The Greatest Show on Earth (1952) Review
4/10
A circus manager is told the season will be cut short unless he makes it profitable. He hires a European trapeze star, which demotes the woman he’s involved with, and the two performers begin competing at increasing risk while a train carries everybody between towns.
The 4 is for a documentary with a plot glued to it. The circus footage is valuable, the acts are real performers doing real work, and everything constructed around them is filler of a quality DeMille wouldn’t have accepted from a writer twenty years earlier.
It frustrates more than plenty of worse films, because the real circus is right there being wonderful while DeMille keeps cutting back to Heston telling everyone he’s married to the show. Two and a half hours of that is a long sit, and the Oscar makes it worse, because it asks me to take it seriously. It’s also the entry on this page where the argument requires no effort at all. High Noon is on this site, is eighty-five minutes, and is one of the ten best American films. It won two awards in total, for the story and for editing, which for a Best Picture winner is a very thin haul. The same ceremony handed High Noon four awards, including Best Actor for Gary Cooper, Editing, Score and Song. That means the film that lost Best Picture went home from that same room with twice as many.
The Circus Footage Is Real
DeMille shot with the actual Ringling Brothers and Barnum and Bailey company, on tour, using their performers, their equipment and their logistics.
That material is the film’s only lasting value and it’s considerable. The tent being raised, the train being loaded, the parade, the acts themselves, and a documentary record of an industry that no longer exists in this form.
The teardown is better than the setup. A performance ends and the film follows the company dismantling everything and loading it in sequence, and the footage of a town-sized operation disappearing overnight is the best thing in the picture. The specificity makes it worth keeping. Eight hundred people, a hundred railway cars, a schedule, and the film spends real time on how a circus physically moves from one town to another overnight.
Anybody interested in that should watch the first forty minutes and the train sequence and stop. The performers are the film and the actors are the interruption. I’d happily watch a cut that was nothing but canvas going up and coming down, and I suspect it would be a better film than the one that won. DeMille narrates it himself, in the first person, over the opening. That tells you what the production thought it was making. A director introducing his own film as a document isn’t describing a drama. He was also two years from The Ten Commandments, which is the film the career is really remembered for, and this win sits awkwardly between a body of work and its best-known entry.
Documentary material inside a fiction is almost always the strongest thing in the work and it can’t carry a story on its own. The trap is thinking the research is the book. Somebody who has learned how a circus moves wants to show it, and showing it’s not a plot.
When your real material is better than your invented material, the invented material has to be smaller. This film gives the drama the same weight and loses.
Process is interesting and it’s not narrative. A hundred railway cars being loaded is worth four minutes and not forty.
The Plot Is Assembled From Parts
Brad manages the circus. Holly loves him and resents the demotion. Sebastian is arrogant and skilled. Angel wants Brad. Phyllis wants Sebastian. A clown has a secret. There’s an embezzling concessionaire.
The film develops none of that. Each strand gets scenes, none of the relationships changes in a way that follows from character, and the resolutions come from events and not from decisions. The trapeze rivalry also resolves off a physical accident instead of a decision. That’s the film’s habit. Almost everything that changes in the plot changes because something breaks or somebody falls.
Charlton Heston is fine and the part is a function. Brad is competent, is married to the circus, and says so repeatedly, and that’s the entire characterization across two and a half hours.
Betty Hutton and Cornel Wilde get the trapeze rivalry, which is the one thread with any tension, and the film converts it into a romance in the final act. That removes the only stake it had established, and it’s where I stopped caring about anybody on the platform. Gloria Grahame and Dorothy Lamour are both in it and neither has anything. Grahame had won Supporting Actress the same year for The Bad and the Beautiful and is here handling an elephant, and Lamour sings a number and disappears for an hour.
James Stewart in Makeup
Stewart plays Buttons, a clown who never removes his makeup, which the film establishes early and explains in the last twenty minutes.
It’s the film’s one good idea. A man hiding inside a costume in a business full of costumes, in a traveling company that doesn’t ask questions, and Stewart plays him as gentle and permanently watchful.
