Thomas Liu

In two experiments, subjects’ task was to decide whether a binocularly viewed target word was evaluatively good (e.g., fame, comedy, rescue) or bad (e.g., stress, detest, malaria) in meaning. Just prior to this target word, a priming word was presented to the nondominant eye, and masked by an immediately following presentation of a letter—fragment pattern to the dominant eye. (Masking effectiveness was demonstrated by subjects’ failure to discriminate the left vs. right position of a test series of words.) In Experiment 1, which used evaluatively positive or negative words as priming stimuli, judgment latency for the evaluative decision task was facilitated by primes that agreed in evaluation with targets, and was retarded by primes that disagreed in evaluation with targets. This result demonstrated that the evaluative meaning of priming stimuli was processed under conditions that prevented subjects from detecting their presence. Primes in Experiment 2 were 2-word strings for which the evaluative meaning of individual words was orthogonal to the evaluative sentence meaning (e.g., the two evaluatively negative words, "enemy fails,“ make up an evaluatively positive sentence). Results of Experiment 2 indicated that masked priming effects were influenced by the evaluative meanings of individual words, but not by their sentence meanings. This result supported the conclusion that the processing of undetected, dichoptically masked words is limited to analyses that are not powerful enough to extract sentence meanings.