Daniel was on a train to Tel Aviv with Ezra when a woman who was clearly not in her right mind started physically attacking a passenger and screaming at anyone near her. People reacted quickly, someone called security, and the security guards initially refused to remove her. They eventually detained her and started taking statements from the people she hit. Daniel wasn't in the immediate vicinity, so his first instinct was protecting Ezra. But he's been turning over the questions ever since. What is actually lawfully permitted in that split second? What are the parameters for a citizen's arrest in Israel? How do you balance protecting your child with the instinct to intervene? And he's noticed a pattern, both in this incident and a bus driver fight he saw previously, where people seem to be waiting for an off-duty police officer to materialize and handle it.
The security guards refusing to act is the detail that sticks with me. You have uniformed personnel whose entire job is train security, they're summoned to a violent situation, and their first move is to say no. That's not just a failure of individual initiative. Something structural is going on there.
The gap between what the law says you can do and what people actually feel permitted to do.
That's the whole episode right there. So let's start with what the law actually says, because the legal framework is surprisingly broad. Most people don't know what they're allowed to do, and that uncertainty is half the reason nobody moves.
The Penal Law from 1977, section 34J, is the self-defense provision. Force is justified when there's an immediate threat to life or limb, but the response has to be proportional. That word, proportional, is doing an enormous amount of work.
It's the hinge the whole thing swings on. Proportional doesn't mean you have to match the attacker's force exactly. If someone is punching a passenger, you're not limited to punching them back with exactly the same intensity. The standard is whether a reasonable person in your position would have considered the force necessary to stop the threat. The law doesn't require heroism, but it doesn't forbid it either.
And the threat to someone else counts. You don't have to be the one getting hit. Defense of others is baked into the same provision.
Right. If you see a woman attacking a random passenger, as Daniel described, the legal justification for physical intervention is there. The threat is immediate, it's happening in front of you, and force to stop it is permitted. But. And this is the but that makes people freeze. The proportionality assessment happens after the fact, in a courtroom or a police station, with the benefit of hindsight and calm deliberation. You're making the decision in half a second while your kid is behind you.
That's the split. Legally permitted in the abstract, terrifying to actually do.
Now the citizen's arrest piece is separate. Under the Criminal Procedure Law, specifically the arrest provisions, any private person can arrest someone committing a crime without a warrant. The conditions are narrow. The crime has to be in progress, the arrest has to be immediate, and the person has to be handed over to authorities promptly. You can't detain someone and then wander off. You can't arrest someone because you think they might commit a crime. It has to be happening right now.
So in Daniel's train situation, the woman is actively assaulting passengers. That's a crime in progress. A citizen's arrest would be legally available.
It would. But here's the part nobody tells you. If you perform a citizen's arrest and it turns out the arrest was unlawful, or you used excessive force in the process, you're personally exposed. Criminal liability, civil liability, the whole thing. The law gives you the tool but doesn't promise to protect you if you use it imperfectly.
And the guards who refused to intervene probably know that better than anyone.
We'll get to the guards. But first, the mental illness dimension, because Daniel flagged it explicitly. The woman seemed clearly mentally ill. She was carrying a cart of what looked like garbage. She was out of her mind. Does that change the legal calculus for a bystander?
My instinct is it shouldn't, for the immediate threat assessment.
It doesn't. And this is one of those misconceptions that trips people up. The law cares about the threat, not the attacker's mental state, in the moment. If someone is swinging at you or at another passenger, the fact that they're psychotic rather than malicious doesn't make the punches less real. Self-defense and defense of others still apply. The proportional force standard is the same.
Where the mental illness matters is afterward.
The attacker's mental state affects their criminal liability. They may be found not guilty by reason of insanity, or diverted into psychiatric care rather than prison. But that's a determination for a court, not for a bystander on a train platform. Your legal duty in the moment is to assess the threat, not to diagnose the person making it.
Though I'll say, practically, it does change how people react. The hesitation Daniel described, the sense that this woman wasn't in control of herself, that's real. People are less willing to use force against someone they perceive as sick rather than criminal.
And that's not irrational. It's human. But the law doesn't require you to make that distinction in the heat of the moment. The reasonable person standard takes into account that you're acting under stress, with incomplete information. The UK's Crown Prosecution Service, for example, explicitly considers whether a bystander acted in the heat of the moment when judging reasonable force. Israeli courts apply a similar principle, even if it's not codified in quite the same language.
You're not expected to calmly weigh options while someone is being attacked three meters away.
