This, I hope, is close to multi-modal in lingual terms. There's potentially a lot to learn from examining where this works/fails :D