I’ve been thinking about something I learned years ago in psychology: negative reinforcement. People sometimes confuse it with punishment, but negative reinforcement is actually when a behavior becomes stronger because it removes something unpleasant.

An obvious example is anxiety. If someone is extremely anxious and discovers something that quickly makes the anxiety disappear, the relief itself can reinforce whatever they just did. The same principle can apply to avoidance. You’re anxious about making a phone call, so you decide to put it off until tomorrow. Immediately, you feel better. Unfortunately, your nervous system may have just learned, “Excellent. Apparently avoiding this thing works.”

That’s where I think psychology gets particularly interesting because it becomes difficult to separate psychology from physiology. Anxiety isn’t merely a thought. Your heart rate, breathing, muscle tension, autonomic nervous system, attention, and interpretation of what is happening can all become involved. Then the relief is physiological as well as psychological, and the brain learns from the entire experience.

You can see versions of this everywhere. Someone repeatedly asks for reassurance because reassurance makes uncertainty disappear. Someone reaches for their phone whenever sitting quietly becomes uncomfortable. In more serious circumstances, rapid relief from anxiety or insomnia can become part of the reinforcing process surrounding certain medications. Obviously these situations aren’t equivalent, but the underlying learning principle is interesting: sometimes we aren’t reinforced because something feels good; we’re reinforced because something stops feeling bad.

Maybe regulation, then, isn’t simply about becoming calm. Part of regulation may be developing the ability to experience some discomfort without immediately treating that discomfort as an emergency that has to disappear.

If every uncomfortable feeling sends us running for the nearest exit, eventually we may forget that sometimes we were perfectly capable of staying in the room.

I’d be curious how other people see this. Where do you think the line is between healthy relief and relief that unintentionally reinforces avoidance?

We spend an enormous amount of time talking about what society should do to people after they commit crimes. How long should the sentence be? Should punishment be harsher? Should someone get another chance? Those are legitimate questions, but sometimes I think we're so obsessed with what happens after the crime that we don't spend nearly enough time asking the more uncomfortable question: What the hell happened before it?

That isn't an excuse for criminal behavior. An adult who hurts another person is still responsible for what he or she does. Childhood abuse doesn't magically erase responsibility, and most people who experience abuse do not grow up to become criminals. That's important, because saying otherwise would be scientifically inaccurate and horribly unfair to survivors.

But acknowledging that doesn't mean we should ignore what the research actually shows.

The relationship between childhood maltreatment and subsequent offending has been studied extensively. A systematic review of 62 prospective and longitudinal studies described the association as well documented, while also emphasizing an important limitation: an association by itself does not prove a simple causal relationship.[1]

A separate meta-analysis of 14 prospective studies involving 20,946 participants found that childhood maltreatment was associated with substantially greater odds of antisocial behavior in adulthood.[2] Another systematic review and meta-analysis found an association between childhood maltreatment and later violence, although the increase in risk was described as modest and there was considerable variation between studies.[3]

In other words, risk is not destiny.

Abuse doesn't insert a criminal program into a child's brain that activates twenty years later. Human beings are considerably more complicated than that. Family relationships, peers, neighborhoods, substance use, mental health, protective relationships and numerous other factors can affect what happens afterward.

But that's precisely why prevention matters.

If experiences during childhood can increase the risk of later offending, helping children after abuse isn't merely an act of compassion. It should also be part of a serious conversation about public safety. Research involving justice-involved youth has found high levels of adverse childhood experiences, and cumulative ACE exposure has also been associated with recidivism.[4]

None of this means that therapy is some magical crime-prevention vaccine. It means that if we're serious about preventing crime, we should be willing to look upstream instead of waiting until somebody commits one.

We should be capable of holding two ideas in our heads at the same time: people are responsible for their actions, and human behavior develops within biological, psychological and social environments. Explanation and exoneration are not the same fucking thing.

I would rather help an abused eight-year-old before he becomes a traumatized eighteen-year-old than wait until he's a twenty-eight-year-old defendant and congratulate ourselves for finally getting tough.

Because by then, somebody else may already have become his victim.

References

[1] Malvaso, C. G., Delfabbro, P., & Day, A. (2018). The maltreatment–offending association: A systematic review of the methodological features of prospective and longitudinal studies. Trauma, Violence, & Abuse, 19(1), 20–34. DOI: 10.1177/1524838015620820.

[2] Braga, T., Cunha, O., & Maia, Â. (2018). The enduring effect of maltreatment on antisocial behavior: A meta-analysis of longitudinal studies. Aggression and Violent Behavior, 40, 91–100. DOI: 10.1016/j.avb.2018.04.003.

