Priming within the ‘Attentional Blink’ in an RSVP Task

Abstract

Recent experiments (eg M M Chun and M C Potter, 1995 Journal of Experimental Psychology: Human Perception and Performance 21 109 – 127; J E Raymond, K L Shapiro, and K M Arnell, 1992 Journal of Experimental Psychology: Human Perception and Performance 18 849 – 860) with RSVP (rapid serial visual presentation) suggest that the attentional blink is caused by local interference. We present data from three RSVP experiments that provide further clues regarding the attentional blink. In experiment 1, subjects detected an ‘X’ and then identified a red letter; in experiment 2, subjects had to say whether the first red target was an ‘X’ and then identify a red letter. In experiment 3, subjects identified two red letters. We systematically varied the lag between the first and second targets. On half the trials, we also primed the second target by placing an identical letter in the lag one position (the position after the first target). In experiment 3, we also examined if the priming effect was semantic with a lower case letter.

The first two experiments suggest that the priming effect is very short-lived and mainly sensory in nature. The priming effect disappears altogether if the first target is not present. More interestingly, we found that when subjects failed to detect the ‘X’, priming could still happen. The third experiment replicates and extends the results of the first two experiments. We also show that priming, albeit in a weak form, may still happen during the time when the attentional blink is supposed to occur. These results suggest that it is not an inhibition that causes the attentional blink and that sensory processing continues during the blink.