Some thoughts in the philosophy of science

When I first got into philosophy as a senior in high school, my interests gravitated towards metaphysics (the branch of philosophy that studies existence) and epistemology (the branch of philosophy that deals with the acquisition and validation of knowledge). I was still interested in ethics, politics, aesthetics and so on, but I was most interested in metaphysics and epistemology. However, in recent years, my interests have shifted to the philosophy of history and the philosophy of science.

With that in mind, these are just some musings on the thought of the Princeton philosopher of science and developer of "constructive empiricism" Bas van Fraassen and one his most vocal critics, the prominent student of Karl Popper, Alan Musgrave.

* * * * *

In his piece entitled “Arguments concerning Scientific Realism”, Bas van Fraassen argues for an alternative to scientific realism that he calls “constructive empiricism”. Van Fraassen holds that the realist position is generally naïve. His interpretation of the realist view runs thusly: “Science aims to give us . . . a literally true story of what the world is like; and acceptance of a scientific theory involves the belief that it is true.”

Generally, antirealist theories, of which van Fraassen’s constructive empiricism is a strain, hold that science does not offer theories that are true in the literal sense or that one need not believe a theory is true to accept it (or both). Van Fraassen’s antirealism holds that theories can be literally construed, but a theory’s truth is not a requisite of it being a good theory. Van Fraassen is also somewhat of a literalist: whatever a theory says, implies or entails then that is what the theory says, implies or entails.

Van Fraassen proposes that science “aims to give us theories which are empirically adequate…” An "empirically adequate" theory “saves the phenomena”. Presumably, that is when what a theory says about observables is true. So, to accept a theory is not to hold the theory to be true, but to hold that what the theory says about the observable world is true.

So, what of the non-observable? Van Fraassen’s position can best be described as agnosticism. One must suspend judgment and withhold belief about the theoretical, the unobservable. A good theory is a theory that is true when discussing observables—if it says anything about unobservable phenomena, the theory (or at least the part of the theory that addresses the “theoretical”) is flippantly dismissed in the typical agnostic fashion: "Who knows?"

This leads right into van Fraassen’s treatment of the observable/theoretical distinction, which is largely a polemic against realist arguments such as those of Grover Maxwell. Given that van Fraassen’s constructive empiricism would require him to at least refrain from dismissing with such a distinction since holding that accepting a theory merely means holding that the theory is “empirically adequate” presumes that one can discern observable from unobservable phenomena.

Van Fraassen points to what he believes is a clear case of an unobservable object. His example involves charged microparticles traversing a cloud chamber leaving behind them a visible condensed-vapor trail. Though the trail is visible, van Fraassen argues, the particles themselves are not visible. Although an observable trail is seen, the particles themselves cannot be said to be observed—one observes only its purported effect, the particle is merely “detected”. It therefore follows, van Fraassen claims, that there are unobservables about which one can be an antirealist (that is, agnostic about the existence of that which one has not observed directly).

Another challenge to van Fraassen’s antirealist constructive empiricism is what is commonly called the inference to the best explanation rule. Briefly, this rule states that given some evidence and two (or more) alternative theories (hypotheses), let’s say A and B, we would do best to infer that B is the correct theory if B does a better job than A of explaining the evidence. The example van Fraassen uses is: given my missing cheese, the scratching noises I hear at night and the hole in my floorboard, I infer the existence of a mouse in the walls of my house. The argument is that if one has good reason to believe B, then one has good reason to believe that any existents entailed in B actually exist, even if they are unobserved.

The realist argument, using the inference to the best explanation rule as a premise, is that we generally follow the rule on such a widespread scale of instances that we must commit to the belief in the unobservable. Van Fraassen argues that the widespread utilization of any such pattern of inference does not prove anything. One can combat such a notion by proposing an alternate pattern that also fits the example. Van Fraassen believes that one will always believe what best explains the empirically adequate.

Hilary Putnam offers another challenge: the “miracle argument” (or “ultimate argument” as van Fraassen calls it). Putnam’s claim is that the only explanation for the continuing success of scientific theories is scientific realism. Otherwise, since we don’t actually believe in the truth of our theories, their continued predictive success can only be attributable to a long line of miracles.

Van Fraassen answer is Darwinian in its structure: many theories struggle to survive and, invariably, those not adapted to their environment die off. Of course, for van Fraassen, the adapted theories are empirically adequate. Empirically inadequate theories don’t survive. He concludes that, as species evolution is not miracle, neither is the death or survival of theories.

