Spontaneous vicarious perception of the content of another’s visual perspective
In order to navigate our social environment we need to be able to represent the world from viewpoints other than our own. People perspective-take all the time, whether they are aware of it or not. Often, for example, we give a person directions from their perspective “take your first right at the traffic lights”, rather than from our own. Similarly, we will intuitively orient a book to upright to another person when showing them the contents so that they can see it well (even if it makes it more difficult for us). In other situations, we cannot avoiding picturing how our embarrassing stumble looked to those around us. But perspective taking might also be more generally useful. By understanding how, precisely, a situation appears to others, we may appreciate better what they are thinking and in this way empathize with them.
The ability to perspective-take may be unique to humans and develops early in life. As they grow up, children initially only realise that what somebody can see might be different from what they can see themselves (e.g if an object visible to oneself is hidden behind a barrier from the perspective of another person). A bit later, they acquire a full understanding of how the objects that others see actually look to these persons. They can then appreciate, for example, that an object to their own left is to the right of a person facing them, or that the letter ‘Z’ looks like an ‘N’ to someone on their right. This ability to take others’ perspectives seems to develop together with more sophisticated abilities for understanding others, i.e. that other people have thoughts, desires and beliefs that differ from one’s own.
One assumption that is often implicit in how people think about perspective taking is that it involves generating a more-or-less realistic mental image of what we would see if we were in the place of the other person and looked through their eyes. Strikingly, however, this assumption has not been directly tested. Prior research has mostly shown that people are aware of what others can or cannot see, or that one’s spatial reference frame shifts if we are in somebody else position (e.g. that our right is their left). It has left open, however, whether people spontaneously form a mental image of what another person sees.
Our experiment asks exactly that: can people “see” through the eyes of another, and, if so, is this virtual view treated by the brain like something that they can see themselves?
Demonstrating such a similarity between perspective taking and visual imagery would be quite important. If somebody else’s perspective is available to us in the same way as what we see through our own eyes, it becomes immediately obvious why we can so easily put ourselves into their shoes, or why we often feel so clearly what we would if we were in their situation. It might also provide some new insights into the developmental milestones that children go through when they develop the ability to perspective taking, and why it may be so difficult for people with an autism spectrum condition, who sometimes find it hard to appreciate others’ viewpoints.
Our experimental paradigm is based on the classical mental rotation task, first pioneered by Shepard & Metzler in 1971. In our version of the task, participants look out onto a table in front of them. On this table, letters appear in different orientations. Participants simply judge, by pressing a button as quickly as possible, whether the letter is presented its standard or mirror-inverted form (“R” vs. “Я”).
The classical finding is that decision times increase the more an item is oriented away from the participant. The reason is that people first have to “mentally rotate” the letter into its typical upright orientation before they can judge whether it is mirrored or not - and this rotation takes longer the more the letter is initially oriented away from the participant. We hypothesized that if people could virtually “see” through the eyes of another, then simply presenting another person in the scenes should disrupt this classical finding: one might not need to mentally rotate a letter if one already sees in its upright orientation through the eyes of another person.
To test this, we sometimes (in 50% of trials) let another person appear in the scenes, who sat to the left or the of the table and therefore looked onto the letters from the left or the right (see panel B in the figure above). A letter that is 90 or 270 degrees rotated away from the participant would appear upright for these persons, and could easily be recognized from their perspective. If participants make spontaneous use of this alternative perspective, then letters that are oriented away from ourselves should be easy to identify when another person is present to whom this letter would appear upright. This is exactly what we found across several of experiments. These results show that people can in a very real sense “look through another’s eyes” and see the world from their perspective, if it helps their own judgments. Importantly, this happens only if participants see another person in the scenes, but not if they see inanimate objects, such as a lamp that “looks” at the letters in the same way as the persons did (see panel B, lower row). Click on the video below to try out the task yourselves.
Importantly, we did not only find that letters that are oriented away from oneself are recognized quickly if they would appear upright to the other person, but also the reverse. Letters become harder to identify if they appear upside down (and therefore hard to recognize) from the other person’s perspective. This means that people sometimes cannot help it but look through the eyes of another, to the extent that the other’s limited view on an object disrupts their own judgments. This form of visual perspective taking therefore does not just happen spontaneously but also somewhat involuntarily — we cannot ignore how other people see the world.