So a few weeks ago, I flew out to Malta to give a keynote at the first academic conference I'd been to in years - the Philosophy of Computer Games. It's pretty daunting standing up in front of a room full of professional philosophers, but I gave it my best shot and figured it'd be interesting to post the talk up here and see what y'all make of it... enjoy!
The unreliable philosophics of The Chinese Room
This is a talk about the games we make and some of the ideas behind why we make them. It’s been said there’s a strong philosophical angle behind what we do, and although that’s something better left for other people to decide, certainly I’ve got a strong personal interest, so it would be strange if that wasn’t the case. Either way, what we do tends to be quite obsessive about things like ambiguity and abstraction, and I’m going to spend the next forty minutes or so just unpacking that slightly and offering some view on how it ended up that way. And it starts with a problem:
Part one: On the paucity of absolutes
The problem – visual representation is an inevitable reductive operation on the field of possibilities, as part of the natural process that actuality is a subset of possibility. That’s a pretty formal way of saying that it’s really hard in a visual medium like games to preserve some aspects of world-building, character, mystery, poeticism, that written mediums like literature or poetry take for granted, and something very important tends to be lost in that. And I guess, what we’re really talking about today, and the things we’ve been exploring in our games, is how to keep that in place.
To explore this as a starting position, consider these two: Ernest Hemingway and H. P. Lovecraft.
So Hemingway famously wrote a microfiction, or short story formed of six words: “For Sale, Baby Shoes, Never Worn”. In just six words a story that has extraordinary power is told.
Although actually, we’re not talking about Hemingway, but William R. Kane, who published a piece in 1917 called “Little Shoes, Never Worn”, itself based on a 1910 article in The Spokane Press.
Anyway, the point being that whoever you attribute it to, these six, or four words convey a complex set of narrative openings. But do they contain an actual story, an absolute, or just invoke a set of highly engaging potentials? That’s the kind of question that can drive narratologists into frenzied rages, so we’re going to bypass it, as it’s less interesting to me than an unarguable and for me, more interesting way of thinking about it: they contain a possibility space that is highly inviting to the imagination.
More so, perhaps, than the actualities presented by the article itself.
It is a mistake to consider that more is more. Poets have been banging on about this for centuries. The surgical positioning of the correct atomic unit can open up a possibility space that extensive description cannot help but to close down.
Let’s take another example. Although Lovecraft tended to the verbose (in some ways, he’s one of those writers who just didn’t know when to stop), what’s really interesting is that his descriptions of his eldritch horrors were full of holes. Let’s consider the Flying Polyps. What does this actually tell us about them?
We know for sure, for absolute, that they can fly, have no wings, are sometimes invisible and other times a bit plasticky, they whistle and they leave large, five-toed footprints. Actually, no, we don’t know that last one, only that the footprints are associated to them, but we don’t know by whom.
But what does this have to do with the paucity of absolutes and what does this have to do with games?
Games are representational engines. They expose, they reduce, they present actualities. They cannot help but be machines that present actualisations. And this reduces the possibility space, and this in term narrows the imagination and this, in turn, the space we have to play with, the realities we can present.
It’s not just about the limits of our visualisations. This shot is from Call of Cthulhu: Dark Corners of the Earth, which is one of those fantastic-but-flawed games that transcends its flaws and ends up just pretty fantastic. This isn’t about either recognising the limitations of a game made over ten years ago, or about a critique of Headfirst’s vision. It’s about how impossible it actually is to visualise a Flying Polyp in a way that does any justice to the source material.
It’s about how we can present a reality in which a flying polyp can exist that retains the weird openness of Lovecraft’s description.
It’s about whether "Little Shoes, Never Worn" and a game can exist in the same representational space.
Part two: Minding the gap
And that’s just GOT to be something worth exploring, right?
