Essay stands on it’s own but for more context:
See part 1
See part 2
Okay. For now, forget about papers.
Let’s assume that papers exist as papers, not as a potential X”; and that science can give a true description of the observable parts of the world because we perceive the real thing. (When we talk about observable features of the world, we talk about trees, water droplets, and everything in-between.) Now, to begin to stir up some more controversies, how should we start thinking about the unobservable parts of the world?
Leptons, quarks, electrons, do they really exist as such? A scientific realist will say, “absolutely!”. However, an antirealist will disagree with the word ‘absolutely.’ To put it a bit more precisely, a modern antirealist will be agnostic with regard to the ‘reality’ of an unobservable entity.
Let me take a step back and stress that both folks on the opposing side of this aisle are metaphysical realists; that is, they believe that there is a mind-independent world in which we can pontificate on its trueness (or the lack of it). That is, with regards to the observable parts of the world, both the science realist and antirealist have no dog in that fight. The reader should note that this position stands visibly in contrast to metaphysical solipsism.
Now that we have gone this far, indeed, we can’t go back. And perhaps you are saying, what is all these fusses about (un)observable entities of the world anyways? Here is the scientific realist argument: 1) given that the theories of the unobservable entities had been highly successful empirically (i.e., accurate predictions); 2) it must be an extraordinary coincidence for it not to be true; therefore, 3) scientific realism is true. There goes the famous ‘no miracle’ argument – to believe otherwise is to believe in miracles (an uppercut for an instrumentalist).
Perhaps not so fast for the scientific realist because we can think of a scenario that will perturb the peace of the premises stated above: ontologically different theories might appeal to precisely the same evidential properties. And if it is the case that they are ontologically distinct, then it can’t be the case that they are both true. Hence, scientific realism is false. So technically, we say that the observational data ‘underdetermines’ the theories – i.e., the underdetermination argument against scientific realism.
Let’s say I am bluffing, except that I am not: take the example of the phlogiston theory of combustion, which argues that during combustion, phlogiston is released. This was believed to be true up until the 18th century. Except that there is nothing like phlogiston – it was all a lie – as subsequent research has shown, burning occurs when substances react with oxygen. This example could be cited by a scientific antirealist, arguing for the position of pessimistic meta-induction. Simply put, scientists have offered theories in the past that has failed or evolved. So why should one think that the best theory now will do better?
In addition, the reader must not miss the elephant in the room, the inference to the best explanation (IBE) motif in the ‘no miracle’ argument – It’s screaming out loud. The argument does not prove that scientific realism is true; it only infers that it is true based on the best explanation of the ‘data.’ And the news isn’t new: the circularity in inductive arguments still hasn’t been defeated. In other words, what is the justification for induction itself, provided we still have circular reasoning in our logical fallacy dictionary? To add insult to injury – even if we grant such and such – the best explanation might just be the best explanation among a bunch of terrible explanations.
Notice how the scientific realism/ antirealism debate leads us back to the theory of truth. A scientific realist is more likely to think of a good explanation as a proxy for truthfulness. In contrast, an antirealist will think of a good explanation as a practical tool, something useful, not necessarily true or real.
Let’s rest our case for now.