Skip to content

You Might Also Be Interested

Streaming killed the video star

A few years ago, I decided I was going to watch all of Robert Altman’s movies and make a zine about them. I’ve never been much of a completist about people whose work I like, so even though he’d been one of my favorite directors forever, it hadn’t really occurred to me to seek out the movies that I’d skipped (mostly because they looked lousy). But I was living in Connecticut, which should give you some idea of how it was going, and I needed a small, good thing in a time like that.

I had no idea that after the delightful but poorly received 1980 Robin Williams vehicle Popeye, Altman (who claimed later that, at the time, no one in Hollywood would return his calls) retreated from the limelight, directing theater and eventually adapting the plays he staged into shoestring-budget films. Rather than disguising the fact that these films were once plays, he emphasized their theatricality, shooting interminably long takes in a single location to the neglect of any cinematic elements. These movies have their moments of surprise and delight, but even the best of them (1984’s Secret Honor, in which Philip Baker Hall rages as a post-Watergate Richard Nixon in a ninety-minute monologue) can be a total slog—they’re claustrophobic films that don’t seem like they were fun to make. “A movie that was a play first,” I wrote in my notes on 1982’s Come Back to the 5 & Dime, Jimmy Dean, Jimmy Dean, “is like a pie filled with cake.” I’m not entirely sure what I meant, but I think I was right.

But a wonderful quality of Altman’s movies is that even the slickest of them feel slack and communal. They turn you, in your role as a viewer, into a passenger on a careening human inner tube where everyone’s weight on either side of the screen makes the ride a little wilder. If we watch movies to be immersed in dreams, each of Altman’s films—from the overt surrealism of 3 Women (1977) to the upstairs/downstairs set pieces of Gosford Park (2001)—feel uniquely lucid, as if we have at least as much control of what’s happening as anyone who made it. If an Altman movie is bad, it’s your fault too.

Which means that although his pies-filled-with-cake aren’t Altman’s best movies, I liked them. I wanted to like them. They seemed to come from a mindset similar to the one I found myself in, as if Altman, too, was in his own private Connecticut. And when you’re low, can I tell you how good it feels to see yourself in someone else? Can I tell you how good it feels to like something?

You Know This Feeling Too

Around the turn of the twentieth century, Thorstein Veblen proposed in The Theory of the Leisure Class that our tastes were determined socially. Our cultural preferences drifted down from the upper echelons of the economic ladder, which in Veblen’s era accounted for the popularity of séances and certain hats. Cultural power was ostentatious and unapologetic; it was gilded in gold.

If you’re reading this, you probably didn’t eat any pickled herring today. Your pants are likely within a narrow range of the sum total of possibilities for what pants can be.

While the midcentury dawn of the consumerist middle class, with its mass advertising and fickle teenagers with disposable income, appeared to democratize tastemaking, it only did so within borders that had long been calcified by the ingrained aesthetics of all your favorite robber barons. Those borders are still there. The choices and mores of our long-dead social betters will always constrain what we think is good to eat and wear and listen to and believe. If you’re reading this, you probably didn’t eat any pickled herring today. Your pants are likely within a narrow range of the sum total of possibilities for what pants can be.

The only plausible avenue of dissent from this form of control comes from the stories we tell. While their aesthetics may still be socially determined, what makes us relate to stories comes from someplace stranger, deeper, more ancient. I can’t add anything to the discussions among sociologists, political philosophers, folklorists, and oral historians as to why some stories are retold and passed down while others aren’t, but I know when one makes my breath catch and my skin prickle. I’m pretty sure you know this feeling too.


In Macario (1960), an itinerant woodcutter’s greatest wish is to, just once, eat an entire turkey by himself. Fearing Macario will follow through on a threat to starve himself if he can’t fulfill his desire, his wife steals and cooks a turkey and sends him off to the forest alone to gorge himself. But while preparing to sate the urge he’s held for so long, three visitors arrive in sequence: the Devil, God, and Death. Each asks to share his feast. He refuses the first two.

