The Infection Narrative From Romero to Contagion Cinema

Every other monster in horror wants something. The vampire wants blood and, usually, company. The slasher wants a specific list of people dead in a specific order. The ghost wants acknowledgement. The infection wants nothing whatsoever — it has no interiority to want with — and that vacancy is the entire reason the form has outlived every cycle it was born into. A monster with a motive can be negotiated with, and therefore dated. A monster with no motive is a mirror, and the film gets to fill it with whatever the decade is frightened of.

That is why sixty years of infection cinema reads as a decent history of Western institutional anxiety, and why the films are so consistently better than their premises deserve.

Matheson wrote it, Romero misfiled it

Richard Matheson published I Am Legend in 1954 and did something nobody had done with the vampire: he gave it a bacillus. The condition spreads, it obeys biology, it explains the folklore retroactively, and the last uninfected man is reclassified as the monster by a society that has finished changing without him. That is the founding move of the whole genre — the infection as the new normal and the survivor as the anomaly.

George A. Romero said openly that Night of the Living Dead (1968) was ripped off from Matheson’s novel, and he was being precise rather than modest. What Romero changed was the vector. His film never uses the word zombie; the script calls them ghouls, the newsreaders call them ghouls, and the cause is a hand-waved Venus probe returning with radiation. The bite transmission is there — Karen Cooper is the proof — and yet the film treats contagion as a mechanism rather than a subject. The subject is a farmhouse full of Americans who cannot agree on a plan, and a Black lead in 1968 who survives the night and is shot by a posse in the morning.

The film’s ubiquity is a legal accident worth knowing. The Walter Reade Organization changed the title from Night of the Flesh Eaters and, in doing so, dropped the copyright notice from the prints — under the law of the time, that put the film into the public domain immediately. Romero’s group never made the money. Everyone else did, forever, and the free availability of the master text is a real reason the form spread the way it did. Romero built the dead as a social mirror and then lost the deeds to it; Night’s descendants have been arguing with it ever since.

The procedural strain is older than the horror

The infection film’s other parent is the institutional thriller, and it predates Romero by eighteen years.

Elia Kazan’s Panic in the Streets (1950) is a noir in which a US Public Health Service officer has forty-eight hours to trace the contacts of a pneumonic plague case through the New Orleans docks before the city burns. It won the Academy Award for Best Story that year. Shot on location with a documentary flatness, it invents the shape every contagion procedural has used since: the expert nobody believes, the mayor who fears panic more than death, and the epidemiological legwork rendered as detective work. Kazan’s plague chase is a noir with a public-health department.

Robert Wise’s The Andromeda Strain (1971), from Crichton’s 1969 novel, took that shape and removed the crime. Five stages of decontamination, split-screen readouts, an automated self-destruct with a rational failure mode: the film’s confidence is that competent people in a good facility are the most interesting thing you can point a camera at. It is the most optimistic infection film ever made, and it is optimistic about procedure. The procedural as science fiction is a genuinely underrated lane.

Romero merged the two lanes in The Crazies (1973). A military bioweapon codenamed Trixie gets into a Pennsylvania town’s water supply and the army arrives in white suits to contain it. The infected turn violent; the soldiers are equally lethal and considerably better organised; the film cross-cuts between a command post that is receiving accurate information and making terrible decisions with it. That is the institutional turn, and it arrives five years after the farmhouse. Romero kept working the same seam — even his witchcraft film is really about a suburb’s institutional boredom.

Cronenberg’s dissent

David Cronenberg looked at the whole apparatus in 1975 and disagreed with its premise.

Shivers was financed by the Canadian Film Development Corporation, which is to say by the Canadian taxpayer, and it puts a designed parasite into a luxury high-rise on an island outside Montreal. The creature spreads by sexual contact and its effect is to abolish inhibition. Cronenberg’s structure gives you the standard containment narrative and then declines to treat the outcome as a tragedy — the parasite’s inventor built it deliberately, believing that mankind had over-thought itself into a coma. The film provoked a genuine national row after Robert Fulford attacked it in Saturday Night under the headline “You Should Know How Bad This Film Is”, specifically on the grounds that public money had paid for it. It also made its budget back several times over. The apartment block is the whole society, compressed.

