Inverse, Inverse! Humans Share an Unusual Facial Processing Limitation with Fish

 
 

New research suggests that some fish may process faces like we do. Emmet Feerick takes a closer look.



Over the years, software companies have spent millions trying to get computers to do what comes naturally to us: recognise faces. We are social creatures, and throughout our evolutionary history the ability to distinguish friend from stranger has proven vital to our survival. While we have no trouble forgetting telephone numbers or even first names, few among us can forget a familiar face.


Despite this, our facial processing has one surprising limitation; we cannot recognise upside-down faces. Study after study has shown that while we can easily recognise inverted objects like chairs and mugs, we are simply unable to tell which of two inverted faces is the original when one has been slightly altered. Those familiar with the Thatcher Illusion will be aware of how startling this effect is. This so-called “inversion effect” is a result of our brain evolving an object recognition area specialised for faces, separate from the generalised object recognition area used for all other objects.


When we try to recognise inverted faces, our brain resorts to the circuitry it uses for normal objects.

The reason for this is straightforward. Over millions of years, we got used to seeing most objects in a variety of orientations – poisonous plants, edible fruit, other animals, and so on. However, faces were always upright. Why take up precious space in the brain with circuitry to recognise upside-down faces when we almost never see them? The result: we are excellent at recognising upright faces and all of the subtle emotions they display, but when we try to recognise inverted faces, our brain resorts to the circuitry it uses for normal objects. As far as our brains are concerned, upside-down faces can be treated as any other objects.


It has been known for some time that we share this face processing quirk with other primates like rhesus monkeys and chimpanzees, and even non-primates like sheep.

It has been known for some time that we share this face processing quirk with other primates like rhesus monkeys and chimpanzees, and even non-primates like sheep. Yet recent research by Wang and Takeuchi was the first to uncover the effect in a non-mammal, namely; the Japanese rice fish. This small fish is highly social, and is most commonly found in and around the rice paddies of East Asia. By alternately obscuring the males’ face, body, and tail with translucent film, these researchers discovered that the females only failed to recognise these males when their faces were covered. This indicates that faces are more important to this fish than other areas of the body when it comes to recognising other fish. They also found that by putting these fish in separate tanks, they were still able to recognise each other by sight. In other words; smell was not a factor.


 

However, the truly surprising finding came when the researchers used a prism to invert the male faces. When this happened, they found that the females were unable to recognise their faces. This was despite the fact that they could easily recognise other inverted objects. Fish had previously been shown to recognise each other by their faces, but this is the first instance where they were shown to be susceptible to the face-inversion effect. This, Wang said, suggests that the Japanese rice fish may be like humans and other primates in that they have a specific brain area for face-processing. This study opens up a whole new branch of future research which will investigate the specific neural mechanisms involved, and the genetic factors which influence their development.

Advertisements