Testing Supernatural Hypotheses
Are supernatural hypotheses testable? This was a question raised by Victor Reppert on his Dangerous Idea blog, and I responded a couple of weeks ago. The subsequent discussion got rather confused and confusing, and I would like to try to untangle things and see if I can get clearer on the issue here. Thanks much to interlocutor Richard Wein whose gentle but persistent probing of my remarks has pressed me to try to state things more precisely and cogently.
What is a supernatural hypothesis? I will limit attention to hypotheses that postulate the existence of supernatural persons or powers. Instances of supernatural persons would include gods, ghosts, demons, angels, spirits (like Ariel), and souls. Instances of supernatural powers would include mana, qi, astrological influences, telekinesis, ESP, and the creative power attributed to God in Genesis where God says “Let there be…” and there is. But what is it for a person or power to be supernatural? By “supernatural” I mean “capable of existing or operating independently of, unrestrained by, or even in violation of, the laws of nature.” Thus, Marley’s ghost passed through Scrooge’s locked door and Christ’s miracle of the loaves and fishes violated the conservation of matter.
Now defenders of hypotheses postulating some of these things deny that they are offering claims that are supernatural; rather, they think that future physics will explain such (purported) phenomena as telekinesis and telepathy. However, our only possible evidence for what physics might someday accommodate is what it currently accommodates, and there is nothing in current physics to offer hope that such purported phenomena might someday be subsumed under physical law. In the ‘70’s New Age types blithely invoked quantum mechanics in hopes that quantum weirdness would somehow make ESP or telekinesis more plausible, but in so doing they only succeeded in demonstrating their own ignorance. So, I will consider such hypotheses to be asserting the existence of genuinely supernatural powers.
So, are supernatural hypotheses as characterized above testable? What do we mean by a “testable?” I mean “testable” in the rather strict sense of “confirmable or disconfirmable by rigorous experiment, experiment of the sort typically employed to evaluate hypotheses in the physical and biological sciences.” The history of science offers a number of notable examples of the robust confirmation or disconfirmation of hypotheses by such experiments:
1) Thomas Young demonstrated the wave nature of light by sending two light beams through narrow orifices so that the beams would spread, overlap and be projected on a screen. In accordance with the hypothesis that light consists of waves, and contrary to the hypothesis that it consists of particles, Young showed that the area of overlap between the projections of the beams showed striped interference patterns. (QM, of course, now accounts for the fact that light can behave either as a wave or a particle.)
2) Antoine Lavoisier demonstrated that the calcination of mercury absorbs the “respirable” element of the air (which we know as oxygen) by confining a quantity of common, respirable air under a sealed glass dome with a sample of very pure mercury and heating the mercury until it was calcined. The air remaining in the sealed dome was demonstrated to be non-respirable. However, when the calcined mercury was reduced again to pure mercury, nearly the same amount of air was recovered as had been absorbed when the mercury was calcined. When this recovered air was recombined with the air that had been rendered non-respirable, it was returned to its common, respirable state.
3) Louis Pasteur kept an anthrax culture at the temperature of 42 to 44 degrees Celsius for eight days in hopes that this would attenuate the virulence of the bacillus. In a famous public demonstration he vaccinated twenty four sheep, one goat, and six cows with the attenuated strain. Twenty nine animals were unvaccinated as a control. A few weeks later he injected all of the animals with a fully virulent anthrax culture. Three days later all of the vaccinated animals were healthy, and all of the unvaccinated ones were dead or very ill.
So, are supernatural hypotheses testable in a comparable manner, i.e., with similar rigor and conclusiveness? In many cases they are! Indeed, a paradigm for such a test is given in scripture, the famous encounter between Elijah and the priests of Baal recounted in I Kings, chapter 18. Elijah and the priests of Baal build altars to present a sacrifice to their respective deities, Yahweh and Baal. First the priests of Baal call upon their god to send fire to consume their sacrifice. After importuning all day, no fire is sent. Elijah, to make his demonstration more dramatic, has his sacrifice soaked with water three times. Elijah calls upon The Lord, and fire falls upon the sacrifice and consumes it. The people acknowledge that The Lord is God, and join Elijah in a celebratory massacre of the priests of Baal.