Though the film is filled with cinematically resonant images (most famously a scene in which Death takes Macario to a cavern full of flickering candles, the length of each standing for how much time an individual person has left), what I think about most after having seen the movie once, fifteen years ago, is a single shot: Macario’s greasy, half-eaten turkey sitting in the grass. It’s a banal, functional image, but it’s become a part of me, like something loved or feared—its inexplicable beauty hanging somewhere in the space between.

Macario isn’t currently available on any major streaming platform, but it’s pretty good!

I’m a Camera

Movies are the perfect medium for immersive, socially connected storytelling, which is weird, since moviemaking itself is so exclusionary. It’s expensive to shoot a film, it involves collaboration between lots of people with different specialized skills, and it usually requires dedicated spaces with costly equipment (and a horde of gatekeepers guarding access to these spaces) to exhibit. Access to digital tools has eroded this restrictiveness somewhat, but the basic language that we recognize as filmic storytelling—like the rhythm of a shot-reverse-shot conversation—was refined by over a century of social and economic elitism in Hollywood.

Intuiting this language makes it hard to know where its rules end and your own perception begins. When I remember my dreams, they flow like movies. I cut to close-ups. I constrict liminal time. I’m myself in my dreams, but I’m a camera too.

A movie theater—social and isolating, bright and dark, silent and deafening—is an inherently paradoxical space in lots of ways, but most of all in that it’s defined by both the godlike power it gives to the film industry and by how it defies that power by creating communities, for two hours at a time, based on shared visions. This contradiction between top-down and bottom-up control is further represented by film projection itself: the rotating shutter that creates the illusion of movement from a filmstrip projects a frame of blackness for each frame of image, which means that half of any film doesn’t exist until you, or I, or we, create it, making the act of watching a movie simultaneously passive and creative. Robert Altman’s movies might be uniquely lucid, but this united, dreamlike quality, providing direct access to the collective unconscious, is in the DNA of all movies. Even if you’re watching them at home.

I was trying to fill in the gaps in my Altman viewing by renting movies via Netflix’s DVD-by-mail service, which was once a cultural phenomenon. A decade earlier, Netflix was shipping nine hundred million physical discs per year, accounting for over 1 percent of all U.S. postal traffic. In 2006, before the rollout of its streaming service, the company approached one billion dollars in revenue. It is rumored that there’d once been a U.S. mailbox strictly dedicated to returned Netflix DVDs outside of the Grand Central post office.

A quick historical timeline of moviegoing shows it shifting away from collectively immersive storytelling and toward social isolation. Home video disrupted the movie theater, the multiplex disrupted the single-screen movie palace, the talking picture disrupted the silent film. Going back further, the silent film disrupted vaudeville, and vaudeville disrupted, like, tying a bear to a pole and poking it with a long stick. A rowdy, participatory audience was evolving into a lone sweatpantsed loser on their couch. At its inception, Netflix could be seen as extending this pattern (you no longer even had to leave your house to rent a movie), but in one important way, its DVD rental service allowed—encouraged, even—an audience’s pushback on the control of the contours of stories told to them.

At first, Netflix’s main competition in the DVD rental market was strip-mall stalwart Blockbuster Video, whose business model was based on guaranteeing that popular new releases would be in stock. This meant that in place of a varied selection of movies, most of a store’s finite brick-and-mortar shelf space housed hundreds of extra copies of Tin Cup (1996). The pace of the U.S. mail system meant that Netflix couldn’t deliver customers the instant gratification of seeing the latest new releases exactly when they wanted to see them but instead had to borrow something like the model of the local independent video store—an institution then on its last legs due to competition from Blockbuster and chains just like it. Especially in cities and college towns, Video Library (or Barn or Vault or whatever) might have had a more varied, less sanitized collection than the chains, but it made no guarantees about what you’d find, mainstream or not. What Netflix realized was that the size of their collection didn’t matter if their typical customer didn’t want to watch the movies they had; if your queue of soon-to-arrive discs was empty, you’d probably cancel your subscription.