[3] Fitton, L., Yu, R., & Fazel, S. (2020). Childhood maltreatment and violent outcomes: A systematic review and meta-analysis of prospective studies. Trauma, Violence, & Abuse, 21(4), 754–768. DOI: 10.1177/1524838018795269.

[4] Astridge, B., Li, W. W., McDermott, B., & Longhitano, C. (2023). A systematic review and meta-analysis on adverse childhood experiences: Prevalence in youth offenders and their effects on youth recidivism. Child Abuse & Neglect, 140, 106055. DOI: 10.1016/j.chiabu.2023.106055.

We spend an enormous amount of time talking about what society should do to people after they commit crimes. How long should the sentence be? Should punishment be harsher? Should someone get another chance? Those are legitimate questions, but sometimes I think we're so obsessed with what happens after the crime that we don't spend nearly enough time asking the more uncomfortable question: What the hell happened before it?

That isn't an excuse for criminal behavior. An adult who hurts another person is still responsible for what he or she does. Childhood abuse doesn't magically erase responsibility, and most people who experience abuse do not grow up to become criminals. That's important, because saying otherwise would be scientifically inaccurate and horribly unfair to survivors.

But acknowledging that doesn't mean we should ignore what the research actually shows.

The relationship between childhood maltreatment and subsequent offending has been studied extensively. A systematic review of 62 prospective and longitudinal studies described the association as well documented, while also emphasizing an important limitation: an association by itself does not prove a simple causal relationship.\\\[1\\\]

A separate meta-analysis of 14 prospective studies involving 20,946 participants found that childhood maltreatment was associated with substantially greater odds of antisocial behavior in adulthood.\\\[2\\\] Another systematic review and meta-analysis found an association between childhood maltreatment and later violence, although the increase in risk was described as modest and there was considerable variation between studies.\\\[3\\\]

In other words, risk is not destiny.

Abuse doesn't insert a criminal program into a child's brain that activates twenty years later. Human beings are considerably more complicated than that. Family relationships, peers, neighborhoods, substance use, mental health, protective relationships and numerous other factors can affect what happens afterward.

But that's precisely why prevention matters.

If experiences during childhood can increase the risk of later offending, helping children after abuse isn't merely an act of compassion. It should also be part of a serious conversation about public safety. Research involving justice-involved youth has found high levels of adverse childhood experiences, and cumulative ACE exposure has also been associated with recidivism.\\\[4\\\]

None of this means that therapy is some magical crime-prevention vaccine. It means that if we're serious about preventing crime, we should be willing to look upstream instead of waiting until somebody commits one.

We should be capable of holding two ideas in our heads at the same time: people are responsible for their actions, and human behavior develops within biological, psychological and social environments. Explanation and exoneration are not the same fucking thing.

I would rather help an abused eight-year-old before he becomes a traumatized eighteen-year-old than wait until he's a twenty-eight-year-old defendant and congratulate ourselves for finally getting tough.

Because by then, somebody else may already have become his victim.

References

\\\[1\\\] Malvaso, C. G., Delfabbro, P., & Day, A. (2018). The maltreatment–offending association: A systematic review of the methodological features of prospective and longitudinal studies. Trauma, Violence, & Abuse, 19(1), 20–34. DOI: 10.1177/1524838015620820.

\\\[2\\\] Braga, T., Cunha, O., & Maia, Â. (2018). The enduring effect of maltreatment on antisocial behavior: A meta-analysis of longitudinal studies. Aggression and Violent Behavior, 40, 91–100. DOI: 10.1016/j.avb.2018.04.003.

\\\[3\\\] Fitton, L., Yu, R., & Fazel, S. (2020). Childhood maltreatment and violent outcomes: A systematic review and meta-analysis of prospective studies. Trauma, Violence, & Abuse, 21(4), 754–768. DOI: 10.1177/1524838018795269.

\\\[4\\\] Astridge, B., Li, W. W., McDermott, B., & Longhitano, C. (2023). A systematic review and meta-analysis on adverse childhood experiences: Prevalence in youth offenders and their effects on youth recidivism. Child Abuse & Neglect, 140, 106055. DOI: 10.1016/j.chiabu.2023.106055.

“When the Punishment Is Irreversible, What the F#ck Are We Doing?” I have been thinking a great deal about capital punishment lately, particularly after reading about the execution of James Broadnax in Texas. The more I thought about it, the more I kept returning to a question that sounds almost embarrassingly obvious: If we’re going to execute somebody, shouldn’t we be really, really f*cking sure we have the right person?

