Forensics Under Fire

By John McCormick

FOR ANYONE WHO QUESTIONS the importance of understanding science, this is the mother of all must-read stories because this could happen to you.

Imagine, if you possibly can, spending half your life in prison convicted of murdering a taxi driver because of thirteen hairs found in a ski mask. Not your ski mask of course. Not even a ski mask found at the scene of the crime. Just a random ski mask found nearby which some “expert” from the FBI testified in court one of the hairs matched yours “in all microscopic characteristics.”

The prosecutor then tells the jury that the chances it came from another person are “one in ten million.”

That is exactly what happened to Santae A. Tribble in 1978, and it happens far too often because of junk science masquerading as so-called forensic science.

Then seventeen-years-old, Tribble was convicted in the killing of a D.C. taxi driver based upon a microscopic examination of Tribble’s hair compared to thirteen hairs in a stocking found near the crime scene. The killer was wearing a stocking mask.

When scientifically tested, it turns out that almost all forensic evidence, even DNA, is only marginally accurate.

But back in the late ’70s, hairs were compared microscopically with a sample and whether someone was acquitted or sent to prison for life depended entirely upon the judgement of the person using the microscope.

In Tribble’s case, twenty-eight years later the DNA on those thirteen hairs was re-analyzed. It turned out that the hairs came from three different people, none of them Tribble, and from one dog.

Tribble was recently awarded $13.2 million, but the question remains whether that can compensate for nearly three decades in prison, heroin addiction, HIV, depression, and hepatitis. Could you or a family member be at risk due to bad forensics?

Circumstances can place nearly anyone under suspicion for some crime but, realistically, just how often does this sort of thing happen?

Almost all the time.

The FBI has only recently acknowledged that forensic hair testimony in at least ninety percent of trial transcripts analyzed by the bureau contained erroneous statements. Twenty-six of twenty-eight FBI hair analysts who were investigated provided testimony or submitted laboratory reports with “grossly exaggerated” data that often helped prosecutors. The analysis looked at cases prior to 2000, the year in which the FBI adopted more stringent procedures for its hair analysis.

The Tribble case is is just one of thousands of cases where later DNA testing has overturned a conviction. Unfortunately this sometimes came after the individual was executed—despite a judicial system based in a country where the first inalienable right is supposed to be “life,” and the second “liberty.”

“Happiness,” was not a right, only a pursuit.

DNA, it turns out, is great for the defense. It is virtually impossible to visit a crime scene without leaving bits of your DNA. And advances in analysis, which permit the detection of even tiny traces, have led to false positives.

Although it can prove someone did not enter a crime scene, the same tool can’t prove someone was present. It is even possible for someone else to be in contact with a person and transfer a detectable amount of DNA to a crime scene merely by accident.

Virtually every forensic tool which has been used in courts and seen in TV and movies for half a century or more is based on nothing but junk science. That’s not my opinion alone; it’s also the conclusion of a 2009 National Academy of Science report, “Strengthening Forensic Science in the United States, A Path Forward”— a scathing denunciation of virtually all forensic “science.”

One reason it took so long to overturn this junk science is that judges wouldn’t permit the science behind forensics to be questioned. It was only on July 25, 2009, in Melendez-Diaz v. Massachusetts when it became possible to cross-examine properly. The Supreme Court ruled that forensic findings are open to interpretation and could be manipulated.

The court decided to rule on forensics because previously the defense could cross-examine witnesses, like fingerprint experts, on what they found, or their qualifications, but they couldn't question the validity of the “science” itself.

Once a standard was generally accepted, that was that, and it was virtually impossible to question the science (or lack of) behind the forensic tool.

We know now that the science behind various forensic evidence such as bite marks, fingerprints, shoeprints, tire marks, and more, is seriously flawed—some completely meaningless. But judges never let the defense (or prosecution) do more than ask how the evidence was collected, whether a search was legal, examine the chain of custody ... Questioning whether the statistics always cited were accurate, or even meaningful, wasn’t permitted.

