Fallacies physicists fall for

In his essay “Quantum Mechanics and Ontology” in his anthology Philosophy in an Age of Science, Hilary Putnam notes that “mathematically presented quantum-mechanical theories do not wear their ontologies on their sleeve… the mathematics does not transparently tell us what the theory is about.  Not always, anyhow” (p. 161).  Yet as Putnam also observes:

The reaction to [such] remarks of most physicists would, I fear, be somewhat as follows: “Why bother imposing an ‘ontology’ on quantum mechanics at all?... [Q]uantum mechanics has a precise mathematical language of its own.  If there are problems with that language, they are problems for mathematical physicists, not for philosophers.  And in any case, we know how to use that language to make predictions accurate to a great many decimal places.  If that language does not come with a criterion of ‘ontological commitment,’ so much the worse for ‘ontology.’”…

[But] to say “We physicists are just technicians making predictions; don’t bother us with that ‘physically real’ stuff” is effectively to return to the instrumentalism of the 1920s.  But physical theories are not just pieces of prediction technology.  Even those who claim that that is all they are do so only to avoid having to think seriously about the content of their theories; in other contexts they are, I have observed, quite happy to talk about the same theories as descriptions of reality – as, indeed, they aspire to be.  (pp. 153-4)

The problem is not confined to the interpretation of quantum mechanics.  The metaphysical implications of relativity theory, or indeed of any theory in physics, is something the physics itself does not reveal.  Then there are more general philosophical questions about science which science itself does not and cannot answer.  For example, what is the relationship between the abstract mathematical representation of nature afforded by physical theory and the concrete reality that it represents?  Is there more to nature than mathematical representations can capture?  What demarcates science from non-science?  What is a law of nature?  Why is the world law-governed in the first place?  And so on.

The tendency of those beholden to scientism, including professional scientists who are beholden to scientism, is to dismiss such questions on the grounds that the only thing worth talking or thinking about is whether the predictions pan out – which entails positivism, or instrumentalism, or some other form of anti-realism.  And yet, when pressed about this implication, or when presenting the findings of science to the layman, the same people will usually insist on a realist understanding of scientific theories – apparently blithely unaware of the contradiction.  And this is an equal-opportunity form of cognitive dissonance, afflicting everyone from whip-smart Ph.D.’s down to the dumbest combox troll.  

You can’t have things both ways.  If you insist that nothing worthwhile can be said about any matter that is not susceptible of experimental testing, then you have indeed ruled out of bounds philosophical questions like the ones just referred to.  But you have also thereby ruled out a realist interpretation of theoretical entities, because realism is not susceptible of experimental testing.  That’s the whole point of the debate between realism and anti-realism – that the experimental results would come out the same whether or not theoretical entities are real or just useful fictions, so that the dispute has to be settled on other grounds.

Indeed, you can’t have things even one way.  For suppose the physicist or the combox troll beholden to scientism sees the problem and, to be consistent, adopts an across-the-board instrumentalism.  He avoids philosophical issues like the ones mentioned, and he also refrains from endorsing realism.  The problem here, of course, is that even instrumentalism itself is a philosophical thesis and not a scientific one – again, the dispute between realism and anti-realist views like instrumentalism cannot be settled experimentally – so he is not really being consistent after all.  

Scientism is simply not a coherent position.  You cannot avoid having distinctively philosophical and extra-scientific theoretical commitments, because the very attempt to do so entails having distinctively philosophical and extra-scientific theoretical commitments.  And if you think that these commitments are rationally justifiable ones – and of course, anyone beholden to scientism thinks his view is paradigmatically rational – then you are implicitly admitting that there can be such a thing as a rationally justifiable thesis which is not a scientific thesis.  Which is, of course, what scientism denies.  Thus scientism is unavoidably self-defeating.

The fallacy is simple, and blindingly obvious once you see it.  So why is it so common?  Why do so many otherwise genuinely smart people (as well as people who merely like to think they are smart, like combox trolls) fall into it? 

