The longtime friends who live in Atlanta are not related. Their ancestors don’t even come from the same part of the world. Malone’s family was originally from the Bahamas and the Dominican Republic. Chasen’s family came from Scotland and Lithuania. It is not the result of some deep dark family secret. However, they look strikingly similar. It’s not just their brown hair, beards and glasses. It is also the structure of their nose, their cheekbones and the shape of their lips.
“Michael and I go way back, and it’s all been kind of a source of a lot of fun for us because over the years, we’ve been mistaken for each other all over Atlanta,” Chasen told CNN’s Don Lemon. “There were some really interesting situations that came out just because people thought we were the other person.”
The two look so much alike, even facial recognition software had trouble distinguishing them from identical twins. But now scientists think they can explain what makes them look so alike — and could explain why each of us can have a doppelgänger.
People who look alike, but are not directly related, still appear to share genetic similarities, according to a new study.
Among those who had these genetic similarities, many also had similar weights, similar lifestyle factors, and similar behavioral characteristics such as smoking and education level. This could mean that genetic variation is related to physical appearance and also, possibly, can influence certain habits and behavior.
Scientists have long wondered what creates a person’s doppelgänger. Is it nature or nurture? A team of researchers in Spain tried to find out. Their results were published in the journal Cell Reports on Tuesday.
Dr. Manel Esteller, a researcher at the Josep Carreras Leukemia Research Institute in Barcelona, Spain, said he had worked on twin research before, but for this project, he was interested in people who look alike but have no real family relationship by turning almost 100 years back.
Art leads to science
So he turned to art to answer a question about science. He and his co-authors recruited 32 look-alikes who participated in an “I’m not a look-like!” photography project by a Canadian artist, François Brunelle.
The researchers asked the couples to take a DNA test. The couples filled out questionnaires about their lives. The scientists also put their images through three different facial recognition programs. Of the people they recruited, 16 pairs had similar results to identical twins identified using the same software. The other 16 pairs might have looked the same to the human eye, but the algorithm didn’t believe it in one of the facial recognition programs.
The researchers then took a closer look at the participants’ DNA. The pairs that the facial recognition software said were similar had many more genes in common than the other 16 pairs.
“We were able to see that they look like humans, in fact, they share several genetic variants. And these are very common to each other,” Esteller said. “So they share these genetic variants that are related in a way that they have the shape of the nose, the eye, the mouth, the lips, and even the bone structure. And that was the main conclusion that genetics put them together.”
These are similar codes, he said, but they are random.
“In the world right now, there are so many people that eventually the system produces people with similar DNA sequences,” Esteller said. This has probably always been true, but now with the Internet, it’s much easier to find them.
Other factors at play
When they looked more closely at the couples, they determined that other factors were different, he said.
“It’s the reason they’re not completely identical,” Esteller said.
When the scientists took a closer look at what they call the epigenomes of the most similar doppelgängers, there were greater differences. Epigenetics is the study of how environment and behavior can cause changes in the way a person’s genes work. When the scientists looked at the microbiome of the most similar couples, they were also different. The microbiome is the microorganisms — viruses, bacteria and fungi too small to be seen by the human eye — that live in the human body.
“These results not only provide insights into the genetics that determine our face, but may also have implications for the establishment of other human anthropometric properties and even personality traits,” the study said.
The study has limitations. The sample size was small, so it’s hard to say whether these results would apply to a larger peer group. Although the researchers believe their findings would change in a larger group. The study also focused on couples who were largely of European descent, so it’s unclear whether the results would be the same for people from other parts of the world.
Dr. Karen Gripp, a pediatrician and geneticist at Nemours Children’s Health, whose research is reported in this paper, said the study is really interesting and validates a lot of the research that comes before it.
Applying science to the real world
Gripp uses facial analysis software in her work with patients who may have genetic conditions to assess her patient’s facial features that may indicate certain genetic conditions.
“It’s a little bit different from the study, but it actually points in the same direction that changes in a person’s genetic material affects facial structures, and that’s really the same underlying hypothesis that was used in this study as actually confirmed. for some other things like the microbiome it didn’t seem to be as relevant,” Gripp said.
As for the nature vs. nurture question posed by the study, Gripp believes both are important.
“As a geneticist, I strongly believe that nature and genetics are very important for almost everything, but that doesn’t mean that nurture is just as important,” Gripp said. “For any person in the world to be successful there are so many contributing factors and the environment is so important that I don’t think it’s one or the other.”
A potential problem
The study, he said, also points out that there are still limits to the accuracy of facial recognition software. While several cities concerned about privacy issues and problems of mistaken identity have enacted rules that prohibit or restrict local police from using facial recognition software, the federal government and some local law enforcement agencies use it more frequently.
A 2021 federal survey found that at least 16 federal agencies use it for digital access or cyber security, 6 use it to generate leads in criminal investigations, and 10 more said they planned to expand its use.
It is also more commonly used in airports. Some companies use it to help make hiring decisions. Some landlords have it installed so that tenants can enter buildings. Some schools use it to track movements in public areas on campuses.
“If you translate this study to the real world, it shows you a potential pitfall that digital facial analysis tools could misidentify someone,” Gripp said.
While the technology is improving, in previous studies, the technology has already been shown to be far less accurate in identifying people of color, and several black men have been wrongfully arrested due to facial recognition.
“If you think about facial recognition software that often opens computer screens and things like that, misidentification is possible. So I think this has taught us something very important about facial analysis tools as well,” Gripp said.
But the study seems to suggest one conclusion. At least physically, we may not be so unique.
“I think all of us right now have someone who looks like us, a double,” Esteller said.
While some would prefer to be unique in their appearance, Malone, who happens to be friends with his doppelgänger, is thrilled that he is not alone in his appearance. His resemblance to his friend has brought them closer, and he thinks if more people knew how similar they were to others, maybe they too could find common ground, especially in this polarized world.
“It made me realize that we’re all connected,” Malone said. “We are all connected because humanity probably starts with a small thing.”
CNN photos by Francois Brunelle
