Erotic and Violent Images Cloud Vision
When people see violent or erotic images, they fail to process whatever they see next, according to new research.
Scientists are calling the effect "attentional rubbernecking."
“We observed that people fail to detect visual images that appeared one-fifth of a second after emotional images, whereas they can detect those images with little problem after viewing neutral images,” said Vanderbilt University psychologist David Zald.
The effect is akin to rubbernecking on the highway, Zald and his colleagues say. Your brain might suggest you watch the road ahead, but your emotions force you to look at the accident on the side of the road.
Research subjects were handed a stack of pictures that included pleasant landscapes and architectural photos. They were told to search for a particular image. Negative images were placed anywhere from two to eight spots before the search target.
The closer the negative image was to the target picture, the more frequently people failed to spot the target.
In a follow-up study, negative images were replaced by erotic shots. The effect was the same.
"This suggests that emotionally arousing images impact attention in similar ways whether they are perceived as positive or negative," said colleague Steven Most of Yale University.
The researchers suspect we can't control the effect.
"We think that there is essentially a bottleneck for information processing and if a certain type of stimulus captures attention, it can basically jam up that bottleneck so subsequent information can't get through," Zald said.
As for rubbernecking on the road, Zald has a caution:
"If you are simply driving down the road and you see something that is sexually explicit on a billboard, the odds are that it is going to capture your attention and – for a fraction of a second afterwards – you will be less able to pay attention to other information in your environment," he said.
The initial study is detailed in the August issue of the journal Psychonomic Bulletin and Review. The follow-up research has not been published. | 