Apparently, that question becomes considerably more complicated once lawyers, appeals, procedural rules, deadlines, confessions, recantations, forensic evidence and the institutional need for cases eventually to end enter the picture. I understand why a legal system needs finality. We cannot retry every criminal case every Tuesday because somebody remembered something different over breakfast. Memories change, witnesses can lie, recantations can be self-serving, and hindsight is notoriously unreliable. Human memory is not a surveillance camera conveniently stored somewhere in the hippocampus waiting for us to press PLAY. But there is one rather significant complication here: we are talking about killing somebody. That strikes me as a particularly bad place for an “Oh sh!t, we got that wrong” moment.

The Broadnax case illustrates why this bothers me. Broadnax was convicted of capital murder for the 2008 killings of two men during a robbery. Shortly before his execution, his attorneys presented a declaration from his co-defendant, Demarius Cummings, who said under penalty of perjury that he—not Broadnax—had shot the victims. Broadnax’s attorneys also argued that DNA evidence supported that account because Cummings’s DNA, rather than Broadnax’s, had been found on the murder weapon and in the pocket of one of the victims.

Does that prove Broadnax was innocent? No. That qualification matters enormously. A person can lie twenty years after a crime just as a person can lie twenty minutes after one. Memories deteriorate, motives change, and hindsight doesn’t magically become reliable merely because somebody signs an affidavit. But what the f$ck? If somebody is about to be executed and substantial new evidence potentially changes our understanding of who actually committed the killing, shouldn’t our response be something closer to, “Hold on, let’s make absolutely sure,” rather than treating the scheduled execution date like we’re trying to avoid a cancellation fee at a restaurant? Imagine calling and saying, “We may have evidence that changes the entire case,” and hearing, “I’m terribly sorry, sir, but executions cancelled with less than 24 hours’ notice are subject to the full charge.” At some point, bureaucracy becomes indistinguishable from a Monty Python sketch, except unfortunately somebody actually dies at the end.

Why Would an Innocent Person Confess?

This is another question that sounds obvious until psychology screws up the obvious answer. Most people intuitively think, “If I didn’t murder somebody, why the hell would I ever say that I did?” I understand that reaction because it seems completely logical. If somebody accused me of murdering a person I had never met, I would like to imagine myself confidently saying, “Nope. Wrong Jonathan. Have a pleasant afternoon.”

Unfortunately, human beings don’t always behave like calm little rational-choice machines, particularly when they are frightened, exhausted, intoxicated, developmentally vulnerable, severely distressed or psychologically impaired. Research on false confessions has demonstrated something deeply uncomfortable: innocent people sometimes confess to crimes they didn’t commit. That doesn’t mean every distressed suspect who confesses is innocent. It means the confession itself cannot automatically end the inquiry into reliability.

Consider an extreme hypothetical. Imagine somebody who is profoundly depressed and suicidal. He is questioned for hours and eventually reaches the point where he says, “Fine. I did it. Execute me. I don’t give a shit anymore.” Well, hold on. If somebody wants to die, then saying something likely to get him killed doesn’t necessarily tell us whether he committed the crime. It may tell us something entirely different about his psychological state. Asking a severely suicidal person, “Would you rather go home or be executed?” and treating “executed” as compelling evidence of murder would make about as much sense as asking me at 6:00 in the morning whether I would rather go to work or be eaten by a rhinoceros and then concluding that I must have some long-standing sexual attraction to rhinoceroses. The answer may tell you considerably more about my feelings toward 6:00 in the morning than it does about rhinoceroses.

This is where I start wondering why psychology isn’t incorporated more aggressively into safeguards for particularly vulnerable suspects. I am not suggesting that every police interrogation needs a psychologist sitting in the corner with a clipboard saying, “Interesting. How did that accusation make you feel?” That would become absurd rather quickly. Detectives have crimes to investigate, psychologists aren’t human lie detectors, and mental-health professionals can make mistakes too.

But when there are serious indications that somebody is developmentally impaired, acutely psychologically unstable, severely distressed or otherwise unable to participate reliably in an interrogation, why wouldn’t we want an independent professional safeguard? Seriously: why the fu%k wouldn’t we? If the purpose of an interrogation is to discover what actually happened, then anything that improves the reliability of the information should help the police as well. An independent safeguard shouldn’t be viewed as somebody ruining the interrogation. The safeguard should help ensure that whatever confession eventually comes out of that room is worth more than the paper on which somebody wrote it.

Finality Is Necessary. Death Is Pretty Darn Final Too.

