A model (mathematical) defines a system as static. It may not be static, in fact it never is is it ? You are in a way trying to take the non-linearities out of the thing, make em linear, or close to it. Sooo... then you can produce proper usable matrices, inputs to outputs, to you know, get something out of it in an analysable sense. Seen ?
Models have regions of legitimacy. As in, within these regions the non-linear nasties have limited effect. Outside of the regions these annoying bastards become too much 'woah..woah.. wtf is it doing.. help me.. help me' for your model to be in anyway valid, or taken seriously by grown ups.
You can, of course use seperate models for different regions and you could build a nice consistent plane to link em together, ideally. BUT, you do need to be able to measure (er,, create sometimes) a region parameter.
now, what we're dealing with in this case is a model that is definitely non-deterministic, non-linear, recursive and adaptive like a bastard. In fact the regions themselves, in terms of their scope (in a legitimate sense) is actually part, if not all, of the outcome variables of the model. So.. its shifting, always, constantly, like some twisted fucked up manifold that goes into spasms everytime you cough near one of its inputs.
Edit : this extrapolation/identification of the problem is based entirely on my own nightmares modelling complex systems. I personally havent actually dealt with the problems of Quantum Mechanics I'm assuming the problems they have are the same. So yeah, I could be horribly incorrect.
1. In science, you build models as per the findings of experiments.
2. The models are helpful in predicting things, but only within certain bounds.
3. Once outside those bounds, things can vary wildly, the model doesn't work so well, if at all.
4. With complex systems, more than one model can be usefully employed, even if the various models are not compatible with each other.
5. The system we are discussing is at the far reaches of our modelling abilities, so the predicting aspect of the models is much reduced, rendering them relatively useless.
6. Thus "religion" or the "religious experience" or whatever we want to call it is, at present, outside the realms of science's capabilities to analyse in the accepted fashion.
Am I close?
