Daniel sent us this one. He was walking down the street, came across a man lying in the middle of the road. Not on a sidewalk, not in a doorway — middle of the street. A woman was checking on him, someone from a nearby shop said he's always like that, another bystander eventually roused him and led him away. The whole thing resolved in about a minute. But Daniel's question is: what if nobody else had been there? What's the actual procedure — who do you call first, and what's the right thing to do when you find someone down in a public space and you can't tell if they're unconscious, asleep, or something worse?
This is one of those situations where most people's instincts are actually correct, but the reasoning behind those instincts is more interesting than you'd think. The short answer is: you call emergency services. In Israel, that's MDA — Magen David Adom — at one zero one. In the US it's nine one one, in the UK nine nine nine, across most of Europe one one two. The number itself matters less than the decision tree that gets you there.
The decision tree starts with "is this person responsive," which sounds simple until you're standing there with your phone in your hand and a stranger at your feet.
The first step in any emergency response protocol is to check responsiveness. You approach, you speak loudly — "are you okay, can you hear me" — and if there's no response, you tap their shoulder or gently shake them. What you don't do is assume they're just sleeping. Because here's the thing that emergency medicine teaches you: an unresponsive person in a public space is a medical emergency until proven otherwise.
"Until proven otherwise" doing a lot of work there.
It really is. The differential diagnosis for someone down on the street includes hypoglycemia, opioid overdose, stroke, seizure, traumatic brain injury, hypothermia, severe infection — and yes, also intoxication and homelessness and exhaustion. But you as a bystander cannot distinguish between those in the moment. Paramedics with years of experience will tell you they've been fooled more times than they can count.
The shopkeeper saying "he's always like that" — which is probably true, by the way, the guy probably does this regularly — that's not actually useful information in the moment.
It's context, but it's not diagnostic. "He's always like that" might mean he has chronic alcohol dependency and passes out in the street regularly. It doesn't rule out that this time he also has a head injury from falling, or that his blood sugar has crashed, or that he's aspirated vomit and is quietly suffocating. The fact that someone is frequently in a dangerous situation doesn't make the situation less dangerous — it arguably makes it more dangerous, because the cumulative risk keeps accumulating.
That's a grim actuarial observation.
Emergency medicine is full of them. There's actually a well-documented phenomenon called "the frequent flyer effect" in EMS — paramedics and emergency departments develop mental shortcuts about people they see repeatedly, and those shortcuts can lead to missed diagnoses. Someone who's picked up intoxicated ten times still has the same probability of having a coincident medical emergency on the eleventh pickup as anyone else. But the documentation shows they get assessed less thoroughly.
Let's walk through the scenario Daniel described, but strip away the other bystanders. It's just you, one person, and a man lying in the middle of the street. Step by step — what do you actually do?
Step one: make sure you're not about to become a second casualty. If he's in the middle of the street, that means traffic. Before you approach, you need to make sure you're visible and that oncoming vehicles can see what's happening. If you have a car, put on your hazards and position it to shield the scene. If you don't, and you're in an area with enough traffic that it's dangerous, that itself might be a reason to call emergency services immediately — because you need help securing the scene before you can safely assess the person.
If it's a quiet residential street at two in the afternoon?
Then you approach directly. Step two: verbal stimulus. "Sir, are you okay? Can you hear me?Not shouting in a way that's aggressive, but loud enough that someone who's merely dozing would startle awake. Step three: if no response, physical stimulus. A firm tap on the shoulder, or a sternal rub — that's rubbing your knuckles firmly on the breastbone, which is painful enough to rouse someone who's deeply asleep but not truly unconscious. Don't do that lightly, but it's a standard assessment technique.
"Sternal rub" sounds like a medieval torture method that somehow survived into modern medicine.
It basically is. But it works. If the person responds — they open their eyes, they mumble, they push you away — you've now established that they're conscious and breathing. That changes the situation considerably. You can now ask them questions: what's your name, do you know where you are, do you need medical help. If they're coherent and decline help, in most jurisdictions you cannot force it on them. Competent adults have the right to make decisions you think are terrible.