I understand the argument for procedural finality better than my initial emotional reaction might suggest. Cases have to end eventually. Witnesses die. Evidence disappears. Memories become less reliable. Victims’ families shouldn’t be forced to relive a murder indefinitely because every new claim automatically restarts the entire judicial process. Those are legitimate concerns.

But execution introduces a rather spectacular problem into the concept of finality: you cannot unexecute somebody. If we discover that somebody has been wrongfully imprisoned, we can release him. We cannot return the years we took from him, and no amount of money can fully repair what happened, but at least we can stop doing it. Execution eliminates even that possibility.

This is where I become uncomfortable when procedural arguments start sounding more important than substantive questions about guilt. Procedure matters tremendously because procedure protects everybody. But procedure is supposed to be a mechanism for producing justice; justice isn’t supposed to become a mechanism for protecting procedure. Otherwise, we’ve created the world’s most terrifying version of customer service: “I’m sorry, sir. We understand that you may be innocent, but your appeal window has closed.” “I’m scheduled to be executed.” “Yes, sir. Is there anything else I can help you with today?” Perhaps they’ll offer him a customer-satisfaction survey afterward. “On a scale of one to ten, how satisfied were you with today’s execution experience?” Unfortunately, selecting zero probably won’t route the call to a supervisor.

At some point the absurdity stops being merely bureaucratic and becomes horrifying.

This Is Also About Proportionality

This connects directly to something I’ve been exploring in my writing about vigilantism and what happens when people begin taking the law into their own hands. One of the reasons we have legal systems in the first place is that anger is absolutely terrible at proportionality.

Suppose somebody walks into my store and steals a candy bar. I’m angry. Maybe I’m furious. Perhaps it was my favorite fucking candy bar. Nevertheless, my anger does not give me the authority to chase the person down the street and execute him. If it did, we’d need armed guards standing beside every Snickers display in America, and Walgreens would start looking like the perimeter of a military installation. “Sir, step away from the Kit Kat. Slowly. We’ve already lost three people near the Reese’s.”

Why don’t we handle theft that way? Because we distinguish among levels of wrongdoing. We distinguish theft from assault, assault from murder, negligence from intent, participation from direct responsibility, and mitigating circumstances from aggravating ones. We ask what happened, what the evidence establishes, how culpable somebody was and what punishment is proportionate. Vigilantism collapses that entire process into you wronged me → I know you’re guilty → I decide what you deserve → I administer the punishment. That’s dangerous because everybody thinks his own grievance is the important one.

But here’s the uncomfortable part: the state needs proportionality too. The fact that a horrific crime occurred doesn’t eliminate our responsibility to determine precisely who did what. If two people participated in a robbery and one actually committed the killings, that distinction matters. If somebody acted intentionally rather than negligently, that matters. If evidence is uncertain, that matters. If someone’s psychological condition affected the reliability of a confession, that matters. Mitigating circumstances aren’t some annoying little loophole lawyers invented because they were bored one afternoon. They are part of what separates justice from vengeance.

The Wrong Place at the Wrong F#cking Time

Here’s the part that scares me personally. Most of us walk around assuming the criminal justice system applies to criminals. Therefore, if we’re not criminals, we’re safe. That’s wonderfully comforting. It’s also not entirely how reality works.

Imagine that you have never committed a serious crime in your life. Then one night you happen to be in exactly the wrong place at exactly the wrong time. Somebody identifies you. Circumstantial evidence happens to point toward you. Maybe you’re psychologically overwhelmed during questioning. Maybe you say something stupid. Maybe somebody else lies. Maybe investigators become convinced you’re guilty and consequently begin interpreting ambiguous evidence through that assumption.

None of this requires evil people sitting around a table saying, “Excellent. Which innocent person shall we execute today?” That’s actually what makes the problem more frightening. Real institutional mistakes generally aren’t committed by a committee of villains wearing black capes and stroking cats while lightning flashes outside the window. If they were, this would be much easier. We could simply institute a rule that anyone who brings a cape, a cat and a giant lever marked EXECUTE INNOCENT PERSON into the courthouse immediately loses prosecutorial discretion.

The actual problem is more mundane: good people operating inside imperfect systems can make catastrophic mistakes. Police officers can be wrong. Prosecutors can be wrong. Defense attorneys can be wrong. Psychologists can be wrong. Judges can be wrong. Juries can be wrong. I can be wrong. You can be wrong. Human beings being wrong is not exactly a shocking discovery. We’ve had several thousand years of evidence. So why would we design an irreversible punishment around the assumption that this particular human institution will never make an irreversible mistake?