In order to compete, Netflix had to expand their subscribers’ taste. They had to transform them from movie watchers to film buffs. You might not be able to rent Hope Floats (1998) right this minute, but maybe you’d be interested instead in Andrei Tarkovsky’s 1966 ethereal medieval epic Andrei Rublev?


Nagisa Ôshima’s Max, mon amour (1986) is about a well-heeled upper-class couple in which the wife (Charlotte Rampling) takes a chimp as her lover, but what’s unique about the film isn’t the concept; it’s how it’s followed up on. When Anthony Higgins’s husband character gets over his initial masochistic fascination with the mechanics of interspecies sexual congress, the film is signaling that we should too. As it becomes a comic domestic drama about a throuple rather than the broad absurdist comedy that the premise suggests, it prompts a viewer to look twice at the easy habit of imprinting our first response to a situation as the best one.

Max, mon amour isn’t currently available on any major streaming platform, but it’s pretty good!

User Scrunch

From its inception, few companies have relied so heavily on algorithmic suggestion as Netflix. “You might also be interested . . .” personalization is ubiquitous now, but at the time, few companies enmeshed it as deeply into the experience of using its product, to the point that Cinematch—the company’s in-house recommendation algorithm—with its increasingly atomized quantification and analysis of your movie selection, began not only to reflect your taste but to define it. Maybe you didn’t realize that you liked, say, violent movies based on books with a strong female lead until you rented The Hunger Games (2012) and Cinematch suggested you rent obvious classics of the genre like Carrie (1976) and Silence of the Lambs (1991), but also semi-obscurities like Lady Snowblood (1973), Let’s Scare Jessica to Death (1971), Morvern Callar (2002), and Woman in the Dunes (1964). And while from a helicopter-shot-level view, this was the same old business of superior entities wielding undue influence on mass culture, for once these entities were encouraging this culture to expand beyond their own preferences. Scrunching users into the deeper corners of its catalog meant the company was asking them to dig within the strange, the useless, the beautiful dead ends, the stories that had either fallen out of the collective unconscious or had never really made it there in the first place, to find something new to love.

In 2006, realizing its importance to the company’s success, Netflix decided to offer a million-dollar prize to anyone who could improve Cinematch’s outcomes by 10 percent, making troves of anonymized customer-rating data publicly available. It took almost three years, but two separate collectives both hit the 10 percent mark virtually simultaneously (an arcane tiebreaker decided who won the money). Despite the fact that they’d discovered a way to vastly improve how customers interacted with their product, Netflix never applied the winning algorithms to their DVD recommendation technology. Our taste, it turns out, isn’t a fixed characteristic. It’s dependent on circumstance.


There’s a section in the based-on-a-true-story independent feature Chameleon Street (1989) in which conman Douglas Street poses as a doctor, referring to a textbook as though consulting an Ikea manual before meeting with patients, and even administering a hysterectomy while interns are instructed to note “the speed and dexterity of Dr. Street’s work.” The underlying idea here is that Street can easily gain expertise, even in supposedly rarified subjects, by attaining credentials that certify that expertise is next to impossible. Successfully performing a surgery makes some people criminals, not doctors. The film’s metanarrative reflects this moral; it won 1990’s Grand Jury Prize at Sundance but, mystifyingly, wasn’t distributed for years afterward, and its director/star, Wendell B. Harris Jr., never made another film.

Chameleon Street is currently available to stream on the Criterion Channel. It’s pretty good!

Listen, I was being reductive when I was talking about the tough time I was having a few years ago. It wasn’t just Connecticut (although Connecticut sure didn’t help). I had a shit job where I was yelled at all the time, so I quit, and then I was broke. I joined a writing group full of old people who workshopped my stories by telling me how much they hated them. My mom died. I went to protests but found myself socially and politically disconnected, and also I realized I was too much of a coward to block I-95. I learned that I was being dismissive about the difficulties of first-time parenthood that lay ahead of me. And though I didn’t know it at the time, cancerous cells were taking over my right nut. So I had some decent reasons to be bummed, but none of these issues would have been unbearable if I hadn’t been far from my friends. That’s the big thing. I was lonely.

