Karl Popper's ideas of science are interpreted as naive and decades out of date by those who study the
philosophy of science. Many scientists still cite Popper's older theories as the ultimate authority in how the "business" of science works, that is:
- Devise hypothesis. (Perhaps a modification of an existing hypothesis)
- Test for falsification. (i.e. Experiment to disprove hypothesis)
- Conditional acceptance. (The new theory "wins" until something better comes along.)
- Repeat.
In reality, Popper's trials do not describe the actual messy "business" of science, which is brilliantly described in
Thomas Kuhn's landmark book,
The Structure of Scientific Revolutions. For example, step 2 can be impossible, or the results could be far from clear-cut. What then?
Consider the case where we have two competing theories which can each explain a subset of observations. The orthodoxy explains many data very well, and is quite accepted. But doubts remain about some "outlier" data. Experimental error? A newer, radical theory explains the "outlier" data but perhaps has outlier data of its own. Is the new theory any good? Will it "grow into" something that explains all? Sometimes, a tension develops between supporters of the new theory and those in the orthodoxy. Whose idea "wins" the hearts and minds of scientists? Do we reach some kind of hybrid acceptance? (E.g., Newtonian physics is still extremely useful.) By this stage we have journeyed far from the territory Popper describes.
Kuhn's book coined the famously abused phrase "
paradigm shift" to describe scientific revolution as a social phenomenon. Popper's ideas are most relevant for incremental changes to the orthodoxy and cannot explain paradigm shifts.
Some
sample paradigms and paradigm shifts.