correctly! Thank you, Dogen!
Gerry
2024-08-22 18:34:04 +0000 UTCHi, Gambel. JP here, Dōgen's business partner. I'm helping manage the messages here on Patreon. I asked Dōgen about this. He said he has not researched rendaku and hesitates to comment on this complicated phenomenon. He strives for accuracy when it comes to phonetics, and would need to commit himself to extensive research before he could confidently comment. I know this might not be exactly what you were looking for, but I hope it helps.
Dogen
2024-06-03 10:04:40 +0000 UTCDogen, I know that it's much more of a phonological process than a phonetic one, but I'm wondering if you have any thoughts on Rendaku. I was actually taking the highest undergrad level of phonology class at my university last semester and ended up writing my final paper on Rendaku, I think it's a really interesting phenomenon. Like I said it's more of a phonetic process (but arguably so is devoicing), I know since you have such an extensive background studying Japanese phonetics (and probably phonology too) I was just wondering if you have any thoughts on it haha, anyways thank you!
Gambel
2024-05-11 03:32:28 +0000 UTCThis reminds me of my several days long adventure trying to figure out how they slur 室 at the end of words. I've more or less settled on "stsu" for now. "Kyoustsu" (I made this comment before I finished the video, haha. I couldn't help relating so strongly to your 行くか)
Anthony
2023-11-22 00:55:51 +0000 UTCI'm glad it helped you take a step forward, and keep practicing your listening!
Dogen
2023-09-11 09:44:54 +0000 UTCThis lesson was great!!! I feel like I've unlocked something, I just need to remember to keep this in mind when listening
Sashin Exists
2023-08-09 06:52:03 +0000 UTCHi Michael! That's a very interesting question. It would seem that way based on a broader theory, but because 尽くす is a single unified word, where as 行く is a word and か is a separate particle, my gut instinct is that it's not ok to treat them the same, and I can't recall hearing any speakers ever saying 行くか with a voiced く. Apologies for not being able to help more with this, and best of luck with your future studies!
Dogen
2023-02-17 04:10:23 +0000 UTCHi, Justin. JP here, Dōgen's business partner. I'm helping manage the messages here on Patreon. I've felt that there can be value for some students to review the pronunciation sub-series, which starts with this lesson, before digging too far into pitch-accent studies in order to better understand the sounds that you are working with. I hope this helps, and I apologize for not catching your question sooner!
Dogen
2023-02-14 06:41:24 +0000 UTCYou mentioned that in the word 尽くす, the reasonくis pronounced could be because there’s an accent on くand it’s okay to voice it if that’s the case. Would that mean it’s okay to voice the くin 行くか？as well since there’s an accent on it?
Michael
2023-02-10 04:37:36 +0000 UTCI've mostly been going through these videos in chronological order, but it seemed that devoicing was pretty important and worth jumping ahead a little bit for (it was). Are there any other topics worth jumping ahead for?
Justin
2022-04-17 18:02:23 +0000 UTCHi Claes! Actually, when we speak in a deliberate manner, we often pronounce “ます” without devoicing. In that case, the downstep clearly occurs on ま. So first, practice the voiced version of ます, then try saying devoiced ます with that “feel” of the downstep! Probably you still won’t hear the downstep in your voice, but that’s totally fine as long as you say it with the “feel”. I believe that’s how native Japanese speakers (including me) pronounce devoiced version of ます whether they’re consciously doing it or not. Incidentally, Dogen did some interesting experiments using spectrograph in lesson 6.4. In the lesson, he compares the word 記者 in 頭高 version with 記者 in 尾高 version. https://youtu.be/H7ZEYrv_pqU?t=583 Even though the pitch of the devoiced ”き” doesn’t appear in the spectrograph, the following vowel ”a” suggests it. Therefore you can hear the difference between 記者（頭高）and 記者（尾高）. But in the case of ます, there is no vowel that follows the devoiced す, so nothing can suggest the pitch of す. I think that could be the reason you can’t hear the downstep in devoiced version of ます. I hope this helps! (^^) -Kimi
Dogen
2021-08-02 01:46:39 +0000 UTCHi Dōgen/きみ, I’ve been noticing the phenomenon since your lesson on devoicing. You say that a devoiced syllable does not have pitch. Does this mean that 食べます does not have a downstep? No matter how hard I try, I can’t make a downstep there. Neither does it sound natural, nor do I hear it in native speech. It makes sense though to have a downstep in 食べました and I have no trouble with it. This is the same for phrases like 学生です and 学生でした. Could you clarify?
