> Air bubbles in water can be modelled as having a negative mass. Forgive the naive take on this, but it's a weird point to me... if air bubbles in water can be modelled as having negative mass, but clearly do not (they are made of particles with positive mass), then they are only "negative" in relation to the material surrounding them. So even if they can be modelled this way successfully, clearly the model doesn't…
The concept of whether something is "real" is a metaphysical amusement. And it is never meant to be answered in modern physics. Physics has come a long way into what it is today from natural philosophy. I believe all physicists in this day and age subscribe to the notion of model-dependent realism in one way or another. https://en.m.wikipedia.org/wiki/Model-dependent_realism Perhaps someday in the distant future we w…
For instance a recent conversation I had was discussing the big bang and the cosmic microwave background radiation. Somebody thought that the CMB was evidence of the big bang. Of course when the CMB was first observed/measured it completely falsified our idea of a big bang due to the horizon problem -- regions of space that should not be causally connected are somehow homogeneous, which should be impossible. But instead of this observation != prediction indicating some fundamental problem we instead add in a magical hyper-inflation period to the model of the big bang so that the model can now "predict" what we see. And now you have people that believe that the CMB itself is evidence of the model's accuracy, when in reality the CMB refuted our models and we were forced to massage it back into the model in a rather arbitrary fashion. There is absolutely no physical reason to accept hyper-inflation other than 'if we add this magic event, then we can keep using this model.'
For a less contemporary example consider geocentricism - the idea that everything rotated around the Earth. Prior to Newtonian mechanics and other 'real' systems of interaction, we simply used models for orbital mechanics. And these models ended up being quite rediculous. You had planets doing magical swirlies, stopping and going backwards, and all other sorts of things. But because you can't falsify models detached from reality, everybody just shrugged their shoulders and accepted it for something on the order of centuries. And as time passes these models become even more difficult to replace because it ends up meaning you'd basically have to toss something on the order of decades, if not centuries, of past work. In this particular case astrology, for instance, was in the past a scholarly and academic field analogous in many ways to psychology today. Refuting geocentricism involved completely scrapping this entire field, and centuries of work within it, due to the fact it took things such as Mercury going backwards as key components of its analysis of human psychology.
The point of this all is that as you accept models, it can be that you're searching for answers at the entire wrong level of a problem. For instance astronomers in times of a geocentric universe may have been trying to explain why Mercury went backwards during its orbit, yet it's extremely difficult to answer something when you start with a wrong assumption -- in that case that the planet does indeed go backwards at some point, planets have these orbital swirlies, and so on. People like to rewrite history to blame geocentricism on the church, but inertia is not limited to divine belief.