| In unconscious semantic priming, a visually masked word (the prime) facilitates semantic classification of a following visible related word (the target). Recent findings challenge a widely held view that analysis of primes occurs at the level of whole-word meaning. Instead, analysis may be sublexical; but even sublexical analysis apparently suffices to produce reliable unconscious priming when masked primes have been practiced earlier as visible targets. Four experiments tested whether reliable priming following practice is driven by analysis of only very small word units (individual letters or short multi-letter sequences). All four experiments used a two-choice, semantic classification task with visually masked, briefly flashed word primes that had been practiced earlier as targets. Experiment 1 found priming from words in which no single letter in its position was diagnostic for the identity of the prime; thus analysis must not be limited to individual letters. Experiments 2 and 3 obtained priming when none of the three two-letter sequences in four-letter words was diagnostic, showing that analysis is not limited to two-letter sequences. In Experiment 4, priming was obtained from five-letter words in which none of the three-letter sequences was diagnostic, showing that analysis is not limited to three-letter sequences. These results support a view of unconscious analysis as sublexical but operating on features potentially spanning the whole length of a word. |