Claes Wächtler
2021-07-29 10:27:13 +0000 UTCWow. I was completely oblivious to the fact that there were rules for this! I didn't even notice that the phrase 残念ですが should pronounce the う portion of す. Thanks so much for this. I have kind of been following Matt's mass immersion approach and delaying speaking, so I hope I won't have to struggle as hard with breaking bad habits, but I feel like I've even been doing listening wrong. XD
Rei
2021-07-16 18:25:45 +0000 UTCAs a beginner, I was also tremendously confused by the "every hiragana always has the same pronunciation" lie; it's tremendously confusing. This helps a bit, to understand that there's at least a pattern to when devoicing occurs. Something which I think is true, but I'm still a little confused about is if this happens as words connect. If you have a word ending in, say, a し, whether the い sound is devoiced depends on whether the next word starts with a devoiced consonant? (Actually, I see now this is answered in the next video, thanks!)
fred drueck
2021-05-30 05:56:53 +0000 UTCGlad to be able to help you connect the dots, Elise! Good luck with your future studies!
Dogen
2021-03-07 04:17:06 +0000 UTCI'm thrilled to have this level-up on pattern recognition. Devoicing was actually one of the first corrective tips I ever received. (That correction was prompted by a "わたくし"-sentence that I dared to utter based on proud self-study -- you can imagine how cringeworthy the wahta-kooshy was!) But the where-when-how has felt hit-and-miss until today. Eureka!
Elise Springer
2021-03-06 16:16:13 +0000 UTCGlad to hear that!
Dogen
2021-01-22 11:24:57 +0000 UTCNice to know my REALLY (like, before I had even memorized the kana) early basic phonetic studies paid off -- I caught on to the lie that they're "always pronounced the same" when I saw that ん had 3 or so different pronounciations.
switz mka
2021-01-18 09:23:19 +0000 UTCHi Victoria! Very glad to hear that you found the lesson so helpful! Indeed, once you properly know how to devoice certain sounds it can make many, many words much easier to say. I personally never struggled with the つ sounds, but I did with phrases like 行くか, as listed in this lesson. Thank you for your support, and good luck with your future studies!
Dogen
2020-09-21 05:15:38 +0000 UTCThis is so helpful omg! I knew there was devoicing going on in a lot of things but I could never figure out exactly what the rules were for what gets devoiced and what doesn't (I mean I never really looked closely to try to figure it out though I suppose). But this explains why I get tripped up so badly with 使う、作る、疲れる。 or, really anything that starts with a つ and is followed by a devoiced constant. I'm trying to voice the つ and it gets messy. I've always been more interested in structure/syntax and semantics/pragmatics in linguistics, didn't care much for phonetics. but Japanese phonetics are so useful to learn! Honestly I feel like we should learn at least stuff like places of articulation for things when we learn foreign languages so we don't just try to use the English sound that's associated with the letter they decided to use to represent a not English sound...
Victoria
2020-09-17 22:19:12 +0000 UTCI see! Thank you!
Strange Guy
2020-09-16 09:39:59 +0000 UTCYes there are, actually! The basic rule is that every other vowel should be voiced such that it's not just a bunch of whispering, and it's generally OK to choose which vowel to voice and devoice—there typically isn't only one correct way to devoice / voice the kind of string you presented. Cheers!
Dogen
2020-09-16 05:08:37 +0000 UTCGreat content! By the way, are there any rules when dealing with a sequence of high vowels + unvoiced consonants? Like some ししきくさ (I made up this word but you got the idea). I mean, I'm not sure that they devoice all vowels, though I may be wrong. Thanks!
Strange Guy
2020-09-15 20:15:25 +0000 UTCMy pleasure!
Dogen
2020-09-08 07:24:13 +0000 UTCWow, I never actually knew there was a set rule for this phenomenon, I’ve always just found myself pronouncing all イ段 and ウ段 kanas much more weakly than other kanas. Really appreciated it!
nivl
2020-09-08 00:25:27 +0000 UTCHi Jasper—there are many factors at play, such as exactly what the speaker is trying to get across, the speakers age, where they are from, etc., but generally speaking while it's certainly not a mistake to voice the う in です or ます, the devoiced pronunciation is more common and has a much more neutral meaning, which is why I personally recommend it for the majority of cases. Voiced う is usually used when the speaker is deliberately trying to get some kind of specific point across, so I would avoiding using it in conversation. Cheers!
