To readers who prefer short posts, I’d like to apologise in advance: this is not a short post. Unforunately, the nature of this extended argument is such that there’s no easy way to break it into 2 or 3 posts without killing the flow.
In a discussion I’ve been having recently with one particular believer, some ideas have repeatedly surfaced. This is not, however, the first time I’ve come across these particular notions. I want to take some time to fully address these ideas and the problems that are imbedded in the ideas. First I’ll simply quote the statements as written, as a group, break down the problematic/vague parts, then address them individually.
- “Can one expect human logic to understand the supernatural realm as easily as it does the natural realm?”
- “Are you saying that you reject the existence of the supernatural because people around you can’t agree on the exact nature of God, or of the Creative miracles?”
- “My fear, for those who choose that route, is that due to the acceptance of methodological naturalism as the defining limitation to science (defining only what can be proven from within nature itself), those that limit themselves in this way and trust that nature itself is “all there is”, will never have the chance to find out if the supernatural actually exists.”
- “Does methodological naturalism include or exclude God?”
The first question one needs to ask when presented with these kinds of statements is a simple one: what the hell are we talking about? There are a bunch of words in here that seem to be doing a lot of work, that refer to some sort of ontology, yet these words are simply assumed to A) have a coherent meaning and B) refer to something that exists, or they (in the case of Methodological Naturalism) seem to imply a misunderstanding of the expression.
The terms that are problematic are:
1. The supernatural
2. Human logic
4. Methodological Naturalism
In order to begin to address the initial statements, we need to some sort of working definition for these terms. While many theists may be operating under their own particular version of these words, I’m going to do my best to work with the mainstream usages. When dealing with this kind of conversation ‘live’, as it were, you should try to get the theist to nail down exactly what they are talking about. Typically, and unfortunately, the theist usually has no clear idea what they mean by these terms, which allows these terms to tie together all sorts of nonsensical arguments.
1. The Supernatural
The best argument in favour of the Supernatural was essentially put forward by Descartes. The concept of the ‘supernatural realm’ is simply an instantiation of the argument in favour of Dualism. The general argument for Dualism is as follows:
1. Not everything that we have witnessed has known causes.
2. The devices we use to detect causes only detect physical causes.
3. These causes cannot be detected by devices designed to detect physical things.
4. Therefore these causes are non-physical
5. Everything in this realm is physical.
6. Therefore there must be a 2nd realm which is non-physical in nature.(e.g. The mind, the soul, ghosts, faeries, god)
This argument is obviously flawed, though Descartes didn’t see the flaws immediately. Here’s the crux of the problem: if the second realm is non-physical, then how does it interact with the physical? How does a non-physical ‘thing’ affect a physical ‘thing’? This seems problematic.
A second (deeper) problem: the devices that we use to study the world are, indeed, restricted to directly observing the physical world. They are, as is correctly pointed out, not placed to directly observe non-physical ‘things’. However, they are perfectly placed to indirectly observe non-physical ‘things’: by observing the effect of the non-physical ‘things’ on physical ‘things’.
Once some sort of event/effect has been spotted, our devices can be used to trace the cause. If the cause is non-physical, we should come up with an apparently uncaused event. That is to say that there are zero antecedent physical events to the physical event being observed. For an example, with regards to ghosts, if a ghost is claimed to often move a table in a particular room, then when the movement of the table is observed, there should be no way to explain the movement of the table by reference to observed physical factors. Historically, however, all these events have been explainable (and I don’t just mean by a nice story) by observed physical factors.
To move into the area of faith: if someone can heal the sick, it should be possible to find people who are sick, put them in the presence of the healer, and then watch the illness mysteriously vanish. When we have attempted this, we have seen people die from lack of medical treatment, but no-one has been healed in accordance with the claims of the various healers. In many cases, the healers have attempted to excuse their failures by (in essence) declaring that their ability is highly unreliable, while claiming to the believers that their ability is highly reliable. Parsimony offers a much simpler answer: liars lie.
Furthermore, there are many events in the physical realm that are highly rare and extremely difficult to predict and/or detect. Nonetheless, we have developed tools that allow us predict and detect these events. That a non-physical ‘thing’ is difficult to predict or detect does not mean that it cannot be predicted or detected. Defining a non-physical ‘thing’ as impossible to predict and detect raises the question of how the theist detected the non-physical ‘thing’ in the first place.
