Friday, December 6, 2013

Naturalism, Falsifiability, and Hiddenness

This post is a bit off the cuff, as it’s mainly in response to a Twitter conversation to elaborate on something you can’t put into a series of tweets.  This was born of a conversation with Alex and Elijiah, and the topic was meaty enough that I wanted to write about it.

The question is whether or not a-priori Naturalism is “reasonable” or at least “not scary”.  That’s basically the starting position in philosophy that: no matter what we observe, we would never accept evidence of something supernatural existing.   It’s rejecting the supernatural a-priori.

Personally, I kind of abhor this line of reasoning, or at least I find it terrible to be in a position where I’d say that there can never be evidence of any kind to prove the existence of a god or other supernatural entities.  

To me, this reeks of a sort of fundamentalism that I’d normally chastise certain religious people for practicing. 

At this point you may be wondering what the issue is.  Surely “naturalism” is different than “a-priori naturalism”, where the former is more of a position we reach by way of conclusion, and the latter is some kind of naive position that you’d expect from a philosophically uninformed wing of internet atheism.

The problem as Alex put it on Twitter is “Clarke’s 3rdLaw”: 

“Any sufficiently advanced technology is indistinguishable from magic.”

The issue comes up when we atheists are asked what kind of evidence we would accept to prove the existence of a deity.

For me, I prefer to point out that it would take the same kind of courtesy that was supposedly extended to the Apostles: The deity could just show up and work miracles.  

I know if some guy just appears out of thin air, can tell me all sorts of things that only I would know (mind reading, knowing private events in the past), start flying around, or start flying me around, or a host of other seemingly physically impossible things - and then they tell me they’re Jesus/god/Xenu then I’m pretty much going to believe.

This sounds simple, but Clarke’s 3rd Law gets in the way.  How would I not know that this isn’t some super-technologically-advanced alien that’s trying to convince humanity to do it’s bidding, and is merely representing itself as a creature from our legends to achieve that goal?  

In fact, on reflection what seems more reasonable: That the supernatural exists, or that the naturalism is true and I’m just witnessing advanced technology?  In fact my life experience to this point would weigh heavily towards the naturalist position!  On further reflection, it seems far more likely that some aliens out there that are using their technology combined with our primitive myths as a way to manipulate us as opposed to Jesus just deciding to show up after literally thousands of years of divine hiddenness.

Given this, it seems reasonable to take the a-priori Naturalism position!  But I have to admit that if I witnessed miracles such a “theory” sounds extremely ad-hoc or worse: an epistemological black hole from which there is no escape. 

That’s what happens when you adopt a position that can’t be falsified.

The problem was that for a while I couldn’t figure out what the problem was with the above reasoning that seemed to force me into adopting a-priori naturalism. 

It’s taken me a lot of thought to come to this conclusion, but I realized that the issue was partially within the definitions I was working under being imprecise, and partially between a distinction in ontology and epistemology.  Let’s start unpacking.


Defining things like “Miracle”, “Naturalism”, and “Supernatural” in a way that isn’t question begging is extremely hard.  I’m going to give it my best shot to define things general enough to be acceptable to both theists and atheists, but specific enough in order to be useful.

I will start with the assumption that things that are physically impossible are still logically possible.

Given this assumption, I will define the terms as follows:

Naturalism – Things that are physically impossible cannot happen in reality.
Supernatural – Entities that can do physically impossible things so long as they are still logically possible.
Miracle – An instance of a supernatural agent doing something that is logically possible but is physically impossible.

I’m hoping that these definitions are acceptable to all parties, and given them, we can identify what would and wouldn’t count as evidence for the existence of the supernatural.

Specific Examples of Miracles

Given my definitions, it becomes evident that there would be different classes of miracles.  The first would be miracles that we limited humans would never be able to apprehend as a miracle and others that would be more evidently miraculous.

A miracle that we couldn’t in principle apprehend:

Pretend that Jesus walks up to you. Jesus then tells you the exact position and momentum of a particular particle somewhere in the immediate area around you.  

Assuming the Heisenberg Uncertainty Principle is correct, this is something that is physically impossible to do, and would be classified as a miracle.   Notice that even if we were able to isolate a particle with the most accurate measuring tools available, we would never in principle be able to verify the fact that this is a miraculous event – we couldn’t actually identify the momentum and the position of the particle at the same time because of the very “law” that makes the event miraculous.

