The retconning of George Floyd: a coda

A response to Coleman Hughes and Avak Howsepian

A memorial for Mario Gonzalez, who was killed by Alameda, California police officers in April 2021. (Photo by Justin Sullivan/Getty Images)

I really hope this is the last time I’ll need to write about The Fall of Minneapolis.

That of course is the name of the propagandistic “documentary” that, unfortunately, has been incorporated into right-wing narratives about the death of George Floyd.

Earlier this year, I wrote a long, three-part series debunking that film’s claims using a column by Coleman Hughes, published in Bari Weiss’s Free Press, as a jumping off point. I did so because Hughes’s column — and the clout he and Weiss have on the right — was getting some traction for the documentary’s conspiratorial claims with more mainstream audiences.

This post will address two responses to my series — Hughes’s, which was also published in the Free Press, and an email I received from Avak Howsepian, the author of a journal article that was credulously cited in the documentary.

I’ll start with Hughes.

---

Hughes’s response includes a defensive introduction by Weiss, which takes issue with both my criticism of his column and a separate piece I wrote for the Unpopulist that’s more broadly critical of “heterodox” thinkers — the cadre of academics and pundits Weiss once dubbed the “Intellectual Dark Web.” I don’t think there’s much in Weiss’s introduction that merits a response. The Free Press did not publish any corrections or clarifications to Hughes’s original column, which I think really only confirms my point that their vows to cover “the world as it actually is” and to pierce “ideological narratives” ring fairly hollow. They then published a response from Hughes that, as I’ll show here, is just as problematic as his original column.

This will not be a line-by-line response to Hughes. I don’t have the time or energy for that, and I’d imagine most of you don’t have the time or the interest to read one. Instead, I have a few general comments, then some analysis of one particularly baffling claim he makes — about what the word homicide means in the context of forensic pathology. Hughes’s entire posture toward Floyd’s death relies on this definition and, as far as I can tell, he appears to have made it up.

Wherever he got it, it’s utterly and completely wrong, yet Hughes asserts it with that familiar, matter-of-fact, just-common-sense confidence of a heterodox pundit. To support his sweeping claim, he then points to a single case he apparently found in his research.

Unfortunately, he appears to have stopped reading about the case once he decided it confirmed his position. As a result, he leaves out important developments in his own poster case — details that dramatically undermine his point.

First, the broad strokes. The overarching theme of Hughes’s response is that his column was not a brief for Derek Chauvin’s innocence, but merely an argument that there was reasonable doubt in the state’s case against him.

This would be fine if that’s what his column actually does — though I’d still disagree. But it is not what his column does. For one, the subhead calls Chauvin a “scapegoat.” But more importantly, Hughes’s original column doesn’t critique the case the prosecution actually presented at Chauvin’s trial. Instead, he critiques the caricature of the prosecution’s case depicted in The Fall of Minneapolis.

Here’s one quick example: Hughes pointed out in his column that “even one of the prosecution’s witnesses conceded” that Chauvin did not have his knee on Floyd's neck for as long as was claimed in media reports — that it was actually on Chauvin’s back for much of the time.

This is only relevant if you think the prosecution claimed that Floyd died because Chauvin was kneeling on his neck, obstructing his windpipe. And indeed, that’s how the documentary depicted the state’s case. It’s also a claim many pundits on the right amplified when the documentary came out.

But this isn’t what the state actually argued. Prosecutors argued at the trial that Chauvin’s weight, the handcuffs, and the asphalt underneath Floyd restricted his ability to expand his lungs. He couldn’t sufficiently inhale. This caused his oxygen levels to drop and the CO2 in his blood to soar, ultimately resulting in his heart and lungs shutting down. Hughes did not address this in his column.

There are other examples in my series. I won’t rehash those here.

In his response to me, Hughes does appear to have attempted to do some of the research he should have done before writing his column. But he tries to retrofit what he found to imply that, actually, he was aware of these things all along — that it was all baked into his column.

I don’t find this persuasive, mainly because it’s clear that he still doesn’t understand these issues.

For example, in his initial column Hughes built up Andrew Baker, the Hennepin County medical examiner who did the autopsy on George Floyd. He contrasted what he descried as Baker’s careful, thorough autopsy with the less thorough analysis done by the medical examiners hired by Floyd’s family. He does this en route to suggesting that Baker was skeptical of the criminal charges against Chauvin, and that Baker’s skepticism was warranted by the evidence.

I pointed out that while Baker’s actions in this case were at times ambiguous, he still classified Floyd’s death as a homicide, and he still testified for the prosecution at Chauvin’s trial. Both of these details contradict the way Hughes depicts Baker in his column, so it’s notable that Hughes didn’t mention either of them.

In his response, Hughes writes:

This argument is a bit rich. From a strategic point of view, the prosecution had no choice but to call Baker as a witness—whether or not his testimony was helpful or harmful to their case. For the state to cede him to the defense would be to create the devastating impression that their own forensic pathologist (and the person who did the official autopsy) agreed with the defense’s case. So strategically, they had to call him no matter what. As for Baker, he would have had no choice in the matter once he was subpoenaed to testify.

