Behavior Versus Intention

[M. J. Young]

I know we've discussed it before; and I know that Ron insists GNS has nothing to do with "intent"; and I know I'm on the record as disagreeing with that.

But the subject has returned, and completely derailed the http://www.indie-rpgs.com/viewtopic.php?t=7307">Criticisms of the Threefold thread, so I'm trying to give it its own space.

Let me present the difference between behavior and intent, as I, as a student of law, would understand it.

Behavior: Joe chose to do this, which addresses the premise.

Intent: Joe chose to do this, because it addresses the premise.

Now, here's what I see in GNS analysis. We say we're collecting data points, looking at all the things Joe chooses to do, and trying to understand how regularly he chooses to address the premise, how often he chooses to meet the challenge, and how much he goes by what is realistic and exploratory. We further believe that for most gamers, a pattern will emerge that shows a tendency to make choices which do one of those things more so than the others.

We also recognize that there are some decisions which could be go any direction:
Joe chose the long bow because it is the best tactical missile weapon and elves get bonused with long bows.
Joe chose the long bow because it seems to him that elves realistically would use long bows in their woodland homes.
Joe chose the long bow because part of the theme of play involves issues of cowardice, and this weapon allows him to face those issues as one who fights from a distance.

All we know behaviorally is that Joe chose the long bow. Therefore, the datapoint of behavior tells us nothing.

However, some datapoints are revealing, as they tend to represent one category over the others. We expect that ultimately we'll find these data points congregating in one area.

At which point, what we know, behaviorally, is "Joe has made more decisions which address premise than which meet challenge or promote exploration."

Which is to say we have learned nothing about Joe or his decisions. We could as easily roll a three-sided die a hundred times and note that three came up more often than one or two. What Joe has decided in the past tells us nothing about what he will decide in the future; they are datapoints, nothing more.

However, when we look at:
Joe has made more decisions which address premise than which meet challenge or promote exploration.
We infer
Joe has made choices in order to address premise more than challenge or exploration.

And at that moment we have intent.

All the data on which we base our decisions must be behavior. Even when we analyze ourselves, behavior is far more telling than any feelings we have. However, the conclusion that someone plays in a narrativist manner either means that they have in the past happened by chance to make decisions which address premise (but are by no means more likely to do so in the future) or it means that we believe they make these choices based on a desire to address premise.

Further, if we mean no more than that their choices happen to address premise, then we cannot possibly have GNS conflict, because there is no reason for choices to conflict. We can only have GNS conflict is few take the step of inferring from that behavior that Joe wants or intends to address premise, and that others in his game have conflicting intentions.

Thus what GNS tells us is that the player intends to make choices based on one of three general goals. The analysis cannot begin with that intention, but must ultimately extrapolate that intention from the observed behaviors.

*****

Part of the problem in this discussion arises because most people (those not trained in law, I suppose) don't understand the nuances of double intention. That is, I intended to aim this gun at him and pull the trigger is not sufficient to support a charge of murder; you also need I intended to kill him. If the gun was supposed to be a prop in a play we were acting, but someone replaced it and I was unaware of this; or if I thought the gun was a toy, or believed it was not loaded or that it had been loaded with blanks; or if we were rather foolishly testing a supposedly bullet-proof vest and it failed; or if he was threatening to kill me, and I wished to disable him by shooting him in the leg or the arm, but instead hit a more vital area--in those cases, I did not intend to kill him.

In the case of GNS, when those of us who argue that intent is part of the model use that word, we mean no more than "Joe intends to address premise when he chooses the long bow." We do not mean to reference the next layer of intent, that is, why Joe intends to address premise. Nor do we have any suggestion that Joe has verbalized the intent, even internally, in these or similar words. All we mean is that narrativism is defined as Joe makes choices which address premise because Joe wants to address premise.

If you do not conclude that all those data points represent an intention on the part of the player toward a particular GNS category, then they have no more meaning than how many stars are in the eastern sky as compared to the western sky. When you say that those data points predict something, you mean that there is a cause behind them which will remain consistent; the only thing that can cause such a pattern in human choices is intent. The alternative is randomness, and randomness cannot be predictive of anything other than future randomness.

Now, let the fur fly.

--M. J. Young

[Marco]

Hi MJ, I agree with your sailent points by and large--yes, to predict future behavior, one must start ascribing some driving force behind the data-points and that's basically gonna be intent.

Stocks have no intent therefore "past performance is no indication of future behavior." I doubt most people would approach the analysis of a fellow gamer^H^H^H^H^H human being the same way.

What got us derailed in the other forum (weirdly) was the flip side:

I get handed a weapon list and I go down the line thinking "what to choose, what to choose ..."

I think at that point I can figure out my intent:

"Ooh! Longbows do 2d6 and get 20-yard range! That's my bow! Come get some!"

Or maybe "Well, the crossbow is superior--but an elf with a *crossbow*--no way!"

Or: "Themes about nature over technology and the question of progress vs. life? Okay--metal is mostly out--and crossbows are more 'advanced' than longbows ... so ... the long bow it is."*

That's a single vaild data-point. Assuming no presence of denial in weapon choice (which, I'd thought, would be a pretty safe assumption--it seems not).

