Researchers Work to Take the Bias Out Of Facial Reconstruction
Instead of relying on European-centric data sets, researchers used a global database to help image a 13,600-year-old woman from Thailand
There’s been something of a boom in facial reconstructions of historical people recently. There's Ava, a Bronze Age woman from Scotland, Context 958, a poor man living in England and Jane, a 14-year-old from the Jamestown colony, who was eaten by fellow settlers. But while facial reconstruction technology has jumped in sophistication in recent years, much of the reference material used to create the faces of the dead are based on averages from European data. A new reconstruction of a 13,000-year-old woman discovered in Thailand, however, used different data to get around that bias, reports Rossella Lorenzi at LiveScience.
According to Bob Yirka at Phys.org, the remains of the woman, including skull fragments, teeth and some bones, were found in Tham Lod rock shelter in northern Thailand in 2002. The sediment around the bones was dated to roughly 13,640 years ago, making her a direct descendent of the first wave of anatomically modern humans to settle in southeast Asia. Analysis of the bones show that the woman stood about five feet tall and was between 25 and 35 years old when she died.
When researchers decided to reconstruct the woman’s face, however, they wanted to test out some new methods since traditional techniques have come under scrutiny. “Facial reconstruction is a very, very popular method, but it has been tested and found to be scientifically invalid since around 2002,” Susan Hayes of the University of Wollongong and co-author of the paper in the journal Antiquity tells Lorenzi.
Instead of relying on forensic data about musculature, tissue depth and other facial features derived mainly from European reference populations, Hayes and her team looked at a global database of statistical information about modern female faces from many cultures. After filtering the data, they used data from 720 contemporary women from 25 different countries to create an average.
The study shows that the Tham Lod woman’s face is similar to the face of Japanese women in terms of height and width. The dimensions of her nose and mouth shares characteristics of African women. Except for the width of the mouth—which is similar to that of contemporary Hungarian women—European characteristics in the reconstruction are absent. Yirka reports that the researchers point out that the image is not an actual portrait of the woman, just an approximation based on averages.
The researchers also tried to avoid making the final image of the face look like a “cave woman.” “The woman is anatomically modern, so you would anticipate an anatomically modern facial appearance,” Hayes tells Lorenzi, pointing out that other reconstructions tend to give their early human subjects a distinctive primitive look. “But this style is not at all supported by the evidence in scientific studies, and instead relates to the pre-Darwinian Christian mythology of the appearance of ‘wild men.’”
While the new method may be more accurate than current facial reconstruction, Hayes tells Lorenzi that it does have one drawback since it takes longer to produce than current techniques. But creating a more precise model is worth the wait.