The question of who is actually a real person, and who is pretending to be a man, has long puzzled scientists and philosophers.
Now, a new paper suggests we may be approaching a milestone in understanding the phenomenon.
“It’s one of the most fascinating problems we have,” says Prof Ian Macdonald from the University of Melbourne, who led the work.
The paper, published in the journal PLoS ONE, suggests a clue to the identity of the real man may lie in the way we process facial features.
In a series of experiments, Macdonald and his team showed how humans learn to recognise facial features by looking at their irises.
When they asked people to see a picture of a man with an eye patch, the participants would immediately identify that man as a man.
But when they were shown a photograph of a face without an eye patches, they showed no such bias.
Instead, they were primed to think that the face was of a woman.
In fact, when they saw the image of a person without eye patches with a full face, they saw that woman as the real woman.
“If you can do this for the other face you can look at the faces in the group and see that they’re the same,” Macdonald says.
Researchers are now exploring whether similar processes drive facial identification.
One hypothesis is that our perception of faces is triggered by the way they look, says Macdonald.
“People who are really good at seeing what’s in front of them will look at their eyes and see a person with full eyes and they’ll see it’s a woman,” he says.
“And people who are less good at this will look more at their cheeks and they will see it has a woman’s face.”
This, he says, “may be because facial features are a sort of proxy for gender.”
Macdonald says it may be possible to use facial recognition technology to identify faces in photographs, as well as in real life.
He and his colleagues are working on facial recognition for facial tattoos, for example.
They are also working on the possibility of scanning people’s faces using ultrasound to determine the identity.
“We’re actually hoping that we can do that in the near future,” Macaugas says.
He says his group will be looking at a range of techniques, including scanning subjects with a laser pointer, which could be used to create a “face-like” 3D image of the subject.
MacDonald’s research also suggests that we could use ultrasound to identify a person in a video game.
For example, in one experiment, the team showed a video of a young woman wearing a bikini wearing a white shirt, to people who had heard the video before.
But when they showed a person playing a video on their computer, they thought that the video was of an Asian woman, not an Asian man.
If you look at a face in the video, you might recognise that it’s the same person, but if you look more closely, you’ll notice the features that you’re more familiar with in a real woman,” MacDonald says.