Not getting "it"

[Paganini]

Regarding hybrid play, I think that I see the term a little differently from the way Mike does. I think he's talking about the "little g, little s, little n" that can crop up in the way I quoted my essay about above. Whereas I see those as "fuzz blips" which aren't really worthy of a GNS name (although I could be wrong about that), and hybrid play as whole shifts among the group, 'way up there in the model, to play This Way for Now, as a supportive mode for another mode which is When We Get Back To It, in which those shifts are playing a specific facilitative role toward that second (main) one.

Incidentally (posted here to avoid thread hijack) this is one reason that so many people not from the Forge don't get, and even hate, GNS.

I so often get "All three of the GNS modes are present in all of my games, Edwards is obviously wrong, and delusional. He insists that my experiences couldn't have happened. Burn him!"

Meaning this in the nicest possible way, Ron, if you payed a little more attention to those little fuzz blips, you might have fewer people running around with the idea that a GNS mode is like a car gear that you automatically flip into when you start to game, driving along in until you hit a brick wall, or get where you're going.

[Ron Edwards]

Hiya,

Little biology humor for you.

An ecologist and a systematist walk into a room, and sticking to the ceiling are all these little rubber balls, different sizes and colors. Right then, the static electricity or whatever lets go, and they all fall to the floor and bounce around.

The ecologist says, "Wow! Look at the diversity!"

The systematist says, "Wow! 9.8 meters per second per second!"

[pause]

[pause. c'mon guys, it's funny, if you're a biologist]

Anyway, my perspective on why and how I started writing about this stuff in the first place arises from the vast confusions and difficulties and useless squandering of funds on same that I was observing throughout my time in the hobby, up through the early 1990s.

Why were people getting mad at each other? Other people play sports where they elbow each other in the eye accidentally-on-purpose and don't get that mad.

Why were people not getting laid all the time through the activity? Some of us were, so it was obviously possible, so why wasn't that hobby-standard?

Why did the products in question seem more and more ill-suited to actual play in such a way that groups could not seem to resolve their internal conflicts, indeed seemed hampered or blind to such resolutions?

When you want to look for solutions to problems like that, you turn to the common points of fracture. You become a systematist and look for the things that make the most sense for most situations most of the time. You shove the ecologist who's rhapsodizing about unique snowflakes out of the room because frankly, he's just gumming up the place, for these kinds of goals.

So that's why I took the approach I did: because the other approach ("let's all get along!") wasn't going to be helpful.

Now for your point about helping other people, or reaching more people, or making things easier for them to understand. Well, you know, this has some history. People previously came to my essays as peers - "Hey, looka what this guy wrote, let's see what he says and see what I think, let's talk about the way he plays and the way I play." It was a point of discussion among people who'd already decided to have a nice learning conversation together.

Now, it's a little different. People sometimes come to THE FORGE where THAT GUY is SAYING THIS THING, and they're all primed to have an opinion, based mainly on self-image. If they want to be seen as free-thinkers, they'll be argumentative with sentences like "The sky is blue." If they want to be seen as sensitive and open-minded they'll get bent out of shape about "labelling people." If they want to be seen as important in the industry or in fandom-culture, they'll be flip and dismissive. And so on.

(This is not to say that cool peers who want to have learning conversations are absent. Far from it, in fact.)

So what's my responsibility regarding my ideas, for everyone else? The answer is, nothing at all. I don't have to make them easier to understand. I don't have to make people feel better so that more people can get it. I really have no sense of evangelism about the whole thing. I just like learning-based conversations: discourse, clarity of thought, and communication among humans. I keep at it because I like what happens.

All that said - am I working on a way to present the material in a more friendly way? Yes. Did I get badly sidetracked from this for two years, writing the three GNS-specific essays, because I needed to clarify my views for the insider crowd? Yes. Have I given thought - a great deal of it - to what should and shouldn't be said to new readers? Yes.

And I've come to the conclusion I came to in the beginning: if a person isn't arriving with the baseline desire to have a learning-conversation, there's nothing I can say, nothing I can provide, no particular nuance of any particular point, and no tone of prose that will help that person understand a damn thing.