This is a misunderstanding of how death investigations work. Baker was called to the stand because he determined Floyd’s death was a homicide, not the other way around. If he had determined it was not a homicide, prosecutors would have had a much more difficult time charging Chauvin. At best, they might have tried to ask another medical examiner for a second opinion. If the second medical examiner disagreed with Baker and determined Floyd’s death was a homicide, they could then call that medical examiner to testify. But they would also be required to notify the defense about Baker’s contradictory determination.

Does this scenario happen often? No. Does it happen? Certainly. In our book about the death investigation system in Mississippi, Tucker Carrington and I found that it was actually common in that state. When prosecutors disliked the conclusions of the state medical examiner, they took the bodies to private medical examiners who were more likely to tell them what they wanted to hear.

My point is that if Baker didn’t really believe Floyd’s death was a homicide, he was ethically obligated to say so. To suggest that he didn’t really think this was a homicide, but testified that it was anyway, is to suggest that Baker perjured himself and violated his professional ethics. (Indeed, some on the right have explicitly accused him of precisely that.)

---

---

This brings me to the portion of Hughes’s response that I think is worth addressing in detail, because it really highlights the chasm between his limited understanding of the death investigation system and the confidence with which he opines on it.

Hughes tries to explain away Baker’s testimony that Floyd’s death was a homicide by arguing that in the world of forensic pathology, homicide doesn’t really mean what we all think it means. He writes:

“For medical examiners, homicide means that other people were involved in a very broad sense. It does not mean that the medical examiner believes that the act clears the legal bar for homicide.”

I have no idea where he got this. He provides no link or documentation. He doesn’t attribute it to any authority. Perhaps this is one of those common sense conclusions that heterodox thinkers just know based on their intuition, learnedness, and unique ability to penetrate media bias.

But Hughes is wrong about this — demonstrably and unquestionably wrong.

Here’s the definition of homicide from the guidelines published by the National Association of Medical Examiners (NAME), the field’s largest professional organization:

Homicide occurs when death results from a volitional act committed by another person to cause fear, harm, or death.

Volitional of course means “of one’s own will or choosing,” or “under conscious control.” It does not mean “involved in a very broad sense.”

This is the same definition used by the CDC’s handbook on death classifications for public health records.

NAME also states in it guidelines that “deaths due to positional restraint induced by law enforcement personnel or to choke holds or other measures to subdue may be classified as Homicide.” At the same time, the guidelines also say that overdose or drug toxicity deaths should be classified as accidents.

In other words, if a death was caused by drugs, that death is not a homicide (unless someone was drugged without their knowledge). It’s an accidental death.

Here’s the Snohomish County, Washington, medical examiner’s office:

Homicide is defined as the action of one person directly causing the death of another.

The Georgia Bureau of Investigation, which oversees the state medical examiner, defines homicide as “[a] death caused by the actions of another person.”

Here’s the Harris County, Texas, Institute of Forensic Sciences: “Homicide is when a person is killed by one or more persons.”

The Office of the New Jersey State Medical Examiner says that a homicide is “the manner of death used when there is a violent death from an intentional act of another individual, whether or not the individual responsible is prosecuted.”

Most medical dictionaries use a similar definition, as do forensic pathology textbooks.

I don’t know of a single medical examiner’s office, professional organization, or other official or quasi-official forensic body that defines homicide as anything other than a death caused by someone else. This notion that it can also mean that other people were merely loosely associated with a death seems to exist only in Hughes’s mind.

It is true that a homicide determination doesn’t always mean that someone will be charged with a crime. Prosecutors might decide that the person or people responsible for a death acted in self defense or in the defense of others. This is a “justifiable homicide.” Prosecutors might also decide that it just isn’t in the interest of justice to bring a charge. But all these are still homicides.

For forensic pathology purposes, a “homicide” is always a death that would not have happened “but for” the willful action (in jurisdictions in which there’s a duty to care, it might also be the willful inaction) of someone else. It does not and has never meant that other people were “involved in a broad sense.”

In support this erroneous contention, Hughes then points to a specific case — the 2021 death of Mario Gonzalez.

He writes:

Consider, for instance, the example of Mario Gonzalez—whose death in Alameda, California, in 2021, mirrored Floyd’s in many ways. In Gonzalez’s autopsy report, the forensic examiner listed the main cause of death as “toxic effects of methamphetamine.” (The contributing causes of death were “physiologic stress of altercation and restraint; morbid obesity; alcoholism.”) But despite the fact that drugs were the main cause of death, the manner of death was still listed as a “homicide,” simply because police were involved in the broadest sense: they were holding him in the prone position at the moment he died from meth intoxication. 

I followed the Gonzalez case when it happened, so I’m fairly familiar with it. I’m astounded that Hughes would invoke it in this context.

