Verdict Problems in Karen Read Trial: Jeopardy Bar?

If you are not a true crime aficionado, or if you don’t live in the Northeastern US, you may not have heard about Karen Read’s murder trial. Last year, she was charged with the murder of her boyfriend, Boston police officer John O’Keefe, whose body was found in the snow near their house after a night in which the two reportedly had a fight at a bar. Karen’s attorneys presented an alternative theory of the crime. According to the defense, a local law enforcement family was involved in John’s death, and the officer placed in charge of the investigation hid evidence implicating them.

A defense strategy that presents an alternative narrative can be very risky, in that it can confuse the jury into comparing stories and choosing the more likely one (rather than deciding whether the prosecution’s story was proven beyond reasonable doubt). But in this case, the alternative theory was carefully pieced together with texts, Google searches, crime scene evidence, and geographic proximity–enough so to provoke considerable debate during jury deliberations. The trial ended in a hung jury and the parties are gearing up for a retrial with a notable difference: one of the jury alternates, who is an attorney, has joined the defense team.

In addition, there are some legal issues of interest: Read’s team is appealing a pretrial decision not to dismiss certain charges to SCOTUS. The story, in a nutshell, is this:

At last year’s trial, the judge declared a mistrial after the jury sent notes repeatedly expressing it was at an impasse. But Read’s attorneys say they later  heard from multiple jurors who indicated they did reach a unanimous decision to acquit Read of her murder and leaving the scene charges. 

Read contends the Constitution’s Double Jeopardy Clause bars her from facing trial on the charges again, given the revelations. 

“Read’s Petition contends that the jury’s decision that she is not guilty of Counts 1 and 3 constitutes an acquittal and precludes re-prosecution on those counts even if unannounced,” her attorneys wrote. 

Let’s figure out what’s what. Here is the original indictment, according to which the three charges are as follows:

M.G.L. c. 265, Section 1. Murder committed with deliberately premeditated malice aforethought, or with extreme atrocity or cruelty, or in the commission or attempted commission of a crime punishable with death or imprisonment for life, is murder in the first degree. Murder which does not appear to be in the first degree is murder in the second degree. Petit treason shall be prosecuted and punished as murder. The degree of murder shall be found by the jury. Second degree murder in Massachusetts requires intent to kill but not premeditation.

M.G.L. c. 265, Section 131/2. Whoever commits manslaughter while operating a motor vehicle in violation of paragraph (a) of subdivision (1) of section 24 of chapter 90 or section 8A of chapter 90B, shall be punished by imprisonment in the state prison for not less than 5 years and not more than 20 years, and by a fine of not more than $25,000. . . Under Massachusetts law, operating a motor vehicle while intoxicated, exhibiting recklessness toward the potential of causing a death, and actually causing a death, constitutes manslaughter.

M.G.L. c. 90 (2) (a 1/2) (2). (2) Whoever operates a motor vehicle upon any way or in any place to which the public has a right of access or upon any way or in any place to which members of the public shall have access as invitees or licensees and without stopping and making known his name, residence and the registration number of his motor vehicle, goes away to avoid prosecution or evade apprehension after knowingly colliding with or otherwise causing injury to any person shall, if the injuries result in the death of a person, be punished by imprisonment in the state prison for not less than two and one-half years nor more than ten years and by a fine of not less than one thousand dollars nor more than five thousand dollars or by imprisonment in a jail or house of correction for not less than one year nor more than two and one-half years and by a fine of not less than one thousand dollars nor more than five thousand dollars. . . Note that this offense, under Massachusetts law, does require the person to have been the cause of the lethal collision or injury, but the actus reus is the departure from the scene.

Of course, all three offenses revolve around the same scenario, but Counts 1 and 2 are the “same offense” for Blockburger purposes – which is to say, the elements of Count 2 are completely included in Count 1. It would be impossible, therefore, to convict Read of both murder and manslaughter. If convicted of murder, she’d have to be acquitted of manslaughter, and vice versa. Determining which is appropriate depends on what the jury finds regarding the mens rea, the mental state of the defendant: for murder, MA law requires intent to kill, and for manslaughter, recklessness at minimum.

