[Olorin]

Originally Posted by JrKASperov I don't agree with this. This is only an interpretation of hypotheses and theories. I for one think that given the nature of hypotheses and theories (described in my post above) we should never hold that they explain anything for the simple reason that they cannot be proven. That said, they do have one striking aspect which makes them very interesting and powerful: they provide simple, elegant and fruitful expressions of observed correlations. That is, they provide an easily understood theoretical framework in order to 'place' certain observations.

Sorry, didn't mean to imply that theories are actually true explanations, only that they are attempts to explain the why behind observations.

Originally Posted by JrKASperov In this case a LAW is a particular form of a theory, namely that which is a singular generalized expression. So: a Law does explain why something happens: there's a law connecting two observables, and that's why things happen like they do. However, what explains the law? This same question can be asked for any theory. So I don't quite get this separation of 'theory' and 'law'.

I've always been taught to distinguish the terms. Laws are for describing universal observations--any two objects attract, + and - charge attract while like charges repel. H and T are a scientists attempt at explain how or why things happen. A good H needs to account for all known observations + make predictions that can be tested and verified. For Einstein, he predicted that light could be bent by strong gravity (later verified by observing starlight that passed near the Sun during an eclipse). He also predicted that time measured by two clocks would be different if they traveled at different speeds (verified later by synchronizing atomic clocks and flying one around on an airline for awhile). Because Relativity accurately predicted new observations, it is considered a solid Theory--but that of course doesn't mean that it's "true". Later observations may contradict it.

Originally Posted by JrKASperov Also, I'd like to point out to you that "(two objects are attracted with a force that depends on their mass and distance)" is not an observed fact. The 'force' named is not observed, only it's effects: an observed object's position at a certain time.

True, all the ways of measuring force are actually a function of observing a change in position. However, I don't tend to worry about that distinction. We can measure force in a number of ways. Sure, our interpretation of what force is, and what the measurements mean is still a theory and can't be proven. But on a day-to-day basis, I don't indulge in quite that level of skepticism.

Originally Posted by JrKASperov Then again, for other people (me for example) Special Relativity (what is what we're debating right now) is something totally different from Newtonian mechanics. You seem to point to the fact that in certain areas, Newton's theory does not deliver a huge difference in accuracy compared to Relativity. However, does this mean that Newton's mechanics were only 'a bit' right? Or was it 'almost completely' wrong? Or is it 'close to the truth'? What do these terms mean anyway?

Take planetary orbits. Using Newton's equations, you can calculate the expected position and speed of any planet at any time. While Einstein showed that these equations were wrong, the only planet where the error in Newton's equations could be measured after a century was Mercury. Even today, most engineering jobs are done using Newtonian mechanics and ignoring the effects of Relativity. To the measured tolerances involved, it simply doesn't make a difference in the calculations. If an architect doesn't Relativity, I don't really care, since buildings don't move and are not subjected to different gravity fields. But a NASA engineer designing satellites better have a grasp of Relativity, or their satellites will have problems.

Originally Posted by JrKASperov You seem to be using a quite familiar logical fallacy which is the 'inference to the best explanation' argument. That is: "it must be true (close to the truth) since it works (it describes phenomena accurately)". Thus, it is intuitively quite easy to say that once Newton's mechanics describe certain observables accurately, we can infer that it's true. However, this is a logical fallacy. So in the end we are left with the question: "what does it mean for a theory to be 'close to the truth?" I for one think, once again based on the nature of theories, that we should not speak of 'true' or 'false' and certainly not in nonsensical terms like 'close to the truth'.

Theories can never be proven to be 'true'. So their utility is based solely on how well they predict things. If I build a machine to accomplish a task based on Newtonian mechanics, and it accomplishes the task, then Newton's theories were sufficient. If I'm extrapolating planetary orbits far into the future, I'll need to account for Relativity.

The reason I don't consider Galileo and Newton's mechanics to be 'false' is that if you compare the equations of motion that account for relativity, they simplify to the older equations in the limit of objects in the same reference frame, or as the change in time approaches zero. Functionally, on the Earth, it is difficult to design an experiment to even confirm the difference (flying atomic clocks on airlines only worked because atomic clocks can measure time in the nano-second range). For almost all terrestrial applications, Newtonian mechanics predicts the resulting behavior well within measurable error. Since it's also mathematically much simpler, most things are still done assuming Newtonian mechanics. The theory works, and that's all it needs to do.

So why even bother with theories? We can never be sure that they are a 'true' description of why and how the universe works. The best we can do with a theory is come up with something that is consistent with all known observations. The reason theories are really useful is that they often predict new phenomenon that haven't been looked for. Special relativity begins with a simple premise: Laws of physics are the same in any uniformly-moving reference frame and the speed of light in vacuum is the same no matter the motion of the observer or light source. These are just generalizations made from known experimental results. However, once Einstein started with these postulates, it led to the prediction of time dilation, length contraction, relative simultaneity, and of course E = mc2.

Without the theory, the attempt to explain the how and why behind observations, Einstein would not have come up with those other consequences. Perhaps we would have eventually observed things like time dilation directly, but the theory got us there faster. Doesn't mean that theory of relativity is 'true' in the philosophical sense, but it has been very useful, because it predicted new behavior accurately (although we can debate whether the discovery of nuclear power was a good or bad thing).