ORIGINAL POST: When I read the breathless headline in the Wednesday print addition of The Bryan/College Station Eagle that is also the title of this post, I shook my head. There staring at me under that banner headline was the DNA based face of the killer of Virginia Freeman in south Brazos County from1981.
It was like reading a summary of an episode of CSI.
The problem? Well, at this point the science behind the composite of the killer's face has not been proven valid. In fact, it is not so much science than speculation based on unnerving extrapolation. You would not know that from The Eagle story, or the story from our local television station KBTX, or a dozen like it from various news outlets and platforms from New England to Arizona - here, here and here.
It is the message rather than the messenger I have the bone to pick with. Sheriff Chris Kirk is one of the finest lawmen I know. As an investigator he was both intelligent and meticulous. As Brazos County's elected sheriff, he has brought stability to a position that for many years had been dangerously unstable. I think we in this county take that stability for granted, but if you lived during the years of upheaval in that office, you appreciate Sheriff Kirk even more.
So, to the so-called science. The second of the stories linked involves the 1987 death of Darlene Krashoc at Fort Carson Colorado. Another, reporter, Carol McKinley, took a different angle for her story about Krashoc's death and the use of the same new technology the subject of the Freeman press conference. McKinley actually spoke to qualified people urging caution about the reliability of the science and false expectations faux science brings:
'Phenotyping is not state of art. It’s not the state of anything,' says retired Denver homicide detective Jon Priest. He is concerned that phenotyping gives false hope to victims families who are desperate for answers.
'This is one of those things that is not proven, supportable, nor accepted,' Priest said, adding he’s not against trying new technology. 'The science may get there someday, but it ain’t there now.'And this:
'It’s giving the family unrealistic expectations,' says Dr. Richard Spritz, director of Human Medical Genetics and Genomics at the University of Colorado School of Medicine. 'Do I think it’s likely having this technology will get them to a perpetrator? No. Because the reliability of this technology has never been subject to critical tests. The likely reliability is low.'
Ellen Greytak of the company, Parabon Snapshot, whose proprietary technology did the phenotyping for the Sheriff's Department, videoed into the presser on Tuesday. As I wrote above, all very CSI. The price tag for the work was paid by the National Geographic Explorers, who, in exchange for paying for it, was present filming the presser as part of a documentary concerning the investigation.
The critical question is whether any results other than heightened expectations leading to crashed realities will be realized.
Greytak, according to McKinley's story, says the technology "[helps] narrow down suspects… gives law enforcement the ability to eliminate them," Yet the headlines and the emphasis is not elimination, it is putting a face with the the DNA. Greytak's statement about elimination is like saying the reason to buy a self driving car is because of its curb appeal - it is just not the sell. Instead, the sell is the technology to put a face on a killer. The science of DNA has proven it can tell you the killer's race, eye and hair color. But provide a face that resembles the killer then and now? NO.
Forensic science is easy to fall in love with, but is littered with examples of being jilted by junk. DNA, the gold standard, has been rocked by changing standards of comparisons and probabilities on mixtures. Bite marks, blood spatter analysis, shaken baby syndrome, all now are forensically questionable.
I believe the science behind this is no better than alchemy at this point.