Originally Posted By: Withering Ice
I think a lot of the problem is just the narrative. In practice, scientists have a question, then perform a couple of experiments, one of which kind of works and one of which fails entirely due to errors or incorrect equipment or sheer difficulty. Then the scientists are at an impasse and don't know how to do something, so they call a bunch of colleagues and fumble around with some more experiments that are largely irrelevant in the meantime. Some of those give interesting but unhelpful results; others plant the seeds of future work. Colleagues call back and propose an elegant experiment that the first scientists don't have the equipment to perform, but they approximate it with something similar but not quite right. At this point, the grants are running out, so they publish what they have.
That's a really messy story to tell. It's not even an acceptable paragraph! Thus, science writers explain that scientists had a question, performed experiments, and got a result. Open and shut.
—Alorael, who likes the retrospective neat and somewhat incorrect version. It's much more encouraging than real science. And the research is never abandoned due to budget cuts.
You don't get funded unless you propose something that is possible, and is worth doing. Scientist that are just fumbling around certainly aren't getting funded, or at least they shouldn't be.
But there are a lot of file-cabinet studies that never get published even though the "failed" experiments themselves are important pieces of information. I think this gives the impression that scientists just do one experiment and everything is revealed from it.