At the Office

Home

On the Couch

At the Office

Mailbox

Neighborhood

Against a Formal Language for Game Design

There have, unfortunately been a number of calls for such a thing. This article discusses why it's a bad idea. The Lindley piece linked below inspired this rant.

Introduction

While the world has already seen far too many articles on the subject of a formal language for game design, for those not familiar with the topic, you may want to check out Doug Church's article "Formal Abstract Design Tools" or the recent piece by Craig A. Lindley on the subject. In a rather tight nutshell, the proposal has been to define a vocabulary for talking about game design that can adequately capture the concepts involved through a rigorous structuring of terms according to formal rules of logic. The benefits, its proponents claim, will be that game design will finally be rendered a field of study within which theories can be hypothesized, tested, and either proved or disproved. With these established knowledge systems in hand, then, we will be able to discuss, study, and teach game design more easily and reliably than we are currently capable of doing.

The problems with this approach are legion. The first warning sign should be that these attempts to clarify discussions of game design are themselves written in jargon-laden, practically impenetrable prose. Physician, heal thyself, indeed. However, even assuming that the presentation of this formal language could be done in reasonably comprehensible form, the project itself is fundamentally misguided. A formal language for game design will not lead to generating better game designs; nor will it create consensus on topics in the field; nor will it make the communication of game design concepts easier. In fact, a formal language will make it more difficult to spread knowledge in the field of game design.

What's in a taxonomy?

Assuming, for the moment, that it were possible to derive a set of logical mechanisms of game design--a grounding in fundamental structures of risk/reward, learning gameplay through example and iterative creativity, variety within a coherent ruleset, and the like--what would this gain us? If it is not readily apparent from the behaviors of the gaming audience (i.e., consumers) that there is no single type of game or gaming experience that appeals to all people, perhaps we should look at the underpinning structure of this proposition. Even if we were able to arrive at a consensus on the axioms of game design, there is ample established proof in the field of philosophy that such an axiomatic system would not be capable of determining whether a given game design would work or not.

If you're not willing to buy that Godel's Incompleteness Theorem would apply in this case, the basis in logical positivism that this project has should provide a clue. For those of you who slept through Philosophy 101, the logical positivists held that as long as we could all agree on what exactly we were saying in a given statement, the truth value of the statement itself would be easy to evaluate. Of course, this was before the post-structuralist revolution in literary theory, which established that not only is it impossible to formalize language such that its meanings are once and forever firmly pinned down, but that the truth value of a statement is only relative to the framework in which it is being used. So, even if we could arrive at a formal language, both its stability in creating meaning and its capacity for establishing truth would be fundamentally undermined by its reliance, of necessity, on a specific context. The "truths" expressed in a formal language of game design, in other words, would only be valid for a particular audience in relation to a specific conjunction of terms.

Furthermore, by creating a formal language, the potential usefulness of its "truths" is restricted to that group which has received sufficient training (or one could be unkind and call it indoctrination) to understand its grammar, syntax, and vocabulary and sufficient practice to use those elements correctly. Put simply, the only people who could use a formal language to talk to each other would be the elite group who had invested sufficient time and money to learn the language. Coming from the school of literary theory, which uses a highly formalized language that it takes students years of time and usually tens of thousands of dollars to learn, I am all too familiar with the limitations of this process. Now, while the language of literary theory does serve a useful purpose, enabling discussions that would otherwise be impractical if not completely impossible, it is, essentially, a professional shorthand that sounds like complete gibberish to anyone outside the field.

At first glance, that may not seem like much of an objection. After all, the proposal is not to create a formal language of game design for the game-buying public, but rather for the professionals in the field of game design. Yes, but that's not a proper analogy. The language of literary theory is not only useless to the book-reading public at large but also to writers. Derrida's theory of differance and Kristeva's abject have never enabled a writer to generate a compelling novel, play, or poem, regardless of how applicable these constructs may be to their work. This is because literary theory is an analytical pursuit; that is, it focuses on descriptive statements, characterizations of forms and structures within literary work. While the generation of a formal language for the study of game design might be a worthwhile project, it is never going to be useful for the production of game designs because that would require a language that formulated prescriptive statements, rules that govern the mechanisms of generating game design.

The limits of rules

Now, while I find Hal Barwood's project of gathering the "rules" of game design a thought-provoking and valuable process, it nevertheless cannot generate either a sufficient or necessary basis for the production of game designs. First, a game design cannot be divorced from its implementation. There is no such thing as a great game design that is ruined by performance issues or lackluster assets. This may seem counter-intuitive, even invalidated by your own experiences. You may be saying to yourself, "That's not true. When I played this one game, it had a great design but it simply wouldn't run." This confusion of "design" with "idea" is a common fallacy in how people think about game design. Sure, there are plenty of games out there that had great ideas that were poorly executed, but an idea is not the same thing as a design. A game design has to take into account whatever limitations there may be with the engine, with the art assets, with the testing process, in short, everything that affects the final experience of the game. Game design simply does not happen in a vacuum. A design that cannot be implemented, represented, or made stable with the time and resources available is not a good design.

