The Major Theoretical Blunders That Held Back Progress In Modern Astronomy

Here's the deal on 'scientific consensus' -- it's not always right, but it is the best guess at the time, supported by the majority of the evidence by smart people who know the subject.

You're right, and I agree that it's generally a safe bet to go with the "scientific consensus."

The issue is that a lot of people (including around here) seem to subscribe to what I'd say is a relatively naive form of logical positivism [wikipedia.org] , otherwise known as that sitcom hit "Everybody Loves Popper." I love Popper [wikipedia.org] too, but Popper's mechanisms to explain scientific progress are a little muddy. According to the naive idea of falsifiability, all scientific theories have to "falsifiable" and theoretically all open to be disproved by superior evidence at any time. The problem with this idea of science is that it doesn't specify how one actually progresses -- how do we choose our research from an infinite number of possible falsifiable statements?

The reality of scientific progress is that real science doesn't always work that way, and in fact no real philosophers of science today tend to think it does. Just to rehash the bits that happened 50 years ago, you have Kuhn's ideas [wikipedia.org] of "scientific revolutions" caused by shifts in research "paradigms," and responses by intelligent philosophers of science, such as Imre Lakatos's cool idea of "research programs [wikipedia.org] .

The point is, the real trajectory of scientific progress is "bumpy," and it needs to be. If everyone were ready to throw out every fundamental theory of science immediately when the slightest bit of new evidence comes along, we'd never be focused enough to do research on specific questions and make further progress. That's where most of science happens -- in fleshing out details of larger theories that are assumed to be true.

Anything else is more likely, not certainly, but more likely, to be wrong.

Yes -- and the times when the "scientific consensus" is actually less likely to be right can uncover some interesting elements about how science works, and can lead to some reasonable critiques. There were long stretches of time historically when the "scientific consensus" was actually "more likely to be wrong" on specific questions by a modern evaluation of the evidence, even assuming the knowledge of the day. But many of these times of disagreement pushed researchers on the other side to pursue evidence of the new theories even more strongly -- thus, arguably, leading to a stronger new scientific consensus on more firm ground once the "paradigm shift" occurred.

People tend to get very nervous when confronted with a "scientific consensus" that was proven wrong, particularly ones that hung around for decades (or, in a few cases, for centuries) even in the face of contrary evidence. But this is a necessary part of the messiness that forms the process of discovery.

It's kind of like having a debate without defining the fundamental terms under discussion. Until those are defined, meaningful debate can't happen. But in the process of debate, we sometimes might come to the conclusion that our initial definitions were inaccurate, or even that perhaps the disagreement can only be resolved by choosing new or different terms. That doesn't mean that the process of debate is necessarily flawed -- if we never started out with our initial terms, we'd never have been able to start making the kinds of distinctions that allowed progress to happen.