Authors:

NORRIS, D. & Kinoshita, S.

Reference:

Year of publication:

2008

CBU number:

6747

Abstract:

In this paper we argue that perception is Bayesian inference based on accumulation of noisy evidence and that, in masked priming, the perceptual system is tricked into treating the prime and target as a single object. Two algorithms are considered for formalizing how the evidence sampled from a prime and target is combined, and only one of the algorithms is shown to be consistent with the existing data from the visual word recognition literature. We then incorporate this algorithm into the Bayesian Reader model (Norris, 2006), and confirm its predictions in three experiments. The experiments show that the pattern of masked priming is not a fixed function of the relations between the prime and target, but can be changed radically by changing the task from lexical decision to a same-different judgement. We conclude with discussion of the implications of the Bayesian framework of masked priming for unconscious cognition and visual masking.