AS sometimes happens in science, the discovery was accidental but no less startling. Dr Clementine Thurgood, a postdoctoral researcher at Swinburne University, was investigating whether there is such a thing as a subliminal effect – the idea that rapidly flashed images on a movie or television screen can influence a person's behaviour, despite the images not being consciously perceived.
With its undertones of mysterious forces exercising mind control, the belief fuelled fears back in the 1960s of government manipulation. It continues to trouble many people who think it could be used by big corporations to persuade an unwitting public to buy their products.
But what if humans were capable of recognising images shown for such a short time that no one thought it was possible? In that case the so-called subliminal effect — literally below the threshold of human consciousness — would not apply because the viewers would know they had seen the images.
Advertisement: Story continues below In her experiment, Dr Thurgood had a photo of her dog Lucy on her computer and could show the image for a thousandth of a second. Associate Professor John Patterson was watching and, to her astonishment, not only identified the image as a dog but also pointed out she was a boxer.
"We didn't think anyone could see that kind of detail in an image when the exposure was so brief," she says. "But when we tried again, we both could see it quite clearly."
Vision researchers had not made this discovery before because they did not have the technology to show an image so rapidly. Generating an extremely rapid flash of light is not so difficult but presenting a complex visual image in a fraction of a second is.
As part of the Swinburne study directed by Professor Allan Whitfield, Dr Thurgood and Professor Patterson invented a device that could show images at speeds never previously possible. They loaded a digital image onto a dark liquid crystal display and, when the image was fully loaded, the screen was backlit with a single flash from an array of light-emitting diodes that reached full luminance almost instantly.
This allowed the researchers to control exactly how long the image could be shown. Called an LED-tachistoscope, and now patented by Swinburne, the instrument can display images as briefly as a millisecond, 1/1000th of a second. This was confirmed using a photodiode — a transistor that conducts electricity only when exposed to light.
To test whether a random sample of people could identify a series of different images each shown at 1/1000th of a second, Dr Thurgood rounded up 63 students, friends, passers-by and anyone else agreeable to take part in the experiment.
She compiled 60 digital photographs of creatures such as elephants and the big cats, owls and eagles, sharks and sea lions. She says the three types were chosen so there would be considerable variability and that animals in general would be well represented.
Half the images were of animals in their natural habitats and half in isolation against a plain white background. Likewise, about half the people taking part saw only photos of the animals in their own environment while the rest saw the animals against the white background. As well, half the photos were shown for one millisecond and the rest for 10 milliseconds – that is, 100th of a second.
The order in which the animals appeared and the exposure times were random. Each person was tested separately, seated in front of the LED tachistoscope with Dr Thurgood at the controls. She explained that photographs of various animals would be shown on the screen although they would sometimes appear so rapidly the viewer might not actually see what they were.
The participants were asked to state, as quickly and simply as possible, the name of the animal or the type it belonged to: "big cat" for a tiger, say. If they could not name it or did not see anything, they were to say "pass" and miss that particular turn.
"We found that at least 83 per cent of the responses were correct when an image was shown for 1 millisecond," Dr Thurgood says. "There was also little difference whether it appeared for one or 10 milliseconds and, even if it seemed the image appeared on the screen more quickly, most people could still see it quite clearly at one millisecond."
She says the findings were unexpected and extended previous research on visual perception in which the researchers had been unable to test such short durations. In fact, the results showed that human brains could consciously process visual information so fast as to stretch the limits of modern measuring instruments.
"Our discovery also has implications in psychology, including evolutionary explanations for uniquely human vision attributes, and spells the end for ideas of mind control via the subliminal effect — at least where brief, unmasked stimulus exposures are concerned," Dr Thurgood says.
"We have yet to reach a limit at which a person cannot see an image presented on the screen although one of our PhD students is developing an instrument that can show images for 10 microseconds — that's 100th of a millisecond! During a test I had a look and I could see the image although not as clearly because it was so fast and now the instrument is being validated."
In her next experiment, she intends to follow or precede the test image with a mask, what she calls a nonsense image that interrupts the neurological processes in the brain and tends to cause the person to forget the target stimulus, "the iconic image". She plans to combine the LED tachistoscope's rapid images with masks but this time expects that those in the trial will not do as well in recognising the pictures of the animals.
So is there such a thing as a subliminal effect — that messages "hidden" in an advertisement, say, could affect a person's behaviour?
"I couldn't establish that in my experiment because the subjects kept recognising the images, no matter how rapidly they were shown," Dr Thurgood says.
"They could still report what they had seen, so the next step is to show the images for even briefer periods and to use masking."