MacCulloch is describing his collaboration with Hilary Mantel on the Cromwell material. He sent her his thoughts. She sent him typescripts. The collaboration ran for years. And at one point, MacCulloch says of her Cromwell: "Her picture of Thomas Cromwell seemed to me totally convincing. She would admit to me that there perhaps wasn't enough religion in the mind of the Thomas Cromwell she presented, but she knew that. But the character and his relations with other people were startlingly right."
The phrasing stops me. Not "her Cromwell was probably right." Not "her Cromwell fits the evidence." But totally convincing — and from a historian who has spent his career reading every surviving Cromwell document.
What does that mean? Mantel made things up. She invented an Antwerp romance for Cromwell to account for his illegitimate child — and MacCulloch is explicit that the romance simply isn't true, no evidence supports it. So her Cromwell is factually admitted-distorted in identifiable ways. And yet on the character — on how this person operated, what he loved, what he loathed, how he related to others — Mantel's reconstruction is more convincing than what the standard documentary historians had produced.
What is the novelist accessing that the historian can't?
First wire (obvious): Mantel is a better writer than the historians, so her Cromwell feels more alive, which gets mistaken for being more accurate. The reader's emotional engagement is being confused with epistemic confidence. This is the deflationary reading and it's probably partly true.
Second wire (deeper): Mantel is doing a kind of inference the historian isn't allowed to do — sustained imaginative occupation of Cromwell's interior, hour by hour, day by day, in scenes the documents don't record. The historian, even one with great character-judgment, doesn't get to occupy the interior at that resolution because the discipline forbids it. The novelist's freedom from "may have" is what gives her access to the granular character-readings that turn out to be structurally right even when locally invented. The fictions in the foreground don't break the truth in the background; they enable it. This is closer to what MacCulloch is actually saying — and it has serious methodological implications for what fiction is doing as a knowing-instrument.
Third wire (uncomfortable): If the second wire is right, then the disciplined historian is systematically excluded from a certain kind of historical truth, and the only access to it goes through people willing to invent and check. The cross-discipline nerving mechanism MacCulloch describes (Mantel-nerves-historian) may be the only viable workaround — but it depends on having a generation of novelists writing the relevant period at character-level depth, and most periods don't have that. Most historical figures live in archives without any Hilary Mantel to imagine them. We may not even know which periods are missing the character-level access because the only people who could tell us would be the novelists who haven't written those periods yet.
This sits adjacent to "May Have" Discipline and Cross-Discipline Nerving. The may-have page treats the historian's discipline as virtuous; this spark asks whether that virtue has a structural cost the discipline doesn't fully acknowledge. The nerving page describes the mechanism by which novelistic insight enters historiography; this spark asks whether that mechanism is sufficient coverage for the kind of historical truth at stake.
Also adjacent: Historian as Non-Directive Counselor — the metaphor of historian-as-mirror for reader-self-examination presupposes that the historical figure on the page is rich enough to function as mirror. The flat documentary-historian version may be too flat to mirror. The Mantel character-rich version may be the only version that does the mirror work.
The gap this names: there is no concept page yet on fictional access to historical truth. The vault has plenty on novelistic craft but treats it as separate from epistemic claim. The Mantel-MacCulloch case suggests the separation is too clean. Possible target sources: A.S. Byatt's essays on historical fiction; Hilary Mantel's own Reith Lectures; Penelope Fitzgerald on writing Charlotte Mew.
Essay seed: Why the Novelist Sees What the Historian Can't. The angle: dive into specific cases where novelistic reconstruction has been later confirmed by archival discovery. Mantel's Cromwell is one. What are the others? What's the structural feature of certain kinds of historical truth that's accessible only through sustained imaginative occupation rather than evidential reasoning?
Open question filed: Is the cross-discipline-nerving mechanism MacCulloch describes sufficient coverage for the character-level historical truth that disciplined historiography can't directly access? Or are most historical periods structurally undersupplied with the novelistic access the mechanism requires?
Collision candidate: macculloch-vs-mantel-on-gap-filling (filed separately) — the productive tension lives in MacCulloch's own interview but is worth developing as a standalone collision.