Small tsu confusion


I’ve understood up to now (Wanikani level 7, learning for N5) that the small つ often works as a kind of apostrophe, marking a double consonant where, effectively, some vowels have been dropped. This has made sense to me.
BUT now I’ve twice experienced a small tsu derived from a normal tsu and this is disturbing and I want to understand why!
活気> かつ+き> かっき??
出血> しゅつ+けつ> しゅっけつ??
Can anyone explain this?
Thank you!


I’m not sure about the apostrophe or dropped vowel. I always learned that the small つ before a letter is just a way of saying that there’s a hard sound there and a double consonant. It’s just how those words are spelled. It doesn’t necessarily guarantee that anything was dropped there, so there’s no rule there.

But perhaps I’ve just never noticed this, and a lot of words with the small つ do come from two things coming together. Interesting!


I guess it’s just because it makes the words easier to say.

かっき is easier to say than かつき, for example


I assume you’re referring to seeing something like 学校, where がく + こう = がっこう.
But this happens very often with つ changing into っ as well, such as in your example. It’s something you should get used to.

As @sigolino said, I think the most likely reason this happens is for ease of pronunciation. がっこう is much easier to say than かくこう. Similarly, しゅっけつ is easier to say than しゅつけつ.


活 is katsu and 気 is ki, but when you combine them, you get 活気 which is pronounced kakki.

出 is shutsu and 血 is ketsu, but when you combine them, you get 出血 which is pronounced shukketsu.

Sorry about using the dreaded romaji, but in this case, I think it helps to make things clear. Often, when the first kanji of a compound ends in つ, it changes to a small っ in the compound word and the pronunciation changes accordingly.


Thank you everyone!
I agree that it’s probably pronunciation simplification and a common occurrence. I will get used to it.
I think the relationship in Japanese between pronunciation and orthography with three alphabets (or four, if you count romaji for us foreigners) is a challenge I’ve not experienced in any of my indoeuropean languages!


Yeah, it was pretty weird to me at fist too. But once you see it a few times, you can not only get used to it, but also guess (with a pretty good success rate) if a reading is going to be shortened or not. Like:

“発表… I know the on’yomi for these kanji are はつ and ひょう, but はつひょう is too hard to say… Hey, I know! It’s probably はっぴょう!”

Just think like this: if you wish the reading was easier, it probably is, because people are lazy.


It is very unlikely to do with “ease of pronunciation” since in many cases a version of the sound exists the other way, there are places where a つ is followed by some “k” sound, etc etc.

Rather than trying to answer “why”, which is more or less impossible to answer, this is simply a lexical feature of Japanese in that it uses gemination to mark lexical differences. It also makes sense that in a language where vowel length is also a lexical feature, the length of consonants would matter as well.

Furthermore loans like スラッガー and キッド point to the fact that this is not a matter of “ease of pronunciation” as native words can not have a voiced geminate like this


You start to develop an intuition for it after a while, it just takes a bit of time. Once I got more used to Japanese say level 15-20 or so, I started guessing the small っ “contractions” correctly in the lessons!

There is one big rule of thumb you can memorize to speed things up (2 parts).

  1. After small っ, unvoiced sounds will more often than not contract.
  2. After small っ, voiced sounds will almost never* contract.

(*As @Syphus points out, voiced “contractions” can happen in loanwords. However, for the purposes of Wanikani, which is mainly teaching you native Japanese words, you can follow this rule).

Not sure about your linguistics background, so just for reference, P, T, K, S, F, Sh, and H are some unvoiced consonant sounds. Their voiced equivalents are B, D, G, Z, V, Zh, and I’m not sure about H lol.

Now with Japanese examples:
活気 = かつ + き. き starts with a ‘k’ sound, which is unvoiced so the っ contracts.
発表 = はつ + ひょう. ひょう starts with an ‘h’ sound, which is unvoiced so the っ contracts. And in this special case ひょ changes to ぴょ, presumably because it’s difficult to “double” an H sound.

発売 = はつ + ばい. ば starts with a “b” sound, so no “contraction” and no doubling. Big つ stays the way it is.

As for why this general pattern exists, I couldn’t tell you. But keeping this rule of thumb in mind should help you develop an intuition for the pronunciation of new words you come across. I hope this helps!


Also you made me think of one thing. It’s fairly consistent with all sounds except for “s”. The “s” sounds seem to go both ways much more often.


Thank you again.
I hadn’t thought of gemination being a lexical marker. But you’re right!
And the voicing issue is also interesting. I had thought of voicing regarding rendaku because voiced phonemes change unvoiced phonemes in many languages but I hadn’t found any real pattern. I’ll have to go look that up.