The Uncomfortable Observation
Hayden White's Metahistory (1973) began from something no historian could deny: the past does not come in stories. It comes as one damned thing after another. Stories have beginnings, turning points, and endings; they have significance distributed unevenly; they have a shape. The past has none of this until a historian gives it some.
So where does the shape come from? Not from the evidence — the evidence constrains but does not narrate. It comes from a limited stock of narrative forms the historian brings to the material. White argued these are literary, and that a historian chooses (mostly unconsciously) among four:
| Emplotment | Shape | The past becomes |
|---|---|---|
| Romance | Hero transcends the world | A story of triumph and redemption |
| Tragedy | Fall, with knowledge gained | A story of noble failure |
| Comedy | Conflict resolved, order restored | A story of reconciliation |
| Satire | No transcendence; the world wins | A story of futility |
Take the French Revolution. Emplot it as Romance and it is liberty triumphing over despotism. As Tragedy, it is noble ideals consumed by the Terror. As Satire, it is a change of ruling class producing a new tyranny in fresh vocabulary. Now here is White's uncomfortable point: all four can be written without a single false statement. The facts do not choose. The choice is the historian's, and it is where the meaning lives.
What This Does and Does Not Establish
The strong claim — history is a branch of literature, its truth-claims no different from a novel's — is the one that caused the wars of the 1990s. It cannot be sustained, and the reason is precise: novelists may invent, and historians may not. A novelist who adds a duke to the room has done their job; a historian who does the same has committed fraud. That asymmetry is not a technicality; it is the whole difference, and no amount of narrativism dissolves it.
The weak claim is devastating, correct, and now largely absorbed: narrative form carries meaning that the evidence did not supply, so historians are responsible for it. The shape of your story is an argument you are making, whether you noticed making it or not.
The Test Case That Settled It
Holocaust denial forced the question out of the seminar room. If a history is just an emplotment among possible emplotments, on what grounds is denial wrong rather than merely differently-shaped? White himself was pressed hard on this and his answers were widely judged inadequate.
The resolution the discipline reached is worth stating exactly. Denial is not a rival interpretation; it is refuted by evidence. The transport records, the camp construction orders, the demographic accounting, the physical sites, the testimony, the perpetrators' own documents — the claim collides with the record and dies. Interpretive latitude operates over an evidentiary floor, and the floor is real.
The settlement, and why it is not a cop-out. Constructedness is not the same as arbitrariness. A cathedral is constructed; you cannot build one out of soup. The evidence genuinely underdetermines the story — several honest histories of the Revolution are available, and choosing among them is a real intellectual act with real consequences. But it does not underdetermine it infinitely. The lesson historians kept: interrogate your own emplotment, because your narrative shape is smuggling in claims your footnotes never made. The lesson they refused: that footnotes therefore do not matter.