Admitting Error Isn’t Weakness

This may be the part that bothers me most. A justice system should not regard discovering one of its own mistakes as an institutional humiliation. Finding and correcting mistakes is what a functioning system is supposed to do.

Science doesn’t become weaker when evidence overturns an old hypothesis. Medicine doesn’t become illegitimate because physicians revise a diagnosis when new information appears. Aviation doesn’t say, “Well, we’ve been building the wing this way for thirty years, and admitting there’s a defect would be terribly embarrassing, so let’s keep doing it.” Hopefully. We investigate the defect because we’d rather experience embarrassment than watch another airplane fall out of the f&cking sky.

Criminal justice should operate with the same humility, particularly when a human life is at stake. This doesn’t mean every confession is false. It doesn’t mean every person on death row is innocent. It doesn’t mean every last-minute affidavit is credible, every prosecutor is corrupt, every police interrogation is coercive, or every conviction should be endlessly reopened. It means something considerably less dramatic: human beings make mistakes, and therefore the greater and more irreversible the punishment, the stronger our safeguards should become.

Perhaps that means greater scrutiny of vulnerable suspects during interrogation. Perhaps it means broader access to independent psychological evaluation under defined circumstances. Perhaps it means better recording requirements, improved interrogation practices, or especially rigorous review when genuinely significant new evidence emerges before an execution. Those are policy questions reasonable people can debate. But the underlying principle shouldn’t be controversial: if the state is going to take a human life, “We’re pretty sure” is a terrifying f&cking standard.

The question shouldn’t merely be whether we’ve reached the end of the legal process. The question should remain, right up until the moment when the answer becomes irreversible: Are we sufficiently certain that we got the truth? Because if the answer later turns out to be no, there is no appeal from death. There is only the institutional equivalent of looking around the room, realizing what just happened, and saying:

“Oh, f*ck.”

Author’s Note: The questions raised in this essay—particularly the relationship among proportionality, institutional rules, psychological vulnerability, and the values we claim a justice system should uphold—are part of a broader set of questions I continue to explore in my writing.

I have been thinking a great deal about capital punishment lately, particularly after reading about the execution of James Broadnax in Texas. The more I thought about it, the more I kept returning to a question that sounds almost embarrassingly obvious: If we’re going to execute somebody, shouldn’t we be really, really fucking sure we have the right person?

Apparently, that question becomes considerably more complicated once lawyers, appeals, procedural rules, deadlines, confessions, recantations, forensic evidence and the institutional need for cases eventually to end enter the picture. I understand why a legal system needs finality. We cannot retry every criminal case every Tuesday because somebody remembered something different over breakfast. Memories change, witnesses can lie, recantations can be self-serving, and hindsight is notoriously unreliable. Human memory is not a surveillance camera conveniently stored somewhere in the hippocampus waiting for us to press PLAY. But there is one rather significant complication here: we are talking about killing somebody. That strikes me as a particularly bad place for an “Oh shit, we got that wrong” moment.

The Broadnax case illustrates why this bothers me. Broadnax was convicted of capital murder for the 2008 killings of two men during a robbery. Shortly before his execution, his attorneys presented a declaration from his co-defendant, Demarius Cummings, who said under penalty of perjury that he—not Broadnax—had shot the victims. Broadnax’s attorneys also argued that DNA evidence supported that account because Cummings’s DNA, rather than Broadnax’s, had been found on the murder weapon and in the pocket of one of the victims.

Does that prove Broadnax was innocent? No. That qualification matters enormously. A person can lie twenty years after a crime just as a person can lie twenty minutes after one. Memories deteriorate, motives change, and hindsight doesn’t magically become reliable merely because somebody signs an affidavit. But what the fuck? If somebody is about to be executed and substantial new evidence potentially changes our understanding of who actually committed the killing, shouldn’t our response be something closer to, “Hold on, let’s make absolutely sure,” rather than treating the scheduled execution date like we’re trying to avoid a cancellation fee at a restaurant? Imagine calling and saying, “We may have evidence that changes the entire case,” and hearing, “I’m terribly sorry, sir, but executions cancelled with less than 24 hours’ notice are subject to the full charge.” At some point, bureaucracy becomes indistinguishable from a Monty Python sketch, except unfortunately somebody actually dies at the end.

Why Would an Innocent Person Confess?

This is another question that sounds obvious until psychology screws up the obvious answer. Most people intuitively think, “If I didn’t murder somebody, why the hell would I ever say that I did?” I understand that reaction because it seems completely logical. If somebody accused me of murdering a person I had never met, I would like to imagine myself confidently saying, “Nope. Wrong Jonathan. Have a pleasant afternoon.”