And that brings us to the thing Daniel actually did, which was protect Ezra. I want to say this plainly because I think it's important. Prioritizing your child's safety is not a failure to intervene. It's the correct first move. You can't help anyone if your kid is now also in danger. The sequence matters. Secure dependents, assess the situation, then act if you can.
Daniel seemed to feel some tension about that. The question of whether he should have done more.
He was there with a one-year-old. Ezra was born in July last year, so he's just barely a toddler. Your first obligation is to the child in your arms. The law recognizes this implicitly. The reasonable person standard includes your circumstances. A parent with a small child is not in the same position as an unencumbered adult.
Let's talk about the waiting-for-a-uniform phenomenon. Daniel described it as people waiting for an off-duty police officer to appear. I think that's exactly right, and it's a specifically Israeli variant of the bystander effect.
The classic bystander effect is about diffusion of responsibility. The more people present, the less likely any individual is to act, because everyone assumes someone else will. But the Israeli version has an extra layer. Universal military service means a large portion of the population has some combat training. There's a cultural assumption that someone with more training than you is always nearby. An off-duty soldier, a police officer, a security guard. Someone who knows what they're doing.
And that assumption lets everyone off the hook.
It does. And it's often wrong. The person with training might not be there. Or they might be there but not intervening, because they're off-duty and unarmed and also weighing the same risks you are. The expectation creates a gap where everyone is waiting for a hero who never arrives.
Meanwhile the woman keeps hitting people.
Meanwhile the woman keeps hitting people. And the security guards, when they do arrive, refuse to act. This is the part I keep coming back to. Daniel said someone summoned security to remove her from the train and they refused. They eventually detained her, but only after some period of refusal. What was that?
Could be training. Could be policy. Could be fear of lawsuits.
I think it's probably all three. Private security in Israel operates in a weird space. They're uniformed and they carry authority, but their actual legal powers are limited. They're not police. They can perform a citizen's arrest like anyone else, but their employer almost certainly has policies about physical intervention. Policies written by lawyers who are more worried about liability than about passenger safety.
The company doesn't want to get sued if a guard hurts someone during a detention.
Or if the guard gets hurt and files a worker's compensation claim. Or if the mentally ill woman's family sues for excessive force. The incentives all point toward doing nothing. Wait for police. De-escalate verbally. Don't put hands on anyone unless it's absolutely unavoidable.
Which explains the initial refusal. The guards weren't necessarily cowards or incompetent. They may have been following their training to the letter.
And that's a systemic failure, not an individual one. If your security personnel are trained to stand back during a violent assault, what exactly are they securing?
The company's liability exposure, apparently.
That's bleak. And accurate.
So we've got this situation where the law says you can intervene, the culture says someone else will, and the uniformed personnel are incentivized not to. It's almost designed to produce exactly what Daniel witnessed.
The question then becomes, what should a person actually do? Daniel asked about that good second when you don't know what's permitted. I think the answer has to start with a simple principle. The law is more permissive than most people think. You are allowed to use reasonable force to stop an assault in progress. You are allowed to detain someone committing a crime. The gap isn't legal, it's psychological and practical.
The psychological part is huge. Your brain in that moment is running a risk assessment it was never trained for. Threat level, your own safety, the safety of people with you, the legality of what you're about to do, the social consequences of being the person who jumps in. All of that in under a second.
And most people have never practiced this decision. They've never thought about it until the moment it's happening. That's why soldiers and police officers train scenarios repeatedly. Not because the techniques are complicated, but because you need the decision to already be made before the situation arrives. If you're figuring it out from scratch while someone is screaming, you're going to freeze.
The freeze is the default.
The freeze is the default for almost everyone. And that's not a moral failing. It's neurology. Your amygdala hijacks your prefrontal cortex and you're running on instinct. The instinct is usually freeze or flee. Fight requires training or a very specific personality.
So the practical advice for someone who wants to be capable of acting is to think about it beforehand. Run the scenarios in your head. Know what the law permits. Make the decision in advance that if you see someone being attacked, you will at minimum call for help and at maximum physically intervene if it's safe to do so.
And the minimum is underrated. Daniel said someone summoned security. That's an intervention. It's not dramatic, but it's real. Yelling for help, calling the police, hitting an emergency button, filming the incident. These are all actions that change the situation. The bystander effect research shows that the single most effective thing you can do is break the silence. Point at a specific person and tell them to call the police. Don't say someone call the police. Say you, in the blue shirt, call the police now.
That specificity cuts through the diffusion of responsibility.
