Is it necessary to model suppositional arguments?
Some well-known rules of inference are easily represented in sequent calculus (?-Introduction, ?-Elimination)
Pollock: Reasoning can lead to a priori conclusions such as p ??p, that do not depend on any input