Although DNA analysis is hard science, it isn't good as a forensic tool because modern techniques are so sensitive they may only reveal that someone—perhaps a policeman—who touched a suspect later walked through a crime scene before evidence was collected.

Fingerprints

The gold standard of police work in the U.S. since December, 1911, was based on the simple conclusion that no two people could have identical fingerprints. But what is the scientific evidence to support this claim? And just how closely do fingerprints have to match to be called “proof?”

It should be obvious to anyone with the slightest knowledge of scientific proof or basic logic that it is not only practically but theoretically impossible to prove that no two people could have the same prints unless you have fingerprints of every human living or dead and compare them.

Yet, people go to prison based on that fiction.

Even worse, fingerprint identification doesn’t involve identical comparisons like superimposing one photograph over another to determine a match, which would be nearly infallible.

Fingerprints are classified based on three major characteristics: loops, whorls, and arches. Because these are biological patterns, they are seldom clearly definable and are always somewhat smudged in collection. The examiner’s personal judgement in identifying each characteristic is also involved. [Left, courtesy of FBI.]

Further, only a certain number of loop, whorl, and arch points are actually required to match for an analyst to say they are identical, and that number varies with different jurisdictions.

There are upwards of 150 ridge identification locations or points on a single full fingerprint. The vast majority of fingerprints found at crime scenes (latent prints as opposed to rolled prints taken by police from an individual) are only partials.

“Forensic Magazine” recently cited a case where only seven matching points were ruled by the court to be sufficient to prove identity; many experts insist more points are required.

In “The Criminal Law Handbook: Know Your Rights, Survive the System,” by Paul Bergman and Sara Berman (Nolo Press), the authors state that many experts require twelve matching points while others want to see twenty before they will testify that a match exists.

A further problem is that newer methods of recording fingerprints, such as with laser scans or fogging superglue, can recover latent prints which are so old they have little or no skin oil left and therefore can’t even be shown to be contemporaneous with any crime. In other words, they could have been made long before the crime occurred.

The FBI, which has a vested interest in supporting their vast archives and sophisticated laboratory, contends that there should be no specific minimum standard for fingerprint matching and a determination can be made based only on the subjective judgement of the examiner. A similar rule applies in the U.K. However, Australia requires a minimum of twelve matching characteristics and both France and Italy require sixteen as the minimum.

So what do we have? First, it is impossible to prove no two people have identical fingerprints because we have not compared every fingerprint from every person. Second, latent prints are never perfect. Third, fingerprints are classified by humans who can have differing opinions or be under pressure. Fourth, even given a close comparison, fingerprints need not match more than a few of the vast number of possible points to be called “identical.”

This isn’t science, it is mysticism. Want more evidence?

Fingerprints are now being used in biometrics to identify cell phone users. In April of 2014, the Samsung Galaxy S5 was released. It uses fingerprint recognition instead of a password to unlock the phone. But recent experiments using electricity-conducting ink and readily available paper created copies of scanned fingerprints accurate enough to fool several Android phones.

Recent experiments have shown that other biometric sensors can also be defeated using the same technique, which costs less than $500.

Footprints, tire tracks, etc., all examples of pattern analysis and comparison, fail statistical tests for validity. Put another way, for decades “scientists” have testified in court that two patterns such as footprints, tire marks, even bullet marks and firing pin marks on cartridges, could be exclusive proof of identity despite being slightly different. The prosecution can merely point out that wear and tear accounts for the differences.

In other words, they have absolutely no basis in science.

DNA

DNA, the supposed gold standard in forensics, can have a useful place in some trial situations. But today, even small amounts of DNA can be tested and lead to false positives. Advanced technologies such as polymerase chain reaction—which can multiply tiny amounts of DNA thousands of times—make it possible to pick up DNA at levels hundreds or thousands of times lower than when DNA fingerprinting was first developed in the 1980s.