Dogen
2020-08-22 03:58:03 +0000 UTCI was taught that the "u" in です and ~ます verb endings is devoiced, making it sound like "des" or "mas", but when watching anime I frequently hear the characters leave it fully voiced. This led me to assume that the full pronunciation of the す was done to make the character seem cute or quirky, since it was only something I had often encountered in anime. However, when I was watching a Japanese news program the other day, a politician was giving a speech and fully pronounced the す! I don't recall exactly what word it was, but I know it was either です or the ます form of a verb. Have I been wrong this whole time?!! I'm pretty sure that politician wasn't trying to sound cute or anything ^^; so is it just a personal preference thing as to whether someone leaves the す devoiced or not? I'd appreciate any input/advice you could offer!
RRainbow
2020-08-21 23:44:03 +0000 UTCIndeed! haha. Hope you enjoyed the lesson!
Dogen
2020-05-23 06:25:57 +0000 UTCI've never been more grateful for my decade long obsession with j-pop than right now. Vowels get messy there haha
Eloise Knuckey
2020-05-22 12:57:31 +0000 UTCHaha glad to hear that it helped so much! Good luck with your continued studies! Cheers!
Dogen
2020-04-12 11:34:38 +0000 UTCI have hated い when first starting to learn hiragana and listening to anime while trying to read along with Japanese subtitles. I knew there was something off about い which would cause me to slip up all the time trying to keep pace with the voice actor/actress. This helps a ton! Perhaps soon I shall except い and maybe one day learn to love い. The $10/month for the relationship counseling between us was worth it! You likely saved our relationship! lol ありがと。
アーロン
2020-04-12 10:42:44 +0000 UTCGlad to hear you found it useful!
Dogen
2020-03-21 23:45:42 +0000 UTCMind blowing!
Pierre-Alexandre Pons
2020-03-20 17:26:47 +0000 UTCHi Qsavr! I don't, but because English phonetics is very well documented it shouldn't be too difficult to find! I would recommend searching on Amazon for resources on 'English Phonetics'. Cheers!
Dogen
2020-03-06 01:52:17 +0000 UTCI really wish I had something like this when studying English. Even after speaking English for many years now, I still feel like I sound noticeably foreign when I record myself, and I have a very hard time making out the exact mistakes I'm making. I've only watched this series up until this episode, so I'm sorry if you already answered this question later on, but do you have any suggestions on similar linguistics/phonetics resources or programs for someone studying English?
QsavrQuest
2020-03-05 22:36:26 +0000 UTCHi Dawson! Glad to hear that you found the lesson so helpful! Cheers!
Dogen
2019-12-06 00:18:59 +0000 UTCMan for a while I kept trying to map the pronunciation of ですか on to ですが and I couldnt figure out why attempting to devoice ですが sounded so wrong LOL. This video was definitely very helpful and im excited to jump further into pronunciation!
Dawson Chen
2019-12-05 20:16:59 +0000 UTCHaha that does make sense, as the 'i' would come between two 'k's. Glad to hear that you found the content useful! Cheers!
Dogen
2019-12-05 01:39:04 +0000 UTCI feel like the final boss for devoicing is pronouncing the "Hurricane Kick" from Street Fighter. The day I can say that correctly is when I'll feel like I've nailed the concept haha!
MaximumPower
2019-12-04 22:36:50 +0000 UTCYes the rules do apply for after し and しゅ！Your examples are also correct, though the final vowel in rekishi is often only partially devoiced. Cheers!
Dogen
2019-11-26 23:27:27 +0000 UTCDoes the devoicing/elision rule apply to し as well? Does 鹿 get pronounced as "sh'ka" and 歴史 as "rek'sh'"? What about しゅ?
Abhi Ilindra
2019-11-26 15:55:36 +0000 UTCHi Abhi! That could indeed be the case, I was really just trying to explain how modern, written Japanese keenly shows the subtle differences between voiced and unvoiced sounds. Cheers!
Dogen
2019-11-26 04:42:44 +0000 UTCI thought the basic kana were created before voiced sounds had evolved in Japanese and later the handakuten and dakuten were made to accomodate them in Middle Japanese or something.
