Not that long ago, famed film critic Roger Ebert stated that games are not, and indeed, cannot be, art.

Recently, following up on a response made by author/game convergence guy Clive Barker at the Hollywood And Games Summit, Ebert clarified his claims.

Ebert’s basic argument is that art requires authorship, and that games abdicate authorship to the player, and therefore cannot be art. This is certainly a compelling and useful argument, as it strikes at the nature of our medium, not at the content we so frequently produce. Ebert is clever not to bother arguing whether or not another boring old rehash of the game where you use 24 different weapons to kill 6 different types of aliens in 18 unique levels and save mankind from annihilation is art or is not. He is going for the kill, and good for him.

As a student of the ‘school’ of design that often loudly and passionately advocates abdication of as much authorship as possible to the player, I guess I need to step up and help Ebert understand what he’s trying to say and what it means.

Ebert claims that because games – or indeed, any interactive medium, abdicate authorship to some degree to the player, that this at least diminishes, and potentially destroys the capacity of an interactive work (or at least a game) to be art. He also tugs gently at the idea that it is potentially in the interaction with the work that the artistry lies, and thus any artfulness in a game or other interactive work is not innate but rather, if it exists at all, is imbued by the audience.

Ebert is wrong for two important reasons.

First, there is authorship in games, no matter how much we abdicate. The form of the authorship is different, and hard to understand, but no matter how much we try to abdicate it, it will always remain. It is undeniably there, and it is inextricable from the act of creating a game.

Second, interacting with a work does not shape the work, it ‘only’ reveals it. Therefore, while there can be an art of expression in the way someone reveals the art, this does not necessarily diminish the art in the design of the work itself.

As both Ebert and Barker acknowledge, we could debate endlessly what is and is not art, or what is and is not a valid definition of art. For the sake of argument, I will accept Ebert’s roughly stated thesis that art requires authorship. In fact, I actually agree with him. I think he just does not understand where authorship lies in games.

Here is how it works:

• I am able to express my ideas, thoughts and feelings through the design of interactive systems
• Because a game is a complete formal system, the entire possible range of outputs from those systems is determined by me
• People interact with those designed systems and receive the outputs I have determined
• People literate in the medium can reconstruct my ideas, thoughts and feelings by experiencing these outputs
• Therefore, by definition, there is an unbreakable chain between my ideas, thoughts and feelings and the player’s experience – I author mechanics that yield deterministic outputs in the game dynamics that lead the player to experience the aesthetic I want them to experience (within a given tolerance)

Now – by way of clarifying this explanation as well as pre-empting some of the counter arguments, I’ll try to lay out the best and most valid ones.

The Epistemological Argument
First there is the epistemological argument – which counters by saying ‘how do you know you are able to express your thoughts and feelings in the design of interactive systems’. Believe it or not, this is the best argument, because the best rebuttal is simply to say ‘I just know it.’ You could ask ‘how do you know the sentence you are speaking is not nonsense?’. I know because I understand it. What I am expressing makes sense to me both intellectually and emotionally. If others do not understand it, it is not really a question of whether I am expressing myself, but rather one of whether I am expressing myself clearly.

Beyond the subjective epistemological edge of this argument there is a requirement that the audience be literate. Hieroglyphics can’t express anything to me, but clearly the people who wrote in them had the capacity to express themselves, and they knew it. In the end, while this is a compelling argument, it degenerates quickly to a purely epistemological one, and arguing on that front is not going to enhance anybody’s understanding of the issue of authorship in games. I’d like to think even Ebert is not interested in pursuing this argument any further.

The Argument to the Incompleteness of Authorship
The next argument is whether or not it is, in fact, true that the entire possible range of outputs from a games’ systems are really determined by me. Well, honestly, no – they are not. Games are extraordinarily complex, and there are many outputs which are not literally determined precisely by me.

If we were to conduct a sort of inverse Turing Test and put me in a room and have me attempt to return the exact same outputs (ignoring ridiculous factors such as speed and precision of calculation) to a player playing Splinter Cell: Chaos Theory as he or she would receive from an Xbox… well, I would fail to pass myself off as an Xbox running SC:CT. I did not write every line of code in the game, nor do I have a complete comprehension of that code. Therefore, the argument could be made that my authorship of the game is only approximate, or incomplete.

The rebuttal to this argument lies in a comparison to film or to music or to any other collaborative artistic creation. Is every single note of a symphony perfectly determined by the composer? Is every single photon of light that enters the retina of every single viewer of a film determined by a director. No. There is noise in these systems too – some of it comes from the collaboration of others, and some of it comes from random noise. We do not deny Ode to Joy its status as art because it is playfully manipulated by a conductor, nor because the 3rd clarinet breaks a reed. It is the same with games. The outputs are broadly determined by me and heavily formalized by a large crew of people working with me to deliver on a promised aesthetic. Sometimes we make mistakes, and there are bugs. Even mediocre art can tolerate this approximate or incomplete authorship, so I am forced to assert that games can too.

