A decade of theory, a sort of recap

This post is more a note to myself than a public statement but I'm prepared to discuss any of the points below if someone disagrees with them. I'm not laying down dogma, I'm trying to organise my thoughts. It's a summary of my current ideas about roleplaying game theory.

1. What is roleplaying?
It is a form of play whose primary medium is a shared fiction.

By play I mean a non-serious, energy-dissipating activity that takes place in a metaphorical magic circle which suspends the rules of the real world. See Caillois etc.

By medium I mean both the means and the substance of play.

By fiction I simply mean any kind of imagined goings-on. This fiction must be shared, that is, I cannot be the sole owner of it. Even if I'm just sharing its ownership with some kind of oracular device or GM-emulator, there must be an element of it over which I have no control. Otherwise I'm simply engaging in daydreaming or phantasy (perhaps in service of world creation, novel writing etc.).

To role-play, means to act as if the fictional content has a realness to it, to pretend the fictional elements and outcomes matter, have weight, to care about stuff you wouldn't care about in the real world. This is done primarily by assuming or advocating the role of a character within this fiction.

2. What is a game?
A game is a (structured) activity where participants have goals (or objects, or aims, or agendas), either set by the game itself or invented by themselves and a set of rules within the constraints of which they try to achieve these goals.

Goals can be explicit or implicit, there can be one dominant goal or a plethora of smaller ones or layers of both.

Rules tend to be explicit, but participants often have their own unspoken or unexamined rules that they follow.

Side note: The conversation
Assuming that the medium of roleplaying games is the fiction, the state of this fiction needs to be constantly communicated and negotiated by the participants (much like pieces on a chessboard must be moved and their state adjusted for the game to proceed). This is done by participants engaging in a conversation (or exchange of information) with one another. The fiction of the game (the medium within which its action occurs) truly exists only within this exchange of information, so the two might possibly be used interchangeably.

3. With what aim are we playing a roleplaying game? (Goals)
While most other games have a pretty straightforward and explicit goal (score more points than the enemy team, etc.) many roleplaying games have traditionally been more obtuse in this context. In addition to any goals set by the game itself (for example "gain treasure to raise yourself to the rank of lord and gain your own castle"), the fictional, character-based nature of roleplaying games allows for all kinds of personal goals for both the character and the player. Temporary goals can arise, be resolved or fall by the wayside during play itself ("we need to gain the lady of the manor's trust", "I have to sneak out before he notices me", "we must survive this combat" etc.). A game will flounder when such long- and short-term goals will not be apparent or emerge naturally from the setup, character creation, setting, etc. A game need not be absolutely explicit about setting its goals as long as the players easily discover satisfying things to do or find themselves in fruitful situations with interesting consequences.

There are also other types of goals, some of them outside the game but they're either not intrinsic or unique to roleplaying or are so diverse and particular to individual games and players it would be a cyclopean task to compile a comprehensive categorisation. There have been attempts at this in the past but I haven't found any of them particularly convincing.

Side note: Big model and Caillois (again)
One of the possible approaches to categorising player goals (although this is a contentious term in this context) is presented by Ron Edwards' Big model and its Creative Agendas. It posits three orientations within moments of play: gamism, narrativism and simulationism. I have made many attempts at a more colloquial definition of the three, the most recent of which would be challenge-oriented play, theme-producing play, and fiction-affirming play. A big confusion regarding this model arises from third-hand accounts and colloquial usage applying it to both individual players and games, while the model originally describes slices of actual play. A serious disagreement exists whether these agendas necessarily exclude each other (as Edwards claims) or if they can coexist or effectively switch back and forth. Furthermore, the model itself has historically struggled to define or qualify simulationism whereas it produced positive and detailed readings of both gamism and narrativism. I've often drawn on the big model in the past but it's not without its limits or problems.

A thing I've noted in the past but never wrote about in this context are Caillois' categories of play: Agon (competition), Alea (luck), Mimicry (suspension of disbelief, pretend), Ilinx (altered perception, vertigo, adrenaline, etc.). Agon maps directly to challenge-oriented play. Alea exists in roleplaying games in the form of dice, although some eschew it for diceless, fiat- or fate-based resolution. It's up to debate if GM fiat counts as Alea in some cases (feeding input into a black box and hoping for a good result to come out) or its opposite in other (everything is deliberate and decided, there is no random chance). Mimicry is the act of roleplaying itself, although traditionally it had not been paired with Alea and Agon before D&D came about. Finally I see a kinship between Ilinx and the various concepts from larp theory (flow, bleed, etc.), the elusive concept of "immersion" and even the simple exhilaration of winning a tough fight in a challenge-oriented game. Unlike the Big model's creative agendas, these categories can be evidently present in various degrees. One thing that Caillois' categories don't cover, but is integral to roleplaying is the production of narratives (whether these are premeditated, intentionally pursued in play or post factum "war stories" told about the game). The premeditated story is not at all unique to roleplaying and can be said to fall under Ilinx (compare to a rush of adrenaline while watching a horror or action movie, ie. getting lost in the fiction) or Mimicry, if one is an actor in such a story (or perhaps a mix of both if you go full method acting). The post factum story is not unique to roleplaying either (you can tell stories about exciting football matches or camping trips or poker games) although there is something to be said for the sheer variance and wonder of stories that can be told about roleplaying sessions as opposed to most other games or activities. This is something that many boardgames have been experimenting with recently, like the Legacy family of games, Cole Wehrle's "experience" boardgames or the modern COIN wargames which all provide interesting and enduring sequences of events and outcomes, ripe for talking about later. The in-play formulation of narrative is considerably more unique here, although it, too, shares functions with other mediums like improv theatre, certain forms of creative writing, collaborative storytelling and so on. Ultimately, I think I'd add a two additional categories here and tentatively call them Poiesis and Catharsis. These are both art and literary terms, which is not surprising since roleplaying games are the conjunction of games and fiction (of which literature is the oldest structured expression, not counting oral folk traditions). Poiesis is the thrill of creation, children drawing is definitely a form of play, that I don't think the other categories properly capture. Ultimately, catharsis is about seeing the outcomes of choices play out, particularly if they are ethically charged choices. In literature seeing the protagonist fall or triumph, seeing the knave succeed or be punished are functions of comedy and tragedy, epos and drama, while in a rpg they are often a function of dice and system, which might put them on the spectrum of Alea, but that would ignore a whole emotional dimension of the affair.

In summary
Roleplaying games are structured activities where participants try to achieve their goals through the medium of a shared fiction and within the constraints of the rules.

To be functional, a game must at minimum enable (and - depending on the game - reinforce) the players to recognise, formulate and pursue both long and short term goals within its medium.

These goals can be wide and varied, from experiencing Ilinx to engaging in Agon, submitting to Alea or indulging in Mimicry, seeking Catharsis or practising Poiesis.

Credit where credit is due:
Roleplaying games as conversations is, as far as I'm aware, a thing Vincent Baker nailed down. Short and long term goals as a thing game design must care about is also his.
Big model and Creative agendas is Ron Edwards stuff, Objects is Vincent Baker again, the reduction of both down to dumb "goals" or "aims" is mine. I'm still waffling on the exact term I want to use.
The designer/game as a participant is from various sources but I think Zak was the first to formulate it as "the game (text) is a player" in a way that was useful and productive.
The stuff I got from Caillois I noted in the text.
Parts of this are also cribbed from the Care-Boss principle "system is the means by which participants agree to what happens in play", Vincent's "system provides unwelcome elements" (paraphrasing).