Shared Imagined Space, Shared Text

[Victor Gijsbers]

In the Provisional Glossary, Shared Imagined Space is defined as:

The fictional content of play as it is established among participants through role-playing interactions.

This is open to different interpretations, as it is not immediately clear what kind of thing the 'content of play' is. However, the very name 'Shared Imagined Space', as well as the definition of the Lumpley Principle:

"System (including but not limited to 'the rules') is defined as the means by which the group agrees to imagined events during play."

seem to imply that we are talking about a space of events, the fictional contents of which are described by those statements made by the players which are accepted (not successfully challenged) by the group. The Shared Imagined Space, then, is a possible world, and play is the process of adding elements to this possible world.

Against this, suppose that play does not so much consist of adding elements to a Shared Imagined Space, but of adding statements to a Shared Text. The activity of roleplaying then becomes like the collaborative production of a text; whenever I say something, it is a proposal of adding this line to the Shared Text.

This is not merely a different way of saying the same thing, for the following reason: there is no one-on-one correspondence between texts and possible worlds. Two subcases of this claim which appear to me to be important to roleplaying are the following: 1) a text can be the description of more than one possible world; 2) a text can be the description of no possible world. In addition, 3) the adding of statements to a text is not the same process as that of adding elements to a possible world. In the rest of this post, I will explicate what these concretely mean for roleplaying. (I believe that the second and third of these are actually interesting for playing and designing RPGs; I have more trouble envisaging this for the first case.)


1. One text, different worlds: ambiguities

Probably the least interesting, if the most common, form of a difference between adding to the text and adding to the world arises in the case of ambiguous statements. Now I am very willing to claim that every statement is ambiguous and that this point generalises to every statement made in every RPG ever played - but that would serve no purpose in the present context. So let us look at statements that have two, clearly very different, meanings. Allow me to take a stupid example:

"There is a fish in a tank."

This could mean that there is a fish in some sort of aquarium, or that there is a fish in a big steel military vehicle. Especially in a weird, absurdist game with animals acting like humans, players might not all interpret the statement in the same way. So it is entirely possible that the statement goes unchallenged; and that half of the players add a huge fish bowl to their Imagined Space, while the others add a military vehicle driven by a fish. The statement is accepted, through the system, but no element is added to the SIS.

Does this mean that something has gone wrong in the roleplaying? No. Suppose that the ambiguity never becomes important in the rest of the game, so the difference in interpretation is never discovered. Then the game will simply move on, entirely functional, and the system has done exactly what it had to do. What was that? Decide whether or not a given line is added to the Shared Text.

Alternatively, suppose that the difference in intepretation is discovered after a while. Then it can be disambiguated. How? By adding a new line to the text, that bars one of the possible interpretations, reinforcing the other. We do not return to a failed act of adding an element to the SIS in order to make it successful. Instead, we decide which line gets added to the Shared Text ("the fish drives at maximum speech" or "the glass of the tank shatters when it is hit by a stone"). The first line still acts as a restriction to what happens next, because it has been successfully added to the text. If it were a failed attempt to add an element to the SIS, it could not restrict what happened next, because it failed; that fact that it does restrict what happens next shows that it has been successfully added to something - and that something can only be the Shared Text.

2. One text, no world: inconsistent fiction, unreliable storytellers

In twentieth century fiction, it is not hard to find stories that are inconsistent. They contradict themselves, and do not resolve that contradiction. No possible world corresponds to such a story, since no proposition can be both true and false in the same possible world. The reader of such fiction may wonder which story is the 'true' one, but this would be a misguided question - there is no true story. The text does not show us a possible world.

The same can happen in roleplaying. (Roleplaying is, as far as I can see, as yet curiously untouched by the more daring experiments in modern art.) I have played a game in which this happened. I quote my actual play report:

Indeed, one of the scenes was even told four times, at least three of which were completely contradictory: the character - who had worked on the ship and truly hated the captain - tried to kill the captain with a knife, and succeeded, the blood splashing over his hands. Cut. Red sand covers the hands of the character, lying in a desert, and he remembers how he tried to kill the captain, who was much faster than him and had easily overwhelmed him. Cut. The character dreamed once again about killing the captain, but he spoke in his sleep, was heard by fellow sailors, dragged onto the deck, faced a mad captain, and hauled off into a boat to be left in the desert.

It was never established which of the versions was the real version: dream, hallucination and reality simply became an inextricable web wherein it was impossible to judge whether any individual scene had taken place or not.

The possibility of contradictory texts being accepted as satisfying roleplaying experiences shows that roleplaying is not the process of creating a possible world, but of creating a text. Whether or not that text is the description of a possible world is left open.

