Header 1

Our future, our universe, and other weighty topics


Friday, April 15, 2016

Fanboy Flub: He's Way Wrong About Scientific Theories

In the New York Times a few days ago there was an article by Carl Zimmer in the Science section. The article was entitled In Science, It's Never “Just a Theory.” Zimmer asserts the following:

Theories are neither hunches nor guesses. They are the crown jewels of science.

Later Zimmer makes this assertion:

A theory, likewise, represents a territory of science. Instead of rivers, hills, and towns, the pieces of the territory are facts.

So generally speaking, scientific theories are made up of facts, and no guesses? Hogwash. Rubbish. Zimmer is apparently a science theory fanboy who we can expect to receive a scientific theory with the same reverent enthusiasm that a trekkie might receive the next installment of the Star Trek franchise. (The Merriam-Webster online dictionary defines a fanboy as “a boy or man who is an extremely or overly enthusiastic fan of someone or something.”) Let's look at some things that conflict with this naive attitude toward scientific theories.

Consider some of the weird scientific theories that are floating about these days. There is the Everett “many worlds” theory of parallel universes, which imagines that the universe is constantly splitting up into different copies of itself, and that there are a zillion copies of you in different alternate universes, where everything imaginable occurs. Then there's the Integrated Information Theory which tries to explain consciousness in a way that implies that your thermostat and cell phone are partially conscious. Then there's the so-called “landscape” version of string theory, which unnecessarily imagines something like ten to the five hundredth power universes, each of which is a different permutation of string theory (a speculative branch of physics that is entirely unproven). Then there's Smolin's theory of cosmological natural selection, which gives us the groundless speculation that certain type of stars spit out new universes when they die. Then there's the panspermia theory that life began on Earth because aliens planted it here. Then there is Freud's theory that you have deep-rooted conflicts caused by your subconscious sexual attraction to your mother. Then there are neurological theories that your self is just an illusion.

Zimmer would have us believe that scientific theories are in general “crown jewels,” but it seems that some scientific theories are more like junk than jewels.

But isn't it at least correct that most scientific theories are true? No, it isn't. Scientists love to churn out speculative theories, particularly physicists and cosmologists. There are countless different versions of what is called the cosmic inflation theory, each making different assumptions. Presumably no more than one of these different versions can be true. There are countless other scientific theories describing different variations of the Big Bang, different configurations of our universe, and different eventual fates for our universe. Probably the great majority of these are false. There are countless different attempts to describe a physics beyond the Standard Model. The great majority of these theories must be wrong, as they all make assumptions that conflict with each other. A similar thing probably holds true about theories in fields such as neurology and psychology.

I may note that there is absolutely no ethic in science that a scientist should regard a theory as being probably true before he publishes a paper suggesting it. Quite to the contrary, there is a “publish or perish” ethic under which it is considered just fine to publish a paper suggesting some oddball theory that probably isn't true, particularly if it is mathematically interesting or is an interesting twist on some other theory or if it is an idea that somehow might cause someone to think of a theory that is probably true. When Harvard professor Abraham Loeb came up with an extremely implausible but interesting and innovative theory about life evolving in the first few million years of the universe's history, his colleagues probably thought it was nice work.

But is it true that we can tell that some scientific idea is solid when it is called a theory? No, it isn't. The claim that scientists only use the word “theory” for well-established ideas with lots of evidence behind them is pure bunk, a groundless piece of folklore typically trotted out when making the case for some theory that has insufficient evidence behind it. What proves this is that the completely unsubstantiated set of ideas called “string theory” is referred to under just that name by scientists and non-scientists, not some other name such as “string hypothesis.” Scientists pretty much use the word “theory” just like ordinary people use it.

But is it true, as Zimmer claims, that in the world of science “theories are neither hunches nor guesses”? As a general claim about scientific theories, that is false. A sizable fraction of modern scientific theories are highly speculative theories largely consisting of guesses. In fact, many scientific theories are guesses piled up on top of other guesses. When the cosmic inflation theory was introduced, it was a wild guess about details of the first second of the universe, which was built upon other wild guesses about the universe (a class of speculative theories called grand unification theories). Neither of these wild guesses has been confirmed, but the cosmic inflation theory is still influential. Heaping speculations upon speculations is typical behavior for the modern theoretical physicist.

Zimmer's generalizations about scientific theories are about as accurate as generalizations such as, “You get quick service when you phone technical support,” and “Middle age men are trim and muscular.” Zimmer claims that in science, it's never “just a theory,” but the truth is in science, it's very, very frequently “just a theory.”

Do any of the theories of science deserve to be called “crown jewels,” the term so carelessly used by Zimmer as a general characterization of scientific theories? Yes, but only a small few. The scientific theories that are “crown jewels” are mainly those that make exact numerical predictions that have been repeatedly proven to be true. Some examples I can think of are the theory of gravitation, the theory of electromagnetism, and the kinetic theory of matter (particularly when it predicts the behavior of gases).

The table below lists some different categories of scientific truth-claims, distinguishing between well-supported claims and claims that are not well-supported. Only the tiniest fraction of scientific theories fall into the first category. What can be particularly confusing is that a particular set of ideas called a scientific theory may be what we can call a composite theory, meaning a theory consisting of different types of truth claims, some being well-supported and others being not well-supported. 

scientific theories

Given such composite theories, it may be quite hard to sort out whether some, all or most claims in a scientific theory should be believed. There is no reliable shortcut to evaluating most scientific theories. You cannot simply rely on a scientific consensus, because there may be strong sociological, ideological and “bandwagon” factors that may cause a majority of scientists in some field to endorse a theory which is not actually worthy of belief, or that contains some unworthy parts. There is no substitute for rigorously analyzing the individual truth claims that make up a scientific theory, and evaluating each such claim in a critical light, demanding supporting evidence for each claim, listening to opposing arguments, and rejecting any part of a theory which is not well-supported by facts or evidence.

It's okay to be a nature fanboy, but don't be a science theory fanboy who reverently receives the latest dubious theoretical claim floating on gossamer threads of thought. Be someone who analyzes scientific theories in the same probing, questioning way that you might analyze the truth claims of a salesman in a used car lot.