Topics Science

Question I have a question about the entities in scientific theories and models. The status of some of these objects seems intuitive. Frictionless planes, for example, though they don't exist, seem helpful enough as an abstraction for understanding how actual planes function. My question is about entities which we (non-scientists) know only through the prism of scientific theory--say, photons or electrons. I know what light and electricity are, on some immediate level, through my everyday experience of these phenomena. But I don't know what to make of the entities that physics tells me compose them. My inclination is to take them as "real", not just convenient notions of the purpose of theorizing and mathematical models. I can't help pictures these tiny, planet-like spheres whizzing around. I know there's something importantly wrong about that image, but I'm not sure what. Furthermore, scientists still talk about particles "spinning" and so forth, I am unable to see in what sense this is an analogy in the way a frictionless plane is to an actual one. The point of my question is that science has a kind of authority to inform our understanding of the natural world even though it is too difficult for the non specialist to evaluate its claims. So as a non-scientist, how should I regard the claim that light is composed of "photons" and also "waves" given that these very concepts seem to rest on a great deal of theory that might potentially change such that the entities we now regard, naively or not, as "real" turn out not to "exist" for the purposes of future theories. From the educated lay-person's perspective, aside from considerations of parsimony, and given the way scientists speak to the public, how can we distinguish between theoretical entities which simply turn out to be convenient (for the moment) fictions, like say, Ptolemy's epicycles or "aether", and entities that exist in some more substantial way? This seems kind of important given the authority science has as the arbiter of how to understand the natural world. Thanks!

Accepted:
June 7, 2012

Comments

The question that you are raising is a venerable and perennial one. In the trade, it is called the dispute between "scientific realism" and "scientific anti-realism." Scientific realism is the view that in science, when a theory is accepted, the unobservable entities that the theory posits are believed to exist and the theory's statements about them are believed to be (approximately) true. Scientific anti-realism is the view that in science, when a theory is accepted, the theory's claims about observable facts are believed to be true, but the theory's claims about unobservables are not believed to be true. Rather, they are believed to make accurate predictions about observables. That is all that science requires of its theories about unobservables.

In your question, you alluded to Ptolemy's epicycles. This is a perfect case in point. In the ancient world, the celestial spheres posited by astronomical theories as carrying the planets along in their orbits (around the earth) were widely regarded not as genuinely existing, but merely as geometric machinery that could "save the phenomena" - that is, that could make accurate predictions regarding planetary positions in the night sky. The ancients knew that many different combinations of celestial spheres could produce the same predictions regarding the positions of the planets as seen from earth. Therefore (they concluded), to regard any of these combinations as more likely than any other to be correct, when they make exactly the same empirical predictions, would be unjustified.

Of course, as you know, it turned out that none of these combinations of celestial spheres was real. But (you asked) what should we think about the unobservable entities posited by today's best scientific theories? Perhaps there are other theories besides the ones that we have managed to come up with that make the same empirical predictions as our best current theories. Should we then remain "agnostic" regarding the truth of those theories, and believe them only accurate regarding all actual and possible observations?

On the other hand, even if two theories agree regarding all of the actual observations that we have made so far, we are surely oftentimes justified in believing that one of these theories is much more likely than the other to be true. Suppose (to choose a notorious example from Nelson Goodman) that every emerald that we have checked so far has been green at the time that we have checked it. Then the theory that all emeralds at every moment are green fits our observations -- but so does the theory that all emeralds at every moment are "grue", which means that all emeralds at every moment are green if the moment precedes the year 3000 and are blue if the moment does not precede the year 3000. This ridiculous theory fits all of our observations (since every moment at which we have observed emeralds has been before the year 3000). Now if we are justified in favoring one of these two theories over the other, though they both agree with all of the observations that we have made so far, then why wouldn't we be justified in believing in the truth of a theory about unobservable entities even if there are (at least in principle) other theories that differ in what they say about unobservables but agree with the given theory's predictions about observables?