It does. It assigns the task to a person who now can't pretend someone else will do it. It's a small thing but it's remarkably effective.
Daniel's train incident had a particular wrinkle though. The woman seemed mentally ill. Does that change the intervention calculus?
In the moment, no. The threat is the threat. But it does change how you approach the intervention if you have time to think. Someone in a psychotic episode may not respond to verbal commands the way a rational attacker would. They may be stronger than expected. They may not feel pain the same way. Physical restraint becomes more complicated.
And the aftermath is different. You mentioned criminal liability. If you perform a citizen's arrest on someone who is clearly mentally ill, the police may not process them the same way. They might be taken to a hospital rather than a holding cell. Your arrest might be questioned more carefully. The scrutiny might be higher.
That's the practical reality. The law doesn't formally distinguish between arresting a rational criminal and arresting someone in a psychotic episode. Both are legally permissible if a crime is in progress. But the system treats them differently once the immediate situation is resolved. And a bystander who used force against a mentally ill person may face more questions, even if the force was justified.
Which circles back to the guards' hesitation. If they've been trained that physical intervention is a last resort, and they see a mentally ill woman rather than a terrorist or a criminal, their risk calculus shifts even further toward inaction.
The mental illness becomes a complicating factor that makes everyone less willing to act, even when the threat is real. And the passengers who were hit, the ones giving statements afterward, they absorbed real harm while everyone was trying to figure out the right thing to do.
There's a moral dimension here that the law doesn't fully capture. In a society like Israel, there's a strong cultural value around collective responsibility. People feel they should act. The social pressure is real. But so is the fear of legal repercussions, and the fear of making things worse, and the simple human fear of physical danger.
The tension between those things is what Daniel was feeling. He wanted to help. He also needed to protect Ezra. He's asking afterward what the right thing would have been. And I think the honest answer is that there's no clean resolution. You do what you can in the circumstances you're in. The law gives you tools. Your judgment tells you when and how to use them. And afterward you live with the choices you made.
The law is a floor, not a ceiling. It tells you what you're permitted to do, not what you must do. Israel doesn't have a duty to rescue law that compels bystanders to intervene. You're not legally required to be a hero.
Some jurisdictions do have that. France, for example, has a duty to rescue that can result in criminal penalties for failing to assist a person in danger. Israel doesn't. The obligation here is moral, not legal.
Which I think is the right approach. Mandating heroism creates its own problems.
It does. You end up prosecuting people for freezing, which is a normal human response. That's not justice.
Let's talk about what happened after the guards finally detained her. Daniel said they were taking statements from the victims, not from the attacker. That's interesting.
It tells you they're treating it as a criminal incident with victims and a perpetrator, even if the perpetrator is mentally ill. The statements are for a potential criminal case. And the fact that they're taking statements from the people who were hit suggests they do consider an assault to have occurred.
Which validates the idea that a citizen's arrest would have been legally justified. The guards, once they acted, treated it as a crime.
Right. The legal framework was there the whole time. What was missing was the will to use it immediately.
Hilbert, you've been quiet on this. You have some experience with this exact situation, don't you?
Hilbert: Worked security on the Tel Aviv to Jerusalem line. Late nineties. Before they privatized everything.
I didn't know that. What was the training like for a situation like this?
Hilbert: Don't touch anyone. That was the training. The company had a list of things we were supposed to do. Call the police. Clear the area. Take statements. Physically intervene was not on the list. They said it was for our safety. It was for their liability.
So the guards refusing to intervene might not be incompetence.
Hilbert: Might be policy. We had a guy on the train once, having some kind of episode. Screaming, throwing things. We cleared the car and waited for police. Took twelve minutes. He broke three windows in that time. Company still said we did the right thing.
Twelve minutes is a long time when someone's breaking windows.
Hilbert: Felt longer. There was a passenger who wanted to handle it himself. Big guy. We had to tell him to stand down. He was legally in the right, probably. We were enforcing company policy.
Did you ever have to make a citizen's arrest?
Hilbert: Once. Drunk passenger assaulted a conductor. I stepped in, restrained him, held him until police arrived. Filed my report. Got a letter of reprimand the next week.
For what?
Hilbert: Exceeding authorized use of force. The restraint left bruises on the guy's wrists. He threatened to sue. The company settled and I got written up. Even though the police said the arrest was lawful.
So the law said you were right and your employer punished you anyway.
Hilbert: The law doesn't pay your salary. I kept the letter. Still have it somewhere.
That reframes the whole train incident. The guards who refused might have been following orders. Or they might have known someone who got reprimanded for doing the right thing.
