Daniel's prompt this week picks up a thread we left dangling. Last episode, when we talked about what actually happens when you get shocked, we mentioned Charles Dalziel and his let-go threshold research. Daniel wants to sit with that anecdote a little longer.
Not the experiment design, not the fail-safes. The people. The volunteers who let researchers run current through them until their muscles locked up and they physically could not release the conductor, all to generate data that would quietly protect billions of strangers. Daniel calls these acts scientific heroism in small. And he's asking us to look at the other examples scattered through the scientific annals, the ones that would never pass an ethics committee today but whose data we still live inside.
The uncomfortable part is that both things are true at once. The research was ethically indefensible by modern standards. And it saved an enormous number of lives.
So today we're going to look at the people who quite literally put their bodies on the line, and ask what we owe them, and what their stories tell us about how science actually progresses.
Let's start with what Dalziel actually did, because the mechanics matter.
Charles Dalziel was an electrical engineering professor at Berkeley. In the forties and fifties, and into the sixties, he ran a series of experiments on human volunteers to establish the let-go threshold. That's the maximum current a person can tolerate while still retaining the muscular control to voluntarily release a conductor.
And the numbers he found are still the backbone of electrical safety standards. Approximately six milliamps for adult men, five milliamps for adult women. Below that, you can let go. Above it, the current causes involuntary muscle contraction. Your hand clamps down on the conductor, and you cannot open it.
Which is the part that makes the experiment itself so strange to read about. The entire study design requires pushing a human being to the exact point where their body stops obeying them.
The setup was deliberately incremental. Dalziel used a variable transformer, and he would raise the current in small steps. The volunteer held a conductor — a wire or a brass rod — and at each step they'd be asked to release it. At some point, the current would cause the flexor muscles in the forearm to contract harder than the extensor muscles could counter. The hand closes. The subject reports they can't let go.
And then what? That's the detail Daniel was curious about.
There was a forced release mechanism. A spring-loaded device that would physically yank the subject's hand off the conductor when they failed to release voluntarily. Dalziel describes it in his papers. The subject's hand was positioned so that when the current reached the point of involuntary contraction, the experimenter could trip the mechanism and break the circuit immediately.
So the volunteers weren't simply abandoned to the current. There was a plan for the moment when their bodies stopped cooperating. But the fundamental design — deliberately inducing involuntary muscle contraction in a human being — is the ethically fraught core.
And here's what makes it complicated. Before Dalziel, electrical safety standards were essentially guesswork. There was no systematic data on what current levels were actually dangerous to humans. Appliance designers were working blind. The difference between a safe design and a lethal one was a matter of intuition and anecdote.
Dalziel replaced the guesswork with numbers. The perception threshold, around one milliamp, where you first feel the tingle. The let-go threshold. The onset of respiratory difficulty at higher currents. The ventricular fibrillation zone. These numbers became the foundation for UL standards, IEC guidelines, national electrical codes.
Every GFCI outlet in your house, every appliance standard, every electrical code on the planet, is built on the assumption that we know exactly where the danger begins. Dalziel's volunteers gave us that knowledge.
And the uncomfortable math is this. How many lives have been saved by having precise let-go threshold data embedded in every circuit breaker and every safety standard? Billions of person-hours of exposure, protected invisibly, by research that seems barbaric by today's standards.
Context matters here. Dalziel's work happened in a period when human subjects research was far less regulated. The Nuremberg Code existed — it was established in nineteen forty-seven, after the Nazi medical experiments — but institutional review boards didn't become standard until the nineteen seventies. There was a genuine belief, widespread at the time, that certain questions could only be answered by direct human exposure.
That's not an excuse. It's a description of the landscape. Dalziel wasn't operating in secret. He published in IEEE journals. His work was peer-reviewed, cited, built upon. The scientific community of his era saw this as legitimate research.
And to be fair to Dalziel, he did implement safety mechanisms. The forced-release system was real. The volunteers were informed participants — they knew they were going to be shocked, they knew the current would increase, they knew the endpoint was the loss of voluntary control. The standards of informed consent were different from today, but they weren't nonexistent.