Count 3 could come hand in hand with either Count 1 or Count 2. What complicates matters is the MA legal definition of the mens rea for this offense: it appears that this offense requires that the death-causing accident occur “knowingly,” which is more than the threshold required for the manslaughter charge. This is important, because here’s what we now know about jury deliberations in this case:

The motion filed in Norfolk County Superior Court said one juror told the defense team that jurors voted 12-0 to acquit Read of second-degree murder and of leaving the scene of an accident in which there was a death. The lawyers also received second- and third-hand accounts about two other jurors, both indicating there was agreement on acquitting Read of second-degree murder. The motion asked for the murder and leaving-the-scene charges to be dismissed.

Jurors were deadlocked, however, on the charge of manslaughter while operating a motor vehicle under the influence of alcohol, the juror told the attorneys.

Had jury instructions been administered properly, the jury would be on notice that they must deliver a separate verdict for each of the three counts. If that were the case, and if the information in the motion is accurate, then the jury would have voted to acquit of murder and of leaving the scene, and would have deadlocked on the manslaughter charge. This, given what we’ve just noted about mens rea, is a logical parsing out of the evidence. It would suggest that the entire jury was in agreement that intent, or even “knowledge”, was not proven here, and that the dispute revolved around at least one of two questions: (1) whether or not an accident had happened at all (which the defense disputed in their alternative story) (2) whether Read was reckless in causing the accident or had a lesser mens rea (such as negligence or less). This analysis means that an acquittal of the murder and leaving-the-scene charges would not bar a retrial on the manslaughter charges, nor would it create any collateral estoppel that would bar those charges (“the killing did not happen through Read’s knowing/intentional actions” does not bar the possibility that it happened through her recklessness). So, in this scenario, going into a retrial, the defense would have to refocus its energy on the alternative story (i.e., the accident did not happen at all) or on showing a mens rea lesser than recklessness.

But as we know, that’s not what happened. What happened was that the jury was not clear on the fact that they should deliver a separate verdict for each count, and therefore we don’t actually know–as in, ink-on-paper know–what their determination was, which the prosecution now argues makes all three charges fair game. It will be up to SCOTUS to decide whether the evidence we now have of jury deliberations is sufficient for a finding that these were, in fact, acquittals. We know from Fong Foo v. U.S. (1962) that an acquittal, no matter for what reason, is an acquittal and thus a jeopardy bar. In this case, we’d have to recur to Downum v. United States (1963), which holds that mistrials must be interpreted based on their circumstances. Mistrials that are, in essence, acquittal avoidances, should be treated as acquittals.

I’m not sure it’s fair to call what happened here an “acquittal avoidance,” because it doesn’t look like the verdict was solicited and procured with some sort of nefarious intent to avoid a situation in which the jury would acquit. After all, if the jury convicted of one of the homicide offenses (murder or manslaughter) it would necessarily have to acquit of the other, and that would’ve been clear to the judge and to the parties. But what we did get here was a vague verdict that stands in place of an acquittal and possesses enough internal logic to suggest what the jury actually thought to the point that I think justice would be served by dismissing counts 1 and 3 and reprosecuting only for the manslaughter charge.

As a coda, I know there are heated debates about this case and people seem strongly persuaded of Read’s guilt or innocence; I’d just say that the prosecution must be very convinced of her guilt to be reprosecuting after this verdict fiasco, after their own investigator was disbarred for how he conducted the investigation, and after the defense’s alternative theory managed to persuade a considerable number of jurors, including one who just joined the defense team. I’ll be following this with great interest.

The Process Is the Reward

About a decade ago we had an interesting conversation at a faculty meeting about some professors’ policy of banning laptops in class. They argued that the students absorbed the material better and were more engaged.