First, a little background:

California, like several other states, still relies on an anachronistic coroner system to oversee death investigations. In these systems, an elected county coroner has ultimate say over manner of death. As you might imagine, entrusting death investigations to a low-ranking, low-paying office that in most places requires little training and few qualifications has created a lot of problems.

But the situation in California is even worse. In 48 of the state’s 58 counties, the elected county sheriff also serves as coroner. Certified forensic pathologists still perform the autopsies in these counties, but they report to the sheriff and, ultimately, the sheriff has the last word about manner of death.

At risk of stating the obvious, entrusting the county sheriff to determine the manner of death of people who die in police custody is a big honking conflict of interest. It’s cognitive bias on steroids. It would be difficult to design a worse system from scratch.

Just last year, an academic study found ample evidence of bias in these counties when it comes to classifying in-custody deaths.

“Our research found that there is a significant discrepancy between the actual number of officer-involved homicides and the number reported by sheriff-coroners in California,” Finch explained. “This underreporting is a major issue that needs to be addressed, as it undermines the public’s trust in the justice system and hinders efforts to hold law enforcement accountable.”

One case in 2017, in San Joaquin County, demonstrated the lengths to which one sheriff would go to protect his law enforcement officers. In this incident, the sheriff-coroner was accused of changing the cause of death from “homicide” to “accident.” The doctors and technicians, including the chief forensic pathologist, Bennet Omalu, resigned in protest, leading the county Board of Supervisors to strip the sheriff of coroner duties.

Testifying before the state Senate Governance and Finance Committee, Omalu said, “California has the most backward system in death investigation, is the most backward in forensic science and in forensic medicine.”

In another example in 2020, a Contra Costa County sheriff-coroner’s office deemed Angelo Quinto’s death a result of “excited delirium,” a term sometimes attributed to sudden unexplained deaths of individuals while in police custody.

According to The Mercury News, Quinto lost consciousness after an officer knelt on his neck for nearly five minutes. He later died in police custody. No charges were filed against any officers involved in the incident; however, it did lead to Angelo’s Law, which bans restraint tactics and facedown holds that could cause asphyxiation in California.

Alameda County, where Mario Gonzalez was killed, is one of these 48 sheriff-coroner counties. (Note that the coroner’s website is nested within the website of the sheriff’s office.)

In a deposition for a lawsuit brought by Mario Gonzalez’s family, Vivian Snyder, the forensic pathologist who performed the autopsy, said that three officials from the Alameda County Sheriff’s Office and the chief of the Alameda Police Department (whose officers killed Gonzalez) were present at the procedure.

This is not uncommon. But it probably should be, particularly for in-custody deaths. A 2022 study found that a majority of the jail deaths in Los Angeles County over a 10 year period were likely homicides that had been wrongly classified as “natural causes.” The same study found that sheriff’s department officials were present at 86 percent of the autopsies. (L.A. County is at least one of the 10 in which the sheriff does not also serve as coroner. Even so, there’s no reason for someone from the department to be present at these autopsies.)

Again, this seems obvious, but it’s a huge red flag for the heads of very police agencies whose officers are under scrutiny to sit in on the autopsy. It’s even more of a red flag if the person performing that autopsy also reports to one of those agencies.

With that context in mind, let’s now look at the death of Mario Gonzalez.

In 2021, a nearby resident called 911 to report a man in a public park who seemed disoriented and confused. The caller said his wife was afraid, but also stressed that the man had done nothing wrong. An Alameda police officer responded, and found Gonzalez, who seemed to him to be dazed and unresponsive. The officer called for backup.

At no point in his interaction with police did Gonzalez commit a crime. There was a Walgreens shopping basket filled with empty liquor bottles near Gonzalez, so police at first suspected that he may have stolen them. But an officer checked with the store. Amanager told them there had been no recent thefts of liquor.

Gonzalez never threatened the officers, and according to the officers’ own testimony, he never struck them.

At worst, the officers might have reasonably concluded that Gonzalez was having a mental health crisis, and thus posed a threat to himself. As it turns out, Alameda County has a Mobile Crisis Response Team for scenarios just like this one.

Yet the officers never called the response team. Instead they tackled Gonzalez. They took him to the ground and handcuffed him. Then, as he lay prone and handcuffed over the next several minutes, at least two and at times three people put their weight on his back.

I use “people” and not “police officers” here because there’s another, especially jaw-dropping detail in this story: At the time of the 911 call, a civilian happened to be accompanying one of the officers for a ride-along. Incredibly, after tackling Gonzalez that officer then instructed the civilian, who weighed about 250 pounds, to lay across Gonzalez’s legs.

Here’s body cam video of the incident.

(Warning: Mario Gonzalez dies in this video. Watch at your discretion.)