The misconception to avoid is that pre-IRB research was uniformly reckless or unethical. It wasn't. Some of it was careful, methodical, and protective of subjects within the limits of what was understood at the time. And some of it was appalling. The history doesn't sort neatly.
Dalziel's story is remarkable, but it's not unique. Once you start looking, the scientific annals are full of these moments where researchers put their own bodies on the line.
Let's start with John Stapp. Air Force physician, nineteen forties and fifties. He was studying the limits of human tolerance to rapid deceleration, the forces a pilot experiences when a plane crashes or an ejection seat fires.
Stapp didn't just design the experiments. He was his own test subject. He rode a rocket sled — a device on rails that would accelerate to high speed and then slam to a stop — over and over again. The most famous run, in nineteen fifty-four, subjected him to forty-six point two Gs of deceleration. That's forty-six times the force of gravity.
For context, a fighter pilot pulling a hard turn might experience nine Gs. Stapp took forty-six. He broke ribs, he broke his wrist, he suffered retinal hemorrhages that permanently damaged his vision. He called himself the fastest man on earth, and he meant it literally.
The data he gathered established human tolerance limits for deceleration. It directly informed the design of ejection seats, seatbelts, and crash restraints. Before Stapp, nobody knew whether a human could survive the forces involved in an ejection. After Stapp, engineers had numbers to design against.
Stapp survived. He lived to be eighty-nine. But he carried the physical damage for the rest of his life. And he did it voluntarily, repeatedly, knowing each run could kill him.
Then there's the Yellow Fever Commission of nineteen hundred. Walter Reed's team in Cuba was trying to prove that yellow fever was transmitted by mosquitoes. The theory was controversial. The alternative was that it spread through contaminated clothing or bedding — the fomite theory.
To settle it, members of the team deliberately exposed themselves to infected mosquitoes. Jesse Lazear and James Carroll both allowed themselves to be bitten. Carroll survived, but suffered lifelong health damage. Lazear died of yellow fever at age thirty-four.
Lazear's death is the starkest version of this story. He was a physician, he knew the disease was often fatal, and he let an infected mosquito feed on him anyway. The experiment worked — it proved mosquito transmission — but the proof cost him his life.
And that proof made possible the eradication campaigns that followed. Once you know mosquitoes carry the disease, you know to target the mosquitoes. The Panama Canal project, which had been stalled by yellow fever deaths, became feasible. Millions of lives saved. Lazear didn't live to see any of it.
The self-experimentation tradition in medicine is its own whole genre. Werner Forssmann, nineteen twenty-nine. He was a young surgical resident in Germany who believed you could catheterize the heart safely. His superiors said it was impossible and forbade him from trying.
So he did it to himself. Local anesthesia, a cut into his own arm, and he fed a catheter through his vein all the way into his right atrium. Then he walked down to the X-ray department and took the picture that proved it.
He got fired from his hospital. And then, in nineteen fifty-six, he shared the Nobel Prize for that work. The procedure he pioneered — cardiac catheterization — is now one of the most common diagnostic procedures in medicine. Millions performed every year.
Barry Marshall, nineteen eighty-four. He was trying to prove that Helicobacter pylori bacteria caused stomach ulcers. The medical establishment thought ulcers were caused by stress and stomach acid. Bacteria couldn't survive in the stomach, they said.
Marshall drank a culture of Helicobacter pylori. A beaker of it. Within days he developed gastritis — inflammation of the stomach lining — and a biopsy showed the bacteria had colonized his stomach. He then treated himself with antibiotics and recovered.
That self-infection was the turning point. Marshall and his collaborator Robin Warren won the Nobel Prize in two thousand five. The discovery transformed ulcer treatment from a chronic condition managed with acid suppressants to a curable infection.
And then there's Humphry Davy, eighteen hundred. He was experimenting with nitrous oxide, laughing gas. He inhaled it repeatedly, at increasing doses, to document its effects. At one point he nearly killed himself — the dose was too high, he lost consciousness, and he was only saved by the fact that the gas wore off before he stopped breathing entirely.
Davy's self-experiments gave us the first detailed account of nitrous oxide's effects, which eventually led to its use as an anesthetic. But Davy was also, by any reasonable standard, reckless. He was inhaling unknown quantities of a gas with no safety protocol, no monitoring, no plan for what to do if things went wrong.