Best,
Ron

[clehrich]

[Ron and Paganini: if this is thread-hijack, split it.  I'm not sure it is, but I'm not sure where the thread goes from here anyway....]

Reading Ron's response here, I think I had a bit of a lightbulb moment.  No, I didn't find the joke uproarious...  but it occurs to me that the way Ron describes GNS is fundamentally scientific in a specific sense.  This may, I think, have something to do with how and whether people "get it," setting aside those who (as Ron rightly carps) have no intention of getting it and just want to throw rocks.

As you must know by now, especially after that big article I posted, I'm an historian of religions.  We sometimes like to think of ourselves as scientists, but I'll happily add the word "soft" before "science" in my own work at least.  Ron, however, is definitely in a hard science -- in both senses.

I've always been confused by GNS because it doesn't seem to fit my sense of a theoretical model.  Seems to me that GNS is diagnostic, in the sense that it describes a situation categorically (in the literal sense), but then it goes on and pushes for specific modifications and alterations of that situation.  For me, those are two different things; this is what I was on about with dividing analysis from synthesis.

Suddenly, it occurs to me that this is a hard-science approach.  You want to describe biological systems, to be sure, but you do this (1) experimentally, and (2) quite possibly to create new things.  For example, you may want to build new enzymes, or develop medicines, or whatever.  These things require you to understand how existing systems really work, but they also require you to formulate some new system or object.

In my end of the Ivory Tower, we don't usually want to do this second thing, and for the first we have to look at "experiments" in progress, i.e. societies that actually exist or have existed.  Sometimes people push for something like experimental construction, but there it's a specifically and deliberately ideological claim: you want to describe society because you want to push for political change, for example.  But many of us are leery of this sort of overt ideological interpretation, because we think it usually skews the readings and besides lends itself to the sort of unpleasantness that early anthropologists got into: let's study the natives to fix their lives by making them like us.

So for me, GNS has a dual purpose, and is internally inconsistent on that basis.  For Ron, I think, GNS's dual purpose is precisely what validates it: if you can only describe enzymes, but can't do anything with them, what's the point?

As long as we're musing on the "state of the field" for GNS, I wonder:

1. Ron, have I remotely accurately described your perspective?
2. Does GNS's dual purpose make it unsuited to some questions and approaches?  Which ones?
3. Is it the "experimental" end of GNS that gets taken as "that guy tells us to do X"?
4. Is a hard-science approach suitable to the object?  That is, for me it's obvious that a social-science or humanities approach best suits a social behavior, but that's because I don't particularly want to advocate change.  Might there be other basic ways of going about things?  What would they be for, and would they be worth doing?
5. If there are wildly different approaches worth taking, are they worth correlating to GNS?  Is it best to leave GNS cohesive on its own slightly narrower grounds, or expand it -- potentially breaking it or rendering it unrecognizable?

Just some thoughts.

Chris Lehrich

[Bankuei]

Hi folks,

The key point that loses people on that particular issue, is mistaking prioritizing for exclusive, 24/7 all in one mode play.  Look at it like you would a diet plan...  Mostly vegetarian doesn't mean you never eat meat, just that you choose to prioritize non-meat products.  Gamism means you choose Gamism over the other two, although they may appear in your play "diet", just as flavoring, not the main course.  Likewise with Nar and Sim play.

Chris

[Ron Edwards]

Hi there,

Chris (clehrich), here are some thoughts of mine about the points you're raised.

I've always been confused by GNS because it doesn't seem to fit my sense of a theoretical model. Seems to me that GNS is diagnostic, in the sense that it describes a situation categorically (in the literal sense), but then it goes on and pushes for specific modifications and alterations of that situation. For me, those are two different things; this is what I was on about with dividing analysis from synthesis.

Suddenly, it occurs to me that this is a hard-science approach. You want to describe biological systems, to be sure, but you do this (1) experimentally, and (2) quite possibly to create new things. For example, you may want to build new enzymes, or develop medicines, or whatever. These things require you to understand how existing systems really work, but they also require you to formulate some new system or object.
...

