To receive the Quarterly Journal, become a member or purchase at our bookstore.
When there are balances to be restored, Vishnu descends to Earth. Our physics can’t contain him, so he must take an incarnation: he appears to mortals as a fish, as a lion, as a Brahmin, or riding astride a great white horse. Sometimes he comes in clusters: 10, 22, or incarnations without number. He comes down — in Sanskrit, ava — and crosses over — tri. In medieval Vedic literature, he is an “avatāra,” a god walking among us. But those are ancient texts, and we’re gods now too. Like Vishnu, we often visit other worlds in infinite bodies, which we also call avatars. We started doing this in 1985, influenced by yet another ancient text — one so old it can only be read with an Amiga, Apple II, or Atari 800 — called Ultima IV: Quest of the Avatar.
Ultima is a series of fantasy video games. In the first three, the player, known only as the Stranger, battles wizards and monsters in the mythical world of Sosaria. The Stranger must be willing to commit violence and thievery; must kill or be killed. The creator of Ultima, a savant named Richard Garriott, went into hiding after the release of Ultima III, reemerging after two years with his most ambitious game, Ultima IV: Quest of the Avatar.In the mid-1980s, two years was considered a preposterously long time to make a computer game — the coding shouldn’t have taken more than a few months. Gamers speculated that Garriott might have run out of ideas, might have gone mad. Garriott milked it. In interviews, he said that Ultima IV was taking so long because he’d received letters from parents about Ultima’s amoral universe, about the negative influence it made on impressionable young players. Ultima IV, he promised, would have higher goals in mind. The Stranger would have no enemies but the self.
Garriott had grown fascinated by Hinduism after watching a television series about the Dead Sea Scrolls that suggested that Jesus Christ had studied Hinduism and become “a very powerful yogi.” Composing Ultima IV, he decided to borrow the concept of avatāra, which resonated with his new game’s objective: to become a manifestation of godliness, an “Avatar of virtue.” Garriott wanted players to feel responsible for their actions, and adding moral stakes lent the primitive graphics a kind of emotional weight. If gamers strived to achieve moral purity and righteousness in-game, he reasoned, it may actually improve them as people. “If someone spends 100 hours playing my game, then I have 100 hours of the input that makes that person what they are,” he explained to his biographer, as only a programmer would.
Ultima IV was different from its predecessors. The kingdom of Sosaria — habitually torn by conflict and swarming with monsters — had been replaced by a new domain called Britannia, which is ruled by a sovereign named Lord British. The Lord sends the Stranger on a quest to master “The Eight Virtues” of Honesty, Compassion, Valor, Justice, Sacrifice, Honor, Spirituality, and Humility, an education accomplished through consultation with mages and bards, meditation at shrines, and the discovery of relevant artifacts across an expansive 8-bit kingdom. Ultimately, the Stranger must travel through a “Stygian Abyss” to secure the “Codex of Ultimate Wisdom” and achieve Avatarhood. Demonstrations of avarice or violence result in penalties. The game is a hero’s quest into godliness.
Ultima IV ushered in anew era of gaming. Its highly intentional moral universe suggested that computer games were no longer dressed-up, pixelated versions of established tabletop role-players like Dungeons & Dragons. Perhaps they might be an art form of their own, with an unprecedented power to influence behavior. Garriott recognized how deeply games burrow themselves into players’ subconscious minds. He felt he had an ethical responsibility to wield that power for good.
Ultima IV was hugely successful, spawning nine sequels, a number of spinoffs, and Ultima Online, an online role-playing world with hundreds of thousands of devotees. But although it introduced the word “avatar” to a generation of keyboard mashers, Garriott’s use of the word was highly specific: an avatar was a player who’d leveled up to a maximum spiritual and moral achievement. That usage didn’t find much purchase outside of the weird mechanics of Ultima IV; the modern-day “avatar” traces its origins elsewhere, in Populopolis, the capital city of a very different computer game. This one was called Habitat.
Habitat was a place to go, rather than a game to play — the first real attempt to build a large-scale, commercial, digital community. Players, known as “citizens,” navigated the virtual city of Populopolis in New Marin, California. With a joystick and some basic typed commands, citizens could interact with objects, one another, and the virtual world itself. The game’s mechanics called for a new kind of relationship between a player’s home computer and the centralized, mainframe computer serving as the game’s host: every citizen dialed into the same host computer, which held the structure of the shared virtual environment, what game designers call a “world model.” And it was a big world.