In short, anyone arguing that ‘the supernatural’ must be taken account of needs, before anything else is done, to explain 1) what they mean by ‘the supernatural’, 2) why it can’t be tested either directly or indirectly and 3) if detection is impossible, how they know of its existence.
There is a second way to define ‘supernatural’. It’s something of a linguistic trick, and a lot more sparse than the above philosophical presentation. Quite simply the ‘super’ means “above or outside of”, so ‘supernatural’ means “above or outside of the natural”.
‘Bingo,’ say the theists, ‘this is the definition we want to use’. But this is an empty definition; it is entirely meaningless.
‘The natural’ is the known. Anything we can point to, that we can explain, that we can physically hold up and ask questions about: everyone agrees that this is ‘the natural’. So what is ‘the supernatural’ on this definition?
Simply those things which we don’t understand. But… not quite. I think it’s reasonably fair to assert that we (or, at least, the vast majority of us) don’t understand quantum mechanics (QM), but it would be improper to place QM within the realm of ‘the supernatural’. Because we can define QM, we can lay out the specifics of where and when it’s relevant, so this is not part of ‘the supernatural’. How about the soul? Definitely part of ‘the supernatural’.
If we follow this kind of categorization, it quickly becomes clear that ‘the supernatural’ refers to ‘words that are not defined, but are used to explain things’. The use of the expression ‘the supernatural’ is a purely epistemic statement: it refers to our lack of knowledge about causes.
‘The supernatural’ is not a metaphysical statement (a statement about an actual object or thing): such a statement requires some kind of definition, an explanation of ‘what the thing is’, but most theists will do their best to push back against such an explanation arguing that ‘no-one can know god’, relegating ‘the supernatural’ to refer merely to “I don’t know what”.
In short, on this second empty definition, if someone says “it was caused by the supernatural”, the sentence is functionally the same as “it was caused by I don’t know what”, but the speaker confuses the this for a knowledge claim. At absolute best, making this statement puts the theist in the position of putting forward Premise 1 of Descartes’s argument, which is subsequently doomed.
2. Human Logic
As someone trained in Philosophy, this one causes me to grind my teeth.
The function of the word ‘human’ in this statement is merely pejorative. It exists to cast aspersions on ‘human thought’, and ‘human thinking’, to emphasize the idea that ‘well, human logic is limited, but the logic of someone else… that’s not so limited’.
In all cases, this is operating on the vernacular usage of ‘logic’ which is radically different from Logic. Logic, in philosophy, is a well-defined science (in the original mathematical sense, not the laboratory sense) about the construction and analysis of statements and how those statements fit together to form arguments.
Logic, in the vernacular, is meaningless babble. Often people will say ‘your logic is faulty’ when really they mean “I don’t like what you said”. Take the following:
1. All clowns are made of porridge
2. Sarah is a clown
C. Sarah is made of porridge
The Logic in that set of statements is air-tight, and irrefutable. The statements themselves may be gibberish, but the argument is Valid (technical term). The argument is not Sound (technical term), because Premise 1 is not Good (technical term, but the meaning should be clear).
In short, ‘human logic’ is used primarily in a pejorative sense to indicate mere disagreement with a speaker.
Furthermore: what other logic is there? I am unaware of dolphin logic, or cat logic, and the rudiments of worm logic escape me. The term ‘human logic’ implies that the speaker is aware of another system of logic that is superior (in some way that is self-evident to the speaker) to mere ‘human logic’: the onus is on them, in that case, to sketch out this superior logic.
Traditionally, however, this argument is just a cover for “you have disagreed with my scripture, therefore you are wrong. God can do anything they like, so neener neener neener”. The confusion lies in that the critic is often raising a problem with the internal logic of the human-written words that the theist is attempting to foist on other humans. Even assuming for the sake of argument that there exists a god that has some superior ultra-Logic, this has nothing to do with the point being raised: that human-written things are subject to ‘human logic’, and that humans need convincing through ‘human logic’. To simply say “I know that what I said is contradictory, but that’s only if you consider human logic” is to (frankly) babble incoherently as the theist is, themselves, a human and has no ‘other’ logic to refer to.
Here I’m going to sketch out David Hume’s argument against believing in miracles, drawn from “An Enquiry Concerning Human Understanding”, and extend it slightly. If you haven’t read this text, you can get it for free through Project Gutenburg.
