[lutusp]

> Many philosophers of science would argue that science is the search for truth ...

Yes, but because the stakes are so high, and because of the critical role of science in modern times, the views of philosophers aren't compelling on science's definition any more.

As just one example, society needs to keep superstition out of science classrooms. To do this, and with the aid of expert witnesses, a court of law has recently defined science this way:

Source: McLean v. Arkansas Board of Education (http://www.talkorigins.org/faqs/mclean-v-arkansas.html)

Summary -- science must have these properties:

1. It is guided by natural law;

2. It has to be explanatory by reference to natural law;

3. It is testable against the empirical world;

4. Its conclusions are tentative, i.e. are not necessarily the final word; and

5. It is falsifiable.

The above is why Creationism isn't presented alongside evolution in Arkansas science classrooms. If you think that's important, then it follows that science should be defined clearly enough to avoid that outcome. BTW the above is only one of many nearly identical legal definitions of science.

Based on the foregoing, a falsifiable idea is not a truth, it is merely an idea that has survived falsification so far.

> My own personal take is that science is a method (and framework of thought) for making statistically accurate predictions about the world.

There's a bit more to it than that. There is the responsibility to test one's ideas against reality, and abandon those ideas that fail the comparison. There is the responsibility to candidly consider alternative explanations for the same observation. There is the responsibility to design experiments sufficiently rigorous that independent, similarly equipped observers are drawn to the same conclusion. And many more.

> Somewhat paradoxically then, "truth" becomes approximations of correct, rather than the traditional inviolable assertion.

For philosophers arguing over tea, no problem. For those who understand the deadly serious role of science in modern times, no way.

[aharrison]

That is an excellent comment and I am particularly amused at the legal definition of science, which I had not heard of before.

Keep in mind that my definition is a simplification of the framework. Everything you said is necessary for generating the most accurate model, but - and correct me if I am wrong here - they are all there to support that final goal.

Note also that this is not how I would describe science to a layperson, or at least not without spending another hour or so going into greater detail of all of the necessary side effects and dangerous pitfalls, but like you are presenting.

[lutusp]

> Everything you said is necessary for generating the most accurate model, but - and correct me if I am wrong here - they are all there to support that final goal.

I don't want to sound like I'm arguing solely to argue, but if your thesis was correct, the Ptolemaic system could have been tuned repeatedly for greater and greater agreement with reality, but without ever reconsidering the basis of the model, the underlying theory.

History records that we instead abandoned the Ptolematic theory, even though it "worked" and would have been perfectly satisfactory by way of refinements, to support modern astrodynamics. But instead we stopped generating more and more accurate (and complex) models and replaced the entire structure.

It was the same with the ether theory -- after difficulties with the theory arose in the late 19th century, instead of "saving the ether" (an actual expression used at the time), we abandoned the entire structure -- this time 20 years in advance of any plausible replacement, which turned out to be relativity. And relativity turned out to be a much better theory in all respects, even though that wasn't obvious at the time.

This is getting into a complex area, one having to do with the law of parsimony (a.k.a Occam's razor) and theoretical elegance, but science isn't about the most accurate model, it is also about beauty and parsimony alongside accurate modeling.

[Locke1689]

This is 99% correct, but you're wrong about the last sentence.

Actually, science is about the most accurate model, but it turns out that what we mean by "beauty," is actually about simplicity, and thus about correctness.

There is a sound mathematical background behind simplicity and its likelihood of being correct, as outlined in Scott Aaranson's paper:

http://arxiv.org/abs/1108.1791

[lutusp]

> ... it turns out that what we mean by "beauty," is actually about simplicity. and thus about correctness.

Not necessarily. The most beautiful theories aren't always the simplest. I'm only objecting to your use of "actually" above as though it were a firm relationship, rather than a desirable outcome that nature sometimes favors.

[nacker]

Great stuff, as usual, lutusp. The paradigmatic example is Kepler, who labored for years on models of the solar system based on the Platonic solids and "perfect" circular orbits.

The reason he is remembered as one of the founders of modern science is that he abandoned the aesthetic criterion in favor of the messier, uglier, facts.

[lutusp]

Good example! It was implicit in my argument, but it deserved to be stated explicitly, so thank you.

[nessus42]

> Based on the foregoing, a falsifiable idea is not a truth, it is merely an idea that has survived falsification so far.

You seem to be unaware of the incontrovertible fact of logic that a proposition can be both true and falsifiable.

Also, you strangely ignore the fact that science aims to produce models that make reliably true predictions. E.g., "If I drop this pencil it will fall to the ground", or "The sun will rise tomorrow morning at 6:27am", or "Information shall never travel from point A to point B faster than a quantum of light."

True predictions are "truths", and science is chock full of them. Furthermore, science's aim is to produce more and more such truths with every passing day.