Four speech production experiments were conducted to examine how adults produce preverbal messages involving comparisons. It was argued that the generation of any message involving a comparison involves three decisions. First, a dimension for the comparison must be selected. Second, a contrasting object for this dimension must be selected. Third, a referent must be selected for the contrasting object. Participants were shown three objects on a computer screen and were asked to compare two

more »

... s along the dimensions of size (Experiments 1 and 2) or hue (Experiments 3 and 4). For example, a participant might be asked to compare the size of a medium-sized snake to either a small fish or a large bird. With each comparison, participants produced a noun (fish, bird) and an adjective (bigger, smaller) that could be repeated or switched from one trial to another. Experiment 1 showed a large tendency to repeat nouns, suggesting that speakers were repeating referents. Experiment 2, however, showed a large tendency to repeat comparisons to objects of the same size, suggesting that speakers were repeating contrasting objects not referents. Experiments 3 and 4 showed that the repetition effect disappeared after one filler trial. This suggested that decisions were made in short-term working memory. It was concluded that these three decisions are both necessary and sufficient for the generation of a preverbal message involving any comparison.