Brian Reed: Eliza — I was on a train yesterday and the person next to me was reading a biography of Napoleon Bonaparte, and I just kept thinking about this episode the whole ride.
Eliza Ward: Because you were thinking about whether Napoleon actually caused anything or just happened to be in the right autocracy at the right time?
Brian Reed: Basically, yeah — and that's actually the question. There's a study — 1,108 leaders across 130 countries — that tracked growth rate changes when leaders died in office. In autocracies with few institutional constraints, the leader's death measurably shifts the economic trajectory. In democracies, the effect nearly disappears.
Eliza Ward: That's the agency-versus-structure debate with actual numbers finally attached to it.
Brian Reed: Right — and what gets me is that it's not resolving the debate the way Thomas Carlyle would have wanted. Carlyle's whole Great Man theory says exceptional individuals are the primary engines of historical change. But the data is saying: the individual matters *contingently*, depending on whether institutions exist to absorb or amplify the decision. Think of it as amplification, not origination.
Eliza Ward: Hold on — that framing still concedes too much to structure. The study shows the causal *weight* varies by context. It doesn't show the decision was irrelevant in democracies.
Brian Reed: So what's the threshold — like, at what point does 'the effect nearly vanishes' become 'the individual caused this'?
Eliza Ward: That's — yeah, that's exactly the problem historians have been stuck on since Carlyle and, I mean, since Marx pushed back on him. And we're not solving it today. But I think the study at least tells you the question isn't metaphysical — it's actually measurable.
Brian Reed: But measurable doesn't mean Carlyle was right — and that's where I want to push. Marx's whole point was that the individual is downstream of material conditions. Napoleon doesn't forge Europe; Europe's collapsing feudal order produces someone who can move fast. The person is a product of the era.
Eliza Ward: Wait — that's actually what the study complicates, though. Carlyle's error wasn't inventing the leader effect. The effect is real. His error was thinking it was intrinsic to the individual rather than gated by the system around them.
Brian Reed: No, I don't buy that as a Carlyle rescue operation.
Eliza Ward: It's not a rescue — it's a refinement. He identified something real. The Annales School and Marx identified something real. The 1960s new social history movement that came out of the Annales tradition, the whole longue durée turn toward geography and demography — all of that was a correction to Carlyle going too far, not a proof that he was entirely wrong.
Brian Reed: Okay, but — hang on — the longue durée point is actually the strongest argument against you here. Geography, climate, demographic rhythms — those forces operate across centuries. Napoleon's decisions are, I mean, they're noise on that signal. The Annales School didn't just shift emphasis. By the 1960s it had basically restructured what historians treated as a legitimate explanation.
Eliza Ward: Restructured historiographical practice, yeah. But that's — wait, that's not the same as settling the causal question. 'Geography does the heavy lifting over centuries' is true and also completely compatible with 'Napoleon's specific decisions in 1805 altered which states existed by 1815.' Those aren't competing claims.
Brian Reed: So you're saying Carlyle had the right instinct and just — what, wrong scope?
Eliza Ward: Wrong scope, wrong mechanism — actually no, the mechanism too. He thought will was primary. The study says the effect is structurally conditioned. But he noticed that individuals sometimes *do* shift outcomes, and historical materialism spent decades trying to explain that away instead of just calibrating it.
Brian Reed: Marx didn't explain it away — he said material conditions produce the conditions for a particular kind of person to act. That's not denying the action. It's locating where the real causal weight sits. And the autocracy data doesn't contradict that. It just shows you the window where individual action can propagate. The window itself is structural.
Eliza Ward: The window is structural — fine, I'll grant that. But the window doesn't make the decision. And that gap is exactly where Hitler breaks every clean framework you try to build.
Brian Reed: Okay — because Mandel actually tried this. His 1986 New Left Review essay, specifically on World War Two. He's a committed historical materialist and he's still saying: social forces are primary, but exceptional individuals retain genuine roles. He won't just dissolve Hitler into the Weimar collapse.
Eliza Ward: Right — and the problem is 'primary' versus 'genuine' is never given a metric. Mandel draws the line and then just... leaves it there.