Imagine, if you will, that whilst happily sitting in a cafe, a man ran into the cafe and declared that he had just seen a person struck by a truck. The death of the person was apparent to everyone on the scene as the person was decapitated in a truly grisly fashion. Minutes later, however, the person got up and walked away. It’s a miracle!
On the face of it, is it more likely that a miracle actually happened, or that there is some other explanation for the situation? The options are really limited to: everything happened the way that the man said that they did, or the man is intentionally lying to us, or the man is mistaken (due to a wide range of possible causes). In the history of the world, people lying to us or being mistaken are common, whereas the actual number of bona fide miracles that have stood up to scrutiny is zero. As we are presented with this testimony, we have no choice but to accept it or to not accept it.
As such, Hume contends, with no information beyond testimony, we should tentatively accept the interpretation that is most probable given the past: the person is either mistaken or lying. What reason could we have for believing that someone who has been killed came back to life, an event that defies everything that we know about how biology works?
Stepping a little beyond Hume (who didn’t say anything more for fear of the compassion of Christians), Christianity puts us in the same position: we have a book that makes claims that fly in the face of everything we know about (for example) the density of water vs the density of a human being, and how many people can be fed with a couple of loaves and fishes, nevermind the occasions when diseased people are suddenly cured, or dead people come back to life.
Given that these are presented as evidence in favour of the existence of god, it is circular to argue that the existence of god makes this possible. For example: I can prove that I had $10 in my pocket this morning, as I have $5 in my pocket now. I bought a cup of coffee, thus the $5 is the change. How can I prove that I bought a cup of coffee? Well, look, I have $5 in my pocket, and since I know that I started the day with $10 in my pocket, this proves that I bought a cup of coffee. Around and around we go.
So the sincere, intelligent and honest theist will not both refer to the existence of god as ‘proof’ that the miracles in the bible are real and claim that the correctness of Christianity is evidence that the miracles happened, because they are arguing that the miracles are evidence that Christianity is correct: A cannot be evidence for B if B is evidence for A.
Something that makes testimony more compelling is when people who are outside of a particular group also repeat the testimony. If, for example, there were Jewish accounts of the miracles of Jesus, or accounts from people who were critical of the Christian claims of Jesus, that would increase the possibility that the events actually happened (and begin to rule out outright lies). But there is no testimony beyond that of the interested parties, and testimony cannot be evidence in favour of itself.
Furthermore, the biggest problem for most theists is that even if the miracles actually happened, they don’t support the idea that Christianity is true, merely that that particular Christian account of that particular miracle is true. If, somehow, that there existed a person named Jesus who actually did walk on water, then this doesn’t provide any evidence that the Christian god created two human beings (in whatever order you feel like cherry-picking) to kick off the whole human race, or that the Christian god somehow exists, and also exists outside of space-time (whatever that means). If it’s true that Jesus walked on water, then it’s true that Jesus walked on water AND NOTHING ELSE.
Additionally, there’s a question to be raised about what, exactly, we mean by “miracle”. A perusal of history will tell us that it (like ‘the supernatural’) is merely a placeholder for “I don’t know how to explain what just happened”. Insofar as ‘a miracle’ is defined as something that circumvents or goes against nature, all arguments regarding ‘the supernatural’ apply to ‘miracles’. ‘Miracles’ don’t happen. ‘Things that we don’t have an explanation for’ happen all the time. To label such things as ‘miracles’ is to pretend that we have knowledge when we don’t.
4. Methodological Naturalism
So unlike pretty much everything else in this essay, Methodological Naturalism actually has a legitimate meaning. It pretty much means that when an event is investigated, we avoid positing ‘supernatural’ events as explanations as much as possible. If we refer back to the discussion on ‘the supernatural’ above, it’s not clear how we could possibly posit a ‘supernatural event’ as an explanation, as it would simply be another way of stating “I don’t know what happened”, and thus fail to explain anything.
Nevertheless, theists often demand that Methodological Naturalism stop “excluding” ‘the supernatural’, under the mistaken belief that ‘the supernatural’ has been excluded merely from some fit of pique on the part of scientists.
In reality, Methodological Naturalism is merely an investigation into events/effects, by looking for correlations between the event-in-question and events that consistently precede the event-in-question. Given the lack of ‘supernatural-detectors’, the definition of ‘the supernatural’ implies that if an event had a supernatural cause, then the event-in-question will consistently have NO preceding event: the event-in-question will, consistently, occur without cause.
