I argue that scientific theories are (essentially) never plain wrong.
The nature of science
To begin, we need to have a clear definition for separates a (generic) theory from a scientific theory. Anyone is capable of making a theory – I could propose that purple unicorns are the cause of our attraction to earth, or that aliens artificially control our every action. A scientific theory is an attempt to explain the basis for phenomena that are observed, and generally makes testable predictions that may lead to confirmation of the theory’s validity. While a theory can be incredibly absurd and/or wrong(as seen above), a scientific theory is subject to the more rigorous criterion listed above. These criterion, along with the scientific method, makes it so that scientific theories are (almost) never plain wrong.
While it may seem counterintuitive that something can be designed in such a way that it is almost never wrong (in violation of Murphy’s Law), this is a result of the rigorous criteria placed by the scientific method. All scientific theories are subject to constant revision and monitoring, and any result must be replicable. In addition, all theories must be consistent with existing data – theories are rarely accepted unless they can adequately explain existing data. Thus, even if a theory fails to explain any new data, the theory will at least be able to explain the original data – hence, a theory can never be completely wrong.
Newtonian physics
A classic example of a flawed scientific theory is Newton physics. Of course, up until the 1900s, no evidence contradicted Newtonian physics, and a number of erroneous conclusions were believed to be true(note 1). However, as the precision of instruments increased, data gathered from many experiment(note 2) soon suggested that these theories, and their resulting conclusions, were inaccurate. In the following few years, physicists reworked many equations until they were consistent with the gathered data. In turn, these theories were subsequently modified(note 3), until another model that was consistent with all the data was created.
Although Newtonian physics was flawed, it is still an incredibly good theory. After all, if you were to ask Richard Feynman (note 4) about a projectile motion problem, he would use Newtonian physics to solve the problem. For a high-velocity object such as an incredibly fast bullet train traveling at 3000 meters/second, Newtonian physics is 99.9999999999% accurate. As such, it is quite reasonable to say that Newtonian physics is quite valid(note 5) – it just so happens that scientists have created a slightly more nuanced description of nature.
Flat-earthing(an exercise in absurdity)
As we’ve seen, a scientific theory such as Newtonian physics is really quite valid, even if it’s a little bit flawed. But let’s take the worst case scenario, perhaps through some (absurdly) bad conclusion drawing.
Suppose a scientist is measuring the land, and observes that the ground is mostly flat. The scientist then repeats this for many locations, and throughout, observes that the earth is essentially flat across each of these locations. Based on the data, the scientist then (incorrectly) concludes that the earth is flat. Supposing that it does become a theory, one might be tempted to regard this as an example where a scientific theory is just plain wrong. However, this is not entirely true – the scientist’s theory is somewhat valid – the earth is (more or less) flat at small length scales.
If we examine a sphere from far away, it is obviously round and curved. But if we zoom in a little further, it looks a little less round, and a little less curved. If we zoomed in even more, the earth would look shallowly curved(note 6). Now, applying this to real life, we rely on this fact every day. When we park cars in parking lots, we don’t bother considering the earth’s curvature at that point. When construction workers use a leveler, they neglect any curvature of the earth. Even though the scientist’s theory is flawed, it still finds uses in everyday life.
In summary, the data-driven scientific method is powerful, rigorous, and ultimately yields powerful results. Any scientific theory must explain existing phenomena accurately, making it intrinsically correct to some degree – even the most absurd theories that are created with the scientific method contain some degree of correctness.
Snarky notes and other useful things
Note 1: These conclusions include infinite maximum speed(replaced by special relativity), infinitely continuous quantities(replaced by quantum mechanics), and simultaneity(replaced by special relativity).
Note 2: These experiments include the Michelson-Morley experiment, muon half-life observations, and gravitational lensing due to the sun.
Note 3: Quantum electrodynamics combines relativity and quantum mechanics. It, in turn, is part of the Standard Model of Physics. For the most part, equations are all math – and I applaud anyone who can truly understand them.
Note 4: Feynman won the Nobel Prize in physics for his work on quantum electrodynamics, and is/was widely regarded as an excellent physicist. He also died a while back, but he did teach undergraduate classes at Caltech(and used Newtonian physics to solve Newtonian problems).
Note 5: A 99% is generally considered to be a very good grade on a test(and indicates that a student is not completely wrong/reasonably knowledgeable), the same logic applies for scientific theories.
Note 6: This may be difficult to visualize, but the link here: http://ned.ipac.caltech.edu/level5/Guth/Figures/figure4.jpeg gives a good idea of local sphere flatness.