GNS may well not *require* an implied motive to get some value out of (although I think your point is well made: the trend is the motive)--but sometimes you do have motive ("Yes, your honor, I chose the long bow for 2d6 damage at long range.") and that's not at odds with the theory either.

So the idea that GNS has "nothing to do with intent" doesn't seem right--rather: it seems to function (IMO) best with known intent and is functional with "implied intent"--a belief that a trend will continue because it has an intellectual cogent choice behind it.

While still not requiring that you know for certain what's in Joe's head.

-Marco
* For Contracycle: it should be noted that I've made the first two decisions cognitively in real-life play--I've chosen less effective weapons because they fit better in some cases and gone for the most effective weapon (step-on-up) in others. The fact that I can be aware of what I'm doing and why and how it suits my mode is both interesting and valuable to me.

[Ian Charvill]

I think the issue is less that intent is involved as that intent is a fairly nebulous term.  What does it mean to intend something?  Sure, as a lawyer, one may have particular specific meanings in mind but one cannot guarentee that these will carry over into general discussion.

Further, if we mean no more than that their choices happen to address premise, then we cannot possibly have GNS conflict, because there is no reason for choices to conflict. We can only have GNS conflict is few take the step of inferring from that behavior that Joe wants or intends to address premise, and that others in his game have conflicting intentions.

I think the fact that roleplaying is a group process means that everyone is all over everyone else's business.  A's tendency towards Narrativist decisions will conflict with B's tendency towards Simulationist decisions.  There will be kind of a mutual inhibition going on, as one persons decisions block or invalidate anothers.  This blocking can be characterised as the source of the dysfunction.

Now, obviously intent is in there: we're talking about human beings - and we're all capable of inferring where the intent is.  But the question is whether intent should be a part of the model.

I suppose it's a question of whether we could have a rigorous definition of 'intent', enforcable on the Forge.  I don't think so without alienating a fair number of people and spawning endless threads about what intent means.  Some of the problems I'd forsee:

Can someone have GNS preferences without knowing about GNS (can you intend to do something without knowing what it is)?
Does a decision have to be conscious to have GNS weight (and how would we ever tell if a particular decision were conscious)?

I suspect there are common sense answers to these which would be completely different to other people's common sense answers.

Intent may prove to be more trouble than it's worth.

Ian

[Valamir]

Intent may prove to be more trouble than it's worth.

Ian

That's my take precisely Ian.  As Gareth said earlier, Marco's idea for using intent assumes people have an accurate ability to assess their own intent and motives.  Given the millions of dollars spent on therapy each year...I figure that assumption is pretty much a shot in the dark.  Self delusion is a human past time after all.

IME statements of intent often contain alot of regurgitation of:
a) roleplaying dogma we've been taught since we started.
b) trying to impress others with being avant garde (or trendily retro, or individually unique, etc.)
c) a decided and unrecognized disconnect between what one actually does during play and what one *thinks* one is doing during play.

In other words I find self assessments of intent and motive to be about as ambiguous and inaccurate as attempting to assess others, and therefor about as unuseable.

I do (and have) acknowledge MJs point about intent being at the root of it all but firmly believe that it is both impossible to determine accurately and completely unnecessary to know for the model to function.

Its like Mike said.  When you get pulled over for speeding, its the behavior that's being measured.  The judge doesn't really care or have any need to know why you were speeding.  People in traffic court give explainations all the time.  Sometimes the judge cuts the fine, or the points.  When he does its after reviewing the drivers past driving history and cutting them some slack for a clean record...not because he's moved by the "reason" given.  In other words, why you were speeding is almost completely irrelevant and unnecessary for the judge to rule on the behavior.

Our own play often doesn't jive with our own perception of our play.

[Ron Edwards]

Hello,

I'm rarely quoted or paraphrased correctly on this issue.

My position is, "The reader may insert or project intent into understanding the process of role-playing as he or she sees fit."

In other words, I do not disallow intent by whatever definition. If you see "intent" here there & everywhere, then so be it. However, one's interpretation, and how one applies it, should not be permitted to carry argumentative weight regarding any particular point.

Best,
Ron

[Valamir]

Sorry Ron, I've got to call foul on that one.  You've said this before and I let it slide, but if you're actually going to draw attention to this statement I'm going to have to shout "boo".

[Marco]

Ralph,

I just want to make sure we're clear on what's being said:

If there's massive dysfunction and a great deal of emotional turmoil surrounding a gaming session--then sure, I can see where one's self-analysis of intent might somehow be clouded by denial or other factors.

But ...

In those circumstances observed behavior is also similarily clouded, isn't it? I mean, denial in general is about denying obvious and extant effects of things that go on in our daily life, isn't it?

If you assume that one's introspection is too clouded to be viable (and remember, this introspection comes with GNS theory--the person analysing their behavior is aware of the three primary modes to which they might apply thier action)--then isn't one's observation equally suspect?