Today you can even collect “touch DNA” from fingerprints on a glass or a doorknob. A mere twenty-five or thirty skin cells will sometimes suffice. But such low amounts don't always prove that someone touched an object or was even present at the scene of the crime. Analysts have picked up DNA transferred from one person to another by way of an object that both of them have touched, or from one piece of evidence to another when two items jostled against each other in an evidence bag.

DNA is widely misunderstood because detailed science doesn’t make for good TV.

The FBI uses two kinds in an analysis: nuclear DNA (nDNA) and mitochondrial DNA (mtDNA). According to the “FBI Handbook of Crime Scene Forensics” (Skyhorse Publishing, 2008), “Nuclear DNA (nDNA) is the more discriminating of the two types and is typically analyzed in evidence containing blood, semen, saliva, body tissue, and hairs that have tissue at their root ends ... mtDNA is typically analyzed in evidence containing naturally shed hair, hair fragments, bones, and teeth.” An mtDNA sample is often preferred if the sample comes from a close relative because it is much more sensitive.

[Right, a DNAAU-1 biologist prepares samples for robotic DNA processing. Photo courtesy FBI.]

Only thirteen matching points (loci) in DNA were formerly required to identify an individual in the U.S. court system. But even that many are now considered insufficient, and the FBI now requires twenty matches.

The latest advance in DNA analysis, forensic phenotyping, attempts to determine physical characteristics of an unknown individual. This is still in the highly experimental stage. Telomerase measurements can be used to estimate age within about a decade.

Telomerase (telomere terminal transferase), is a ribonucleic acid (RNA) based enzyme attached to the end of chromosomes found in fetal tissue, adult germ cells, and tumors. It can be used as a forensic or medical tool because as cells replicate over time the telomerase becomes shorter and shorter. In fact, at one time it was thought to be the key to stopping aging.

Isotopic analysis is another forensic tool which has been successfully used to identify bodies. Depending on where and when someone grew up, their bones and especially teeth have different levels of oxygen and other isotopes.

The Future

With virtually every forensic tool we have all seen on TV and in movies now shown to be faulty, what is the future of forensic testing?

Fortunately for law enforcement, as old methods come under extreme scrutiny there are new methods being developed.

Each brand of lipstick has its own combination of chemicals that gives it color, but analysis requires extensive testing only capable at a few labs, and is therefore expensive. Recently, researchers described a new, inexpensive technique that could allow scientific identification of the exact brand and color of a sample at the 251st national meeting of the American Chemical Society.

The simpler method strips away common chemicals, so gas chromatography can identify the unique organic components in the sample. When a library of the chemical components of different lipstick brands is developed, the technique can be more useful in the real world.

Public Enemy, Phone Home

One forensic tool which may or may not be useful came to prominence lately because it involved breaking into Apple iPhones used by terrorists in California. The FBI requested Apple give them a patch for the software which would let the government crack the password. Apple fought this in the courts.

A few weeks ago the government dropped its case saying that a third party (unidentified) had come forward with a way to crack the password (leaving the data still encrypted). However, some experts believe this could be a ruse intended to disrupt communications from terrorists and criminals. If suspects think the government can open their phones, then they will be less likely to communicate on these devices. But, If true, then iPhone users are now on notice that their private lives may be compromised.

One bright spot in the phone battle for privacy is that a court order would still be required to hack into a private phone to get usable evidence. This didn’t come into dispute in the terrorist case because the phone in question was actually a loaner from the terrorist’s employer, a government agency itself. However, because the FBI probably won’t divulge its methods in this case, the information can’t be used as “evidence,” anyway. Still it may be useful if they communicated with other terrorists.

Tracking Money

Tracing transactions involving bitcoins is another new forensic tool. Used in various countries, even for government transactions, and accepted by a number of online retailers such as WordPress and Overstock, bitcoins have a total value in the billions of dollars. They are now being used in an attempt to track criminal activity.

