GNS theory is a part of this model of roleplaying developed by Ron Edwards that we call the Big Model. GNS is perhaps the most famous part of the overall theoretical framework, the most contested and the most used for different purposes. It’s also rarely understood very well, which makes it a worthwhile topic here. A discussion at Story Games convinced me that there is still a need for yet another introductory article – I’ll make a point of writing mine in a very exact and simple manner, perhaps it’ll be useful to somebody.
Sourcing the theory
Ron originally introduced the GNS theory in this short article he first published in 2001, I think. The basics of the theory are there, but I don’t think that it’s very fruitful to use that as a current reference – GNS and other matters of Role-playing theory launched a lot of interest in GNS and rpg theory in general, so the theory has been polished and argued over a lot since then.
The tricky part is that, as is the wont of Internet, all this polished GNS discussion is actually just a bunch of shards of discussions that have gone of for years in a multitude of different channels. The best bet of a newcomer to this discourse really is not to go shift through these archives, but rather to ask somebody familiar with the material to explain it, or to read one of these handy summation articles. Like this one I’m writing here, for instance.
Other people who’ve written about this topic include Vincent Baker, Ben Lehman and M.J. Young, to name the few that pop to mind first. Ideally, assuming that we’re all understanding the theory, our writing basically says the same things in all the important points. We might explain ourselves differently, though, which is the motivation for writing more of this introductory stuff – if you think that you understand one of us, but another one seems to be saying something different, then the chances are that you’re not understanding either explanation of the theory. Or we’re disagreeing on something, that’s a possibility too, I guess.
GNS in the Big Model
As I intimated above, Ron also has this thing called the Big Model, which is a more general rpg theory. GNS is often described as a part of the Big Model or an outdated precedessor, but I don’t think that either of these is very accurate. GNS details a very important part of the larger-scale Big Model in detail and provides some unique claims that are not available in the Big Model, either – one might compare this to the relationship between the general theory of evolution and the more specific theory of sexual selection: Darwin used the latter to explain how the former might instantiate in practice among the prominent sexual life-forms. Similarly GNS theory models how Creative Agenda (more on that later) forms cohesive structures out of the Coherence properties defined by the Big Model.
My particular goal here in writing this GNS introduction is to write it specifically with the help of Big Model terminology. Reading the original GNS essay is so difficult partly because Ron didn’t then have the words for describing what he was trying to say in a very exact manner – he only came up with the necessary terminology a couple of years later, when he formed the Big Model. Thus it’s a pretty good idea to write some texts that take the whole of GNS and explain it again in modern terms.
To explain GNS in a modern manner, I first need to explain Creative Agenda, which is the sort of the object GNS categorizes. The Big Model defines roleplaying as this activity that we engage in that utilizes a Shared Imagined Space – roleplaying is not defined by having a role or having adventures or having a story, but simply by having imaginative content we create through the process of play.
(I should note here that my readership will inevitably include folks who’re not reading so much to learn about GNS but to find out whether they can agree with the Big Model. If you find that you disagree with the above definition of roleplaying, that’s cool; for example, I myself find the Nordic theory definition of “immersing in character” to be logically usable, even if the field of roleplaying it implies is narrow, leaves out much of what I find interesting in roleplaying and takes in many things I wouldn’t recognize as the same. For our purposes we’re going with SIS, above.)
As the definition of roleplaying is so wide, different sorts of roleplaying can emerge to utilize the Shared Imagined Space (SIS from here on; I’m being old school here, most people have adopted Ben Lehman’s suggestion of calling this thing simply “fiction” in everyday discourse). I like to compare this to the sort of holodeck they have in Star Trek: anybody on Enterprise can say that somebody is “on the holodeck”, but that doesn’t tell us why or what they’re doing there – they might be using it as a learning tool, using it to play a game, using it to adventure in a historical milieu… all sorts of possibilities here, and most significantly, all sorts of reasons for the people to do those different things with the holodeck, too.
If SIS is the holodeck, then Creative Agenda is the answer to what somebody is doing on the holodeck. What’s more, it’s the specific answer to what they’re doing on holodeck that they need the holodeck for. This latter detail is only important in the larger context of the Big Model, but for now we can assume that our roleplayers are using the holodeck because there is something they want to accomplish that the holodeck specifically allows them to – they’re not messing around with a SIS because they’ve nothing better to do or because they want to impress the girlfriend or any of those other possibilities, but because they genuinely want to do something imaginative and shared with their friends. Then we can say that Creative Agenda is the answer to what they want to do.
So Creative Agenda is a declaration such as “I want to create a great story” that explains why somebody is messing around with a Shared Imagined Space. It doesn’t have to be verbalized, and often isn’t – the notion of “agenda” here is that people can want things they don’t know they do, or that they don’t know how to achieve. I’m sure most people have had this sort of experiences in many different facets of their life, when they realize only afterwards that they wanted something that they didn’t realize was missing. Like all other points here, I can write about volition vs. Creative Agenda in more detail later if necessary, for now let’s accept that a person who wants to roleplay generally speaking has a Creative Agenda to go with this wish, explaining why he wants to roleplay.
