Over time words and even entire languages change their sounds. Mechanics as we use them, especially how we choose to use them, also drift over time, adjusting to a group’s or person’s preferences. And just like in phonological drift, some mechanics can be dropped, some merged (or synthesized into new procedures), and some changes produces a chain of subsequent changes.

Distinctions and their mental categories often shift consciously or unconsciously. A mechanic is a way of indicating a distinction, of judging a situation’s presence within a mental category. So a group can struggle to maintain distinct judgements after a mechanic’s application changes, either going back and forth on the change, or adjusting the other mechanics to accommodate the shift.

What can cause that shift in the first place varies. In some cases it can be an error or misremembering of a procedure which turns into a new mechanic. In other cases, it can be a fully intentional change. It frequently occurs as a result of the immediate recollections of the procedure being subsumed over their social and fictional applications so that other parts of their categories are forgotten. In essence we train ourselves to a new set of categories and distinctions, evolving constantly during the course of play.


Spoken languages are often built on a surprisingly small collection of sounds, most likely because the more sounds a language includes in the range of human hearing and vocalization the more difficult it becomes to adequately distinguish among them. Communication already lies in a vast sea of ambiguity with uncertainty about experiences, points of view, mood, and even our senses looming over any attempt to express ourselves. In that light we can see why effective languages distinguish things like the sound “ka” and the sound “da”.

Distinctions give us a foundation to build on, to expand on the basic sounds, ideas, or moments that are clearly delineated. In a game, the most common way to generate a distinction is mechanics*. In a D&D variant invoking the attack roll mechanic clearly describes the moment and intent as an attack being attempted, as a opposed to simply describing something which may be interpreted as an attack, a feint, or even an expression of raw skill to impress. And because this is delineated in a D&D game, and not so much in, say, a game like Prime Time Adventures, it indicates that what is and what is not a physical attack matters far more in the language of play of D&D and its variants than the aforementioned Prime Time Adventures, which has its own distinctions.

*Indeed, I find it particularly useful to treat the concept of mechanics as defined by introducing distinctions in play. The indication that something is a mechanic is that it requires a category judgement, which could be a discrete decision.

At first it can be disheartening when designing games to realize that immediately or eventually the game played among a group of people will no longer adhere to the original design. This is the difference I described earlier between the prescriptive model of play presented by a design and the descriptive model of what happens during play. Language similarly evolves both slowly and quickly due to errors, conscious and unconscious coordination, and the influence of other languages. However this evolution is not simply random, often following trends which have been the root of the field of evolutionary linguistics. RPGs fall prey to some of these same forces, so it is reasonable to search for trends in how play changes over time.

The ways that RPG play derive from a static design is a process I’ve called induction, after the electro-magnetic process by which moving magnets cause the creation of electricity. Induction approaches perhaps the core dilemma of an RPG designer, “how do you get people to play your game correctly?” since this is presumably the best way to play, and turns it on its head. Instead it asks, “how do you get people to reliably evolve my game into a better one?” This turns the design and publishing focus on simplification, precision, and clarity into one of reliability, robustness, and most importantly imperfect communication. Ultimately it accepts the gap between prescriptive and descriptive, between what is written or spoken and what the audience interprets from it.

One of the key dichotomies in understanding both language and RPGs is the distinction between prescriptive and descriptive models, whether these be grammars, game systems, or some other representation.

Prescriptive models are often idealized, they are bases for things like best practice and notions of correctness. In RPGs, we are in the prescriptive domain when we talk about whether we are playing such and such game, or we are hacking or drifting it into a different game. Prescriptive models play heavily into identity politics in both language and RPGs, since they serve as means to identify whether or not language use or RPG play is right or wrong. From the RPG perspective, an example of a prescriptive model or grammar of play is a reasonably complete game system description.

Descriptive models on the other hand, instead of driving the usage are derived from the usage of a language or a game and serve to encapsulate and predict the behavior of some or all the communities involved. Descriptive models are extremely useful for understanding how language or games are used, but they are built on ever-changing data and observations. From the RPG perspective, an example of a descriptive model or grammar is a well-constructed actual play report or “example of play”, at least one focused on how things happen in play rather than in presenting the events of play entertainingly.

Within the separation between the prescriptive and the descriptive is the fruitful and frightening domain where communication evolves. Like languages, RPGs change as we play them, and further as a community experience the same or different games. One important frontier in designing and teaching RPGs is how to embrace this dynamism, instead of trying to constrain it.

In the past few years I’ve been finding it increasingly useful to consider RPGs as languages. At their root, RPGs communicate ideas, emotions, and stories among their players, perhaps to a particular end or just to enjoy the collaborative experience. In any case, communication is a central part of playing RPGs, even “solitaire” ones. In this light, we can see the rules, procedures, and components of an RPG as words and structures within a language in which the social and fictional worlds of play are built and communicated.

One useful idea that derives from this is the notion of grammar – a set of rules which generate what can be expressed through play. Grammars can serve many purposes in language. They can capture the syntax or structure of a communication or they can be used to translated into other languages of games or otherwise, or even sketch out the semantics or meaning of events during play. They can be prescriptive, defining the correct way to use the game, or they can be descriptive, describing the way the game is played by a particular group or community.