Abhi Ilindra
2019-11-25 16:19:35 +0000 UTCHaha Sorry about that Chris! Will try to find something a bit more soothing on version 2 of the series!
Dogen
2019-11-17 07:38:12 +0000 UTCLife changing information. I just wish you’d found some more soothing segue sound than that chuka chuka sound, it goes right through me! Lol
Chris Lane
2019-11-16 21:26:23 +0000 UTCHaha some information can really change the way you hear Japanese, right? Glad you’re getting so much out of the series basics!
Dogen
2019-07-14 10:38:30 +0000 UTCI tend to shadow every bit of Japanese I hear, so I guess it's only natural that I had already been doing this all this time, but it's still crazy to think about! There were a few moments when I'd think about why some words sounded that way, but quickly dismissed it and just accepted it. This series just keeps dropping crazy knowledge bombs!
thebasedpisces
2019-07-14 07:37:32 +0000 UTCHey Craig! Glad to hear you were able to avoid a lot of bad habits early on—they can be really difficult to fix. Thanks for your support and good luck with your future studies!
Dogen
2019-06-05 09:06:21 +0000 UTCLooks like I was definitely doing a bunch of this already but 100% due to osmosis. I'd learn something new, go wait that's not... Oh must be an exception like 「すき」and would confirm its pronunciation through forvo etc. I'm not sure why my other study's haven't explicitly stated this when it came up so long a go when you learn 「です」...
Hannah
2019-06-05 05:30:58 +0000 UTCHey Callysta! Glad to hear that you found it so useful! Devoicing is definitely one of the game-changer concepts in Japanese Phonetics. I had a similar experience when I learned about it—suddenly everything was much, much easier to say. Thank you for your support and good luck with your studies! ^^
Dogen
2018-07-15 01:35:34 +0000 UTCThis has been so great. I've actually been trying to learn this rap from a Japanese song and noticed I kept falling behind at some parts. After watching this, I realized they were all parts that should have been devoiced. Now I keep up perfectly haha! I know in the future this will help immensely, but I was surprised at such immediate results! Thank you soo much!!
Callysta
2018-07-13 20:05:22 +0000 UTCHi Virtually. I absolutely agree with this statement. Nearly every Japanese person I've met who speaks English well, and nearly ever foreigner I've met who speaks Japanese well is or was a musician. For what it's worth I took piano lessons for about six years, and music is still a major part of my life—just sent my $1200 custom made earphones in for repair as I can't stand listening to most songs with earbuds haha. I definitely don't have perfect pitch, and I'm definitely on the less-talented end among the musically inclined, but I still strongly believe the experiences I've had with music have only helped in improving my spoken Japanese. Cheers!
Dogen
2018-06-27 01:08:54 +0000 UTCHi Dogen, first of all, thanks for your work on this series, I've felt myself improve and feel more and more ready to speak with natives. Throughout the pitch accent series, I reflected on my absorption of the topics and my recordings and felt like my pronunciation, and ability to critique myself was of a good standard. While listening to this devoicing episode, I realised I'm already mimicking this phenomenon (of course, one needs to be aware of it, and this episode is pure gold) and enjoying speaking with a more naturally adroit tongue. My reason for saying this, is that I think my background as a musician has greatly aided my study of Japanese phonetics (as well as having a sensei with a knack for explaining phonetic phenomena). As part of an orchestra, one must listen to themselves as well as others, and I feel that this conscious listening has enabled me to become aware of phonetics and hear the truth of spoken Japanese as it hides behind the rules we're taught. I'm hoping it's not my ego talking, and that musically inclined individuals might genuinely find it easier to study phonetics. Regardless, I wanted to share my thoughts as it might be something you've considered over your years of study. All the best <3
Virtually in Japan
2018-06-25 19:11:22 +0000 UTCGlad to hear that it helped so much! ^^
Dogen
2018-06-18 07:50:27 +0000 UTCWow. Game changer. This explains so much of what I've heard but then really struggle to say!
Tom
2018-06-17 22:50:49 +0000 UTCditto for 主張...additionally this is why I'm always getting these two mixed up, it's because I can't say them because I'm trying to voice the しゅ。
tensaimon
2018-05-05 13:15:59 +0000 UTCLightbulb moment: this is why I can't say 出張, I'm always trying to voice the しゅっ！
tensaimon
2018-05-05 11:27:42 +0000 UTCHi Niklas. Thanks for the comment! I’ll be sure to cover 失礼 in the follow up. This word in particular can be a bit difficult because of devoicing and pitch-accent location. Will cover devoicing of other vowels a bit as well ^^ cheers!
