(Prepared in response to this entry of Jesper's blog, which began with the discussion question of whether Pac-Man "really is" its source. This has been included as part of Vol. 7 of my free lessons in hobby videogame development.)
I should note before continuing that I found your Pac-Man source posting 15 months late coming from a link in Ian Bogost's "Videogames are a Mess - DiGRA 2009 Keynote on Videogames and Ontology". Having read that immediately prior, and that being the context in which I found your message, I'm somewhat unable to pry my thoughts on the two matters completely apart. I apologize in advance if I quixotically tilt against a few seemingly unrelated points while attempting to break down my perspective.
I agree, that "games on a fundamental level consist of their rules".
I disagree with the common argument that videogames on a fundamental level consist of their rules.
This disagreement is based in doubt with the unstated assumption that videogames are completely a subset of games.
I propose that at the most fundamental level, the only specific description of a videogame ("what it is" in the most relevant sense) is from the game itself running, not the assets/source nor any description of rules. This is rooted in 6 considerations. In enumerating these, I hope to highlight that the issue mentioned arises from treating videogames as games (even while, in many important ways, videogames are certainly like games):
1. Unnatural laws in videogames are equivalent to rules in videogames
Rules are relevant for turn-based games by constraining action.
Rules are relevant for human-enforced games, because the game otherwise does not exist, and without every player's regard for the rules people would simply walk off the field or away from the table.
However, in a programmed environment, what we create and experience is closer to alternative natural laws, or more suitably, unnatural laws. The way things work and interrelate does not need to be respected for the game to function, because the way things work and interrelate is no more optional than gravity or the inability of physical objects to intersect in the real-world. We do not, so far as I think, regard gravity, the way friction interacts between shoe and ball, or human bodily fatigue as rules to describe soccer, but these sorts of natural laws must be simulated in a videogame interpretation of this or any sport, on equal gameplay significance to what happens each time the ball enters the goal area.
In other words, there are no rules in videogames. Things are not possible in a videogame that rules or penalties would need to restrict in a non-digital game. There are only unnatural laws, a fraction of which fulfill the same function that rules would serve in a non-digital game.
(For sake of clarification on word usage, rules are invented in the cases where human players make agreements outside the mechanics of a videogame to only use pistols, never use the brakes, etc. Because these can be violated, without penalty if uncaught, and are more restrictive than what is explicitly possible, this construction seems much closer to the human enforced rules of non-digital games.)
2. Physical sports games are like videogames on a unmentioned engine
The seeming trouble in capturing rules of Pac-Man definitively imposes a degree of rigor on the game that we cannot possibly consider appropriate for physical sports. For sports, we can trivialize away all kinetics, space, time, and behavior into unexplained convention: real physics/biology/psychology. However, the kinetics/space/time/AI of every videogame must be constructed beginning from a empty, timeless void. We are unable to describe the rules of soccer even as exactly as we can the rules of Pac-Man, provided we are fair by not giving soccer the presumption of a complex background universe to function within.
The rules as they concern me as a developer, user, or critic are specifically described by the actual gameplay of the game, as it runs on its original hardware. This is at once far more complicated than the list of rules we would make for soccer to be played, yet far less complicated than the list of rules we would need to specifically describe soccer. The difference is that we normally ignore describing the engine (to use the videogame term) that the rules for soccer takes place on, but we have no such luxury in specifically describing a videogame.
3. Videogames are not their medium
A particular movie is not a reel of thin, perforated strip of cellulose. It the same movie, at least as it generally concerns consumers, creators, theorists, or critics, whether watched on DVD, VHS, cable TV, theater projector, or MPEG4.
Isn't it neither fully described by its images alone, nor its soundtrack and audio alone, but only as these different forms of media are presented together, whether they're coming from a spool of film or digital data on an optical disc? Certainly, the screenplay would also not suffice as a specific description of the film, as it would overlook the significance of actors, cinematography, sound, and everything else that makes films different than a book of text.
It is the same distinction that I wish to make about videogames being inseparably different than a traditional game described by (to such a point that it is at its most fundamental level) explicit, enumerable rules.
The operational behavior of a videogame can be specifically describes by the original hardware, by emulation, or by a code port made with excruciating attention to detail to reproduce original behaviors resulting from the original hardware, but like a movie, nothing but the active phenomenon experienced by the consumer can describe the most essential parts of what it is.