The Arguments from Noise and Nonsense
The next argument would be that audiences cannot reconstruct the meaning I intend them to by way of interacting with systems. At very least there is a legitimate challenge here in saying "maybe they could reconstruct it, but they can also construct so many random meanings or contrary meanings, or generally just bring so much noise into the experience as to dilute the meaning to the point where it is not legible."

This is an interesting point, but again, this is not unlike other works of art. People can watch Citizen Kane and fast-forward the ‘boring parts’. They can watch it in ten sittings or fall asleep during parts of it and forget what it’s about. This does not strip the film of its status. So, yes, I would concede that because (some) games offer the player so many ways to play, players might well ‘miss’ the meaning, but I don’t think the likeliness of missing the meaning should determine the artfulness of the work. There are some paintings hanging in far away back rooms of the Louvre that would likely take four hours of walking full speed non-stop to get to – this does not diminish their artfulness. If the audience does not participate with the work they may never perceive the art, but that does not mean it is not there.

And what about those nonsensical interactions? The argument here is similar to the one above, and suggests that people can interact with games in all kinds of silly ways that don’t support or develop the meaning the creator intends them to experience. That’s true. I can play GTA 3 purely as a racing game, without ever doing any of the non-racing missions. How does that support the game’s central meaning (which I take to be about freedom and consequence)? Well, the answer is, it doesn’t. But hold on… I can also use War and Peace to prop up the broken leg of my couch, or view Warhol’s Campbell’s Soup Cans in a pitch black room and then say I don’t get the message. The point is that the audience must always interact with a work on some level (in games, this is very literalized), but their ability to interact with a work in nonsensical ways does not diminish or destroy the art.

You could say that in the Tolstoy example, the book is not being read, and in the Warhol example, the painting is not being viewed – but in my GTA example, the game is being played… and that’s the difference. I would actually simply say that is inaccurate. When you’re only playing one tiny part of GTA, you’re not really playing it at all, any more than you are reading Bram Stoker’s Dracula if you are only reading the sexy bits.

The Argument from Legitimacy
Another argument against the existence of real authorship in games is the argument about the legitimacy of the kind of authorship I am talking about. In his responses to Barker, Ebert says:

“If you can go through "every emotional journey available," doesn’t that devalue each and every one of them? Art seeks to lead you to an inevitable conclusion, not a smorgasbord of choices.”

In other words, he is questioning the nature of an authorship focused on providing a single perspective, versus an authorship that affords the player a range of perspectives.

I agree with Ebert that there is a difference in these two kinds of authorship. Romeo and Juliet is one love story. The play gives us a specific and authored understanding of what Shakespeare wants us to think about love in this specific case, between these two specific characters, in this specific set of circumstances. But Ebert is wrong to suggest that games – in affording a different kind of authorship somehow do not lead to inevitable conclusions, or that they offer a simple ‘smorgasbord’ wherein ‘every emotional journey’ is somehow possible.

In a sense, the kind of authorship in a movie or a play requires an inductive approach to understanding the meaning. If I start with the specific love of two star-cross’d teenagers in Renaissance Verona whose families despise one another, and then try to generalize what love means, I am likely to end up with a lot of useless conclusions. It’s an inductive and error-prone process to move from the specific to the general. With all of the inductive errors possible, what can I truly understand about love in general from Romeo and Juliet alone? I could reasonably induce that real love can only exist in the convoluted set of circumstances of the play where one lover drinks a sleeping potion, and the other – distraught at seeing their love ‘dead’ – suicides with poison, so that the first can then prove their love equal and true by also drinking poison on awakening.

Certainly that would be an absurd notion of love, and that problem with induction is fundamentally why we continue to write love stories at a breathtaking rate… because when you are providing singular examples of love, and trying to facilitate the audience feeling love, you will always have room for more examples. You can write unique and powerful love stories forever and never exhaust the infinite scope of the material ‘love’. Examples piled on examples forever will never pile up to infinity. Regardless – and for good reasons – this ‘inductive’ form of artistic expression is the kind of authorship that we see in most media (and it is wonderous).

In games, it is different. The artist does not only create the specific case of the convicted criminal suddenly set free when his prison transfer bus is ambushed… it does not only tell the story of one criminal learning about the importance of liberty and the consequences of unchecked freedom. The artist is also capable of creating an entire expressive system space that explores a potential infinity of different notions of freedom and liberty. Where most other media require the audience to induce their meaning, games afford the audience at least the possibility of deducing their meaning.

In other media, ‘supporting material’ that is coherent with the central themes of the work is pushed to the side in a B-plot… in games, this supporting material affords the artist ways to illuminate the meaning from many, many possible directions, allowing the player to explore the meaning the artist is trying to provide. Potentially, because the game designer is able to express himself in systems rather than in examples, infinities can be examined.