Including lying in the middle of the street.
Including that, yes. Though at that point you might shift to a welfare check approach — "I'm concerned about you being in the road, can I help you move to the sidewalk at least" — rather than a medical emergency approach. But if the person is unresponsive, or if they're responsive but confused, unable to answer basic questions, or showing signs of distress, you're now in emergency territory.
Then you call.
And here's a detail that matters: when you call, the dispatcher is going to ask you a specific set of questions in a specific order. This isn't bureaucracy — it's a structured assessment protocol. They'll ask for the location first, because if the call drops, they can at least send someone. Then they'll ask what happened, then about the person's condition — are they conscious, are they breathing normally. The dispatcher is trained to triage your answers and decide whether to send an ambulance, police, or both.
What about the photo Daniel mentioned? He took a photo to capture the geolocation. Is that actually useful, or is it just what we do now because we have phones?
It's not standard protocol, but it's not a bad instinct either. The dispatcher will want a precise location, and if you're in an unfamiliar area, a photo with GPS metadata can help you give cross streets or a landmark. But the far more important thing is to stay on the line and follow the dispatcher's instructions. They may walk you through checking breathing, putting the person in the recovery position if they're breathing but unresponsive, or starting CPR if they're not breathing at all.
Let's talk about the recovery position, because I've seen it described in first aid manuals but I'm not sure I could execute it under pressure. What's the actual maneuver?
The recovery position is designed to keep an unconscious person's airway clear. When someone's unconscious, the muscles that keep their tongue from falling back and blocking their throat relax. They also can't protect their airway if they vomit. The recovery position uses gravity to solve both problems — the person is on their side, head tilted slightly back, which keeps the tongue forward and allows fluids to drain out of the mouth instead of into the lungs.
You do this yourself, as a bystander, before the ambulance arrives?
If the person is breathing normally and you don't suspect a spinal injury, yes. The dispatcher will walk you through it. The basic sequence: you kneel beside them, straighten their legs, place the arm nearest to you at a right angle to their body with the palm facing up, bring the far arm across their chest and hold the back of their hand against their cheek, then pull up the far knee, and roll them toward you by pulling on that bent knee while keeping their hand pressed to their cheek. Then you adjust their head to keep the airway open.
That's a lot of steps to remember while someone's unconscious at your feet.
It is, which is why the dispatcher stays on the line and talks you through it in real time. You don't have to memorize this — you just have to be willing to follow instructions. And that's actually the most important thing to understand about this whole scenario: the system is designed so that the bystander doesn't need to be an expert. The expertise is at the other end of the phone.
That's genuinely reassuring. But it raises a question about the specific scenario Daniel described — someone who appears to be homeless, someone the locals recognize. Does the calculus change when the person is known in the area?
Not in the first few minutes, and here's why. The assessment protocol doesn't change based on housing status. Unconscious is unconscious. But there's a second layer here that's worth addressing, which is what happens after the immediate medical question is resolved. If the person is conscious, alert, and refusing transport, but you're still concerned about their welfare — they're clearly vulnerable, they're in an unsafe situation — many jurisdictions have a non-emergency option.
What is that in practice?
It varies enormously by city, which is part of the problem. In Jerusalem, where Daniel was, there's the municipal welfare hotline — one zero six — which can dispatch social workers or coordinate with NGOs that do street outreach. In many US cities, there's two one one for social services. But the landscape is fragmented, and there's been a major shift in the last few years toward what are called "alternative response" or "community response" programs.
Alternative to police, you mean.
The idea is that when someone is in crisis but it's not a criminal matter and not a medical emergency, sending armed police officers isn't the right tool for the job. Instead, these programs send teams of mental health professionals, social workers, or peer support workers — people trained in de-escalation and connection to services. The most famous example is CAHOOTS in Eugene, Oregon, which has been doing this since nineteen eighty-nine, but there are now dozens of programs across North America and a growing number in Europe and Israel.
CAHOOTS — that's the one where they send a medic and a crisis worker in a van.