As I pointed out in my series, the prone and handcuffed position has long been known to be dangerous and potentially life-threatening — ever more so when police also put weight on the suspect’s back. Groups like the NYPD, police training organizations, and the International Association of Chiefs of Police have warned about the dangers of the prone and handcuffed position for years, especially with people who are obese, experiencing a mental health crisis, or under the influence of drugs or alcohol. There was good reason to think Gonzalez was hampered by all three.

One officer claimed in his report that he and his colleagues continued to put weight on Gonzalez’s back because while he was on the ground Gonzalez “violently” resisted them. But as a law enforcement trainer argued in an affidavit for the family, that claim doesn’t appear to be backed up by body camera video. Moreover, officers are taught that when someone puts up resistance while in a handcuffed and prone position, it’s often not because the suspect presents a threat — it’s because they’re struggling to breathe.

Moreover, the Alameda PD was already on notice about the risks of putting weight on someone in this position. In 2018, three years before Gonzalez’s death, Alameda officers confronted Shelby Gattenby, a veteran with PTSD who was in the midst of a mental health crisis. The officers tased Gattenby and put their weight on his back while he was handcuffed and on his stomach. He died. In June 2020, the city paid a $250,000 settlement to his family.

Hughes is correct that Snyder concluded both that methamphetamine toxicity was the primary cause of his death and that his death was determined to be a homicide. But Hughes is wrong about what this means.

He writes:

Whether you agree with this explanation of Gonzalez’s death is beside the point. The important point is that it is coherent for a medical examiner to list the manner of death as “homicide” while also listing drugs as the main cause of death—so long as other human beings were “involved” in a very broad sense. That is how broad the concept of “homicide” is in the lexicon of forensic pathology. So to infer anything of direct relevance to a murder charge from the “homicide” determination is a fallacy. 

First, again, despite Hughes’s supercilious tone, there is no “lexicon of forensic pathology” that agrees with his bizarre definition of homicide, and I don’t know of a single forensic resource that does.

Second, it’s fairly unusual for an office to classify a death as a homicide while also citing drug intoxication as the primary cause of death. Again, drug-related deaths are typically classified as accidental.

So even if Hughes were correct that the homicide designation means that Snyder and the sheriff’s office for some reason used his bizarre definition of homicide instead of the one used by everyone else in the field, it would still be inappropriate to draw such a sweeping conclusion about “the lexicon of forensic pathology” from one opinion in one case in a county where manner of death is actually ultimately determined by the sheriff.

But it turns out that Hughes is also mistaken about what happened in this particular case.

Snyder did not write in her autopsy report that the officers were only involved in Gonzalez’s death in a “very broad sense.” She wrote that their actions were “a significant condition in causing death.” In her deposition testimony, Snyder explained that meth intoxication, Gonzalez’s health, and the officers’ actions caused his death. In other words, she testified that if Gonzalez hadn’t ingested meth, he would likely still be alive today. But he’d also likely still be alive if the police officers hadn’t engaged in a physical altercation and applied the restraint they did. Here too, she never suggests that the officers were involved only “in a very broad sense.”

In depositions for the lawsuit filed by Gonzalez’s family, the opposing experts hired by each side did disagree — often contentiously — about whether Gonzalez had enough meth in his system to be fatal. Overdose deaths from meth are rare, and Gonzalez had .907 ng/liter of the drug in his system — well within the parameters of recreational use. (Some reports note that for some toxicology tables, this level falls into the category of “toxic,” but “toxic” in this context doesn’t not mean fatal. It only indicates illegal use of the drug.)

Defense experts for the police officers argued that Gonzalez’s other health ailments likely exacerbated the effects of the meth, while experts for Gonzalez’s family disputed whether he actually suffered from some of those ailments. They also disputed that those he did have played any role in his death.

Consequently, there was also disagreement over whether Gonzalez died solely from positional asphyxiation caused by the hold police put on him, or from the combination of his struggle with police, the meth in his system, alcohol withdrawal, and a possible heart condition.

Here’s the important part: There wasn’t a single expert for either side who argued what Hughes is arguing. No one claimed that the “homicide” classification could be interpreted to mean that the police were only loosely or tangentially involved in his death.

A couple defense experts did think that the police did not cause Gonzalez’s death. But because of that, they also thought Gonzalez’s death should never have been classified as a homicide in the first place.

This is the only case Hughes cites to ascribe his own, odd definition of homicide to the entire field of forensic pathology. Yet in their sworn depositions, none of the expert witnesses — for the plaintiff or the defense, forensic pathologists or otherwise — agree with him.

Ultimately, whether the officers should be held criminally or civilly liable for Gonzalez’s death has nothing to do with how you define homicide (mostly because that isn’t really up for debate). It turns on whether the officers’ actions were justifiable — on whether or not this was a justifiable homicide.

To make that determination, we need to answer questions like whether the officers had reasonable suspicion to detain Gonzalez, had probable cause to arrest him, if the tactics they used were in violation of Alameda PD policy, and if those tactics are generally accepted in policing.

