In proof theory, the structural inference rule
Cut differs from other rules encountered in sequent calculi in an important way. In the other rules, e.g., a sequent calculus for classical logic, inferences preserve subformulae from upper to lower sequent. For example, in applications of an elimination rule, every subformula in the auxiliary formula appears in the principal formulae. In the above representation of cut, however, the formula (the ‘cut formula’) and its subformulae are completely absent in the lower sequent (unless they appear in some side formula).
Hence, it has historically been an important task to prove that a sequent calculus allows cut elimination, that is, that when cut is not taken as a primitive inference, cut is still an admissible inference. When a sequent calculus has cut elimination, any proof that employs the cut rule can be emulated by a proof in which cut is not employed, that is, if a sequent is provable, then it has a cut-free proof. Logician Gerhard Gentzen (1909–1945) proved cut elimination for his sequent calculi for classical and intuitionistic logics in the famous Hauptsatz or cut elimination theorem.