And the data on these programs is encouraging. A study out of Denver found that their STAR program — Support Team Assisted Response — handled nearly three thousand calls in its first two years with zero arrests and zero injuries to responders or the public. The calls that would have gone to police instead went to a mental health clinician and a paramedic, and the outcomes were better across the board.
Zero arrests and zero injuries is the kind of statistic that makes you wonder why this isn't the default everywhere.
Partly because it requires political will and sustained funding, partly because the infrastructure of emergency response is incredibly sticky — nine one one systems were built around police, fire, and EMS, and adding a fourth branch is technically and bureaucratically complex. But the momentum is real. I saw a report from the Vera Institute that tracked a forty percent increase in alternative response programs between twenty twenty-three and twenty twenty-five across the US.
If Daniel had called one zero one in Jerusalem, what would he have gotten?
MDA would have dispatched an ambulance if it was reported as a medical emergency — someone unresponsive in the road would certainly qualify. But if the person woke up, was clearly just sleeping, and refused assistance, MDA might assess and clear the scene without transport. At that point, the question becomes: is there someone else to call who can address the underlying vulnerability? And that's where the system gets thin.
That's the gap, isn't it? Between "not a medical emergency" and "still a human being lying in the street.
That's exactly the gap. And it's the gap that alternative response programs are trying to fill. In Jerusalem specifically, there are organizations like Yad Sarah and various municipal welfare teams that do street outreach, but they're not integrated into the emergency dispatch system the way CAHOOTS is in Eugene. You'd have to know to call them separately, and most people don't.
Which brings us back to the individual bystander making a judgment call in about thirty seconds. What's the actual moral calculus here?
I think the moral calculus is simpler than people make it. If someone is unresponsive or in distress, you call for emergency help. You're not diagnosing, you're not adjudicating, you're not deciding whether they "deserve" an ambulance. You're reporting what you see and letting the system triage. The potential harm of calling when it wasn't necessary is a wasted ambulance trip. The potential harm of not calling when it was necessary is death or permanent disability. Those are not symmetrical risks.
Yet people hesitate. He got out his phone for geolocation but didn't dial. And he had the presence of mind to reflect on it afterward and ask what the right procedure is. Most people just keep walking.
The bystander effect is real, and it's been replicated in dozens of studies since the original Kitty Genovese research in the nineteen sixties. But what's interesting is that the bystander effect almost completely disappears when someone knows what to do. The hesitation isn't usually callousness — it's uncertainty. People don't act because they don't know the right action, and they're afraid of doing the wrong thing.
Knowledge is the antidote to inaction.
It really is. There was a study in twenty twenty-three — published in the Journal of Emergency Medical Services, I believe — that looked at bystander intervention rates before and after a public awareness campaign about overdose response. The campaign basically taught people three things: recognize the signs of an overdose, call nine one one, and if you have naloxone, administer it. Intervention rates went up by over sixty percent in the areas where the campaign ran.
Three pieces of information, sixty percent behavior change. That's a remarkable return on investment.
It applies far beyond overdoses. The basic framework — check for responsiveness, call for help, stay with the person — covers the vast majority of street-level emergencies. You don't need a medical degree. You need to know three steps and be willing to take them.
Let me push on something, though. The scenario Daniel described — middle of the street, shopkeeper says he's always like that, another bystander eventually leads him away — there's a version of this where calling an ambulance would have been overkill, and everyone on the scene knew it except the hypothetical outsider with the phone.
I understand the instinct, and I think it's worth distinguishing between two situations. Situation one: the person is responsive, coherent, and declining help. In that case, calling an ambulance against their will is not appropriate, and may actually cause harm — it can escalate the situation, it can involve police, it can create a confrontation that nobody wanted. Situation two: the person is unresponsive or clearly impaired, and you can't get a coherent refusal. In that case, implied consent applies — the legal and ethical assumption is that a reasonable person would want emergency medical help if they were unable to make their own decisions.
Implied consent — that's the legal doctrine that protects the bystander from being sued for calling an ambulance for someone who didn't want one?
