Purchase this article with an account.
Alejandro Lleras, James T. Enns; Negative compatibility in masking: unconscious inhibition or new feature priming?. Journal of Vision 2003;3(9):769. doi: 10.1167/3.9.769.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent reports have described a negative compatibility effect (NCE) in visual masking (Eimer & Schlaghecken, 1998, 2002). The sequence of stimulus events for NCE includes a brief prime stimulus, then a pattern mask, followed by a target that receives a speeded bimanual response. Of critical importance is that the prime-target relation can be either response compatible or incompatible. The main finding is negative compatibility: response time (RT) to targets following an incompatible prime is shorter than to targets following a compatible prime. This is interpreted as evidence of unconscious inhibition of the masked prime.
We hold a different view. Rather than the inhibition of unconsciously processed primes, we believe prime-mask interactions give rise to NCE. If the features of the mask include those of the prime and new target-relevant features, then positive priming can occur. The newly visible features of the mask will prime those of the subsequent target.
We tested this hypothesis in three experiments. Two different types of masks were used that were equal in overall energy: RELEVANT masks shared features with the prime and the target; IRRELEVANT masks did not share any prime or target features. In a prime-discrimination task, observers tried to discriminate between the two masked primes; in a target-discrimination task, they made a speeded response to the target.
The results were that (1) primes were less visible when followed by relevant than by irrelevant masks, (2) irrelevant masks yielded a standard positive compatibility effect in target-discrimination, whereas (3) relevant masks yielded the negative compatibility effect.
These results indicate that NCE is driven not by the identity of the prime, via unconscious inhibitory processes, but rather by the most recent set of task-relevant features to be displayed prior to the target. An interpretation of these results is given in terms of reentrant visual processing.
This PDF is available to Subscribers Only