(It’s notable that I’m not discussing the general properties of the group of Creative Agendas in much detail here. One might ask whether there is an infinite number of different CAs, for example, and many other similar questions. We can discuss these things in more detail at some other time if you’re interested.)
Now that we know what a Creative Agenda (CA; I’m going to start abbreviating this, too, to make this lighter to read) is, we come to another central Big Model concept also necessary for explaining GNS in a modern manner. Coherence is a quality of the Creative Agenda that pertains to how players interact when playing the game – it’s ultimately a social issue having to do with communication and appreciation of interaction. (Remember that we’re ultimately defining the whole act of roleplaying as manipulation of a Shared Imagined Space: genuine interaction is necessary for this to happen in the first place for a space that is “shared”.)
Defining Coherence comprehensively is a very complex issue, so what we’re going to do here is to give a good faith working definition and work from there. The following might raise some nit-picky questions, but it should be enough to help realize what we’re going for.
Two players have a Coherent relationship of play if and only if their interaction is fulfilling for both of their Creative Agendas – fulfilling meaning that they both are fulfilling the goals their CAs depend on. We call a play relationship that is not working towards this sort of fulfillment for either or both parties incoherent.
So Coherence is basically the same as being able to genuinely work together for the benefit of both players simultaneously. More generally, we say that a whole group of players is playing in a Coherent manner if they manage to work towards the fulfillment of the whole group’s Creative Agendas. Indeed, we even say that two CAs are Coherent towards each other if they effortlessly and naturally fall into this sort of mutually beneficial relationship.
(I should note here that we also call many other things Coherent based on this basic property. For example, a game text might be Coherent if it instructs players to come to the table with Coherent CAs and incoherent when it does not comment on the topic or, worse, displays support for several, mutually incompatible CAs.)
When a whole group is playing Coherently, we call their common appreciation of their mutual goals the Creative Agenda of the whole group. In this manner an individual’s Creative Agenda, when it matches that of other players in the group, transforms into a common agenda for all. For example, if all the players want to create a story about the vampiric struggle with the Beast within, then we, ceteris paribus (and man o man what a reservation this is in this particular example…), say that the group has a Coherent Creative Agenda. All the players are not necessarily wanting to do the same exact thing (like one might prefer being the GM while another wants to play the protagonist), but their exact roles in the game will still support each other, and they appreciate the goals and actions of each other, understanding them.
GNS theory: CA modes
Now that we know what a Creative Agenda is and what it means when Creative Agendas are Coherent, we can finally look into what, specifically, the GNS theory claims. The first of these claims concerns the concept of CA mode. A short definition for that, too:
A Creative Agenda mode is a set of Creative Agendas that are all Coherent with each other and incoherent towards other Creative Agendas.
So that’s very simple – we call a set of CAs a mode if and only if they form a well-defined equivalence class regarding the Coherency relationship; any Creative Agenda is either in the mode or outside it, there are no liminal cases.
The first claim of the GNS theory (although one often interpreted as a necessary result of the Big Model in general as well) is that such things as CA modes exist at all. This is not obvious by any means – many critiques of the theory can basically be interpreted as attacking against this often implicit claim. Thus it’s perhaps best if I spend a few paragraphs arguing for why modes can be said to exist.
The first, and most obvious, argument is empirical: we can examine our own roleplaying experiences and accounts of others to form an overall landscape picture of how and why people actually roleplay. Then we can perhaps recognize groups of Creative Agendas that work better or worse together, and ultimately come to find a mode that is reliably Coherent internally, but has some sort of incoherence trouble when interacting with other sorts of CAs. This was pretty much what Ron originally did when forming the GNS theory, he observed play and recognized a number of CA modes.
A second way for establishing that CA modes exist is to define one and demonstrate how it fulfills the requirements above. This is possible if we’re willing to work in concrete sociopsychological terms – we have certain human drives that are fulfilled by certain sorts of roleplaying, and we can determine how these goals interact with other CAs in a social setting. This sort of discussion has been worked out after the empirical stage after Ron first established his model – the CA modes he recognized can be affirmed to exist to some degree by further theoretical models based on widely agreed-upon understanding of how humans work.
A third, more theoretical way for establishing the existence of CA modes is to analyze the Coherence relationship and display that it fulfills the set theory requirements for an equivalency relationship – that is, if one can demonstrate that Coherence is transitive, reflexive and symmetric, then the set of Creative Agendas necessarily forms one or more CA modes. This relates more to examination of CA Coherency than GNS per se, though, so I’ll leave it for another time.
Because this is a basic account and I don’t have all day, we’ll leave in-depth examination of this question for some later date. For now it’s enough to note that although the GNS theory is usually formulated in absolute terms that use the concept of CA mode, most of the later results hold some relevance for weaker formulations as well – we don’t, strictly speaking, need such a clean and simple structure in our human Creative Agenda landscape to make use of the related observations; even a critic who remains unconvinced of the existence of CA modes after reviewing the in-depth arguments would usually agree that the CA modes posited by GNS are, if not pure modes, then at least groups of CAs that display a remarkable amount of practical Coherence towards each other and some serious challenges to such towards other CAs. So even if the claims GNS theory makes prove not to be hard-coded into our human relationships as 100% certain and absolute truths as GNS claims, minor exceptions won’t impact on the overall usability of the theory’s results.
