The use of physical characteristics as identity evidence has a long history in forensic investigation, predating the DNA era. Bertillon measurements, fingerprints, and distinctive body marks were all used in nineteenth and early twentieth century identification. The Anderson case generated extensive use of these methods, applied retrospectively to the question of whether Anderson’s physical characteristics matched those of the real Anastasia.
Ear shape analysis was conducted because photographs of the real Anastasia and photographs of Anderson showed characteristics that some analysts said were consistent. Anthropological analysis of facial structure pointed to similarities. Scars on Anderson’s body were compared to accounts of injuries that Anastasia might have sustained.
These analyses were methodologically problematic in several ways. Photographs of Anastasia were taken at various ages under various conditions and provide an imperfect record for biometric comparison. The characteristics being compared were not uniquely identifying in the forensic sense — many people share similar ear configurations or facial structures. And the comparison was being conducted by analysts who could not be fully blinded to which case they were analyzing.
Contemporary forensic standards would require a substantially more rigorous framework for such analyses to carry significant evidentiary weight. In the 1920s through 1960s, the standards were less developed, and the analyses were more easily assimilated into either the belief or the skepticism structures that shaped how observers approached the evidence.
What the physical evidence could not do, under any analytical framework available at the time: definitively identify or exclude Anderson. It could produce impressions of similarity or difference. It could not produce a biological fact of the kind that DNA analysis produces. The transition from impression to fact is what the 1994 testing accomplished.