The same happens when the roleplaying uses unreliable or potentially unreliable narrators. In such a case, we are not establishing what happens in the SIS, for everything remains open to question. It always remains to be seen whether the (explicit or implied, fictional or actual) speaker of the text spoke the truth. Although in this case we can have an implied possible world ('that what really happened'), no statement added to the text can be said to establish anything about the possible world. What we are doing is adding lines to the Shared
Text.

3. Adding to the text, changing the world

Yesterday, I played a game of Universalis, in which a little girl got stuck in a haunted house (on a fair). Then, suddenly, a monster laid a hand on her shoulder and screeched that it wanted to drink her blood, after which she fought the monster and tried to escape.

Next scene: set in the past. A drunken farmer enters the haunted house, sees a little girl, puts his hand on her shoulder and schreeches that he wants to drink her blood.

What has happened? There was an element in the SIS, on which everyone agreed: there is a monster attacking the little girl. Now, the new lines of the text suggest that no such thing happened - indeed, it appears as if the little girl only mistook the farmer for a monster. (Things are complicated by the fact that this was not explicitely stated.) It is certain that adding these new lines to the text does not have the simple effect of adding new elements to the SIS. It also takes some old items out, or at least radically transforms them. Does this mean that adding these lines to the text entails that the authority of the previous narrators (those who told the first scene) to add elements to the SIS has been called into question? Or does it mean that their authority is instead respected - this authority being the authority to add lines to the text (which are now given a different interpretation, but respected as part of the text)?

I suggest that the second reading is much more natural. Indeed, if we - the players - had taken the first reading, we should have seen the whole first scene as a very eleborate challenge to the coins spent in the first scene. (Which in addition would have been against the rules of Universalis.) We did not play it as such. No one ever thought of playing it as such. It would have been bizarre to insist on that. Does this not prove that the dynamics of such a game of Universalis can be better understood as the creation of a Shared Text than as the creation of a Shared Imagined Space?

(This whole example could be interpreted as an instance of an unreliable narrator - the little girl being the implicit narrator of the first scene. It could be argued that all of my three points are fundamentally the same point. But let's not focus on such questions.)


My suggestion, then, is that one can make better sense of actual roleplaying experiences by substituting the notion of Shared Text for that of Shared Imagined Space, and reformulating the Lumpley Principle accordingly. In addition, it would stimulate the creation of more 'avant garde' roleplaying games using the complex techniques of contemporary fiction.

And my question is: what do you think of these suggestions?

[Callan S.]

Hi Victor,

Just a quick question: Are you suggesting something like the text is the only shared property amongst a group? And any imaginary scene conjured by it in a readers/listeners mind is only a derivitive of the property which is shared (the text), rather than this imaginary stuff actually being a shared property itself?

Sounds reasonable so far.

[Vaxalon]

Yes, this sounds perfectly reasonable...

Except that (in a face to face game) the text is usually virtual at best.  People don't write down what they say, except when it's really really important, like a riddle or a mission.

Each person, in addition to having an unshared imaginary image of what's happening, have an unshared memory of what has been said, and can be mistaken about it.

[Victor Gijsbers]

Hi Callan,

Just a quick question: Are you suggesting something like the text is the only shared property amongst a group? And any imaginary scene conjured by it in a readers/listeners mind is only a derivitive of the property which is shared (the text), rather than this imaginary stuff actually being a shared property itself?

Well, partly. I am not suggesting that the text is the only shared property amongt the group. There are surely many instances of play in which there is a Shared Imagined Space; and there are many actions in roleplaying that add an element to this Shared Imagined Space. So what I am saying is not that there is no SIS, nor even the rather uninteresting thesis that we cannot know what our fellow players are imagining. However, I am suggesting that:

[*] Roleplaying always involves the creation of a Shared Text, but only sometimes involves the creation of a Shared Imagined Space.
[*] Recognising this allows us to make better sense of roleplaying experiences involving inconsistent texts, unreliable narrators, ectetera; and will help us in designing games that use these elements.
[*] There are successful acts of roleplaying that amount to adding a line to the Shared Text, but cannot be interpreted as adding an object or event to the Shared Imagined Space.
[*] System cannot be anything but the means by which the group decides to add lines to the text; and recognising this allows us to make better sense of roleplaying experiences involving recontextualisation, such as my Universalis example above.
[*] Hence, we might want to talk about the Shared Text instead of the Shared Imagined Space, both because it allows us to make better sense of roleplaying and because it allows us to think clearer about game design.
[/list:u]


Vaxalon, I must apalogise for not making clear enough what I mean by 'text'. Following the use of this term in much of the Humanities and continental philosophy, I mean by 'text' any body of signs, whether written down, spoken, or even abstract. So in the case of the usual face-to-face RPG (but not in that of an IRC or forum-based RPG), this text would be a spoken text.

