A column about curious stories from the worlds of science, technology, and film. Today’s “random movie of the day” is Kalifornia, the 1993 thriller that marked Dominic Sena’s directorial debut, starring Brad Pitt, David Duchovny, Juliette Lewis, and Michelle Forbes. It premiered in Montreal on August 27, 1993, and won two festival awards. The film opened in U.S. theaters on September 3, in 359 cinemas, debuted at number seventeen with $1.1 million for the weekend—and finished its run with $2.4 million against an $8.5 million budget. A classic box-office corpse.
Now for another date. On April 6, 2009, the FBI publicly showed journalists the workings of its Highway Serial Killings program for the first time. Bureau analysts had linked hundreds of unsolved murders: bodies were being found along interstate highways, and over and over, the suspects turned out to be long-haul truckers. By then, the database held more than five hundred dead women—mostly women whose disappearance nobody was likely to notice: trailer-park prostitutes, hitchhikers, drivers who had run out of gas. Over three decades, the victim count, by the Bureau’s estimate, ran into the hundreds.
The overlap between the film’s plot and the statistics is not superficial. Kalifornia is the story of a journalist who drives across half the country researching serial murders—and lets a serial killer into his car. In 2009, the FBI was stating much the same thing, only in the language of agency reports: the road really does work as an incubator for these stories, and the scale was not something screenwriters had imagined. I looked into where that leads—and found something better than a random coincidence: an Easter egg the filmmakers hid in the geography, and an investigative mechanism Hollywood has shown on screen more than once without noticing that it had started working for real.
Tim Metcalfe wrote the screenplay in 1987 with Stephen Levy and gave it a terse title—simply Kalifornia. The concept was unexpected: not a horror movie, but a black comedy with Hitchcockian suspense. Metcalfe put his aims this way: “to scare the audience, comment on our national obsession with ‘true crime’ stories—and punish myself for my unhealthy fascination with murder and murderers.” The script sat idle for another three years, until Propaganda Films bought it in the fall of 1990—at the request of director Dominic Sena, who had fallen for the overall concept and for the character of Early Grayce.
Metcalfe did two rewrites—and was fired in March 1991: the producers found his drafts hopelessly bland. There was no money for another screenwriter, so Sena and two producers spent another year hammering out ten drafts themselves. That revision produced the journalist’s voice-over and a change of genre: the black comedy became a hard-edged thriller, and the violence stopped hiding. Along the way, the budget rose from the originally proposed $4.5 million to $8.5 million. Metcalfe later publicly disowned the result: he called the violence in the final version tasteless and “cartoonish,” in the Tarantino vein—and praised Juliette Lewis alone.
The casting came together with startling precision. Pitt was hunting for a role against the “nice guy” image he had acquired in Thelma & Louise—and Sena had noticed him there. Lewis was cast while she was finishing Scorsese’s Cape Fear; she and Pitt had been a couple for several years by then—it had started on the set of the TV movie Too Young to Die?, when she was sixteen and he was twenty-six—but Sena, by his own admission, had no idea they were together. Michelle Forbes was found at an audition in Los Angeles: the director needed her cold detachment. Duchovny got the role last, after several reads with all four actors. For Forbes, it was her feature-film debut.
Filming in the spring of 1992 took place in Georgia and rural California; the schedule was compressed, and some locations were allotted a single day. Producer Steve Golin later marveled at the endurance of the four leads: “They work unimaginable hours under conditions stars should not have to endure. Desert hotels are not luxury: there’s a swamp cooler in the room, no air conditioning at all. A tough shoot.” Then, on June 28, 1992, during a shift at a gas station in rural California, the set was shaken: the press at the time reported a magnitude-7.5 quake. Today we know the event as the Landers earthquake—under the modern USGS catalog, its moment magnitude was 7.3. The figures differ not because somebody got it wrong, but because the scales are different.
