Minnesota’s fourth competency evaluation was built to justify detaining, medicating, and continuing to prosecute Matthew Guertin. Measured against the U.S. Supreme Court holdings that bind every state, it does the opposite of its job — on point after point, the fact the State needs is refuted by the State’s own document. Every finding is inverted, and every inversion runs one direction, which is how you tell carelessness from design.
Nature of this record.
The two companion reports in this series read the fourth Rule 20 competency evaluation on its own face and read the statute that evaluation serves. This record measures both against the one standard neither the examiner nor the legislature can rewrite: the federal constitutional floor that binds every state. The method is the same as the rest of the series — the document under examination supplies the evidence against itself — and here it does so with unusual force. On point after point, the fact the State needs in order to detain, medicate, or continue prosecuting Matthew Guertin is contradicted by the State’s own report, by the examiners who actually met him, and, at the threshold, by the statute’s own applicability date. Every Supreme Court holding quoted below is reproduced verbatim from the opinion; every claim about the report is checkable against the report. Where a proposition is this record’s argument rather than settled law, it is labeled as argument, and the honest limit of each claim is stated alongside its strength — because the strong version survives only when the overclaim is left out.
1. The floor that binds every state
Nothing in the argument that follows requires the fine print of Minnesota’s competency chapter. It is enough to measure the fourth report against the federal constitutional floor — the line of United States Supreme Court holdings that binds every state and that no state statute, and no court-appointed examiner, may drop beneath. That floor begins with a definition, and the definition is the whole reason the rest of the report collapses.
Competency is a functional question, not a medical one. It asks whether the defendant can understand the proceedings and work with counsel — nothing more.
The “test must be whether he has sufficient present ability to consult with his lawyer with a reasonable degree of rational understanding — and whether he has a rational as well as factual understanding of the proceedings against him.” — Dusky v. United States, 362 U.S. 402 (1960)
Read what is not in that test: no diagnosis, no dangerousness, no opinion on the right medication, no view on where the defendant should live. A person can be floridly ill and competent; a person can carry no diagnosis and be incompetent. Illness is not the question. This matters here because almost nothing in the fourth report is addressed to Dusky’s question. Its content is medication, a substance-use label, a dangerousness finding, and an opinion on the defendant’s living arrangements — the four subjects the competency test excludes. The instrument was ordered to answer one narrow question and answered four different ones instead.
The sections that follow take each of those four excursions and measure it against the floor. On every one, the same thing happens: the finding the State needs is refuted inside the State’s own document.
2. The threshold question: does this chapter even apply to Guertin?
Before any of that, there is a question that comes first in time and first in logic, and it is the newest point in this series: whether Minnesota’s competency chapter — §§ 611.40–611.59, the machinery the fourth report runs on — even governs this case.
The dates decide it. Guertin’s competency track was initiated in early 2023:
March 10, 2023 — the first Rule 20 evaluation (Dr. Rogstad) is submitted into the record.
July 7, 2023 — the first contested competency hearing is held.
July 13, 2023 — the first finding of incompetency is entered.
The new chapter did not yet exist as operative law. Enacted in 2022 (Laws of Minnesota 2022, ch. 99) and refit before it ever took effect (Laws 2023, ch. 14, which swapped “restoration” for “attainment”), the chapter carries an operative date of April 1, 2024 — pushed back from an original July 1, 2023 — and it applies, by its own terms, only to competency determinations initiated on or after that date. Guertin’s determination was initiated more than a year before the trigger. His own first examiner said as much in the March 2023 report, in plain words: “These statutes do not exist in Minnesota at present.”
That sets a clean dilemma, and its value is that both horns cost the State something.
If the old rule governs — Minn. R. Crim. P. 20.01, because the determination was initiated before the operative date — then the old rule contained no § 611.49 “supervision or monitoring” escape hatch from the three-year felony dismissal. Under the old rule the charges simply die at three years: no notice-of-intent-to-prosecute stretching the clock to ten, no perpetual court-ordered supervision. That is the better outcome for Guertin, and it is the outcome the statute’s own applicability text points to.
If the new chapter governs anyway — then the State is applying it against its own applicability clause, and it does not get to enforce the convenient provisions while ignoring the one that says the chapter should not reach this case. Read as written, the chapter still fails him in the ways the companion record on the statute lays out.
The honest counter, stated so it can be met. Courts frequently treat competency procedure as “procedural” or “remedial,” a category that can let a new procedural scheme reach cases already pending when substantive law would not. The State will reach for that rule.
The rebuttal. This chapter is not merely procedural. It changes substantive outcomes — whether and when the charges must be dismissed, whether the State may compel medication, and how long a person may be held under court control. A scheme that decides those things is not a neutral change of forms; and an express applicability clause is an express applicability clause. The legislature wrote a trigger date into the law. Guertin’s case sits on the wrong side of it. (That the clause controls is the settled part; how a court resolves the procedural-retroactivity question is the argument.)
