#772 Physicists & Philosophers Reply to the Kalam Cosmological Argument (Pt III)

Dr. craig’s response

---

A

As we continue our examination of this video, the discussion now turns to the scientific evidence for the beginning of the universe based on the universe’s expansion. Much of this discussion, Paul, will doubtless be beyond you, but your perception of obfuscation and unprovability are quite correct. It is amazing how any model, no matter how speculative, no matter what the evidence ranged against it, is thought to be preferable to a model with a beginning. My responses are in blue italics.

C. FIRST SCIENTIFIC CONFIRMATION

NARRATOR: The infinite future is not the only infinity supporters of the kalam tend to embrace. They've also claimed that the big bang's singularity proves a beginning of time.

WILLIAM LANE CRAIG: . . . arrives at a state of infinite density at some point in the finite past. This state represents a singularity at which space-time curvature along with temperature, pressure, and density become infinite. It therefore constitutes an edge or boundary to space-time itself.

DANIEL ISAACSON: The argument by Lane Craig that the fact that the infinite past is impossible and therefore that conception of the universe is untenable is incompatible with their claim that there is a singular point at which the universe originates because the properties of that singularity are entirely infinite.

ADRIAN MOORE: Precisely what a singularity is, as John Barrow has forcefully emphasized, is something that involves a kind of actual infinity – for physical reasons, it involves a kind of actual infinity. And this is precisely what these very same thinkers said that they were suspicious of. So to put it very bluntly and very crudely, are they trying to have their cake and eat it? Accepting an actual infinity in one context while trying to dismiss it and being suspicious of it in another context.

Response: The initial cosmological singularity is a mathematical artifact of the standard model. That’s not to say that it is a physical state of affairs. I think it is an idealization. Nor does it involve infinite quantities in the Cantorean sense. As Quentin Smith points out, it is a case of division by zero: any mass over zero volume yields infinite density. Cosmologists often “cut out” the initial singularity so that the universe, though finite in the past, has no boundary point. Some cosmological models, like the Hartle-Hawking model, “round off” the initial part of spacetime, so that, although the past is finite, it has no boundary point. Having a beginning does not imply having a boundary point. See my and James Sinclair’s essay, “On Non-Singular Spacetimes and the Beginning of the Universe,” in Scientific Approaches to the Philosophy of Religion, ed. Yujin Nagasawa (London:  Macmillan, 2012), pp. 95-142. 

NARRATOR: Mathematicians do not think infinity is contradictory. What may be contradictory is claiming that infinity is both a contradiction and not a contradiction – claiming that the infinite past is incoherent when the infinite future is embraced, and claiming that the beginning of time is confirmed by the big bang singularity which itself involves infinities. A further problem for the kalam argument is that it assumes an eternal universe would have had to traverse infinite time in order to get to a unique now. But in relativity there is no such notion. Time is relative. In order to tackle this, Craig appeals to what is known as the neo-Lorentzian interpretation of relativity which does restore a unique sense of now.

Response: This is another red herring. Einstein’s original formulation of special relativity involved ordinary three-dimensional objects enduring through time. It was wholly compatible with a tensed theory of time and temporal becoming. The tenseless spacetime interpretation of Hermann Minkowski came a few years later. The soundness of the kalam cosmological argument does not stand or fall with a neo-Lorentzian interpretation of relativity (even though that is my preferred view).

25:17: CARLO ROVELLI: The discovery with Copernicus is that the Earth is not the center of the universe. Well, but you can always rethink that. You can imagine that. You can add a center. Nobody prevents you from doing that. You just complicate life for yourself. And it's the same for special relativity. You can add a preferred reference frame. It’s not visible, it’s not detectable but you think it’s there. That’s the neo-Lorentzian interpretation of special relativity. What do we gain? The only thing we gain is that we can do the new physics with the old mindset. But I think we should adapt our intuition to the new physics, not adapt new physics to our intuition. When we discover the Earth is round – that’s up, that’s down, but if you’re in Sydney it’s the other way around. You can think, yeah, yeah, but in reality there is a true up and a true down. You can, but what do you get? I mean, you have to adapt your intuition about that discovery, not the other way around. Physicists are not particularly keen about neo-Lorentzian interpretations of special relativity. I mean such that when you go to general relativity and have a neo-Lorentzian interpretation it is much harder because in special relativity you just pick a frame and in general relativity there is no global frame. I think everything in physics is telling us that there's no preferred frame, there is no preferred time slice, there's no preferred time.