(The next paragraph is not really central to my argument, and may be skipped by those happy to think of the text as a spoken text.)

Strictly speaking, that is not true either. Perhaps I should have called the Shared Text the Shared Imagined Text, because it is a text that is imagined by the people and not written down, recorded, or perfectly remembered from spoken instances. This can not only be seen from the fact that people may have an imperfect memory of what has been said, but even better by the following example. Suppose John plays a character Frank, and says: "I decapitate the orc." If his statement is accepted, the line that is added to the text is not "I decapitate the orc", but "Frank decapitates the orc". So there is some grammatical transformation taking placeto transfrom the spoken statement into the line that gets added to the Shared Imagined Text. The SIT, then, is an abstract text that has its real basis in the minds of the players. (So I am not proposing a behaviourist RPG-theory.)

[lumpley]

Victor, I think you're right.

I'd even suggest that the most profitable reading of Ron's glossary and essays treats "SIS" as precisely the text.

Certainly the loody poody is concerned with what's communicated, not what's imagined.

(And every time I say that someone comes back with "communication starts and ends with imagination." Yeah, I know, you're right, please read corresponding nuance into my "not concerned with.")

-Vincent

[Victor Gijsbers]

Hi Vincent,

Certainly the loody poody is concerned with what's communicated, not what's imagined.

The word 'communication' is so ambiguous that I am bit careful in concluding that we are saying the same thing. My claim is that the Lumpley principle is concerned with which sentences get added to the Shared Imagined Text, not with which objects and events get added to the Shared Imagined Space. Both cases are cases of communication; that which is communicated might even be the very same thing. The question is rather: what are we constructing, a text or a possible world?

At least one article in your blog made me believe that you would defend the latter idea, namely that the act of roleplaying is the act of adding elements to a possible world. If you'll allow me to quote your post Established but Unknown in order to make sure that we are saying the same thing:

All the players can control the unknown-but-past, like in Univeralis, or just a few or one can, like in Dogs - it depends on what you want out of this game. But only the group's informed agreement can possibly control the unknown-but-future.

What this passage seems to suggest is that there is a qualitative difference between adding elements to the possible world's future and adding elements to the possible world's past. Such a difference may exist if roleplaying is the act of adding elements to this possible world; but I find it hard to see how it could exist if roleplaying is the act of adding elements to the text. There is, after all, nothing special about statements concerning earlier states of affairs coming after statements concerning later states of affairs in any given text.

When we talk about the SIS, the 'unknown-but-future' is the unknown future state of the possble world. But when we talk about the SIT, the 'unknown-but-future' is the unknown future state of the text; that is, all the lines that will get added to the text in the course of the game. This means that there is no 'unknown-but-past', because the past is simply the text that has already been agreed upon (and is thus known). (Wild suggestion: the existence of a definite 'past' text, a text which can no longer be changed, is the difference between roleplaying and collaborative story writing. Let's not discuss this suggestion here.)

My suggestion would be that qualitative differences in the experience of roleplaying (except those that result from particular systems) can only have to do with the past/future-distinction on the level of the text, not with the past/future-distinction on the level of the possible world. This is a straightforward application of the broader theory I outlined above. If you agree with my posts, then, and I am reading your blog correctly, would you say that what you wrote in your blog is mistaken? Or has my discussion laid bare a difference between us?

[Alan]

Hi all,

Forgive me if I'm being pedantic, but I'm confused by semantics here.   The word "text" to me refers to printed words on paper.  I've looked this up in the dictionary and it agrees.  Since there is usually no text produced from the events of a roleplaying game, I don't know what text you're referring to.  

Is this thread using the word text in some special way?  Is it possible that there's a better word to use?

[Vaxalon]

Vaxalon, I must apalogise for not making clear enough what I mean by 'text'. Following the use of this term in much of the Humanities and continental philosophy, I mean by 'text' any body of signs, whether written down, spoken, or even abstract. So in the case of the usual face-to-face RPG (but not in that of an IRC or forum-based RPG), this text would be a spoken text.

This really doesn't alter what I'm saying; the text exists only in the moment in which it is spoken; after that, all the roleplayers have is their memories of the text.  Those memories are just as imperfect and unshared as their imaginations are.  If you define the "shared text" as the subset of the individual texts which all of the participants have in common, it could be said to exist on a semiotic level, but it's not something that the participants experience or interact with.  It's more of an ideal that they work towards.