In virtually every jurisdiction, if you act in good faith to summon emergency help for someone who appears to need it, you're protected by Good Samaritan laws. Those laws vary by country — in Israel, the Good Samaritan law explicitly protects bystanders who render aid or summon emergency services, as long as they're not grossly negligent. In the US, all fifty states have some form of Good Samaritan protection, though the specifics differ.
The legal risk of calling is basically zero, and the moral risk of not calling is catastrophic. That's a pretty clear asymmetry.
But I want to complicate it slightly, because I think the hesitation people feel isn't just about legal risk. There's a social dimension. Nobody wants to be the person who overreacts, who causes a scene, who — and I think this is the real fear — imposes a medicalized response on someone whose actual problem is poverty.
That's the real tension, isn't it? Calling an ambulance for someone who's homeless and sleeping isn't just "wasting resources" — it's potentially subjecting them to a system they may have very good reasons to distrust.
And this is where the alternative response movement gets its moral force. The argument isn't "don't call anyone" — it's "call the right someone." If your city has a non-police crisis response team, and the situation is clearly social rather than medical, that's the better call. If your city has a street outreach hotline, that's the better call. The problem is that most cities still don't have these options, or they exist but aren't widely known.
The practical advice for a listener is: know what's available in your city before you need it.
Spend ten minutes looking up your local emergency numbers — not just the universal one, but the specific ones. Know the difference between the medical emergency line, the police non-emergency line, and any social services hotline that operates in your area. Put them in your phone. The moment you're standing over someone in the street is not the moment to start researching.
That's a very Herman Poppleberry approach to moral philosophy — do your homework in advance so your conscience has a user manual.
I'll take that as a compliment.
It was backhanded and you know it.
I choose to receive it warmly.
Let's go back to the specific details of Daniel's scene, because there's something else in there that I think is worth unpacking. The shopkeeper came out and said "it's okay, he's always like that." And that statement — which was probably meant to reassure — actually makes the situation more confusing for the bystander. Because now you have local knowledge telling you this is normal, but you're looking at something that is manifestly not normal.
This is a classic information asymmetry problem. The shopkeeper has context you don't have — they've seen this person before, they know the pattern, they've developed a mental model of what's happening. But as we discussed earlier, that mental model might be wrong, and even if it's right about the pattern, it doesn't rule out an acute problem this time around.
There's a social dynamic too. The shopkeeper is asserting local authority — "I know what's going on here, you're new, stand down." It takes a certain amount of social confidence to say "I appreciate that, but I'm still going to check on him.
That's a real barrier, and it's not trivial. But here's a reframe that might help: you're not contradicting the shopkeeper. You're adding information to the situation. "Thanks, I'm glad he's known here — I just want to make sure he's responsive." That's collaborative rather than confrontational. You're acknowledging their expertise while still doing your due diligence.
"Collaborative rather than confrontational" — that's the Jerusalem approach to basically everything.
It's not a bad default. But to be clear, if the person is unresponsive, you don't stop to have a debate with the shopkeeper. The social niceties are secondary to the medical reality.
What about the other bystander in Daniel's story — the one who roused the person and led them away, saying they'd find somewhere for them to sleep? That's a different kind of intervention. Is that good?
It depends entirely on who that bystander is and what their relationship is to the person. If it's someone who knows them, who has a genuine connection, and who's actually going to help them get to a safe place, that's a positive outcome. If it's a stranger who's just moving the "problem" out of sight, that's not help — that's displacement.
You can't tell the difference in the moment.
You often can't. That's another reason why staying with the person until professional help arrives — if that's the path you're on — is important. Handing someone off to a stranger who claims they'll help is not the same as handing them off to a paramedic.
We've covered the immediate response — check responsiveness, call if unresponsive, stay with the person, follow dispatcher instructions. What about the aftermath? Daniel's prompt is really about the moment of decision, but I think there's a second question lurking in there: what do you do with the experience afterward?
That's a thoughtful reframe. And I think it matters, because these encounters can be distressing. You see someone in a situation that's not right — a human being lying in the road — and even if it resolves without tragedy, it stays with you. The question of "did I do the right thing" can loop for days.
The moral hangover.
