The Predictable Distortion of Global
Basic to any understanding of the dynamics of science is the concept of the paradigm, elucidated by Thomas Kuhn in a number of academic papers in the 1950s, and more comprehensively in the oft-quoted The Structure of Scientific Revolutions, first published in 1962 and a steady seller since then. (Indeed, more than 40 years later, it is still in print.)
But Kuhn's brilliant analysis of scientific behavior was developed at the same time that a remarkable shift took place in the public support of science. Thanks to its success, and thanks to the scientific bureaucracy that it engendered, the Manhattan Project, which coordinated American resources in a crash program to develop the atomic bomb, laid the groundwork for massive federalization of scientific research. Kuhn's work appeared before the impact of that federalization on his thesis could be assessed. Indeed, it makes Kuhn's view even stronger.
Kuhn defines science paradigms as bodies of knowledge that “define the legitimate problems and methods of a research field for succeeding generations of practitioners.” Paradigms are fractal. They have the same dynamics whether they refer to an expansive notion, such as relativity, or a smaller one, such as the notion that we have the quantitative skills to confidently predict the future climate of the earth from first physical principals.
Some paradigms endure, others do not. Most are highly explanatory and only are replaced when fatal flaws are uncovered and a replacement paradigm exists. The physics of Isaac Newton will land you at a precise point on the moon, and will work just fine as long as your velocity is much less than the speed of light, whereas proper prediction and control require the newer paradigm of relativity. The Ptolemaic universe was quite adequate until Copernicus and Galileo blew it to bits. The flat earth couldn't survive the voyage that Magellan didn't survive.