Unfortunately, human beings don’t always behave like calm little rational-choice machines, particularly when they are frightened, exhausted, intoxicated, developmentally vulnerable, severely distressed or psychologically impaired. Research on false confessions has demonstrated something deeply uncomfortable: innocent people sometimes confess to crimes they didn’t commit. That doesn’t mean every distressed suspect who confesses is innocent. It means the confession itself cannot automatically end the inquiry into reliability.

Consider an extreme hypothetical. Imagine somebody who is profoundly depressed and suicidal. He is questioned for hours and eventually reaches the point where he says, “Fine. I did it. Execute me. I don’t give a shit anymore.” Well, hold on. If somebody wants to die, then saying something likely to get him killed doesn’t necessarily tell us whether he committed the crime. It may tell us something entirely different about his psychological state. Asking a severely suicidal person, “Would you rather go home or be executed?” and treating “executed” as compelling evidence of murder would make about as much sense as asking me at 6:00 in the morning whether I would rather go to work or be eaten by a rhinoceros and then concluding that I must have some long-standing sexual attraction to rhinoceroses. The answer may tell you considerably more about my feelings toward 6:00 in the morning than it does about rhinoceroses.

This is where I start wondering why psychology isn’t incorporated more aggressively into safeguards for particularly vulnerable suspects. I am not suggesting that every police interrogation needs a psychologist sitting in the corner with a clipboard saying, “Interesting. How did that accusation make you feel?” That would become absurd rather quickly. Detectives have crimes to investigate, psychologists aren’t human lie detectors, and mental-health professionals can make mistakes too.

But when there are serious indications that somebody is developmentally impaired, acutely psychologically unstable, severely distressed or otherwise unable to participate reliably in an interrogation, why wouldn’t we want an independent professional safeguard? Seriously: why the fuck wouldn’t we? If the purpose of an interrogation is to discover what actually happened, then anything that improves the reliability of the information should help the police as well. An independent safeguard shouldn’t be viewed as somebody ruining the interrogation. The safeguard should help ensure that whatever confession eventually comes out of that room is worth more than the paper on which somebody wrote it.

Finality Is Necessary. Death Is Pretty Damn Final Too.

I understand the argument for procedural finality better than my initial emotional reaction might suggest. Cases have to end eventually. Witnesses die. Evidence disappears. Memories become less reliable. Victims’ families shouldn’t be forced to relive a murder indefinitely because every new claim automatically restarts the entire judicial process. Those are legitimate concerns.

But execution introduces a rather spectacular problem into the concept of finality: you cannot unexecute somebody. If we discover that somebody has been wrongfully imprisoned, we can release him. We cannot return the years we took from him, and no amount of money can fully repair what happened, but at least we can stop doing it. Execution eliminates even that possibility.

This is where I become uncomfortable when procedural arguments start sounding more important than substantive questions about guilt. Procedure matters tremendously because procedure protects everybody. But procedure is supposed to be a mechanism for producing justice; justice isn’t supposed to become a mechanism for protecting procedure. Otherwise, we’ve created the world’s most terrifying version of customer service: “I’m sorry, sir. We understand that you may be innocent, but your appeal window has closed.” “I’m scheduled to be executed.” “Yes, sir. Is there anything else I can help you with today?” Perhaps they’ll offer him a customer-satisfaction survey afterward. “On a scale of one to ten, how satisfied were you with today’s execution experience?” Unfortunately, selecting zero probably won’t route the call to a supervisor.

At some point the absurdity stops being merely bureaucratic and becomes horrifying.

This Is Also About Proportionality

This connects directly to something I’ve been exploring in my writing about vigilantism and what happens when people begin taking the law into their own hands. One of the reasons we have legal systems in the first place is that anger is absolutely terrible at proportionality.

Suppose somebody walks into my store and steals a candy bar. I’m angry. Maybe I’m furious. Perhaps it was my favorite fucking candy bar. Nevertheless, my anger does not give me the authority to chase the person down the street and execute him. If it did, we’d need armed guards standing beside every Snickers display in America, and Walgreens would start looking like the perimeter of a military installation. “Sir, step away from the Kit Kat. Slowly. We’ve already lost three people near the Reese’s.”

Why don’t we handle theft that way? Because we distinguish among levels of wrongdoing. We distinguish theft from assault, assault from murder, negligence from intent, participation from direct responsibility, and mitigating circumstances from aggravating ones. We ask what happened, what the evidence establishes, how culpable somebody was and what punishment is proportionate. Vigilantism collapses that entire process into you wronged me → I know you’re guilty → I decide what you deserve → I administer the punishment. That’s dangerous because everybody thinks his own grievance is the important one.