The contest between Elijah and the priests of Baal is as elegantly designed an example of an experimentum crucis as one could ask for. Why don’t we see more such tests? It would take only a few such demonstrations, which could now be broadcast worldwide, to convince most of the unbelievers of the world (at least, if adequate safeguards against trickery could be assured). Since experimental demonstrations are the epistemic gold standard when it comes to the evaluation of hypotheses, we would expect, prima facie, that defenders of supernatural hypotheses would employ such tests whenever feasible.
When we examine the actual history of efforts to subject supernatural hypotheses to experimental tests, however, we find that they frequently elude such testing. One obvious problem is that many hypotheses postulate the existence of supernatural persons who may choose to cooperate with our testing procedures or not. How can we test claims about ghosts or gods if they simply refuse to be tested? There seems to be no way to compel the cooperation of such beings, and there appear to be no discernable “laws of supernature” that license precise predictions about when and where such beings may be expected to act in ways that would reveal their existence. True, the shrine at Lourdes has many discarded crutches (but no dis
carded artificial limbs, as one wag pointed out). Yet, there seems to be no place where one might go to reasonably expect to see a miracle (a real miracle, not “the miracle of birth” or something like that). Worse, scripture actually seems to forbid putting God to the test (Deuteronomy 6:16; repeated by Jesus in the Gospels at Luke 4:12 and Matthew, 4:7), which seems to imply, or is often taken to imply, that it is sacrilegious even to attempt an experimental verification of God’s existence.
When we look at supernatural claims, they are definitely a mixed bag with respect to testability. Some do not even appear testable even in principle (e.g., very vague hypotheses such as that carrying a four-leaf clover is “lucky”). Others, while in principle testable, are presented in conjunction with auxiliary hypotheses that have the effect of preventing any actual testing. Others are testable and have been tested. Examples of this last category are many of the claims of parapsychology. There have been numerous well-designed, rigorous, and controlled experiments testing claims of extrasensory perception, clairvoyance, or telekinesis. (Unfortunately, some of the best known such experiments, such as those of J.B. Rhine at Duke University, were not properly designed so as to preclude cheating). The problem is that such tests, when properly conducted, consistently give results that are negative, insignificant, or irreproducible. (See, for instance, Terence Hines’ extensive review of laboratory parapsychology up to the late ‘80’s in Pseudoscience and the Paranormal, Prometheus Books, 1988).
The most interesting supernatural hypotheses are those that are can be tested, but, for some reason or another, always seem to elude actual testing. Consider the theistic hypothesis, the hypothesis that the God of theism exists. This hypothesis can be tested, and, as we noted above, according to scripture has been tested in the past—with spectacularly positive results. The problem, of course, is that all those alleged public demonstrations of divine power occurred long, long ago, in what Hume called “ignorant and barbarous nations.” In short, it is eminently reasonable for the skeptic simply to deny that such events ever occurred. What we need, then, is something now, something very public and conclusive. As I say, an Elijah-like test could be broadcast worldwide now. Or, if such a display is considered vulgar, there could be rigorous, reproducible results performed in a scientific setting and verified by the qualified parties. So why not?
Now, some theistic philosophers, such as Victor Reppert, say that the theistic hypothesis is testable, and, I surmise, they hold that it passes the tests. However, as far as I can tell, what Victor means by “testable” is something much looser and less stringent than I do. By “testable” he seems to mean something like “rationally discussable vis-à-vis the evidence.” By “testable” I mean “capable of being subjected to rigorous experimental verification of the sort exemplified in the above three examples.” Now I have no doubt that the theistic hypothesis is rationally discussable in the light of evidence, but I think it is an innocuous and unobjectionable observation to note that recent theistic philosophers and apologists have not invoked rigorous experimental tests to support their hypothesis. Their failure to do so is not per se a grievous fault. Very many important questions can be settled, even settled decisively without appeal to performed experiment. Indeed, scientific fields such as paleontology, archaeology, and anthropology incorporate non-experimental, but still rigorous, methodologies.
