Imagine you're creating a new language.
You'll be given pairs of items and asked which you would call “bouba”, and which you would call “kiki”.
After each question, you'll be shown what everyone else's answers were.