Facial Recognition Isn’t in Fact Only a ‘Human’ Ability

It had been assumed that humans recognize faces thanks to their humongous brains, but Israeli scientists discovered that our most primitive brain parts are involved as well. Enter the crocodile

Ruth Schuster
Ruth Schuster
Send in e-mailSend in e-mail
Send in e-mailSend in e-mail
A crocodile, whose "brain" is actually like the primitive part of ours.
See crocodile. See crocodile run. RunCredit: Yaron Kaminsky
Ruth Schuster
Ruth Schuster

See Spot. See Spot run. You recognize Spot’s face as opposed to that of other dogs, and your mother as opposed to other humans, and everybody else using the “higher” parts of the brain, the cortex. This is true. But to some consternation among neuroscientists, it turns out that primitive sections of the brain collectively called the subcortex are also involved in facial recognition and other “higher” cognitive processes.

These “higher” cognitive processes were widely assumed to be the product of our uniquely large brain, but it ain’t necessarily so, according to groundbreaking research led by Prof. Shai Gabay, head of the Institute of Information Processing and Decision Making at the University of Haifa. The findings have been published in a series of articles.

Moreover, people have a tendency to see faces where there are none – “smiling” cars, a deity on toast. It’s called the pareidolia effect, and we knew that. But now the scientists’ research shows that the primitive subcortical parts of the brain are involved in that too.

The implication is that we humans aren’t that special after all, Gabay says. Mammals, reptiles, fish, frogs – we all have the primitive subcortical structures. “The subcortex developed early in evolution” – hundreds of millions of years ago, he explains.

The “higher” sections of the brain called the cortex evolved later. The cortex is the outer layer of the brain surrounding the primitive subcortical structures nestled inside.

Our cortex is a massive beast, but it isn’t the only one out there. Still, there is an evolutionary scale – mammals have a more complex cortex than reptiles and fish have none. (It had been postulated that fish don’t feel pain, or at least don’t feel it like humans do, because they lack the “higher” brain parts, but that convenient assumption is widely contested.)

In any case, the cortex is where neuroscientists usually seek sophisticated human cognitive abilities – but there is something of a bias. They’re looking for our advanced cognitive abilities there, Gabay says, and now he and his team have looked at the subcortex too.

Prof. Shai Gabay: Humans aren't as special as they thought they wereCredit: Anna Vedenichev

The wasps and the bees

The bottom line is that neuroscience has assumed that the “higher” cortex is responsible for humans’ sophisticated cognitive abilities such as the ability to think, not just react reflexively; the cortex enables us to monitor our actions and have a degree of executive control over various functions.

And it had been thought that “advanced” abilities include facial recognition. “One reason for considering face perception as an advanced cognitive ability is that social processes are considered a hallmark of human ability,” Gabay says.

“From birth to old age, social connections are imperative to our survival and well-being. Despite the existence of many social species in nature, it has been argued that humans’ unique social disposition is central to our experience and consciousness, and what accounts for the evolutionary development of our large brains and cortex.”

But now Gabay and his colleagues have identified involvement of subcortical structures in that ability. What, does that indicate that a sheep, for instance, could possibly recognize faces? It does.

This is a smack in the face for humans’ sense of superiority and our assumption that we have unique cognitive abilities. Maybe we do, but apparently recognizing Spot isn’t one of them. Even insects such as bees and wasps have been found capable of recognizing human faces.

“Bees can be trained to recognize human faces, so long as the insects are tricked into thinking that the faces are oddly shaped flowers …. The insects use the arrangement of facial features to recognize and distinguish one face from another,” one study reported in 2010.

See Spot think. Mammals have a more complex cortex than reptiles. Credit: funkyfrogstock / Shutterstock

Asking if all this means we’re recognizing Spot et al. with our crocodile brain augmenting our magnificent cortex, Haaretz was crushed to learn that the crocodile brain isn’t a thing – it’s a euphemism for the basal ganglia that are responsible for the four Fs: feeding, fighting, fleeing and fornicating. But the basal ganglia are part of the subcortex where we process primitive functions.

“What we do in our lab to shows that even complex behaviors and cognitive functions are guided by the subcortical stuff,” Gabay says.

Why was it even assumed that facial recognition was a function exclusive to our noble cortex, which is, truthfully, bigger than that of most other animals, since eyes have been around for half a billion years? Fight or flight is subcortical, but first you have to recognize that what you’re seeing is objectionable.

“To me it’s obvious but not to everyone,” Gabay answers. “A lot of researchers think that cognition is the product of cortical involvement. The ability to recognize faces was taken to be a higher ability – cortical stuff.” Face perception was expected to be one of the things humans are best at, he says.

Got it. Speciesist hubris led us to assume that we have unique cognitive abilities stemming from brain regions unique to humans. We may have unique cognitive functions that are the property of our extraordinary cortex. But if facial recognition involves the subcortex too, this isn’t one of our unique qualities.

Want more humbling evidence? When we’re talking with someone, we tend to track their gaze – if they suddenly look in one direction, so do we. Higher cognition this is not: Archer fish do the same thing, tracking the gaze of other fish, the lab showed in separate work. Also, cats have recently been outed for being cued by gaze, as dogs are.

Cats have recently been outed for being cued by gaze, as dogs are – and people.Credit: Darrin Zammit Lupi / Reuters

See Spot, if looking with both eyes

Supporting the researchers’ thesis is the following weirdness. Gabay and the team have shown that people better distinguish different faces when they’re shown to the same eye, but if different faces are presented to different eyes, the subjects had difficulty seeing that they were different!

Which is why? The information from the left eye travels along different neurons than the information from the right eye. This segregation is in the subcortical regions. The visual information from the different eyes converges at the cortex, where unitary perception is created.

This is why the sensitivity observed in Gabay’s lab regarding the eye of origin is a marker of subcortical involvement.

There is definitely also a cortical area devoted to faces; people with damage to that area of the cortex can’t identify faces. It’s just that the primitive animal brain is involved too, and what does that mean for Spot? Can Mittens identify our faces? The sheep, the fish, the frog hopping around Central Park?

“The question is where it developed during evolution,” Gabay says. “It strengthens the claim that facial recognition is based on the primitive areas of the brain, which means that a lot of organisms that have only subcortical structures, or no advanced cortex, can also recognize faces.” And that explains the research showing facial recognition among bees and the like.

All this also explains why we have a tendency to see faces everywhere where there are none – motor vehicles, electric sockets. “We claim that it’s lower brain processes that encourage us to see faces in a bus, Jesus in toast, the Man in the Moon. We have a system that’s constantly looking for faces,” Gabay says. “There are whole industries based on this.”

Click the alert icon to follow topics:

Comments