Dogen
2017-12-23 05:41:17 +0000 UTCAs usual - great lesson!One word I was desperately waiting for during the lesson was 失礼 - most of the time I hear that as 'shtre' and would count that as devoicing the i of shi and simply dropping the s and u of tsu - but why oh why? Laziness? Another thing I have heard a couple of times now is a mysterious half-voiced え appearing where the u has been devoiced in です: two girls introducing themselves, first one 'A des', second 'B dese' - is this a dialect?
Niklas Michalopulos
2017-12-22 15:36:14 +0000 UTCThanks for the message HB. Really glad to hear it was so useful for you. Will use your input as well—stay tuned! Thanks for the support!
Dogen
2017-12-22 02:20:48 +0000 UTCI’m a relative beginner—just started learning this year—and this was a huge help for me. I had been trying to figure out why the ひ in ひと sounded so different than in ひらがな. Suddenly it all makes sense. I would definitely appreciate a follow up.
HB
2017-12-21 18:00:59 +0000 UTCHi Aleksey, Thanks for the in-depth reply. Was very interesting to read! As far as the pronunciation of devoiced vowels, I think that both camps put forward interesting points. I personally fall into the 'devoicing camp' rather than the 'dropped vowel camp', precisely because of what you mentioned at the end of your post. That is to say, even if vowels are, in fact, dropped—as opposed to devoiced—it is the vowel that changes the pronunciation of the preceding consonant, i.e. dropped vowel consonants still have vowel-specific pronunciation. In other words, I think that there are two ways of thinking about it: 1. Vowels become devoiced. 2. Vowels are dropped, BUT the preceding consonant is said as if the vowel is about to be said, such that a 'dropped き' sounds like a 'dropped き', rather than a 'dropped く'. While I believe that 2 is more accurate, I also think it's a bit intimidating for people who aren't familiar with phonetics. Moreover, and I think might be where we disagree, I personally feel that the 'devoicing mindset' does in fact lead to the correct pronunciation, whether that involves a devoiced vowel, or a fully dropped vowel preceded by a properly altered consonant, though I can certainly understand the benefits of covering both sides. Does this answer your question? Please let me know if this isn't clear! Cheers!
Dogen
2017-12-21 03:33:26 +0000 UTCWriting this down again (2nd time) because the first time I couldn't figure out how to send the message because I couldn't find a "send" button but then I accidently closed the browser (lol) so now it's all lost. ------------ Hey, I knew all of that! I actually spend about 2-3 months when I first started "studying" Japanese on just a bunch of listening and learning all that I could find about Japanese phonetics, so this phonetics series for me is more of a test, will see how I do! :D I wanted to add a few things. First is that vowels can still be "devoiced" when are between an unvoiced and a voiced consonant, it's also very common. For example ”ですね" has U in す devoiced even though there's a voiced consonant that comes after it. Another thing is that い and う are not the only vowels that can be devoiced. え also can be devoiced, but it's extremely rare, so I think that's why you didn't cover it in this lesson. I myself have only heard it a couple of times in about one and a half year of actively listening to Japanese. ------------ Now, there's one thing I actually wanted to ask you about before, but didn't feel like it was (appropriate? Fitting? Can't seem to find the right word for it.) to ask it because you were doing the pitch accent series, but now that you are doing the phonetics series and going over the devoicing part of it, I feel like it's the best chance I can possibly have to ask it. It's about how devoicing actually happens, as in what you actually do, to devoice something. On almost all recources that I used to learn about phonetics, they tell you that devoicing happens this way - In order to devoice a vowel, you pronounce it as if you were to whisper something, the tounge, the lips, everything is in the same position, you still pronounce the vowel, you just don't add voice to it. A bit after I read about it and started trying to copy that pronounciation and actively listen for it, I started to have doubts about it. First when a speaker was "devoicing" U in words like です or ます I never heard or saw that U. The lips didn't move into the position for U, neither could I hear it. Then, one thing didn't make sense, how