"Pay attention to that alley over there. A gorilla in a clown suit is going to come out of it in a second." If I said that to you, and the ape subsequently appeared, you would (rightly) conclude that I'd helped you see the beast by focussing your attention on the right place at the right time. Suppose, though, that I'd said to look at the alley after the gorilla rushed by. And then you realized, hey, I saw a gorilla over there! That would be weird right? After all, if awareness is a straightforward record of what we've seen, then you either see the gorilla or you miss it. Outside of dreams, there's not supposed to be "miss the gorilla, look where it was, then change the record to say that you saw it." But, this paper suggests, intuition is wrong, and this dreamlike revising is indeed a part of everyday waking perception. In other words, it isn't always seeing a gorilla that makes you focus on the alley. Sometimes, focussing on the alley makes you see the gorilla.
In the study, published last month in the journal Current Biology, Claire Sergent and her colleagues had 18 volunteers look at a computer screen on which were two circles. After a moment, one of the two circles filled with fuzzy parallel lines for a mere 20th of a second. The volunteers then had to say which circle contained the lines and how they were oriented (vertical, horizontal or various kinds of diagonal). People's ability to answer correctly was much enhanced if they had a cue—if the correct circle dimmed a bit just before the lines appeared—thus drawing their attention.
Remarkably, though, people also did better on the task if the relevant circle dimmed about a half second after the lines had appeared. As Sergent et al. write, "this suggests that, contrary to a commonly held assumption, postcueing can influence perception itself."
You might think that this cueing effect was a sort of hallucination—that people could have imagined lines only because, and wherever, they'd seen the circle dim. But Sergent et al. eliminated that possibility by dimming both circles in one set of tests. In this case, people still saw the lines in the place where they had really appeared before the cue. They weren't following the cue alone to imagine something wherever it appeared; rather, the focussing device was leading them to see something that really had appeared—but which was no longer visible when they realized they had seen it.
There has been a lot of research over the years on perceptions that never reach awareness. For example, people whose visual cortex is damaged are unaware of seeing anything, yet they often react to things in their field of vision. Sergent et al.'s study is significant because it's not about these kinds of perceptions outside of consciousness. Instead, it's a demonstration that the mind can edit perceptions before they reach consciousness. Therefore Sergent et al. also had to eliminate the possibility that these results were a kind of blindsight, in which people were giving correct answers without knowing how or what they knew.
To address this, the researchers ran the experiment on another 18 people, and this time, added a measure of their awareness: in addition to saying where the lines were and how they were oriented, volunteers also had to rate how visible they were. There was a certain amount of projection, or "response bias," here, with people reporting better visibility wherever there was a cue (even when no actual lines had appeared). Despite that noise, though, there were "drastic improvements" the reported visibility of lines that had appeared and vanished just before the attention-focussing cues. In other words, post-event cues weren't just causing people to see lines in the right places; those cues were also causing people to know that they'd seen something. Which suggests that the reality you see around you isn't a "raw feed" of data coming from the eyes and visual centers of the brain, but a product that has been edited—and (as Dan Dennett points out here) is still being edited, by "you," even as "you" consume it.
Follow me on Twitter: @davidberreby
Sergent, C., Wyart, V., Babo-Rebelo, M., Cohen, L., Naccache, L., & Tallon-Baudry, C. (2013). Cueing Attention after the Stimulus Is Gone Can Retrospectively Trigger Conscious Perception Current Biology, 23 (2), 150-155 DOI: 10.1016/j.cub.2012.11.047