Podcast and Liveblog: “Play in the Age of Computing Machinery” with Miguel Sicart

Miguel Sicart is a games scholar based at the IT University of Copenhagen. For the last decade his research has focused on ethics and computer games, from a philosophical and design theory perspective. He has two books published: The Ethics of Computer Games; and Beyond Choices: The Design of Ethical Gameplay (MIT Press 2009, 2013). His current work focuses on playful design, and will be the subject of a new book called Play Matters (MIT Press, 2014). Miguel teaches game and play design, and his research is now focused on toys, materiality, and play.


Play is at the center of our culture, but we don’t really know what it is. What can play be, and how does it relate to computers? Play Matters (Miguel’s forthcoming book) is a manifesto, a romantic take on play. It takes on three modes of expressing play in the world: play as submission, play as resistance, and play as (computational) expression. Instead of talking about the engineering part of play, Miguel thinks we can take lessons from literature. His background is in philosophy and comparative literature (before he moved to game studies), and books are a very good lens for seeing the relationship between play and computers. There are three in particular that guide this talk.

Don Quixote is the story of a man who goes crazy because he reads too many books. His reality is constantly clashing with fiction. It’s a sympathetic book, though. The book ends with him recovering his sanity, and for a reader it’s a sad ending–he should have died in the world that he created in his imagination. As such, Don Quixote is a metaphor for creative and resistive play.

Toby Shandy is Tristram Shandy’s uncle. He is an empathetic character, because his too-logical attempts to understand the world break down. Toby Shandy helps us explain gamification.

Finally, Ulysses is an attempt to create the reality of one day in one particular location, using language. But instead of using language to show what the world was or is, it uses language to create reality: it generates the world. Ulysses explains many things about the way we understand computing and computation in the world.

These are three lenses through which to explain the relationship between play and computers.

Play and Computation

As soon as we had computers, we start wanting to play with them. If it weren’t for World War II, games would probably have been the first thing we made with computers, and we would have started making them even earlier.



SpaceWar! happened very early on in the history of computing machinery. Then as soon as we had network computers, we had text generating games like Colossal Cave. And as soon as networks became better, we had MUDs. Once we had 3D rendering, we had Myst. Then when networking technology got even better, we had MMOs like Ultima Online. We then used these networked computers to create social networks, and games for them (e.g. Farmville). And with even more increased bandwidth, we now broadcast the games, performing them for other people (Twitch.tv). This evolution in the way that we play with computers has happened because computers themselves evolve.

Screenshot of Twitch website


The evolution of technology and computation is always hand in hand with play, and the things we do with play. Network games are always about the spectacle (showing how good you are), but Twitch.tv changes the game because you can broadcast it. Computing and play are still hand in hand.

What can computers do? Computers are very silly, dumb machines. They can do four things: they can perform calculations really fast. They can store and manipulate data. They can sense the world (translate analog input into digital data)–it’s no longer a dumb box; it’s fairly aware of where it is and what kind of environment it’s in. Finally, computers can network, increasing their capacities.

These four capabilities drive us to think about reality in a very particular way, in a way I will call the Ulysses paradigm. We’re thinking about them in the same way Joyce thought about language. We are thinking about a world can be sensed and translated. It is the same ontological reduction that Joyce wanted to use with language; he pushed language to create a reality (i.e. Dublin in 1916). We need to let computers create that world, or we need to create that world for them. They are ontologizing machines.

Interestingly enough, that’s the same thing that we do with play. The purpose of play is to create a sensible and translatable world. This the a historical idea of the magic circle. The magic circle would be a formalized structure that helps us see what play does, but it’s more an idea of a creation than a space, or a co-creation. We create this world together. So there is an analogy between Ulysses, computers, and play, as each is creator of reality. With each, we generate a world in which some actions are relevant, some are irrelevant, and these actions are meaningful within that context; they are upheld by their own nature. The Ulysses paradigm is a connection between ontological machines and play as a creator of worlds.

