Checkers is a social construct. It is a game, played on a board with sixty four squares, with certain rules about how pieces are moved. However, although it is a social construct, it is possible for any observer to determine whether or not the game of checkers is being played. Of course, it possible to develop other games. These games might be more or less enjoyable than checkers, but they are just other games.
Now, let us consider another board game: Combat Commander Europe. CCE is a board game, and in this sense, it is similar to checkers. However, CCE purports to simulate what combat in war is really like. Therefore, while it is possible to change the rules of CCE, these rule changes will either make the simulation more accurate or less accurate. Such changes might make the game more fun or less fun, as well, but they will also change how well the game simulates real warfare.
Scientific theories are, in a sense, ‘games.’ They are social constructs, yes, in the sense that humans have proposed them. However, they are constructs which are intended to describe reality. And in this sense, we can test them to see which ones are accurate – or, if we want to be very skeptical, we could say, we can test them to see which theories are useful for predicting future outcomes.
Thus, for example, Newtonian physics is useful for mechanical predictions, whereas Aristotelian physics is not useful.
It is absolutely true that the particular form which a theory takes is ‘arbitrary.’ As an example, every theory of classical mechanics can be made in the form of Lagrangian mechanics, which involves a geometric space called a configuration space and an ‘action,’ which involves variables which describe the position on the configuration space as well as the tangent vectors on the configuration space. On the other hand, every theory of classical mechanics can also be put into the form of Hamiltonian mechanics, which occurs on a phase space, which has a symplectic form, and which involves a Hamiltonian. In Lagrangian mechanics, the evolution of a system is determined by extremizing the action. In Hamiltonian mechanics, the evolution of a system is determined by the Poisson brackets with the Hamiltonian (these are determined by the symplectic form).
I realize that many people may not know what these terms mean. Suffice it to say that Lagrangian mechanics is one approach to classical mechanics, which describes the mechanics in one way, while Hamiltonian mechanics is another approach which describes it in a completely different way. (It turns out these approaches are related by the Legendre transform for most cases in classical mechanics – complexities arise when considering systems which are particularly symmetric).
As an analogy, consider describing a situation in English versus in German. The particular words used are a social construct, but the English language system is useful for describing the situation, as is the German language system. There are other conceivable languages which might be completely inadequate for describing a given situation. As another analogy: consider how arithmetic can be described by three methods: Arabic numerals (1, 2, 3, etc.), Roman numerals, and tick marks (0, 0′, 0”, etc.). We could do arithmetic with any of these conventions, but the same results would apply when we ask what 2+2 equals.
So, similarly, some theories are useful for making predictions, and others are not. These are the ‘useful social constructs.’ The theory of Newtonian gravity is useful for making astronomic predictions; the theory that a flat earth is carried on the back of a tortoise is not (among other things, we can look for the tortoise, and observe that it is not where such a theory says it should be).
This is why although scientific theories are social constructs in one sense, they are not arbitrary social constructs, nor are all theories of equal value. Some are objectively instrumentally valuable for the purpose of predicting future outcomes, whereas some are not valuable for doing so (recall that there is a difference between instrumental and intrinsic values). Some are downright bad for doing so, such as the Aristotelian theory of gravity (which is internally inconsistent anyway, as Galileo demonstrated).
The purpose of the scientific method is to help in figuring out which social constructs are useful for predicting future outcomes, and which are not useful. A ‘scientific theory’ is an hypothesis which passes the scientific method of testing to see whether it is useful for predicting outcomes. And this is why we accept those theories which meet its criteria, and reject those which do not. Thus, physics and biology have proved themselves to be useful, whereas the theory of ‘humours’ has not proved itself useful. Nor have the theories put forth by tribal medicine men. While some critical theorists have claimed that, say, African magic is an equally valid way to understand the universe, it is not – or at least, it has not proved itself to be, and to prove that it is, it would have to satisfy the same rigorous testing criteria as our physics and biology.
This is the reason that scientists reject such theories. It has nothing to do with racism or Eurocentrism. In fact, many scientific breakthroughs are made by non-Europeans (and, those seeking to advance pharmacology often look at plants which natives use medicinally). However, we go with what proves itself to work (I say we, although I am a mathematician, not a scientist). Some social constructs prove themselves useful as predictors, and some do not. Scientific theories are not arbitrary, even though they are social constructs in a sense.
I write this because it is true that science is a social construct, but there are some who think this means that science is arbitrary or biased. When science is done properly, it gives us a useful method for predicting outcomes, and it rules out non-useful ways of predicting the future. There may be other useful ways to predict outcomes, but these are generally just translations of the scientific theory into ‘other languages.’ And there may be more accurate ways to predict outcomes, just as Einstein’s theory of relativity supplanted Newtonian gravity. The point is that science is objectively a superior way to predict the outcomes than other methods. I have focused on physics here, because it is the area which is most familiar to me. However, the same arguments apply to other areas of science – when they are properly done.
Sometimes, people will claim something as ‘scientific’ when it is, in fact, pseudoscience. This is the case with such areas as phrenology or critical theory (which hovers between claiming to be scientific and claiming to reject scientific methods as ‘Eurocentric’). We cannot blame ‘science’ for the failures of pseudoscience, because pseudoscience is a failure to properly follow the scientific method. However, at this point we run quite far afield.
Suffice it to say: scientific theories are social constructs, yes, but they prove themselves to be useful at making predictions, rather than being arbitrary constructs of no more value than any other constructs.