So for me, GNS has a dual purpose, and is internally inconsistent on that basis. For Ron, I think, GNS's dual purpose is precisely what validates it: if you can only describe enzymes, but can't do anything with them, what's the point?

Actually, I like to think of hard sciences as divided up into Basic, Applied, and Engineering. Basic = study it 'cause it doesn't make sense as we currently use or "know" it; Applied = study it because it must be relevant to some widely-recognized problem; Engineering = develop a specific, novel, and usable item.

As it happens, professionally speaking, I'm awfully tightly ensconced in Basic research (although I avoid the biased self-description of "pure" research, which some of my colleagues use). But that's probably not very relevant here at the Forge. What is relevant, I think, are these observations:

- Basic conclusions are often very very useful as a foundation for addressing Applied questions, and Applied conclusions are often very very useful for addressing Engineering questions. The transition is not at all transparent or predictable, and usually takes a long time - and the people involved are usually not collaborators or even have any sort of contact at all.

- Sometimes the flow goes the other way, in that an Engineering-type invention or whatever prompts a whole door being flung open at the (say) Basic level.

All my written work regarding role-playing tends to scoot up and down the scale in both directions, which is very different from my (or anyone's) professional work as scientists. I also think that the flow is more equally two-way for this topic. All of which means: (a) GNS Discussion and RPG Theory are a lot like the interface between Basic and Applied, and (b) Indie Design and Actual Play are a lot like the interface between Applied and Engineering, and (c) Publishing is a lot like Engineering by itself.

1. Ron, have I remotely accurately described your perspective?

Not really, I'm afraid. I think you were kind of focused on Basic vs. Engineering, and identifying them with soft vs. hard sciences, which I don't think is correct. My claim is that science has all three levels in it. Here's my real answer to your question, then.

I don't think the soft/hard distinction matters much. I distinguish instead between a unified vs. a dis-unified field of study. Biology, chemistry, and physics are unified not because everything in them is rock-solid sure (far from it!) or because they have such notable Engineering outcomes (e.g. technology), but rather because conclusions in them are always constrained to be rigorous - to make sense in terms of any other conclusions out there (or to acknowledge that you're challenging those other conclusions). Disciplines like psychology and anthropology have instead a tendency to "live and let live" (if this phrase may be applied to such extreme contention) - meaning that if conclusion A and conclusion B are incompatible, that's philosophically OK, because they can sort of "hyper-co-exist" in the shared space of the discipline without one or the other being modified or going belly-up eventually.

I do think I've brought my preference for the unified expectation into the Forge, and made it a cornerstone of the discourse here. Again, it's not a matter of everyone agreeing, but rather everyone committing to see where the points of agreement are or are not. And my hope is that it can be found up and down the entire scale of Basic/Applied/Engineering, in terms of role-playing as an activity and as a publication/sales item.

2. Does GNS's dual purpose make it unsuited to some questions and approaches? Which ones?

Too abstract for me, I'm afraid.

3. Is it the "experimental" end of GNS that gets taken as "that guy tells us to do X"?

Yeah, that's a problem. Again and again, some point I'm making at the Basic level gets inappropriately seized upon as a direct and specific recommendation (or even proclamation) about the Engineering level.

4. Is a hard-science approach suitable to the object? That is, for me it's obvious that a social-science or humanities approach best suits a social behavior, but that's because I don't particularly want to advocate change. Might there be other basic ways of going about things? What would they be for, and would they be worth doing?

Well, that's tricky. I don't think of what I'm doing at the Forge/etc to be hard science, in any way. My preference for unified discourse is about the only thing I can think of that translates over.

And I'd go a little farther and state that I consider all disciplines of study to be suitable to any topic - that topics themselves are not disciplines. Given that each discipline is sufficiently strong to yield fruit, I should be able to talk about anything in biological, social, artistic, economic, physical, political, chemical, etc, etc terms. I consider disciplines to be angles of attack, not limitations of topics.

5. If there are wildly different approaches worth taking, are they worth correlating to GNS? Is it best to leave GNS cohesive on its own slightly narrower grounds, or expand it -- potentially breaking it or rendering it unrecognizable?