Part of the reason is precisely because it is so common and so simple.  Again, as Putnam complains, even many professional scientists (by no means all, but many) commit the fallacy.  So, when you call someone out on it, there is a strong temptation for him to think: “If my critic is right, then I and lots of other scientists have been committing a pretty obvious fallacy for a very long time.  Surely that can’t be!”  They think that there must be some way to avoid the contradiction, even if they are never able to say what it is, and always end up doing exactly what they claim to be avoiding, viz. making extra-scientific philosophical claims.  Paradoxically, the very obviousness and prevalence of the fallacy keeps them from seeing it.  As Orwell famously said, “to see what is in front of one's nose needs a constant struggle.”

Then there is the element of pride.  You have to be smart to do natural science.  Combox trolls usually are not very smart, but they think of themselves as smart, because they at least have the capacity to pepper their remarks with words like “physics,” “science,” “reason,” etc. as well as to rehearse whatever science trivia they picked up from Wikipedia.  So, suppose you are either a scientist or a combox troll who has gotten your head full of scientism.  You are convinced that philosophers and other non-scientists have nothing of interest to say.  Then one of them points out that you are committing a fallacy so simple that a child can see it.  That can be very hard to swallow.  And if the person pointing out the self-defeating character of scientism happens to be religious, the blow to one’s pride can be absolutely excruciating.  “Some religious nut is going to catch me out on a blatant fallacy?  No way in hell!  I refuse to believe it!”  One’s pride in one’s presumed superior rationality locks one into a deeply irrational frame of mind.

A third factor is that, though the fallacy is pretty simple, you have to have at least a rudimentary understanding of certain philosophical concepts – realism, instrumentalism, self-contradiction, etc. – and a basic willingness to think philosophically, in order to be able to see it.  Now, suppose you not only don’t know much about philosophy, but are positively contemptuous of it (as those beholden to scientism often are).  Then you are not going to know very much about it, and you are not likely to be able to think very clearly about even the little bit you do know.  Your prejudices keep getting in the way.  You are bound to be blind even to obvious fallacies like the one in question.

The bottom line is that if you cannot help doing philosophy – for again, the very act of denying that one needs to do it itself involves one in a philosophical commitment – but at the same time also refuse to do it, then you are inevitably both going to do it and do it badly.  

The clueless reactions I have seen to these simple points over the years only reinforce their validity.  For example, many defenders of scientism will, in response to the claim that extra-scientific philosophical commitments are unavoidable, demand that you produce an operational definition for this or that philosophical concept, or experimental evidence for this or that philosophical thesis – thereby adding begging the question to the list of fallacies of which they are guilty.  For of course, such demands presuppose the correctness of scientism, which is exactly what is at issue.

My favorite response is the suggestion that a philosopher who criticizes scientism has gotten too big for his britches.  “How dare you suggest that scientists don’t know everything!  How arrogant!”  Scientism, it seems, kills irony along with basic critical reasoning skills.

In his recent book Enlightenment Now, Steven Pinker summarizes some cognitive science research on bias, and notes that there is a special kind of bias to which those who detect bias in others are prone.  He calls it “bias bias” (p. 361).  The idea is that when you are keen to ferret out biases in others, you are often blind to the biases that influence you as you do so.  As Pinker also points out, people who are well-informed about a subject are also often prone to certain biases, precisely because the interest in the subject that leads them to learn a lot about it also makes it more difficult for them to be objective about it.  As Pinker writes:

[A] paradox of rationality is that expertise, brainpower, and conscious reasoning do not, by themselves, guarantee that thinkers will approach the truth.  On the contrary, they can be weapons for ever-more-ingenious rationalization.  (p. 359)

Pinker also judges, absolutely correctly in my view, that “the major enemy of reason in the public sphere today … is not ignorance, innumeracy, or cognitive biases, but politicization” (p. 371).  When you turn an idea into a political cause to promote, with allies to the cause needing to be recruited and enemies of the cause needing to be defeated, etc., then you are bound to let reason give way to rhetoric, to lose the capacity for dispassionate evaluation, and so forth.

These factors account for why defenders of scientism are often so dogmatic and nasty in their dealings with critics, often prone to ridicule and ad hominem attacks rather than the calm and rational discourse you’d think their purported commitment to reason and science would commend to them.  Scientism has become a political cause, and those beholden to it tend to delude themselves into thinking that their loud condemnations of cognitive bias and rationalization somehow make them immune to these very foibles.  There is no one in greater danger of irrational and unscientific thinking than the fanatic who screams “Reason!” and “Science!” in your face at the top of his lungs.

