Japan's Must-Read Magazine

Japanese Pronunciation Pt. 4: non-initial “consonant” sounds

In this installment I’m going to look at 3 more features of Japanese pronunciation that all involve non-initial “consonants.” I’m putting the word “consonant” in quotes here because these sounds are actually quite difficult to simply categorize.


First, we have the semi-vowel spelled as y in romanized Japanese, and found in the hiragana characters や、ゆ and よ. I discussed this sound used as an initial “consonant” in the last installment—in that case, it is pretty much exactly the same as the English y sound as in “yes” or “yellow”–a very short [i] sound. No difficulties there.


This sound does become problematic, however, when it follows an initial consonant sound, as in words like “Tokyo” or “Kyoto,” as well in many other words like kyoka (permission) or gyuunyuu (milk). When English borrows words from Japanese, we tend to render this y sound found in the middle of a word as a pure vowel sound “toh-key-oh” or “Key-oh-toh.” This is fine in English, but it can be a problem if carried over to Japanese. In Japanese the sound is exactly the same as it is when it begins a word: it’s a consonant (or more properly, a semi-vowel) that does not add an extra syllable to the word. So, it’s not “Key-oh-tow,” it’s more like “Kyoh-toh.”


This use of the “y” sound as a non-initial consonant rarely occurs in English, but you can find it in several common words like “curious” and “cute.” For example, it’s not “key-ewt,” it’s “kyewt” [kyut]. So, when you say words like gyuunyuu, just keep the word “cute” in mind, and you should do fine.


The second non-initial consonant issue involves doubled consonants, written in the kana systems as っ or ッ (the small versions of the symbols pronounced tsu when written full-size). This is not a sound at all, but rather an indication of a doubled consonant sound. It usually occurs in the middle of a word with more than one syllable: it can be found twice in the phrase chotto matte. The romaji practice of using a doubled consonant letter is quite useful here, because the consonant really does get repeated twice. This is quite different from English spelling, though, because in English doubled consonant letters usually indicate a change in the vowel sound—compare the a sound in “lady” vs. “laddy” and you’ll get the idea.


Actually, in Japanese, it does change the vowel sound too, but in a different way. The vowel sound is shortened, although “cut-off” would be a better way of describing it. In the word matte, first you get a very short mat-, followed by an instant of silence (considered a separate syllable in Japanese!), followed by te. This distinction is important because there are a number of pairs that are distinguished only by whether the consonant is doubled or not: kite vs. kitte (“coming” vs. “cutting”) or mitsu vs. mittsu (“honey” vs. “three”)


Finally, there is the n used as a non-initial nasal sound—either in the middle or the end of a word, written in katakana as ん. This sound is particularly confusing, partially because it’s usually written with the letter n but it’s not the same sound as the initial n sound, and partially because there’s nothing quite like it in English.


First, to get the idea of んas a final sound, put your tongue up flat against the roof of your month and say “nnnnn.” It’s kind of a humming sound, a lot like the English “mmmm” used to show agreement, except for it’s definitely a more n-like sound. One crucial distinction is that in Japanese, it’s much longer than the English final n, to the point where it’s actually considered to be a separate syllable. So, for example, the word zen actually has two syllables: ze-n. If you listen to Japanese people speaking slowly and carefully, you’ll notice that the n used as a final sound is actually pronounced quite separately from the sounds that precede it. Singing is another area where it is easy to hear ん pronounced distinctly by itself. In faster speech, the sound doesn’t sound as separate, but you should still be able to notice that it is much longer than the English n sound.


When used in the middle of a word, when preceding the consonants ch, d, j, n, r, t and z, the sound made by the non-initial n sound is quite similar to the English n sound, except longer. Beware, though, because it can precede the normal initial n in words like onna (female), it’s quite important to make the sound quite long in order to distinguish the two sounds. A good example is konnichiwa: the other day, I saw an album by an English-language rock band entitled “konichiwa,” which is pretty much how many English-speakers pronounce it: “koh neechy wah” with no ん sound at all. Actually, it’s ko-n-ni-chi-wa, with the first n being ん, and the second one being the normal initial n.


By the way, the technique of tapping out a beat while practicing Japanese pronunciation that I described in earlier installments of this series works quite well here—try saying konnichiwa as above with a separate beat for every syllable, and you’ll start to get it right.


also changes before b, f, m and p, or, more properly, becomes a different sound entirely. There is a district in Osaka usually written in hiragana as なんば, but spelled as both Namba and Nanba in the Roman alphabet. What’s going on here? Actually, in these cases, んjust becomes an m sound, albeit considerably longer than the English m. This really isn’t so weird: even if you pronounce Nanba as in English, the second n sound will still come out sounding like an m, unless you speak very slowly and carefully.


Likewise, before k and g, ん becomes something like a long version of the English ng sound. Again, this should seem too odd from an English-speaking perspective: compare the n sound in “ban” vs. “bank” or “sin” vs. “sink,” and you’ll see that English does exactly the same thing. In Japanese, a good comparison would be ken vs. kenka or bun vs. bunka. As always, just remember to make the nasal sound stretch out for a whole syllable.


What is fairly odd from an English-speaking point of view is that ん actually becomes a vowel sound before several sounds, namely s, sh, y, w, h and other vowels. For example, in the word densha (train), ん becomes something like an incredibly nasalized i sound, sounding something like “dayshah,” with the y sound stretched out very long and coming through the nose. This is one area where I really recommend just listening to how Japanese people pronounce words and imitating them as much as possible.


In the meantime, the full-syllable n sound that ん makes when it is the last sound of a word is a reasonable substitute until you figure out the nasal vowel sounds. Just don’t use the short English n sound, because it will often be impossible for Japanese people to understand. Taking the example of densha again, if you say it as in English as “denshah” with a normal short n sound, it will sound to Japanese people more like like dencha (this is because nch as in “bench” is a possible combination of sounds in English, but nsh is not!).


Really, as you can see, ん is not a single sound at all, but rather a generic representation for a group of syllable-length nasal sounds that can all either end a word or come in the middle of a word. Confusing, yes, but ultimately not that hard as long as you are aware that it is quite different from the English final n sound and spend some time carefully listening to how Japanese people pronounce it in various words.


The next and final installation of this series will look at two more aspects of Japanese pronunciation: syllable-timing and the “pitch accent” system, both which provide quite a contrast to their English equivalents.