-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
It's a good ramble. But to be clear, modeling and simulation (M&S) is not about, nor has it ever been about, finding a "better model". Perhaps that's arrogant on my part. How about M&S has never been about finding a better _anything_ for me. M&S is about thinking. Models[*] are instances of extended physiology, much like a spittle bug's foam. Models are extensions of our brains. I can go further and assert that general purpose computers[+] are _not_ about improving things or finding better things, either. The basic idea is "the bigger and more complex your brain is, the more you think". And, to an extent, it's true. What you've identified, the premature fixation on a particular model, was inherited from the already-present structure of our brains. I've heard people assert that it was advantageous for a hunter-gatherer to be able to react quickly without/before engaging the higher structures of the brain. I don't really buy that; but I do buy the concept that we have evolved (for whatever reason/explanation) to prefer our own perspective over others'. Fixation on _a_ model or a _better_ model is just an artifact of that. M&S, as a discipline, is about _avoiding_ such a fixation, fleshing out the mechanisms and explanations that satisfy some set of conditions and helping you break out of the limitations of your own perspective ... discovering alternative "truths" you haven't yet thought of. This is the M&S elevator pitch. [*] I'm not including "mental models", here, because they are different beasts. I'm talking about externally identifiable artifacts we typically call "models". [+] It's more difficult to make that assertion about embedded systems devices because at least 1 particular model must be inscribed into the device. So, one can successfully argue that they are more like, closer to, sensors or motors. Steve Smith wrote circa 11-05-04 10:40 AM: > Glen, et alii - > > I have listened to the chatter on this mail list for years now and with > only a few exceptions, we don't seem to put our models where our mouth > is. We are all "here" ostensibly because we know something about > Complexity and Modeling, or at least that we are Complexity Groupies of > some sort... and yet... but still... > > I do know (of) Nick's, Eric's, Owen's, Stephen's, and Shawn Barr's > collaboration on the MOTH (Myway Or The Highway) model of the prisoners > dilemma: > http://ccl.northwestern.edu/netlogo/models/community/LogoMoth > > but not much otherwise (I'm sure there are more, I just don't know of > them!). > > I too share the feeling that complex systems studies/modeling might help > in the problems of governing. Doug, I know, has had a good taste of > building models and running studies for Government Agencies and > Policy/Decision Makers who are known to directly disregard highly > refined results and make decisions on (apparently) an entirely political > basis. > > I still do work on decision support systems (mostly for Gov't customers) > and still hold the opinion that our job is not to help decision makers > make more informed decisions, but rather help prevent them from making > uninformed decisions. This may sound like splitting hairs, but my > experience is that decision makers often have already made their > decisions and decision support systems (including simple human analysis) > are often used to justify the decisions already chosen. By building > well motivated models with a clean stream of source data, I believe that > at best, our systems occasionally prevent an overzealous decision (or > policy) maker from making a rash, uninformed decision based on personal, > political, or in some cases whimsical opinions. Sometimes they might > pull back because they are enlightened by our enhanced > information/analysis, but more likely because they know that others will > be enlightened by it enough to question their decisions. > > While I think I understand Glen's point about delegation... there are > reasons for a representative government... I think that making *more > explicit* a shared hierarchical model of the things we are making > decisions about, "we the people" can make better decisions about our > popular (or electoral) support for issues and candidates. While I think > the professional decision maker deserves better tools, better support, > such may only be adopted if the unwashed masses have access to the same > information/tools... > > The news stream and more aptly, the opinion/analysis cloud that trails > it, is an attempt to do this. One might think that we could make this > more formal? Perhaps Tom Johnson and the School of Analytical > Journalism might be able to steer us to something in that regard? > > In an ideal world, I would like to think that with enough transparency, > enough broadly disseminated information, enough understanding at many > levels, the answers (or decisions) would flow easily from the > questions. I believe this to be the main way things happen science... > the question is refined until the answer becomes obvious. > > Unfortunately, I think there is something more going on, and it is "the > stories we tell" which I believe formally is "the models we choose". > My friend and colleague, David Thompson who some of you might know from > his time at Bios Group, has a blog around his new work in "Story > Resolution". In particular, this entry on "Affording to Know" seems > fairly apt for the current discussion: > http://storyresolution.org/2011/03/the-value-of-affording-to-know/ > > What leaps out at me from David's postings and our private discussions > is that we can fit many models to the same data, or we can tell many > different stories based on the same observed events. Multiple observers > can report the very same events but tell radically different stories > about them. > > This sobers my thoughts that if we could just "write a better model" we > could understand the world more clearly, make better decisions, etc. > But then, it is not uncommon for us to challenge eachother with "tell a > better story" when we are using the facts to make ourselves miserable or > abusive.... The important question (as always) might be "what do we > mean by better?". > > At the risk of sounding newage (rhymes with "sewage"?), I do think the > models we have of the world, the stories we tell, are critical in > defining the world we are generating. "Visualize Whirled Peas" comes to > mind. The pragmatists (or pessimists, or warmongers or hawks or ...) > might say that we have ample evidence that humans are always in violent > conflict and that any semblance of peace is an illusion or at best > fleeting, etc. The Optimists (or Polyannas, or Candides or > LaLaLanders) might be accused of ignoring important facts when choosing > to view the world through rose colored glasses. But we do know that to > some extent we find what we look for, we optimize what we measure, our > big hammers make everything look like nails, etc. What about that, really? > > This seems to be a meta-modeling question, and one I suspect many of us > have answered (partially) in relatively limited but more technical > domains. Is there broader application? Is it actionable? > > Sorry for the ramble... I suspect only a dedicated few could stick with > me on this one, but if I make the effort to go back and tighten it up, I > will probably just delete it. - -- glen e. p. ropella, 971-222-9095, http://tempusdictum.com -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iD8DBQFNwZqMpVJZMHoGoM8RAgTWAKCItxldVUhaKkQhgMs1aCCLrMKYcwCfcq2i VjS/VFuvGX5tj7InRALXPJk= =Ogfc -----END PGP SIGNATURE----- ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
