In this work I present some theoretical considerations about what I consider to be a permanent and ever-present feature of verbal irony, namely, inferred contradiction, which has to be distinguished from plain, direct (non-inferred) contradiction as well as from indirect negation, for a contradiction which is directly expressed cannot be interpreted as ironical (since it lacks a crucial component: inference), and an indirect negation may or may not be ironic (depending on the situation), and thus cannot be considered a permanent feature of the phenomenon.