Comprehension and Recall of Sentences

Preliminary experiments showed that 5"s better recall a noun pair if they generate their own linking sentence for the pair than if they merely read an equivalent linking sentence. Initial attempts to explain this effect in terms of memory search activities or idiosyncratically high-associative mediators proved unproductive in later experiments reported here. The hypothesis was then offered that the generate vs. read conditions differ in comprehension of the sentences and that comprehension aids retention. Subsequent experiments on incidental learning showed that recall is excellent when S is set to process a sentence in different ways designed to promote comprehension of its meaning, whereas equivalent exposure to or mouthing of the words in control sentences without comprehension produces relatively little recall.