Cover Story

Many of us have had the experience of walking into a familiar room--a friend's apartment, say--and immediately sensing that something is different, without quite being able to identify what that difference is. A moment later, the change pops into focus: "Oh yeah, they moved the sofa."

Now one researcher says that such sensations are evidence that people can be aware of--and can use--visual information without necessarily consciously "seeing." In a study published last year in Psychological Science (Vol. 15, No. 1), University of British Columbia psychologist and computer scientist Ronald Rensink, PhD, postulates that we have a distinct, fast-acting mode of visual perception that takes in visual input and leads us to sense rather than see a change in a scene.

That system, which Rensink calls "mindsight," could, he says, help explain why people sometimes have an intuition or "sixth sense" about a situation before they can consciously understand why.

"In a way, it's like a 'first strike' system," Rensink says, "that we use without conscious thought."

But some psychologists remain skeptical of Rensink's idea, saying that his data could more easily be explained by scientists' knowledge of how people use their ordinary visual system.

"This is a remarkable claim, and in order to make a remarkable claim you have to have really unimpeachable data," says Andrew Hollingworth, PhD, a psychologist who studies visual cognition at the University of Iowa. "I don't think we have that here."

The study

Rensink's mindsight theory originally grew out of his work on a different, well-established phenomenon: change blindness. In a series of studies over the past decade or so, researchers have found that, under the right circumstances, people can have a surprisingly hard time identifying even very large changes in a scene. For example, when people are asked to view two images flickering quickly back and forth--such as a picture of a dog in front of a bench and then that same picture but without the bench--but there is a distracter such as a brief gray screen between the images, many people will have a hard time identifying even such a seemingly obvious change.

Change blindness research forced cognitive psychologists to rethink some of their ideas about how vision works: In order to "see" something, it seems, we must do more than just look at it, we must also pay particular attention to it.

In 1997, Rensink was conducting a change blindness experiment when he noticed something interesting. Many of his participants would tell him that although it took them a long time to visually identify the change in the scene, they much more quickly had a sense that something about the scene was different.

Intrigued, Rensink decided to investigate this further. He ran the same experiment again: Participants watched two images flicker back and forth on a computer screen. One image depicted a statue in front of a wall; the other showed the same statue with the wall removed. Each imaged appeared for 240 milliseconds, and in between each a blank gray field appeared for 80 milliseconds.

Rensink asked participants to press a response key twice--the first time when they had a sense or feeling that something was different in the scene, and the second when they saw the change and could identify what it was. If they did not have a sense or feeling before they could identify the change, the researchers instructed them to press the key twice in immediate succession.

Each participant completed 48 trials. To ensure that the participants weren't simply guessing about the scene changes, Rensink included six "catch" trials in which the scene actually didn't change. Any participant who falsely claimed to see a change in more than 50 percent of the catch trials was labeled a "guesser."

Overall, Rensink found that approximately 30 percent of the participants reported that they could sense the change before they saw it in five percent or more of the trials, and there was often a time lag of several seconds between the moment they sensed and the moment they saw the change.

Those "can-sense" subjects, Rensink says, were using a sort of secondary visual system that works in parallel with our everyday vision. They were taking in visual information before they were consciously aware of it--and before they consciously saw it or had a "picture" of it in their minds. Instead, the visual information produced a nonvisual "sense" of change.

An alternative explanation

But University of Illinois-Urbana-Champaign cognitive psychologist Dan Simons, PhD, isn't convinced that any such system exists.

"If Ron's right in his interpretation, this is really fascinating and revolutionary," says Simons--but he adds that he doesn't think Rensink is right. The two are friends and colleagues; they've collaborated on review studies for journals in the past. But in this case, Simons thinks that his colleague's interpretation has exceeded the limits of what the data reveal.

"Any claim this extraordinary requires that we take a really close look at it," he says, "and in this case I just don't think the evidence is there."

Simons has no argument with Rensink's data. In fact, in a study that will be published in an upcoming issue of Psychological Science, he replicates Rensink's findings almost exactly. But, he says, the findings can be accounted for with a prosaic explanation: Participants who said they sensed the change before they saw it were simply taking time to verify their conscious detection of the change. In other words, they pressed the "sense" key when they first thought they saw the change and then pressed the "saw" key when they were absolutely certain of what they'd seen. Participants who never indicated that they sensed the change, Simons says, were simply more conservative than those who did--they wanted to be absolutely certain before they pressed any key.

"It's not necessarily clear to the participant in this task exactly what's meant by 'sensing,'" Simons says. "Their experience is that they're searching around, trying to figure out what's changing--Is it the dog? The mailbox? And there's a frustrating feeling of 'I know something is changing.' Ron was trying to tap into that feeling. But the question is, is that feeling based on some sort of real information, or is it just frustration?"

Simons particularly points to the fact that, in his study, participants who said they experienced mindsight were also more likely to give false alarms--saying that they saw a change in a catch trial--even if they didn't rise to the level of the guessers.

Also, Simons says, the cutoffs used in the original study were arbitrary--defining "can-sense" participants as people who indicated sensing in more than 5 percent of the trials. Change that number to 10 percent, he points out, and the number of people who demonstrate mindsight drops precipitously.

But Rensink remains convinced that his interpretation is correct.

"To me, it doesn't make sense that people would talk about 'sensing' and 'seeing' as two different experiences if there wasn't really a difference there," he says.

He says he's actually pleased that Simons' data confirmed his own. "It shows," he says, "the effect is robust. It all comes down to how you interpret the data."