Scientific Realism: Some Arguments

Philosophy of Science

The first three arguments are all more or less present in, or at least suggested by, Maxwell's essay. The fourth is also discussed by van Fraassen, but doesn't seem to be explicitly discussed by Maxwell.

1. Argument from Continuity

1. We should believe in observable entities
2. There is no sharp dividing line between observable and unobservable
So, 3. We should believe in "unobservable" entities also.

Van Fraassen's reply: this is a slippery slope argument (or, to use a fancier term, a sorites argument). It's no better than Sextus's argument that incest is OK, or the argument that everyone is bald.

2. Argument from Observability in Principle

1. We should believe in the existence of the observable entities that our theories invoke
2. "observable" means "can be observed"
3. Everything can be observed (if our technology or our sense organs improve enough)
So, 4. We should believe in the existence of every entity our theories invoke

van Fraassen's reply: van Fraassen would reject either 2 or 3, depending on how we understand "can be observed." 2 is correct if "can be observed" means "can be detected by contemporary humans with unaided sense organs." But in that case 3 is incorrect. 3 would be correct if "can be observed" meant something like "can be detected by humans with the aid of present or future technology," but then 2 would be incorrect.

Also, according to van Fraassen, advances in technology will not make more things observable. On his view, we don't observe things through a microscope, for instance. The only change that would make something previously unobservable become observable would be a change in our sense organs, or else bringing into our epistemic community other beings with better sense organs than ours.

3. Argument from Epistemic Unimportance

1. We should believe in the existence of the observable entities that our theories invoke
2. The observable/unobservable distinction has no epistemic importance
So, 3. We should believe in the existence of all the entities invoked by our theories

van Fraassen obviously rejects premise 2. Is there a way to argue about 2 in a more persuasive way than "yes it is" vs. "no it isn't"?

4. Argument from Inference to the Best Explanation

1. We should believe the theory that best explains the evidence ("inference to the best explanation" or "abduction")
2. The theory that best explains the evidence invokes unobservable entities
So, 3. We should believe in the existence of unobservable entities

van Fraassen rejects premise 1. He argues that we should accept the theory that best explains, but acceptance doesn't require belief in all of the theory, just in the claims the theory makes about observables.

This is related to van Fraassen's rejection of the idea that explanatory power is an empirical virtue: he classes it instead as a pragmatic virtue. What makes him a (constructive) empiricist is the idea that only empirical virtues give reasons for belief.

Pragmatic virtues may give practical reasons for accepting one theory over another, but that doesn't give us a reason to think that the theory's claims about unobservables are literally true.


Curd, Martin, J.A. Cover, and Christoher Pincock, Philosophy of Science: The Central Issues, Second Edition. New York: Norton, 2013.

van Fraassen, Bas. "Arguments Concerning Scientific Realism." In Curd et al., 2013. (First published 1980, in van Fraassen's book The Scientific Image.)

Maxwell, Grover. "The Ontological Status of Theoretical Entities." In Curd et al., 2013. (First published 1962.)

Last update: September 12, 2016
Curtis Brown | Philosophy of Science | Philosophy Department | Trinity University