If you are acting as a magistrate (Mike's analogy)--impartial, un-involved in the game--then perhaps you can judge the behavior you observe--but similarly, a choice of weapons that offends no one and invokes no emotion would be similarily neutral, wouldn't it?

GNS isn't meant to only be applied if you're an observer on the side-lines--and I can't see how basing one's analysis on what you think you observed is any more accurate than basing it on what you think you intended.

The difference is that with the intent you can analyze an atomic instance of play instead of having to generalize trends. I think that's a lot of power to ignore, the baby, if you will, being thrown out with the bath-water.

Edited to add: I'm not talking about *statements* of intent by others, either. I'm talking purely about a self-assessment of why my character chose a weapon/course of action/etc. So the idea of a "statement of intent" that's designed to impress someone only makes sense in this context of "I'll take the short sword instead of the broadsword so that people will see what a cool anti-munchkin I am!"--which, you know, is worth examining--it's a non-GNS behavior that impacts play.

If there's a lot of that sort of thing going on, analyzing it is useful, IMO.

-Marco

[Ron Edwards]

Hello,

I refer people to GNS, intent, and motivations, as well as the thread from which it spawned.

I can't be any clearer. Ralph, we're disagreeing. What I wrote above is what I'm saying, and what I've said. It's the only position I take on the matter. If any phrasing led people to other conclusions, then I'll take responsibility for it, but this really is what I am saying.

I don't know where the "boo" is.

Best,
Ron

[Valamir]

Hello,
I can't be any clearer. Ralph, we're disagreeing. What I wrote above is what I'm saying, and what I've said. It's the only position I take on the matter. If any phrasing led people to other conclusions, then I'll take responsibility for it, but this really is what I am saying.

I don't know where the "boo" is.

Best,
Ron

Well, the boo is with regards to the total non statement of your comment.  Essentially your position boils down to (and please clarify if I missinterpreted it) as "go ahead and insert intent into your own understanding of GNS if it makes you feel better, but I don't and I'll dismiss out of hand any arguement or point you make in future discussions that rely on ideas of intent as lacking argumentative weight"

Or in other words "discuss till your heart's content but I'll completely ignore whatever conclusions you come to".

I don't see any other way of interpreting:

The reader may insert or project intent...However, one's interpretation, and how one applies it, should not be permitted to carry argumentative weight regarding any particular point.

That's what I cry foul on.  If you're going to make a statement that something lacks argumentative weight, then IMO it had better be backed up by something more tangible than "because I said so"; which is all your statement above profers.

The way I see it there are 4 possible positions being discussed here:

A) That intent is at the root of GNS and value can be gained by discussing it more explicitly (even if it is an ambiguous thing)...which I believe is Marco's position.

B) That intent is at the root of GNS but it is so ambiguous that little to no value can be gained by discussing it explicitly; and that the net result of intent aggregates out into peoples behavior anyway so behavior captures it sufficiently and therefor the issue of discussing intent becomes moot.  I.e. its there but not worth talking about...which is my position.

C) That intent does not have anything whatsoever at all to do with GNS, never has, never will; and behavior stands alone as seperate and distinct from intent entirely.  Which is what I used to think you were saying, although your position has grown decidedly less clear over time.

or D) This is not an issue I'm interested in discussing or prepared to discuss at this time so I'll recuse myself from the conversation.

Is there an E.  What really are you saying with regards to intent and its relation to GNS.

[Ron Edwards]

Hi Ralph,

I'm close enough to (B), with you, for government work. I suspect that all we're wrangling over is a preferred set of phrasings.

Best,
Ron

[Valamir]

Cool, I'm good with that.

[Marco]

I'd amend A to say "more explicity with regards to your own intent" (i.e. I wouldn't try to impute intent in other people with any expectation of accuracy either).

-Marco

[Gordon C. Landis]

B is also what basically what I thought Ron was saying, with the additional thought that while adding intent as an explanation is basically harmless, building or developing an expansion or clarification to the theory that relies on the insertion of intent is unwise.

I also think that if I'm allowed to add "as demonstrated by your actions" to the "your intent" part of Marco's ammendation of A, A also avoids most of the pitfalls we try to ward against in B.  The real key is (IMO) to have it be the actions and results that MATTER, not the intent.  To the extent that talking about intent helps with actions and results, talking about intent is cool.  But people can be very good at using intent to avoid talking about actions and results, so I don't think GNS, Ron and/or others are wrong to be a little concerned about an over-focus on intent.

Gordon

[M. J. Young]

If I understand it correctly, B is saying that it is difficult to actually measure intent, but that by examining data points we infer intent, which is the basis for GNS decisions but cannot be measured directly with any reliability; we therefore work from the data points, observing the pattern, and projecting the pattern forward from the implied intent to predict future behavior.

O.K., leave it to a lawyer to make it sound more complicated.

But I'm good with B, too; and if we are all agreed that GNS infers motive from data, and future data from motive, and that thus Gamism, Narrativism, and Simulationism are in some nebulous sense intended courses by players which drive play decisions, then I think we may have come to consensus on the issue (finally).

If so, I think this thread is done; and I hope it will bring an end to the disruption on the other one.

--M. J. Young