What happens when we play games with computers? When humans and computers overlap, we get Huizingan games. The danger of the Ulysses paradigm is that rules are epistemologically invulnerable. You cannot change or modify the rules; anyone who breaks the world is outside the game, as it destroys the reality. It’s much the same with computers: if you don’t give it the necessary rules, it cannot compute. Computers need to operate in that way, because they are fairly dumb machines. We are not dumb machines, yet we are creating Huizingan games. We are borrowing from computing rules and translating them to the rules that we play by. This is a problem.

Computable worlds tend to be playable worlds because they have clear rules, similar to the rules we use when we play. Both play and computation reduce the world; they create these circles of magic, bounding the world. It’s not only a reduction of the world, but it’s a required reduction for the creation of worlds. We reduce and create. It is also interesting to think of play as a language. Computers require languages to generate the world, and computers have the same traits as language, the same ontological capacities.

The Seductions of Computable Play

In both theory (game studies) and practice (game design), there is a deep fascination with computers and play, and this is slightly problematic.

Toby Shandy can help explain. In the Battle of Namur, Toby Shandy injures his leg. So in the novel, he is constantly creating a reproduction of the Battle of Namur. He insists that if he can reproduce the battle, he can understand the reason for his injuries, and therefore heal. Likewise, if we reduce the world to buttons we can press, we think we can understand the world. But this submission to computational play is a negative thing.

There are three particular seductions. The first one is that the world is computable. This is what Toby Shandy is doing. It assumes that computation does not have social, political, or ethical implications. If we want the world to be computable, we can’t bring in those kinds of messes; these are at odds with computation. We presume that we can build models with predictability, that models will respond to our predicted behaviors. Computers are so good at storing and producing data that seems clear and incontrovertible that anything that can be modeled will be predictable.

The second seduction is that play is akin to computation, that it can reduce the world to repeatable patterns, with clear goals, behaviors and rewards. The world of play is extremely clear because it’s very close to computation. This type of play is always engaging: the world is dull and complicated, but if we reduce it to these patterns of engagement, compulsion, and pleasure, then we will engage and play with the world in different ways. Since play has a history of being a positive thing, we will get all the benefits of it.

The third and perhaps most dangerous seduction is that we can solve problems by computing them, that computational play can be an agent of change. We can engage and motivate if we play; what would happen if, instead of work, we could play? If we earned our promotions as badges, and explored via rewards and points? What if we could actually change the world through play? This takes the idea of a computational world, couples it with play, and results in real change.

But What is Play?

A Brazilian critical theorist named Paulo Freire wrote about the banking model of education, saying the problem with education is not in the teacher or student, but in the system. The teacher is the bank, and we go to the teacher to extract knowledge. Similarly, the game is an authority; the computer is an authority. A banking model of play is a submission of ourselves to material reality. We are not really playing, but giving a lot of agency and epistemological value to a computerized vision of the world. We go there to extract knowledge and get value out of it. Do we really engage or play with games, with computers?

This is defining play as epistemically invulnerable. Computers can’t deal with ambiguity, but we submit ourselves to that seduction, to the idea that rules cannot be questioned or contested. That is a marginalized, bounded type of play. The very authority of the computer game would mean that questioning it would break its own validity. Toby Shandy was gamifying his injury; he had a logic, and once the world was logically reproduced, he would be able to understand his injuries and therefore heal.

Maybe these are actions that change the world. We can call it play, but it’s more like computational play. It is a reduced, uncritical type of play. This particular understanding of play is limiting our understanding of play, falling into the Huizingan form of play, in bounded environments. But what if these rules are not invulnerable? What if play is like a language, and we use the capabilities of play differently?

Quixotean Play

The first part of Don Quixote was published in 1605, and it immediately took off. It became a universal novel, a universal way to understand the world. In 1614 Cervantes published a sequel, because somebody else had written an apocryphal sequel. In Don Quixote 2, Don Quixote meets characters that have not only read part 1, but also the fake sequel.

What transpires in part 2 helps us understand the world of play and computers. A group of dukes decide to build the world that Quixote envisioned–they built the medieval world of chivalry that he had dreamed of. But Quixote is not happy, because it’s not the world he’s created. Somebody else is putting worlds in his own madness, and he’d rather create the fantasies on his own. The dukes were just constructing this alternative world, one in which he had no agency or creative capacity. It was a banking model of insanity. He wanted to create the world he imagined, and this is what we do in Quixotean play; we are creating worlds, and not strictly by the rules we are given, but by acts of submission and rebellion.

