Abstract In a paper published in this journal, Golan argues for the thesis that there is no tenable notion of global metainferential validity. The conclusion rests on the premise that metainferential validity for propositional languages should be closed under uniform substitution of arbitrary formulas for atoms. This paper looks into Golan's justification for this premise in terms of the formality of logic. It is argued that neither formality as schematicity nor formality as topic-neutrality implies closure under uniform substitution of arbitrary formulas for atoms as requirement. It is also argued that the justification of such a requirement in terms of the formality of logic makes Golan's conclusion irrelevant for logical research.
Andreas Fjellstad (Tue,) studied this question.