The officers do not appear to have had probable cause for arrest. They’d later claim they arrested Gonzalez for “public intoxication,” but his blood-alcohol level at most was below .02 (the minimum level the test can detect). For comparison, the legal limit to drive is .08. The police did not give Gonzalez a breath or sobriety test at the scene.

The officers also suggested that the basket of liquor bottles near Gonzalez was a violation of state open container laws. But that’s a cite-and-release crime in California. It is not cause for a violent arrest.

The experts for Gonzalez’s strongly disagreed with the conclusion that meth intoxication contributed to his death. But even if Snyder was correct on this point, we’re back at eggshell theory. If I rob a bank and a bystander with severe heart disease panics and drops dead from a heart attack, that bystander’s death would be classified as a homicide, and I’d almost certainly be charged with a separate crime for causing that death. If, alternatively, I’m merely doing my banking and accidentally drop my cell phone, which makes a loud noise and causes the same customer to have a heart attack and die, that death would not be a homicide, and I wouldn’t be charged for it.

In evaluating Gonzalez’s death, then, we know that police are trained to not put weight on prone, handcuffed people. We know they’re told that doing so is especially dangerous for obese people, people showing signs of mental health distress, and people showing sings of alcohol or drug impairment. We know that these dangers were widely and publicly amplified after the death of George Floyd, and that Alameda police in particular were already on notice after the death of Shelby Gattenby. We also know that they lacked probable cause to make an arrest.

Despite all of this, in 2022, Alameda County DA Nancy O’Malley announced that she would not be charging the officers responsible for Gonzalez’s death. Importantly, even O’Malley did not use Hughes’s expansive definition of homicide in reaching her decision. She defined homicide as “the killing of one human being by another,” and acknowledged that for the purpose of manner of death analysis, “the officers’ actions during the altercation and restraint were a substantial factor in Mr. Gonzalez’s death.”

O’Malley cleared the officers because she concluded that when they confronted Gonzalez, tackled him, handcuffed, him, and held him under their weight for several minutes, they were acting within the scope of their training and department policy. She also concluded — and here I strongly disagree — that Gonzalez’s death was not a “direct, natural, and probable” consequence of the officers’ actions. In other words, while the officers’ actions were indeed a direct contributor to his death, his death was not a foreseeable and probable consequence of the restraint they used.

As we’ve already discussed, I think there’s ample evidence to contradict her conclusion, here. Cops are trained on this stuff.

O’Malley’s decision not to charge these officers is why Hughes thinks this case supports his position. And he’s right that there are a lot of similarities with the death of George Floyd. And in this case, at least from the links he provides, it appears that despite the homicide classification, state officials decided that these officers were not criminally liable for Gonzalez’s death.

But all the coverage of the case that Hughes links to in his response is from 2021. He probably should have kept reading. Because theres’s quite a bit more to the story.

O’Malley left the DA’s office in January 2023. She was replaced by Pamela Price, who ran as a reformist DA, and specifically made police accountability a key part of her campaign.

That message clearly resonated with voters, and with good reason. Alameda County has a long history of both police misconduct, and aggressive, rules-bending prosecutors. Alameda County also regularly leads Bay Area counties in annual payouts to victims of police violence and their families.

The city of Alameda itself also has a reputation. It’s effectively an island adjacent to Oakland. Alameda the city has about half the black and Latino population of Alameda the county, and the city has an ugly history of racial covenants, zoning laws to restrict multifamily housing, and racism in its police department. There’s a sentiment in the county that Alameda, the town, wants to keep Black and Latino people out.

Despite Alameda County’s history of police abuse, in her 14 years as DA, O’Malley only brought criminal charges against police officers for an in-custody death a single time. In 2018, she came under fire for accepting a $10,000 campaign contribution from the Fremont police union while she was investigating the fatal of shooting of a pregnant teenager by three Fremont officers — one of whom was president of the union. O’Malley cleared those officers of criminal wrongdoing.

Pamela Price reopened the Gonzalez case shortly after taking office in 2023, and in April of this year she filed involuntary manslaughter charges against the officers responsible for his death.

This is where the criminal case stands today.

Those criminal charges would not have been possible if Gonzalez’s death hadn’t been classified as a homicide. They also wouldn’t have been possible if, despite that classification, the Snyder had told prosecutors that the officers were only “involved in a very broad sense.”

But there’s more. In December 2023, the city of Alameda settled with Gonzalez’s family for $11.35 million. It’s the largest pre-trial settlement for an in-custody police death in California history, and one of the largest police abuse settlements in the U.S. overall.

Alameda has a population of 76,000. The $11 million the city agreed to pay represents 4 percent of the city’s annual budget. Just for perspective, if nearby Oakland payed a settlement equal to 4 percent of its budget, it would come to about $880 million.

Cities do sometimes settle lawsuits they might otherwise win if they conclude that doing so would be less expensive than litigating. But cities do not pay out $11 million unless their attorneys are pretty certain they’re going to lose. They only pay out $11 million settlements if they think jury will likely conclude the city’s officers were clearly culpable — and not that they were merely “involved in a very broad sense.”

