Thursday, July 18, 2013

Misunderstandings regarding Craig, Kalam, and Relativity

Jonathan Pearce is having a "Why I am a Christian" series and in the inevitable discussion on Vincent Torley's piece it seems Vincent has done some criticism of my Countering the Kalam series.  Vincent alleges that I've gotten my science wrong and that I have misrepresented William Lane Craig.

I believe this is all based on a misunderstanding of some of the science involved and in a superficial reading of Craig's work in which he's talking about how different philosophies of time impact the cosmological argument.

The accusations from Vincent

Vincent made a lot of claims against me that I'd like to address in order of complexity, starting with the simplest.

Defining Lorentz Invariance

Vincent misunderstands me when I say that “Lorentz Invariance” means Time Dilation and Length Contraction." I was talking in the context of a quote from Craig where he says "Lorentz Invariance is apparent, not real".  In that context, the evidence we have for Lorentz Invariance is Time Dilation and Length Contraction. That's how we observe Lorentz Invariance, I'm not trying to do a definition there.

John Bell

Vincent also makes illusions to Craig's views getting support from noted physicist John Bell, which is misleading.

As I've addressed previously, Bell was talking about the Lorentzian view of Special Relativity as one possible solution to the tension Bell's Theorem causes in physics between quantum mechanics and relativity. Bell didn't actually endorse the Neo-Lorentzian view, he had mentioned it as one possible way of many theoretical ones that can solve the issue.

Bell did advocate teaching the Neo-Lorentzian view of Special Relativity before then teaching the standard Einstein-Minkowski space-time interpretation so that students could better understand the amazing implications of relativity.

Timeless Creation

Vincent takes issue with my charge of "timeless causation" against the Kalam.  He states:

Finally, I simply don't buy the argument that a timeless God cannot create the universe.  Craig's reply that God exists timelessly without (or beyond) the universe, with a timeless intention to create a universe with a beginning, makes perfect sense to me. In any event, is it any less coherent to say that the multiverse (which is also outside time) created the universe? If you accept the latter, I can't see why you'd balk at the former.

But here Vincent is confused about the science involved, and the different understandings of time and causality at play.  This confusion can ironically be cleared up by reading some of William Lane Craig's own writing on the issue!

On the B-Theory which is strongly supported by Einstein's relativity, our space-time always exists.  This is the view of universe Stephen Hawking endorses, in fact when Hawking talks about "imaginary time" he's really referring to space-time realism.  In this way time can be finite, but still always exist.

Similarly, on this view, causality isn't inherently time dependent (Craig explicitly discusses this) and if you still hold to theism the universe becomes "contingent upon god". The issue here is that this view contradicts the idea of creation ex-nihilo (out of nothing), which Craig objects to. 

The problem for the Kalam here is that once causality is no longer time dependent, Craig loses all impetus to say that there must be a cause if time is finite. There still could be a cause, but you can't argue that there must be one.  At this point the Kalam argument fails, and Craig has to switch to a Leibniz Cosmological argument which deals with things in terms of necessity and contingency.  The problem there is that there are refutations of that and such an argument can't even pretend to appeal to scientific support.  The entire appeal of the Kalam is that it supports the vague notion that "Science really shows that god must exist!" that is popular in apologetics.

But what about my objection to Craig's appeal to timeless causation?  Aren't I being hypocritical?

The issue here is my objection to Craig's version of timeless causation on his very specific version of the A-Theory.  Craig himself discusses the problems with his own view:

So on the view I propose, God exists timelessly without the universe with a timeless intention to create a universe with a beginning.  He exercises His causal power, and time as a result comes into being, along with  the first state of the universe, and God freely enters into time.  It all happens co-incidentally, that is, together at once.  This is, I own, a mind-boggling conclusion, but makes better sense to me than the alternatives.

When Countering the Kalam, I don't hitch my proverbial wagon completely to the B-Theory of time.  What I did argue in my philosophy video is the idea that on Craig's view an actual infinity is preferable to timeless causation. This is because Craig's arguments against an actual infinite fail, and some objections raised against them have been unaddressed for nearly a decade.