Rene Magritte would have been an interesting game designer. There are people who carry his absurdist principles into game design for sure. Bennett Foddy and Pippin Barr are two that spring to mind, or you’ve got the high surrealism of Sky May Be, the DOOM mod that’s like a bad acid trip with a shotgun.
But to steal his most famous image, this is not a digital system.
And this is not a brain.
Games – computer and video games to be clear, in case the board or paper RPG or LARP designers and players in the room get upset – run on digital systems so simple that they can fit in a box. The soul of the underlying technology is a 0 and a 1, a yes or a no, a true or a false, an actual or a non-actual. We talk emergence like it’s a natural, inevitable thing, but the reality of emergence is that, like consciousness being a by-product of neurons firing, emergence is a by-product of systems that are not, in themselves, emergent. Unlike players.
That’s not entirely true, or entirely fair of course. But we’re talking about the paucity of absolutes, so we’ll rest there for a moment.
I have the rocket launcher equipped or I do not. I fire it or I do not. There is ammo in the clip or there is not. The missile hits the Cyber Mancubus or it does not. There is no condition in this scenario for Lovecraft. There is no allowance for the rocket to both hit and miss, for it to both have ammo and no ammo. There is no Schrodinger’s DOOM.
And of course, it doesn’t actually matter, right?
Years ago, before I left academia, I spent a lot of time thinking about ritual, particularly Victor Turner’s work on liminal spaces, and how reduction was an inherent part of the power of ritual. A ritual relies on codified, formalised, symbolic behaviour – that is to say, reduced behaviour. In order to take part in the ritual, an actual, prescribed set of actions are boiled down from the overall set of possibilities. No reduction, no ritual. It’s a set of rules, of mechanics.
Games are so enjoyable, in part, perhaps, because they are ritual spaces. They map out a set of reduced available actions and we buy into them. DOOM is so much fun precisely because we can flow through a simplified set of responses and predict successfully how the system will respond.
It’s one of the reasons, as an obsessive FPS player, Fallout 3 annoyed the crap out of me. I was presented with an FPS system that forced my shots wide because even though I was good enough, my character wasn’t. It broke the ritual because my reduced behaviour wasn’t met with system predictability. So let’s not throw the baby out with the bathwater. Tread carefully. Mind the gap.
Part three: Into unreliability
After all, the question isn’t about fixing games. A world without DOOM would be a far poorer world. This is a conversation about whether games can bridge the absolute and open up the possibility space for the player. How do we get the power of the gap, the whole range of the player’s imagination into a system that is inherently reductive and absolutist.
It’s easy, actually. This is a non-problem. You’d think.
Let’s talk about Pyramid Head. In fact, let’s talk Silent Hill in general. The genius of Silent Hill is that it’s a mess, narratively speaking. It sort of adds up, but it’s also full of holes and counter-turns and contradictions and stuff that just seems to have landed in there because it felt right for that moment whether or not it really fitted well in the overall grand arching narrative or world mythos.
Silent Hill is non-sense, in the proper sense of the word. It’s a world of collapse, of fragments, of the gap. It escapes its mechanical genetic blueprint and flees into unreliability. And it’s a wonderful, wonderful game because of it.
My favourite poet is Galway Kinnell. I love his work because I don’t understand it, it refuses to boil to an absolute, it keeps shifting meaning, but it stays firmly fixed into my mind. When he writes
And a wind holding / the cries of love-making from our nights and days / moves among the stones, hunting / for two twined skeletons to blow its last cry across
the power does not lie in the literal meaning, but the cascade of imagery, the fluid shift from possibility to possibility.
In Everybody’s Gone to the Rapture, Stephen throws the shoes of Howard, a war veteran in the deep throes of post-traumatic stress disorder, over a telephone wire to make him stay in the junction box. He does this to make Howard stay, even though it makes no real sense. He does this because he knows he is sentencing his lover and their child to death and somewhere in his mind, if Howard stays, Lizzie will call and he can warn her of the approaching bombers.
