The meaning of if

Abstract

In line with the arguments presented in Part I of the book, this chapter argues that it is implausible and indeed unnecessary to offer a linguistic semantic account of conditionals. The bulk of the chapter revisits the relation between conditionals and the logic of Material Implication (MI). I argue that the perceived mutual exclusivity of pro-MI and anti-MI approaches is only apparent and stems from the anyway problematic a priori assumption that linguistic semantics exists/should exist. Once problems with this assumption are acknowledged, we can begin to consider other ways in which conditionals may ‘have something to do’ with the logical operator MI, without evoking linguistic semantics or compromising MI’s claims to supra-personality. Moving beyond MI, the chapter also deals with an assumption made by both truth-functional and non-truth-functional approaches: that a conditional is false if the antecedent is true but the consequent false – what I’ll refer to as the ∼(p & ∼q) constraint. Focusing on conditional promises and threats, I show that the ∼(p & ∼q) constraint does not uniformly apply to all conditional inducements; whether it does or does not depends on holistic assumptions of the interlocutors. Thus the ∼(p & ∼q) constraint cannot be a linguistic semantic constraint. I put forward a socio-cognitively grounded Inverse Effect Hypothesis and argue that it offers a more explanatory and more uniform account of conditional inducements than one which assumes MI or the ∼(p & ∼q) constraint as the encoded semantics of conditionals.