That said, lets say you disagree with that preference for actual infinities. The stronger point I demonstrate is that on Craig's view of time, at the very least there is simply no clear way to decide between which of the two "absurdities" is true, which leaves the Kalam not establishing it's conclusion.

Similarly, when I discuss quantum nucleation cosmological theories that are proposed by Alexander Vlienkin, Alan Guth, and popularized by Lawrence Krauss I'm talking about theories that are compatible with both the A and B Theory of time, so it's not a case where if we ever found out that the B-Theory was not likely to be true, we're not stuck dealing with the Kalam again.

Misrepresenting Craig

Vincent alleges that I've misrepresented William Lane Craig's views when I've claimed that Craig does not believe that Time Dilation and Length Contraction are only "apparent" and "not real".  This is despite the fact that Craig explicitly states this about his neo-Lorentzian theory. To support this charge, Vincent links to this footnote in an article by Craig on god and time where Craig says the exact opposite:

This is, in fact, the modern Lorentzian interpretation of STR, which holds that velocity affects one's measuring devices so that moving rods contract and moving clocks run slow. Such an interpretation does not commit one to a substantival aether, but merely to an aether frame, i.e., a privileged frame of reference. That the Lorentzian interprets length contraction and time dilation as not merely apparent, but real, cannot be cited as a disadvantage of the theory, since the Einsteinian also must posit real contraction and dilation...
These are contradictory statements, did Craig change his mind?

The short answer is no he hasn't changed his mind, Craig has not contradicted himself either, and I have not misrepresented his views.  Craig is simply talking about different things that span both science and philosophy of time.

The long answer involves getting into some science. Lets do this.

Breaking Relativity

Einstein's relativity combined with quantum mechanics forms what's known as the Standard Model for particle physics.  In fact the reason finding the Higgs Boson was such a big deal is because it was the last particle predicted by the Standard Model that we hadn't found evidence for yet.

To put it mildly, the Standard Model is a pretty big deal in physics.  If one of the underlying theories that form the basis of the Standard Model (like relativity) were to be proven false it would trigger a massive shift in physics.

This massive shift wouldn't necessarily be a bad thing, it could provide us with unique insights which would indirectly lead us to understand quantum gravity and ultimately come up with a more complete theory that describes all of physics.

This is what Craig is referring to in the footnote Vincent linked to. This is the efforts by some scientists to test the assumptions of Einstein's Special Relativity to see if there is something wrong with them.

Standard Model Extension

The Standard Model Extension (SME) is a theoretical framework which would explain all available data that we currently have, but it allows for Lorentz symmetry to be violated.   If Lorentz symmetry was violated, it would mean that one of the critical assumptions of Einstein's special relativity would be violated, which would trigger that massive shift in physics I mentioned earlier.

The key thing here is that this is a theoretical framework, but that it also makes predictions that can in principle be tested.  These tests are exactly what's being done by scientists studying this area.  It's actually pretty rough since the theory is that violations of Lorentz symmetry would be extremely small and thus hard to detect.

Here's the thing, so far the tests have all failed, and Special Relativity (as well as the Standard Model) still stand.

What they're looking for

The assumption Einstein made that is being tested is that "That the laws of physics are constant across all reference frames."  This basically means that there's no privileged reference frame where the laws of physics don't work the same way that they do in all other reference frames.  The idea is that physics is the same no matter where you are or how fast you're moving. 

In the SME, the effects of Lorentz invariance (ie. time dilation and length contraction) are as "real" as they are in the Standard Model.  The idea is that they actually happen, time actually slows down, rods actually get shorter, etc in a majority of reference frames.  The kicker is that there would still be a reference frame where this didn't happen, a privileged frame where the laws of physics themselves were different. 

How is this different than Craig's Neo-Lorentzian view?

