Chapter 34 — Key Takeaways

A scannable one-page card. For the full argument and the figures, see index.md.

The core claims

• A wrongful conviction is a cascade, not a single villain. It is almost always a convergence of several reinforcing errors: a mistaken eyewitness names a face, a detective's tunnel vision narrows the investigation to confirm it, a false confession is fed scene details across the interrogation table, and a forensic examiner — knowing the wanted answer — reads ambiguous evidence toward it and testifies to a "match." Each error borrows credibility from the others; no one need be dishonest.
• DNA audited the system's past by exclusion. Post-conviction DNA testing could be applied retroactively to stored biology, and a non-match excludes close to conclusively. That is the §1.6 asymmetry turned backward onto old convictions — the key that opened doors a century of appeals could not.
• The causes are patterned and they combine. The same handful of contributing factors recur in the same combinations — which means they are diagnosable and, in principle, preventable.
• Forensic science is a routine participant — present in roughly half of the DNA exonerations. Not the most common factor (that is mistaken eyewitness identification), but among them, and uniquely dangerous because it wears the prestige of science (the CSI effect makes a jury stop questioning).
• Forensic failure splits into two sins, with two different fixes. First sin: methods never valid (bite marks, hair "matching," arson folklore) — fix is exclusion. Second sin: valid methods overstated ("is" for "is consistent with"; "zero error rate") — fix is calibrated language and measured error rates. Conflating them produces bad reform.
• The path out is obstructed by design. Ordinary appeals test process, not factual innocence; evidence is lost or destroyed; there is no federal constitutional right to post-conviction DNA testing (Osborne, 2009); a clean exclusion is still not automatic freedom. Exoneration happens when the truth emerges and someone fights long enough to make the system act.
• Reforms work, unevenly — and most reduce to "let the evidence refute the theory." Double-blind lineups, recorded interrogations, foundational-validity requirements, blind/context-managed analysis, Brady enforcement, and conviction integrity units. Each is a door propped open against the cascade's failure to admit disconfirming evidence.
• The cost is double. The innocent imprisoned (some, like Willingham, executed) and the guilty left free — a failure of public safety, since the true perpetrator sometimes offends again. Getting the right answer and protecting the public are the same project.

The method-validity verdict (NAS 2009 / PCAST 2016) — as the exonerations reveal it

This chapter does not validate a new method; it reinterprets every method through the exonerations. The verdict is on the two sins of forensic failure.

Where this leaves the field: humbled, not cynical. The same DNA that freed the exonerees is the field's rigorous best; the bite marks and overstated prints that convicted them are its worst — and the difference is not degree but whether the core claim is validated and whether the testimony stayed inside it. The validity spectrum (Theme 2) made into a matter of innocent lives.

What you can honestly say on the stand

• On an examiner's duty (any method): "I can state what this comparison supports — that the defendant cannot be excluded as a possible source — and I must state its limits: I cannot identify him to the exclusion of all others, and this method carries a known, non-zero error rate." (The verb tracks the validity; "is consistent with," not "is.")
• On the meaning of agreement: "Corroboration is only as strong as the independence of the things corroborating. If the confession, the identification, and my comparison all trace to the same original lead, their agreement is not independent confirmation — it is one possible error repeated."
• On a junk method: "This method has no established foundational validity for the claim being made; no degree of examiner skill or sincerity can supply what the method lacks."
• What you must NOT say: "to the exclusion of all others" or "zero error rate" for a human-judgment method; "this proves he did it"; "the experts all agree, so it's certain"; or any conversion of a weak, class-level inclusion into proof.

Key terms (one line each)

• Contributing factors to wrongful conviction — the recurring causes (mistaken eyewitness ID, false confession, flawed/misapplied forensics, incentivized informants, official misconduct) that combine rather than act alone.
• Exoneree — a person convicted of a crime and later officially cleared of it; the human being to whom an exoneration happens.
• Post-conviction review — the legal/investigative channel by which a final conviction is reexamined for error after ordinary appeals are exhausted; built to resist the reopening exoneration requires.
• Conviction integrity unit (CIU) — a division inside a prosecutor's office that reviews innocence and serious-error claims in its own cases; effective only with genuine independence, resources, and will.

The cold-case line

Read the cascade against the early Mill Creek investigation and it maps almost perfectly: the comfortable "accidental fire" frame, the narrowing onto a convenient local, and an eleven-hour interrogation that produced a confession whose "non-public" details came from the interrogators. The science — the autopsy timeline (Ch. 11), the cell-site data (Ch. 25), and the DNA exclusion (Ch. 7–9) — excludes Cody Renner at full strength. The science exonerated an innocent man. Naming who it points toward instead is the capstone's job (Ch. 39), not this chapter's.

The themes this chapter advanced — all four converge here

• Exclusion over proof (Theme 1) — DNA's clean exclusions opened the cells; the overstated inclusions filled them. Reserving "proves" and letting exclusion do its surer work prevents the most of these failures.
• The validity spectrum (Theme 2) — bite marks are not DNA; the gap, not any examiner's competence, separated the methods that convicted the exonerees from the one that freed them.
• Cognitive bias is the chief threat (Theme 3) — the cascade is a bias mechanism; the examiners were sincere, which is why blind, context-managed analysis, not exhortation, is the fix.
• The CSI effect cuts both ways (Theme 4) — juries over-trusted weak forensic testimony confidently presented; communicating uncertainty honestly is the skill that stands between weak evidence and a wrongful conviction.