Megan Skiendel: Read a Yelp review lately written by the restaurant owner? No, I mean — you can usually tell. The facts check out, technically. But everything chosen, everything left out — it's calibrated. That's actually what a primary source is.
David Sterling: A primary source. Documents, records, letters — produced at or near the time of the event.
Megan Skiendel: Indispensable to historians. And never neutral. That's the tension this whole episode lives in. Structural bias in sources isn't an occasional problem — it's a permanent, inescapable property of the evidence itself. You can manage it. You cannot eliminate it.
David Sterling: And the timeline on this goes back — how far, exactly?
Megan Skiendel: Twenty-three hundred BC. The Victory Stele of Naram-Sin — an Akkadian king — is one of the earliest surviving official records we have. It praises his military campaign. Doesn't describe it. Praises it. That's not a modern problem with spin culture. That is the origin point.
David Sterling: Wait. The very first record is promotional material.
David Sterling: So what do you actually do with that? The toolkit historians use — contextual analysis, triangulation — those are systematic responses to that problem. Who wrote it, for whom, under what constraints, toward what end.
Megan Skiendel: Right. And source triangulation is the move — compare multiple independent accounts, convergences point toward fact, divergences point toward bias or propaganda. Sounds clean.
David Sterling: Until you get Jallianwala Bagh.
Megan Skiendel: 1919. British colonial records — 'restoring order.' Indian nationalist accounts — imperial brutality. Same event. Irreconcilably different primary source narratives. Triangulation doesn't resolve that.
David Sterling: And that's — well, that's where I get skeptical of the whole framework, actually. Because triangulation is only as good as the independence of your sources. If every surviving account comes from the same literate, powerful class — convergence isn't confirmation. It's corroboration of a shared bias.
Megan Skiendel: Elite echo.
David Sterling: Richard Evans has this — historians disagree about a mountain's appearance, but they must all be painting the same mountain. I find that reassuring up to a point. Interpretation varies, facts converge. But the question I can't get past is — what if the mountain you're painting was pre-selected by who survived to write it down?
Megan Skiendel: That's exactly it. And Lord Canning is the case that breaks it open for me. 1857, Indian Mutiny — he writes privately to Queen Victoria admitting 'indiscriminate vindictiveness' by British forces. Terrified. Strategy failing. Needs permission to escalate. Then his public dispatches, forty-eight hours later — controlled, measured, inevitable. Same man, same events.
David Sterling: Two completely different historical records.
Megan Skiendel: I mean — I've sat in rooms where that's the email chain versus the earnings call. It's not dishonesty, exactly. It's audience-appropriate calibration. The real stakes only live in the gap between the two versions. Which means if the private letter doesn't survive, you've lost the actual story entirely.
David Sterling: Louis XVI's proclamations are the same architecture. Crafted to shore up monarchical authority — not to document what was happening. So the absence isn't random. It's load-bearing. What doesn't get written is a decision.
Megan Skiendel: Forensic accounting, honestly. The real story was never what got recorded — it was when something stopped being documented, who got dropped from the email copy. The gap is the data.
David Sterling: Collingwood called it — wait, his exact phrase — 'imaginative construction stretched between certain fixed points.' The fixed points are Canning's letter, the date, the money moved. Everything else is bounded imagination. Which is fine. Until you ask whether even that bounded construction escapes present-day assumptions about which gaps matter.
David Sterling: And that's — frankly, that's the thing I can't put down. Because if present-day assumptions shape which gaps matter, then it's not just history with that problem. Lippmann — co-authored the first systematic study of bias in newspapers, analyzing the Times' coverage of the Russian Revolution. Journalism. Same structural source-bias problem. And then you extend it: climate models, medical research, financial reporting — anything where the evidence base is incomplete and filtered through whoever had the resources to generate the record.
Megan Skiendel: The postmodern critique isn't wrong about that. Power shapes what survives.
David Sterling: No — and I'm not dismissing it. But if you accept it fully, then triangulation, contextual analysis, absence reasoning — the whole toolkit Collingwood and Evans built their framework around — that's not a history methodology anymore. That's the methodology. For any evidence-based claim. And I don't know whether that's reassuring or... I'm still turning that over.
