1985 Volume 56 Issue 5 Pages 269-276
This study examined Ortony's (1979) model of asymmetric similarity between constituents of metaphors In Experiment I, 35 subjects rated the similarity of topic (subject) to vehicle (predcate) in 96 metaphorical and anomalous sentences (Noun-A is Noun-B). Another group of 32 subjects listed features (attribute-adjectives) of each topic or vehicle, and the salience of features was measured by their frequency of mention. Multiple regression analysis indicates that the saliences of common features and of distinctive features predict the rated similarity well. Multiple discriminant analysis indicates that metaphor, reversed metaphor and anomalous sentence are distinguished in the relative degrees of salience of common features, as predicted by Ortony's model. Experiment II manipulated the salience of common or distinctive features by applying nine types of adjective modification to topic or vehicle (e. g., Adjective Noun-A is Noun-B). Twenty-eight subjects ranked these metaphors according to the similarity of topic to vehicle, and another group of 28 subjects ranked the same set of metaphors according to the comprehensibility as well as the intensity of interest they felt in each metaphor. Results show that the adjectives which cue (make more salient) common features increase similarity and comprehensibility, while the adjectives which cue distinctive features decrease them, but increase metaphor interest. These results are consistent with Ortony's model.