Rabid (1977) refined the joke with Marilyn Chambers as a woman whose experimental graft grows a weapon and whose victims spread rabies through Montreal under martial law. Rose is a patient rather than a monster, and the film insists on it. The Cronenberg position — that the infection may be a correction rather than a catastrophe — is the single most useful heresy in the subgenre, and his flesh-and-machine project follows directly from it.

The speed argument

The most-litigated question in the form is pace, and the received history is wrong.

Danny Boyle’s 28 Days Later (2002) is credited with the running infected, and it earns its reputation for other reasons: Alex Garland’s script relocates the threat from the infected to the surviving soldiers by the second half, and Boyle shot on Canon XL1 mini-DV cameras, which let a small unit close central London at dawn for a handful of mornings and produced the smeared, low-resolution look that became the film’s signature. The Rage virus is also honest about its timescale — twenty seconds, not seven days — which fixes the genre’s oldest plausibility problem.

The precedent is Italian and earlier. Umberto Lenzi’s Nightmare City (1980) had irradiated infected sprinting, using weapons and driving vehicles two decades before Boyle, and it is a mess for reasons unrelated to the innovation. Lenzi got there first and nobody sends him a card. Jorge Grau’s Let Sleeping Corpses Lie (1974) had already supplied the ecological cause — an agricultural pest-control machine emitting ultrasonic radiation — which is a better idea than anything in the American cycle. The eco-zombie was a Spanish-Italian invention.

The speed argument matters because it changes the genre’s politics. A slow infection is a film about neighbours, hoarding and rota systems. A fast one is a film about the twenty minutes before the state arrives. The walk-versus-run question is really a question about what kind of story you want, and the meanings keep shifting underneath it.

The economics, and what they did to the form

The infection film is cheap, and the cheapness is load-bearing rather than incidental. The antagonists are extras with contact lenses. The locations are places the world has emptied — a mall after hours, a farmhouse, a train. There is no star, no creature suit, no third-act rig. Romero shot Night in Pittsburgh for a figure usually given as around $114,000 with a cast drawn from local theatre and his own advertising firm’s clients.

This produced a genuine flowering. Lucio Fulci’s Zombie Flesh Eaters (1979) was made to trade off Dawn of the Dead’s Italian release as Zombi, and Fulci spent the freedom on a shark and a splinter. Zombi 2 is a landmark built on a marketing loophole, and the Italian boom that followed ran on the same arithmetic. Michele Soavi’s Cemetery Man (1994) used it to make a philosophical romance. Dellamorte’s melancholy is the strangest thing the cycle ever produced. Shin’ichirō Ueda’s One Cut of the Dead (2017) was made for roughly ¥3 million with unknown actors and grossed something in the region of a thousand times that. Its structure is the joke and the joke is the point.

It also produced the glut. When the barrier to entry is a field and forty friends, the median output is dire, and the subgenre has spent forty years being judged on its median.

The case against my own thesis

Here is the honest problem: the zombie is a terrible model of an actual epidemic, and the films know it.

Real contagion is invisible, statistical and dull. It has an R number rather than a face. Steven Soderbergh’s Contagion (2011), written by Scott Z. Burns with epidemiological consultants including Ian Lipkin, is the only major film to accept this, and it is structurally an anti-horror picture: no monster, no chase, a global cast of experts doing paperwork, and a bat. It performed respectably in 2011 and then became one of the most-rented films on earth in early 2020, which is the most damning review any of its competitors ever received. Audiences under actual threat reached for the film with no zombie in it.

What survives that charge is the work that stops pretending to be about disease at all. Yeon Sang-ho’s Train to Busan (2016) is a film about a fund manager learning, too late, what he owes other people, and the infection is a device for compressing the lesson into a carriage. Its heart is the reason it travelled. Bruce McDonald’s Pontypool (2008) makes the vector the English language itself, traps three people in a church basement radio station, and becomes the most formally daring film in the entire canon on a budget you could raise locally. A horror film built from words is the logical endpoint. Bob Clark’s Deathdream (1974) had already proved the form could carry a single grief — a son back from Vietnam who needs blood — with no epidemiology at all. The homecoming is the horror.

The infection is a solvent, and the genre keeps using it to dissolve whatever institution it wants a look inside. That is why the form will still be here when the current cycle is a memory, and why the good ones — Kazan’s, Cronenberg’s, McDonald’s — share nothing except a refusal to be about a virus.