From Pennsylvania's death row, convicted murderer Daniel Dougherty, 51, is hoping to be freed based on his contention that "junk science" led a jury to convict him of setting a fire that killed his two sons.
And with mounting proof that eyewitness accounts are often wrong, New Jersey is examining how such evidence should be used, while a Pennsylvania committee is expected next month to make recommendations aimed at preventing wrongful convictions.
Fingerprint and firearm analyses, arson investigations, and forensic conclusions about bite marks, footprints, and hair and fiber comparisons have, in varying degrees, all been called into question. There are even growing concerns about what was once considered unassailable evidence of guilt - a confession.
"People in the criminal justice system, including judges, have a lot more skepticism than they once had," said Philadelphia Common Pleas Court Judge Benjamin Lerner, who has presided over criminal cases for more than 20 years.
The legal challenges around the country are an outgrowth of the revolutionary impact of DNA testing, which has freed 267 prisoners nationwide, including 10 in Pennsylvania and five in New Jersey, and caused four states, including New Jersey, to end the death penalty.
Eyewitness testimony was a factor in the convictions of 75 percent of those exonerated through DNA testing, according to the New York-based Innocence Project. Faulty forensic evidence contributed to 50 percent of the convictions. The project said that 25 percent of the defendants later found innocent had confessed or pleaded guilty, and that jailhouse informants were a factor in 15 percent of the overturned cases.
Despite the popularity of TV's CSI shows, many of the forensic methods used in modern crime-solving also took a hit when a National Academy of Sciences report said that all too often, forensic evidence was presented in court "without any meaningful scientific validation."
The need for improvement was "tremendous," said the 2009 report, which called for uniform standards, certification, and peer-reviewed research to ensure greater reliability. It said that in fires, for example, "rules of thumb" pointing to an accelerant being used "have been shown not to be true," and limitations of fingerprint identification may also have led to wrongful convictions.
The idea of bad forensics is nothing new to James McCloskey, the founder of Centurion Ministries in Princeton, which has won the release of 40 prisoners since 1983.
"Forensic evidence is like gold" in courtrooms, he said, and it can be extremely difficult, even impossible, to successfully challenge faulty scientific conclusions after a jury's verdict. Pennsylvania, he said, is one of the toughest states for a defendant to challenge a conviction based on new evidence.
David L. Faigman, a professor at the University of California Hastings School of the Law, said one day people would look back with dismay on all the defendants convicted of crimes based on "pseudoscience" and old-fashioned techniques.
"Our children's children . . . will say the legal system was profoundly ignorant about science," said Faigman. "We're turning a corner, but like a luxury liner, it takes a long time to change direction. And there are a lot of hurdles."
He said few states are "forward thinking," though New Jersey's decision to review eyewitness identification makes it a standout.
Law enforcement officials defend the integrity of many long-used investigative techniques. If anything, "the technology has been enhanced by advancements in science," said Upper Darby Police Superintendent Michael Chitwood, a former homicide detective in Philadelphia.
Dauphin County District Attorney Edward Marsico, president of the Pennsylvania District Attorneys Association, said he didn't buy all the National Academy of Science findings and still had faith in evidence like fingerprint and ballistic analyses.
"Are there false confessions? Sure. Are there bad eyewitness IDs? Absolutely. What's the best way to combat that? We're not sure," Marsico said.
That is precisely the dilemma facing courts.
While courts were relatively quick to accept the reliability of DNA testing, they are moving cautiously on challenges to other types of evidence.
"What strikes me is how resistant we are as an institution to change," said David Rudovsky, a law professor at the University of Pennsylvania, noting that juries here are still told that a victim's certainty can be a sign of reliability, though so many DNA cases have proven otherwise.
Psychological studies have shown problems with false confessions and eyewitness identification.
Such studies have shown that suspects give false confessions for reasons ranging from suggestive personalities to pressure from police. Studies on eyewitness identification indicate that a witness' memory can be affected by the passage of time, the presence of a weapon, and other factors, such as trauma or being asked to identify a person of a different race.
Demonstrating its concern about eyewitness identification, the New Jersey Supreme Court last year appointed a special master to make recommendations on its use.
