Validity, Inference and Implicatures II

Abstract

In the first part of this paper, published in the previous issue of Crítica, the notion of logical validity and Grice’s theory of conversational implicatures were discussed a propos the validity of the logical rule of addition. This rule had been contested by Dr. Bunge in previous papers published in Crítica. T.M. Simpson had defended the validity of the rule against an objection of Strawson, The objection was that the conclusion of the argument P /∴ P or Q expresses a doubt not expressed by the premiss. Simpson understands “or” as “v” and has a clear idea of what “valid” means in logic. So, he concludes that the argument is valid. Then, he tries to show what is wrong with Strawson’s objection. He thinks that the argument is logically valid but psychologically implausible. A valid argument and a psychological inference are not the same thing. But this distinction does not prove that Strawson’s objection is incorrect, since it fails to show that a sentence of the form P or Q can be true when the speaker does not have the doubt expressed by the sentence, and so, that it can be true when P is true and the speaker knows it. In fact, the solution to this problem is to be found in a theory such as Grice’s theory of conversational implicatures. Another objection to Simpson’s solution is that it implies a false relation between valid argument and the process of change of beliefs called inference. The second part of this paper, published in this issue, discusses the relation between valid arguments and the psychological process of inference, understood as rational change of belief. Following an idea of Gilbert Harman, the paper criticizes the opinion that valid arguments are forms of inference. Different versions of this opinion are discussed in order to show that they fail to explain why any change of belief is rational. Lewis Carroll’s parable of Achilles and the Tortoise is used to show that any valid argument, taken as a rule of inference (i.e., as a rule warranting changes of belief) is incomplete. Achilles cannot write all the premisses of his inference down in his notebook: he has a complete, valid argument, but the Tortoise finds a belief without which the inference is not warranted. So Achilles writes this belief down in his notebook as a premiss of his inference, but, again, the Tortoise finds another belief that Achilles needs in order to infer the conclusion.

Later on, Harman’s holist view of inference is compared with partialist theories. The main reason not to accept partialist theories is that it is not possible to assess the rationality of any inference if only part of the beliefs of the reasoner are considered. There are two objections to a holist view that are discussed: the fact that we are not conscious of the totality of our beliefs when we infer, and the fact that not all our beliefs are relevant in every inference. The answer to the first is that the process of inference is not totally conscious, and is even sometimes completely unconscious. The answer to the second is that the sense of “relevant” that makes this objection plausible is not equivalent to the sense of “premiss” when we say that all the beliefs of the reasoner are premisses of his inferences.

[Summary by Hugo Margain]