That one's too abstract for me as well. I think first of all I should ask whether you're talking about my Big Model or about GNS (Creative Agenda) per se.

Best,
Ron

[clehrich]

- Basic conclusions are often very very useful as a foundation for addressing Applied questions, and Applied conclusions are often very very useful for addressing Engineering questions. The transition is not at all transparent or predictable, and usually takes a long time - and the people involved are usually not collaborators or even have any sort of contact at all.

Fair enough, and similarly for your comments about Basic, Applied, and Engineering.  I guess it's really the experimental nature of "hard" science that I'm referring to: you can construct an experiment to test theories, but I can't because the data has to be "out there" already.  That is, I can't invent a new religion to test my theories (though that would sort of be interesting in a weird way...).

All of which means: (a) GNS Discussion and RPG Theory are a lot like the interface between Basic and Applied, and (b) Indie Design and Actual Play are a lot like the interface between Applied and Engineering, and (c) Publishing is a lot like Engineering by itself.

Gotcha.  Same page, actually, but you put it better.

Disciplines like psychology and anthropology have instead a tendency to "live and let live" (if this phrase may be applied to such extreme contention) - meaning that if conclusion A and conclusion B are incompatible, that's philosophically OK, because they can sort of "hyper-co-exist" in the shared space of the discipline without one or the other being modified or going belly-up eventually.

I don't quite agree, but that's really irrelevant here.  I'll get back to it in a minute.

3. Is it the "experimental" end of GNS that gets taken as "that guy tells us to do X"?

Yeah, that's a problem. Again and again, some point I'm making at the Basic level gets inappropriately seized upon as a direct and specific recommendation (or even proclamation) about the Engineering level.

Again, we're on the same page.  I do wonder, though, whether distinguishing among the three types more rigidly might help.  I doubt it, actually, because I suspect most folks who transform a description into a prescription and then denounce it on that basis have no interest in "getting it" (to get back to the thread's title).

I'd go a little farther and state that I consider all disciplines of study to be suitable to any topic - that topics themselves are not disciplines. Given that each discipline is sufficiently strong to yield fruit, I should be able to talk about anything in biological, social, artistic, economic, physical, political, chemical, etc, etc terms. I consider disciplines to be angles of attack, not limitations of topics.

Let me clarify this one, actually.  I agree that a discipline in this sense can be taken as an "angle of attack"; what I'm wondering is whether the angle of attack you've used in GNS/Big Model (see below) is capable of dealing with all questions.  I suspect your answer would be "no, of course not, and that's not inherently a weakness," with which I would agree.  I wonder what the limitations of such an approach are, however, and what questions it necessarily leaves aside.

5. If there are wildly different approaches worth taking, are they worth correlating to GNS? Is it best to leave GNS cohesive on its own slightly narrower grounds, or expand it -- potentially breaking it or rendering it unrecognizable?

I think first of all I should ask whether you're talking about my Big Model or about GNS (Creative Agenda) per se.

Quite right.  Big Model.  Sorry.

I realize it's a very abstract, open-ended sort of question.  My point is that I think the Big Model represents a sophisticated and coherent descriptive result of a particular type of approach, one I associate with the "hard" sciences.  I see this as a good thing -- it's not a criticism.  But that sort of approach and model cannot answer every question, or describe every phenomenon; indeed, it must constrain itself to avoid certain types of question as unanswerable from this sort of perspective.  I wonder if anyone has thoughts about what questions are thereby rendered out-of-court.  Further, I wonder whether it's reasonable to try to bend the Big Model to cover currently excluded question-types, or whether other approaches should run in parallel.

See, this is where your above remarks about "hyper co-existence" really come to the fore.  My inclination, since I come out of the very soft sciences, is to let the Big Model deal with what it deals with, then formulate quite different -- and contradictory -- approaches to deal with excluded issues.  I think that this would generally be unpopular among hard scientists.

I'm also thinking of some of the discussion of synthesis/analysis in my Ritual essay.  Many seemed to feel that such a distinction ultimately invalidates the analysis, and what I contend here is that this claim of invalidation arises from a scientific perspective.