The biggest difference is that in the SME, this privileged reference frame, if it existed, could be detected and described.  The idea in SME is that Special Relativity is correct (ie reflects what really happens) in most reference frames that we can observe, but that there is a special frame where it doesn't apply.

In Craig's Neo-Lorentzian view, he takes the idea that the equations in Special Relativity incorrect in that that they merely explain what appears to happen, not what actually happens.  The key distinction is that Craig's view is unprovable and unfalsifiable, where as SME predicts things that if correct, we can verify them.

What Craig Wants

What William Lane Craig really wants is for the A-Theory of time to be true (ie. that space and time are absolute), and he wants the evidence we have against that view to be shown to be problematic.   Right now the evidence we have from Special Relativity shows that space and time are relative, not absolute.

To get around this problem Craig endorses the Neo-Lorentzian view, the ultimate punt into metaphysics which states that the question of space and time being relative or absolute is something that in principle can't be answered by science.

However, Craig would be equally happy to have SME turn out to be true, so that his Neo-Lorentzian view becomes unnecessary.  Incidentally, this is exactly the same kind of excitement he showed when it looked like relativity may have been over turned by undermining the other assumption - that nothing travels faster than the speed of light.  Unfortunately for Craig both efforts to undermine relativity have come up short, so all he can do is make extravagant unfalsifiable assumptions about relativity to try and keep the A-Theory of time as a live option.

The problem of course, is that so far Einstein and the Standard Model have held up, so Craig is left appealing to his unfalsifiable privileged frame to try and pretend the Kalam is a live option in anything but the minds of people who already commit to god's existence, or an esoteric philosophy of time.


I realize that there is a lot of material here, but I want it to be clear that I've not at all misrepresented Craig or gotten the science wrong.  Hopefully I've done that.


  1. Good post. One thing I noticed when Craig says that god has "a timeless intention to create a universe" is that I think it destroys the cosmological argument from contingency or at least shows it's logic to be faulty.

    The reason is because if god's intention is to create the universe, and this intention exists timelessly, there was never a possibility that god could not have had this intention, since there was no moment when he made this decision. So the universe cannot be contingent, in that there is no possibility that god could have created another universe or no universe because for a decision to be contingent it cannot exist timelessly.

    Logically it looks like this, if A exists necessarily, and if A exists, B necessarily exists, then B exists necessarily too, because there is no possible way that B could not exist.

    The only way out of this is to say that god makes decisions in temporal terms (since metaphysical time has no basis)and that forces god to be in time before creation.

    I wrote a short blog about exploring this idea here:

    1. Definitely interesting stuff. If I remember correctly there are problems along these lines that get brought up if you follow the Principle of Sufficient Reason (PSR) through all the way.

      For those who don't know the PSR is the basis for the Leibniz Cosmological Argument, and it's one of the reasons some apologists like Peter Van Inwagen reject that argument.

  2. The problem for Craig is to explain the muon. In cosmic rays, which are studied at the Pierre Auger Observatory in Argentina, the muon's half-life is greater than its lab sibling, the difference being in pefect accord with Special Relativity. If time dilation is an illusion, then everything we observe is an illusion. Craig would have to believe we live in a matrix, or we are brains in a vat.

    1. Wow, that seems like some pretty significant evidence. Is there a resource to look at that describes this in some more detail?

      That said, with the kinds of appeals Craig makes in that Neo-Lorentzian interpretation I wouldn't be surprised that he can account for this discrepancy.

    2. A simple explanation of the muon half-life in cosmic rays and the role of Relativity is given here:

      As to the neo-Lorentzian, it postulates an aether that can't be detected, hence the theory is unfalsifiable. So it has very little scientific merit.

  3. The Thinker,

    Yes that was always the problem right off the bat with positing a God who decides to create the universe.

    In his argument from the beginning of the universe to a God, Craig simply posits a mind that can do the things that will simply get him out of the quandary and create a universe.