Oddly enough, this is exactly how we have discovered molecules, and atoms, and sub-atomic particles (and so on): an event occurred that was not adequately explained by the science of the time (sometimes known as the “huh, that’s weird” moment). The investigators poke around, and try to get the event to recur. Should they get it to recur, they will look at the context of the event, and look for the possible triggers. This usually involves holding as many environmental factors as possible stable, and playing around with the frequency of one (or more) of the environmental factors (this is more generally known as ‘an experiment’). Once the event-in-question can be made to occur (and prevented from occurring) consistently, the event-in-question is considered to have been explained.
Notice that we can’t control ‘the supernatural’. It is therefore logical that if the event-in-question has ‘a supernatural cause’, then the event will never be explained by the manipulation of environmental factors. The investigators will be required to run through all possible environmental factors, and create more and more hypotheses in an attempt to guide the experimentation. When all of these things fail, when all possible changes have been ruled out, then it’s possible that ‘the supernatural’ is responsible for the event-in-question. Or, y’know, the investigators made a mistake. Which is more likely, I wonder?
Notice that if we invoke ‘the supernatural’ early in the experimental process, two things have occurred. 1) We have declared, simply, that we don’t know what caused the event-in-question, and 2) the experimentation stops before all environmental factors have been tested. This means that it’s possible (and, given the historical trends, highly probable) that ‘the supernatural’ isn’t responsible at all, but that an untested environmental factor is responsible. By positing ‘the supernatural’ as a cause, we shut down the experimentation. We stop looking for answers. We stop thinking.
A perfect example of this: Isaac Newton invented a whole new form of mathematics in order to try to explain the cosmos, but found that he was unable to account for the stable orbits of the planets. He threw up his hands and declared god to be responsible, constantly stepping in to adjust the orbits of the planets. It was necessary for Pierre-Simon Laplace to step to rectify this mistake.
In short, everytime someone has posited ‘the supernatural is responsible’ (i.e. “I don’t know what the cause is, but let’s stop looking anyway”), someone else has stepped in, done some additional work, and managed to provide a useful explanation. Until that second someone stepped in, whatever body of work was underway stalled while people uselessly declared in favour of ‘the supernatural’.
To put it another way: Methodological Naturalism is an investigation into effects. As soon as ‘god’ starts affecting things, science will start the investigation. Until that point, however, positing ‘god did it’ is on par with ‘faeries did it’.
Many of the above criticisms about ‘the supernatural’ apply equally to the concept of ‘god’. If we simply say that ‘god is a supernatural entity’, then the issues we have above with ‘supernatural’ come into play: ‘god is I don’t know what’.
Historically there have been several rigorous attempts to define ‘god’, and I’ll outline a few below, but first I think it’s important to note that these are rigorous definitions; that is to say that they are the definitions of the philosopher and the theologian. They are not the vernacular for the theist-on-the-street: the above definition (“god is I don’t know what”) is the everyday definition for most theists. When pressed, a theist will often throw out a buzzword or phrase half-heard from a theologian or philosopher (“god is the most perfect being”, “god is love”) without much thought as to what they are actually expressing. The typical response at this point is to make a claim about the limits of “human logic”, or to assert that “god is unknowable”: in essence, the theist affirms that they have no idea as to what they are talking about.
Let’s begin with Descartes (taken from his Meditations).
Imagine a perfect being, Descartes tells us. Now, all else being equal, is “exists” a more or less perfect quality for something to have than “does not exist”? “Exists” is more perfect, declares Descartes, thus the MOST perfect being possible MUST exist.
There is, to put it mildly, a bucket of problems with this, not the least of which is the notion of “possibility: what are the grounds for arguing that a timeless, formless being is possible? None. Note that I’m not arguing that such a being is impossible, but that the theist hasn’t provided even a framework within which the notion of a thing-that-is-not-a-thing makes any sense. Given that I can’t even accept the framework of Descartes’s argument, his argument necessarily fails.
Another criticism that demonstrates how the conclusion does not follow from the premises is that I can imagine a great number of things, such as a unicorn. Surely the MOST perfect unicorn MUST exist, as it likewise fits Descartes’s logic? If we reject this argument for the most perfect unicorn, then we must also reject the argument for the most perfect being.