[Victor Gijsbers]

Is this thread using the word text in some special way?  Is it possible that there's a better word to use?

I tried to answer this question in my second post; is it still unclear? Perhaps you can find a better word, but my use of the word 'text' is that of many 20th century writings in continental philosophy, structuralist and post-structuralist literary theory.

This really doesn't alter what I'm saying; the text exists only in the moment in which it is spoken; after that, all the roleplayers have is their memories of the text. Those memories are just as imperfect and unshared as their imaginations are. If you define the "shared text" as the subset of the individual texts which all of the participants have in common, it could be said to exist on a semiotic level, but it's not something that the participants experience or interact with. It's more of an ideal that they work towards.

The Shared Imagined Text is an ideal, non-empirical concept just as much as the Shared Imagined Space is. It is, if you wish, that part of the constructed text which is memorised by the players or written down for later reference. This means that it is not a perfect record of that which has been said. In all of this, there is no significant difference between the SIT and the SIS.

I hope I have made clear that the reasons for choosing to use the SIT instead of the SIS are not that it is more 'empirically accessible' or 'perfect' or 'shared' than the SIS. If not, please tell me.

[Ron Edwards]

Hello,

Victor, your use of the term "text" is synonymous with my "shared imagined space." Or to put it better, I chose to describe the characteristic spoken/communicated text of role-playing as "shared imagined space."

So it seems to me that what you're doing is simply clarifying your own understanding of my term by realizing that I was talking about text (as we are using it here) the whole time.

I hope people realize that such a text would be affected by, and thus include, all communications among the people involved, not just the ones which describe imaginary things. That's what the "shared" is doing in there.

Also, as a side note, I hope people can now see why I consider separating "in-game" vs. "out-of-game" typed-text in IRC play to be damaging to the enjoyment of play. I vastly prefer mixing it all together as in Code of Unaris chat-play.

Best,
Ron

[Victor Gijsbers]

Hi Ron,

So it seems to me that what you're doing is simply clarifying your own understanding of my term by realizing that I was talking about text (as we are using it here) the whole time.

Everything I say and write should be understood as an attempt to clarify my own understanding - and hopefully that of others as well. Nevertheless, I am a bit surprised to hear you say that my SIT is a synonym of your SIS, given that the "Provisional Glossary" contains such expressions as:

sufficient to introduce fictional characters, places, or events into the Shared Imagined Space.

Obviously, one cannot add characters, places or events to a text - one can only add sentences to it. Perhaps this is merely a somewhat sloppy but more easily comprehensible way to say what I have been getting at in the previous posts, but given the different readings supported by the "Provisional Glossary", spending a topic to clarify the distinction between a possible world and a text does not seem like wasted time.

I hope people realize that such a text would be affected by, and thus include, all communications among the people involved, not just the ones which describe imaginary things. That's what the "shared" is doing in there.

Here we radically part ways. Indeed, what you are stating here is in stark contradiction with the lumpley principle. "All communication among the people involved", that means that there can be no selection as to what statements are accepted into the SIT/SIS, as every statement that is made is communication and hence in it immediately. This seems to me a radically different use of the term SIS/SIT than is usual on this forum. Would you agree with that? If you mean SIS to include all communication, I feel that 'SIS' becomes synonymous to 'roleplaying session/campaign'.

I meant the term SIT to specifically describe a text that is constituted in play through a process of offering and selecting statements, where the selection is governed by the system, conform the lumpley principle.

[am]

Victor,
I am very interested in reading your article (book?) regarding the Poetics of Roleplaying. I am currently working two papers for my thesis regarding rps and performance theory. I haven't started the writing process yet, just the research process. Could I get a copy of that work whenever you finish?
Your postmodern role-playing game also perked my interests. How exactly would that work? Which postmodern theorists are you drawing from for this? The idea sounds amazing.
~ Annamaria

[Victor Gijsbers]

(I have answered Annamaria through private messaging.)

[Sean]

Hi Victor,

I thought along these lines a while back. I decided the reformulation was unnecessary for reasons I'll get to in a second, but given that you're onto these issues at all I tend to think that you're in the same 'neighborhood of understanding' as Ron and Vincent with respect to this part of roleplaying theory. The exact phraseology still seems to me to be up for grabs, if it matters.

I don't like 'text' because of the connotation of physical text that's bugging Alan. Also, I think this word is so hideously ambiguous, as it is used even by experts in postmodern theory and the Derridean wing of philosophy, that it's unhelpful as an analytic term - certainly less (or no more) helpful than 'communication'. YMMV.

