Maya Chen: Nathan, you know the thing that got me this week — I was thinking about how we talk about wrong theories versus right theories, and whether that distinction even holds the way we think it does.
Dr. Nathan Hayes: Mm. You're not going to tell me miasma was fine, are you.
Maya Chen: Not — no, not exactly. But here's the specific thing: nineteenth-century cholera treatment included fumigation campaigns, driven entirely by miasma theory, bad-air theory, completely wrong. And those campaigns moved people away from contaminated water. People lived. Wrong theory, real lives.
Dr. Nathan Hayes: The mechanism for those survivals was microbial — interrupted Vibrio cholerae transmission. Full stop. The miasma hypothesis predicted nothing correctly about why it worked.
Maya Chen: Yeah, and I'm not — I hear you, I'm not saying the theory was right. I'm asking something different. I'm asking: if the outcome was good and the explanation was wrong, what does that do to how we evaluate theories at all?
Dr. Nathan Hayes: It's a real tension. And it maps directly onto Semmelweis — he reduces puerperal fever mortality dramatically in the 1840s with chlorinated lime handwashing, and he cannot tell his colleagues what he is actually washing off. He has 'cadaverous particles.' He has a protocol. He does not have germ theory. And they destroyed him for the protocol anyway.
Maya Chen: He had the evidence without the scaffolding to hold it up in public.
Dr. Nathan Hayes: Now, Leeuwenhoek had seen microorganisms under a microscope in the 1670s — the actual entities, physically observed. And that observation sat essentially dormant for nearly two hundred years before Koch in 1876 proves Bacillus anthracis causes anthrax, proves it rigorously, builds the postulates. And then — those postulates fail almost immediately for viruses. Fail for conditional pathogens. The canonical framework couldn't hold its own scope.
Maya Chen: Hold on — so the first formal proof that a specific germ causes a specific disease happens in 1876, and the framework is already breaking by the time anyone applies it widely?
Dr. Nathan Hayes: Essentially, yes. Which puts us somewhere uncomfortable — if the wrong theory produced real survival, and the right theory required immediate amendment, and the empirical tools for all of this existed two centuries before anyone used them — what are we actually claiming when we say we understand causation?
Maya Chen: And that's — okay, that question is the one I can't get past. Because there's a version of this where miasma theory looks almost *rational*. Like, you're a city planner in 1830, you notice people get sick near the river where the latrines drain. You move the latrines. People stop dying. From the inside of that story, you were right.
Dr. Nathan Hayes: The medieval city planner who says 'the stench is killing us, move the latrines' — wrong about mechanism, correct that the river was killing people. The intervention worked. That's the image that stays with me.
Maya Chen: Which is almost more unsettling than being simply wrong.
Dr. Nathan Hayes: Now here's what I want to land precisely, because it matters — miasma wasn't just folk superstition. Disease genuinely clustered near filth. The observations were real. The framework was internally consistent. If all you have is observation and no microbiology, miasma is a *reasonable* inference.
Maya Chen: And then Semmelweis walks in. Vienna obstetric wards, mid-1840s. He introduces chlorinated lime handwashing, mortality from puerperal fever collapses — and he cannot tell anyone *why* it works.
Dr. Nathan Hayes: He had 'cadaverous particles' and a protocol. No validated mechanism. And colleagues rejected him — not because the mortality data was hidden, it wasn't. They rejected him *with the data in front of them.*
Maya Chen: That detail still stops me cold.
Dr. Nathan Hayes: Now John Snow in 1854 is doing something structurally different — he's not working from miasma, he's not working from a confirmed pathogen either, he doesn't know what the specific organism is. But he asks *which pump*, not which vapor. He has its handle removed. The Broad Street outbreak stops. That's a causal inference move, not a germ theory move.
Maya Chen: Wait — so Snow gets credit for germ theory, but he wasn't actually using it?
Dr. Nathan Hayes: Correct. Which is — actually, no, it's more precise to say: the *logic* he used predated the validated theory. The win at Broad Street belongs as much to epidemiological reasoning as to anything Koch would formalize twenty years later.
Maya Chen: So what we're really looking at is — wrong theory drives sanitation, right causal logic drives the pump handle, and then germ theory arrives and takes credit for all of it retroactively.