4. We experience and study videogames as "living" things
A dead dog is made of the same tissues as a live one, and in a very similar arrangement. If it died very recently, it may even mechanically and chemically be very much like a living dog.
However, surely, from the perspective of someone that seeks to breed, play with, judge, or theorize about dogs, we are most likely concerned almost exclusively with how they are while alive, and generally disinterested in the composition of their physical parts. Rex's body is not Rex after death (though we may use the name to refer to the body), and Tetris is not Tetris when it isn't running (though we may use the name to refer to the cartidge/machine/file).
5. Changing the assets or behavior of a videogame breaks its identity
If calm music is replaced with heavy metal, or if the player's ninja sounds/visuals are swapped out for the sounds/visuals of a bear that walks upright, there is little question that the videogame has undergone a significant change and will be experienced much differently by the user. This change is not really much different than a change to the game's source code.
Graphically, quite apart from how a videogame's rules (or its significant, non-continuous events from its unnatural laws, as when a player touches spikes) are written, as a matter of affordance the player will make very different choices and movements based on what colors, forms, and enemy movements communicate to the player about safety, intention, potential for meaningful interaction, etc. I offer as a very crude and minimalist example Protected, an experimental project I threw together to illustrate how a rule can be suggested into a player's mind to steer behavior even if there is no consequence in the explicit rules/mechanics.
A videogame is not the same videogame as soon as any of the presented behaviors or content are changed. By elimination, then, I propose that what the videogame "is" must be the videogame when absolutely none of those outwardly observable/measurable presentation elements or user interactions are noticeably changed. Thanks to the combination of hardware mass production and simple digital duplication, this is description not exclusive to any one hardware or software platform, and is not as geographically bound as, say, trying to discuss architecture based only on blueprints and photographs.
6. Rule-based games are not a superset of videogames
Whereas we can fully play Poker in a complete sense within a videogame, or meaningfully simulate the rules and physics affecting games like Pool, we cannot fully play Super Mario Bros in a complete sense with playing cards, or on a granite table with sticks. Although we could model a card or table based on a very limited set of SMB's significant events, it would fail to account for the game's unreal spatial, kinetic, temporal, and representational particulars, which to my view are every bit as important to accurately describing SMB as (if not more important than) which rule-based events occur during operation. To synthesize or analyze SMB using a rule-centered approach would miss much of what makes SMB what it is to the player or developer, from the timing/structure involved in battling winged koopas to the subtle visual environment clues that suggest access to hidden warp zones.
A real-world environment of tangible and physically constrained objects manipulated within a few dozen explicit rules, accounting for input events at most once every few seconds (that is: cards, board games, D&D), cannot fully reproduce the behavior of a digital system that manipulates arbitrarily transformed expressions of imagination and interrelated dynamic systems involving tens of millions of operations happening every 60th of a second.
However, the inverse is often true. Might the issue of searching for specific descriptions of rules as a fundamental part of the videogame come from applying questions specialized to games to things which are not fully contained within the category/subject of games?
As someone focused specifically in videogames (but not games more generally), the minimal specific description of a videogame's behavior and what-it-is seems to me to be that videogame in active operation, as opposed to either its development source/assets, physical delivery media/hardware, or its complete software while inactive.
The videogame is the presentation and dynamic relationship between the intended input mechanism and output over time.
If this relationship and presentation could be produced somehow without the other hardware/software/asset layers of its implementation and existence it would still be the same videogame; if this relationship and presentation were other than they are or non-existent, none of those other layers refer to the videogame itself but only the means of creating or distributing generic data. To clarify, a fully functional emulator playing a PS1 game on a PC with a PS1 controller through an adaptor "really is" that game, but playing an Atari game with a computer mouse is not, nor is a sufficiently scratched game disc still that game.
Endnote, disclaimer on semantics
I use the word "videogames" reluctantly here as shorthand for "interactive digital media" to minimize awkwardness in wording, despite the loaded historical meanings of its stem words. I would prefer some more neutral term like "interactive digital systems"/IDS or perhaps "input+output+processor"/IOP. However it seems as though every time people attempt to coin terms for videogames-technologies to free them from the connotations of play, triviality, fiction, and psychosocial moratorium, these new terms invariably get mistreated as a euphemism for videogame by consumers, producers, and critics asserting the values of entertainment in judging or discussing it.
Sept 25, 2009