One great example is Twitter bots. Twitter is a wonderful social network, but we build automatic machines that post in it. We take over a social network, and turn it into an antisocial network. We do not embrace the insanity of having to express ourselves; instead we take over that world and show its own foolishness. We feel and project love, sorrow, humor, creativity, and passion in these bots. The act of building a bot is a window into the world of Twitter, and it helps us to understand it; that is play.

There are four characteristics of play that help us to rebel, in a way, against computational play.

Play is appropriative. To play is to take over a particular situation, to modify and spin it. Sometimes we submit, but sometimes we build a new world around it. Skater culture is a good example of appropriation; even when skate parks are built for them, they continue to skate outside of the parks, showing that the world is always playful.

Play is also always expressive. It has the capacity to create worlds through expression. It is a language. Therefore, it is always deeply personal. There is no detachment, no magic circle.

Finally, play is always autotelic. It is always for a particular purpose that stops when we are not playing. We think of play as being for play’s sake, but actually we are playing through a negotiation and engagement. The purpose of play is to play, but we are always discussing what play is. We know when we cease to play.

This helps us see where play exists far beyond games, in other activities that are much more interesting. Computing history has been about submitting to the machine, but in other fields, we find playful appropriation of technology. Contemporary artists like Nam June Paik can help us think about a more subversive way to play. We don’t have to submit to play, but we can engage with technologies in a playful way. We can play based on the characteristics of play, rather than the characteristics of the devices; by using play as a way of creating the world, sustained or supported by computers, rather than by submitting to them.

Computers themselves are machines of play. Computers are playing with machines. The Turing machine is a machine of appropriation; it takes binary processes and turns them into a machine. So even computers are appropriative, by definition.


(Pic: newstweek.com)

Newstweek is an art project by Julian Oliver and Daniil Vasiliev. It is a small, self-made device that is designed to be placed in public, open wireless networks. It then modifies the headlines of news sites for any computers on the network. Isn’t this what computers should be doing? We are relying on computers to process the world for us, trusting that they are true, but any computer can be appropriated. Newstweek has a strong political and aesthetic message, but also a playful one. They play with the computer, but not against the computer; it requires deep technical knowledge (therefore collaboration with a machine), but not submission. Play allows us to see computation as production, and not just consumption.

Play is probably the dominant way of being in the modern world. We have positive associations with play, we put it on a high pedestal. If this is true, and we can claim that we see play in this way as a dominating interaction with the world, then we should also see computation in itself as play. Information ethics reontologizes the world. One argument is that computers have done that, but this is a dangerous idea; instead we can think about play as a reontologizing activity. Play is a way of taking over the world and making sense of it, of putting deeply complex assemblages together. In the era of computing machinery, we need to think about play as a way of constructing these realities. We need to live in the era of Quixotean computation. Even if computers are epistemologically invulnerable, we always need to try to crack them, to play with them.


Jason Haas: I’m confused as to how, in front of computers, humans are suddenly submissive. Is it really the computers? Aren’t capitalist institutions and computers are being fused in this model?

Miguel: I’m reacting to the idea of submitting to the game as an object from which we have to derive truth from. We are not seeing an idea of negotiation of rules when it comes to computers; because of their computational nature, we cannot discuss it. We can streamline the behavior so we are no longer obsessed with this. This is why HCI is obsessed with seamlessness. Play is always negotiating with materiality, but it should be made aware and present rather than hiding it.

Eduardo Marisca: In Jorge Luis Borges’s “Pierre Menard, Author of the Quixote,” Menard painstakingly rewrites Quixote word for word. Isn’t that a form of play that is limiting? Isn’t trading banking play for Quixotean play just a slightly larger box?

Miguel: It might be, but Pierre Menard is also about interpretation. It’s doing an act of appropriation in an extreme, radical, artistic way. It’s taking an original text and reshaping it, reappropriating it, rereading it. It’s not quite following the model of authority.

Scot Osterweil: You talked about play broadly. What about non-game forms of play? Do we need to privilege those?

