What Do You Mean by "Supernatural"?

Q: kentlyon asked: “What Do you Mean by Supernatural?”

A: My notion of “natural” comes from what I know of science. Since science progresses, that is a moving target, but I think the idea of regular laws will continue to be central to science. A claimed phenomenon is being treated as “supernatural” if there is no serious attempt to subject it to scientific investigation in order to establish its existence and explain it by the principles of regular science. On the other hand, if something exotic–say a claim of mind reading–is subjected to scientific investigation, it will either be (a) drawn within the orbit of science and what we count as “natural,” (b) shown to be a hoax, or © investigated by “science” done so badly that we can justifiably call it pseudo-science.


  1. Very often, how statistical issues are handled makes the difference between good science and bad science. For example, people often point to selected anecdotes as persuasive evidence even in situations where large-sample, even-handed evidence is, in principle, available. 
  2. If someone makes a claim that would require new physics, the one making the claim should be making an attempt to convince the physicists. More generally, the immediate objections of the relevant experts should be registered alongside whatever claim is made.
  3. In cases where a scientific discipline itself is off target, it may be necessary to find a jury for a claim made up of people capable of understanding a scientific discipline who are enough on the outside that they can have some impartiality. Typically, it will be best to choose such a jury for judging an entire discipline as a panel of academics in other disciplines. 
  4. Though particular scientific disciplines may be badly off target, I disagree with any claim that science and academic inquiry as a whole are off target.
  5. Scientific disciplines such as economics often have only weak evidence one way or another on questions people care a great deal about. In that case, it is important to distinguish between scientific statements that are backed by a large, informative body of evidence, and scientific statements that are simply the best one can make of grossly inadequate evidence. In this, I want to emphasize that evidence is often grossly inadequate even after valiant efforts to gather as much evidence as possible. Scientists in disciplines that often have grossly inadequate evidence sometimes forget to remind each other just how uncertain the answers to the questions they are debating really are, given the inadequacy of evidence.
  6. One should be especially slow to accept a claim that requires violating a principle of a scientific discipline such as physics where most statements are backed by a large amount of evidence. 
  7. To the extent that there is no conceivable way to test a statement, then it is neither natural nor supernatural. (Indeed, it may not be a “phenomenon.”) Following Wittgenstein, let’s call such a statement “metaphysical.” Metaphysical statements can be important as framing, but they should not be confused with claims of the supernatural. Often, a way to quickly tell the difference between a supernatural claim and a metaphysical claim is that a supernatural claim usually has a hint of something that (if true) would seem like genuine magic (at least outside of religious contexts that desensitize people to the magical element), while a metaphysical claim points to at best metaphorical magic–which isn’t really magic at all.  
  8. Just because a claim is extraordinary doesn’t make it a supernatural claim. For example, take the claim that Earth might be observed by intelligent aliens from listening posts inside our own Solar System, and that these aliens make intentional efforts to hide their presence. If the claim is that these aliens came to our Solar System in at only a small fraction of the speed of light, this does not violate any laws of physics. There is no particular implausibility to the idea that intelligent aliens would be interested in studying us scientifically. (Presumably, some species of intelligent aliens would be interested in such a scientific investigation, even if others wouldn’t.) And there is no implausibility to the idea that they might want scientific data about our behavior that was not contaminated by awareness of their existence.
  9. To give another example, there are conspiracy theories that are very implausible, but do not violate the laws of physics. I would be inclined to call these claims natural. That tells me that in my distinction between “natural” and “supernatural,” I am privileging my understanding of the laws of physics over my understanding of social science. This is in line with the phrase “the natural sciences.” It is my impression that pretty much all the principles of the natural sciences can be derived from the principles of physics and basic logic. (Leaving aside any broader claims of evolution, the narrow version of the principle of natural selection is just a matter of the logic of selection–it is like counting.)

Update. This afternoon, Noah Smith tweeted links to two articles that are wonderful on exactly the issue here: