TIA on falsifiabilityFebruary 11, 2008 — Deacon Duncan
I’m up to chapter two in Vox Day’s e-book, The Irrational Atheist. This chapter deals with defining science, how to distinguish science from non-science, and the relationship between science and religion. After some introductory material, we find this little gem:
The need to separate real science from non-science can also be seen in the way that the phrase “studies show” has become a secular form of making a vow, a useful means of reassuring the skeptical listener that the speaker is swearing to the truth of his words despite any doubts that the listener might harbor.
Rather ironic, considering that Vox just got done claiming that “Low Church Atheists” are worse off than believers because “[s]tudies have shown that those without religion have life expectancies seven years shorter than the average churchgoer, are more likely to smoke, abuse alcohol,” etc., etc. But where he really gets himself into trouble is when he tries to tackle the notion of what falsifiability is, and what it means for science.
For a hypothesis to be falsifiable, it must be theoretically possible to make an observation that would disprove the subject… But can Popper’s concept of falsifiability really be taken seriously as a dividing point between science and not-science? It appears more than a little flawed to me.
Vox has a point. Falsifiability, defined in those terms, is flawed. That definition, however, does not quite capture the essence of what falsifiability is, and why it is important to science. Remember, science isn’t just some arbitrary set of rules that some guys just got together and made up, the way someone once invented the rules for Texas Hold ‘Em. Science is a process of discovery, including a discovery of which scientific principles actually function effectively and reliably as a means of distinguishing between truth and error. Finding a flaw in one man’s description of falsifiability is not the same as finding a problem with falsifiability itself.
Before we get into falsifiability, let’s look at Vox’s examples. He begins by referring to “All swans are white” as an example of a falsifiable statement, and then applies that to a parallel phrasing, “All gods speak Aramaic.” Both are statements that are at least theoretically falsifiable, in that they meet Vox’s criterion that it is theoretically possible that one could make an observation which would disprove the subject. But are they both falsifiable hypotheses?
It seems Vox has fallen into the converse fallacy once again. All scientific hypotheses must be falsifiable, therefore Vox takes all falsifiable statements as scientific hypotheses. But that’s not the case in real life. Falsifiability is a necessary condition for a scientific hypothesis, but not a sufficient condition. All Vox is doing is showing that there are some statements which are hypothetically falsifiable without being actual scientific statements.
Falsifiability is quite simply the requirement that you must state your hypothesis in such a way that it is possible to reliably and objectively look at the evidence and determine whether or not the evidence is consistent with your hypothesis. To be scientific, in other words, there must be some objective way to compare your hypothesis with actual, relevant evidence, to determine whether your hypothesis is right or wrong. It’s not really a hard concept to grasp, and it explains Vox’s examples pretty easily enough: the statement “All gods speak Aramaic” is hypothetically falsifiable, but it’s not actually falsifiable because there is no actual evidence against which the statement can be compared. It’s like the statement “All unicorns have white horns.” Neither gods nor unicorns actually show up in the real world, and therefore there is no realistic possibility of evaluating either claim scientifically.
It’s pretty significant that Vox would begin his chapter on science by attacking falsifiability, since the principle of falsifiability lies at the heart of science’s ability to distinguish between correct answers and incorrect ones. Later on in Chapter 2, Vox is going to deny that anyone is trying to interfere with business as usual in the science department, and yet here he is, right up front, trying to sabotage the idea that science needs to be able to distinguish between right and wrong answers. Without the requirement that all theories must be falsifiable, without requiring that science be able to tell when a hypothesis is wrong, science becomes mere superstitious politics. Anything you see, you can ascribe to any cause you like, and nobody can ever prove you wrong—as long as you can get a lay majority to support your interpretation anyway.
This is indeed the danger posed by religious challenges to science. Real-world facts are not consistent with what the Gospel claims about God, and so in order to protect their dogma, Christians need to either become liberals and spiritualize everything, or else castrate science so that it no longer has the ability to recognize and reject false conclusions. Only then can conservative Christians console themselves with the thought that “science” cannot prove them wrong.