And that's the taxonomy question Daniel is really asking about. What distinguishes these acts? They share features — the researcher is the subject, the risk is real and known, the data is otherwise unobtainable, and the results have outsized impact. But they differ in crucial ways.
Lazear's death is heroic. He knew the risk was death, he took it anyway, and the knowledge gained was irreplaceable. Marshall's self-infection was a calculated gamble with clear scientific logic — he had a hypothesis, he tested it on himself, and the result was unambiguous.
Davy's nitrous oxide experiments were closer to recklessness. He was exploring, not testing a specific hypothesis, and the safety margins were nonexistent. The fact that it worked out doesn't retroactively make it careful.
So the taxonomy isn't just heroism versus recklessness. It's a spectrum. And the position on the spectrum depends on things like whether the risk was proportionate to the knowledge sought, whether there was a clear hypothesis, whether alternatives existed.
The ethical double-bind is this. We cannot endorse these methods as a model for today. IRBs exist for good reasons. The history of abuses — Tuskegee, the Nazi experiments, the radiation experiments of the Cold War — shows exactly where unregulated research leads.
But we also can't dismiss the results. The let-go threshold data is real. Stapp's tolerance limits are real. The mosquito transmission proof is real. Pretending the knowledge would have emerged anyway, through some cleaner path, is a comforting fiction.
Some of it might have. Animal models, computational simulation, epidemiological studies — these tools exist now and can answer questions that once required human exposure. But in nineteen hundred, there was no animal model for yellow fever transmission that could settle the debate. In nineteen fifty-four, there was no simulation that could tell you whether a pilot could survive forty Gs.
So what do we owe these subjects? The volunteers in Dalziel's study, Stapp, Lazear, Carroll. They weren't coerced. They understood the risks, to varying degrees. And they contributed to a body of knowledge that has saved lives at a scale they could not have imagined.
Does that retroactively justify the methods? Or does it simply mean we should honor the sacrifice while acknowledging the ethical problems?
I keep coming back to the volunteers in Dalziel's study specifically. Not the researchers, the subjects. They were probably graduate students, or lab technicians, or local residents recruited for a few dollars an hour. They sat in a chair, held a conductor, and let someone run current through them until their hands stopped obeying.
There's a kind of quiet courage in that. It's not the dramatic courage of Lazear facing a fatal disease. It's the mundane courage of showing up, knowing it's going to hurt, and doing it anyway because the data matters.
And the data does matter. Every time you plug in an appliance, every time a GFCI trips and saves someone from electrocution, you're benefiting from what those volunteers endured.
The misconception about the let-go threshold is that it's a single universal number. It's not. Dalziel found it varies by gender — six milliamps for men, five for women — and it also varies by body size, by the path of the current, by the frequency. That's why safety standards build in substantial margins.
The five milliamp figure for women isn't because women are weaker. It's a physiological difference in average muscle mass and body composition. Dalziel's data showed the distribution, and the standards were designed to protect the most sensitive portion of the population.
And that's another thing the volunteers gave us. Not just a number, but a distribution. An understanding of how the threshold varies across a population. You can't get that from a single heroic self-experiment. You need a study with multiple subjects.
Which is why Dalziel's study design, ethically fraught as it was, produced something that individual self-experimentation couldn't. Statistical power. A range. Confidence intervals.
I'm not sure about this part, but I think the replication studies continued into the eighties in some university labs. The numbers were so foundational that people wanted to verify them under different conditions, with different equipment.
Wait, the eighties? IRBs were well established by then.
They were. But the research still happened, with IRB approval, under conditions that would look very different from Dalziel's original work. Paid volunteers, informed consent documents, medical monitoring. The question had been asked, and the answer was important enough that people kept asking it.
Hilbert: We ran one of those replication studies. Nineteen eighty-five, university electrical engineering department. I was a lab tech, twenty-two years old, no ethics training to speak of.
Hilbert: My job was to calibrate the current delivery system before each session. And to test the emergency release mechanism. That was the part I took seriously, because I'd seen what happened when it didn't fire cleanly.
Hilbert: It was a spring-loaded clamp. Heavy. When the subject failed to let go, you hit a release and the clamp physically yanked their hand off the conductor. We tested it with a weighted dummy hand before every session. If the spring was worn, the hand would hang on for half a second longer than it should.
