A Tool Built to Protect Kids Gets Read Backwards
A Child-Testimony Credibility Tool, Read Backwards documents Criteria-Based Content Analysis — an 18-point checklist forensic psychologists use to judge whether a child's testimony reflects a real memory — reproduced in The Behavior Ops Manual as a guide for what features to plant when constructing a false memory in someone.
A Child-Testimony Credibility Tool, Read Backwards documents Criteria-Based Content Analysis — an 18-point checklist forensic psychologists use to judge whether a child's testimony reflects a real memory — reproduced in The Behavior Ops Manual as a guide for what features to plant when constructing a false memory in someone. False Confession Psychology documents the opposite direction of the same underlying fragility: how legitimate interrogation pressure can make an innocent person generate a false confession that reads, to everyone in the room, exactly like a true one.
CBCA exists because a real memory and a coached one don't look the same up close — real memories have specific, sometimes odd, "checkable" texture (an unexpected detail, a mentioned smell, a self-correction) that fabricated ones tend to lack. That's the whole reason the test is trusted in a courtroom: it's supposed to be hard to fake convincingly. Turn the checklist around and it becomes a build spec — plant this many time-boundaries, plant a couple of decontextualized details that go nowhere, plant one self-correction, and now the fake memory passes the same test built to catch it. False confessions run the identical inversion from the interrogation side: a technique built to extract a true account (sustained pressure, repeated retelling, suggestion of what "probably happened") can manufacture a confession with all the surface markers of authenticity, in someone who did nothing.
Both cases point at the same underlying fact: any diagnostic built to distinguish real from fabricated psychological content is, by definition, also a specification for fabricating content that passes the diagnostic — the more precisely a test names what "real" looks like, the more precisely it also names what to fake. That's not a flaw specific to CBCA or to interrogation technique. It may be a structural property of any test whose criteria are public and legible enough to be useful in the first place.
- A third example from a domain with zero overlap with psychology or interrogation — a forgery-detection method in art authentication, a plagiarism-detection method in writing, a fraud-detection method in finance — where naming the detection criteria publicly is shown to hand forgers/fraudsters the exact recipe for passing detection. If that pattern holds outside psychology entirely, this becomes a real epistemics finding (something like "legible tests are gameable tests") rather than a coincidence of two adjacent forensic-psychology cases.