And I think part of the answer is exactly what Daniel is doing — asking what the right procedure is, so that next time, the uncertainty is reduced. But there's also something to be said for debriefing. If you were with someone else, talk about it afterward. If you called emergency services, it's okay to call the non-emergency line later and ask what happened — you may not get an answer due to privacy laws, but sometimes you can.
If you're alone and you just walked away after the situation resolved?
Then I think it's worth reflecting on whether there's a gap between what you did and what you wish you'd done, and closing that gap with preparation. Look up the numbers, learn the basic assessment steps, maybe take a first aid course if you're so inclined. Not because you failed — you didn't — but because these situations are inherently disorienting, and having a mental script reduces the disorientation.
A mental script. That's actually a useful way to think about it. Most of us have a script for "someone says hello" or "someone cuts in line," but we don't have a script for "human being horizontal in the roadway." And without a script, we freeze.
That's exactly what the research on bystander intervention shows. People don't fail to act because they're bad people — they fail to act because they don't have a response pattern for the situation, and in the absence of a pattern, the brain defaults to inaction. It's not a moral failure, it's a cognitive one.
Give us the script. If you had to boil it down to something someone could remember six months from now, standing on a street corner with their heart rate climbing.
One: ensure your own safety first — don't become a second casualty. Two: check if the person is responsive — speak loudly, tap their shoulder. Three: if unresponsive or in distress, call emergency services immediately and follow dispatcher instructions. Four: stay with the person until help arrives. That's it. That's the whole script.
If they're responsive and decline help?
Respect their autonomy. Offer to help them move somewhere safer if they're in danger. If you're still concerned, you can call a non-emergency welfare line if one exists in your area. But you cannot force help on a competent adult who refuses it.
What about the photo Daniel took? Is that something you'd include in the script?
I wouldn't make it a formal step, but it's not a bad supplementary action. If you're in an unfamiliar area and need to give a precise location to the dispatcher, a photo with GPS data can help. But don't let taking a photo delay calling — call first, then if the dispatcher needs help with the location, use whatever tools you have.
There's also a privacy dimension to photographing someone in a vulnerable state. Is that something to worry about?
It's a legitimate concern. If you're taking a photo specifically for location reference — a wide shot of the intersection, not a close-up of the person — that's defensible. If you're taking photos of the person themselves, especially if they're unconscious or in distress, that's exploitative and potentially illegal depending on the jurisdiction. The rule of thumb: the photo is for helping, not for sharing.
"The photo is for helping, not for sharing" — that's a good boundary. Let's zoom out a bit. The scenario Daniel described — homeless person, public space, bystanders unsure how to respond — this is playing out thousands of times a day in cities around the world. Is the system actually set up to handle this well?
The system is fractured. Emergency medical services are designed for acute medical emergencies — heart attacks, strokes, traumatic injuries. Police are designed for law enforcement. Neither is well-designed for the gray zone of chronic homelessness, mental illness, and substance use that presents as "person down in public." And the social services that are designed for those things are chronically underfunded and siloed from the emergency response system.
The individual bystander is being asked to navigate a system that the system itself hasn't figured out how to navigate.
And that's both unfair and unavoidable. It's unfair because we're asking ordinary people to make complex triage decisions with incomplete information. It's unavoidable because until the system is reformed, the person on the street is the first responder whether they want to be or not.
That's a heavy thing to lay on someone just walking to get lunch.
But I think the weight is lighter if you're prepared. And preparation doesn't mean becoming an expert — it means knowing the four steps, knowing the numbers, and having made the decision in advance that you're the kind of person who stops.
"The kind of person who stops" — that's a choice, not a personality trait.
And it's a choice you can make in advance, so that when the moment comes, you're not deciding from scratch. You're just executing a decision you already made.
That's a framing I haven't heard before. Usually we talk about these situations as tests of character, revealed in the moment. You're describing it as a policy decision you make about yourself, ahead of time.
Because character is just the accumulation of decisions you've already made. If you decide now that you'll stop, you'll probably stop. If you leave it to the moment, the moment has a lot of ways to talk you out of it.