I was hesitant to implement something like this, not only because I was transitioning to an electronic casebook, but because, as an English-as-a-third-language speaker I am aware of how slowly I write compared to my typing speed. I assumed (correctly) that people would want to take notes in class, and that handwriting would slow them down.

Then, I decided to experiment on myself. I had to attend a conference and decided to try both note-taking styles. Each day included talks that were of comparable professional interest to me. The first day, I wrote notes by hand. The second day I typed. A week later, I remembered plenty from the first day and absolutely nothing from the second.

Despite my <groan> “lived experience” </groan> I decided not to ban laptops in class. I shared the results of my experiment with my students, and while some decided to try taking notes by hand, most kept their laptops. But the whole thing taught me something valuable about the process of learning. Plenty of psychological and educational research confirms this: when note-taking by hand, people carefully consider the material rather than type it unprocessed, which improves recall, even if the end product is less pretty.

Two recent experiences made me think about this. First, I got to read this piece, which is a Jeremiad about what AI has done to postsecondary education (I agree, which is why I will no longer teach paper-based seminars). What was most interesting about the piece, however, was the author’s effort to explain why it was that students recurred to AI to cheat. Sure, it’s easy, and people panic and procrastinate, and all the usual answers. But there’s something else going on:

In Anarchy, State, and Utopia, Robert Nozick writes, “Suppose there were an experience machine that would give you any experience you desired. Superduper neuropsychologists could stimulate your brain so that you would think and feel you were writing a great novel, or making a friend, or reading an interesting book. All the time you would be floating in the tank, with electrodes attached to your brain.” Nozick asks whether you would plug into such a machine for life, with all your experiences no more than preprogrammed illusions. Hedonists would surely reply, “heck yes, the sooner the better!”

Nozick was anti-hedonism. He argued that as appealing as the experience machine initially seems, we would not want to spend our lives attached to it. There are two key reasons why: first, we want to do things with our lives, to accomplish things. It is not merely the artificial feeling of accomplishment that we crave, but the actual fact of it, the truth that we have done something with our time in the world. It is better to write the Great American Novel than just have the false conviction that you have written it, it is better to be loved and have real friendships than buy the machine’s lie otherwise.

Second, the experience machine not only strips us of a connection to real accomplishment, but prevents us from becoming authentic persons. Are you courageous, kind, intelligent, witty, loving, generous? Plugged into the experience machine of phony pleasures, you aren’t anything. Instead, you are simply a passive blob, a receptacle of pleasurable (but fraudulent) sensations. If you prefer to do and have done things and to become a real and genuine person, then you should reject the allure of the experience machine.

That’s Nozick’s argument. Nozick is wrong. AI is the birth of the experience machine, and students prefer it. That’s why they cheat.

Writing is not a tool to express our thoughts—the act of writing is the act of thinking. Sitting down and making words, pondering not merely le mot juste, but how the argument works, what metaphors are appropriate, what really is the conclusion you want to reach, that is the creation of your own point of view. A sophomore telling ChatGPT or Gemini or Claude to write a critique of euthanasia is not developing her own ideas about euthanasia. But it does give the illusion.

Finally, having cheated all the way through college, letting AI do the work, students can have the feeling of accomplishment walking across the stage at graduation, pretending to be an educated person with skills and knowledge that the machines actually have. Pretending to have earned a degree. If Nozick were right then AI would not lead to an explosion of cheating, because students would want the knowledge and understanding that college aims to provide. But in fact many just want the credential. They are hedonists abjuring the development of the self and the forging of their own souls.

