Why does the wavefunction 'collapse' in relational approaches to quantum mechanics?

I argue that there is a straightforward way to understand the occurrence of wavefunction collapses or 'quantum events' in relational approaches to quantum mechanics: we necessarily encounter a discontinuity in our description when a system interacts with the reference relative to which we are describing it, since the reference system cannot be described relative to itself. This makes it clear how recent concerns around quantum events in relational quantum mechanics should be resolved. However, the solution requires accepting that quantum mechanics is not a complete description of all physical facts, and moreover I argue that this is most likely inevitable if we want to be able to give a precise description of quantum events.