In a first-of-its-kind study, German researchers used an electroencephalogram to track how the brain reacts to others' faces as they are exposed to them in different ways over time, finding more pronounced neural reactions corresponding to familiarity when people viewed the faces of individuals they had met in person than when they viewed faces they had indirectly observed in the media or in static images.
The findings, published Monday in the Journal of Neuroscience, revealed that long-term memories and contextual information may help humans instantly recognize familiar faces and distinguish them from those of strangers. The study also helped elucidate the difference between facial identification — the ability to spot a face in the crowd and recognize it as a particular person — and familiarity — the sense of knowing someone based on prior interactions with that person. The team's results suggest that familiarity and facial identification may emerge independently of each other.
More specifically, the researchers observed different surges of brain activity in participants' right temporal cortices, which they said corresponded with the degree to which a person had familiarized herself with a particular face. Meanwhile, activity in the brain's bilateral occipital lobe appeared to more closely coincide with participants' identification of various facial features, a phenomenon that the researchers think is distinct from familiarization.
The strongest familiarization response occurred when participants observed the faces of those whom they'd spent time with in person. In the age of the coronavirus, when many human interactions have migrated to the digital sphere, the results may illustrate the importance of a speedy return to in-person communication when it becomes safe.
Prior research had already revealed differences in how the brain reacts to familiar versus unfamiliar faces. But the German researchers wanted to know more about the gradual process of building familiarity with the face of another — the neurological steps that can transform an unrecognizable stranger into someone with whom one may feel a deep sense of personal knowledge or understanding.
Gyula Kovács, a professor of biological psychology and cognitive neurosciences at the University of Jena and the last author on the paper, suspected that our facial recognition abilities might vary greatly under various conditions. For instance, one may see a building doorman, butcher or barista nearly every day at her place of work. But outside that context — on the street or the bus — the same person may be completely unrecognizable.
Meanwhile, fans of a celebrity like Angelina Jolie, despite having never met her in person, might be able to recognize the actress with a high degree of precision. "You saw her in many films," Kovács pointed out in an interview with The Academic Times. "You developed some information about her personality. You heard stories that she adopted children. You heard stories that she divorced Brad Pitt."
He thinks it's these emotional, biographical and memory-based connections that make it easier for humans to recognize figures they may have only encountered in the media. Meanwhile, our ability to quickly locate and feel connected to the faces of our friends and family is potentially rooted in the prompting of complex, long-term memories and other higher-level cognitive processes.
The study considered how three groups of participants would react to faces, depending on different levels of prior exposure to each of them. All three groups were exposed to images of unfamiliar faces as part of the experiments. The first group simply completed a sorting activity to develop some degree of perceptual exposure to a set of faces. Remarkably, at this purely visual-based stage, participants' brain activity in response to photos of people to whom they had already been exposed was very similar to their brain activity in response to photos of people they had not already seen.
In the second stage, participants watched an entire season of a TV show — either "The Americans" or the "The Bridge" — and were sent daily text messages with trivia and personal details about the lead actors from each show. Brain scans indicated that these participants showed some degree of familiarity with the actors when viewing images of their faces, having built something of a one-sided relationship with them by watching a show in which they starred as well as by learning about their lives.
A final group spent face-to-face time with two research assistants, playing quiz games and engaging in informal discussions, with the aim of forming a natural relationship. The assistants also shared autobiographical details in order to help provide more concrete contextual information for the participants. These in-person relationships led to "even stronger signatures of face familiarity that spanned electrode clusters over the bilateral occipital and temporal cortex," when viewing the faces of the familiar assistants, the researchers wrote.
In other words, people displayed the highest degree of neural familiarity following live, interpersonal exchanges, compared to the other exposure tactics. Those neural signatures were most pronounced in one brain region, the occipitotemporal cortex, about 400 milliseconds after the person viewed a face, suggesting that the region may play a key role in registering a face as familiar.
While face-to-face interactions may lead to more robust brain activity, the same levels of connection may be more difficult to achieve on platforms like Zoom. Two-dimensional digital interfaces offer limited interactivity and immersion and present artificial constraints that are rarely present in the real world.
"Normally, when you are talking to somebody, you don't see yourself, continuously," Kovács said. "When you are having a personal meeting, you stand up, you walk, you drink a coffee, you look around." But Kovács cautioned that more research would need to be conducted in order to determine the extent to which online interactions impact the brain's ability to form facial connections.
By focusing on facial familiarity — rather than identification alone — the researchers learned that the process by which we build familiarity with someone's face may take longer than previously thought. This could have implications within the scientific community itself, as researchers frequently employ visual-based training methods to test human facial recognition. "Short-term, perceptual training — what we are using regularly in the lab — is probably not enough," Kovács said. "It's probably measuring different mechanisms, compared to the ones we thought."
The study "Getting to know you: Emerging neural representations during face familiarization" published May 24 in the Journal of Neuroscience, was authored by Géza Gergely Ambrus, Charlotta Eick and Gyula Kovács, University of Jena; and Daniel Kaiser, University of York.