Note the miracle here is that Jesus would know both the position and momentum of the particle, and not that he would be able to know one value that would otherwise take an incredible amount of technology and time to measure.  That may be classified as its own miracle, but that is not the one under consideration.

Assuming this happens and we were able to verify one or the other value to the point where we accept that Jesus got it at least half right, we’re left with the following possibilities that would allow the naturalist to insist this wasn’t a miracle:
  1. The Heisenberg principle is actually false, and it is physically possible to know the position and momentum of a particle, but we just can’t replicate it.
  2. Jesus is actually wrong about the value that we can’t measure, but we can’t prove that in principle.
A miracle that we could apprehend: 

Pretend you park your car, get out, and Jesus approaches you.  Jesus tells you who he is and what powers he possesses, and you ask for a demonstration.  Jesus then makes your car spontaneously burst into flames, destroying the car.  You’re able to verify that your car is not working and inspect the wreck.  Jesus then has you step back and your car somehow reconstitutes itself, and you’re instructed to get in and go for a drive while Jesus rides along (because he wants to be your co-pilot).  Assume everything in the car works.

We can go a bit further and have the scenario happen under controlled conditions, so that we could ensure that the car was within a physically isolated system and that no outside energy could get where the car was stored.

There are technically two miracles here, first the car spontaneously catching fire, which would be easy enough to imagine technology being used to bring about.  Let’s leave that aside and concentrate on the car reconstituting itself, a dramatic violation of the second law of thermodynamics among others.
At this point the naturalist is left with the following possibilities to insist a miracle hasn’t taken place:
  1. The second law of thermodynamics is actually false and it is physically possible for entropy of a closed system to decrease.
  2. That Jesus has used something to induce a hallucination in you and others witnessing the event.
In either case of miracles we have some commonality – either we’re wrong in our understanding of what is physically impossible, or we’re being deceived in some way. 
Let’s tackle both cases.

Clarke’s Third Law

Notice in both cases, Clarke’s Third Law boils down to the first possibility – what has happened isn’t actually physically impossible, it’s just that our understanding of what is physically impossible is wrong.

The problem is that such a position seems to beg the question in terms of naturalism, since it would always be possible to state that we just have a poor understanding of the laws of nature.  This would effectively broaden the definition of “naturalism” to encompass “supernaturalism”, since the demonstration of any previously thought physically impossible act would then just broaden the definition of naturalism beyond any useful capacity. 

At this point even “god” becomes “natural”.   The entire enterprise ends up devolving into a useless kind of tautology where “anything that is possible in reality is possible”.  

This also misses another key area in discussing miracles, which is the difference in capability between humans and the supposed divine.  So even if we broadened the definition of naturalism so far that even god is natural, we still haven’t established that at some point we humans, or any other being for that matter, could do the miraculous actions.

From this, I’ve come to the conclusion that Clarke’s Third Law is just another way of stating that we can’t ever really have “absolute certainty” about miracles/magic.  The issue is that when it comes to epistemology we’re already aware of problems with “absolute certainty”.  It’s always about degrees of certainty.

This leads into the next area.

The Deception Option

In both cases of miracle we’re alleging that Jesus has not actually done something physically impossible, but rather has deceived us into thinking he has.

Before I get into the problems here, I want to highlight the difference between the deceptive options in both miracle cases.

In the first case, the naturalist would allege that Jesus is simply lying to us about something we could never verify the truth of empirically even if the miracle was occurring.  

This is quite different from the second case where the only option left for the naturalist is to assert that their senses are no longer reliable, since the effects of the miracle would be empirically verifiable at every stage.
The issue here isn’t only that this position can’t be falsified; the problem is that it’s an objection that could be raised against any empirical observation.  It’s the first step on the road to solipsism, with Jesus eventually becoming a Cartesian Demon of sorts.

The key bit is that in the second scenario, the miracle is empirical and presumably Jesus would be willing to repeat it (and other miracles) as often as we’d like.   Eventually the fact that Jesus could work supposed miracles would become as empirically established as well as we could establish anything empirically.  In this scenario the person insisting on being deceived in every case of Jesus’s miracles could be dismissed as delusional as a Young Earth Creationist.

