A Rally Built for the Cameras

Leni Riefenstahl had been a dancer and an actress in German mountain films before she directed. Hitler admired her work, and after a first, shorter film of the 1933 party rally he commissioned her to film the next one. The congress held at Nuremberg in September 1934 became Triumph of the Will, released in 1935.

The film has no narrator. There is no voice telling you what to think, only what was in front of the cameras: Hitler's plane descending through cloud over the city, crowds, marching columns, speeches, the massed formations in the Luitpold Arena, torchlight. That absence is the basis of the defence Riefenstahl made for the rest of her long life. It was, she said, a documentary: a record of a historical event, with nothing invented and nobody explaining it, and therefore history rather than propaganda.

The defence fails, and how it fails is the interesting part.

It fails first on the facts. The congress was not an event that happened to be filmed. Riefenstahl's crews had camera positions built into it, with tracks and towers, and Susan Sontag argued that how the congress was staged was partly decided by the decision to film it, so that history had become theatre before a camera turned. And by Albert Speer's account in his memoirs, some footage of the speeches was unusable, so the shots were filmed again in a studio, on a set Speer built to reproduce part of the hall, where party leaders delivered their speeches a second time to a camera. A film defended as pure record contains reconstructions made for it.

It fails second on the principle. Leaving out a narrator does not leave out the argument. Every documentary still selects, frames and cuts. Triumph of the Will shoots its leader from below and its crowds from above, alternates the individual and the mass until they feel like one body, and arranges its sequences so that the movement flows towards a single figure. Those are arguments, made in the most effective way an argument can be made: without anyone appearing to make it. A narrator can be disagreed with. A sequence of images that seems simply to have happened is much harder to resist.

Set it beside Nanook of the North and the comparison is precise. Both films were arranged. Flaherty staged scenes to recover a way of life that had passed, with the cooperation of the people in them. At Nuremberg the event staged itself for the camera, and the film's claim not to have intervened was the device that made it work. The difference is not the amount of staging. It is what the staging served, and what the film claimed about it.

Riefenstahl's earlier film of the 1933 rally makes the same point from another side. It was withdrawn after the Night of the Long Knives in 1934, because one of the men most prominent in it, Ernst Röhm, had been murdered in the purge. A film that only recorded history needed recalling when the history changed.

Sontag's essay Fascinating Fascism, published in 1975 as Riefenstahl was being rediscovered as an artist, argued that the craft cannot be separated from what it was built to glorify. The tracking shots, the geometry of massed bodies, the light on banners are studied by filmmakers to this day, and the imagery has travelled far beyond its source: the walk down a wide aisle between massed ranks is regularly compared to the medal ceremony that ends Star Wars. Formal brilliance is a reason to study a film. It is not evidence that the film only recorded.

Riefenstahl died in 2003, at 101, still insisting she had been an artist rather than a propagandist. The film itself is the best evidence against her. What it shows is real, and almost none of it is neutral.