Facial approximation, also known as facial reconstruction, is a method that attempts to recreate the likeness of an individual’s face from the features of their skull. Facial approximations may be used by forensic investigators, anthropologists and archaeologists to assist in the identification of unknown victims of crime, to portray the faces of historical figures or illustrate the features of fossil human ancestors. However, facial approximations rely as much on artistic interpretation as anatomical knowledge and many researchers consider the method too subjective. As a result, facial approximation remains controversial and is generally considered only when other identification techniques have failed.
If an investigation resorts to facial approximation, two and three-dimensional approaches are available. Two-dimensional approximations involve sketching the facial features on an overlay of a photograph of the skull. This may be done by hand or by using specialized graphic design software. Tissue depth markers may be attached to the skull before it is photographed to assist the artist in rendering the soft tissues. Two-dimensional techniques are relatively fast, efficient and economical. In addition, they allow small adjustments in the hair or face to be made quickly and easily. However, two-dimensional methods depend heavily on the skill of the artist, the completeness of the skull and the quality of the original photography.
Increasingly, three-dimensional techniques are being used to recreate individual faces using either clay sculpted over the skull or computer rendering software. Clay sculpting methods are visually compelling representations, but are also time consuming and expensive. In addition, unless a cast of the skull can be made, clay modelling techniques restrict access to the skull for other purposes and are not recommended. Three-dimensional software applications provide an alternative to clay that allows material to be scanned and manipulated without compromising their integrity. Using mirroring tools, computer modelling can also compensate for incomplete material more easily than other techniques. However, as with two-dimensional photography, digital images are susceptible to distortion and care must be taken to ensure the remains are not incorrectly rendered.
Photographic superimposition is another identification tool sometimes associated with facial approximation. However, superimposition differs from approximation in that it does not involve an ‘unknown’ face. Instead, it attempts to ‘confirm’ the identity of an individual by comparing photographs or radiographs of a skull with antemortem photographs of a presumed victim. Here, investigators already suspect the remains to be a particular individual and are attempting to achieve a positive, rather than a circumstantial identification.
As mentioned, not all researchers accept the validity of facial approximation and significant challenges remain. First, the correlation between the bony features of the skull and the soft tissue of the face remains weak. Many facial muscles are highly variable in both structure and presence and some do not attach directly to the skull at all. Consequently, it is difficult or impossible to rebuild some muscles from the hard tissue. In addition, there is no way to predict the level of subcutaneous fat in the face, yet this clearly impacts the final appearance. A second problem relates to tissue depths. All approximation techniques use average tissue depths calculated from reference samples. Unfortunately, these data are limited both in number and in the coverage of ages, populations and body types. Consequently, a ‘finished’ face will look different depending on the set of standards used. A final issue relates to accuracy and reliability. Legal rulings in both Canada and the United States now require methods used in criminal investigations to meet specific criteria for scientific validity. This ensures that evidence provided in court is based on testable methodologies that use accepted standards and known error rates. At present, facial approximations do not meet these criteria.
Overall, facial approximation techniques provide a crude means of generating new leads or sparking public interest in a case that may lead to a tentative identification. Through dramatic representations of the past, they may also encourage interest in new archaeological material and promote additional research. However, facial approximations should not be mistaken for authentic representations of past individuals and the methods should be used cautiously until further research can validate them.