Semantics vs. proof theory
Dialogues themselves do not provide the information to determine the consequences of a set of premises, but dialogue strategies
Dialogue strategies are however only declaratively defined, just as extension-type semantics
So why are dialogue-style definitions called proof-theoretic?