Been somewhat busy lately...
"I think it lacks punch."
So you think I'm copping out, not taking a stand?
"I'd like to have seen you combine this response with the 'Darwinian selection' attribute of theories.......if you base your reasoning mostly on the margin of error compared to observation, then you have no way to distinguish 'Darwinean selection' attribute of theories with approximate truth. All you can say, as far as I can tell, is that theories are keeping pace with observational sensitivity (i.e., fairly so). It doesn't tell you the latest theories are more approximately true than Ptolemy or Aristotle."
Yes it does! Absolutely it does! Modern theories explain observed phenomena not explainable through Aristotlian or Ptolemic philosophy and physics. So the theory explains what P and A tried to, and encompasses more information, and does a better job of doing so. Therefore, it is more accurate. And I do combine my view with the darwinian selection attribute of theories, but not to the same degree. I don't toss out old theories as incorrect and embrace new ones as correct. I toss out old theories as less correct and embrace new ones as more correct. The more use statements and observations a certain theory satisfactorily accomodates, the more valid it can be said to be. You can't look at this from a black-and-white perspective. You're trying to seperate theories into two types: CORRECT and INCORRECT. This is unviable. I can tell you right now that the only theories mankind is likely to see fall into the INCORRECT category, relative to the total information in the universe. So you have to see the whole thing in shades of grey. I'm just trying to find the theories that are the least grey, I think it's futile to look for pure white, and I think that's where you're misunderstanding what I'm saying. I consider the amount of observational data that a theory describes to be a good test for a theory's 'shade.'
"How can you justify that GR is an approximate truth versus more suitably selected for the current set of observations?"
I hold that any theory that accurately holds for a set of observations has a real bearing in the outside universe. That is, any theory that can put use statements together is approximately true for the universe. And the degree of accuracy (how one theory can be held to be more valid than another) is decided upon by how the theory expands to deal with new use statements that the older one couldn't deal with.
"as the 'Darwinean selection' attribute of theories indicate, you cannot say you are talking about the way reality 'is', you are only talking in terms of our observational sensitivity."
Yeah, I'm saying that the more observational data a theory can handle, the more bearing it has on how reality actually 'is.' Newton may have been less correct than Einstein, but he was right to a lesser degree. Since Einstein explained more phenomena, his theory is more valid. NOT correct, just better.
"Let's suppose that I walk in a room where people are playing poker. Upon walking around and observing closer, I see that they are playing 5 card poker, but upon closer examination I see that it isn't 5 card poker, but a game that is very similar to 5 card poker. Upon further examination I notice that it isn't a game at all, but rather it is some way that these people are making their living (like commodity trading). Upon further examination I notice that they are not making a living because I find that they are AI machines. Upon further examination by interviewing these AI machines I find that they have some sort of program that makes them appear to participate in that 'game'. Upon further examination I see that it is not a program running them, but some sort of wireless communication, etc etc."
When you saw the people playing poker, your assumption that poker and playing cards and people were involved were basically false, constructs needed to support your theory (in the same way that Newton's approximately true theory had the false connotation that acceleration past the speed of light was possible, no time dilation, etc.) However, the building blocks of the theory, that there are a certain number of participants, that they are interacting with each other through a medium involving cards in one way or another, were approximately true. Would you suggest that the postulation "there are a bunch of wireless AI machines engaging in commodity trading" has NO more bearing on the truth than "A bunch of people are playing poker"? The most recent postulation may be wrong (and probably is) but it describes much more, and deals with more observation than the first. So, it's more valid. Since Einstein deals with observable time dilation and Newton does not, I consider relativity more accurate than Newtonian physics.
"What is the difference between 'approximate truth' and our observations approximately matching with our best theoretical interpretation of the evidence?"
Nothing, I'm saying that our observations have tangible bearings in reality, no matter how small. An external stimulus is required for our observation.
"If they are the same, how do you know that your theoretical interpretation is not just a pure human invention made from the millions of potential human myths that could fool us into believing that reality is approximately like the way our theory says it is?"
Our theories are certainly human inventions. But since they (I'm arguing) have real sources in the external, objective universe, they have some claim on reality.