Given that a naturalist is generally very big on science and empiricism, I think we can conclude that the second type of miracle I listed is something we would have to accept without also having to give up the very basis we have for science and empiricism.

A Real Problem for Theists

Given my definitions and my example of an empirically verifiable miracle, naturalism becomes falsifiable in a sense that’s trivial for any being that wanted to falsify such a belief.

One of the roots behind this discussion is the argument from divine hiddenness, and an often asked question by apologists: “What would count as evidence?”  This is typically asked when atheists object to arguments from Natural Theology as supposed “evidence” for the existence of god.   

The issue I have is that “arguments” don’t really hold up well as evidence, and I think that even theists, or at least Christians, would have to agree to this.

Whether it’s Yahweh in the Old Testament or Jesus in the New Testament, they didn’t just show up and start spouting off arguments from natural theology.   They did actual, empirically verifiable miracles!   The entire argument for the resurrection hinges on a miracle that, if it actually happened, was at one point empirically verifiable.  

This is the key behind divine hiddenness, supposedly god did all these empirically verifiable miracles, but that was all in the past, and it never seems to happen in any circumstance that’s verifiable today.  This puts any access to miracles firmly in the first category I used, the inaccessible kind of miracle.

The problem for theists is that it’s trivial to imagine a world where god provides us with empirically accessible miracles all the time.   This isn’t even something that’s just cooked up by us atheists, it’s something that comes straight out of the New Testament in Mark 16:17-18:
17 And these signs will accompany those who believe: In my name they will drive out demons; they will speak in new tongues; 18 they will pick up snakes with their hands; and when they drink deadly poison, it will not hurt them at all; they will place their hands on sick people, and they will get well.”

We can go further! Any god could simply act like any deity that’s in a D&D or Pathfinder campaign, bestowing spells on faithful clerics that attempt to channel his energy in a way that is consistent with god’s desires.  If there was only one true god, then that god could make it so that no other malevolent spirit could channel similar abilities, so that only followers of the true god exhibited these powers.

There would be no history of supernatural claims being continually shown as natural phenomenon, within the reach of any normal human.  There would always be an empirically verifiable gulf in ability between believers and non-believers.  The fact that this isn’t happening in reality is a very real problem for theists.

The final question for theists is that if the god of Christianity supposedly wants nothing more than to establish a relationship with us, then why are we denied access to the best kind of evidence we could ever have of its existence?


  1. If "arguments" don't hold up well as evidence, then what of the divine hiddenness argument? It seems to be in the same "class" as the natural theology arguments - it points to some feature of reality, and argues that this feature is evidence that the proposition "God exists" is false. This doesn't seem much different in structure from, say, the fine tuning argument.

    1. An excellent point, and one worth expanding on.

      I should preface this with the idea that if there was an argument whose premises were true and it was logically valid, I'd have to accept the conclusion. The problem with natural theology arguments is that the premises are not more likely to be true (in the way the arguments require) so as to establish their conclusion. They are controversial at best, and that's just for the deistic god.

      The argument from hiddenness only counts against deity that supposedly wants nothing more, and created all of the material universe, so that it may have a personal, loving relationship with us humans.

      This leads into some reflection on the argument itself. One thing we do have extremely strong inductive evidence for is the idea that the best way to identify true features of reality is via empirical observations. The immense progress of science and the empirical method it entails demonstrates its superiority to the a-priori metaphysical thinking that preceded it.

      Finally, there is a bit of an asymmetry in the position of the theist and the atheist, and this is at the heart of the argument from hiddenness.

      The theist is asserting the existence of something that given what we have access to, we have no empirical evidence to back up its existence - but it is NOT the kind of thing that we can't ever have empirical evidence of in principle.

      Given this, all an atheist can do is adopt the neutral position and try to show contradictions in the properties of the supposed being. In principle, if the atheist is correct, the most an atheist could ever do in principle is give arguments.

      A supposed deity, at least one that wants us to believe and love it, does not suffer this limitation if it does exist. It could even refrain from direct appearances, and allow it's believers to channel it's power under certain conditions, ala Mark 16:17-18 or D&D/Pathfinder.