come speakers can hold that S at the end of desu/masu if they still pronounce it the same way but just don't use their voice? That shouldn't be happening... So then I figured, that if I hear and see a speakers fully whispering, then I'll actually see if that statement about how devoicing happens is true or not. Because in theory, if the speaker is whispering, since everything is going to be whispered, one would pronounce all vowels and there won't be any devoicing in it, because literally everything the speaker says is already devoiced. So you should be able to hear the U in です, the I in 来た and so on. But you know, you don't just go around casually whispering and recording yourself and then putting yourself on the internet don't you? :D So didn't really have a chance to test that idea. And so I was about to give up on it for the time being, but just about at that moment, Joey (aka the anime man) made an ASMR video (or however those are called) in which he was whispering throughout the whole video in Japanese. And here's what I heard and saw - There was no devoicing, there was vowel dropping. If it was truly devoicing, I'd hear all the "devoiced" vowels, but I didn't. If it was truly devoicing, I'd see the mouth taking the position for those sounds, but I didn't. There was none of that. So after that I started to think that probably, the vowels are actually fully dropped, and not devoiced like I previously thought. And so I googled some more on that topic and I found a little bit of information that confirmed it. That yes, the vowel are indeed dropped and not "devoiced". I'd also like to add that, I was trying to learn to "devoice" vowels properly for about 3 months at that point, and I just couldn't do it no matter how much I tried. It always sounded a bit off and felt very unnatural. But after I tried dropping the vowels, everything started to finally sound right and speaking finally got significantly easier. I'm sorry if it's a bit unclear where I'm going with this, my thought are a little bit all over the place. I guess to put this in a more neetly tied question I'd like to ask your thoughts on it. How do you think the "devoicing" is actually properly pronounced? Sorry for a long post! I really can't think of a way to make it shorter. ------ I'd also like to add that, some people might say that if I just drop I in き and not "devoice" it, it'll sound wrong because I'll just say K, which would sound just like the english K in isolation. But that's not true. What I think is happening - the consonants that precede い are palatalized, so if you drop the い they still stay palatalized, so that きた doesn't end up sounding like くた.
Aleksey Kachnovich
2017-12-20 08:57:07 +0000 UTCHi Enrico, Thanks for watching! Glad to hear that you enjoyed it. The hand gesture was, indeed, a small comedic interlude. With regards to your question, I've actually heard conflicting theories on this one, some people say that the accent stays in place but is unvoiced, some say that it stays in place and is voiced to a certain degree (I tend to agree with this most, and some say if shifts (in which case there are usually multiple correct pitch-accent patterns, as in 父). Hope that this helps, and thanks again!
Dogen
2017-12-20 02:57:21 +0000 UTCHi Carol. Glad to hear that the lesson proved to be beneficial! There aren't a whole lot of questions so far (perhaps because the concept is quite straight-forward once you hear about it), so I'm still debating about what to do for the follow up. Will keep everyone in the loop, however! Thanks for watching!
Dogen
2017-12-20 02:54:19 +0000 UTCAbsolutely my pleasure!
Dogen
2017-12-20 02:53:15 +0000 UTCIt does indeed! Provided that the speaker is talking relatively quickly, that is. Cheers!
Dogen
2017-12-20 02:53:09 +0000 UTCLike others are saying, great lesson! I was in Japan recently and found myself stumbling over some phrases and wondering how native speakers seem to speak so much "faster" than I could - and now I know it was because I was not devoicing! Looking forward to the follow up!
carol kong
2017-12-19 23:27:39 +0000 UTCI have a question. I have my own guess, but I'd like your informed opinion on the topic. Do vowels also get devoiced when they are in between words? For example, in "毎日コンビニに行く", the final "i" of 毎日 is surrounded by two unvoiced consonants (ち and こ), so does the "i" become devoiced? Thanks!
Nicolas Guillemot
2017-12-19 21:50:13 +0000 UTCI was blind but now I see. Thank you for another eye-opening presentation!