How about Samuel Clarke? (Samuel Clarke: A Demonstration of the Being and Attributes of God: And Other Writings (not on Project Gutenberg, sorry))
Can you imagine god simply not-existing? No? Then god MUST exist! (I’m not making this up, this is the cornerstone of Clarke’s argument)
We can go through history, through all the various Christian, Jewish and Islamic conceptions of God, and they all fall apart. Any criticism of the these vapid notions is criticized in turn of being merely ‘human logic’, as if the fault for a crappy argument lay with the person pointing out the vapidness of the argument.
Another way of pointing out the lack of content in the word ‘god’ is to simply substitute in a different word like ‘wizard’ or ‘fairy’ and see how the sentence changes. In terms of understanding the cause of an event, there is no significant difference between the sentiments ‘god did it’ and ‘faeries did it’: neither of these claims explain anything at all, they don’t tell us how to avoid the event (if it were negative), nor do they tell us how to repeat the event (if it were positive): we need a real object for these words to refer to, an object that we have some kind of understanding about. For example, “wolves ate the sheep” gives us some clear directions on how to prevent the problem (the disappearance of sheep) because of our knowledge ABOUT wolves. We have zero knowledge about ‘god’ and zero knowledge about ‘faeries’ (just lots of fanciful stories), so these attributions of cause are worthless.
The final way that god is typically defined is negatively: not mortal, not limited, not material. The issue here is that without any positive assertions, it’s still entirely unclear as to what we are talking about. Furthermore, simply writing down a lot of words on a piece of paper doesn’t demonstrate that said list of words actually comes together to exist somewhere/when.
To finally address the opening statements:
1. “Can one expect human logic to understand the supernatural realm as easily as it does the natural realm?”
We expect explanations, which are given by humans, of ‘the supernatural realm’ to meet basic logical standards. If those explanations are illogical, then the theist has failed to adequately explain their perspective. This is not anybody else’s problem.
2. “Are you saying that you reject the existence of the supernatural because people around you can’t agree on the exact nature of God, or of the Creative miracles?”
I reject the claims of theists regarding ‘the supernatural’ because I’ve heard these claims before, from a variety of other religions and fairy tales. When there’s nothing to differentiate the claims of a theist from mere wishful-thinking then the problem lies with the theist. When a scientist says “hey, want to see a piece of ceramic float in the air with no visible means of support”, we demand evidence (and lo! The MagLev train was born). When a theist claims that someone can walk on water, they have the same bar to meet. Failing to meet that bar leaves the theist in the same position as The Brothers Grimm.
3. “My fear, for those who choose that route, is that due to the acceptance of methodological naturalism as the defining limitation to science (defining only what can be proven from within nature itself), those that limit themselves in this way and trust that nature itself is “all there is”, will never have the chance to find out if the supernatural actually exists.”
The theist is asserting that ‘the supernatural realm’ exists. The theist, at this point, needs to explain A) what they are talking about, and B) how they happen to know that it’s real.
4. “Does methodological naturalism include or exclude God?”
Since methodological naturalism includes anything that has a measurable effect on reality, ‘god’ will start being included as soon as ‘god’ starts having a measurable effect. Up until this point, ‘god’ has not had any noticeable effect on reality. Once ‘god’ does something noticeable, they’ll be included.
Please notice: it takes significantly more work to break down these statements than it does to make them. This is why most scientists and science-minded people do not engage with the drivel spouted by Theists: the theists, by and large, demand that rather than putting in the intellectual legwork themselves, that the non-theists drop whatever they are doing and explain everything THAT HAS ALREADY BEEN EXPLAINED again. Caps for emphasis.
If there’s any message I would like any theist reading this to take home, it’s the following: rather than raising your eyes to the sky and demanding that other people take time out of their life to explain things to you, take a walk to the library instead and crack open a book. Rather than talking to the echo chamber that is a church, do some research. You don’t have time to do this? Guess whose problem this is…
Failing to meet that bar leaves the theist in the same position as The Brothers Grimm.
Actually, the BeeGees, as pioneering folklorists and linguists, have earned significant respect in the history of science.
Those asserting theophenomena so far have earned squat.
