Insight into a logical relation

Two experiments are reported which aimed to investigate factors affecting the gain of insight into the logical relation of implication. In the first experiment, subjects had to make a series of inferences about either a conditional sentence or a quantified sentence, both of which had the same underlying logical form. Under one condition the sentences had to be proved true, and under another condition, false. Proving a sentence false facilitated gain of insight, but the linguistic form of the sentence exerted no significant effect on the main dependent variable. In the second experiment, implication was not expressed as a sentence but was inherent in the structure of the task. The experimental material differed in complexity and allowed the cognitive load imposed on the subject to be varied. Results suggested that insight was not all-or-none. It was spontaneously gained when the material was simple, but temporarily lost when it was complex.