What does the BGV theorem say?

My problem is that instead of plausability (which I have no problem with) you kept claiming that (emphasis mine):

These are saying that the BGV theorem is not only positive evidence for the beginning of time, but strong positive evidence for the beginning of time. This is very different from “plausible” which from the very beginning I have no problem with, as I kept saying, like in these two instances for example:

Also, before I forget, you will have to retract this statement:

As under your broadened version of “objectively confirmed”, there should be no more question that the statement

In general, theorems in classical physics cannot be used in the quantum regime
is “objectively confirmed”.


What little I had previously known about the BGV theorem came from this webpage by a CERN physicist and what I had heard from William Lane Craig. (Of course, while I greatly respect Dr. Craig, a Christian philosopher, I take what he says about physics with a grain of salt.)


I will reiterate what @dga471 said:

Evidence (in the Bayesian sense) is cheap, strong evidence is hard. Do you think the fact that “Most Americans speak English” constitutes good evidence that “Most Chinese speak English”?

Also, in the article @MStrauss wrote:

Other ideas about the origin of the universe like those proposed by Lawrence Krauss4 or Sean Carroll5 do not have real scientific evidence to back them up. They are conjecture.

This is unrelated to the point I am trying to make. Arguments against the BVG theorem being used to conclude that the Universe has a beginning has nothing to do with other theories of Universes without beginnings. Instead, it is fact that the BVG theorem is, by itself, sans any other theories of the Universe, not proven in the regime relevant for the beginning of the Universe.

1 Like

I may not have been clear, or may have gotten it mixed up somewhere in the discussion, but what I’m trying to argue is that BGV can be used as evidence in a cumulative case to infer a beginning.

I think before we go any further, though, I’d like to get some distinctions sorted out that I think effect the judgments (hopefully the correct term) of subjective inferences both in terms of plausibility and probability. This may be getting off topic again, but I think it is relevant to some degree since we’re talking at present about BGV in regards to its probability/plausibility. Speaking of which, I’m not sure exactly how probability and plausibility should be defined.

I would define plausibility as an inference in light of the availability of various types of evidence to something(?) that is within reason of what human experience and objectively confirmed reality generally would allow. Probability would be an inference as to how likely it is for something(?) to happen in light of the availability of various types of evidence. Does that sound right?

I think first I’d like to address the issue of built in assumptions that would affect the plausibility/probability judgments. For example, I would assume as a scientist in your probability judgments, because of MN, you are assuming naturalism is true. Now on my side, if I assume that naturalism is false, which I think I have ample good reasons to believe is the case, we start off with quite significant differences in our judgments right out of the gate.

The way I see it, particularly in conversations between scientists and nonscientists, I think the only fair position to take on both sides is a neutral position, since neither position can be proven true. That would apply to all assumptions, regardless of what underlying view of reality they are based on. That, I suggest, would provide a level playing field from which to base the initial judgment concerning the background knowledge. I think it would facilitate a much more robust discussion and avoid unnecessary misunderstandings.

Next, another issue arises with subjective inferences. Correct me if I’m wrong, but from what I can tell, virtual particles, their popping in and out of existence, violations of the law of energy conservation of energy, and quantum instability are inferences, I’m assuming probabilistic in nature, basically based on either indirect empirical evidence or mathematical proofs, i.e., uncertainty principal and observations of oscillation patterns.

Now I would say, between the two, empirical evidences carry more weight in that they are actual observances of actual reality or its effects. Whereas mathematical proofs (which I think would fall into the logically confirmed category?) are abstract in nature. So if an inference made based on a mathematical proof contradicts empirically confirmed evidence, I would say that the empirically confirmed evidence is a good indication that the inference based on the mathematical proof is wrong.

Now whether or not you agree with that could also make a difference in how I would set the probability factor of the inference, and how you would. So I guess what I’m trying to say is, to avoid misunderstandings, there’s more that needs to go into a fair representation of a subjective inference in these types of discussions in regards to its plausibility and probability than is necessary when doing such on a purely scientific, or nonscientific level. Does that sound like a valid point to you?

Plausibility is just that X is within reason, or even it is reasonable to think that X. Probability analysis is quantitative. Without probability analysis, we cannot say that X is more likely than Y. We can only say that X is within reason, or it is reasonable to believe that X is true, i.e. plausible. I have no problem with this. Do you really think that I believe that the Universe having a beginning is unreasonable? After all I kept saying that my personal believe is that the Universe do have a beginning.

Where does this enter the picture? I do not subscribe to MN (or at least some formulation of MN). I believe that science cannot detect the supernatural, but that does not mean that I subscribe to MN.