Still, nothing persuades like a good experimental demonstration. Pasteur’s anthrax vaccine experiment was so dramatic that some recent critics such as Bruno Latour have accused Pasteur of being a showboat and a meretricious self-promoter. (However, my considered professional opinion, as I have argued in two or three books, is that Bruno Latour is full of shit.) So, where are our latter-day Elijahs? Of course many will piously invoke the scriptural prohibition against putting The Lord to the test. However, such a proscription did not deter Elijah, and, frankly, and I hope not too cynically, I suspect that today’s defenders of theistic hypotheses could likewise put aside their qualms if they were as confident of results as Elijah was.
Perhaps, though, God has good reasons not to permit his existence to be demonstrated in some conclusive, unambiguous way. Such a claim, added as an auxiliary hypothesis to the theistic hypothesis, would make theism untestable. Such a claim would tell us that God, for good reasons, will not cooperate with any attempted experimental verification of his existence. But what why would God not want his existence made plain to all? Two reasons that sound plausible are these:
1) If God were to make his existence indubitable then there would be no room for faith. God does not want us merely to acknowledge his existence, as we would acknowledge the existence of gravity or neutrons, but to live in a trusting, faithful, loving relationship with him. Faith involves an element of will or choice; we commit ourselves despite doubt. True trust involves risk, and there is no risk in acknowledging the obvious.
2) If God’s existence were plain to all, then the human capacity for goodness would be compromised. Knowing with certainty that God exists, people would abstain from sin out of prudence, not for genuinely moral reasons. Nobody will steal hubcaps with a cop standing right there. Only if God maintains some degree of “hiddenness,” i.e epistemic distance, will we be called upon to struggle with temptation and resist evil because it is the right thing to do rather than because we want to avoid punishment.
Are these good reasons for thinking that God would not want his existence experimentally demonstrated? I do not think so. First, many theists do in fact say that God’s existence is plain and indubitable. Most famously, Paul, in the first chapter of Romans says:
For the wrath of God is revealed from heaven against all ungodliness and wickedness of men who by their wickedness suppress the truth. For what can be known about God is plain to them, because God has shown it to them. Ever since the creation of the world his invisible nature, namely, his eternal power and deity, has been clearly perceived in the things that have been made. So they are without excuse; for although they knew God they did not honor him as God or give thanks to him but they became futile in their thinking and their senseless minds were d
arkened (Romans I: 18-21).
For Paul and many other eminent Christian thinkers, therefore, the purported obviousness of God’s existence cannot be an impediment to faith. Further, unbelievers reject God despite the obviousness of his existence. Advocates of above reason (1) therefore need to argue with Paul, not me.
Further, experimental confirmation of hypotheses need not rule out doubt. I personally would believe in God if there were as much experimental evidence for God as there is for, say, protons. I am a scientific realist. I think experimental evidence strongly supports the existence of protons, neutrons, and even neutrinos. On the other hand, many philosophers of science advocate various forms of antirealism whereby experiments confirming hypotheses about theoretical objects merely show that those hypotheses are empirically adequate, i.e., that their predictions are reliable, and not that the theoretical entities they postulate actually exist. Perhaps experimental confirmation of the God hypothesis would permit a similar antirealist attitude. Such hypotheses could be taken as empirically adequate, but not as demonstrating the existence of a theoretical entity, i.e., God.
Most fundamentally, merely knowing that God exists need not abrogate or compromise a relationship of trust or faith with him. Marriage is a risky commitment of trust and faith though, supposedly, something one knows for certain is that one’s spouse exists. Acknowledging God’s existence does not entail being in a trusting, loving relationship with him. As scripture somewhere says, even the devils know that God exits and tremble.
As for morality being compromised if God’s existence were plain, I see no evidence for this at all. Moral enormities are constantly committed by people who have no doubts whatsoever about God’s existence. Indeed, they often invoke God as the justification for their crimes. Besides, hell is always for the other guy—atheists, gays, liberals, evolutionists, infidels, feminists, Democrats, etc.—people like that. Hell is never for YOU. So I see the above reason (2) as having no justification at all.
Are there other, better reasons why God would not want his existence experimentally demonstrated? I don’t know, and will leave that dear (and patient) reader for you to tell me. If no such good reasons can be found, what do we conclude? In practice, when defenders of a hypothesis shield it from rigorous testing by invoking arbitrary and ad hoc auxiliary hypotheses, that insulated hypothesis is regarded as discredited. Are there creditable reasons for not experimentally testing theism?