Nicolas Guillemot
2017-12-19 20:26:51 +0000 UTCGreat lesson! As you said in the video, I've learnt devoicing over the years just by osmosis, but actually studying it a few weeks ago was still somewhat eye-opening as it gave me a theoretical framework, and I found your lesson extremely well done in that regard, so I very much enjoyed it despite already being aware of the contents... I very much wish your lessons existed several years back so I could retroactively profit from them! =P Also, the hand gesture for 行くか at around 10:02 floored me. Not sure whether it was intended as a brief comic "interlude" or if just came out that way, but in any case it made my day. Regarding whether to do another follow-up video on devoicing, I can't really think of any questions I'd like to be answered on the matter, but I was wondering if you had any plans to cover how a devoiced mora can shift pitch-accent for some speakers, for example 来た being pronounced (mostly by older speakers, as far as I know) with the downstep after た due to the devoicing of the き, or 父's downstep shifting after the second ち for the same reason. Again, it's not a request, I was just wondering if you had thought about covering this at some point, though I imagine it might be best saved for the advanced series. Thanks for all the great content, and keep up the good work! =)
Enrico Paolini
2017-12-19 15:59:34 +0000 UTCNot a problem. I can certainly hear a difference, and just confirmed that lip placement at least is different with a Japanese native. That said the difference isn't massive, so I can understand having difficulty from audio alone. Do you have any native around you who can voice these in person? I imagine you'll notice a difference immediately when looking at their lips!
Dogen
2017-12-19 10:33:36 +0000 UTCSorry for my spelling and laziness to switch to the Japanese keyboard, which caused the cofusion. Of course, I can distinduish [ɕ] from [s], but I was talking about [ɕi] and [ɕu] with devoiced vowels. For example is there a difference in the pronunciation between 失禁 and 出勤?
Dominik
2017-12-19 09:37:12 +0000 UTCHo Dominik, Thank you for the feedback. I'm not familiar with the terms 'idab' and 'udan', can you elaborate on these a bit more (I may have learned said concepts with different terminology)? With regards to differentiating between shi and su, would the following also sound similar to you? desu deshi Thanks again for the comment, feedback, and support!
Dogen
2017-12-19 09:22:04 +0000 UTCCouldn't agree with you more Miles—this should be in every first-year textbook! I definitely struggled with 七時 until I learned about devoicing!
Dogen
2017-12-19 09:18:13 +0000 UTCHi Bertrand, Thanks for the comment. To answer your question, I do believe that native English speaker's have more difficulty with devoicing, primarily because they have the tendency to map a strong accent onto the wrong part of the word, which affects how the rest of the sounds interact with each other. Glad to hear that you're excited for the upcoming lessons as well. I have a feeling you'll know most of the things I intend on covering, but hopefully the explanations are still beneficial!
Dogen
2017-12-19 09:17:06 +0000 UTCThank you for the feedback, Gilbert! There's one!
Dogen
2017-12-19 09:05:49 +0000 UTCThanks for the video! I'd be interested in palatalization. This would probably also tie in nicely with a further discussion of devoicing, as it seems to be the distinguishing feature between devoice イ段and ウ段 morae. While I can distinguish between e. g. devoiced き and く, devoiced し and しゅ sound basically the same to me.
Dominik
2017-12-18 22:31:26 +0000 UTCThis is one of those things that I think beginners should absolutely know. When I heard native speakers doing it, I always thought it was laziness, so I continued pronouncing words in what I thought was the "proper" way, only to realise that it wasn't laziness at all. I found that the on'yomi for 七 in particular is one you absolutely can't say properly unless you devoice the i in し
Miles Gard
2017-12-18 20:57:39 +0000 UTCGreat stuff! I consider myself an advanced learner, so, as you said, I'd already become aware of this through osmosis years ago, though I never realized this wasn't so much a shortening of the vowels as a devoicing phenomenon. Didn't know the rules either. It's always fun to learn the actual rule. Do you reckon 'overvoicing' is mainly a native English speaker's issue? Possibly because of English stress accent? As far as I can recall, I think I picked up on devoicing fairly quickly even though I was never formally taught it. The patterns seemed easy to recognize, and I don't think any of my French friends who learned Japanese ever struggled much with devoicing either. So it came as a surprise to me when you said that you yourself had trouble even hearing devoicing at first back in the day. It's one of those things I figured people realize very quickly even if they have a bad ear. Guess we all have our blind - or rather, deaf - spots! Speaking of... Pitch accent, on the other hand, is something I never realized until one of my teachers made me aware of it. [EDIT] Also, I'm very eager to see what other pronunciation lessons you've got up your sleeve. Most of it is stuff I should already know, if subconsciously, but it'll nice to get down to the nitty-gritty.
Toberu
2017-12-18 20:55:17 +0000 UTCI would love to hear more on devoicing going forward, but please also enjoy your holiday season!
Gilbert Y
2017-12-18 15:19:14 +0000 UTC