Dr. Nathan Hayes: That's the category error I want to name clearly. Attributing the miasma sanitation wins *to the theory* rather than to the practice it inspired — that's imprecise. When you redirect people away from contaminated water because you believe in noxious vapors, the mechanism that saved them was microbial. You blocked Vibrio cholerae transmission. The vapor hypothesis predicted nothing correct about *why* it worked.
Maya Chen: And that distinction — between the practice working and the explanation being right — that's not just a historical footnote, is it. Because if we collapse those, we're basically saying it doesn't matter what we believe about causation as long as something good happens. And that feels like it leads somewhere dangerous.
Dr. Nathan Hayes: It leads exactly there. And that's why the consolidation matters — because germ theory didn't just give us better outcomes, it gave us a logic for *why* the outcome worked. Pasteur's fermentation experiments in the late 1850s, early 1860s — he's not doing medicine yet, he's watching yeast drive fermentation, watching microorganisms cause decay, and he kills spontaneous generation as a hypothesis. That's the first empirical move. Not a cure. A mechanism.
Maya Chen: And each of these is a separate move. Pasteur doesn't hand Lister a complete theory. Lister takes microbial principles and just... applies them to a table in an Edinburgh surgical ward in 1867.
Dr. Nathan Hayes: Carbolic acid. On the instruments, the wound, the dressings. Surgical mortality dropped. And the surgeon doing that amputation — he doesn't have Koch's logic yet, Koch is still nine years away — but he has something he didn't have before. A named enemy.
Maya Chen: That's — yeah, that's the thing that gets me. That surgeon in 1867 Edinburgh, carbolic acid on the table, doesn't fully understand *why* it works, but for the first time there's a reason to believe the invisible thing can be countered. Not just endured.
Dr. Nathan Hayes: Agency. Mechanistically unvalidated, but real. And then Koch in 1876 is the consolidating move — he takes Bacillus anthracis, he grows it in pure culture, he reproduces anthrax in a healthy animal, he re-isolates the same bacterium. One specific organism, one specific disease. That's the first time anyone had done that rigorously.
Maya Chen: Wait — so Lister is already doing antiseptic surgery for nearly a decade before Koch formally proves any of this.
Dr. Nathan Hayes: Yes. Correct. The intervention precedes the proof. Again.
Maya Chen: And then Koch formalizes it — the Henle–Koch postulates, named partly for Friedrich Henle who'd theorized living contagia animata on purely theoretical grounds — and suddenly there's a *logical structure*. Isolate, culture, reproduce, re-isolate. And that structure is what drives everything after: sewage systems, hospital sterilization, vaccination scaling.
Dr. Nathan Hayes: That's the downstream application. Once you can say 'block this specific organism's transmission,' prevention becomes logical, not just hopeful. Municipalities aren't just moving latrines anymore — they're engineering water systems because the model tells them *why* the water is the vector.
Maya Chen: So the identity shift is real. Pasteur gives doctors a world with causation in it. Lister gives them something to do with their hands. Koch gives them a method for knowing they're right. That's — mm, that's not one discovery. That's three different gifts arriving separately.
Dr. Nathan Hayes: And worth noting — what gets canonized as 'germ theory' is actually the postulates arriving right at the moment they're already failing. Viruses didn't fit. Mycobacterium tuberculosis caused disease in some hosts, stayed latent in others. We'll get to exactly how badly that breaks things — and what it means that we kept teaching the postulates as gospel anyway.
Maya Chen: Hold on — canonized *as* it was failing?
Dr. Nathan Hayes: Almost immediately. The framework was formalized, institutionalized, made the standard — and the exceptions were already accumulating. That's not incidental. That's a structural pattern in how we treat knowledge once it starts winning.
Maya Chen: And that's the part that actually disturbs me — the postulates weren't just challenged later, they failed on contact with the next category of pathogen. Like, almost immediately.
Dr. Nathan Hayes: Right — so here's the mechanism problem in plain terms. Koch's postulates require four things: isolate the pathogen from every case, grow it independently in pure culture, introduce it to a healthy host and reproduce the disease, re-isolate it from that host. Four steps. Clean logic.