Critics split in a way that rarely happens. Roger Ebert gave it four stars: the film was “so recklessly honest and so well acted that I dropped all my distance and simply watched it as the lives of real people”; he called Pitt’s and Lewis’s work “two of the most chilling and convincing performances I have seen.” Leonard Klady of Variety compared it to Laughton’s The Night of the Hunter. At the other pole were Peter Travers of Rolling Stone, who found elements of the screenplay absurd; Mark Caro of the Chicago Tribune (“a pretentious B movie”); and Owen Gleiberman of Entertainment Weekly. The aggregators recorded the final split: 59% fresh on Rotten Tomatoes, with an average of 6.2 out of 10 and a consensus calling it a “tonally uneven thriller with a predatory performance from Pitt,” and 49 out of 100 on Metacritic. Consolation prizes: two Montreal festival awards and three Saturn Award nominations, including Best Horror Film.
The film did not survive its theatrical run—but it had a long afterlife on home video. PolyGram released VHS and DVD editions on November 18, 1997, in two versions at once, unrated and theatrical; MGM reissued the DVD in 2000, put out the unrated version on Blu-ray in August 2010, and on March 5, 2019, Shout! Factory assembled a two-disc collector’s edition with a new Sena interview—only then did the theatrical version reach Blu-ray for the first time. There was a third artifact, the most curious of all: DC Comics commissioned a 32-page comic adaptation—script by Chuck Dixon, art by Duncan Fegredo—tied to the home-video release. The project vanished along with distributor Gramercy Pictures: they never finished coloring the comic, it was never published, and today it exists only online—black and white, with Fegredo’s original cover.
Now for the promised Easter egg—almost nobody noticed it. The locations where the plot places its murders were named after the actors, and IMDb lists them: Lewiston Ranch and the town of Mount Juliet in Texas—after Juliette Lewis; Forbes, Tennessee—after Michelle Forbes; the Davidson Mine and Dew Cove, Nevada—after David Duchovny; the Bradbury textile warehouse in Pittsburg—after Brad Pitt. Each of the four got a “crime scene” bearing their name. Given the genre, the joke is exactly as dark as the plot demanded.
There is plenty of small-grain texture besides. Pitt personally went to a dentist and had a front tooth chipped for the role. And the film’s quietest “message” is hidden in the ending. The action is long over: Brian and Carrie live in a house by the ocean in Malibu, a Venice gallery is interested in her work, they are getting ready to go out and celebrate—and Brian, sorting through tapes from the recorder, accidentally leaves it running. At the end of one cassette, there surfaces a “thank you” left by Adele—the killer’s naïve girlfriend, whom he shot. With that scene, the film acknowledges this: those erased along the roadside leave a voice behind, too.
Let us jump from 1993 to early 2004. Terry Turner, a senior criminal intelligence analyst with the Oklahoma State Bureau of Investigation, had hit a wall with a series of seven murders along I-40: all the victims were trailer-park prostitutes, killed and left by the roadside. Each murder was in a separate jurisdiction, and no county could physically see what was right under its nose. Turner sent the cases to ViCAP—the Violent Criminal Apprehension Program, the FBI’s only national database for serial crimes. Program analysts found links to unsolved murders in other states; the recognizable figure of the suspect—a long-haul trucker—became the trigger: in 2004, the FBI launched the Highway Serial Killings Initiative, with support from the trucking industry itself.
The FBI posted its official breakdown of the initiative later, on April 6, 2009, and it opens with a sentence nobody would want to read aloud in HR: “If there is an ideal occupation for a serial killer, it is probably long-haul trucking.” Then comes the necessary caveat from analyst Christie Palazzolo, who ran the program for years: it is an honorable profession, and the overwhelming majority of drivers are not killers. But the pattern, she said, was clear: “We had an inordinate number of victims and offenders from one fairly narrow occupational pool.” The anatomy of the problem is simple. A killer behind the wheel of a rig has perfect mobility: it is easy to pick up a victim at a gas station, leave the body in a neighboring state, be in another county by morning—and every county investigates its own murder and has almost never encountered crimes of this kind.