In a voluminous report, retired Appellate Judge Geoffrey Gaulkin said judges should instruct jurors about the vagaries of eyewitness identification and hold pretrial hearings at which prosecutors must establish the reliability of the witness. The New Jersey Supreme Court is expected to hear oral arguments on the report this year.
Pennsylvania has a distance to go. In 1995, the state Supreme Court barred defense lawyers from calling expert witnesses to testify about the unreliability of eyewitness identification.
And among the 10 Pennsylvanians who have been found innocent by DNA analysis was Bruce Godschalk, who spent 15 years in prison after confessing to two rapes and being identified by a victim.
The state's Wrongful Convictions committee, composed of 48 people from all sides of the justice system, is expected to recommend extensive changes next month. They include the taping of all police interrogations, new lineup procedures where eye witnesses are involved, and establishment of a statewide forensics board.
"There's a lot at stake here," said Duquesne University law professor John T. Rago, who chairs the committee.
Questions about fingerprint analysis, long considered nearly sacrosanct in the justice system, began surfacing in the late 1980s. Then, in 2004, a mistaken identification in a terrorism case drew international attention when Oregon lawyer Brandon Mayfield was taken into custody after the FBI said his print was on a bag of detonation devices found during the investigation of the Madrid train bombings.
Two weeks later, Mayfield was cleared after an Algerian national and known terrorist was identified as the source of the print.
In Philadelphia, Kenneth Mapp is hoping to show that fingerprint analysis was flawed in his case, slated for trial Monday. Mapp was charged in connection with the 2009 robbery of a pizza-takeout restaurant after police lifted a print from the cash register.
Detectives said that an analysis showed it was a positive match for Mapp. He also had been a military marksmanship expert, had a recent gun license, and had a criminal record for a small drug-possession case in Montgomery County, according to court records.
In court papers, his lawyer, Assistant Defender Jordan F. Barnett, cited 21 cases of mistaken identifications by print examiners since 1990, including a Delaware County murder case in which a defendant, jailed for two years, was freed after a panel of experts agreed that his fingerprint was not at the crime scene.
Arson investigations are another area of review. The National Academy of Sciences report found serious deficiencies in how fire investigators decide whether a fire had been intentionally set.
Locally, lawyers for Daniel Dougherty are arguing that that kind of flawed science led to his death-row conviction for setting a 1985 fire that killed his two young sons, ages 3 and 4. Prosecutors, however, say their case remains strong.
Dougherty contends he was asleep on a sofa in his Northeast Philadelphia home after a night of drinking when the fire erupted. He ran outside and grabbed a garden hose, a neighbor said, trying to douse the flames. But the fire was relentless, and smoke killed his two sons asleep upstairs.
"My name's mud. I have to die for what I did," a police officer quoted the distraught Dougherty as having said that morning. A fire marshal concluded burn patterns showed fires had been started in three places - in a sofa, in a love seat, and under the dining room table.
More than one point of origin is classic evidence of arson, but arson experts, after years of examining how fires burn, now know that accidental fires involving "flashovers" - giant accelerations - can leave such burn patterns.
Dougherty was arrested 14 years after the fire when his second ex-wife, embroiled in a custody battle with him over another son, told police he had confessed to setting the fire. Two jailhouse informants also testified that Dougherty had admitted igniting the blaze.
His trial lawyer did not consult an arson expert, and his current lawyers, David Fryman, Shannon Farmer, and Robert B. Dunham, contend Dougherty deserves a chance to show that no one following scientifically accepted protocols on arson could have concluded the fire was intentionally set.
"This man is innocent. This man is entitled to have a court review his claim," Dunham told U.S. District Judge Juan Sanchez during a recent hearing. "We plead with you, Your Honor, to do the right thing and let him have his day in court."
Retired New Jersey Supreme Court Justice James H. Coleman Jr. said he expected it would take some time - and lots more courtroom battles - to sort out the trouble spots in forensic evidence and eyewitness identification.
"It will likely be slow," said Coleman, who wrote the precedent-setting 1999 opinion that requires judges in New Jersey to tell jurors about studies showing that people of one race often have difficulty identifying a perpetrator of another race.
That opinion came in the case of McKinley Cromedy, an African American defendant convicted of raping a young white woman who had said she was certain he was her attacker.
He spent five years behind bars before DNA testing finally exonerated him.
Contact staff writer Emilie Lounsberry at firstname.lastname@example.org.