The Gonzalez settlement came six months before Hughes published his response. It was widely reported in outlets across the country. The indictments of the officers who killed Gonzalez came two months before Hughes published his response. Those too were widely reported.

Hughes doesn’t mention either the charges or the settlement in his response. As with the important details Hughes left out of his original column, I won’t claim to know if he left these details out to intentionally mislead his readers or if he just stopped reading once he thought he found a case that supported his position.

But they’re pretty important details!

Mistakes happen. I’ve certainly made them. The important thing is to acknowledge and correct them when they’re brought to your attention. When you refuse, readers have good reason to doubt your credibility.

Hughes, Weiss, and the Free Press refused to acknowledge the errors in Hughes’s original column. Here’s hoping they’ll acknowledge the errors in this one. But I’m not holding my breath.

---

On to the other response.

After my series concluded, I received an email from Avak Howsepian, the psychiatrist who wrote a journal article arguing that George Floyd may have suffered from a condition called acromegaly, and that this may have been the cause of his death.

As I wrote back then, the Fall of Minneapolis documentary cites Howsepian’s article to support its narrative that there was something corrupt and conspiratorial about the autopsy.

“As the camera zooms in on the article’s title, narrator Liz Collin says it was published in a ‘peer-reviewed journal,’ and that it identifies ‘17 errors’ in Baker’s autopsy report.”

As I wrote at the time, the “peer-reviewed journal” here is obscure — no one I spoke to had heard of it. The medical examiners I consulted for the series were also skeptical of Howsepian’s theory.

I also pointed out that this article appears to be the only time Howsepian has published about forensic pathology, and that instead, he frequently published on culture war issues like homosexuality, conversion therapy, and gender dysphoria in right-of-center publications.

So here’s Howsepian’s email, with my responses.

(1) I do not say that Mr. Floyd suffered from acromegaly, only that he possibly did, as reflected in the very title of my article, ‘Possible Acromegaly Complicating the Death of George Floyd.’

Fair enough! I will amend this in the original piece, along with a link to this post.

(2) Much of what I write about is about errors in Dr. Baker's autopsy. One might make errors in an autopsy in at least two ways. First, one might not do what one ought to do in any forensic autopsy that involves examination of the brain. Second, one might not do what one ought to do in a forensic autopsy when a certain condition is possibly present, because one does not suspect the condition in question. There are both kinds of errors in Dr. Baker’s autopsy of Mr. Floyd. I will point out only two examples of each. (1a) There was no documented examination, either gross or microscopic, of Mr. Floyd’s pituitary gland — a gland that ought to be examined in every forensic autopsy that involves examination of the brain. (1b) There was no documented examination of Mr. Floyd’s dentition, something which should be noted in every forensic autopsy. Now, it seems apparent that Dr. Baker, a pathologist, did not suspect acromegaly — but he should have suspected it. After all, a psychiatrist from Fresno, California suspected it without having conducted an autopsy on Mr. Floyd, simply by way of what is known about his history as found in the media, his pictures (and family members’ pictures) in the media, and other clues from Dr. Baker’s autopsy. Were Dr. Baker to have suspected acromegaly, (2a) he clearly would have biopsied the patch of hyperpigmentation on Mr. Floyd’s skin, and (2b) he clearly would not have called Mr. Floyd’s extraadrenal paraganglioma ‘incidental’ and left it at that. Rather, Dr. Baker would have further investigated, in multiple ways, the role of this tumor in Mr. Floyd’s death.

The medical examiners I spoke with were of mixed opinion about the thoroughness of Baker’s autopsy. One, who consulted for the prosecution, told me a number of histology slides were missing. But others said Baker’s autopsy was pretty standard and has no major deficiency. Only the autopsy report itself is publicly available, not the slides and tissue samples. So I can’t confirm or refute Howsepian’s suggestion that Baker did not examine Floyd’s pituitary gland or teeth.

Baker does briefly mention Floyd’s teeth in his report — he writes that they are “native and in good repair.” Perhaps this is why he didn’t subject them to further analysis.

But more generally, there are thousands of different conditions for which a medical examiner could do further testing after an autopsy. You can’t test for all of them. Medical examiners make decisions about additional testing based on signs and symptoms they find during the autopsy.

The documentary highlights what Howsepian calls 17 “deficiencies” in Baker’s autopsy. It gives the implication that these are standard procedures that should be done in all autopsies. But most of these were procedures that a pathologist would only do if they suspected the condition Howsepian suspects. No one I spoke to said this is a condition Baker should have suspected. Therefore, they did not consider the fact that he didn’t do those tests to be erroneous.

(3) Referring to my article as ‘bizarre’ is uncharitable and misleading in the extreme. What is so ‘bizarre’ about it?

My characterization of Howsepian’s article is an opinion based on my discussions with other medical professionals.