Hilbert: The subjects were graduate students. Paid. I don't remember the exact rate, but it was enough that nobody had trouble recruiting. There was a kind of machismo about it. Guys would come back from a session and joke about how far they'd gotten before they couldn't let go.
Hilbert: One of them was my roommate. He did three sessions. The money went toward his wedding. His wife still doesn't know exactly what he did to earn it. I don't know why I'm telling you this now. I've never told anyone.
The fact that your roommate volunteered for three sessions — that's the thing. The volunteers weren't passive victims. They were making a calculation. Discomfort in exchange for money, and maybe a story they could tell.
The calculation was informed enough. They knew what the endpoint was. They knew the release mechanism existed. They knew they could stop at any time.
Hilbert: The researchers said that. You can stop whenever you want. But we need you to not stop until you can't. That's the line I remember. It was honest, in its way. They weren't pretending it was pleasant.
Hilbert: The data was solid. I don't have any doubts about that. The numbers we generated matched Dalziel's, within the expected variation. The safety standards that came out of it are real and they matter.
Hilbert: But I've never been able to shake the image of those students, hands strapped to electrodes, waiting for the current to climb. The moment where they'd try to open their hand and it wouldn't move. You could see it in their faces. The body stops being something you control.
That's the part the abstract ethical discussion can't capture. The lived experience of the moment when voluntary control fails. It's not just pain. It's a kind of betrayal by your own body.
Yet your roommate went back twice more.
Hilbert: He did. Said it wasn't that bad after the first time. You know what to expect, so it's easier. He was more annoyed about the electrode paste than the shock, as I recall. It took forever to wash out of his arm hair.
Hilbert: I think about him sometimes. He's still married. Two kids. He's a structural engineer now. Designs bridges. I wonder if he ever thinks about what he did for that wedding money.
The volunteers in Dalziel's original study probably had the same attitude. It's a few hours of discomfort, the pay is decent, and you're contributing to something. The fact that the something turned out to be the foundation of global electrical safety — they couldn't have known that.
That's the thing about these stories. They're not abstract. They're about real people who made real choices, with real bodies, and the choices had real consequences. Some of them died. Some of them carried damage for decades. Some of them just had a story they never told their wives.
The open question is whether the end justifies the means when the means are voluntary. We can't run Dalziel's study today. An IRB would reject it instantly. And that's correct — the risks to subjects, the lack of therapeutic benefit, the availability of alternative approaches, all of it would fail review.
But the knowledge is real. The safety standards are real. The lives saved are real. And we can't unknow what we know. We can't retroactively decide that the data doesn't count because we don't like how it was obtained.
The future implication is that we're going to face this again. Brain-computer interfaces, advanced prosthetics, gene editing — these fields will eventually require human data that can't be obtained any other way. The Dalziel story is a reminder that the answers are never clean.
The volunteers in these studies are not heroes in the conventional sense. They didn't run into burning buildings. They sat in chairs and held conductors and let their bodies be pushed to the limit. But their willingness to endure pain for strangers they would never meet is a form of heroism worth remembering.
Even as we're grateful that modern ethics review would never allow it.
The single most common wrong belief about today's topic is that pre-IRB research was uniformly reckless. The truth is messier. Dalziel had a forced-release mechanism. Stapp had medical monitoring. The volunteers were informed, within the standards of their time. The ethics were different, not absent.
The other wrong belief is that the let-go threshold is one number. It's a distribution, and the safety margins exist precisely because the distribution is wide.
What stays with me is Hilbert's roommate. The structural engineer who paid for his wedding by letting researchers shock him until his hand wouldn't open. He's out there right now, designing bridges, and the safety factors in those bridges are informed by the same kind of data he helped generate.
I wonder if he'd volunteer again, knowing what he knows now. I suspect he would. The calculation wasn't wrong. The discomfort was real, but the outcome was worth it. For him, and for everyone who's ever been protected by a GFCI.
Thanks to Hilbert Flumingtop for producing, as always.
This has been My Weird Prompts. Email us at show at my weird prompts dot com with your own stories of scientific heroism in small. We'll be back soon.
