Semantic Encoding and the Effects of Formal Intralist Similarity
Benton J. Underwood, Arnold M. Lund · The American Journal of Psychology · 1980
The central question was whether subjects would use semantic encoding to avoid interference. The learning task consisted of eight word triads. On test trials the subjects were given the 24 words and their task was to assemble the triads. In one set of conditions the intralist similarity was increased by increasing the number of repeated words making up the triads; in a parallel set pairs of homonyms replaced the repeated words. Semantic encoding could (ideally) eliminate all interference resulting from the acoustic-articulatory identity of the homonyms. If the effect of using homonyms is to bring semantic encoding into heavy use, the effects of intralist similarity should be less for the homonyms than for the repeated words. The results showed that performance was poorer for the homonym lists than for the repeated-words lists, indicating minimal use of semantic encoding as a means of reducing the interference.