But none of that really matters, not really.
What really matters is the image, the shoes thrown over the wires. The possibility space of that image. It was Galway Kinnell who threw those shoes over the line.
Games don’t generally like the unreliable, because it damages the ritual. My aim was unreliable in Fallout and I hated the game for it. I hacked the game and boosted my weapon skill to 99, not even thinking I was cheating, but because I was frustrated that the system wasn’t playing fair. It had introduced unreliability into a system that was supposed to be reliable.
QWOP is based on this idea. It messes with your head by making something that is fundamentally simple in all games impossibly difficult and it makes it fun. It’s is, to paraphrase Jeroen Stout, Magritte’s Walking Simulator.
So it’s not a given that unreliability isn’t fun, whether we’re talking worlds or mechanics. There’s an inherent tendency towards reliable, predictable mechanics and worlds in games because:
- they are made from technology which is fundamentally about predictability and reliability, and
- predictability and reliability are the rewards for reduced behaviour in a ritual system. You only pray to gods you trust after all.
But it’s not a given.
Part four: You are not here (well, not really)
QWOP messes with the mechanics.
Silent Hill messes with the world.
Both exploit the gap between reliability and prediction, and the reality of reality, whether real or fictional.
Let’s take another quick detour...
This is an academic conference on philosophy and I got asked to come along as an ex-academic, so I get to play fast and loose with philosophy. That’s my pay-off for being part of the ritual. As Galway Kinnell said “Only the cow, the cow of such hollowness, mooing down the bones”.
So here goes then.
I’m a big fan of China Miéville’s fiction. I love his use of language, his precision and his flair, the way he surrenders sense and absolutism to a free flow and collapse of images. Like Lovecraft, his skill is in presenting the illusionary almost-there. I was reading "Embassytown" and was struck with one of those little tidy revelations that I was operating as if I was visualising the novel, as if I knew what these places, these characters, the hosts and their world looked like, and of course I didn’t, I’d just been happily cruising through the novel on the assumption I did.
All of which sounds dangerously close to Fodor’s mentalese, right?
Novels, poetry – linguistic modes of expression rely on translation from abstract symbol to reference. Games, like other visual media, do not. They present the reference itself. They reduce the interpretative gap through visualisation.
This is looping back to the Flying Polyps. The power lies in the lack of absolutism, the power of the imagination to occupy. This isn’t new.
But it’s interesting that ritual is normally used to tackle the unknown, to contain it, rather than to invite it over the threshold. Ritual is social containment. The whole point is it’s reliably predictable. Like Philip K. Dick’s reality, it doesn’t go away when you ignore it. DOOM is always going to be there in the same way when I boot up the application.
But of course buy a consciousness scholar or cognitive scientist enough drinks and they’ll probably slip and start talking about how actually we’re assembling reality from fragments and predictions in our daily lives anyway, just like how, whenever I’m away from home and next see my son, it’s always a shock how much he’s grown because it’s an illusion that I really see him every time I look at him without supplementing this image with my memory. And then we’re back at Roger Schank and his scripts, and Bartlett and his schema, and it’s like I’ve never gone away.
Part five: Let's assume the fractures, shall we?
So we’ve done a lot of digressing and picking up fragments and some of you might even be thinking that there’s no actual coherent thread through this lecture at all, just the illusion of one being created by a rapid series of ideas and images that seem to orbit around a common thread that’s never actually been absolutely defined, and that’s interesting given the subject of this lecture.
So let’s continue to avoid talking about the thing, and continue to line up our little shoes, never worn, around the absent absolute and let’s shift gear and talk a bit more about the games we make and why we make them like we do.
Let’s be frank about this, our games make some people really, really angry. Like obsessively enraged, furious, foaming at the mouth, assemble a mob with flaming torches and march on the castle angry.