But OK. If the point is that roleplaying rules generally and for the most part directly constrain what gets let into the spoken/vocalized/public arena and not what you've got in your head, I think most of us probably agree about that. The problem - as Vincent's irritated aside seems to me to indicate - is in the idea that this exists independently from what's going on in individual people's heads.

You get yourself into the metaphysical soup on this one. If you don't like the Heideggerian literalism that's possible to read into 'shared imagined space', you might go to 'text' and pat yourself on the back for being a better materialist than was Heidegger. But then you think about it more and say "well, but they don't really have the same text in their heads either, so..." and then you're off to the races with "Shared Imagined Text" and simulacra of simulacra and more metaphysical nonsense. I find it easier to just punt on the twentieth century and talk about words and thoughts and physical actions and imagination but I think we're all talking about the same stuff here so it's only a question of how well the vocabulary helps us do it.

What we imagine for ourselves matters crucially; rules don't constrain that, but what's socially communicated in the public space of the game does. (In theory, yeah, sure, whatever, you could come up with arbitrary mental pictures and stories to match what's said in play, but, well, grow up. That's not how human communication works even if it's logically possible that it could work that way. Proust and model theory are just as wrong as Heidegger.) And so if rules constrain the latter there's still a second-order constraint, so the rules wind up exerting a kind of constraint on what we imagine for ourselves as well after all.

It's true that there are a lot of things in rp where it really doesn't matter how you're thinking of it vs. how the guy next to you is thinking about it, and that's cool and OK. And so if you want to say there's no SIS because of that possibility I don't think anyone who understands what's going on here will disagree with you. On the other hand if you say 'well, but there is a text here', you're off to the races again, because then what's the text? ("Hmm...well, we don't really all have the same text..."...epicycles, or relations between relations in scholastic philosophy. You can go on with this regress forever.)

What we're looking at is a social activity which includes individually imagined stuff and acts of communication ('text', it's the roughly the same in ambiguity IMO) and physical actions and a framework of rules which helps constrain, guide, encourage, and orchestrate all these different parts of what goes on in an RPG.

The 'shared' in the 'shared imagined space' is the principles of coordination and mutual editing for our individual imagined spaces. These principles are usually second-order because most rpg rules and props don't say 'imagine this', they say 'imagine something that would fit with this'. (But consider the difference with showing your group a picture vs. stating what just happened here.) But 'something that would fit with this', while it's not totally constraining - and hence there's no guarantee that all the stuff individuals imagine will be isomorphic even when they are all isomorphic to the public statements in the game (you could say 'the SIS underdetermines the IISs') - is also not totally unconstrained in the way that analytic philosophers besotted with model theory and continental philosophers drunk on Derrida would have you believe. Back in the real world, most of the time, most of what most of us are imagining is pretty close, and you can tell by the way we check each other and keep tacking back to the same points in what we say publically.

So anyway. It's experientially different , to be working with a group of people in a way that involves your own imagination, then to just lie on the couch daydreaming and imagining stuff for yourself. It's also different getting into dialogue with others and negotiating out the elements of imaginative exploration than it is say telling a story or filling up a binder with stuff about your rpg world at home.

If you want to call this difference the difference between a shared and an individual imagined space, I don't see why that would bug us, unless we wanted to do metaphysics.

On the other hand, if you wanted to avoid the term 'shared imagined space' because of the metaphysical connotations, that would be cool too. Then you could talk about principles of coordinating what people say and the way they establish constraints ('things said') for everyone's individual imagining. But then if you talk about this as 'rules' or 'text' or 'public constraint' you get the opposite metaphysical confusion, the idea that playing doesn't really involve individual imagination at all - behaviorism. But then if you don't like behaviorism you say 'well, we've got to bring the individual back in, individual interpretations of the shared social stuff matter, so it's not just a text, it's a shared imagined text'. Layers upon layers. You can repeat these moves forever.

Best,

Sean

P.S. I think the idea of multiple contradictory interpretations of the same event in an RPG is cool, and maybe deserves the sobriquet 'postmodern', but it does not indicate that there is no possible world under discussion. At most one such account can be true of any imagined space, that's all. If you refuse to decide which, that's fine too. But all you've done here is broaden yourself to a class of indicated possible worlds at most one of which can hold.

In other words the contradictory stories case is just the same thing as the ambiguity case, except that there are some ambiguity cases where they could both be true, and in the contradictory stories case they can't all be true - unless you invent more story to explain how. But anyway, that's my take on that.

[Ron Edwards]

Hello,

I'll step out of this conversation. I think the distinctions being drawn are largely meaningless. However, if others are getting somewhere with it, then please continue.

Best,
Ron