The Subtitled Horror Film and the Anglophone Blind Spot

In 2009 Magnolia put Let the Right One In out on disc in the United States and shipped it with a subtitle track nobody had seen before. The theatrical translation had been replaced with a new one, and the new one was flatter in a way that mattered — dialogue with menace in it came out neutral, exchanges built on what a character declines to say came out as information. Viewers noticed within days, the complaint went round the film internet at speed, and Magnolia eventually agreed to make the original theatrical track available on replacement discs. Nobody involved was malicious. Somebody had simply decided that the wording could be made easier, and the film was quietly de-tuned on its way to the shelf.

I keep coming back to that incident because it is the whole argument in miniature. The anglophone blind spot toward subtitled genre cinema is usually described as an audience failure — a laziness, a refusal to read, a provincialism you are supposed to shake your head at. The evidence points somewhere else. For most of the last seventy years, the English-speaking viewer was never offered the choice. The barrier was constructed downstream, by the people importing the films, and the audience has been blamed for a decision made in a distributor’s office. I unpick the general trade-off in subtitles or dubbing; this is about who built the wall.

What the importers did

The standard pipeline for a foreign genre film reaching Britain or America was renovation rather than translation.

The Italian industry got the worst of it. Dario Argento’s Profondo Rosso ran around two hours in Italy in 1975 and reached English-speaking audiences roughly twenty minutes shorter, retitled Deep Red, with a large amount of the character material and comedy stripped out — the very things that make the film breathe, as I argue in Argento’s perfect giallo. Fulci’s Zombi 2 was itself an opportunistic Italian sequel-in-name to Romero, and arrived in Britain as Zombie Flesh Eaters and in America as Zombie, a retitling shuffle I get into in Fulci’s shark versus zombie landmark. Honda’s Gojira lost its atomic argument and gained Raymond Burr. Bava’s films were re-ordered, re-scored and re-named so routinely that establishing what an English-language viewer actually saw in 1965 requires forensic work, which is part of why the father of Italian horror took thirty years to be recognised as one.

Then Britain criminalised a chunk of the rest. The Video Recordings Act 1984 and the DPP’s list of prosecutable titles removed dozens of films from legal circulation, and the list leaned heavily European. An entire generation grew up with a horror shelf whose foreign section had been curated by a prosecutor.

So when someone says the anglophone audience ignored world horror, ask what they were ignoring it with. A butchered print, a retitled sleeve, a dub recorded by a man in New Jersey, or nothing at all.

The dubbing question is more complicated than purists admit

I have to prosecute the subtitle position here, because the purist case collapses on precisely the films the purists love most.

Italian genre cinema was shot without synchronised sound. Dialogue was post-synced in the studio afterwards, in Italian, over performances by casts who on set were frequently speaking English, Spanish, German and Italian at each other in the same scene, counting numbers where the lines would go. The Italian track is therefore a dub too. There is no sacred original vocal performance being betrayed by the English track, because the Italian version is also a construction laid over a silent shoot, and in some cases the English track carries the actual actor’s voice while the Italian one does not.

That undermines the reflex entirely. For Bava, Fulci, Argento and most of the Eurohorror canon, “watch it subtitled” is a preference between two artificial soundtracks, and the honest answer varies film by film — which cast, which mix, which script, whether the English track was written by someone who had read the film or someone paid by the reel. I go through the specifics in the dubbing of Eurohorror. The purist rule works cleanly for Japanese, Korean and French cinema, where the production sound is the performance. Applied to a 1971 giallo it is superstition wearing a critic’s hat.

The other channel, and the other blind spot

There was a second import route and it did almost as much damage by being respectful.

Kwaidan (1964), Onibaba (1964) and Kuroneko (1968) all reached the West through the art-house circuit. They were subtitled properly, reviewed seriously, and filed under Japanese cinema, which meant they were placed on a shelf the horror audience never walked past. Kobayashi made three hours of the most beautiful ghost stories ever committed to colour and they were sold to people who wanted Cannes, as I argue in ghost stories as painted theatre, while Shindō’s reed-field films sat in the same category despite being unambiguously horror, which I get at in the reed field and the demon mask.

The exploitation channel mangled the films and delivered them to the right audience. The art channel preserved the films and delivered them to the wrong one. Between those two errors sits nearly everything an English-speaking horror viewer of the 1970s never got to see, and neither error was theirs.

