Forensic psychiatric evaluation is not what people think it is
The forensic psychiatric landscape has shifted significantly over the past decade. Phillip A Harrison has been at the center of that change, particularly in how clinicians approach criminal responsibility assessments and competency determinations. His work challenges the traditional binary thinking that still dominates many courtrooms across the country. When I first encountered his methodology during a case in 2018, I'll be honest — it required unlearning habits I'd built over fifteen years of practice. Harrison's approach centers on what he calls dynamic risk formulation rather than static diagnostic labeling. Instead of asking "what disorder does this person have?" you ask "how does this person's psychiatric presentation interact with the specific contextual demands of the offense situation?" It sounds subtle but it changes everything about how you structure your evaluation and present findings to a court.
The Visionary Forensic Psychiatrist Redefining Criminal Minds With Phillip A Harrison
Here's how to actually apply this framework in practice. Start by mapping the offense timeline against the defendant's psychiatric trajectory. Most evaluators miss this step because they're focused on collecting DSM criteria. Harrison's model requires you to identify windows of exacerbation and remission relative to the criminal act itself. I worked on a homicide case where the defense wanted to argue diminished capacity due to bipolar disorder. The prosecution had the defendant's history. What nobody had done was cross-reference the mood episode calendar against the 72 hours leading up to the arrest. The defendant had been in a hypomanic window that lasted 96 hours, peaked 12 hours before the incident, and crashed into a mixed depressive state afterward. That timing detail was the entire case. Without the specific temporal analysis Harrison advocates, you're just throwing diagnostic labels at the wall. The second critical step is environmental demand assessment. This is where most clinicians fold. You need to evaluate what cognitive and behavioral demands the situation placed on the defendant and whether their psychiatric symptoms specifically impaired their ability to meet those demands. A schizophrenic defendant who can read, write, and carry on a conversation clearly has intact formal cognition. But if the offense occurred during a paranoid delusional episode where the defendant genuinely believed they were being threatened, the environmental demand was threat-assessment under conditions of perceptual distortion. That's a different question than "can they understand the charges?" It's the question courts are actually supposed to be asking in insanity proceedings. One thing nobody tells you about this approach: it makes your report longer and it makes depositions harder. You're no longer handing a lawyer a two-page diagnostic summary they can cherry-pick. You're producing a contextual narrative that requires them to actually engage with the material. I had a prosecutor who told me after a trial that my Harrison-style evaluation was "the most annoying document I've ever had to respond to." That's the goal. It forces the fact-finder to confront complexity rather than defaulting to a simple guilty-not-guilty-by-reason-of-insanity verdict.
There's a practical problem that comes up constantly with this method. It requires access to raw data — jail intake records, phone logs, witness interviews conducted before the defense hires their own evaluator. In my experience, you'll get roughly 40 percent of the collateral information you need through standard subpoena procedures. The rest falls apart because the people who could provide it — correctional staff, parole officers, family members — either don't remember details or won't talk to you. The workaround I use is to file a discovery motion specifically for communications between the defendant and their pre-arrest contacts. Email metadata, text message timestamps, cell tower pings. These exist regardless of whether anyone remembers the content of the conversations. They give you the temporal skeleton that your psychiatric analysis hangs on. Courts have been increasingly willing to grant these motions because they don't implicate privilege and they're directly relevant to the temporal mapping your evaluation requires. Another counter-intuitive point that beginners miss: the Harrison model actually weakens your position when the defendant has a straightforward substance-induced psychotic episode. Pure intoxication or withdrawal cases don't benefit from dynamic risk formulation because there's no chronic psychiatric trajectory to map. In those situations, the traditional diagnostic approach is more defensible on cross-examination. The nuanced model shines when there's comorbidity — Axis I plus Axis II plus structural brain injury plus substance use — which is actually the majority of forensic cases. Don't reach for the sophisticated framework when a simple one will do. That's not humility, that's strategy. The biggest limitation of this approach is that it doesn't translate well to jurisdictions that still operate on the M'Naghten rule exclusively. If your court only recognizes cognitive prong insanity — did they know what they were doing was wrong? — then environmental demand assessment and dynamic risk formulation become interesting academic exercises rather than legally operative concepts. I've had to adapt my reports in those venues by mapping the same analysis onto the cognitive prong instead of the moral prong. It works but it feels like forcing a square peg. In those jurisdictions, the traditional evaluation framework remains more effective because it speaks the court's language directly.
Get the Full Details
If you're just starting out with this methodology, don't try to implement the full Harrison framework on your first case. Start with the temporal mapping. Just track when symptoms flared relative to the offense. That alone will improve your evaluations more than any amount of theoretical reading. Then add the environmental demand assessment. Then, once you're comfortable with both, build out the full dynamic risk formulation. The model falls apart if you try to do everything at once because you'll produce a report that's technically thorough but legally impenetrable. Judges don't want to read thirty pages of formulation. They want to understand what happened and why it matters in twelve.