Say, that was very interesting. It would definately be great to have a local youth group go over this before they head out for an evening of proselytizing. However, your article could be difficult for some teenagers to grasp. Which isn’t a bad thing, just that some concepts are better understood by educated people, ie. most adults. It makes me think about how sad it is that millions of American kids get indoctrinated so easily into religion at an early age, but it takes so much more effort to show them the reality of our world. I agree completely with your article, but also makes me sad that it had to be written, ya know? Perhaps the local youth pastors will read it and just quit their job before they prep the kids for that evening of proselytizing… That’s my prayer, uh, I mean, wishful thinking… Great article!
I would love it if a local atheist group read this post before going out and approaching and debating the local street preachers, or teen evangelists, etc. Or even if a group merely printed out this post in a nice, pamphlet form, and handed it out for free to anyone who wanted one, at local community events (like farmers markets, in front of the local drugstore, etc.)
That would be AWESOME!
Thanks for this Brian and welcome.
Hah! I was reading posts in reverse order after several days away, and read to the end of this without realizing you were not Ian. And given how much I love Ian’s writing, you should take that as a compliment.
Anyway, what you said. It drives me nuts when people refer to “the supernatural”, as if there were some other realm that’s completely separate from dreary commonplace physical reality, and yet determines how we must live in this reality. Sh-what?
That is definitely a compliment. Thank you! 🙂
Diligent, thorough, and eloquent.
As far as I’m concerned you’ve eliminated the entire “god” argument! 🙂
And the funny thing is, as soon as ‘the supernatural’ is understood and explained, it ceases to be, and becomes ‘natural’.
Interesting post (and not too long, in my view).
On the issue of methodological naturalism, if by “methodological naturalism” you mean avoiding positing “supernatural” events as explanations, and by that you mean avoiding saying “I don’t know what happened” as explanations as much as possible, I would say that that has to be avoided always – it’s never an explanation to say “I don’t know what happened”.
That said, it seems theists make a claim that somehow some kind of entities are excluded from science, as a method, so I’m not sure whether it’s useful to say that one adopts methodological naturalism, while one is using the words to mean something they do not mean (which is not to say that they mean something, but in any case, they do not mean “I do not know what happened” when they say that the cause is “supernatural”).
So, I’d say that science looks for explanations, but I’m not sure why say that it adopts methodological naturalism. It seems to me that that would give a mistaken impression of what one is claiming, in all likelihood.
There is another way of defining “supernatural”: to say something like “ghosts, witches, human-like vampires, demons, and all sort of weird entities described in fictional and/or religious tales are supernatural”.
That gives you a very imprecise concept – i.e., there would be considerable variation in usage between different speakers -, but at least, it’s probably usable in some contexts.
Of course, it’s not the case, in that sense (or those senses), that science assumes methodological naturalism: scientists reject explanations by, say, human-like vampires, just as they reject a claim of a planet made of cheese elsewhere in the Milky Way: they know enough about the world to know that those are false claims – silly claims even.
However, if there were vampires roaming around like they do on TV, of course scientists would have no objection to study them. The same goes for other entities usually called “supernatural”.
Granted, some of those might be difficult to study if they don’t cooperate, but the same goes for, say, advanced extraterrestrials, so that does not have to do with whether the entities in question are supernatural.
I think that theists claim that their “theories” are being unfairly discounted beforehand, but that is not so: they’re being discounted as any silly claims are discounted.
And if someone actually wants to put some such claims to the test (when they are testable), then whether or not the event is “supernatural” would not be a problem, it seem (non-testability is of course a problem, but it affects hypotheses about advanced stealthy aliens as much as it affects hypotheses about shy ghosts).
So, in short, I would be inclined not to say that science adopts methodological naturalism.
Regarding the concept of “supernatural”, I do agree that there are serious problems with the concept: the vernacular concept is at best too vague (more precisely, there are many concepts treated as one, resulting in miscommunication), and philosophical concepts are often not coherent – if they are, they don’t seem to match any of the common usages, and they too are varied.
I don’t think that “the natural” is the known, though. I think “natural” has the same kind of conceptual difficulties as “supernatural” does.
So, in short, I would second your suggestion of asking anyone who makes a claim that there is a supernatural cause what they mean by “supernatural”.
If they go with the definition by examples (e.g., as I did above), then point out that science does not adopt methodological naturalism, but rejects supernatural claims just as it rejects claims of planets made of cheese.
If they posit an incoherent definition, one can point out that their definition is incoherent.
Else, one can discuss under the given definition. But this third possibility is very improbable in my experience.