But here’s the uncomfortable part: the state needs proportionality too. The fact that a horrific crime occurred doesn’t eliminate our responsibility to determine precisely who did what. If two people participated in a robbery and one actually committed the killings, that distinction matters. If somebody acted intentionally rather than negligently, that matters. If evidence is uncertain, that matters. If someone’s psychological condition affected the reliability of a confession, that matters. Mitigating circumstances aren’t some annoying little loophole lawyers invented because they were bored one afternoon. They are part of what separates justice from vengeance.

The Wrong Place at the Wrong Fucking Time

Here’s the part that scares me personally. Most of us walk around assuming the criminal justice system applies to criminals. Therefore, if we’re not criminals, we’re safe. That’s wonderfully comforting. It’s also not entirely how reality works.

Imagine that you have never committed a serious crime in your life. Then one night you happen to be in exactly the wrong place at exactly the wrong time. Somebody identifies you. Circumstantial evidence happens to point toward you. Maybe you’re psychologically overwhelmed during questioning. Maybe you say something stupid. Maybe somebody else lies. Maybe investigators become convinced you’re guilty and consequently begin interpreting ambiguous evidence through that assumption.

None of this requires evil people sitting around a table saying, “Excellent. Which innocent person shall we execute today?” That’s actually what makes the problem more frightening. Real institutional mistakes generally aren’t committed by a committee of villains wearing black capes and stroking cats while lightning flashes outside the window. If they were, this would be much easier. We could simply institute a rule that anyone who brings a cape, a cat and a giant lever marked EXECUTE INNOCENT PERSON into the courthouse immediately loses prosecutorial discretion.

The actual problem is more mundane: good people operating inside imperfect systems can make catastrophic mistakes. Police officers can be wrong. Prosecutors can be wrong. Defense attorneys can be wrong. Psychologists can be wrong. Judges can be wrong. Juries can be wrong. I can be wrong. You can be wrong. Human beings being wrong is not exactly a shocking discovery. We’ve had several thousand years of evidence. So why would we design an irreversible punishment around the assumption that this particular human institution will never make an irreversible mistake?

Admitting Error Isn’t Weakness

This may be the part that bothers me most. A justice system should not regard discovering one of its own mistakes as an institutional humiliation. Finding and correcting mistakes is what a functioning system is supposed to do.

Science doesn’t become weaker when evidence overturns an old hypothesis. Medicine doesn’t become illegitimate because physicians revise a diagnosis when new information appears. Aviation doesn’t say, “Well, we’ve been building the wing this way for thirty years, and admitting there’s a defect would be terribly embarrassing, so let’s keep doing it.” Hopefully. We investigate the defect because we’d rather experience embarrassment than watch another airplane fall out of the fucking sky.

Criminal justice should operate with the same humility, particularly when a human life is at stake. This doesn’t mean every confession is false. It doesn’t mean every person on death row is innocent. It doesn’t mean every last-minute affidavit is credible, every prosecutor is corrupt, every police interrogation is coercive, or every conviction should be endlessly reopened. It means something considerably less dramatic: human beings make mistakes, and therefore the greater and more irreversible the punishment, the stronger our safeguards should become.

Perhaps that means greater scrutiny of vulnerable suspects during interrogation. Perhaps it means broader access to independent psychological evaluation under defined circumstances. Perhaps it means better recording requirements, improved interrogation practices, or especially rigorous review when genuinely significant new evidence emerges before an execution. Those are policy questions reasonable people can debate. But the underlying principle shouldn’t be controversial: if the state is going to take a human life, “We’re pretty sure” is a terrifying fucking standard.

The question shouldn’t merely be whether we’ve reached the end of the legal process. The question should remain, right up until the moment when the answer becomes irreversible: Are we sufficiently certain that we got the truth? Because if the answer later turns out to be no, there is no appeal from death. There is only the institutional equivalent of looking around the room, realizing what just happened, and saying:

“Oh, fuck.”

Author’s Note: The questions raised in this essay—particularly the relationship among proportionality, institutional rules, psychological vulnerability, and the values we claim a justice system should uphold—will be discussed briefly in my upcoming book, The Values We Pretend to Have. I also plan to expand these ideas into a larger book-length project and, ultimately, explore them as the basis for a film, much as I have been developing related questions about vigilantism, proportionality, and the breakdown of legal boundaries in The Permission Structure.