Response: Rovelli admits that we can, with Lorentz, add a preferred reference frame. There is no scientific problem here. Rovelli’s only objection is, “What do we gain?” For my part, our gains are metaphysical: a more credible view of the world than that offered by the tenseless spacetime interpretation. But there are some prominent physicists like J. S. Bell, Franco Selleri, Antonio Valentini, and others who think scientific gains are to be had as well. In particular, the so-called Bell inequalities seem to require a preferred reference frame. See my Time and the Metaphysics of Relativity (Dordrecht: Kluwer Academic Publishers, 2001) and my and Quentin Smith’s anthology Einstein, Relativity, and Absolute Simultaneity (London: Routledge, 2007).

NARRATOR: Craig has suggested there is a universal clock that gives us a unique now – that is, cosmological time, the time since the big bang.

WILLIAM LANE CRAIG: Cosmic time is the same for every hypothetical observer in the universe regardless of his state of motion. In other words, cosmic time is a kind of reinstatement of Newton's absolute time.

CARLO ROVELLI: Well, cosmological time is a fake. Why? Because matter, gravity slows time so inside the galaxy clocks go slower than outside. Point is there are many different clocks in the universe which they don't agree with one another and there are many times in the universe which don't agree with one another. The idea of the cosmological time is just one arbitrary definition of an average, but I can give a different definition of it.

Response:  For shame! Rovelli knows that the slowing of clocks in a gravitational field has absolutely nothing to do with the time parameter that measures the duration of the universe. Cosmic time is not arbitrary but charts the evolution of hyper-surfaces of homogeneity.

NARRATOR: What about cosmology? Doesn't the big bang imply the universe had a beginning? The Penrose-Hawking singularity theorems that supposedly proved this have assumptions that almost nobody in cosmology take seriously anymore. In particular, most physicists think Einstein's theory of gravity will need to be modified to take into account quantum or other effects that arise in the extreme conditions of the big bang. This isn't just the consensus of cosmologists but the views of the very authors of the singularity theorems – Roger Penrose and Stephen Hawking.

STEPHEN HAWKING: The real lesson of the singularity theorems is therefore that we need to combine the general theory of relativity with quantum theory in order to understand the origin of the universe.

ROGER PENROSE: It is certainly a view quite commonly expressed now that we shouldn't simply give up that the big bang is a singularity – we need a theory which describes that. Most people will say it's a form of quantum gravity. I have a view which is different from that, but nevertheless I have in common with that that we need a new theory.

Response: In my published work I have treated quantum gravity models such as the Hartle-Hawking model, which features a non-singular beginning of the universe. Penrose’s conformal cyclic cosmology does not involve a beginning but is not taken seriously by cosmologists because it is based on an alien physics, not on known physics.

29:06: GEORGE EFSTATHIOU: I don't think anybody believes the universe started off with a singularity. That just tells us that Einstein's classical theory of general relativity breaks down so that classical theory of gravity doesn't apply when you get to very high energies. So it's replaced by some consistent quantum theory of gravity. String theory is a candidate.

29:29: ALEXANDER VILENKIN: Standard big bang cosmology which was a hot big bang, right - you start with a very hot dense universe and which begins at singularity. So that picture I don't think anybody believes.

Response: Right! So what becomes of the above objection that the singularity involves an actual infinite? It wasn’t a problem after all! As I said, the beginning of the universe does not require a singular boundary point.

NARRATOR: Many scientists have taken our best candidates for a quantum theory of gravity and analyzed the big bang. Models inspired from string theory often show that the big bang was not the beginning and a frequent result is that there was a contracting era that preceded our expanding universe.

29:58: GABRIELE VENEZIANO: There had been an evolution of the universe which was some kind of mirror image of what happened afterwards. So in the same way as the post-big bang era where we live, in our model will last forever and I think it also lasts forever in ordinary inflation. OK? For us, because of the symmetry, this duality, also the past is infinite. The future and the past are both infinitely long.

NARRATOR: String theory's main rival – loop quantum gravity – seems to show the same thing.

30:35: ABHAY ASHTEKAR: If you reverse the film, so to say, of the evolution of the universe and went back in time the universe was contracting and then instead of reaching a zero volume in fact when it reached the curvature became Planckian. Then new quantum effects come into the picture and they cause the universe to bounce.

30:59: MAIRI SAKELLARIADOU: Any theory that has some kind of discrete space then by definition there you don't hit the singularity. That could be an explanation of why all these models they somehow go to this bouncing solution when we arrive at a very small scale.

