Karl Popper made falsification the key to scientific legitimacy. But as others have pointed out, scientists do not spend much time trying to falsify theories. Instead, they work to confirm and extend theories. Moreover, an observation that goes against a theory doesn’t falsify the whole theory; it creates an anomaly that can be dealt with in various ways–for example, search for hidden factors, modify the theory slightly, or discount the observation.
It is only when a superior theory arises that explains anomalies and everything else an older theory explained that scientists take note. So there can be a period of instability as some people question the theory and others try to defend it. This has happened many times in the history of science, from the geocentric-heliocentric debate, to the origins debate of today.
What I’d like to suggest is that falsification shouldn’t be the motivation regarding a theory which has some evidence for it. The question should be: What are the limits to the theory? The fact is that all theories have their validity limits (as Fritz Rohrlich calls them). Why? Because theories assume simplifications of reality, construct isolated systems, and are based on limited data.
While scientists posit theories that are nominally universal, that scope is merely a default in place of the unknown limits that will be discovered later. Science is both optimistic that its theories cover a wide number of cases and open to findings of the limits of theories. Promoters of science seize on the optimistic part and downplay or ignore the limitation part.
In the 18th century enthusiastic Newtonians were very influential in making a clockwork universe the common mindset. Their mistake was taking the word “universal” in universal gravity literally as if Newton’s theory had no problems. In recent years promoters of universal common descent have been very influential in taking the word “universal” literally in evolution, even as the limits of natural selection (and other mechanisms of change) are becoming more known.
Science should search for the limits of every theory. That can be done by finding out the conditions under which it is false, or it breaks down, or works poorly. This sets the stage for a superior theory, that is, one with a larger extension. It also puts all theories on the same level: they all can have their uses but they are always limited.
Scientific theories are not falsified; they are limited, and their limits become known over time.