Thursday, January 27, 2011

Things I Know #3 Revisited

Things I Know #3 states: Models are not reality. Models are expressions of belief about reality.

Things I Know #3 was born of my ruminations on climate models and Anthropogenic Global Warming theory, but it certainly is not limited to climate science.  It applies to all kinds of Scientific models.

Ok let's just get this out of the way right now.  As I plow my way through Mr. Godwin's book, it's probably going to spawn some postings, and this is one of them.

He put it this way:
[scientists] forget that they are only using a simulated abstract model of their own invention, with no actual external referent.  In short: "one begins by abstracting from concrete existence, and ends by attributing concreteness to the abstraction." - (Wofgang Smith)
And
The properties we ascribe to our object of interest and the questions we ask about it reinforce the original metaphorical image and we miss aspects of the system that do not fit the metaphorical approximation.  - (Lewotin)
In other words, the model is incapable of showing us anything based on things we hadn't already assumed when we created the model.  The model will tell us what we told it to tell us (whether absolutely true or not at all), and it won't take into account anything we didn't tell it to tell it to take into account (whether it is relevant or not.)

That was just too good not to bookmark for future reference.

2 comments:

Severian said...

Interesting that you should note these...

I have conversations like this with my liberal buddy all the time. He busts out statements like Wolfgang Smith's, and I agree -- models reflect back their assumptions (like the old-school computer science guys say, GIGO: Garbage In, Garbage Out).

The difference is, I take this to be a useful comment about epistemology and a warning on not reading too much into stuff. He, on the other hand, takes it more in an "everything is relative and there's no such thing as truth" direction. You know the drill: "everything is just a social construction," and therefore we should immediately enact seventeen different varieties of liberty-squashing hard-conforminst leftwingery.

This, I get. How they jump from "everything is a social construction" to "'social justice' requires this that and the other thing," though, still beats my pair of jacks. (Social whatsthatnow? Didn't you just say it's all constructed....?)

I guess it just goes to show you that for every step of the liberal "thought" process that's comprehensible, there are ten more that shall ever remain mysterious.

philmon said...

Hey, "Social Justice" is just a social construct, so who is to say THEIR social construct is better than mine? Hmmmmmmm?????? ;-)

But that's pretty much what your poit was as well... I'm just reiterating it.

My retort to the "everything is relative and there is no such thing as truth" is,

"Are you sure about that?"

So that's the truth, eh? Just wanted to get that straight.

As I say in my dissertation about Things I Know #3 ... models can be useful in predicting things provided they contain enough truth and not too much untruth.