3. The report defeats its own dangerousness finding
Set the applicability question aside and grant, for argument, that the chapter applies and that Guertin is mentally ill. The fourth report still fails, because its most consequential finding — that he poses “a substantial likelihood of physical harm to self or others” — is contradicted by the report’s own contents. This is the point at which the document becomes its own best rebuttal.
Start with what the Constitution protects. A State may not confine a person merely because he is mentally ill; it must show something more, and willing family support is not a symptom but a shield:
“a State cannot constitutionally confine without more a nondangerous individual who is capable of surviving safely in freedom by himself or with the help of willing and responsible family members or friends.” — O’Connor v. Donaldson, 422 U.S. 563 (1975)
Now read the report’s own observations against that sentence. The examiner’s only direct findings are that Guertin was calm, appropriately dressed, well-groomed, and socially appropriate. Add the conduct no one disputes: he drove himself to the county government center, arrived on time, understood the proceeding well enough to knowingly decline the interview, is seeing his own treating providers, is taking his prescribed medication, is complying with the terms of his release, made bail, and has a mother who takes him in. That is not a sketch of a man who cannot survive safely in freedom. It is a point-by-point portrait of exactly the person O’Connor places beyond the State’s reach — assembled entirely from the State’s own report.
Against that, the dangerousness finding is contradicted three ways, all inside the record:
Every evaluator who actually examined him found the opposite. Across all three prior reports the record is uniform — he “disclaimed any suicidal or violent thinking,” had “no history of assaultive behavior,” and was “not deemed an imminent risk” — and it held even in the third report, produced by an examiner who had never met him. The fourth evaluator flipped that record to “Yes” on a brief observation, after he attended but declined the interview, while introducing not a single new fact about harm.
There is no evidence of harm to others anywhere. No victim, no assault; the sole underlying incident was a man firing a weapon into the air — “into the sky and trees,” in a prior evaluator’s own words — to summon police. “Dangerous to others” has nothing in the record to attach to.
The only realized risk is self-directed — and the report itself explains it away. The report attributes the February 2026 suicide attempt to Guertin’s “fears he would be sent to the state hospital.” That is the circular tell: the State’s own threatened outcome produced the crisis the State now cites to justify that outcome. The dangerousness is manufactured by the process, not exhibited by the man.
The federal floor requires more than a checkbox here, too. Confinement on a mental-illness rationale is permitted only while the State can prove the person is presently both ill and dangerous:
“the acquittee may be held as long as he is both mentally ill and dangerous, but no longer.” — Foucha v. Louisiana, 504 U.S. 71 (1992)
A retrofitted “Yes,” contradicted by three prior examiners and driven by the State’s own threat, is not the proof of current dangerousness Foucha demands.
How to deploy this — and how not to. The temptation is to argue that O’Connor wins automatically because Guertin has willing family. It does not, and overclaiming it hands the State its answer: O’Connor’s shield protects the nondangerous, and the State checked the dangerousness box precisely to escape it. The correct argument is narrower and stronger. It is not “O’Connor applies”; it is that the dangerousness finding is unsupported and internally contradicted, so it cannot carry the weight the State needs it to carry — and once that finding falls, O’Connor and Foucha bar the confinement it was built to justify. The report needs the finding to be true in order to reach past the constitutional floor. The report is the best available evidence that it is false.
4. The family shield, inverted
The dangerousness finding does not stand alone. It is one half of a single maneuver, and the other half is visible in one sentence of the report: “Concerns were raised that his mother was also mentally ill and that living with her is exacerbating his symptoms.”
Hold that against O’Connor again. The Supreme Court named “willing and responsible family members or friends” as a reason a person cannot be confined. The report takes the exact fact the Constitution makes protective — a family willing to house and care for him — and re-casts it as a reason to confine. It does not merely ignore the holding; it turns the holding’s protective fact into its opposite.
Why would a report do that? Because O’Connor has two exit conditions, and a family willing to take the defendant in closes one of them. Read the two moves together:
Neutralize the family — by asserting the mother is “also mentally ill,” so her willingness to care for him can be dismissed rather than credited.
Manufacture the dangerousness — the finding examined in §3 — so the “nondangerous” half of O’Connor’s protection is defeated as well.
Either one left standing, and O’Connor still bars confinement. Only by doing both at once does the report clear the holding. The mother sentence and the dangerousness retrofit are not two separate defects that happen to appear in the same document. They are the two halves of one engineered work-around, aimed at the two things O’Connor says the State must overcome — and the report supplies the evidence for both halves being contrived, because the same treating record it cites shows the prescriber crediting the mother, not diagnosing her.