Scientism is, by the way, self-defeating in more than just the way already identified.  Consider that scientific methodology involves both the construction of mathematical representations of nature, and the experimental testing of those representations.  If you think carefully about either of these components – including even the second one – you will see that it cannot be correct to say that we can have no rationally justifiable belief in what cannot be experimentally tested.  

This is most obvious in the case of mathematics.  Even those beholden to scientism will typically admit that even those parts of mathematics that do not have application within empirical science constitute genuine bodies of knowledge.  And even the parts of mathematics that do have application within science operate in part by distinctively mathematical rules of reasoning rather than being evaluated solely by experimental testing.  

Now, defenders of scientism are often willing to expand their conception of what counts as “science” to include mathematics.  But there are two problems with this.  First, once they do this, then they can no longer consistently criticize philosophical claims for not being susceptible of experimental testing.  For their admission of mathematics into the fold concedes that there are rational forms of discourse that don’t involve empirical testability.  Second, the thesis that empirical science and mathematics exhaust the genuine forms of knowledge is not itself a proposition of either empirical science or mathematics.  Admitting mathematics into the science club simply does not suffice to save scientism from self-refutation.

Turn now to the notion of experimental testing.  Obviously, this presupposes that we have experiences.  Now, the fact that we have experiences, and certain very general features of experience, are themselves known through experience.  However, these particular facts are not susceptible of experimental testing.  The reason is that experimental testing – and in particular, the possibility of falsification – requires that experience can go in one direction or another.  We predict that it will go in direction A rather than B – that we will observe this rather than that – and then try to set up an experiment or observational scenario in which we can see whether this prediction pans out.

But not everything that is true of experience is testable in this way, not even in principle.  To take an example beloved of us Aristotelians, consider the proposition that change occurs.  We know this is true from experience.  But that does not mean that it is empirically testable in the sense of falsifiable.  It is not falsifiable.  For the very possibility of testability or falsifiability presupposes change.  You predict that you will have such-and-such an experience and see whether it happens, and that procedure itself involves change.  You go from thinking “Let’s see if this happens” to thinking “Ah, it did happen” or “Oh, it didn’t happen,” and either way a change will have occurred.  The thesis that change occurs is, accordingly, not falsifiable or empirically testable.  And yet we know it from experience, and the very possibility of empirical testing presupposes it.  Any appeal to empirical testability thus presupposes that we know at least some things that are not empirically testable (such as the reality of change).  Which is precisely what scientism denies.  Hence, once again, scientism is self-refuting.

Those beholden to scientism don’t see this because they conflate empirical with experimentally testable.  And these are not the same thing.  Again, the proposition that change occurs is empirical in the sense that we know it via experience, but it is not experimentally testable or falsifiable.  Aristotelian philosophers like Andrew van Melsen and Henry Koren characterize propositions like this as grounded in “pre-scientific experience.”  They are grounded in experience in the sense that we know them empirically rather than a priori.  They are pre-scientific in the sense that science involves empirical testability or falsifiability, and these propositions concern facts about experience that are deeper than, and presupposed by, anything testable or falsifiable.  

Hume’s Fork famously holds that all knowable propositions concern either matters of fact or relations of ideas.  The logical positivists drew a similar dichotomy between analytic and synthetic propositions, and contemporary naturalists often claim that all significant propositions concern either empirical science or conceptual analysis.  These are all variations on the same basic idea, and scientism typically appeals to one or another of them.  But as I have argued elsewhere, they are all self-refuting.  Hume’s Fork is not itself true either by virtue of relations of ideas or by virtue of matters of fact.  The positivist’s principle of verifiability is not itself either analytic or synthetic.  The naturalist’s dichotomy of empirical science and conceptual analysis is not itself knowable either by way of empirical science or conceptual analysis.  Like the adherent of scientism caught in his self-refutation, none of the adherents of these related views has much more to offer in response than a shit-eating grin.

Anyway, propositions of mathematics, propositions grounded in “pre-scientific experience,” and philosophical propositions (such as the thesis of scientism itself, which is philosophical rather than scientific) fall into a third (and indeed, perhaps a fourth, a fifth, etc.) category beyond the two that these self-defeating views are willing to recognize.

Metaphysics, as Gilson said, always buries its undertakers.   Or it would do so if those untertakers weren’t so busy burying themselves.