Miguel: What I’ve presented today is a rhetoric of play. It does not think about games as the privileged form of play, but as one more incarnation of play. What I’m saying is that we live in this culture of games, and I don’t want to play.

Sasha Costanza-Chock: Wouldn’t the opposite of a banking model of play be a pedagogy of play? To teach people to create play systems?

Miguel: That’s exactly the type of play I’m interested in. I grew up with RPGs rather than computer games; here the act of negotiation is crucial to this approach. Given the boundaries we face with computers, can we have that type of playful relationship with them?

Sasha: So Newsjack might be a better example than Newstweek because it’s more bottom-up, inviting users to subvert the system themselves.

Miguel: Maybe, but I like the “dark play” aspect of Newstweek.

Erik Stayton: What about computer viruses as play?

Miguel: I’ve been thinking about the boundaries of play with regard to botnets, and perhaps they are play devices of a dark kind. It’s a type of play we are not used to. We create worlds with play, but computer viruses are destructive machines. It’s hard to unroot play from a positive, creative angle. Newstweek is not completely negative, whereas viruses are.

Todd Harper: Saints Row IV is a very “meta” game. Your goal is to break the VR world by playing within and with its rules. The simulation gets weaker as you get farther. But you’re still submitting to the machine. What space is there in this rhetoric of play for play that is not necessarily transgressive?

Miguel: Here we are negotiating what kind of submission we are engaging with. This is sometimes very satisfying. The extreme of submission is competitive play; it is the deepest submission of rules you can imagine. There is a pleasure in this, because you’re better at reading/performing the rules than anyone else.

T.L. Taylor: I’m surprised that you don’t use the term “actor.” Computers can’t be cultural actors. They can shape culture, but they’re not interpretive. But you shied away from that language.

Miguel: For me it’s obvious that computers are fairly dumb machines.

T.L.: Your model of play is much more individualistic than mine is. I think of play as a cultural act even when you play alone.

Miguel: I agree, even though I come from a collectivist culture.

Todd: But can computers be social actors in the sense that we can empower them to be social actors?

T.L.: They are social actors, but maybe it’s another thing to be a culturally interpretive actor.

Miguel: I agree, but on some occasions we can give cultural agency to computers within frames that we have agreed upon and that are predefined. The object is not a cultural agent, but we’re allowing it to be.

Jason: I’m curious about how you see Newstweek as something different from a Toby Shandy act. It is a remediation of the world that the designers of Newstweek want, a similar imperial act.

Miguel: The reason Newstweek works is because we already live in an imperial world where we accept these computers. Newstweek forces us to reread the world from a non-Toby Shandy world. So for the designer it might be, but for the user it’s the opposite. If it were in an art gallery, it definitely would be a Toby Shandy act. It can only work if it’s non-consensual, hidden, and disruptive.

Wang Yu: The divide between submission and transgression reminds me of Stuart Hall’s preferred reading theory.

Miguel: Sometimes there is pleasure in submission to whatever the creator has given us, and acknowledging that the creator exists. But I don’t want this to be a universal model.

Wang Yu: In Minecraft, creativity is acceptable.

Miguel: I don’t think of Minecraft as a game, more as a playground. It’s a space where you can build games; it has some rules but it is not a game. It’s an environment created so you can play.

Yu: When paratext comes into the playground, is that Quixotean?

Miguel: Don Quixote doesn’t go against the rules all the time, just when they negotiate the extent of his madness. We need to negotiate the extent of our involvement with play.

Yu: Isn’t any textual analysis of a game a renegotiation?

Miguel: I have a problem with the idea of games as texts. A text implies a particular mode of interpretation, but we don’t play texts. Play is a way of making meaning that we don’t have with texts. I would define the object by the activity. Games are anything we play with, whereas texts are things that we read.

Todd: I’m interested in the intersection between Stuart Hall and fan work. What about fan work that supports rather than fights an existing canon? What about non-productive play?

Miguel: I share your skepticism about non-productive play. It comes from Kant, who thinks about it in binaries, where work is productive and play is non-productive. Huizenga takes this uncritically. Play is nonproductive but at the same time the generator of culture. In the “serious games” movement, there is a lack of critique of these two positions. The productive side is not tied to the activity of play, but instead an externalized notion of it.



Share this Post