GNS theory: the three modes
The actual meaty, strongly argumentative core of the GNS theory is the claim that there exist three recognizable CA modes, which are also named and defined by the GNS theory. The theory doesn’t claim that these are necessarily the only possible CA modes, only that these three are the ones easily observable in rpg history. The three recognized by GNS are as follows:
- Gamism: this mode includes all CAs where the players concern themselves with recognizing and resolving challenges, enjoying the excitement of struggle and the social esteem, however playful, that accrues to the victor. The reason for why Gamist agendas tend to be highly Coherent towards each other is tentatively that players interested in this sort of performance-based entertainment are equally interested in their own performance as they are in that of others; the personal achievement is largely established in relation to how the others are doing. This is also why Gamism doesn’t cohere with other sorts of CAs: if the other players do not provide the social esteem you’re looking for in performing well (that is, if they’re not interested in your performance), then it’s all a waste of time. Just like playing Chess against somebody who doesn’t care to try for real.
- Narrativism: this mode includes all CAs where the players concern themselves with establishing thematic questions within the SIS and then answering them to create story dynamically through player interaction. This mode is difficult to understand for many, and it’s often misinterpreted as being a generic “I like story!” agenda; my personal theory is that this is because we didn’t actually have anything in our culture that consistently provided Narrativist enjoyment before roleplaying, so we’re having to struggle with learning to distinguish a wholly new thing here. Narrativism tends to not play well with other sorts of CAs because the creative structures used in dynamic story creation (as opposed to having the GM prepare a story beforehand, you understand) tend to be impossible to use if the whole group is not on board with the effort – and, of course, if the rest of the group doesn’t play along, then you can’t create story together, but have to do it alone instead.
- Simulationism: this mode includes all CAs where the players concern themselves with experiencing the SIS matter they introduce as what it is, without bringing in external ambitions apart from directing where the experience is going. Simulationist CAs tend to crash and burn if the other players are not on board because players with other sorts of agendas are usually willing to bend the SIS to serve their own CA in ways that seem shallow and pointless from the Simulationist point of view – there is no such thing as perfect realism of game experience, but the places where a Simulationist decides to save on effort are often quite different from where others might, resulting in others being bored by the content a Simulationist produces and the Simulationist being frustrated by the disinterest.
Now that we have named some CA modes, it should be noted that people will often use expressions such as “my Creative Agenda is Gamist” to mean that their current CA falls into the Gamist mode. This is pretty clear, but people also might say “I am Gamist” to mean the same thing, or perhaps that “most of my roleplaying is done to fulfill a CA that falls within the Gamist mode”. This is all just language use and shortcuts that make sense, but probably quite a bit of the misunderstandings related to GNS are based on misreading these statements. This is especially true when we get “game text X is Gamist”, which means “game text X primarily provides support for and encourages a CA that falls within the Gamist mode” and “our group is Gamist” and all sorts of other transferences of the terminology. These are not at all difficult to understand if you understand what the CA modes actually are, but if you don’t, it’s easy to interpret them as some sort of gamer or game text classification scheme.
GNS claims that not only are there Creative Agendas, but also that some of them work better together than others, in a phenomenon called Coherence. Furthermore, CAs form discrete families called CA modes, which allows us to learn about the internal structure of the field of CA. Based on practical observation GNS claims that the field of CA is divided between three major CA modes called Gamism, Narrativism, Simulationism.
As can be seen, the GNS theory itself is really simple, at least if you paid attention when I defined Creative Agenda and Coherence, above. What makes this stuff so important is the fascinating proposition of examining roleplaying from the viewpoint of social agenda, and bravely imposing structure upon that chaotic field by determining structural rules of human behavior that manifest as GNS modes. It’s a very ambituous model that has something most other models of roleplaying have lacked, an unified explanation of the phenomena of actual play interaction.
GNS, when properly understood, can be used as an analytical tool for developing game design and personal play practices in different ways. It is also useful as a more generic framework for working with Creative Agendas, even when the work happens on a detailed level within individual GNS modes – the modal theory includes plenty of detail on how and why we humans want to engage in play within one of the modes anyway, so understanding those parts of the theory for your preferred mode is useful for fine-tuned rpg work within that mode. For example, understanding the Egri-derived premise-setting model of Narrativism that Ron developed in his further work on Narrativism is useful for developing Narrativist games and becoming a better player within this mode – and similar specific modeling is available on the other modes as well.
If you know your GNS theory and disagree on some point, feel free to correct me in detail; I reserve the option to fix this article for factual accuracy if that becomes necessary, just because people will be reading this later and referring to this whether it’s accurate or not. If you don’t know this stuff and have questions, feel free to ask them – I might well write more about the Big Model if there’s interest in this stuff out there.