Here’s why I link to David Coppedge:
This [is] not to say that we must doubt what a car is, and whether we are driving down the road, or what will happen if we drive into a brick wall. But when discussing scientific objects like a quark, or the interior of a star, or a black hole, or a species, or a population, or a gene, or global climate change, or laws of nature – and the causes of their interactions and how we explain them — scientists get onto slippery philosophical ground very quickly. A finite number of instances cannot necessarily be extrapolated to a class. A time sequence does not necessarily imply a cause-effect relationship. A majority of scientists can be wrong (bandwagon fallacy). And today’s best theory is not necessarily a good theory (Macbeth’s best-in-field fallacy). These are just a few of the caveats that can be levied against even simple, down-to-earth explanations in our everyday experience. The problems become even more significant in astronomy and cosmology, where the line between theory and observation is often very fuzzy.
This entry about globular clusters reminds us that scientists can never simply assume that their theories correspond to Reality with a capital R. No matter what the textbook says, or how solid the scientific consensus appears, everything in science is subject to revision. Some things are subject to revolution. Whenever a scientist says “We now know,” that’s the time to get out the red flag and say, “We heard that line 100 years ago, and now look.”