You’ve probably had times in your life when you were dealing with similar essential sorrows—disconnects between you and the world you aspire to—just as viscerally, or more so, as I was at the time. I was lucky. I knew how to reestablish that connection. It didn’t even require much effort, which is essential when you’re too bummed to get off the couch.

I’ll say it again: When you’re low, can I tell you how good it feels to see yourself in someone else? Can I tell you how good it feels just to like something?


A text image reads “Netflix and Nil.”

Use It or Lose It

By 2009, Netflix had begun to shift customers to their streaming service (then called “Netflix Instant”), which gave the company access to a bounty of data. Before, they didn’t know if someone watched a movie all in one sitting, what time they watched it, or what else they were considering afterward. One group competing to improve Cinematch began categorizing people into different identities based on when, where, and who they were watching with—defining “Eric alone” as a completely different person from “Eric on a date.”

I’ll say it again: When you’re low, can I tell you how good it feels to see yourself in someone else? Can I tell you how good it feels just to like something?

The shift between movies as discrete pieces of physical media to something intangible seems like a lateral move. It shouldn’t make much of a difference how a movie gets to your device—the screen is the same, who you’re watching it with is the same, the snacks are the same. But Netflix had stumbled upon an easily exploitable difference between the two platforms. Whereas Netflix DVDs, out of necessity, encouraged a user to interrogate the depths of their own taste to explore a vast catalog, Netflix Instant set out to disguise a paltry streaming library by wresting conscious control from the viewer. By 2016, titles autoplayed after just a few seconds of inactivity, guiding a customer, through apathy or inertia, to watch a title whether they wanted to or not. The thumbnails displayed for its titles changed based on what the algorithm perceived to be individually enticing (my personalized thumbnail for the 1993 John Grisham adaptation The Firm currently shows Gary Busey looking agitated in front of some Venetian blinds; Gary Busey is billed tenth in The Firm’s credits). Netflix Instant even actively discouraged the barest hint of nuance in the response it asked for from viewers, downsizing a one-through-five-star user rating system for DVDs (including an option for written user reviews) to a binary, thumbs-up-or-down request, diminishing a customer’s input on what they watched to something more akin to the feeder bar on a lab rat’s pellet dispenser.

The platform was now disincentivizing exploration, with all its dead ends and meandering discovery, its user interface having become about implementing a “frictionless” experience. It made the path from deciding to consume a product to actually consuming it so easy as to be almost unconscious.

There would be no surprises, and your taste was moot; it was convenience choking out idiosyncrasy. Your queue of DVDs had been aspirational: it was for connecting with who you wanted to be. Streaming was for forgetting who you really were.

There’s a part in Wade Davis’s ethnographic study The Serpent and the Rainbow that uses the Haitian concept of the zombi and the belief systems that undergird it to talk about the construction of consciousness. He describes how sailors used to navigate by looking at Venus, clearly visible even by day. This planet is still there. It hasn’t exploded. It’s still in the same relative position in the sky at the same time. But according to Davis, with the advents of the astrolabe, more accurate sea charting, and satellite GPS systems, humans can no longer see it. As far as collective perception goes, if we don’t use something, it doesn’t exist.

The Serpent and the Rainbow was (very) loosely adapted into a fictional horror movie by Wes Craven in 1988. It’s available from multiple streaming services for $3.99. It’s pretty good!

Our Brave New Slippery World

I believe that there’s value in obscurity. Unearthing what’s been buried and what’s been lost is how you find other voices: the ones muffled because they come from people whose identities are actively being suppressed; the ones that make fun of naked emperors; the ones belonging to freaks and weirdos and losers that speak of things too depraved to stay on the surface for very long. It grants second chances. It channels the dead, letting them speak once again. This isn’t to reflexively trash what’s popular—the hits often rule!—but to suggest that the popular and the obscure can complement each other, like sun and stray planetoid.

