Masked primes can be genuinely semantically processed: a picture prime study.

Van den Bussche and Reynvoet (2007) argued that since significant priming was observed for novel primes from a large category, subliminal primes can be processed semantically. However, a possible confound in this study was the presence of nonsemantic effects such as orthographic overlap between primes and targets. Therefore, the first aim of this study was to validate our previous claim when nonsemantic influences are avoided. The second aim was to investigate the impact of nonsemantic stimulus processing on priming effects by manipulating target set size. The results showed that when nonsemantic effects are eliminated by presenting primes as pictures and targets as words, significant priming emerged for large stimulus categories and a large target set. This cannot be explained by nonsemantic accounts of subliminal processing and shows that subliminal primes can be truly semantically processed. However, when using a limited amount of targets, stimulating nonsemantic processing, priming disappeared. This indicates that the task context will determine whether stimuli will be processed semantically or nonsemantically, which in turn can influence priming effects.