There was also a short essay with the above title once, but got lost.
Below is an email excerpt that prompted it, and the various drafts of
Models and Change -
|
|
|
6/29 |
09 |
Vague Def / Stan |
|
0.8 |
Correspondence of model & reality become vague rather than assured couldn't make models with vaguely defined parts is one of the problems. I'm pointing out that sometimes you can predict when models will become undefined as nature evolves new form in the system being modeled. That situation is confronted if the model correctly implies that the physical processes being modeled will become unstable at some point. You get that with any set of assumptions that results in an unbounded exponential or a Zeno's paradox problem with systems implied to have unbounded divisibility. |
|
|
|
6/29 |
09 |
Vague Def / Stan |
|
0.8
|
The gist was: Take the question confronted when a model implies a process having tipping points that will destabilize the physical process represented, and then make the relation between the natural system or mechanism and the model become undefined. As the physical system approaches that threshold the = signs in the equations become ? marks instead, and your abstract representation of the system actually becomes suddenly very vague. The equations then are not equations any more, having become undefined. Nature herself doesn't have that problem perhaps, but our way or representing nature definitely does. Becoming undefined is generally bad for a model, right? The question for the modeler is "Were you ready for that?", and "Had you pre-studied the new formulas that nature might adopt after that?" Stan - Dammit! When are you going to get this stuff published!! It would be useful to have some warning of that, and a chance to think vaguely about the range of possibilities approaching instead of having to wait for enough mishaps to accumulate, right? Nature herself might actually seem to make system transformations and build new systems by constructively linking complementary differences between things that no one would ever think of until seeing them. This could not be anticipated, hence not modelable. For us knowing when to start preparing and looking out for such unpredictable change at least makes the model maker the first rather than the last to know
|