Let's get concrete for the listener who's been following along and wants to be prepared. What numbers should they have in their phone, and where do they find the ones specific to their area?
Universal emergency numbers first: one one two across the European Union, nine nine nine in the UK, nine one one in the US and Canada, zero zero zero in Australia, one zero one in Israel for MDA, one zero zero for police, one zero two for fire. But beyond the universal numbers, I'd recommend looking up three things specific to your city: one, the non-emergency police line — for situations that need official attention but aren't immediate threats. Two, the mobile crisis team or mental health crisis line if your area has one. Three, any street outreach or homeless services hotline. These numbers are usually on your city or county website, or you can find them through two one one in the US.
If your city doesn't have those specialized services?
Then the emergency number is your only option, and you should use it without guilt. The fact that the system is inadequate is not your fault, and it doesn't relieve you of the responsibility to respond to the person in front of you. Call, explain clearly what you're seeing, and let the dispatcher make the triage decision. That's their job.
What about the argument that calling an ambulance for a homeless person who's just sleeping is an inappropriate use of emergency resources — that it takes an ambulance away from someone having a heart attack?
That argument assumes you know the person is "just sleeping," which — as we've established — you cannot know from observation alone. It also assumes that EMS systems don't have triage protocols, when in fact they do. When you call nine one one, the dispatcher assigns a priority level to the call. An unresponsive person in a roadway will get a high-priority response, as it should. If the person wakes up and is fine, the ambulance may be diverted to a higher-priority call en route, or may arrive and clear quickly. The system is designed to handle uncertainty — that's literally why triage exists.
The "wasting resources" concern is mostly a post-hoc rationalization for not wanting to get involved.
I think it often is, yes. Not always — there are people who worry about system strain, and that's a legitimate concern in abstract. But in the specific moment, with a specific person at your feet, the concern about hypothetical resource allocation should not override the duty to the actual person.
Let's talk about one more dimension before we wrap. Daniel's scenario took place in Jerusalem. Does the fact that it's a dense urban environment with a particular social fabric change anything about the response?
Urban density actually makes intervention more likely, statistically — the bystander effect is weaker in cities where people are accustomed to proximity and where help is physically closer. But Jerusalem has some specific dynamics. There's a strong culture of mutual aid in many neighborhoods, particularly in religious communities. There's also a complex overlay of political and social tensions that can make people hesitant to intervene across community lines.
The street Daniel was on — there are areas in Jerusalem where someone lying in the road could be a medical situation, a social situation, or a security situation, and the distinction matters enormously.
That's true, and it's a layer of complexity that most cities don't have to the same degree. But for the individual bystander, the protocol doesn't change. You assess the person, not the geopolitical context. If they're unresponsive, you call. The professionals who respond will be better equipped to handle whatever additional dimensions exist.
I think that's a good place to land. The core message here is surprisingly simple: know what to do, so that when you encounter the unexpected, you're not starting from zero. The four-step script, the numbers in your phone, the advance decision to be someone who stops.
The recognition that the system isn't perfect, but perfect isn't the standard. The standard is: did you do what a reasonable person would do with the information you had? If the answer is yes, you can sleep at night.
Even if the person you helped is sleeping in the street.
Now: Hilbert's daily fun fact.
Hilbert: In the eighteen eighties, the Ottoman Empire commissioned a series of detailed maps of Eritrea for administrative purposes, but the cartographers — trained in Islamic geometric traditions — kept inserting mathematically perfect but geographically nonsensical hexagonal grids into the desert regions, apparently because they found the blank spaces aesthetically offensive and believed ordered emptiness was a kind of lie.
The cartographic equivalent of filling every silence with small talk.
Ordered emptiness as a lie. I need to sit with that.
Thanks to Hilbert Flumingtop for that. This has been My Weird Prompts. You can find every episode at myweirdprompts dot com, or wherever fine podcasts are distributed. If you've got a situation you don't know how to react to, send it our way — apparently we have opinions.
We always have opinions. Whether they're correct is a separate question, but we'll leave that to the listeners.
Until next time.
