Debate over scientism often consists of critics arguing that certain areas of knowledge are beyond the domain of science. The realm of morals is a common example, as are ‘why’ questions and the supernatural. Philosopher Massimo Pigliucci can be relied upon to play the role of critic, for example writing:
[Richard] Dawkins and [Jerry] Coyne … insist in applying science to the supernatural, which is simply another form of the same malady that strikes [Sam] Harris: scientism, the idea that science can do everything and provides us with all the answers that are worth having.
This claim that science cannot deal with the supernatural is widely accepted even among scientists. For example, the website “Understanding Science” says in its introductory “What is science?” account:
Science cannot support or contradict the existence of supernatural entities. It deals only with natural phenomena and explanations.
The claim is particularly widespread in America, partly as a political tactic to avoid science appearing to contradict religion. By intentionally limiting science, the hope is to avoid a clash that might imperil support for science amongst a highly religiose populace. Any attempt by science to talk about the supernatural or gods is deemed ‘bad science’, and any attempt by religion to contradict factual scientific findings is labeled ‘bad religion’. Thus the American National Academy of Sciences declares:
Science is a way of knowing about the natural world. It is limited to explaining the natural world through natural causes. Science can say nothing about the supernatural. Whether God exists or not is a question about which science is neutral.
It also produces a religion-friendly booklet Science, Evolution, and Creationism that says:
Because they are not a part of nature, supernatural entities cannot be investigated by science.
In contrast, biologist Jerry Coyne has argued several times that science can test the supernatural; physicist Sean Carroll tends to agree, as does philosopher Russell Blackford, while skeptic Michael Shermer disagrees, saying that “Science operates in the natural, not the supernatural”.
So how can intellectuals of similar world view (everyone named above is an atheist) reach opposite conclusions on this point? The answer is that ‘supernatural’ is an ill-defined, colloquial word, and thus the disagreements amount to different interpretations of what the claim “science can/cannot deal with the supernatural” amounts to.
So what does ‘supernatural’ mean? Dictionaries are not much help, defining ‘supernatural’ as ‘not natural’, but giving little guidance on how to distinguish natural from supernatural. One possible answer is that natural things are those that do exist whereas supernatural things do not. However, those who believe in and postulate supernatural entities do regard them as existing, and deducing whether or not something exists should be the result of examining evidence, not an a priori declaration.
A similar, slightly facetious, definition of ‘supernatural’ would be ‘things that science cannot deal with’, but, again, this doesn’t help in establishing whether a particular postulated entity can or can’t be addressed by science.
To understand ‘supernatural’ we should realise that it is an ill-defined folk category of likely imagined beings that are usually envisaged as variations on natural beings. As emphasized by scholars of folk religion, such as Pascal Boyer in his Religion Explained, supernatural beings are not totally alien beings unrecognisable to us, they are like us, but with a limited number of differences. Thus ghosts can walk through walls and cannot die, but otherwise are recognisable as humans, with human-like attributes and behaviour.
Similarly, comic-strip heros, Superman, Spiderman and their ilk, are humans with a limited number of added super-attributes, and also some human weaknesses. As with all science fiction, it is their overlap with real humans that makes the stories interesting to us. Gods have added powers, just like comic-strip heros, but are human in their attitudes and motivations. Supreme gods might theoretically by omnipotent and omniscient, but they aren’t actually like that in the stories about them, they are beings with feelings and desires, beings you can talk to and make deals with; they are extrapolations of humans, disembodied tribal leaders with added powers.
Thus, supernatural entities may obey laws somewhat different from the natural laws, but they do obey laws, they exhibit regularities that are comprehensible, and have natures, ones that we humans can understand. That is the only reason they are interesting to us, the reason they are part of our mythology.
So what is it about the supernatural that might mean that science cannot deal with it? Well, non-existence, perhaps, but that conclusion needs to be the result of investigation, and we can only arrive there if science can investigate the supernatural. And I doubt that by “Whether God exists or not is a question about which science is neutral” the National Academy of Science meant that, of course, gods don’t actually exist.
