Eleanor Crane: I've been thinking about a man — sixty-seven, healthy his whole life, and now his body is starting to turn on him in these specific, predictable ways. And I keep asking: did evolution just miss this? Just fail to patch it?
Ben Okonkwo: And your answer was no.
Eleanor Crane: My answer was — actually, no, it's worse than that. Consider that elephants have twenty copies of the TP53 tumor suppressor gene. We have one. Evolution is not incapable of building more protection in. It made a different investment in elephants and it worked for their lifespan. So when it comes to us, the question isn't capability. It's priority.
Ben Okonkwo: Twenty copies — and they cap out around forty years of life. We get eighty on one copy. So the redundancy doesn't obviously buy longevity.
Eleanor Crane: Right, which is part of what makes this so hard to read cleanly. But the deeper issue — the one I want us to work through today — is whether aging is something selection neglected, or something it chose. And Haldane was the one who first put a name to the shape of it.
Ben Okonkwo: 1941. He's looking at Huntington's disease — onset in the forties, fatal, hereditary — and he's asking why it hasn't been eliminated. And the only answer that works is: by the time it activates, the person has already reproduced. Selection literally cannot reach it.
Eleanor Crane: And then Williams in 1957 goes further — it's not just that these alleles slip past. They're maintained. There's early benefit holding them in the genome.
Ben Okonkwo: Which is the claim that aging is not a bug evolution failed to fix. It is a structural byproduct of how selection actually works.
Eleanor Crane: And the thing I keep sitting with — what is the early benefit keeping a Huntington's allele in the pool? Because if that's load-bearing for Williams's whole argument, I want to know what it's actually resting on.
Ben Okonkwo: Okay, so — that's exactly where the evidence gets genuinely thin. And I think that tension is what this whole episode is about.
Eleanor Crane: Thin in a way that matters, though — because the popular version of this story stops at Medawar, and people don't realize there are actually two separate claims here.
Ben Okonkwo: That's the thing I want to cut through. Medawar in 1952 is saying: selection goes blind late in life. Bad alleles accumulate because nothing removes them. That is real. But it's — it's a story about negligence. A blind spot.
Eleanor Crane: And Williams says something structurally different in 1957.
Ben Okonkwo: Actively different. His paper — 'Pleiotropy, Natural Selection, and the Evolution of Senescence,' published in Evolution — the argument isn't just that harmful alleles slip through. It's that selection is writing loans. It keeps funding early reproductive payoffs, and the debt — the late-life cost — comes due after the founders have already reproduced and cashed out. The system isn't blind. It's doing this on purpose.
Eleanor Crane: Wait — so the allele isn't just surviving. It's being kept.
Ben Okonkwo: Positively selected. Maintained. And that — okay, that's the step that changes everything, because it means you're not fighting a failure of the system. You're fighting the system working exactly as it was shaped to work.
Eleanor Crane: That's the moral shift, isn't it. Medawar gives you negligence. Williams gives you structural inevitability.
Ben Okonkwo: And testosterone is where you feel it in the body. Early life — male reproductive success, secondary sexual characteristics, the whole competitive advantage. Later — elevated prostate cancer risk. Same allele. Same molecule. It didn't change. The context around it did.
Eleanor Crane: The man at forty is not a different biological organism than the man at seventy. He's just past the repayment window.
Ben Okonkwo: Right — but the part I want to flag, because I think it gets glossed over, is that Williams's claim is actually harder to prove than Medawar's. Medawar just needs to show selection weakens with age. Williams needs to show the allele is actively maintained *because* of the early benefit. Those are not the same standard of evidence.
Eleanor Crane: So when we say testosterone 'proves' this — what are we actually proving?
Ben Okonkwo: We're showing a plausible case. The correlation is real, the early benefit is real, the late cost is real. What we're not showing, not cleanly, is that selection is actively holding the allele in the population *because* of the early side. We've got the pattern. The mechanism underneath it is — mm, still partly inferred.
Eleanor Crane: Which means the 1957 paper is still the founding document of the idea — but the genome is still mostly dark on which specific alleles are actually doing this work in humans. We have the theory. We have the examples. The full inventory? We don't have it.
Ben Okonkwo: And that darkness — the fact that we can't name the alleles — that's actually the tell that the confident version of this story is getting something wrong.
Eleanor Crane: Say that more directly.
Ben Okonkwo: The take that's circulating — in popular science, in longevity writing — treats antagonistic pleiotropy as basically a confirmed mechanism. Proven. And it's not. It's an elegant hypothesis that produces patterns consistent with what we see. But Medawar's mutation accumulation also produces patterns consistent with what we see. Both theories predict aging. So how do you run the experiment that separates them?
Eleanor Crane: You can't just point at the pattern.
Ben Okonkwo: Right — but that's what the clean story does. And testosterone is the flagship example, which is actually... I mean, I want to stress-test that one, because the causal chain is genuinely entangled. Testosterone correlates with prostate cancer risk. The benefit is real, the cost is real. But between the molecule and the tumor there are androgen receptors, inflammation pathways, cell proliferation rates — it's not one gene doing two clean things. It resists the simple one-gene, two-effect framing that Williams's hypothesis needs.
Eleanor Crane: So using testosterone as the clean proof case may actually be overstating what the data show.
