#ReadWithClutch · Book Review
Junk Science and the American Criminal Justice System, by Innocence Project attorney M. Chris Fabricant, documents how forensic disciplines with no scientific validation, bite mark analysis chief among them, sent innocent people to prison and death row, and how the legal system’s reliance on precedent kept that evidence admissible long after scientists had repudiated it. It is the rare true crime book where the institution, not the perpetrator, is the subject under examination.
Fabricant, the Innocence Project’s Director of Strategic Litigation, writes from inside the litigation: in 2012 the organization began searching for prisoners convicted on junk science, and three men convicted of capital murder became his clients.
Bite mark identification has never been validated by any governmental scientific body, yet courts admitted it for decades, and the Innocence Project counts dozens of exonerations among people convicted on bite mark testimony.
The book’s central mechanism is legal, not scientific: admissibility precedent compounds. Once one court accepts a technique, later courts cite the acceptance rather than the science.
The cases of Keith Harward, Eddie Lee Howard, Levon Brooks, and Kennedy Brewer show the pattern at full scale: confident expert testimony, conviction, decades of incarceration, and eventual exoneration when DNA or reinvestigation exposed the original evidence.
For readers of institutional accountability work, the book functions as a case study in how systems protect their own prior decisions, a dynamic that extends well beyond forensics.
The Book and Its Author
M. Chris Fabricant is the Director of Strategic Litigation at the Innocence Project, a former public defender, and a former clinical law professor. Junk Science and the American Criminal Justice System, published by Akashic Books in 2022 and reissued in an expanded paperback edition, is his account of the litigation campaign to remove discredited forensic methods from American courtrooms. The book follows courtroom battles from Mississippi to Texas to New York, anchored by three capital cases the Innocence Project took on after it began systematically searching for prisoners convicted on unvalidated forensic evidence.
The book has been widely reviewed and endorsed by figures who do not hand out praise casually in this space, including John Grisham and Pulitzer finalist investigative reporter Spencer Hsu. But endorsements are not the reason this book belongs on a Clutch Justice reading list. The reason is methodological. Fabricant does what this site argues all accountability work should do: he lets the record convict the system. The transcripts, the expert reports, the appellate opinions, and the exoneration filings carry the argument. The outrage, where it appears, is earned by the documents rather than asserted over them.
What Junk Science Actually Means
The term is not rhetorical. The book’s targets are forensic disciplines that were admitted into evidence for decades without ever being subjected to the validation that defines actual science: controlled studies, known error rates, demonstrated reproducibility. Bite mark comparison is the book’s central example, and the record on it is now unambiguous. Reviews by multiple governmental scientific bodies have concluded that bite mark analysis has no demonstrated scientific basis, and research has shown that certified examiners could not even reliably agree on what constituted a human bite mark.
The broader landscape was mapped by two federal reports that frame the book’s argument. The National Academy of Sciences concluded in its 2009 report, Strengthening Forensic Science in the United States, that apart from nuclear DNA analysis, no forensic feature-comparison method had been rigorously shown to reliably connect evidence to a specific individual. Seven years later, the President’s Council of Advisors on Science and Technology reached similar conclusions about pattern-matching disciplines, finding that several lacked the foundational validity required for courtroom use. The science had spoken twice, at the highest institutional level available. The courts, as the book documents, largely kept admitting the evidence anyway.
The failure Fabricant documents is not primarily a failure of individual experts, though the book contains those. It is a structural failure of gatekeeping: judges, not scientists, decide what juries hear, and judges defer to what prior judges decided. Scientific validity and legal admissibility are separate systems, and the second updates far more slowly than the first.
The Cases That Carry the Book
Keith Harward: Six Experts, Zero Doubt, Wrong
The Harward case is the book’s most complete demonstration of how expert consensus can form around a false conclusion. Keith Allen Harward was a sailor aboard a Navy ship in dry dock in Newport News, Virginia, when a 1982 home invasion left Jesse Perron dead and his wife raped. Because the assailant wore a Navy uniform, sailors were required to give dental impressions. Forensic dentists testified that Harward’s teeth matched a bite mark on the victim, with testimony framed in terms of scientific and medical certainty. In total, six forensic odontologists identified Harward as the source of the bites, including experts consulted by his own defense. Notably, the initial screening dentists had excluded him, then reversed their opinions after board-certified specialists made the identification.
Harward served 33 years. In 2016, DNA testing excluded him and identified another sailor from the same ship. The Virginia Supreme Court granted a writ of actual innocence, and he walked out of Nottoway Correctional Center in April 2016. Six credentialed experts, two juries, and three decades of appellate review had all ratified evidence that was wrong. That is not a story about one bad expert. That is a story about a discipline.
Eddie Lee Howard: Death Row on Invisible Bite Marks
Eddie Lee Howard was sentenced to death in Mississippi for the 1992 murder of Georgia Kemp, an elderly woman in Columbus. The forensic pathologist’s initial autopsy report noted no bite marks. Dr. Michael West, then regarded as a leading bite mark expert, later examined the body and reported finding bite marks not visible to the naked eye, which he attributed to Howard without qualification. That testimony was the state’s most important evidence.