Hundreds, even thousands, of citizens could experience Habitat at once, manipulating some 200 discrete classes of objects. Every citizen was given an apartment, or a “Turf,” which they could decorate with objects familiar from the real world — doors, garbage cans, books — as well as magic wands, teleportation booths, and a gender-switching device called the Change-O-Matic. Each citizen was also given an “Avatar.” These were not paragons of virtue, as with Garriott’s Age of Ultima Avatars. They looked like boardwalk caricatures: dot eyes, wacky hairdos, comic-book speech bubbles floating over interchangeable heads. An internal 1987 LucasFilm promo video announced that Habitat, which “might become the first game to have 10,000 online players,” would “connect people from all over the country into a single, imaginary, cartoon reality.” The first manual sent to beta testers proposed a tenuous historical link between the noble Ultima Avatars of yore and the goofy pixel pals of Populopolis:
The early Avatars were adventurers like no other …
But as the years passed, Avatars changed. The spirit of adventuring died away, and we became more and more content to do nothing …
You see, the Avatars, when left to themselves, have become basically lazy creatures that would only be too happy to sit or sleep all day and night, lounging in their hot tubs, reading magazines or books, or chatting with friends, but never getting out to see the world because it would take too much effort.
The game ran on Commodore 64 home computers scarcely more powerful than pocket calculators, and the graphics were rudimentary, but Habitat offered a multitude of ways to interact, with very few rules, and that cocktail created a strangely compelling simulation of real life. Citizens nudging their Avatars around with joysticks engaged in all the mundanities of existence — redecorating their apartments, shopping, reading the newspaper — as well as its dramatic highs. They waged wars, fell in love, engaged in complex in-game politics, started religions, and even died.
These “lived” experiences prompted an unprecedented transference of self onto the cartoon renderings. Players were unsure: were their Habitat avatars just game characters, or were they personal projections? F. Randall Farmer, one of the game’s developers, documented some anecdotal data from a face-to-face meeting with a group of about 50 Habitat citizens in 1990: “Half of them said they thought of their Avatar as a separate being,” he wrote, while “the others said it was their self.” The distinction between these two ways of thinking about an Avatar is subtle, but in the case of Habitat, it marks a huge shift.
Think of it this way: say I’m sitting down to play Monopoly with a group of friends. We pull the box off the shelf. It contains a board, a couple dice, some colorful paper money, and a collection of pewter figurines. These are game pieces — our Avatars in this particular game. My choice between top hat, a Scottie dog, a thimble, or a boot confers zero advantage to me as a player. That’s because the pieces are really all the same; they only represent each player’s position on the board. When I pick my favorite game piece, saying, “I’m the hat!” everyone understands, for the purposes of this Monopoly game, that whenever I roll the dice, the little pewter hat will move accordingly. Nobody believes, least of all me, that I am actually a hat, nor that my choice speaks to some inner hatness at the seat of my soul.
The same holds true for many games. Playing Tetris, I don’t confuse myself for a tetromino; I know full well Super Mario isn’t me, even on a good day. Questing across the kingdom of Britannia in Richard Garriott’s Age of Ultima IV, solving riddles and righting wrongs to mold my player into his best self, I still don’t confuse myself for the figure on the screen. But beyond the simple lack of visible representation, games don’t always encourage the ontological leap required to see oneself in an avatar. In a traditional board game like Monopoly, a transference of self serves no purpose. But it’s not really a question of how we play. It’s who we play with.
In May 1990, Chip Morningstar and F. Randall Farmer, the two programmers who built and maintained Habitat for LucasFilm, were invited to speak at the First International Conference on Cyberspace at the University of Texas at Austin. At the time — before the Mosaic browser made the web visible, before search engines like AltaVista or Yahoo made finding anything online even remotely practical — “cyberspace” was an empty concept. The word was a neologism borrowed from science fiction, coined by William Gibson in his short story “Burning Chrome,” and later popularized in his novel Neuromancer;in Gibson’s noir world of high-tech low lives, “cyberspace” was something you jack into, a “consensual hallucination experienced daily” by “console cowboys,” hackers, and “billions of legitimate operators.”
Gibson formed his general idea of cyberspace after watching a group of teenagers play arcade games in Toronto in the late 1970s. He wasn’t conversant in computing and famously wrote his early novels on a typewriter. But he got the sense, watching those kids, that they wanted to go inside the machine. The idea resonated with other writers, who made quick work of turning Gibson’s vision into a robust subgenre all its own. Academics found it fascinating as well. What was digital space? Was it something visible, physical, spiritual? “Cyberspace” was catchy enough to be memorable and vague enough to be mutable. And so, in March 1990, when 50 architects, literary theorists, media critics, and computer scientists descended on UT Austin to debate and share ideas about cyberspace, opinions were as divided as they were colorful. The theorist Marcos Novak explained that navigating cyberspace “is to become a leaf on the wind of a dream.” The conference organizer, Michael Benedikt, an architecture professor at UT Austin, reveled in its non-material qualities, noting that “many of the axioms of topology and geometry […] can be violated or reinvented, as can many of the laws of physics.” But the pioneering Virtual Reality artist Nicole Stenger might have put it best. “Cyberspace,” she wrote, rapturous: “the dessert of humanity!”