Anyway, this is just abstract musing.  If I'm the only one whom this bothers, I'm not going to push it.

Chris Lehrich

[Rob Carriere]

I'm also thinking of some of the discussion of synthesis/analysis in my Ritual essay.  Many seemed to feel that such a distinction ultimately invalidates the analysis, and what I contend here is that this claim of invalidation arises from a scientific perspective.

If I may chime in as someone from the engineering end of the spectrum, I think there are two issues running here. The first is that in the hard sciences theories are required to be capable of predictions (as opposed to being purely descriptive). Because there's this wonderful field called Optimization Theory, any analytical theory capable of prediction is automatically rendered capable of synthesis, will ye or nill ye. I think that may be behind a lot of the sentiment of the form `non-synthetic means poor analysis'. The people saying that are thinking `non-synthetic means non-predictive' and `non-predictive is not a theory'. The former is correct, the latter depends on a specific jargon meaning of theory. That's indeed different people's perspectives getting in the way of clear communication.

The second is the pet peeve of anyone who's worked in optimization for too long: words like equal, better, best are meaningless unless you specify the criterion of evaluation. This means that there is a wide chasm between a synthesis-capable theory and a prescription; the name of that chasm is `purpose'. A synthesis-capable theory of ritual wouldn't tell anyone how to improve their ritual. In order to do that, you would have a to add a description of what you consider to be improvements.

I agree completely that telling the Roman Catholics how to optimize Mass would be presumptuous and out of place for a scientist. But a synthesis-capable theory of ritual wouldn't do that. It wouldn't do anything at all until you added the evaluation criterion.

SR
--

[clehrich]

If I may chime in as someone from the engineering end of the spectrum, I think there are two issues running here. The first is that in the hard sciences theories are required to be capable of predictions (as opposed to being purely descriptive). Because there's this wonderful field called Optimization Theory, any analytical theory capable of prediction is automatically rendered capable of synthesis, will ye or nill ye.

Can you clarify what you mean by prediction here?  I find myself going around in circles trying to work out this distinction between capable-of-synthesis and not-so-capable.

For example:

A synthesis-capable theory of ritual wouldn't tell anyone how to improve their ritual. In order to do that, you would have a to add a description of what you consider to be improvements.  ...  I agree completely that telling the Roman Catholics how to optimize Mass would be presumptuous and out of place for a scientist. But a synthesis-capable theory of ritual wouldn't do that. It wouldn't do anything at all until you added the evaluation criterion.

How would you do synthesis without such criteria?

With things like ritual, prediction itself is tricky, because you can't usually set up experimental situtations to test such predictions.  The data, like the Truth, is Out There.

I do think that the distinction you draw between predictive and descriptive covers most of what I meant in terms of analysis and synthesis, since it seems as though a predictive theory, necessarily capable of synthesis, tends to get validated precisely in the predictive/synthetic application.  

In the very soft sciences, we tend to think that a theory is good when it seems to make sense of a large expanse of data without required loss; that is, you don't have to throw away part of the data in order to make it fit the model.  In the end, such theorizing usually comes down to explaining the intangible, such as what large groups of people are thinking below the surface, in a way not intelligible to themselves.

Chris Lehrich

[Rob Carriere]

Can you clarify what you mean by prediction here?  I find myself going around in circles trying to work out this distinction between capable-of-synthesis and not-so-capable.

Oops. didn't mean to send you into a tail-spin. Let me see if can express myslef more clearly.

A prediction would be a statement of the form, `if X occurs, then Y must result', or as us engineers tend to put it, `if you do X, Y will happen'.

A prescription would be a statement of the form, `if you want Y to happen, you must do X.'

Thus, a prescription is a prediction turned backwards. That means that if your theory can predict, then you can find prescription statements by searching along all possible X until you find one where the theory predicts Y. (Possibly searching more cleverly than by exhautive search, but that doesn't really matter for the argument.)

Note the above inversion procedure is predicated on you having a Y to go looking for. That Y is not part of the theory, it arises from your desires/goals.

