Search Ace Linguist

June 25, 2024

Karuta, a Competitive Phonetics Game?

In the anime “Chihayafuru,” a team of high schoolers plays a card game called ‘karuta’. It’s based around hearing a reader read a poem aloud, and then finding the card corresponding to that poem on the floor. You can see an example in the following illustrative video:

There are 100 poems in the version of karuta played in the anime. The phonological qualities of the poems matter - for example, 7 of the poems begin with unique syllables (u, ho, me, mu, sa, se, su), so if you hear one of those syllables, you don’t have to wait to hear the rest of the card. Others share the starting syllable, so you have to wait longer, such as with the ‘chi’ cards.

One of the conceits in the game is that the main character, Chihaya, has amazing hearing. Other high-ranked characters also have great hearing. Some of the examples of this hearing:

  • Chihaya can hear the sound before the f.
  • The meijin can hear the difference between “su” and “se”

The ability to be able to anticipate a sound before it is completed grants you a competitive advantage. Is this actually linguistically possible?

Because of anticipatory coarticulation, it may be possible that the ’s’ in ‘su’ and the ’s’ in ‘se’ sound different. There is lip compression involved in the Japanese ‘u’ sound that is not present in the ‘e’, which could theoretically affect the ’s’ in ‘su’. Someone with very good hearing may be able to notice it.

What about ‘the sound before the ‘f’? This must be interpreted as some sort of artistic license, as there is no sound before the production of the Japanese bilabial fricative. Perhaps what is meant by this is that Chihaya can tell based on a very short sample and no vowel that the sound is going to be a bilabial fricative and not a glottal fricative or devoiced vowel.

Is there linguistic evidence that coarticulation affects Japanese consonants this way? I couldn’t find any articles on anticipatory coarticulation affecting consonants in Japanese, so I can’t tell you. Looking at the linked video, you can see that the participants do react rather quickly. If you'd like to see a high-level match commentated with English subtitles, you can look here also:

It would be interesting to ‘port’ this game to different languages with a different set of poems, and see how high-level players react. I also wonder if you could make a game like this that's cross-lingual - IPA recognition? Karuta isn't just about listening, but also about card layout and reflexes, so you could use that as a base to avoid making the game just about who has the better listening recognition.