Today, obscure movies are more accessible than they’ve ever been. You don’t have to live near Los Angeles’s Vidiots or Hamden, Connecticut’s Best Video to see something wild. (Something Wild [1986] is available to stream on Prime Video. It’s pretty good!) Streaming services from the Criterion Channel to Mubi to Night Flight curate selections of arthouse and cult movies for a small subscription fee. You can pay a couple bucks to stream thousands of films from Amazon’s surprisingly large library. If you don’t mind the occasional lousy transfer, you can go on YouTube and in a matter of seconds start watching anything from Aladdin and His Magic Lamp (1967), a Soviet children’s movie that beautifully retells the Arabic fairy tale using low-budget practical special effects, to Bill and Coo, a 1948 rom-com starring birds.

The problem is that subscribing to niche film services and hunting down bird-centric romances on YouTube are the acts of a self-selecting group—it takes money or effort or both. This ghettoizes lovers of the strange into a “fandom,” that most pejorative of internet taxonomies. Classifying love as fandom damns the most affecting and holy of reactions to art—recognizing yourself in others—into just another mode of material consumption. It turns lovers into shoppers and then sells them Funko Pop figurines of John Cassavetes. It turns connection into loneliness.

But looking for new old stories could be for everybody. It doesn’t have to be exclusive; you might not even have to try hard to do it. It can be for your parents, your neighbor, the dudes on the stoop, all the people you left your small town to get away from. In whatever infinitesimal way, if we all did it, it might make our world a little more like us.

In 2006, the new Morgans and Vanderbilts of Silicon Valley hadn’t yet realized that algorithms would confirm their place as our social betters, in control of the boundaries of our taste. It wouldn’t become clear to them until they realized how personal mercuriality added an untenable level of friction to our brave new slippery world. So they had to remove it.

Five years ago, 45 percent of adults responded that they “very often” or “always” used a second screen—i.e., their phones—while watching TV, voluntarily fracturing the immersiveness that’s always been a part of seeing movies. I’d guess this number has only increased. In internet-speak, the word Netflix has been memed to become part of a phrase that means you’re going to ignore the movie you’re watching to fuck in front of the TV. On its streaming platform, Netflix currently provides access in the United States to under seven thousand titles (down from around one hundred thousand available on DVD). The company produces over half of them through its in-house studio. It doesn’t do Netflix any good to incentivize us to plumb the depths of their library; the more our taste is constricted, the less we’ll question if we’re actually interested in what they say we are. And if we forget the shape of our collective dreams, it’ll be a lot easier to take us out of the dream-making process entirely.


Brewster McCloud (1970) is one of Robert Altman’s uncanny and wonderful early movies. A Harold and Maude-era Bud Cort is hiding in the bowels of the Houston Astrodome while he works on a set of homemade wings. He’s being hunted by a Bullitt-esque cop who suspects him of a string of murders. The film is punctuated by lectures delivered by a professor who is turning into a bird. Shelley Duvall talks memorably about diarrhea.

Looking for new old stories doesn’t have to be exclusive; you might not even have to try hard to do it. It can be for your parents, your neighbor, the dudes on the stoop, all the people you left your small town to get away from.

Though it’s way, way more unrestrained than Altman’s uptight stage films a decade later, Brewster McCloud shows a lot of the same concerns. Brewster’s ambition isn’t limitless. He doesn’t want to fly in open sky but in the closed confines of a domed baseball stadium—a boundary on possibility just as confining as restricting a film to the size of a stage set. Imagination, dreams, creation, even magic: none of these help us to escape ourselves, but instead enable us to go deeper into the everyday, to help us understand what’s right there in front of our faces. And maybe to transcend it, just for a moment.

There’s an idea that dreams exist so that we can process the parts of our lives that are difficult to make sense of consciously. We tell ourselves stories in our sleep so that we can integrate the incomprehensible into our waking lives.

Who—or what—are we letting define the borders of our dreams?

Why are you so sad? Why am I?

You can currently stream Brewster McCloud on Prime Video or Apple TV for $3.99. It’s pretty good!