    2. Something I forgot to mention:

      I'd just reiterate the point I raised in the article - even according to Christians, arguments weren't sufficient to establish the existence of the deity in order for it to gain followers. In order to get to the existence of Yahweh or Jesus being god, the being had to manifest itself in miraculous ways.

  2. Going off on a bit of a tangent...

    If arguments for theism are sufficient in principle, and it's just in practice that they fail, then the atheist need not demand the miraculous evidence written about in the bible - it would be enough to demand (at least) arguments that are sufficient in practice.

    I had another paragraph here about the theist/atheist asymmetry involving time machines, but I deleted it because it turned out to not work out so well. Oh well.

    1. There's nuance there I think you're missing.

      An argument could in principle establish the existence of some kind of deistic clockmaker god, but I don't think it could work to establish the kind of god who wants a relationship with us.

      However even then, the argument is only as good as its premises. When it comes to direct experience of empirically verifiable miracles, that's far and away the best kind of evidence we could have.

      The point is the arguments become superlative, they shouldn't even be necessary.

  3. I've been thinking about this for over a week now.

    I've decided that in principle, Clarke's Law would still make any apparently supernatural event *either* point to a) God, or b) advanced natural life, due to our ignorance of unknown physics. In practice however, there might be specific events that make the god more likely, since we can always ask the question 'What is most reasonable to believe on the evidence at hand?', rather than 'What is possible on the evidence at hand?'

    Perhaps something akin to true believers in that god developing non-natural abilities in virtue of their belief, like the power to heal amputations, or some other similarly natural-law breaking event. Or something akin to the powers of Biotics from Mass-Effect. Putting the power to effect that kind of change into the hands of a human seems to nudge it slightly away from the possibility that an alien hoaxster could be responsible (at least in my intuition).

    1. I can agree completely, it's either a) god or b) advanced natural life.

      The fact that it could be either, logically speaking, IMO, makes things worse for the theist. The thing is, by doing these miracles, god isn't "forcing" someone to believe in some absolute way.

      Interestingly enough, I've been doing some reading over at about the definition of "faith" and one of the ways he uses the word there is as if faith were a way to rationally analyze evidence and come to an inductive conclusion.

      In this case, if that was what "faith" is, then it would still take faith to believe in light of these miracles, though obviously far far less than it takes to believe in the historical account of miracles that we have to evaluate today in a universe where it appears no miracles seem to happen.

  4. This comment has been removed by the author.

  5. I think I have spotted an error. Given my experience in physics, I know that the second law of thermodynamics says that the probability that the entropy of a closed system decreases is very low. It does not say that it's (physically) impossible for it to decrease -- not at all!

  6. I suggest checking out A Leibnizian Theory of Miracles. Your definitions reduce the theist to arguing in god-of-the-gaps style, which is very ugly. Furthermore, it is actually extremely hard to know what is and is not physically impossible; see the struggles philosophers of science have with defining "natural law". Also, you might enjoy The Computational Theory of the Laws of Nature.

    1. Oh I've read the theory of miracles paper, and it entails bringing on a-priori all sorts of Aristotelian thoughts of causation that I reject as being unnecessary and outdated.

      If you've got to bring in all sorts of stuff a-priori to save your god or position, then I think that is a very big problem for you.

    2. Wait a second. The way you've set things up, the only way to prove God exists is for a god-of-the-gaps event to happen, and stay god-of-the-gaps forever. Or do I have it wrong? What could possibly suffice? Perhaps we'll be able to resurrect people three days dead, 10,000 years into the future.

    3. Well yes, I'm saying what I need is for god to show up and start demonstrating the whole omnipotence, omniscience, and Omni-benevolence thing.

      It wouldn't "stay god of the gaps" so much as it would be something that really would be "physically impossible". As such it would necessarily be something that would always be beyond our reach - and an omnipotent god would understand that to pick the right kind of thing to demonstrate! He could even explain why certain things aren't miracles by showing us why/how things are physically impossible.

      At some point you're stuck between two options - retreating into almost total skepticism about what is and isn't physically possible, or through repeated interaction with the deity, you believe, and if Omni-benevolence is as good as it should be, worshiping this being would follow.