As envisaged, supernatural beings can interact with the natural world (there would be little point in envisaging them if they couldn’t): ghosts can be seen, can appear in photographs and can cause the temperature to drop; sprites can cause good luck or bad luck; imps can mischievously cause things to go missing; demons can possess people; vampires kill people; gods can heal the sick, or cause your lottery numbers to come up, or determine the outcome of football games.
These interactions with the natural world mean that supernatural beings are entirely accessible to science. After all, that’s what science does, it investigates things that can have discernible effects on anything that science can measure. Very little about the natural world is observed directly, and most of the time science is observing things indirectly, through their effects on something else.
For example, science cannot observe a neutrino directly, but we deduce its nature from, say, its effect on an electron, with that electron then affecting other electrons in a liquid, which causes bubbles to form along the electron’s track in a bubble-chamber, and those bubbles can be detected by their interaction with photons, with those photons then interacting with photographic emulsion; and then, after the photograph is developed, other photons bounce off the photograph and into our eyes. Most physical material is only known by a highly indirect chain of effects on other particles that affect other particles that … eventually have some effect on our sense data.
Thus science pursues evidence trails, trails of evidence that follow back along such chains of causation, spreading out into all aspects of the observable world. Science doesn’t know or care whether some cause at the end of a chain is ‘natural’ or ‘supernatural’, all it cares about is that something is having a discernible effect on something else, and from that handle alone science can investigate its nature.
Thus science doesn’t rule anything out as an a priori assumption, and issues of whether an entity exists, or what attributes and powers it has (whether ‘natural’ or ‘supernatural’), are to be determined by scientific enquiry, by following trails of evidence wherever they lead.
The issue of whether intelligent agents (“gods”) are interfering in the world is no different in principle from whether chimpanzees or gorillas (or humans) are affecting the world. Yes, gods may be capricious and hard to understand, but then so are chimpanzees and humans!
The claim that ‘supernatural’ entities cannot be pursued by science because they obey no laws and no regularities is false because no supernatural beings are ever envisaged that way. Even if some cause were entirely random and exhibited no regularities, that could still be accommodated within science, as in, for example, the randomness currently thought to be inherent in radioactive decay and other processes governed by quantum mechanics.
The only ‘non-natural’ beings that science could not investigate are those which are postulated as being unable to have any effect at all, even in principle, on anything in the natural world. But that is not how supernatural beings are envisaged, and the possibility of such beings, entirely causally disconnected from our universe, is largely uninteresting.
Religious believers don’t conceive of gods as unable to communicate, or unable to help you, or to heal the sick, or to do anything at all. Indeed such helpless beings are the very opposite of the super-capable and even omnipotent beings of religious belief.
Thus believers are faced with a dilemma: either admit that your god is utterly impotent, totally uninteresting, and incapable of having any effect whatsoever on us or on our world, or admit that your god is a hypothesis about the real world that is just as amenable to scientific investigation as any other.
Even if the religious claim is that God doesn’t interfere in nature, but limits himself only to revealing knowledge in the mind of the believer (the “other way of knowing” or sensus divinitatis beloved of theologians), that is still an observable effect on the natural world — it must be, since our knowledge is manifest as physical patterns in our material brains, and the material state of our brains is in-principle observable. Further, any such revealed knowledge could have clearly observable consequences for that person’s real-world behaviour.
Thus, if you want to place your god beyond the domain of science, you have to accept that you can have no knowledge whatsoever of this god.
From the point of view of the scientist, we should assert that ‘supernatural’ claims are not a priori beyond the scope of science. Anything envisaged as having a possible effect on anything at all in the natural world — and supernatural claims always are envisaged that way — is within the proper domain of science.
Whether such entities exist is then to be determined by science (aka evidence), with the burden of proof being, of course, on the person claiming that the supernatural entity exists. Trying to claim that science is not the right tool to answer that question is really just an admission that you have no good evidence.