This is depressing, and it confirmed something I experienced a couple of years ago when I taught pre-law undergrads across the bridge. It was an online class, and I prepared a lot of interactive assignments in small groups, but very quickly it turned out that the class would not work as planned. Students would return from breakout rooms and report that everyone else in the group was nothing more than a black, unresponsive rectangle, and they couldn’t discuss the assignment. Defeated and disappointed, I had to revert to straight lecturing, which is my least preferred method of conveying information online, because otherwise the few folks that tried to be engaged would just face endless frustration. What the heck? I thought. If they’re not in class, why did they log on? And if they logged on, why won’t they participate? Some of this classroom dysfunction was demystified on December 26th when, to my surprise, I received an email notifying that “Melany’s OtterPilot Has Joined Your Meeting.” It was our regular class time, but of course it was the day after Christmas and the semester had ended weeks prior. When I logged on out of curiosity, a black rectangle representing the student was there. I turned on my camera and said “well, hello there,” and the rectangle disappeared. As per an Internet search, OtterPilot is an AI service that attends virtual classes for you and summarizes them. If it was in wide use throughout the semester and I spent four months lecturing to dozens of bots, it would explain not only the scant participation during online lecture, but also the the lackluster performance and milquetoast answers on the final.

I’m not a megalomaniac and I’m sure my class wasn’t the single most scintillating experience on the planet. But what could 160 19-year-olds have going on on Tuesday and Thursday mornings at 9:30 that was so much better than talking with their friends about jury selection and murder cases and ethical dilemmas about crime–topics that interested them enough to pick the course as an elective? What was so much better on the phone or on the screen during those particular 100 minutes? Wasn’t it a shame that they missed out?

Anyway, now that I’m a student myself again, it’s been interesting to see how my colleagues–highly accomplished second-career folks who are studying something they are passionate about enough to burden themselves with a rigorous program on top of full-time jobs and families–are taking to the panoply of AI tools, which was not around last time when they were at the grad school rodeo. Folks are having their otters and other AI rodents take notes during class and sharing them. I take my notes myself, so I have some basis for comparison. The AI captures some of what happens in class, but when I read it, it seems very removed from the class I attended. Had I not paid attention and taken notes, I would have absorbed nothing from the AI summary. When we have an assignment, hefty as it is, I don’t freak out because the material, which I summarized by hand, is fresh in my head–I don’t need to reconstruct it from disembodied notes.

On the other end of the spectrum is the second experience: my awesome drums class (easily the best two hours of my week outside of family time). I learn drums in a group of marvelous people, in person, from an in-person teacher, not from an app. Everything is very immediate, very hands-on, and very conducive to learning. As someone with a more than decent sense of rhythm and plenty of experience playing instruments in bands and ensembles, I expected drumming to become rewarding more quickly (imagine, as a woodwind player, skipping the long months you spend just working on getting a decent sound out of your instrument! Hit the drum and it plays!). To some degree it’s true. But it does turn out that mastering drumming skills is a multi-step process, and after doing this for about a year, I can distinguish them. Step one, which is usually not an issue for me, comes being able to comprehend and absorb the pattern in your head. Step two is translating what’s in your mind to your hands and legs, i.e., to get it in your body. And step three is the physical ability to perform the pattern (e.g., how fast can one kick the kick drum, how accurately you can play 16ths on the hat and the snare, can you time the hat foot with the kick foot). Now that we’re playing patterns that disconnect the kick foot from the dominant hand hitting the hi-hat, things take a bit longer to digest, and we all spent several weeks “in the soup”, as it were, clumsily trying to separate foot from hand. It was a very physical sense of plateau in step 2. At some point, I felt that I got over that, and then realized, to my dismay, that even as the mind is strong, the flesh is weak: quads were cramping, forearms were hurting, etc. It was only a couple of weeks ago that I had a transcendent sense that we all, one way or another, got over the step 3 hump and accomplished something. Folks, it was exhilarating. I’m sure there are more learning fits and starts in our future, but I can feel us getting better.

Even as I understand, cognitively, the allure of the experience machine, it makes me deeply sad that people don’t want to experience the profundity of the learning process. It’s so interesting and satisfying to see what your one and only brain can do, where it stumbles, how it overcomes, how shortcuts are assembled and dismantled, how neural pathways are forged and overwritten. What possible diploma or grade could be better than rewarding yourself with the experience of your own cognition, experiencing yourself as a full human, and seeing what you’re capable of?

When it’s over, like Mary Oliver, I want to say: all my life I was a bride married to amazement. Don’t you?