See, people think that design is all about having that great idea, but it's not. The reason why there are millions of people out there with great ideas for games but only hundreds of game designers (and even fewer good game designers) is that design is about working within the restrictions you have. You could have a great idea for a game that happens in a four-dimensional space, but unless you can design a way to have the player experience those four dimensions in the three dimensions we're used to living in (or the two-dimensional representation of that three-dimensional space that most games are limited to), it's just an idea. Designing a game that outreaches the ability of the team to execute it is poor design. Sure, there are plenty of things beyond the designer's control that can prevent a designer from adjusting a design when a problem crops up (marketing commitments, shortened schedules, mandates from on high, etc.), but it still results in bad game design.

This is where the "rules" of game design fall short. There is simply no way to generate a ruleset for the production of game designs that can invariably fit every circumstance of a game's production. What may be a good rule at one stage of development may be a bad rule due to technical or time limitations at another. If the "rules" of game design could provide a sufficient basis for game design development, you wouldn't need a designer; anyone could sit there with a list of the "rules" and simply go through and check that they had all been satisfied, applying trumps where necessary, making adjustments where needed. "Game editor" would have a completely different meaning.

Just as the game development process is never the same thing twice, there is no one-size-fits-all game. Ask a dozen people what the best game they ever played was and you'll likely get ten different responses (not twelve; after all, there are "schools" of gaming thought). Many people will defer, actually, for the same reason that most people don't have one favorite song or one favorite movie; there are simply too many to choose from, and one can be different from another without being "better". The most you will get out of them is a "best" in a particular genre or time period. If your household is anything like mine, you've probably watched tons of those "100 Greatest Movies" or "20 Best Rock Videos" or whatever shows; heck, VH-1 seems to have based its viewership model on that one proposition. The one common point of all these lists is that at the end of the show, people will disagree with the rankings.

Why even try? No serious academic would hold forth that such-and-such a book is the greatest book ever (or poem, or play, etc.). There is no great chain of being, no linear ordering of experiences that everyone can agree runs from best to worst. Games, like other forms of entertainment, are subjective experiences that generate radically divergent responses from diverse groups of people. The "rules" of game design are no more necessary to an individual's enjoyment of any particular game than a specific metrical scheme is to an arbitrary reader's experience of reading a poem. At best, the "rules" establish points of commonality between diverse games that have been successful in creating positive experiences for large groups of people. As such, they are eminently useful for game designers to think about, as I said at the beginning of this section, but in no formal logical sense are they determinative of the field of good game design.

In the same vein, any formal language of game design will fail to generate consensus on game design issues. Regardless of how rigorously defined the terms are, irrespective of the precision with which the grammar can establish relationships, notwithstanding the unfailing stability of its syntax, there are going to be simple differences of opinion about what elements of game design contribute to "good" game design, and no matter how abstract the formal language may be, the ultimate judgment of game design will not fall on its conception (which is, after all, only an idea) but rather on its execution.

What's the point again?

Somebody once said, "Those who do not learn from history are condemned to repeat it." Those who call for a formal language of game design should look back on Bertrand Russell and his project of a formal language for mathematics and its spectacular failure in a field far more tied to logical syntax, grammar, and vocabulary than the field of game design is. Or, simply return to the field of literary theory, one that is near and dear to my own heart. In neither of these cases did the production of a formal professional language make communication about its field more easy or more broadly accessible. Rather, these formal languages are available precisely to the fewest of the few, the elite who have the luxury of being able to invest years of study and training into the pursuit. If anything, they have succeeded in keeping large swaths of the population away from their respective fields, because they are artificial, alien constructs that have little if any relationship to the functional language most people understand.

Nor were these languages ever brought into being by conscious fiat. In fact, if you look back at the history of these (and other) professional languages, what you will see is that they evolved organically. At some point, there was a canonical text (in both these cases, going back over thousands of years to the Greeks) that established formal terms as part of a description of particular phenomena. Over the course of these many years, they have borrowed (or "appropriated", depending on your point of view) terminology from a variety of cultures, from a vast array of writers and thinkers, as that vocabulary captured important concepts. The utility of the language is what has preserved it, not some arbitrary value of formal logical coherence.

There will be a professional language of game design. It will emerge, like the professional languages that preceded it, from the everyday uses of the people in the field; they will pick up on terms from discussions with their colleagues, from articles or books that they read that capture a concept with grace and perhaps even wit. In fact, it is already here. Ask one of your non-gamer friends if they think the concept of a "GUI-based AI scripting tool for MMORPG's" is a good idea. The blank stare they will give you is confirmation that you are already one of the elite, the few who have studied enough to speak this strange, private language.

So, let us put to rest, once and for all, the call for a formal language of game design. The only purpose such a thing could serve is to further divorce the few from the many, which is only a good thing if you're trying to credential yourself at an institution of higher learning. For the rest of us, we have a language we can use. The last thing we need is mandatory training in some sort of design-oriented esperanto.

Back to Top

Home

On the Couch

At the Office

In the Library

Through the Mirror

Mailbox

Neighborhood