In August 2020, the Mississippi Supreme Court vacated Howard’s conviction, writing that an individual perpetrator cannot be reliably identified through bite mark comparison. DNA testing of crime scene evidence, including the murder weapon, excluded Howard. The district attorney dismissed the charges, and Howard was exonerated in January 2021 after 26 years on death row. West himself had stated in a 2012 deposition that he no longer believed in bite mark analysis and that it should not be used in court. The technique’s own leading practitioner had disavowed it nearly a decade before the courts finished unwinding what his testimony had built.
Brooks and Brewer: The Same Failure, Twice, in the Same County
The paired cases of Levon Brooks and Kennedy Brewer in Noxubee County, Mississippi, show what happens when discredited forensics meets investigative tunnel vision. Two three-year-old girls were abducted, assaulted, and killed 18 months apart in the early 1990s. Brooks was convicted of the first murder and sentenced to life; Brewer was convicted of the second and sentenced to death. In both cases, the prosecution’s forensic foundation was bite mark testimony from Dr. West, working with medical examiner Dr. Steven Hayne, with West testifying in the Brooks case that no one but Brooks could have made the marks.
DNA testing in the Brewer case identified Justin Albert Johnson, a local man with a documented history of sexual assaults who had been an initial suspect in both crimes. Johnson confessed to both murders. Both men were exonerated in 2008, Brooks after 18 years in custody and Brewer after 15, including time on death row. Experts retained during the reinvestigation concluded the marks on the victims were not human bite marks at all. The same sheriff’s office and the same district attorney handled both cases, and twice the actual perpetrator was overlooked in favor of a man implicated by pattern evidence that could not survive scrutiny.
Read together, the three case clusters establish the book’s thesis empirically. The errors were not isolated. They were produced by a repeatable process: an unvalidated technique, an expert willing to testify to certainty, a court willing to admit it, and an appellate structure built to defer to all of the above.
The Lab is the Clutch Justice interactive tools suite: FOIA request generator, judicial report builder, decision trees, and a plain-language glossary for navigating court records and forensic claims.
Open The LabThe Mechanism: Precedent as Preservative
The most valuable analytical contribution of the book is its account of why correction takes so long. Forensic admissibility in American courts runs on precedent. When a technique is admitted once and survives appeal, that ruling becomes citable authority. Subsequent courts do not re-litigate the underlying science; they cite the prior admission. Over decades, a discipline can accumulate an imposing wall of legal validation that contains no scientific validation at all. Even as exonerations mounted, no state or federal appellate court had categorically ruled bite mark evidence inadmissible at the time of Harward’s 2016 release.
This is the same structural dynamic Clutch Justice documents in other institutional contexts. Systems are built to defend their prior outputs. A judicial disciplinary body that ratified conduct once has an institutional interest in not revisiting it. A court system that admitted a category of evidence for forty years faces an uncomfortable arithmetic when it considers excluding it, because exclusion implies that every prior conviction built on that evidence is suspect. The incentive structure rewards continuity over correction, and the cost of that continuity is paid by people like Harward, Howard, Brooks, and Brewer, in decades.
The exonerations in this book did not come from the system’s self-correction machinery. They came from outside pressure: DNA technology the original courts never anticipated, and a litigation organization willing to spend years per case. Where external forcing mechanisms do not exist, the book implies, the errors remain in place.
Assessment
As a reading experience, the book is structured as narrative nonfiction and moves like it. Some reviewers have noted that it is a litigation memoir woven through case studies rather than a systematic treatise on forensic reform, and that is a fair description of its shape. Readers seeking a comprehensive policy framework will need to supplement it with the NAS and PCAST reports themselves. What the book offers instead is something those reports cannot: a documented, insider account of what it costs, in years and in lives, to remove a single discredited technique from a system designed to keep it.
For the Clutch Justice readership, the book earns its place on the shelf as a methods text. It demonstrates pattern recognition across jurisdictions, sourcing discipline under adversarial conditions, and the long-horizon persistence that institutional accountability work requires. The villains, where they exist, are less interesting than the architecture that empowered them. That is the correct emphasis, and it is rarer in this genre than it should be.
Bluebook: Williams, Rita. #ReadWithClutch: Junk Science and the American Criminal Justice System by M. Chris Fabricant, Clutch Justice (July 6, 2026), https://clutchjustice.com/2026/07/06/readwithclutch-junk-science-fabricant/.
APA 7: Williams, R. (2026, July 6). #ReadWithClutch: Junk science and the American criminal justice system by M. Chris Fabricant. Clutch Justice. https://clutchjustice.com/2026/07/06/readwithclutch-junk-science-fabricant/
MLA 9: Williams, Rita. “#ReadWithClutch: Junk Science and the American Criminal Justice System by M. Chris Fabricant.” Clutch Justice, 6 July 2026, clutchjustice.com/2026/07/06/readwithclutch-junk-science-fabricant/.
Chicago: Williams, Rita. “#ReadWithClutch: Junk Science and the American Criminal Justice System by M. Chris Fabricant.” Clutch Justice, July 6, 2026. https://clutchjustice.com/2026/07/06/readwithclutch-junk-science-fabricant/.
Continue Your Investigation
If this reporting raised more questions, use the Clutch Justice ecosystem to keep going.