Fortunately enough people sit at the opposite end of the emotional spectrum for us to be a viable business and keep making that first group angry, because the reason they get so incensed is usually one of two things. The first is that “it’s not a proper game” which is intensely boring as a discussion so we’re not going to get into it here, and the second is that “they don’t make sense”, more usually presented as the worlds or stories starting out engaging and interesting but then degenerating into this pretentious mess of ideas and images that don’t settle into a tidy resolution.
For example, people that hated Rapture hated it because it was too slow and the ending was disappointing because rather than close things up, it just went off into a philosophical pontification about the value of existence and what it meant to be human and why death doesn’t matter and – well, actually it seemed to be more about this scientist woman banging on about seeing fields from the window of a plane and butterflies in sunlight and so on, and it just plain refused to make sense.
But for me, stories with answers are just dull compared to stories with questions.
Margaret Atwood’s "Oryx and Crake" is a stunning novel because it leaves on a question. The third part of that trilogy, "MaddAddam", ends on a sense of resolution and I like it less as a result. I didn’t want to know about the piggoons and refugees from the apocalypse finding common ground, just as I didn’t want the gaps in understanding about the apocalypse itself filled in. Snowman leaves the first novel in a state of tension – infected, dying, creeping towards the only humans he’s seen. “Zero hour” “Time to go”.
Like the Strugatskys' "Roadside Picnic", Tarkovsky’s "Stalker" and GSC Game World's Shadow of Chernobyl, the power of "Oryx and Crake" (like the truth Jimmy and Crake seek) hides in a pixel, not the image, but a tiny component part of the image. A field of symbols is presented without the dots pre-joined, and the underlying tension as a reader that the dots may not actually join or, at the least, that the sense we are joining them is an illusion. The ritual is undermined, we are left in a state of tension, like Snowman, because the predictability, the reliability is undermined.
It’s this sense we wanted to create with Everybody’s Gone to the Rapture. You missed the apocalypse and you can never know it. Why should you? What kind of an apocalypse would it be if it was easy to bottle up and visualise, like a Flying Polyp. We should recognise it not by it’s form, but by the five-toed footprints some claim it leaves behind.
Take aim, but the VATS system will force your shots wide. It’s about challenging the illusion of absolute.
Other people that hated Rapture said it was because it was too linear. Rapture is an open-world game.
In fact, there’s a correlation. Players who hated it explored less, followed the obvious golden path. Players who loved it just went wandering in the world without requiring strong reference points. They ignored the map. They went exploring and they didn’t expect the world to present them with neat, reductive conclusions.
Part six: By-products
I’ve often said that that most powerful tool you have as a designer is the player’s imagination. But remember Magritte. You don’t get absolutes for free. You might up the fidelity, but you do this at the expense of the fracture.
We work in the fracture. The key principle for us is to preserve that. Sometimes it works really well, other times it works less well.
Amnesia: A Machine for Pigs is a very flawed game, although we’re not going to talk about most of those flaws today because they loop back to the question of whether a game is all about mechanics, and hopefully as I’ve intimated, the relationship between games and mechanics is like the relationship between reductionism and ritual and we should move away from that and leave it hanging, like Kinnell’s hollow cow.
What I want to talk about is this choice. In Machine for Pigs, the Manpig enemies chase you around, they are the bad people, the monster under the bed. On the superficial, absolute level they are there, they pursue you. You hide or run or you are killed. They are true or false.
We wanted to push this further. We wanted to have the pigs get escape velocity from the absolute, to do a Silent Hill in reverse. So we made a choice, and it wasn’t an obvious or easy one, to create a sequence in the game that showed the monsters as victims. You see one attempting to stack coloured blocks, like a child. Another cries near a smashed mirror, understanding what it is. Another is bullied by a larger monster.
The by-product was that, after experiencing this, many players found the enemies less frightening. Actually, I think that was really a design issue, the hunt spaces were too easily demarcated, the game was far too predictable. We fell into absolutism on a level design level.