There are times when I’ve looked back at an emotional reaction and wondered why something affected me as strongly as it did. Sometimes the immediate event really wasn’t enormous. Somebody said something, another driver behaved recklessly, a family member became frustrated, or an old joke landed badly. Yet I’ve increasingly come to believe that asking whether the immediate event was “big enough” to justify the reaction can miss something important. We don’t necessarily meet each new stressor empty-handed. Sometimes we arrive carrying everything that came before it.

I was thinking about this recently during Yom Kippur. There was a line in the prayers about being led astray, and I asked my rabbi how that fit with the idea of free will. If we’re responsible for our choices, shouldn’t we also have the capacity to keep ourselves from going astray? His response was essentially that we’re human. We make mistakes, and there are limits to what human beings can handle. Under enough stress, maintaining the self-control we might have under easier circumstances can become considerably harder.

Thanks for reading! Subscribe for free to receive new posts and support my work.

That immediately connected with something I’ve spent a great deal of time thinking about psychologically: capacity. A new stressor doesn’t always arrive by itself. It may arrive on top of fear, exhaustion, anger, disappointment, humiliation or unresolved experiences that were already there. The proverbial straw that breaks the camel’s back doesn’t explain very much if we study only the straw and ignore everything the camel was carrying before somebody put it there.

I’ve experienced this in situations that, viewed from the outside, might seem overwhelming at the time. For instance, when a driver in a truck cut dangerously close in front of me shortly after I had already been through a stressful car accident. Because of my recent accident and what I knew from my insurance agent about how rear-end collisions are typically faulted—meaning I could easily be blamed if someone cut me off and I hit them—that moment completely dysregulated me. It tapped directly into lingering fear, panic, and a very real vulnerability about being held responsible for an accident I didn’t cause.

Looking backward, I realize I could have handled that stressor much more proactively. If I had installed a dashboard video camera beforehand, it would have captured the proof and recorded the license plate, ensuring the insurance company and police had solid evidence. Had I taken that practical step in advance, getting cut off would still have been annoying and inconvenient, but it wouldn’t have thrown me into such an intense, reactive panic because I would have felt confident that the situation was resolvable.

Perhaps that’s one of the ways capacity develops. We sometimes talk about emotional regulation as though it means acquiring a larger psychological gas tank: somehow becoming tougher so that more things can be dumped into us before we finally explode. I’m beginning to think that’s incomplete. Capacity can also come from learning how to carry less unnecessary weight into the next encounter. That may mean processing something painful, setting a boundary, solving a practical problem, recognizing an old trigger, changing an interpretation, asking for help or simply understanding why a particular situation affects us so strongly.

The same principle can operate in relationships. When somebody explodes over one disappointment, the argument may ostensibly be about what happened today while emotionally incorporating ten things that happened before today. A parent frightened by another setback in an adult child’s employment, for example, may not be reacting solely to that particular job. Fear about earlier disappointments, uncertainty about the future and frustration over problems that never felt resolved may all enter the conversation simultaneously. Understanding that accumulated load doesn’t make yelling or hurtful behavior acceptable, but it can help explain why the apparent size of the trigger and the intensity of the reaction don’t always match.

That distinction matters to me because explanation is not exoneration. If I understand why I’m angry, I haven’t therefore earned permission to mistreat somebody. In fact, understanding the reaction may give me more opportunity to take responsibility for what I do with it. Instead of saying, “That’s just how I react,” I can begin asking, “What exactly am I reacting to? How much belongs to this moment, and how much did I carry into it?”

That brings me back to Yom Kippur. Perhaps the prayer I want to make isn’t simply, “G-d, don’t let difficult things happen to me.” Difficult things are going to happen. Nor is it, “G-d, make me so strong that nothing ever affects me.” I’m human, and things are supposed to affect me. What I find myself wanting to ask for is something more complicated: Help me understand what I’m carrying. Help me resolve what can be resolved, prepare for what can be anticipated, and develop enough awareness that yesterday’s pain doesn’t automatically decide tomorrow’s behavior.

Maybe that is one way of understanding what it means not to be led astray. Free will doesn’t require pretending that history, stress, fear and exhaustion exert no influence over us. It means trying to create enough space between what happens to us and what we do next that we still have some say in the matter.

And perhaps capacity isn’t the ability to carry an unlimited number of straws.

Perhaps it’s learning, whenever possible, to put some of them down before the next one arrives.

https://preview.redd.it/6bjqxosf58rh1.jpg?width=1408&format=pjpg&auto=webp&s=af75a70c8aac78c6c75201a5ea70412f14d9f3b9