Mother of three Joann Parks smelled smoke one night in her home. Slowly, she rose, thinking it coming from outside, and opened the bedroom door to find a raging inferno. She heard her children screaming and did what she could to get to them, but the fire was so intense she couldn't get through the flames. So she immediately ran to the house next door and called the police and fire department, screaming for them to hurry.
Unfortunately, it was too late as the fire was too intense, and all three of her children were killed.
Fire investigators analyzed the scene and decided the fire had two points of origin, and as such it was arson and not accidental, since an electrical malfunction was unlikely to start a fire in two separate locations. Joann was convicted of the murder of her children and sentenced to life in prison.
Twenty-two years later, fire science looks completely different, and various non-profits set up by the ACLU and defense attorney organizations (you're welcome public) are in place to review cases of arson convictions. What the expert panel of fire science academics found was that the fire that killed Joann's children had "jumped" to a new location, making it look like it had two points of origin: a trait of fire the initial investigators didn't know about.
The panel decided that if the trial were held today, the fire investigators likely wouldn't have been able to testify because of their lack of basic fire science knowledge.
You would think she was immediately released with some money for 20 years in prison, right? Wrong. She is still sitting in a California prison, with people like Pulitzer Prize winning journalist Edward Humes fighting for her release. Hopefully one day California will right this wrong.
But Joann's story is far from unique. For decades, all across the country, innocent people have been falsely convicted of crimes they did not commit based on the testimony of "experts" in fields that are found to be scientifically inaccurate ten or twenty years later.
FORENSIC SCIENCE IS AN ART, NOT A SCIENCE
Much of forensic science is junk, plain and simple.
Before you call me crazy, let me first say that DNA is absolutely an amazing tool that has freed hundreds of innocent people from incarceration. People that were in prison 20 or 30 years have been able to prove their innocence and be released. It's been a huge step forward for the entire justice system. So there is some forensic science that is sound and does us good.
But what is forensic science exactly?
Forensic science is essentially the use, or we should say the attempted use, of scientific principles applied to criminal investigation by law enforcement or law enforcement ancillaries acting like law enforcement (Surprise! CSIs are not cops and couldn't arrest a suspect to save their lives). Scientists are called to trial as experts, and juries, through many detailed studies, have been shown to place more weight on their testimony than any other witnesses in trial. This makes the experts and the methods they use extraordinarily important. If we want to avoid innocent people being imprisoned or executed, or guilty people going free, the science behind forensics must be sound.
The problem is forensic science isn't like biology or chemistry, where modern theory is built on the foundation of the past. Fields of forensic science that have been used for decades to garner convictions are constantly being found to be junk science and inaccurate. For example, shaken baby syndrome, bite mark analysis, and fire investigation have all been found to be scientifically unsound, or outright false (in the case of bite mark analysis). Believe it or not, firearm and bullet analysis are close to being declared unreliable and we're seeing less and less ballistics experts testifying in trials.
I could fill pages with the number of people falsely convicted using these outdated and inaccurate methods.
But, as a representative example, let's take a look at how one specific field has changed: hair analysis.
We've all seen CSI and how hairs found at a scene match the suspect and it leads to David Caruso putting on his sunglasses while The Who starts blaring (CSI: Miami is the best of them, change my mind). The FBI for decades held three conclusions when comparing hair found at scenes or on bodies to suspects:
1. Inconclusive: Both similarities and differences were observed between the hair sample and the subject's hair and therefore a determination can't be reached with certainty that the hairs match
2. Exclusion: The hairs are not the same and are not a match
3. Conclusive: The hairs are sufficiently similar on a microscopic level and are a match
Big deal right? If you were on a jury and an experienced FBI agent or technician took the stand and said the hair found on a murder victim's body matches the defendant's hair, who claims he's never met the victim, would you convict? You would. Probably just based on that testimony.
So how did the FBI determine the hairs match? Some space-age looking machine powered by super-computers like on CSI?
The FBI trace evidence technicians would LOOK at two pictures, one the hair found at the scene and the other the subject's hair, and GUESS if they looked sufficiently similar to be a match.
I'm not kidding, they would compare pictures.
So what if the tech had bad eyes, was having a terrible day, had just gotten divorced and was distracted, was a smoker and had irritated eyes, had contact lens that had expired and he hadn't gotten new ones yet, had a previous head injury that affected his perception, was color blind, had some alcohol to drink, was on prescription meds that affected his vision or cognitive functioning...? What if a case was right on the borderline between match and no match; how would the technician decide which way to conclude? Flip a coin? Just a default position of matching?
This method of hair analysis is no longer used for obvious reasons. The National Academy of Sciences put out a report finding that there are no scientifically accepted data on the frequency of specific characteristics of hair distributed in the population. In other words, there's no evidence our hair is different enough to be used to identify us. They recommended only using what's called the STR test, which uses the hair root as comparison (most violent murders that don't involve a gun are likely to have hair pulled out which will contain the root).
What if there's no root? Then the NAS recommends mitrochodrial DNA testing to match hairs. But here's a problem they just found: with mitochondrial DNA testing of hair, close family members are an exact match. Meaning if your brother Charlie goes out and kills someone and leaves hair behind, that hair will be an exact match to you.
Tell me that isn't scary.
I didn't write this post to terrify anyone. The fact is, with shows like MAKING A MURDERER and the growing awareness of the flawed lab procedures used by law enforcement, forensic science is going through a transformation. Painfully and slowly, but it's trying. It's trying to actually follow scientific principles. Because for far too long we've used guesswork as science, and the people truly suffering are people like Joann who lost everything they had because the test used wasn't accurate. Hopefully forensic science actually gets its act together and we can stop destroying the lives of innocent people.