What the barrier costs in practice

I can put a number on the gap, roughly. Bong Joon-ho made his one-inch-barrier remark at the Golden Globes in January 2020, and six weeks later Parasite became the first non-English-language film to win Best Picture. That is seventy-odd years of the Academy, and the film that broke it is a genre film — a home-invasion thriller with a basement, which I would happily programme next to The Host. The interesting part is what happened next. Korean cinema had been extraordinary for two decades before 2020, and the anglophone market had been treating it as a supply of remake rights, which is the subject of the remake as cultural translation. The blind spot had a business model.

Streaming has changed the mechanics without settling the question. Netflix defaults non-English titles to an English dub for most viewers, on the reported logic that dub-default improves completion rates, which is almost certainly true and is also how you build a generation who have technically watched a lot of world cinema and heard none of it. The boutique disc labels went the other way, restoring the long cuts, commissioning fresh translations and putting both audio tracks on the same disc, which is one of the reasons the physical media cult turned out to matter rather more than its detractors expected.

The craft nobody credits

Understand what a subtitler is actually doing and the Magnolia incident stops looking like carelessness.

A subtitle is governed by reading speed. The working standards used across the industry put a line at roughly forty-two characters and cap the rate at something close to seventeen characters per second for adult viewers, with a minimum duration below which a caption reads as a flicker. That arithmetic turns the job into compression. The subtitler decides which of the several things a sentence is doing can survive, and the first casualty is always register — the hesitation, the politeness level, the word chosen to avoid another word.

Japanese and Korean punish this harder than most, because both languages encode social relationship in the grammar of ordinary speech. A character who shifts from formal to plain address has just done something enormous, and there is no English word to put in the caption. The subtitler either finds a rhythm that implies it or drops it, and dropping it is faster, cheaper and invisible to anyone checking the work against a script.

That is what happened to Let the Right One In. Somebody was asked to make the captions easier, obeyed, and removed the sub-text that the reader was supposed to do the work on. The vandalism was upstream, in a translation brief written by someone who thought the words were the content.

The honest cost of subtitles

Now the part the subtitle evangelists skip, because there genuinely is a price and it is highest in exactly this genre.

Reading takes your eyes to the bottom of the frame. In dialogue cinema that costs you a little. In Bava and Argento it costs you the film. These are directors whose meaning is carried by colour, by what is placed in the top-left corner of the shot, by a slow reveal at the edge of focus while somebody talks — the argument I make in colour timing as horror and in Argento’s colour, glass and the killer’s glove. A subtitle drags the eye to the safest, emptiest region of a composition that has been designed to be scanned. You are reading a book about a painting while standing in front of it.

Dubbing costs you performance and gains you the frame. Subtitles cost you the frame and gain you performance. There is no free option, and anybody who tells you otherwise is selling a position rather than describing an experience. The reasonable practice is to choose per film: subtitle Miike, because Audition lives on a vocal register that no dub has ever caught; consider the English track for a 1972 giallo where the frame is doing the work and the Italian audio is a studio artefact anyway.

The films that never got offered

Consider what the blind spot actually withheld, because abstraction makes it sound like a matter of taste.

Jaume Balagueró and Paco Plaza made [REC] in Spain in 2007, a found-footage film with an unusual willingness to actually run rather than shuffle, which I go at in the found footage film that actually sprints. Sony’s American remake Quarantine arrived barely a year later, in 2008, and in a number of territories the original’s own release was complicated by the copy. The remake stood in the doorway the original needed to walk through.

Nacho Vigalondo shot Timecrimes in 2007 on a budget that would embarrass a music video and built a time-loop thriller tighter than most studio science fiction of the decade, as I set out in Spanish time travel on a shoestring; a Hollywood remake was announced almost immediately and the film’s own profile in English-speaking markets stayed a festival matter for years. Banjong Pisanthanakun and Parkpoom Wongpoom’s Shutter (2004) is one of the most efficient ghost films of its decade and is known to most English speakers, if at all, through a 2008 American version — the pattern I trace in the Thai ghost photography landmark.

Three films, three countries, three remakes inside about eighteen months. The market’s answer to a good foreign genre film was to buy the idea and lock the film in a drawer, and the audience was then told they preferred it that way.

What the blind spot really was

The comfortable story is that anglophone audiences could not be bothered. The record shows an industry that cut the films, renamed them, dubbed them badly, prosecuted a portion of them, filed the beautiful ones under the wrong heading, and then remade the survivors rather than release them. The audience inherited that wall, and a good deal of the fun of watching genre cinema now is discovering how much was standing behind it the whole time — which is where a list like Korean genre cinema, ten to start with earns its keep.

The one-inch barrier is real. It was mostly built from this side.