(4) As alluded to in (2) above, not all acromegaly is due to pituitary tumors. Your stating, therefore, that acromegaly (p. 34) is a condition “in which a tumor on the pituitary gland causes overproduction of human grown hormone” is false, since acromegaly could also be due to a tumor that is not of pituitary origin (as noted above, might have been the case with Mr. Floyd). Furthermore, the tumor to which you allude is not ‘on’ the pituitary (as you note), but ‘of' certain pituitary gland cells.

Acromegaly itself is rare — it affects around 6 people per 100,000. Acromalegy caused by non-pituitary tumors is extremely rare, even for acromegaly — it occurs in less than 1-2 percent of overall acromegaly cases.

So Howsepian speculated that Floyd (may have) suffered from an extremely rare form of an extremely rare disease, that the effects of said disease kicked in just as Floyd was being pinned by Chauvin, that Baker’s failure to look for signs of the disease were “deficiencies,” and that all of this escaped the notice of the myriad other certified medical examiners who reviewed Baker’s autopsy of Floyd.

I suppose all of that is possible. But it seems unlikely. And it’s misleading at best for the documentary to highlight Howsepian’s list of Baker’s alleged errors to support its narrative that the autopsy was so deficient as to be suspicious, and perhaps part of an FBI-led coverup.

Also, here again eggshell theory makes this discussion irrelevant. If Floyd would not have died but for the officers’ actions, then his death is still a homicide. Unless you’re going to argue that Floyd’s health conditions would have caused him to die that day even if he hadn’t been pinned under Chauvin’s knee — that the timing was mere coincidence — the acromalegy claim would not exonerate the officers for his death.

Howsepian’s point in the final sentence of the passage above seems rather pedantic to me. But feel free to judge for yourself.

Moving on:

(5) The medical examiners with whom you spoke are correct regarding acromegaly’s being irrelevant to whether Mr. Floyd’s death was a homicide (as you note on p. 34). However, that itself is irrelevant. Not all homicide types are treated alike under the law (e.g., murder and manslaughter are both degreed offenses, and one might even commit a homicide that is wholly legal in the case of a homicide due to self-defense). Even more to the point, sentencing by the Court for whichever type of illegal homicide of which one has been convicted is supposed to take into consideration multiple factors, including the health status of victims insofar as more medically compromised victims might die as a result of significantly less violent actions on the part of defendants than one who is healthy. Furthermore, such medical compromise might even bear on the specific type of homicide with which one might be charged or for which one might be convicted. If someone, for example, gets in a pushing match with someone who dies from being pushed down because that individual is medically compromised, the defendant in that homicide might not be charged with as serious of a crime as one who knifed or shot the victim in question.

I’m not persuaded by the hypothetical. Unless someone is standing on the edge of a cliff or on the precipice of a crocodile-filed moat, death is not a foreseeable consequence of pushing. Death is a quite foreseeable consequences of putting an obese man who is experiencing a mental health crisis in a handcuffed, prone position with weight on his back. Martin Tobin, the pulmonologist prosecutors put on the stand, testified that even a healthy person would have died under the conditions to which Floyd was subjected.

I don’t disagree with much of the rest of this particular passage. But none of it means Floyd suffered from acromegaly. It also doesn’t mean that if he did, the officers were wrongly charged. And it doesn’t mean that Baker’s failure to test for the condition was part of a coverup. And that is context in which the documentary cited Howsepian’s paper.

(6) Pointing out that I am a psychiatrist rather than a forensic pathologist — and the way this is pointed out — seems to me to be an ad hominem diversion. I am, after all, a medical doctor, and earlier in my career (prior to medical school) I assisted in hundreds of forensic autopsies. (My current psychiatric practice is wholly a consulting practice in which I act as a forensic psychiatric consultant, hence, my familiarity — more than most psychiatrists — with the law, and my ongoing interaction with discovery materials involving autopsies in those many homicide cases on which I have worked as a psychiatric expert.) Still, none of this matters so very much. What matters is whether my criticisms and inferences in my paper are, in fact, thoughtful and well-grounded. You, yourself, stated that you are not in a position to evaluate the merit of my theory (p. 34). What I was hoping to do by publishing this paper was to begin an intelligent discussion with experts in the field — forensic pathologists, cardiologists, attorneys, and endocrinologists — regarding whether Mr. Floyd suffered from acromegaly and, if so, whether that might have contributed to his death, and, if so, whether that might have had any legal implications.

I’d argue that as a layperson trying to assess the validity of an article in an academic journal, it’s entirely relevant to consider the author’s qualifications to write about the topic. In this case, the author was venturing outside his discipline to offer opinions on a hot-button case. Meanwhile, in his own field, he published other work advancing positions that others in that field find dubious.

These aren’t the only factors to consider. But they’re certainly relevant.

(7) Your referring to one of my papers as “linking trans acceptance to pedophilia” is unfortunate.  I doubt whether you read the paper to which you are referring. In that paper, I compare and contrast the American Psychiatric Association’s depathologizing of pedophilia with its depathologizing of gender dysphoria. I say nothing about “trans acceptance.” You would know this had you read my paper, which I invite you to do.