Van Fraassen’s attack on scientific realism in his The Scientific Image (from which “Arguments Concerning Scientific Realism” is taken) has garnered quite a lot of criticism of its own. Notably, the book Images of Science contains Alan Musgrave’s critique in the article “Realism Versus Constructive Empiricism”. The article contains critiques of several points of van Fraassen’s constructive empiricism, but I will focus on two in particular: van Fraassen’s account of the connection between truth and empirical adequacy and his account of the observational/theoretical distinction.

The constructive empiricist recommends that we accept an empirically adequate theory about unobservables, but not believe it. This is because, the constructive empiricist holds, all we can know about a theory is whether the aspects of it relating to observable phenomena are judged true. Any parts dealing with unobservables can’t be judged true or false and so any empirically adequate theory that entails unobservables is to be accepted but not believed as to take the realist position that one can believe a theory about unobservables is true is unnecessarily riskier.

Musgrave argues however, that these worries apply also to the constructive empiricism. He argues that an empirically adequate theory must “save” (that is, explain) all observed phenomena past, present and future. Therefore, judging a theory to be empirically adequate requires more knowledge then we can have at any given time.

Another issue Musgrave confronts van Fraassen on is his account of the observational-theoretical distinction. Though van Fraassen does not defend the observational-theoretical distinction (he believes that “observable” is too vague a predicate to demonstrate that there is a distinction between cases where it applies and cases where it doesn’t) he refuses to acknowledge that a seamless transition from observation to detection is harmful to the dichotomy. Observing microbes under a microscope and observing the trails of microparticles in a cloud chamber are fundamentally different to van Fraassen: the latter are not observed but detected. Van Fraassen holds that the limitations on what is observable are what “a final physics and biology” tell us are our limitations as humans. Musgrave argues that such limitations vary from person to person and are species-specific and, therefore, cannot be granted special significance.

Unfortunately, the best arguments against van Fraassen’s brand of scientific antirealism can be found in the writings of Hilary Putnam. I say "unfortunately" because they do not address van Fraassen on a fundamental level and show why he is wrong. Most other philosophers, either addressing van Fraassen directly or being the object of van Fraassen’s polemics, though they may be well-intentioned in trying to dismantle antirealism, often fail…horribly. This includes Musgrave. Van Fraassen, however, doesn’t escape Putnam’s “miracle argument”. His answer of evolving theories is almost a non-sequitar.

More unfortunately, and more fundamentally, the entire field of philosophy of science (and the philosophy of mind and several others while we’re at it) appears to be suffering from a kind of Rationalism (ideas divorced from reality). That is, the philosophers in the field seem so divorced from reality and the need to tie their theories to the facts of reality that philosophical debate almost invariably devolves into nit-picking highly specialized pet theories with no references to the real world.

In regard to van Fraassen in particular: that his work is taken as seriously as it is, is ominous. Without risk, science stagnates. Every advance, every great leap in a new, unexpected direction, every new enlightenment came with risk. Lavoisier’s work in chemistry provided more proof for the atomic theory of matter than any work in any field before him. Atoms were still “unobservables”, but do we ignore the evidence? Do we sit in the agnostic limbo, throw up our hands and say “Well, who can say for sure?” Faraday’s work united the fields of electricity and magnetism, forces which worked largely as “unobservables”. Should the world have ignored his work? What could be empirically verified is empirically verified, but what then? We accept the non-empirical portion of these theories but if they aren’t true or false, and we couldn’t know anyway, what does it matter?

We can’t see black holes. We know they are there by the effects they have on their surroundings. Often the proof that something exists is not actually seeing it, but seeing what it, and only it, can do. What else explains the vapor trails in the cloud chamber? And if we’re wrong, we adjust and fix it. Reality is reality, everything has an identity, causality exists. Mistakes in knowledge are not disasters, a philosophy which evades reality in order to make truth or falsehood superfluous, is.

Scientists and philosophers (and, it seems, almost every scientist-philosopher) so often make the mistake of believing that scientific theories have philosophical implications. This inverts the relationship between science and philosophy. Philosophy is the fundamental discipline. It comes first. The Uncertainty Principle didn’t have an enormous effect on metaphysics or epistemology: Heisenberg’s bizarre personal philosophy lead him to a bizarre interpretation of his research.

One hears talk of the end of physics. This may be prophetic, though not for the reasons the prophets believe. Everyone, even scientists, has a philosophy. It may be implicit and unstated, swallowed dogmatically and unquestioningly, but it is there. It guides everything we do. With the state of modern philosophy, with philosophy nearly being on its deathbed since Kant, is it any surprise that the sciences are crumbling?

Ghouls & slugs--

Matt