So is the BGV theorem. Also, it is unclear what you mean by “indirect empirical evidence”. Most of physics is “indirect empirical evidence”. This of course depends on what you mean by direct or indirect. Is LIGO detection of gravitational waves direct? Certainly a lot of analysis is required to obtain the LIGO result; the only thing “directly” measured are interferometric patterns from two lasers. Yet, the physics community agrees that their measurement is “direct”, and the team won the Nobel prize specifically for the “directness” of this measurement.

Also let’s not forget that I do not need to go that far to demonstrate that quantum mechanics reign supreme over classical physics (of which GTR is a subset). As I mentioned, all I need is to look into an oven and point to the fact that it does not explode via the ultraviolet catastrophe of classical physics. What is more of a “direct empirical evidence” than that?

Not always. This depends on the uncertainty in the observations of the empirical evidence versus the uncertainty of the observations that support the theoretical basis from which the mathematical proof flows from. The proper way to do this is to be agnostic about ranking empirical evidence vs inductive evidence vs abductive etc a priori and just plug everything into Bayesian inference.

Edit: Another thing, if we are talking about probability analysis (i.e. Bayesian inference),

Evidence, as I mentioned before, is cheap. What is necessary is strong evidence. Just any old evidence can be nearly worthless, as your conclusion (posterior) can be dominated by what you previously assumed is true (prior). In this case, the conclusion (the posterior) that the universe has a beginning is not changed significantly by the BGV theorem (the evidence), and is mostly driven by the fact that you already assume previous to knowing about the BGV theorem that the universe has a beginning (the prior).

1 Like

Thanks for that clarification. And I realize as a Christian you would believe in the beginning of the universe. I’m just a bit ignorant about the distinctions between probable and plausible. Thanks for your patients with me on this and on all my other ignorances in general. I do appreciate it. :slight_smile:

Glad to hear that. Science cannot detect the supernatural seems like a much more neutral position. If I’m following what you’re saying correctly I think you may be the first scientist I’ve come to realize doesn’t subscribe to MN. I’ve even dialogued with a YEC scientist, and an OEC scientist and they both defended MN in the context of doing science.

I’m trying to work this out. I’m not sure how to distinguish between something directly detected, and something directly observed. I guess detection would be in a sense direct access, but just very limited access. So I guess there would be degrees of being directly accessible, some providing a more complete picture than others.

But, when I say indirect empirical evidence, I essentially mean empirical evidence used to infer something that cannot itself be objectively verified. Empirical evidence used to infer possible explanations of how past events like OOL could or couldn’t have happened would be an example of this.

Though it might be more clear in some sense to just use the already available term of scientific fact, when I say direct empirical evidence, I mean an entity/event that is directly accessible to objectively verifiable observation or detection, and has been empirically confirmed by multiple observations or detections.

E.g., the existence of Pluto is an established fact, i.e., objectively confirmed through multiple direct observations using a telescope. There may need to be more distinction made here, but that’s the general gist of it.

To approach it from another angle, there are subjective inferences, and objective confirmations. Subjective inferences would be dependent on indirect empirical evidence and/or logical proofs. Whereas objective confirmations would be a result of direct empirical observations and/or logical proofs. In other words confirmations are established facts that are not susceptible to subjectivity.

I’m not real familiar with LIGO, but I would say a detection is direct access, but in the most limited sense. It provides a very small piece of the whole. Now LIGO may be described as a multiple instrumentation detection as opposed to a single instrumentation detection, like, say, a microscope.

But whether or not that distinction is necessary I’m not sure. Guess it would depend on the context. So with LIGO I would say if it is indeed shown to be accurate, it confirms the existence of CGW, but only provides a very limited picture of their existence.

Now I’m no philosopher, nor do I pretend to understand all of the scientific terminology, but I’m just trying to come up with a general way of making distinctions in order to be able to discuss these issues in a way that I, and maybe others like me, can follow and make sense of what’s going on in a discussion.

I’m not claiming this is the perfect way to go about the distinctions, and there are probably adjustments that would need to be made. It’s just an attempt on my part to provide a somewhat coherent way to keep a conversation on track.

I’m not sure I understand what you’re saying. I’m mainly concerned with established scientific facts and confirmed human experience which would mean the highest degree of certainty. Are you saying that we should be neutral to whether or not these facts could be overturned by subjective inferences?

I’m not quite sure I follow you here. I think you’re saying that in light of my prior assumptions, based on the evidence I’ve already considered, though BGV would add additional weight to support my conclusion, it wouldn’t significantly affect it. Is that correct?

