The pronunciation of intervocalic ん is a topic by itself, with the surrounding vowels tending to become nasalized, and ん itself maybe being pronounced as a nasal vowel. But that's not what I'm asking about here.

In 千円{せんえん}, I sometimes hear a "y"-sound inserted after the first ん, although I believe I don't pronounce it that way myself. On the other hand, in the city name 田園調布{でんえんちょうふ}, I don't hear this "y". (I hear this word less often, though).

At first, I thought it had something to do with whether the え came from ゑ or not, but it seems both 円 and 園 used to be with ゑ.

Am I just imagining things, or are the cases where this "y"-sound is and is not inserted? If so, what determines this?

  • do you have other example than 円 when the 'y' is pronounced ?
    – oldergod
    Commented Feb 17, 2012 at 4:26
  • 1
    I don't know about phonology, but in 千円 (SEn.en or HLLL), the accent is on [se], and if the [n] is weakened, then can it be considered that insertion of [y] is a strategy to repair hiatus? On the other hand, the [n] in 田園調布 (deN.ENCHOofu or LHHHHLL) has accent, so it is not weakened, so a hiatus does not arise. Insertion of the glides [y] and [w] seem to be generally used to resolve hiatus in Japanese.
    – user458
    Commented Feb 17, 2012 at 4:26
  • @sawa. That's an interesting theory. I don't think I pronounce the "y" myself, so I have to speak from memory, but I don't remember hearing it in 三円, probably because the preceding vowel is different, making it less necessary to insert the "y" to prevent えんえ turning into one long nasal e-sound. As for low-pitch えんえ, the only other example I can think of is 扁桃腺炎, but I don't hear this very often, so can't really be sure. I feel a "y" would be unnatural here, though. Looking up 炎, it didn't clasically have a ゑ, so I don't know if that could be relevant.
    – dainichi
    Commented Feb 17, 2012 at 6:05
  • With 三円, the preceding vowel [a] is a back vowel, so the glide to be inserted, will not be [y], but will rather be [w], if any. In hiatus resolution, the inserted glide seems to assimilate in backness with the preceding vowel. So after the front vowels [i, e], the front glide [y] will be inserted, and after back vowels [u, o, a], the back glide [w] will be inserted.
    – user458
    Commented Feb 17, 2012 at 6:33

2 Answers 2


Update: This might interest you: On the Phonological Derivation and Behavior of Nasal Glides (R. L. Trigo Ferre, 1988). (PDF available here, I assume legally.)

I claim that nasal ‘absorption’ occurs when the occlusion of a nasal stop is removed or weakened considerably to the point where it is a glide. A nasal glide without any place features, [N], is shown to derive by a process which simultaneously reduces obstruent stops to glottal stops in Japanese, Choctaw and Malay.

I do not have the background to evaluate the whole of Trigo Ferre's argument, but my understanding from scanning the parts that relate to Japanese is that she would argue that the "y" you are hearing is the intervocalic /N/, or at least all that remains of it (apart from some vowel nasalization) in the word /seNeN/.

Original comment: I am not an expert on Japanese phonetics, but this page has some data on similar words that suggests that nasalization is the key here (edit: which, of course, you did bring up; but I'm suggesting that it is the cause of what you're asking about).

  • 真一 /siNiti/ → [ɕiĩtɕi]
  • 本 /hoN/ → [hõ] or [hoɴ]
  • 翻案 /hoNaN/ → [hoãɴ]
  • 本意 /hoNi/ → [hõĩ]
  • 電話 /deNwa/ → [deẽwa]

Maybe what you are hearing is not a glide, but rather the onset/release of nasalization (in [seẽɴ] or [seẽeɴ] or whatever).

Or, to restate sawa's suggestion in comments, it could be that the rule "/N/ turns into a nasalized vowel" is leading to a two-vowel sequence like [eẽ] or [ẽe], and people are inserting a glide to keep the two vowels apart (prevent "hiatus").

The difference between 千円 and 田園調布 could be due to accent as sawa says, or other factors (e.g. 千円 is arguably two words, 田園 is not; 千円 is very common, 田園 is not; etc.)

Incidentally, although I am not a native speaker... when I listen closely to myself saying 千円 what I hear is a narrowing (but not a stop!) in the uvular region, which is then released back into something more like a regular [e], and it is that release which sounds something like a "y". But it is definitely distinct from the palatal [j] in, say, 矢 /ya/.

  • thanks for the writeup, but as I hinted in the question, I'm quite aware of the nasalization. Could definitely be useful for others, though.
    – dainichi
    Commented Feb 17, 2012 at 7:06
  • Yeah, sorry, I didn't mean to seem patronising. I do think that what you describe is emergent from vowel nasalization -- the writeup was intended to make it clearer exactly why (e.g. phonetic phenomena arising from the shift from nasalized to non-, or vice versa).
    – Matt
    Commented Feb 17, 2012 at 7:17
  • No no, no need to be sorry, just not exactly what I was looking for. And I definitely agree that the "y" sound comes from the nasalization. Without it, /eNe/ might turn into one long 3-mora nasal e-sound. And with /eNeN/ it could be even worse, depending on how you pronounce the last /N/. I was just wondering whether this had been studied, since I wasn't able to find anything about it. Unfortunately, the page you link to doesn't have any ンエ examples.
    – dainichi
    Commented Feb 17, 2012 at 7:43
  • @dainichi Even though 円【えん】 (your example) and 電話【でんわ】 (one of Matt's examples) are written differently in Japanese (even in supposedly phonetic hiragana), I think they are written more similarly in... actual phonetics/IPA. Therefore I think it's still possible that the example of 電話 is relevant.
    – atlantiza
    Commented Feb 17, 2012 at 7:55

I think a word like sen does have some nasalisation, but I also think there is a bit of a velar approximant (a faint ng) at the end, and it's this slight movement of the tongue that cause a subtle palatal y to come up, though not the full one found in English.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .