The Cut That Travels

China is not an open market for films. A limited number of foreign titles are imported each year on revenue-sharing terms, which is the only route to the large money, and each one needs approval from the state's film authorities before it can be shown. Everything else, including what will be approved, is partly written down and partly guessed.

In 2012 an agreement between the two governments raised the revenue-sharing quota from twenty films a year to thirty-four, with the extra places reserved for 3D and large-format releases, and raised the share of the takings that foreign studios could keep. Chinese box office was growing fast enough that it looked set to overtake North America. For a studio budgeting a blockbuster, China became a market worth editing for before anyone asked.

That is the distinctive feature of this lesson's censorship: the edit is made in advance, by the studio, on a guess.

The clearest case is the remake of Red Dawn. It was shot in 2009 about a Chinese invasion of the United States. Its release was delayed by the studio's financial troubles, and in the meantime the producers decided the villains were a liability, so the invaders were changed in post-production into North Koreans: flags and insignia altered digitally, dialogue rewritten and dubbed. Nobody in Beijing had issued an instruction. The film that reached American cinemas in 2012 had been altered to suit a different audience altogether.

Doctor Strange, in 2016, is the contested case. In the comics the Ancient One is a Tibetan man; the film cast Tilda Swinton as a Celtic woman. Marvel said it wanted to avoid an old and offensive stereotype of the wise Asian mentor, which is a real concern. The screenwriter, in an interview, gave a different reason: that acknowledging Tibet risked the Chinese market. Both may have been at work. The point is that a reason of the second kind is never announced, so we usually learn about it by accident.

Top Gun: Maverick is the case that ran backwards. The first trailer, in 2019, showed Maverick in the leather jacket from the 1986 film, but the patches bearing the flags of Taiwan and Japan had been replaced with other designs in similar colours. The Chinese company Tencent was then a co-financier. It withdrew before release, and when the film came out in 2022 the original patches were back. It was not released in China, and it took nearly $1.5 billion around the world regardless.

Now compare this with the Breen office, because the shape is the same.

Breen's censorship was effective because it worked on scripts. Writers learned what would come back marked, and stopped writing it, so the finished films carried almost no evidence of what had been removed. Pre-emptive cutting has the same property. A scene is never written, a villain's nationality is settled in a meeting, a map is redrawn before the shot is designed. There is no letter from a censor to object to, because none was sent.

The difference is reach. A cut made for Breen changed American films for the American audiences the Code claimed to protect. A cut made for China is usually made to the one version released around the world, so the restriction travels to audiences it was never aimed at. A viewer in London or São Paulo watches a film shaped for a reader in Beijing, and has no way of knowing it.

The bet did not always pay. Approval was never guaranteed, release dates could be withheld without explanation, and from the late 2010s Hollywood's share of the Chinese box office fell sharply as Chinese productions took over the top of the chart. A guess about an opaque process can be wrong in either direction, and a studio has no way of telling which of its cuts bought anything.

That leaves the lesson this book has been building. The censorship that works is rarely the kind with a statute and a pair of scissors. It is a reader somebody anticipates, at the point where a film is cheapest to change, which is before it is made. The Breen office was that reader for thirty years. A market can be one too, without owning a single cinema.