Now, I guess this is kind of hard to wrap your head around, but surely this is a concept Ebert can understand. Many filmmakers, from Taratino to Inarritu to Haggis and dozens more have been increasingly attempting to explore stories from multiple angles in an attempt to mimic – in a medium severely limited for this purpose – what games can do innately. If Haggis’ Best Picture winning Crash was 100 hours long, and contained 100 different interconnected plots all echoing the same themes of racial tension from different perspectives, would it suddenly lose its status as art? It probably wouldn’t be a very good movie, because 100 hours of movie is painful. In any case, no matter how long you make Crash, you will never fully explore the domain of the themes of racial tension in modern America. 100 hours is just 50x what the movie already offers, and is no closer to the infinite depth of the theme than is the existing 2 hour film.

GTA: San Andreas on the other hand – which I played for a good 100 hours or so, gave me such a world transforming view of racial tension and inequity in early 1990’s California, that I have been shaken to the core, and have been forced to re-examine a huge part of my world view.

In a response to a reader letter dated November 27, 2005 Ebert said:

"To my knowledge, no one in or out of the field has ever been able to cite a game worthy of comparison with the great dramatists, poets, filmmakers, novelists and composers."

Well, here you go. Let me state it clearly and for the record:

Taken as wholes, GTA: San Andreas is a more compelling, meaningful and important work of art than Crash.

Admittedly, not everyone will agree, and admittedly, I have a high level of literacy in reading systems. The point here is not to enter a subjective debate about what is a superior work of art, rather the point is to say that – yes – if a game is offering a smorgasbord of unrelated mechanics that are neither supporting each other nor driving toward a coherent theme, if they are not providing the player with a broad range of perspectives on a specific meaning that the creator is trying to express, then Ebert is right.

But if a game creator does have something specific he is trying to communicate, and he designs his game well, and the mechanics and dynamics are coherently supporting that aesthetic, and providing the player – more or less whatever he does (assuming it is not wilfully nonsensical) – with insight into that meaning, then yeah… it’s art.

The Argument to Migrated Authorship
The final argument that I see remaining is the one that asks ‘who is the artist here anyway?’ Ebert says:

“I believe art is created by an artist. If you change it, you become the artist.”

This is a much easier point to tackle simply because there is a fallacy in Ebert’s argument. He is implying that interacting with a work is the same as changing it. But this is not true. My ‘paint’ is not ‘what the player does’. My paint is ‘the rules that govern what the player can do’. The way the player expresses himself is a form of artistic expression (or a least it can be), but it is impossible for him to change the rules or even to express himself outside the domain of the rules that I have created. And it is not simply a case of saying ‘people who make paint are necessarily artists, while painters can be artists’. I do more than ‘make paint’. If all I did was ‘make paint’, I would concede the point in Ebert’s response to his reader where he states games cannot move "beyond craftsmanship to the stature of art."

If I created ‘magic paints’ that could only be used to paint flowers that appeared highly vaginal, would I be less an artist than Georgia O’Keefe? While some painters using my paints could make crappy paintings of vaginal flowers that were not artful, and others could create beautiful masterpieces – I think it would remain clear that I was an artist for having created ‘paints that constrained the set of possible paintings achievable to those that dealt with a set of themes I had chosen’. There is a statement there – a statement about flowers and vaginas (and paint) and it is as important as any statement in any of O’Keefe’s works. It may even be as important as all of the statements made in all of O’Keefe’s works combined… but we don’t need to get ahead of ourselves here.

The best analogy here, again, is that of a symphony. There is an art in the composing of the symphony itself – the creation of the song and the recording of the instructions for reproducing that song using a symphony orchestra. Yet, because of the comparative fuzziness of the transcription, there is often a high degree of malleability in interpretation of those instructions, and the ability to interpret those instructions well and to facilitate an orchestra to actually play the symphony is an artistic task. There are ways to do it non-artistically. Technically, a metronome and sheet music could do the job of conducting an orchestra, but we make a lot of room for conductors because in their art, they can add a tremendous amount of beauty to what is already a beautiful work.

Ebert’s fallacy extends to suggest that Ode to Joy is not a work of art once a composer conducts it, and that the conductor’s art is not art for long anyway, because as soon as a musician plays it, he or she becomes the artist. The reality is, they can all be artists, and when an artfully composed work is conducted by an artful conductor, and played by artful musicians, the full perspective of the work is truly appreciable.

It is the same with a game and a player. Technically, a game could be played by a computer – many games are played by computers. While the beauty of the particular play experience might be diminished or even destroyed by doing it that way, it does not diminish or destroy the artfulness inherent to the game itself.

Such is the relationship between the designer and the player as I see it. The designer is the composer, the player is the conductor. The orchestra is the hardware and the sheet music is the software.

Conclusion
I think that about sums up my rebuttal to Mr. Ebert. When he made his first statements months ago, I was mostly just upset and insulted, but with his latest clarification of those original statements, he has not only looked more carefully at the real and relevant question, but he has opened the door to have the real nature of authorship in games described. I hope – should he happen past here and read this – he’ll finally understand the scope of the issue and admit that he’s wrong.