By the time of the public announcement, the FBI database contained more than five hundred dead women whose bodies had been found along trucking routes, at rest stops, and in motels, along with data on dozens of truckers charged with—or suspected of—murders and rapes near highways. Since 2008, the database has been available online to local police: they can enter their own cases and search for matches themselves. Michael Harrigan, who oversaw the initiative, described its work in the same words screenwriters like to use for detectives: connect the dots. He added a caveat of his own: “We don’t want to scare the public into thinking that every time they stop for gas they need to look over their shoulder. A lot of the victims were making bad choices, but that doesn’t mean they deserved to die.” His colleague, supervisor John Molnar, was convinced the real number was “grossly underreported.” By 2016, the ViCAP list, according to the Bureau itself, had grown to more than 750 people killed along highways and nearly 450 suspects; on a map in the Quantico office, more than five hundred cases stood as red dots—and the overwhelming majority still had not been linked to a single identified offender.
What connecting the dots looks like can be seen in the cases from a Los Angeles Times report. Nashville, July 2007. Detective Pat Postiglione watches footage from a Truck Stops of America camera: rigs come and go like airplanes at an airport. The only thing that catches his eye is a yellow 18-wheeler that remains there for around thirty minutes, while the other drivers spend an hour or more fueling and resting. The next morning, the detective spots a similar truck in a rundown part of town and knocks on the cab door. Out comes Bruce Mendenhall: average height and build, with pinched features, an unbuttoned shirt, and no shoes. There is a drop of blood on his thumb, several more on the door.
Mendenhall calmly agreed to a DNA swab and a voluntary search. In the cab’s sleeping compartment, officers found a bag with bloodstained women’s clothing, along with a phone and bank card belonging to a young woman who had vanished from Indianapolis twelve hours earlier. By the end of testing, blood and DNA had tied Mendenhall to at least seven victims; he was charged with four murders. He pleaded not guilty and, as the newspaper wrote in April 2009, was awaiting trial in Nashville. The detective recalled that the FBI’s reconstructed timeline of his routes showed the intervals between murders shrinking. “He was heading for a catastrophe.”
The second case shows the network effect. A month after Mendenhall’s arrest, trucker Adam Leroy Lane parked his rig in a Boston suburb, slipped through an unlocked door into a house, and stood over the homeowners’ fifteen-year-old daughter with a knife. The father burst from the bedroom, put the intruder in a chokehold, and the mother snatched away the knife. Lane had three blades, a coil of wire, and a throwing star on him; in his cab was a DVD of the film Manhunter. Six months earlier, a Massachusetts police sergeant had attended an FBI presentation in Reno, and now he sent the Bureau a one-sentence email: “Just want to make sure this guy is on your radar.” That was how Lane’s case was connected to murders in two other states; in Massachusetts, he pleaded guilty and received fifty years. J. Patrick Barnes, a New Jersey prosecutor handling one of the linked cases, stated the moral without theatrics: “You can’t connect the dots if you don’t know where they are.”
Most of the dots on that map remained unconnected. A typical case has sat in the database since 2005. In central Illinois, near Interstate 55, ATV riders found a decomposed human leg in the woods; a little farther on, another. The painted toenails suggested the remains belonged to a woman—and the case stalled: no name, no leads.
Three years later, an FBI analyst spotted part of a tattoo on one of the legs and helped state police identify the remains: Lindsey Harris, twenty-one, a sex worker from the Las Vegas Strip. She had disappeared roughly two weeks before the discovery—1,400 kilometers from Illinois. Harris became the third sex worker from Las Vegas whose dismembered remains had been found along highways between 2003 and 2005; a fourth young woman from the same street vanished, is presumed dead, and her remains have still not been found—at least as of the report. Mike Jennings, an Illinois State Police agent who worked with the FBI on the identification, admitted that he would soon retire and that this case would weigh on him to the end: “My gut tells me there was a trucker behind the wheel.”
And there lies the honest answer to the question posed by the hook. Kalifornia predicted nothing—it merely described, earlier, a sample that county police could not physically assemble: a mobile offender, hundreds of jurisdictions, victims few people were looking for. In 1987, Metcalfe wanted to mock the national obsession with true crime—and two decades later, that same obsession, scaled up to the federal level, became a working tool: a database, a map with red dots, and a sergeant’s one-line email. And the film’s ending now sounds suspiciously contemporary: the recorder is accidentally left running, and in a place nobody listened through, suddenly there is a “thank you.” Five hundred such cassettes lie in the database at once—working like they do in that ending: waiting for someone to listen all the way through.