5. Overriding the doctors who actually examined him
The through-line of the report is an inversion. Guertin is already in treatment and complying — seeing a therapist and a prescriber, taking his medication. He declined to sit for the court’s examination, which is his right, so the examiner conducted no clinical interview — only a brief observation before he declined — and built the report almost entirely from records. From that non-examination the report concludes that he needs to come off his Adderall, out of his mother’s home, and onto more neuroleptics — overruling, on paper, the clinicians who see and examine him in person.
Clinically, that inverts the basic hierarchy of medical evidence. Direct, ongoing contact outweighs a records review; it is not close. The people who examine him repeatedly (Plymouth Psych Group) reached the opposite conclusions, and each of the report’s overrides collides with a specific examining professional:
On Adderall: the treating prescriber re-wrote the same Schedule II prescription Guertin has held for years — and did so after a suicide attempt, the single most cautious moment on any prescriber’s calendar. A physician stakes his federal prescribing registration on that decision. It is a high-conviction clinical judgment that the ADHD is real and the medication is not the problem. The report re-labels the same medication “Stimulant Use Disorder.”
On living with his mother: the treating record shows the prescriber called the mother twice, as collateral — crediting her account, not treating her as a source of contagion.
On more neuroleptics: the treating team diagnosed acute stress and ADHD, not the psychosis the court’s label requires. The report recommends antipsychotics to “restore” a competency the examiner never tested, for a disorder the examining clinicians did not find.
Two lanes run through this, and keeping them separate is what makes each one hold.
The first lane is the standard of care — a professional and tort concept (practice guidelines, enforced through malpractice law), not a constitutional rule. There is no Supreme Court holding that “treating doctors always outrank a record reviewer,” and a forensic competency examiner is not per se barred from disagreeing with treating providers; the examiner answers a legal question, not a treatment question. So the honest statement is that a record reviewer overruling the clinicians who examine him in person runs against ordinary clinical practice — not that it is, by itself, unconstitutional.
The second lane is constitutional, and it is where the override becomes fatal — because the report does not merely diagnose, it recommends compelling neuroleptic medication. Forced antipsychotic medication sits under three holdings, and all three require that the drugs be in the patient’s own medical interest:
forced antipsychotic treatment is permissible only “if the inmate is dangerous to himself or others and the treatment is in the inmate’s medical interest.” — Washington v. Harper, 494 U.S. 210 (1990)
“The [Fourteenth Amendment] affords at least as much protection to persons the State detains for trial” — Riggins v. Nevada, 504 U.S. 127 (1992) — which would have required a showing that the medication “was medically appropriate and, considering less intrusive alternatives, essential for the sake of Riggins’ own safety or the safety of others.”
before ordering medication to restore trial competence, “the court must conclude that administration of the drugs is medically appropriate, i.e., in the patient’s best medical interest in light of his medical condition.” — Sell v. United States, 539 U.S. 166 (2003)
Sell also warned that such orders “may be rare.” This is where the caveat from the first lane turns and cuts for Guertin. A reviewer who conducted no clinical examination of him, recommending forced neuroleptics over the diagnoses of the clinicians who actually examined the patient, cannot satisfy Sell’s fourth requirement — “the patient’s best medical interest” — because the very treating record the report cites establishes the opposite: ADHD, acute stress, and a re-prescribed stimulant. The State’s own attachment refutes the medical predicate the Constitution demands.
One point of precision on timing. Sell and Harper bite hardest at the moment the State actually moves to compel medication, not at the moment an evaluator recommends it. The fourth report is not yet that motion. But it is the document that builds the predicate for it — and it builds a predicate that cannot be met, because the record it assembles argues against the very order it recommends.
6. Every inversion runs the same direction
Step back from the individual findings and a single move repeats across all of them. The report does not make scattered mistakes. It makes the same mistake, over and over, and the mistake is always an inversion: it takes a fact that points one way and enters it into the record pointing the other. The sections above have already traced most of them:
Willing family, which O’Connor names as a reason a person cannot be confined, is entered as a reason to confine him.
A calm, compliant, non-assaultive man — with three prior no-risk findings and no victim anywhere in the record — is entered as a “substantial likelihood of physical harm.”
The clinicians with the most contact, who examine him in person, are outranked by the reviewer with the least, turning the basic hierarchy of medical evidence upside down.
A stimulant re-prescribed after a suicide attempt by a physician staking his federal license on it — the strongest clinical vote of confidence a prescriber can cast — is entered as “Stimulant Use Disorder.”
Guertin’s own 2023 account of acute stress, which his treating team later reached independently, is entered as the delusion.