I would argue that “trans acceptance” and the “depathologizing of gender dysphoria” go hand in hand. Or to put it another way, pathologizing gender dysphoria goes hand in hand with discrimination against trans people.

(8) Very recklessly, you also say that I have “devoted quite a bit of time to defending conversion therapy — the idea that gay people cane be ‘cured’ of their sexuality” (p. 34). Please cite even one quote from me in the professional literature in which I do what you say. You will not find any such quote. My sense is that you have either gotten this information from a bad source or you have simply illicitly inferred that because I am a Christian who considers 'sodomy to completion' to be unethical, and because I have published articles on homosexuality and belong to an organization which defends therapeutic choice, the 'Alliance for Scientific Integrity and Therapeutic Choice’ (hereafter, ’the Alliance’), that I have "devoted quite a bit of time to defending conversion therapy.” 

I’m fascinated by the curiously-worded phrase “I am a Christian who finds sodomy to completion unethical.” I have questions! But not now.

Here’s a select list of other articles Howsepian has authored:

• — “Psychiatry's Dysphoric Turn: Psychophysical Dysmorphia, Transgender Euphoria, and the Rise of Pedophilia”

• — “Treating Homosexuality: A Response to Yarhouse”

• — “Sexual Modification Therapies: Ethical Controversies, Philosophical Disputes, and Theological Reflections”

• — “The American Psychiatric Association's Guidelines Regarding Possible Conflict between Psychiatrists' Religious Commitments and Psychiatric Practice: A Conceptual Critique”

• — “The family physician and adolescent homosexuality”

• — “Military position on homosexuality”

• — “Fetophilia” (In this article, Howsepian argues that the “widely held liberal principles concerning abortion and sexuality” allow for “the moral permissibility of sexual contact with live human fetuses.”)

As for “the Alliance,” journalist Madison Pauly recently wrote in Mother Jones about attending one of that group’s conferences.

The day before the conference, I received a packet of materials from Alliance board member Keith Vennum, a psychiatrist who specializes in “helping men develop their heterosexual potential,” according to his profile on Focus on the Family’s Christian Counselors Network.

Dr. Howsepian also makes an appearance!

They included an article by a gender care specialist who turned against youth medical transition, reading suggestions from Cretella on how to “heal” “transgender belief” in children, and an essay by Fresno psychiatrist Avak Howsepian arguing that supporting “diversity and inclusion” means supporting pedophilia. I packed my bag and flew to Las Vegas.

Pauly also points out that the “Alliance” isn’t fond of the term “conversion therapy,” likely because most of the public finds the idea repugnant. Instead, they now use terms like “therapeutic freedom,” or “treating unwanted same-sex attraction.”

Here’s the FAQ from the group’s website. Feel free to look it over and make your own decision about whether they’re pushing conversion therapy by another name.

Back to Howsepian.

(9) You also say that I serve “on boards of groups who promote it [conversion therapy]” — plural. I do serve on the Research Committee of the Alliance. If that is what you mean by a ‘board’ then that much is true (however, ‘boards' typically are part of a more general organizational structure than the ‘Committee' on which I serve). And there is only one such organization in which I am involved, not more than one. 

In addition to “the Alliance” research committee, Howsepian is affiliated with the American College of Pediatricians, which you should definitely not confuse with the more reputable American Academy of Pediatrics. The former promotes conversion therap . . . er . . . “thereputic freedom.” The latter does not. And the former was founded in 2002 explicitly in protest of the latter, more reputable, similarly-named group’s support for gay and lesbian adoption.

However, at least as of 2018, Howspeian was merely a “consultant” for this group, not a board member.

Howsepian is also on the editorial board of the Journal of Human Sexuality, a publication that, it turns out, isn’t nearly as fun as its title suggests.

According to its website, the journal “particularly seeks articles relevant to the understanding and care of persons who experience unwanted same-sex attractions and/or gender dysphoria and the clinicians who provide this care. Reasoned perspectives and theoretical approaches that may be unwelcome for political and ideological reasons within mainstream psychology are encouraged.”

The lead article for the journal at the moment is called, “Debunking a Fallacy: New Study Shows Therapy for Undesired Same-sex Attraction Can Be Effective, Beneficial, and Not Harmful.

Sure seems like this is another group promoting conversion therapy! However, the journal is actually published by the Alliance for Scientific Integrity and Therapeutic Choice — "the Alliance” — or the same group for which Howsepian serves on a research committee.

So Howsepsian is a consultant for a group that promotes conversion therapy, serves on a research committee of another, and is on the editorial board of a journal that takes a similar position. But the journal is published by one of the other two groups.

For these reasons, Howsepian is correct. It would be accurate to say he is on the “board” of a group that promotes conversion therapy, but I used the plural “boards” and “groups.” I regret the error.

---