    Ask him HOW this magic power works and he has no answer. It's not a mechanistic explanation, you see. (Hence, magic). But then once that door is opened to positing causes for which you need give no account of HOW they work, the counter explanations are limited only by our imagination. A timeless turtle caused the universe. A timeless napkin. A timeless flower. A timeless quantum particle.

    If Craig demands an explanation for HOW any of those things could cause a universe we can retort: It's simply a magical attribute they have, and attributing this ability explains how they could be posited as the cause of the universe. Since Craig does no better and has no explanation for how this God's magic works to cause the universe - he just gives it the magic attribute and slaps the label "willing" or "deciding" on it - then he has no grounds to demand from us an explanation for how the above magic entities cause universes. That's what you get when you allow yourself to play the magic card. (And the reason science has progressed our knowledge of reality while theism has not, is that we've recognized these liabilities with magical explanations).

    We always have to call theists on this special pleading, that they can not open the door and let only their preferred bit of magic through, without causing a flood of magic through the door.

    This is one reason why I have a bit of a problem with the way so many atheists who debate the Kalam, especially with Craig, get sucked into trying to debate the initial premises. It does seem that one should be able to cut the legs out of the argument right in those premises, and hence obviate any following moves the theist makes to infer a Personal Cause. The problem is, and this plays out so often in Craig's debates, to counter the first premises inevitably drags both sides into areas of physics in which, typically, neither side is admittedly an expert. And even when you do have people on both sides professing some expertise, it's above the expertise of the audience, so it's hard to land some decisive blow that everyone easily comprehends.

    Whereas attacking the move from cause to personal cause is far, far easier. The moves there are so blatantly gratuitous it's easy to highlight them in neon for an audience.

    (BTW, the only time I've ever seen Craig get close to dealing with his special pleading for a Personal Cause for the universe - how this personal cause COULD have the means to cause - is to attribute it with Libertarian Free Will. Which to Craig is simply playing the magic card again, since He doesn't feel he has to have a causal/mechanistic explanation for how that works.
    He has then tried to say "it has Libertarian Free Will, just like us" but that, again, just begs the question, giving every human a magical ability for which there is no actual evidence).


  4. Unrelated to Relativity, one of the other problems with the Kalam Cosmological Argument is the equivocation of the words "begin to exist":

    P1) Everything that begins to exist has a cause for its existence.

    P2) The universe began its existence.

    C) Therefore the universe has a cause for its existence.

    In P1, the phrase 'begins to exist' refers to objects within the universe that are a rearrangement of existing energy and matter. Trees, tables, people, etc. begin to exist under this definition. This is creation ex materia.

    In P2, the phrase 'begins to exist' refers to creation out of nothing. This is creation ex nihilo.

    Hence the conclusion C is invalid. If not creation ex nihilo in P2, then the universe has a cause but it's a rearrangement of existing energy and matter as per definition in P1.

    1. Hey Joe,

      Craig technically doesn't equivocate here, but he does do some sleight of hand.

      Check out the first two posts/videos on the Kalam here:

      The short version is that he uses a very general definition of "cause" that can mean efficient OR material causation.

      So since he also defines "universe" to mean "all of material reality", he then eliminates the idea that we can have any material cause for the universe - if you accept premise 2 of the argument.

      The problem for him is that there's no evidence whatsoever that "all of material reality" has an absolute beginning, preceded by nothing - which is what he's trying to establish in Premise 2.

    2. Fine, though I was referring to KCA per se.

      In the case of Craig, his argument can easily be defeated by pointing out that the parts have certain properties -- in this case, "began to exist" -- which does not mean the whole has the same properties. The counter example is hydrogen and oxygen: both have properties; together they form water, with entirely different properties. All things in the universe began ex materia, but for the universe, either it has the same type of beginning, therefore it already existed but in a different form, or has a different type of beginning, ex nihilo, that is, the whole has a different kind of beginning, but then you need magic to create it. And as you say, there's little evidence for that.

      Either way, P1 + P2 does not lead to C.

  5. Hey bro, I found this the other day while surfing. Would love to listen to ur critique.