Hilbert: You learn fast. The system tells you one thing and the company tells you another. Most people go with the company.
Because the company signs the checks.
Hilbert: And the company has lawyers.
So we've got a situation where the law gives citizens broad powers to intervene, but the institutions that employ uniformed personnel actively discourage using those powers. And the public absorbs that message. If the guards won't act, why should I?
Hilbert: That's about the size of it.
Did you ever see a passenger intervene successfully?
Hilbert: Few times. Usually off-duty soldiers. They don't worry about company policy. But they also don't always know what they're doing. Saw a guy put someone in a chokehold once. Effective, but dangerous. Could have killed him. He was lucky.
The training gap. Soldiers are trained to neutralize threats, not to perform citizen's arrests. The techniques aren't the same.
Hilbert: Different tool for a different job. A citizen's arrest is supposed to be restraint and detain, not neutralize. Most people don't know the difference.
So we're back to the same problem. The people with training have the wrong training, and the people with the right legal powers are discouraged from using them.
Hilbert: The law says you can intervene. The system punishes you for it. So people wait for someone else to be the hero.
That's... really the whole thing, isn't it.
Hilbert: I had a colleague who handled it differently. Older guy, been doing security for twenty years. When something kicked off, he'd walk over and talk to the person. Quiet voice. Calm. Nine times out of ten he could de-escalate without touching anyone. He said most people having an episode just need someone to acknowledge they exist.
That's a skill though. Not everyone can do that.
Hilbert: No. He was good at it. Company never trained for it. He just knew how.
De-escalation training should be standard for transit security. It's not a physical intervention, so it doesn't trigger the liability concerns. And it actually works in a lot of cases.
Hilbert: Cheaper than settling lawsuits too. But nobody asked me.
Hilbert, you said you still have the reprimand letter. Where is it?
Hilbert: In a box. With some other things from that job.
What else is in the box?
Hilbert: My badge. A timetable from ninety-eight. Four unused complaint forms.
Four unused complaint forms.
Hilbert: We were supposed to hand them out to passengers who wanted to file complaints. Never gave out a single one.
Because nobody complained or because you didn't offer?
Hilbert: Both.
That's a whole separate episode. The accountability structure that doesn't exist.
The complaint forms that sit in a box for thirty years.
Hilbert: They're probably still good. Paper doesn't expire.
I don't think that's how complaint forms work.
Hilbert: They'd take them.
I think they'd take them.
All right. So where does this leave us? We've got a legal framework that's actually quite permissive. We've got a cultural expectation that someone more qualified will step in. We've got institutional incentives that push security personnel toward inaction. And we've got ordinary people who freeze because they've never thought about this before and they're scared of getting sued or hurt or both.
The misconception I want to name is the idea that waiting for an off-duty police officer is the safest choice. It's not. The law doesn't require you to wait. And in Israel specifically, the expectation that someone with training is nearby can lead to dangerous delays while an assault continues. The law gives you the tools. The question is whether you've prepared yourself to use them.
The other misconception is that a citizen's arrest is a simple, risk-free legal tool. It's not. It carries real legal exposure if the arrest is deemed unlawful or excessive force is used. That doesn't mean you shouldn't do it. It means you should understand what you're doing before the moment arrives.
Daniel's question was how to react in that split second. I think the answer has to be that you can't figure it out in the split second. You have to figure it out beforehand. Know what the law permits. Decide what you're willing to do. Accept that there's no perfect option. And then, if the moment comes, act on the decision you already made.
And if the decision is to protect your child first, that's a valid decision. That's not a failure. That's triage.
The open question I keep coming back to is whether there should be clearer legal protections for bystanders who intervene in good faith. Right now, the law says you can act, but it doesn't shield you from the consequences if your judgment is questioned afterward. Some countries have Good Samaritan laws that provide immunity for reasonable interventions. Israel's framework is more ambiguous.
And as public transit expands and these situations become more common, that ambiguity is going to cause more problems. The population is growing. The trains are getting more crowded. Mental health crises in public spaces aren't going away. We need a clearer social contract about who is expected to act and what protections they have when they do.
The next time you're on a train and someone snaps, what will you do? If you haven't thought about it, the answer is probably nothing. And that's worth sitting with.
Thanks to Hilbert Flumingtop for producing, and for the box of unused complaint forms that I now desperately want to see.
This has been My Weird Prompts. If you want to send us your own questions, email the show at show at my weird prompts dot com.
We'll be back soon.