How would you do synthesis without such criteria?

You wouldn't. That's what I tried to express by using the phrase `synthesis-capable'. It's a little like instant soup: add a criterion and the theory will synthesize.

With things like ritual, prediction itself is tricky, because you can't usually set up experimental situtations to test such predictions.  The data, like the Truth, is Out There.

Of course. Us hard scientist and engineers picked the easy fields of study, the ones where you can experiment. I would guess, without the hindrance of actual experience :-), that the only type of experiment open to you would be subdividing the data, builiding the theory off the one half and then testing to see whether it can predict the other half. (Which incidently is how neural nets, expert systems and the like tend to get trained; they often suffer from a similar problem of no experiments allowed).

In the very soft sciences, we tend to think that a theory is good when it seems to make sense of a large expanse of data without required loss[...]

I'd never seen anybody describe it like this. This is conceptually actually quite similar to the work I did in my dissertation days, signal modeling. How much model do you need to `explain' the signal adequately. Of course, being engineers, we were phrasing this in terms of the predictive power of the model, but that's actually just the mental crutch we were most familiar with.

I seem to be rambling, I hope I've managed to be slightly less obtuse?
SR
--

[clehrich]

Thus, a prescription is a prediction turned backwards. That means that if your theory can predict, then you can find prescription statements by searching along all possible X until you find one where the theory predicts Y.

I get this.  Interesting way to look at it!

I would guess, without the hindrance of actual experience :-), that the only type of experiment open to you would be subdividing the data, builiding the theory off the one half and then testing to see whether it can predict the other half. (Which incidently is how neural nets, expert systems and the like tend to get trained; they often suffer from a similar problem of no experiments allowed).

Yes, quite true.  In fact, this has led to lots of problems in its own right, because the choice of subdivision gets subtly aligned to the desired predictions.  For example, Emile Durkheim subdivided the data so as to privilege the "simplest religions," arguing that what was additionally present in other religions must be an accretion above and beyond the original simplicity.  Setting aside the fact that the chosen "simplest religion" wasn't anything of the kind, you have to wonder how Durkheim decided which religions were simplest.  Well, in essence he had already decided which elements of religion were the core, and thus he chose to privilege religions for which those elements were apparently dominant.  At the conclusion of the study, he could turn around and say that he'd demonstrated his theory about core elements.  But he hadn't: he'd demonstrated that he'd started with an assumption.

The same could be said of a really vast range of theorists, actually, and is arguably not something anyone can ever entirely get out of, but whatever.

Anyway, I never thought of this in terms of data-subdivision as experimentation.  Verrrry interesting.

In the very soft sciences, we tend to think that a theory is good when it seems to make sense of a large expanse of data without required loss[...]

I'd never seen anybody describe it like this. This is conceptually actually quite similar to the work I did in my dissertation days, signal modeling. How much model do you need to `explain' the signal adequately. Of course, being engineers, we were phrasing this in terms of the predictive power of the model, but that's actually just the mental crutch we were most familiar with.

Yes, well, this is I think probably the normal way of thinking of it in the humanities and so forth.  For example, Durkheim's approach described above is already suspect because he has to discard 99% of the data: he started by throwing away almost everything, claiming that it was late accretion.  If you have a theory that covers 90% of known cases, including all the little nibbly bits, you have a really good theory (for us, anyway).  I don't know of such a theory for ritual, which is why we always fight about it, but then again we haven't even decided which data we ought to be looking at in the first place (which would require a definition of ritual, and so on in circles).

I seem to be rambling, I hope I've managed to be slightly less obtuse?

Well, I don't know if anyone else is much interested, but I have to say I'm finding this fascinating.  I may be being obtuse, but you certainly aren't.  Thanks!

Chris Lehrich

[Rob Carriere]

Well, I don't know if anyone else is much interested, but I have to say I'm finding this fascinating.

Second that motion. I've always been fascinated and exhilirated by what happens when different fields of study and their different perspectives get together, so I could go on for quite a while, but this should probably be turned back into a discussion that actually touches on getting GNS?

SR
--