Response: For an examination of string models, including Veneziano and Gasperini’s pre-Big Bang inflationary model, as well as Steinhardt and Turok’s cyclic ekpyrotic model, followed by an examination of loop quantum gravity models, including Bojowald’s cyclic model and Ellis et al.’s asymptotic model, see discussion by James Sinclair in our “The Kalam Cosmological Argument,” in The Blackwell Companion to Natural Theology,  ed. Wm. L. Craig and  J. P. Moreland (Oxford:  Wiley-Blackwell, 2009), pp. 158-74. As Alex Vilenkin has repeatedly emphasized, none of these models can really be past eternal (Alexander Vilenkin, “Did the universe have a beginning?” http://www.youtube.com/watch?v=NXCQelhKJ7A.) (Say, I wonder why the interviewer didn’t ask Vilenkin about the viability of these models! Hmm!)

NARRATOR: With the physics community abandoning its belief in the applicability of the Penrose-Hawking theorem, kalam advocates switched to a newer theorem known as the Borde-Guth-Vilenkin theorem, or BGV. This theorem was developed because the period of rapid expansion known as inflation that's believed to have taken place in the early universe violates one of the assumptions of the Penrose-Hawking theorem - that gravity is always attractive. Inflation is also thought to be eternal into the future, creating an infinite multiverse.

WILLIAM LANE CRAIG: In 2003, Arvind Borde, Alan Guth, and Alexander Vilenkin were able to prove that any universe which is on average in a state of cosmic expansion throughout its history cannot be infinite in the past but must have a past space-time boundary. What makes their proof so powerful is that it holds regardless of the physical description of the very early universe. So, in fact, the Borde-Guth-Vilenkin theorem does imply an absolute beginning of the universe.

ALAN GUTH: I was working with Arvind Borde and Alex Vilenkin to understand what we can learn about how inflation might have started and how far back it could have gone and in particular once we realized that inflation could be eternal into the future it seemed like a very natural question to ask – could inflation have also been eternal into the past? And what we found was that inflation could not be eternal into the past. What we basically managed to achieve was proving a theorem which says that any expanding region of space-time that has some minimum expansion rate can only go back so far and not infinitely far. So that means that inflation must have had a beginning. It doesn't really say that the universe must have had a beginning, but it says that the universe could not have been expanding forever up until the present time.

Response: Proving that “the universe could not have been expanding forever up until the present time” implies a beginning unless you can craft a tenable model according to which the universe was previously contracting or was static prior to expansion, which has not been done.

INTERVIEWER: So we interviewed Abhay Ashtekar. And you know in their model they have a bouncing universe. So with that, that means the universe could be eternal in the past, and that [the BGV] theorem doesn’t prevent that. Is that right?

ALAN GUTH: Yeah, that is right.

ALEXANDER VILENKIN: The theorem proves that inflation must have a beginning, right? The universe as a whole – the theorem doesn't say that. It says that the expansion of the universe must have a beginning, right? But it opens the door somewhat for alternatives.

Response: The BGV theorem has the condition that the universe be on average in a state of expansion throughout its history. Bouncing or cyclic models may not satisfy that condition and so are not precluded by the theorem. But as Vilenkin has repeatedly explained, such models face other insuperable obstacles that preclude the universe’s being past eternal. Again, I wonder why they didn’t ask Vilenkin about that?

NARRATOR: As we've seen, one way to avoid the BGV is to have a past contracting universe which is also a common prediction in multiple approaches to quantum gravity. But some suggest that this collapsing phase will be messy and unstable and therefore not be able to transition to our smooth homogeneous universe.

Response: Right, like Vilenkin, quoted above! Funny they never asked him to explain it to them!

34:23 AURÉLIEN BARRAU: We have absolutely no idea of what exists in the universe before the bounce. So it's very speculative and very difficult to decide whether the situation is stable or not. And even in the case where the situation is unstable, usually what it means is that the simplifications we use in standard cosmology cannot be used anymore. But, you know, nature doesn't care about the fact that it is hard to calculate. So when we say that something doesn't work, most of the time it means that the model we use to describe these things doesn't work anymore. For example, you cannot use a Friedman equation because the universe is so homogeneous that the Friedman equation is not usable anymore. So what? The universe doesn't care that we have to invent another equation. I think the correct way to think is not to go from the past to the future, is to start from what we know. What we know is what we see around us today. The universe is expanding. We have stars, we have galaxies, we have gravity. We have a theory. And then we go backward in time and we see what happened. And in our theory, when we go backward in time we do not end up with a singularity. We end up with a bounce.

