Semantic Satiation: Replication and Test of Further Implications

Abstract

A set of six basic list words with corresponding lists of: I. close semantic similarity (S-words); 2, word associates (A-words); and 3. controls (C-words) with neither of the foregoing properties; was selected. Three experimental lists containing two words of each type of correspondence were constructed. A nine-scale semantic differential representing the three major factors of connotative meaning was assembled. Thirty male undergraduates were randomly assigned to an experimental list. Ss rated the basic list and the corresponding list words on the semantic differential first. Words and scales were presented individually on cards for the satiation condition, S first repeating a B-word for 15 sec., then immediately rating a corresponding word against a single scale (54 pairings altogether). Immediately after all the pairings were completed, S rated the B-words on a standard semantic differential form again. Polarity difference scores (difference between pre- and post-satiation ratings) were obtained for all words for all 5s. The binomial expansion was used to test for differences in rating change from zero for the basic and corresponding list words. Sign tests were used to contrast the different word types. There was a significant satiation effect only for B-words, but since even these were not significantly different from the controls, one hesitates to admit wordspecific satiation effects. An exact replication of Lambert and Jakobovits' experiment is dictated by the very weak support found in the present experiment. If the satiation phenomenon is found to be replicable on a Minnesota sample, then Osgood's theory of representational mediation processes should be questioned as the satiacion effect should generalize to semantically similar words-which it definitely did not in the present study.

Keywords

Affiliated Institutions

• University of Minnesota System US

Related Publications