Ben Okonkwo: Overstating, or at least — collapsing the complexity in a way that feels more settled than it is.
Eleanor Crane: And then there's Kirkwood — 1977 — who adds a third frame. Disposable soma. The argument that somatic maintenance is just evolutionarily expensive when you're likely to die from a predator anyway, so the body doesn't bother investing in it. Which is... actually, wait — that's not the same mechanism as Williams at all.
Ben Okonkwo: No — Kirkwood's is a resource allocation story. Not a two-benefit-one-allele story. And a 2024 PLOS Biology paper had to propose a hierarchical unified framework just to reconcile all three — mutation accumulation, antagonistic pleiotropy, disposable soma — into something coherent. That paper exists because the theoretical boundaries between these mechanisms are still genuinely unsettled among specialists.
Eleanor Crane: A 2024 paper. Sixty-seven years after Williams.
Ben Okonkwo: Sixty-seven years. And they're still trying to build the unified model. That's the tell.
Eleanor Crane: Which makes the confident popular retelling — the one that treats the 1957 paper as the answer rather than the opening question — more than a little premature. And for longevity medicine, the drugs being developed right now to suppress these pathways — this gets genuinely uncomfortable, and we'll get to it.
Ben Okonkwo: The theory is elegant. The evidence is sparse. And we keep forgetting those are different things.
Eleanor Crane: And that's the part that keeps me up — because 'we'll get to suppressing pathways' sounds clinical. But there's a woman who's forty-two, fit, no medications, and she's just gotten a polygenic report back. It flags a variant. Early fertility benefit, late-life cognitive decline. What does her doctor actually say?
Ben Okonkwo: The doctor can read the finding. That part is solved. What they cannot tell her is whether suppressing that pathway's late-life effect — the cognitive piece — pulls out something her body is still actively using at forty-two.
Eleanor Crane: Still using now.
Ben Okonkwo: If Williams is right — if the allele is being maintained because of the early benefit — then the pathway isn't dormant. It's doing something. And we don't have a clean way to suppress only the late-life end of a pleiotropic allele. The biology doesn't come apart that neatly.
Eleanor Crane: So medicine is being asked to solve a problem evolution not only didn't solve — it never had to face. Because organisms didn't survive long enough post-reproduction for long-life-and-health to be a selection target at all.
Ben Okonkwo: That's the structural mismatch. Natural selection optimizes for early reproductive success. Longevity medicine is trying to optimize for long life and health simultaneously. Those are — I mean, those aren't just different goals. They're goals selection never needed to invent tools for.
Eleanor Crane: Which is where the TP53 story stops being reassuring.
Ben Okonkwo: Walk me through what you mean.
Eleanor Crane: The elephant comparison gets used to make the point that evolution could build more protection in. Twenty TP53 copies. But elephants live roughly forty years in the wild. We live up to eighty on one copy. So the redundancy in elephants isn't a model for longevity — it may be responding to a completely different selective pressure. Body size, cancer risk per cell division, something else entirely.
Ben Okonkwo: And we don't actually know whether our single TP53 copy is being actively maintained by selection — or just tolerated. Those are different claims with different intervention implications.
Eleanor Crane: Tolerated is — wait, that's actually a more unsettling word than negligence.
Ben Okonkwo: Right — because if it's tolerated, maybe you can supplement around it. But if selection is actively holding it because single-copy TP53 is doing something in early life that we haven't identified yet — fertility, immune priming, something — then any intervention touching that gene is flying partially blind.
Eleanor Crane: That forty-two-year-old woman is the first generation being asked to make that bet. Not in a clinical trial. In a doctor's office, with a report she paid for, with a question her physician doesn't have the evidence to answer.
Ben Okonkwo: And the tools that would answer it — the ones that could distinguish active maintenance from passive tolerance — those tools don't exist yet. Because evolution never needed to invent them, and medicine is only just understanding why it has to.
Eleanor Crane: Williams wrote his paper in 1957. Sixty-seven years later, a woman is sitting in a clinic holding evidence he'd recognize — and the field is still reaching for the instruments to tell her what it means.
Ben Okonkwo: And that's — I mean, I keep turning this over. Williams gave us the reframe in 1957. Aging isn't a failure, it's structural. That part has held. But knowing it's structural doesn't tell you how to touch it. Whether you work with the trade-off, try to find the seam where early and late effects can be pulled apart — or whether you just step outside that space entirely and build something evolution never needed to construct. Those are genuinely different bets.
Eleanor Crane: And we don't know which one longevity medicine is actually doing right now.
Ben Okonkwo: No. And the unsettling version of that — actually, wait, let me say it plainly — if it turns out we're already building tools outside the space selection ever explored, we don't have a model for what we're risking. There's no evolutionary precedent for an organism that extends its lifespan while keeping early-life functions intact. Selection never needed to invent that. So whatever costs come with it, we won't see them coming from the theory.
Eleanor Crane: The first generation that lives substantially longer might not know whether they succeeded because they understood Williams's logic — or because they got lucky stepping past it.
Ben Okonkwo: Yeah. And I don't think that question resolves on a short timeline.
Eleanor Crane: I'm going to sit with that for a while. Thank you for the company on this one.