Stenger created the first immersive film, Angels, in 1989, a VR experience that ran on a high-end Silicon Graphics machine and featured polygonal angels beckoning and singing in colorful environments. At the time, virtual reality was a cumbersome endeavor involving heavy head-mounted screens, dozens of computers, and touch-sensitive gloves, but the image of a virtual reality “pilot” with hands outstretched, reaching into an unseen matrix of light, was core to many conceptions about the future, and indeed about the nature of cyberspace itself. Habitat’s Morningstar and Farmer were skeptical. “The almost mystical euphoria that currently seems to surround all this hardware is, in our opinion, both excessive and somewhat misplaced,” they wrote. It seemed to them a seductive distraction from the real work of homesteading cyberspace.
Habitat’s cartoon world ran on the inexpensive, consumer-level Commodore 64, a computer that even Morningstar and Farmer acknowledged was “ludicrous,” “vicious,” and basically “a toy,” connected to a central game server by a 300-baud serial telephone connection. Their “virtual reality” didn’t run on top-of-the-line machines, but Habitat’s 8-bit world was more immersive than anything being dreamed up in high-tech VR labs. How else could they explain the fact that one player, a Greek Orthodox priest in real life, had opened up a church inside the game,The Order of the Holy Walnut? Or that another player had willingly taken on a job as the sheriff of Populopolis? That their users logged countless hours, despite the cost? They could only draw one conclusion, which they called their primary principle. Like Soylent Green, virtual reality is made of people.“A multi-user environment is central to the idea of cyberspace,” they argued:
This stems from the fact that what (in our opinion) people seek in such a system is richness, complexity and depth. Nobody knows how to produce an automaton that even approaches the complexity of a real human being, let alone a society. Our approach, then, is not even to attempt this, but instead to use the computational medium to augment the communications channels between real people.
We’re a long way from programming virtual characters or virtual worlds that can interact with players with anything near the unpredictability, candor, or strangeness of other actual human beings, they argued, so why even bother? Habitat proved that, given the right circumstances, human connection colors the picture so convincingly that it can transmute even the humblest hunk of silicon into a city populated by reflections of each player’s truest selves. The intense bond that Habitat citizens forged with theirAvatars revealed that the reality of a digital space has nothing to do with processor speed or pixel resolution. Reality — like hell — is just other people.
This was underlined by what Morningstarand Farmer called “the great debate” of Habitat, the question of whether or not Avatars should be allowed weapons. Death was possible in Populopolis, but impermanent. A fatal blow would only send an Avatar back to where it started, stripped of whatever object it was carrying at the time of death — closer to a setback in Chutes and Ladders than any real violence. Still, it affected how the game was perceived and played. Did death, violence, and theft have a place in Habitat, as it does in the real world? The question hinged on the ongoing debate about the nature of the Avatars. Were they people? Or were they “Pac Man-like” critters “destined to die a thousand deaths”? Was Habitat murder a crime? Or was it all just a game? Opinions split on this point along the exact same lines as in the debate about Avatars: half of Habitat users believed that “murder was a crime and shouldn’t be a part of the world, while the other 50 percent said it was an important part of the fun.”
It’s not surprising that these questions were divisive. In order for the fantasy to hold together, players had to invest themselves completely in the collective world model — its virtual consensus reality. This made dying in Habitat more offensive than dying in Pac-Man; community creates accountability, and accountability requires trust. To know and trust someone in a shared virtual environment, you must be relatively certain that their Avatar expresses a persistent selfhood. It’s one thing to die in a game — it’s part of the agreed-upon risk — but dying in a chat room is something else entirely. In Habitat, Avatars weren’t Pac-Men or Monopoly pieces. They were people.
The Habitat pilot program was shut down in 1988 and rebranded as “Club Caribe,” a virtual island nation for users of Commodore 64’s online service, Quantum Link. LucasArts then licensed the software to Fujiutsu; Fujitsu Habitat launched in 1990 and enjoyed many years of popularity in Japan. Habitat was a predecessor to what are now called MMORPGs, Massively Multiplayer Online Role-Playing Games. Where Habitat failed, games like Second Life, World of Warcraft, EverQuest — even Ultima Online, a descendant of Richard Garriott’s Ultima games — would succeed. With the exception of Second Life, these MMORPGs were (and are) largely task-oriented, with risk and reward commensurate to user investment in ongoing narratives, quests, and battles.