Add the word at the center of the recommendation. “Restoration” presupposes a capacity that was lost; the report applies it to a capacity that was never lost, and quietly redefines it as a change of belief. Even the governing verb is inverted.
Any one of these, standing alone, could be an error. Together they are not — and the reason is a single feature they all share: they run the same direction. Every inversion, without exception, moves toward the same three outcomes: detain, medicate, keep the case open. Not one of them cuts the other way. That is the fact that separates carelessness from design. A careless report inverts at random, and random inversions would sometimes fall in the defendant’s favor — a missed risk factor, an overcredited strength, a diagnosis softened by mistake. Here nothing ever falls in his favor. A pattern of errors that never once helps the person being evaluated is not a pattern of errors. It is a direction — chosen, and held.
And that direction is what makes the document self-defeating. Each inversion is contradicted by the very record it sits inside: the treating notes, the prior reports, the examiner’s own observations. So the more consistently the report inverts, the more completely it testifies against itself — because a truthful record would not have to be turned over on every page to reach the conclusion this report reaches. The consistency is not the report’s strength. It is its confession.
7. Conclusion: the report as its own rebuttal
Measured against the federal floor, the fourth report needs four things to be true. It needs Guertin to be incompetent under Dusky — but its content never reaches Dusky’s functional question, and the man who drove himself in and knowingly declined an interview plainly meets it. It needs him to be dangerous under Foucha — but every prior report found the opposite, and the only self-harm in the record is one the report blames on the State’s own threat. It needs his willing family to be disqualifying rather than protective under O’Connor — but the Constitution names willing family as a reason he cannot be held, and the treating prescriber credited the mother the report tries to discredit. And it needs forced medication to be in his best medical interest under Sell — but the treating clinicians it cites diagnose acute stress and ADHD and re-prescribed the very medication it attacks.
On each of the four, the fact the State requires is refuted by the State’s own document. That is not the profile of a competency evaluation. It is the profile of a conclusion written first and a record turned to fit it — turned, as the section above shows, in one direction every time. The most damaging witness against the fourth report is the fourth report. It was produced to justify continued control of Matthew Guertin, and read closely, against the only standard that outranks it, it does the opposite: it rebuts itself.
Authorities
Federal constitutional floor (holdings quoted verbatim above; links to the Cornell Legal Information Institute).
Dusky v. United States, 362 U.S. 402 (1960) — competency is a functional test (rational and factual understanding; ability to consult counsel), not a diagnosis. https://www.law.cornell.edu/supremecourt/text/362/402
Jackson v. Indiana, 406 U.S. 715 (1972) — no indefinite confinement on an incompetency label; the person must be restorable within a reasonable time or released. https://www.law.cornell.edu/supremecourt/text/406/715
O’Connor v. Donaldson, 422 U.S. 563 (1975) — mental illness alone cannot justify confinement; a nondangerous person capable of surviving with willing family may not be held. https://www.law.cornell.edu/supremecourt/text/422/563
Washington v. Harper, 494 U.S. 210 (1990) — forced antipsychotics require dangerousness and the patient’s own medical interest. https://www.law.cornell.edu/supremecourt/text/494/210
Riggins v. Nevada, 504 U.S. 127 (1992) — a pretrial detainee is entitled to at least as much protection as Harper’s convicted prisoner. https://www.law.cornell.edu/supremecourt/text/504/127
Foucha v. Louisiana, 504 U.S. 71 (1992) — continued confinement on a mental-illness rationale requires proof of current mental illness and dangerousness. https://www.law.cornell.edu/supremecourt/text/504/71
Sell v. United States, 539 U.S. 166 (2003) — forced medication to restore competency requires, among four findings, that it be “medically appropriate, i.e., in the patient’s best medical interest.” https://www.law.cornell.edu/supremecourt/text/539/166
Minnesota authority
Jarvis v. Levine, 418 N.W.2d 139 (Minn. 1988) — non-emergency involuntary neuroleptic administration requires prior judicial approval.
Statute
Minn. Stat. §§ 611.40–611.59 (competency proceedings); enacted Laws of Minnesota 2022, ch. 99; effective April 1, 2024, and applicable to competency determinations initiated on or after that date. Official text: https://www.revisor.mn.gov/statutes/cite/611.40
Former Minn. R. Crim. P. 20.01 (governing determinations initiated before April 1, 2024).
Record
Rule 20 Evaluation Report (Dr. Barbo), dated June 22, 2026; filed July 10, 2026 — the document analyzed throughout.
Prior Rule 20 Evaluation Reports: Rogstad (Mar. 10, 2023 — “These statutes do not exist in Minnesota at present”); Milz (Jan. 11, 2024); Cranbrook (Dec. 20, 2024).
Contested-competency hearing transcript, Mar. 5, 2025 (Tr. 9–11).
Pulling for you, Matt.