Putting interaction theory to the empirical test

I report an empirical study deriving from a Gestalt-Interactionist approach to metaphor. Both the type of figurative expression (metaphor or simile) and the form of the expression (A is B or B is A) were manipulated in a factorial design. Subjects were asked to evaluate a given figurative expression both with regard to complexity and interest, and in terms of the degree of imageability of the tenor and the vehicle. As hypothesized, the design factors interacted in their influence on these ratings. Specifically, both the metaphor in standardform and the simile in reversed form received relatively higher ratings in degree of interest aroused and degree of complexity, while receiving relatively lower ratings in degree of vehicle imageability.