Maya Chen: That sounds airtight.
Dr. Nathan Hayes: Step two collapses for viruses. Entirely. Thomas Rivers — virologist, early 20th century — articulates it directly: viruses cannot be grown independently of host cells. They require a living cell to replicate. So the second criterion, pure culture, independent growth — it simply does not apply. The postulates assume a world of free-living microorganisms. Viruses are not that.
Maya Chen: Wait — so we taught Koch's postulates as the gold standard for proving disease causation, knowing viruses don't satisfy them?
Dr. Nathan Hayes: For essentially a century, yes. And it's not only viruses — conditional pathogens. Mycobacterium tuberculosis causes active disease in some hosts, stays latent in others indefinitely. That breaks postulate three. You can't reliably reproduce the disease in a healthy host if the host's immune state is doing half the causal work. The framework required modification almost from the start, and we just... kept teaching the original.
Maya Chen: So the framework was authoritative because it was useful, not because it held universally.
Dr. Nathan Hayes: That's exactly the structural problem. And now — important — historians will tell you it's more accurate to speak of plural germ theories in the late 19th century. Not one coherent revolution. Real incoherencies about transmission routes, about the role of the host versus the organism. That messiness is scientifically legitimate.
Maya Chen: Mm — but that same messiness gets used to argue something much harder?
Dr. Nathan Hayes: It gets weaponized. There are minority positions — I mean, actively circulating now — arguing bacteria and viruses are the result, not the cause, of disease. And the people making that argument point directly to these real historical incoherencies, the postulates' limits, the plural theories. But the argument is citation-poor. It does not engage with the experimental record from Pasteur and Koch forward. The nuance is real; the conclusion they draw from it isn't.
Maya Chen: So the legitimate qualification and the denialism are using the same historical evidence, just — one of them actually follows where it leads.
Dr. Nathan Hayes: Correct. And the place where genuine extension is needed — not denial, extension — is chronic disease. Non-communicable disease is now roughly 70% of disease burden in developed countries. The CDC's Preventing Chronic Disease journal has flagged this directly: no germ-theory equivalent exists for this dominant category. No single microbial etiology. The one-pathogen-one-disease logic doesn't map.
Maya Chen: Which is — okay, that's where this becomes something other than just a history problem. Because if the framework we canonized was already provisional at the edges, and now the majority of what's killing people doesn't fit it at all... we're not missing a Broad Street pump. We might be missing the question entirely.
Dr. Nathan Hayes: That's the tension worth sitting in. The messiness that denialists exploit is real — the postulates did break down, the theories were plural, the framework was always provisional. What isn't real is using that to say specific microbes don't cause specific infectious diseases. Those are not the same claim, and collapsing them is where the actual danger lives.
Maya Chen: And that's — we're right back where we started. Wrong theory, real lives. I said it like it was a puzzle to solve. But Semmelweis and Snow kind of prove it isn't a puzzle. It's just... the condition. You act on what you have, the mechanism catches up later, or it doesn't.
Dr. Nathan Hayes: Both of them got the intervention right before anyone could tell them why it worked. Snow removes the Broad Street pump handle in 1854, no pathogen identified, no validated framework — just causal logic pointing at a water source. Semmelweis a decade earlier, same structure. The proof followed the result.
Maya Chen: And we still don't have a postulates-equivalent for chronic disease. That's — mm, that's not a small gap.
Dr. Nathan Hayes: No. And the honest version of where that leaves us — we may be locked into looking for single causal agents in terrain that is genuinely multicausal. The Broad Street pump equivalent for, say, type 2 diabetes or autoimmune disease... it may not be a pump. It may be twenty pumps, interacting.
Maya Chen: Which is exactly what the miasma theorist couldn't see — not because they were stupid, but because the framework only had room for one thing to be the cause.
Dr. Nathan Hayes: Right — but the part that doesn't fit neatly is: we know this about ourselves now. We have the history. And I'm not sure that makes it easier to see the blind spot while you're inside it.
Maya Chen: No. I think it probably doesn't. That city planner in 1830 moving the latrines was pretty confident too.
Dr. Nathan Hayes: Good place to stop, I think. Genuinely — thank you for pushing on the mechanism side of this.
