What Logical Consequence Could, Could Not, Should, and Should Not Be

In ‘Logical Consequence (Slight Return)’, Gillian Russell asks ‘What is logical consequence?’, a question which has vexed logicians since at least the twelfth century, when people first began to wonder what it meant for one sentence (or proposition) to follow from another sentence (or proposition, or set of sentences, or set of propositions), or whether it was possible to put down rules determining when the relation of ‘follows from’ (or ‘is antecedent to’) holds. Her aim is threefold: (1) to explain what an answer to the question ‘What is logical consequence?’ would need to be able to do in order to be a satisfying answer; (2) to identify previous answers to the question; and (3) to demonstrate why these previous answers are inadequate to do what the answer needs to be able to do, and to offer a new answer. In the present paper, I respond to these aims in two ways. The first is to say something about where Russell’s central question even comes from, because this is not a topic that is often discussed by twentieth- and twenty-first-century logicians, and even historians of logic tend to not have had much to say about when—and why—this question even comes about in the first place. The second is to evaluate the accounts proposed and discussed by Russell, including her new proposal. In the end, I will argue that she has reached the right account of the nature of logical consequence, but not necessarily for the right reasons.