A parallel evolution of the avatar was more explicitly social. In the text-based chat environments of the early net — Multi-User Domains and Internet Relay Chat Channels — a kind of textual avatarhood was developing. Though this version of the avatar was expressed via detailed third-person description, it caused similar confusions between character and self, as the freedom to create highly specific character descriptions inevitably became a form of self-expression, or a sublimation of latent characteristics not expressed in day-to-day life. Gender play was, of course, rampant; in one early Multi-User Domain, a text-based world called LambdaMOO, users could choose between no less than 11 genders, and often vacillated between genders depending on the moment. In this sense, the textual avatar was never static; rather, it was an ever-shifting reflection of identity, desire, aspiration, and idealistic projection. In an early study of identity on IRC the Australian scholar Elizabeth Reid noted that “the chance to escape the assumed boundaries of gender, race, and age create a game of interaction in which there are few rules but those that the users create themselves.” And even if such forms of online representation weren’t completely accurate — someone cruising around a chat environment calling themselves SexyLady69 might well be a teenage boy — they nevertheless allowed early Internauts to plumb their identities, create virtual embodiments of hidden personality traits and desires, and to “try on” new selves.
With faster connection speeds came visual chat rooms where users could express their personalities more overtly through graphic avatars, pixel-resolution representations ranging from inanimate objects to cartoonish “dollz,” animals, and abstract swizzles. In a popular mid-’90s chat environment called The Palace — in many ways an inheritor of Habitat —avatar customization was something of an art form, with entire rooms devoted to the making and trading of novelty avatars, props, and accessories. The psychologist John Suler, one of the earliest academics to make an in-depth study of online behaviors, taxonomized Palace avatars by the personality types they expressed (narcissistic, histrionic, masochistic, and so forth) and noted that many users built collections of avatars, often numbering in the hundreds, to reflect myriad distinct aspects of their personalities. The most intimate avatar in such a collection was the “real face av,” an image of the user’s human face, generally only shared once users gained a certain degree of trust. During “face nites,” Palace users might “step out of their masks and out of their anonymity,” wanting to be as “real” as possible — a poignant experience of virtual intimacy for the users involved.
As with Habitat, in the Palace, anonymity brokered intimacy. Only after interacting extensively from behind masks did users feel comfortable revealing their true selves to one another, and only briefly at that: for just a moment, to cement a friendship, or for a night, in the company of trusted virtual friends. Today, of course, we show our faces more freely; avatars are not a significant part of online life for most people. This is both a cultural trend, normalized by the ubiquity of social media, and a consequence of the market, since one of the few ways to make money from a free social platform is to turn its users into the product, selling demographic information and targeted ad space to advertisers. The commercialization of the web has sent us down an inexorable path toward a culture of authentication. The avatar has been replaced by the Twitter handle, the Facebook profile, the Instagram account. These platforms provide a different kind of opportunity for performance but deny users the freedom conferred by an avatar — the freedom to be the character you want to be, to try on new identities. This particular kind of freedom disappeared when users began to migrate toward platforms that tether our real-world identities to our words and actions onscreen.
Few technologies have separated words from flesh so thoroughly as the internet. In the early days, bodies were a mutable concept, limited only by the imagination. But if we are to participate today, we must do so publicly, our identities tethered to our names in mandatory fun. As ever, those who do seek anonymity often do so to exploit the precarious privacies of our digital lives to shame, troll, and slander without consequence (to say nothing of the vast morass of bots, with their manufactured-to-seem-real identities). That shift — in which pseudonymous avatars no longer liberate people to explore themselves but serve instead to obscure the identities of hateful users and facilitate a culture of unreality and uncertainty — is the single most fundamental shift in network culture since text gave way to image. We all have “real face avs” now, but they confer very little about our true nature. Our faces have become the masks.
Chip Morningstar and F. Randall Farmer, architects of the avatar, understood this early on — that a low-resolution space with limited options for self-representation can be more revealing, more true, and more profoundly immersive than one explicitly designed to feel “real.” Strong-armed by our social platforms into giving too much of ourselves away, we have responded by doubling down on the superficial — meals, whereabouts, signifiers of class — and have lost the capacity to express our identities with any imagination. This kind of playfulness persists in video games, albeit within commercial constraints — the average Fortnite player has spent $84.67 on cosmetic items to personalize their characters, with the lion’s share of that money going to outfits — or through mobile avatar-creation apps like Bitmoji, which charge for extended customization. Perhaps we shouldn’t be surprised. “Cyberspace may indeed change humanity,” Morningstar and Farmer wrote, back in 1996, “but only if it begins with humanity as it really is.”
Claire L. Evans is a writer and musician. She is the author of Broad Band: The Untold Story of the Women who Made the Internet (Penguin Random House).