Response: This is mere hand-waving, an appeal to the unknown. The bounce needs to result in the low entropy condition of our early universe. Vilenkin and others have shown that that cannot happen. So beginning from what we know, we go backwards in time and find that you cannot get here from there. And even if we could, postulating an earlier contraction does nothing to restore a past eternal universe.

NARRATOR: The BGV theorem relies on principles of relativity. For example, signals cannot travel faster than light. But this is problematic for Craig's Lorentzian relativity, which allows for faster than light travel. And with the authors of the BGV explicitly stating it does not prove the universe had a beginning, the kalam advocates have begun using yet another theorem developed by Aron Wall.

Response: The narrator can’t have her cake and eat it, too. If Lorentzian relativity is, as she thinks, false, then it’s no obstacle for the BGV theorem. Lorentz’s theory does not in any case imply that there are superluminal particles, nor does relativity preclude them, so long as they do not accelerate to achieve superluminal velocities.

35:59 ABHAY ASHTEKAR: Wall is assuming that some version of second law of thermodynamics is really fundamental and should be a fundamental ingredient in the series we determine whether there will be singularity or not. And I don't agree with that. I think that it really is something that would emerge in approximate sense in the semi-classical regime.

NIAYESH AFSHORDI: Generalized second law that was proven by Aron Wall for a quantum singularity, it assumes an infinite of space. And if you don't allow for infinities in your theory then you cannot take that theorem seriously either.

Response: Gosh, why do you suppose they didn’t go to the source and ask Aron Wall about this? So I did. He tells me that while his generalized second law of thermodynamics (GSL) implies an initial cosmological singularity only in an infinite space, nevertheless in a finite space at some time t in the past “the GSL would simply reduce to the ordinary second law of thermodynamics (OSL).” Then “you can still argue from the ‘ordinary’ second law of thermodynamics that most likely either (a) the universe had a geometrical beginning, or (b) the arrow of time must have reversed sometime in the past” (personal correspondence, January 27-28, 2022). In the latter case, one still has a beginning of the universe, since both time directions are future-oriented.

NARRATOR: Craig also quotes a paper by Anthony Aguirre and John Kehayias that he claims closes the door to a past-eternal quantum gravity era. But in most models the quantum gravity era is brief anyway. And the paper only referred to a particular model known as the emergent universe.

Response: I quoted them simply to show that a quantum state cannot persist changelessly from eternity past and then produce a new effect. What he says below confirms that.

36:58: ANTHONY AGUIRRE: But there are other possibilities, too, as to how the universe could be in accord with the Borde-Guth-Vilenkin theorem but could, I think, really be past-eternal. If you define a past-eternal model, you can; it self-consistently works but you do have an edge to it. You can ask what's beyond that edge and it's another copy of the same universe. So I think where the debate lies – past-eternal models exist. I think some of them are flawed like the emergent model. I think you can't really make a self-consistent model that stands up. Past-eternal inflation or the steady state model I think you can, but you have to do it carefully and there's an interesting then answer to what is this Borde-Guth-Vilenkin theorem pointing to. It's pointing to something very interesting; it's just not an initial time.

Response: Jim Sinclair has already responded to these proposals in his Blackwell Companion article. In short, it is more plausible to interpret the BGV theorem as Vilenkin does as implying a beginning of the universe than to try to save a past eternal universe by postulating two non-interacting universes at a past boundary or a reversal of time’s arrow at some point in the past.

NARRATOR: Just as there are theorems that prove the universe had a beginning, there are also theorems that imply the opposite.

37:58: JOÃO MAGUEIJO: I think you can prove anything. Theorems are very strange things because you can have a theorem to prove anything as long as you assume the right things. In a way that's a problem. You can violate any theorem as long as you violate the assumptions.

Response: Pardon me, but you can’t just prove anything. Your assumptions must be credible in light of the evidence. The BGV theorem rests on the single, unremarkable assumption that the universe is on average expanding throughout its history. If it’s not, then the theorem doesn’t hold. But in that case, as Vilenkin says again and again, you encounter other insuperable problems.

38:13: FRANCESCA VIDOTTO: Of course there are the singularity theorems but you can always interpret them as telling you that a quantum theory of gravity should not have one of the hypotheses that you are using to formulate this theorem.

Response: Fine, so bring on the quantum gravity models, like the Hartle-Hawking model or the Vilenkin model, both of which involve a beginning. Quantum gravity doesn’t restore an infinite past. As I expect will come up later, the question will be whether according to these models the universe can come into being from nothing, as both Hawking and Vilenkin suggest.

This brings to a close the discussion of the scientific evidence for a beginning of the universe based on cosmic expansion. Next time we’ll look further at the evidence of thermodynamics in support of a beginning.