Because many people talk about models being more or less faithful to some underlying absolute reality. That suggests a notion of the model being an inferior copy of the thing modeled.

Thanks for the pointer. From Wikipedia: The Cosmological Principle is a principle invoked in cosmology that, when applied, severely restricts the large variety of possible cosmological theories. It follows from the observation of the Universe on a large scale, and states that: On large spatial scales, the Universe is homogeneous and isotropic. Or simply put, the universe is the same everywhere on a large scale.

I read that as a rule to allow only certain kinds of theories and to exclude others that don't qualify. In stating and applying such a rule we don't literally create some sort of universe that is the same everywhere, one that may or may not correspond to the genuine article.

Think of what you would do if on the first trial the robot overshot the hole. You might remeasure and tweak the system to zero in on a good putt. If you did manage to get it right on the first try, then it means you already acquired all the relevant control values earlier or by other means and knew how you had to make use of them.

You are describing a system that is sensitive and responsive to those aspects of the environment--the club, the ball, the green, and the golfer or golf-robot--that are required to sink the putt. Perhaps you, the programmer had to measure and enter the data. You become part of the loop then.

What constitutes the model here? Is it the robot itself as it sits idle? Is it the robot in the act of taking the shot? Is it the program and data residing in flash memory? Is it the program and data bytes being munched on by the CPU? Even the operating system code? And for whatever your answer is, how is it that it is a model of a golf putt (and say, not a model of something else)?

Perhaps the robot is built with a neural network or networks instead of a traditional computer system and was repetitively trained to sink putts (which amounts to the system self-adjusting weights and perhaps other variables among the neurons). There is no model of a green, a club, or a putt inherent in the system. Rather training pretty much wires it to efficiently produce particular sorts of responses to particular sorts of inputs.

Watch golfers prepare for a putt. There is a lot of observation and care. Up until the shot they may glance back and forth at the hole and at the ball. Practice has made them into a physical being, head-to-toe, that is sensitive and responsive to the characteristics of the course, the ball, the club, and their bodies.

Well, some athletes do consult with a psychologist to get themselves as ready as possible, but let's just consider the hundreds or thousands of hours repetitive practice and drills, of keeping at it until it is done well.

Yes, models can be subsets of what they model. But we can also simplify our tasks by focusing on the relevant to the task at hand. The child learning to catch a ball is taught to "keep his eye on the ball." He doesn't need to construct and consult a model of ball flight. He learns to point his eyes at the ball. When we carry out a task using Newtonian mechanics, perhaps calculating the trajectory of a ball, we attend to limited features of the environment. We make a few measurements and then proceed.

Not necessarily. Any means that guides the golfer to making better putts is fair game. The computer system has to provide an error signal of sorts that the golfer uses to adjust his putt.

I could incorporate your position into my thinking better if I understood in what ways scientific theories are models and how being in possession of a model is useful. Surely there must be clear-cut examples of models and a way to show that theories fall under the same category. Keep in mind that I fully understand why we liken the achievements of science to being in possession of models that probably are not fully faithful to what they model. Such notions are handy, but, in my opinion, can be safely set aside for philosophical discussions.

I do find it surprising that I seem to be the only one here arguing that science keeps us in fuller contact with the world than we would have without it. Others seem to be suggesting that the elaborate theories of science have an isolating or insulating effect, that a theory is a sort of barrier to some sort of absolute truth that is said to be lying beyond it.