The makeup rule is also established properly. He’s never seen without it, including in the sleeping car, and the film sets that up early enough that the audience has stopped noticing by the time it matters. The reveal is that he’s a doctor wanted for the mercy killing of his wife, and a police officer has been traveling with the circus looking for him all season. What the film does with that is nothing. He’s exposed, he saves a life, he’s arrested, and the film treats it as a subplot resolution when it had built a tragedy. Nothing else in the picture makes me angrier, because a better film is visible inside that material and DeMille walks right past it. Stewart deserved that picture.
Stewart was also at a strange point in his career, two years past Winchester 73 and Harvey, taking a supporting part under heavy makeup in somebody else’s spectacle. He took the job because DeMille asked, by his own account.
One good idea in a weak work is worth identifying clearly, because it shows you what the film could have been about. A man in permanent disguise inside a company of performers is a whole picture. The failure is proportional. This gets perhaps twelve minutes in a hundred and fifty-two.
If you’ve constructed a tragedy, resolve it as one. Exposure, a heroic act and an arrest is a plot sequence and not an ending. Ask which of your subplots would survive being promoted. If one of them is better than your main line, you’ve made a structural error instead of a writing one.
The Train Crash
The set piece everybody remembers is a deliberate derailment, staged with real rolling stock, at considerable expense, about two hours in.
It’s well executed and it’s enormous. Carriages overturn, animals escape, and the sequence runs for several minutes with real scale and no process shots for most of it. The animal footage in that sequence is uncomfortable to watch now for reasons the production wouldn’t have considered, and the film treats the escaped cats and the panicking horses as spectacle. What it lacks is consequence. The circus performs the next day, the company recovers, and the crash is a spectacle followed by a demonstration that the show continues. That’s the film’s stated theme and its least examined one.
That one sequence contains the film. Enormous resources, real skill, no dramatic result, and a return to the status quo with a speech about the circus going on. I admire the wreckage and feel nothing about anyone in it, and for a disaster sequence that’s about as damning as it gets.
Heston also does the sequence with a real injury, having been thrown from a vehicle during the staging, and the footage of him being carried is real. That kind of detail is the only thing in the second half with any weight to it.
High Noon
Best Picture in 1953 went to this. The other nominees were High Noon, The Quiet Man, Ivanhoe and Moulin Rouge. Spielberg has said repeatedly that this was the first film he saw in a cinema and the reason he became a director, and he recreated the train crash in his own work more than once. That’s the clearest evidence of what a child takes from a film regardless of whether the film is good.
High Noon runs eighty-five minutes in near real time, is a McCarthy-era allegory about a town refusing to help a man who has protected it, and won four awards including Best Actor for Gary Cooper and Best Editing. It didn’t win Picture or Director, and Zinnemann lost to John Ford for The Quiet Man.
The political context is documented and is part of the explanation. Carl Foreman wrote High Noon, was called before the House committee during production, refused to name people, was blacklisted, and left the country. A vote for that film in 1953 was a statement and a vote for a circus picture wasn’t.
DeMille was also on the other side of that argument publicly and was a central figure in the industry’s anti-communist activity. That’s context the record keeps and which nobody discussing the result can reasonably ignore.
A work can lose on politics and the politics will be invisible in the record twenty years later. What survives is the result and not the reason. That’s worth knowing because it means the record isn’t a judgment about quality even when it appears to be. Eighty-five minutes against a hundred and fifty-two, and the shorter film is the one still watched. Length is the most common unforced error in ambitious work.
Compare the two films’ subjects. One is about whether a community will act and the other is about whether a show will go on, and only one of those is a question.DeMille also won the Irving G. Thalberg award at the same ceremony. That’s the detail that explains a great deal. A career award and a Best Picture in the same night for a man of seventy-one is a valediction instead of a judgment about a film.
Should you watch The Greatest Show on Earth?
Watch the first forty minutes and the train sequence, and skip the rest without guilt. You get a real documentary record of a circus that no longer exists, one good idea about a clown who never removes his makeup, and a derailment with no consequences. The 4 is for that footage and for Stewart. Everything DeMille built around them is filler, and it’s damn hard to forgive it for beating High Noon, which makes it the least defensible Best Picture of the studio era for me.
