Perceiving the logical status of sentences

Abstract

Two experiments confirmed Osherson and Markman's (1975) finding that children find it very difficult to perceive when the truth or falsity of sentences is self-evident (contradictions and tautologies) and when it is open to testing. The study also addressed two claims made by these authors: (a) that one source of difficulty is with taking an ‘objective’ mental stance towards the sentences, treating them as objects for assessment distinct from their referents; (b) that it is the nonempirical nature of the sentences which is responsible for much of the difficulty, not the ‘logical words’ (e.g., ‘and’, ‘not’, ‘or’) which the sentences contain. Claim (a) received some partial support: although manipulating the objectivity of the sentence presentation medium did not affect performance, removing the referents did. Hypothesis (b) received no support for the following reasons. First, the subjects also interpreted empirical sentences nonempirically, frequently justifying their decision. Second, the tautologies were far more difficult to assess than the contradictions, despite the fact that both kinds of sentence were equally nonempirical. The crucial difference between the tautologies and the contradictions was that the former contained disjunctions and the latter conjunctions. Subjects' responses on the empirical sentences suggested that they were interpreting ‘either … or’ constructions empirically. The possibility was discussed that the semantics of disjunction could explain this strategy. Finally, this semantic explanation of the problem with tautology was contrasted with a cognitive account of it.