Brian Reed: Which is why Lindenfeld's taxonomy is interesting — he tries to actually classify the causes. Empowering causes versus constraining causes. And here's the part that stops me: the structural conditions that installed Hitler — the economic crisis, Weimar instability, the class dislocations — Lindenfeld has to rank those as constraining, not empowering, because Hitler didn't control them.
Eliza Ward: That's — wait. That's a counterintuitive result from his own framework.
Brian Reed: Exactly. The conditions that made him possible get classified as constraints. Which — I mean, that should make us suspicious of any taxonomy that produces that answer and calls it settled.
Eliza Ward: Think about a junior Weimar civil servant in 1932, processing the final electoral returns that hand the NSDAP a plurality. She is executing a structural outcome — that part is almost purely mechanical. But what Hitler does in the cabinet negotiations immediately after? No structural model predicted that with any precision. The decision space was real.
Brian Reed: No, I think that's actually — hang on — that's where Giddens's duality of structure is supposed to do the work. Structure and agency mutually constitutive, not opposed. The civil servant reproduces the structure; Hitler's negotiation operates within and then reshapes it.
Eliza Ward: 'Mutually constitutive' is true. I'll give Giddens that — it's a real philosophical advance over the binary. But it does not tell you how much causal weight to assign Hitler's specific decisions in 1941 versus the conditions that preceded them, and that's the problem historians actually need solved — we'll get into whether there's an operationalization path out of that, because I think it gets genuinely hard.
Brian Reed: The operationalization path is real though — the leadership mortality research found a way to measure the gap. The question is whether historians are willing to use it.
Eliza Ward: Okay but 'willing to use it' skips the actual problem — even if historians adopt the mortality study's methodology, Giddens's duality of structure still doesn't give you a number. It gives you a circle. Structures produce agents, agents reproduce structures, and you're back where you started trying to assign Hitler's invasion of the Soviet Union in June 1941 to a column.
Brian Reed: Wait — are you saying Giddens is useless, or just not sufficient?
Eliza Ward: Not useless — that's actually where I'll concede. The duality of structure is a real advance over the old binary. Pure Carlyle voluntarism, pure historical determinism — Giddens dissolves that false choice. That's genuine philosophical work. I'll stop calling it useless.
Brian Reed: Okay, that's — yeah, I'll take that. But the broader gap?
Eliza Ward: Stays open. 'Mutually constitutive' is not a calibration tool. Mandel admitted exceptional individuals retain genuine roles — that's the moral accountability problem right there, because if you can't weight Hitler's 1941 decision against the structural conditions that preceded it, you either dissolve his responsibility into Weimar or you're back to Carlyle. Neither survives the evidence.
Brian Reed: Right — but the mortality study is at least a proof of concept that the weighting problem isn't permanently unsolvable. The effect size difference between autocracies and democracies, that's the beginning of a calibration metric. Not the end, not Giddens's replacement — the beginning.
Eliza Ward: Beginning, yes. But historians still have to get from an aggregate mortality effect across 1,108 leaders down to one decision, one month, one invasion. That distance hasn't closed.
Brian Reed: So the gap is real — what changes that? What's the next actual test?
Eliza Ward: Honestly? I think the test is the accountability question — and not as philosophy. If you miscalibrate toward historical determinism, you've handed every future perpetrator a ready-made defense. The system made me do it. The structural forces were primary. And there's no agreed-upon metric that lets you say no, actually, the causal weight in *your* case sat with the decision, not the conditions.
Brian Reed: And the other direction is just as bad — overweight agency and you let every complicit institution walk. Weimar's collapse doesn't just evaporate because Hitler chose. The epistemological question and the moral question are tangled up, but they're not the same question, and I don't think we've found a way to hold them apart cleanly. Nobody has.
Eliza Ward: No agreed metric. That's actually where this lands, isn't it.
Brian Reed: We keep arguing about who holds the pen, but maybe the real problem is we can't agree what counts as writing.
Eliza Ward: That's going to sit with me for a while. Thanks for the train ride, essentially.