But what was interesting for me was that the Wretches moved towards the Flying Polyps. They started to occupy a space that both was and wasn’t. They were objects of pity and fear simultaneously. We might even have felt sorry for them. The possibility space opened back up, we introduced a fracture into the absolute. Our failure was not then supporting this fracture with good, unpredictable enough supporting challenge design. But this uncertainty we introduced is core to why Machine for Pigs works. It is utterly contradictory. The narrative, the world cannot make sense.
Which, given it’s a game about madness, makes sense.
Part seven: Failure of meaning does not mean failure
Fear and absolutes don’t mix. Lovecraft understood this; so does Silent Hill, at least initially.
There are images in Machine for Pigs which are deliberately contradictory, which break the meaning. This is completely deliberate.
A Machine for Pigs is a game where the primary victim is truth. But it struggles to escape the truth system that enables it to exist.
We’re working on a game at the moment which is currently called Total Dark (even though it won’t be for much longer).
In many ways, it’s an absolutist game for us. It’s got mechanics and everything. It started life as a paper-based RPG, then a board-game, then a computer RPG. It’s had bits of survival horror and action-adventure.
It’s even got an inventory, a dialogue system and ballistics in it. So it’s pretty traditional.
The question for me is, with all of these absolute systems in place, how do we keep opening things out? How do we enable the Little Shoes to co-exist with a predictable, reliable, traditional system.
And like Silent Hill and Machine for Pigs, this is where the story comes in.
Total Dark takes place simultaneously on the inside of a pocket universe 19 inches across, on the Welsh island of Anglesey and in the mythical Welsh otherworld of Annwn. It takes place simultaneously in 1974, 2016, 1887 and 1941. It’s about a librarian who becomes a fictional king, a magician who accidentally might start doing real magic, or might just be tricking us, and a spy who doesn’t exist because her mother made her up.
The key principle is to make a game with Flying Polyps in it. Can we push the player into the ritual where they can reliably predict the system but live in the fractures, embracing the uncertainty at the same time? Can you be mechanically sound but fictionally rootless?
That’s got to be an interesting question.
Eric Zimmerman says games suffer from cinema envy. That’s true of many mainstream games, but what’s more important is that games also suffer from the same problems with absolutism that film does. We come back full circle to the start of this lecture: visual representation is reductive, absolutist and although that enables the core ritualistic joy of gaming, it restricts the possibility space that poems and novels exploit as a core feature.
Why can’t we have both? What would games be like if they suffered from poetry envy instead?
Part eight: Why the tidy must be destroyed
The core does not exist. We talk endlessly about structures and architectures in game development. We spend months and months ironing out bugs and behaviours.
We obsess over emergence but we’re terrified about unpredictability. We talk change and endless regurgitate conservatism. We’re all about innovation, but we’re political throwbacks in the narratives we spin and the way we spin them.
We are an industry of contradictions. But we run terrified of contradiction in the things we create.
None of that is absolutely true, of course. Because absolutism doesn’t really exist, which is why rituals are profoundly unnatural, artificial things. And that is their power. But they are not absolutely things to aim for. We should always be trying to destroy the tidy.
Crake destroys the world, fundamentally, because it won’t do what he says, and strands Jimmy the Snowman in a fetid junkyard of possibilities that is terrifying because it has no structures left.
Silent Hill accelerates its terror by presenting non-sense.
Kinnell and Atwood and Miéville and Lovecraft and not-Hemingway-but-Kane all operate in the fragments, suggesting the absolute whilst resisting it utterly. They understand the operation of the momentary glimpse that we construct sense from, and how to exploit the will to resolution that appears hard-wired into consciousness.
Our games, at their least interesting, are simply mirrors to the technologies that enable them. At their best, they are mirrors to ourselves.
There is no inherent reason why we cannot find our little shoes, never worn, in a game. It’s a question of will, of aspiration, not technology. We have the choice of what games we want to make.
Zero Hour, time to go.