They are popular in online transactions. They can be either carried in an electronic wallet, or a physical representation of a bitcoin can be printed as a QR Code.

The actual bitcoins are stored on multiple servers (distributed storage) so even the loss of some mean nothing. What bitcoin holders (owners) actually have are the private decryption keys to the bitcoins. Without that unique identifier, no one can sell or use a bitcoin. (A backup is a good idea. One individual reportedly lost millions of dollars when he accidentally discarded his hard drive containing the only record of his private key.)

Although they are merely electronic records of encryption files and never pass through a normal bank, all bitcoin transactions must pass through Internet-connected server banks. So all exchanges are technically public.

There is no name attached to the bitcoin. Once an individual’s unique private key is identified, any of his or her transactions can be traced through the network, sometimes even linking them to a specific IP address.

But while the IP address of a physical computer can be linked to a general geographic location, it is far from accurate. (For example, computers normally think I am located in a city which is actually sixty miles away because that is where the Verizon hub is located.)

The Tor (anonymity) network, was formed not by a criminal organization but by various private and government agencies (including The U.S. State Department). It was created at the Naval Research Center for use by intelligence agencies. Current sponsors include the National Science Foundation and the U.S. State Department. Early sponsors included DARPA and Google. Tor makes it impossible for network snoops to bug your Internet traffic. With a simple click of a mouse, your IP address can be virtually relocated on the Net to foreign countries. That is the trick you sometimes see simulated on various TV shows and movies where the origin of Internet phone calls appear to be constantly shifting around the world.

The Tor Network can also be used by law enforcement in stings and it is a useful tool in hiding the location of dissidents and, one would presume, spies or diplomats.

Who’s Bugging You?

The basis for another new forensic tool exploits the novel concept that people could be uniquely identified by the precise analysis of the various bacteria which live on every person (microbiome). There are over one thousand different species found on humans.

Time of death can be determined using the other kind of “bugs,” that is, measuring the progression of blowflies, carrion beetles, millipedes, etc., as they attack the body. There are a lot of variables involved, so this isn’t very accurate.

But the latest twist is that it may be possible to determine time of death much more accurately or reliably by following the order in which various microbes colonize a body after death.

In Conclusion

It is interesting to note the comments of B. Manning, former Deputy Chief with the Appeals Bureau of the Bronx District Attorney’s Office. She retired in 1997 after twenty-three years on the prosecution side.

When asked about how juries viewed forensic testimony vs. eyewitness reports, Manning told “Perihelion” that forensic experts who use technical language had probably done so many times. “That didn't make it any less essential to the case. Eyewitnesses were frequently poor and less educated, often with an imperfect command of English, brought into an unfamiliar and perhaps intimidating setting, testifying about something that roused their emotions to some extent. I would imagine jurors would be more interested in eyewitnesses for the same reasons reality shows are so popular.

“The police and prosecutors would love to have every bit of evidence, forensic and otherwise. The reality is that, aside from the autopsy report or drug testing essential to prove a case, or an attempt to get a fingerprint match in major cases, it isn't possible because of budgetary constraints. By the same token, the police don't have the time to chase down and interview every potential eyewitness.

“Twenty years ago the only forensic evidence was routine and uncontested for all intents and purposes—tests for drugs, the medical examiner’s report, a breathalyzer test. It was very, very rare that useful DNA or fingerprint evidence was found. Ballistic evidence was almost unheard of. Forget shoe prints, fiber comparisons, and other exotica. I do remember one case involving handwriting analysis. I think blood splatter evidence might be a figment of someone’s vivid imagination. Maybe the Feds get into stuff like that.” 

John McCormick is a physicist, science/technology journalist, and author with more than 17,000 bylines to his credit. He is a member of The National Press Club and the AAAS. He recently launched the canine celebrating website, A to Z Dogs.