@Jim, as I’ve shown by quoting the latest version of the KCA by Craig & Sinclair, they do not endorse using ONLY BGV to argue of the universe. Instead, they dutifully go through several popular quantum cosmology models (which are indeed supposed to apply to the beginning of the universe), and attempt to argue that a beginning is demanded even by these models. BGV is just one ingredient. They also have philosophical arguments for the beginning of the universe. Do you agree with C&S that all of these elements are needed to argue for the beginning, instead of BGV on its own? Currently, it seems to me that you’re defending a stronger version of the argument that C&S would not endorse themselves.

@PdotdQ can speak for himself, but I think that in general, scientists would be less disturbed if you supported the comprehensive argument of C&S, while also admitting that our knowledge of quantum cosmology is still relatively uncertain and undeveloped. Certainly less disturbed than if you tried to use the BGV as a magic bullet to argue for a beginning.

But that fundamentally makes no sense. Theories mechanistically integrate evidence. They aren’t evidence themselves.

I’m sorry. I seem to have given the wrong impression, or got mixed up about it somewhere along the line. I agree that BGV is just one aspect of a cumulative case for a beginning. I still think it provides more weight than I think scientists want to acknowledge. But on it’s own, I don’t think it’s sufficient to make the case.

By the way, I came across this interesting comment from Craig in his debate with Carroll. It pretty much shows that even the quantum regime would have had to have a beginning based on the current estimate of the age of the universe.


Edit: Additional quote. Think about it. If there is such a non-classical region, then it is not past eternal in the classical sense. But neither can it exist literally timelessly, akin to the way in which philosophers consider abstract objects to be timeless or theologians take God to be timeless. For this region is in a state of constant flux, which, given the Indiscernibility of Identicals, is sufficient for time. [13] So even if time as defined in classical physics does not exist at such an era, some sort of time would. [14]

But if the quantum gravity era is temporal, it cannot be extended infinitely in time, for such a quantum state is not stable and so would either produce the universe from eternity past or not at all. As Anthony Aguirre and John Kehayias argue,

It is very difficult to devise a system – especially a quantum one – that does nothing ‘forever,’ then evolves. A truly stationary or periodic quantum state, which would last forever, would never evolve, whereas one with any instability will not endure for an indefinite time. [15]

Hence, the quantum gravity era would itself have to have had a beginning in order to explain why it transitioned just some 13 billion years ago into classical time and space. Hence, whether at the boundary or at the quantum gravity regime, the universe probably began to exist.

Just dropping in to point out that there’s a difference between saying the universe had a first moment of time(it’s beginning), and saying the universe literally came into existence where before it didn’t even exist.

I assume you’re pointing to the possibility that the initial event may not have had a cause. Right?

The atheist (militant?) Sean Carroll does not seem to subscribe to MN (Dump the Metaphysics — How About Methodological Regularism? - #3 by Patrick)

For example, when the New Horizon team took this beautiful picture of Pluto, all that is directly detected are voltages on CCD chips (the same things that are in your phone cameras). Then, they need to use quantum mechanics to interpret these voltage readings to make the picture. In this case, is Pluto directly or indirectly observed? For more information on this problem in general, you can take a look at the Duhem-Quine thesis.

Correct, we should be neutral to “whether or not these facts could be overturned by subjective inferences”.

First, there are no such thing as scientific facts that cannot in principle be overthrown, even for those in the “highest degree of certainty”. After all, classical mechanics was thought to be in “highest degree of certainty” once.

Second, again you are a priori ranking “subjective inferences” as worst than “direct inferences”. This is not true in general. Every scientific inference has an error bar. What you might call “direct inference” will have an error bar from its observational errors. In the sciences, logical inferences flow mathematically from a theoretical foundation which are empirically confirmed, again with observations that have their own error bars. If there is evidence where the logical inference contradicts the direct inference, these two error bars have to “fight it out” to see which come out on top. I am not sure how else to explain Bayesian inference aside from this.

Not only it might not significantly affect it, I am not even sure that it affects it at all.

I think one should not even quote the BGV theorem. The problem is that I am not convinced that the BGV theorem is strong enough to warrant inclusion in this list of “comprehensive argument”. In this case, adding it into the arguments is just a Gish Gallop. It’s like saying “I believe that most Chinese people speak English”, because:

A) I did a survey of X Chinese people and found that Y% speak English
B) Oh, and also, most Americans speak English

Just use A) to argue and drop B).

First, just because we have not been clever enough to find a model that “does nothing ‘forever’”, does not mean that it is not true.

Second, no one claims that the quantum Universe will “[do] nothing ‘forever’” or “does not evolve”. It might, for example, oscillate or balloon back out in the negative time direction.

Third, this argument does not use the BGV theorem, which is the “evidence” under contention in this thread. Like I said, if you have better evidence that the Universe has a beginning, use that instead of the BGV theorem, unless you are interested in doing a Gish Gallop.

This is a good point. The beginning of time might be explained by physics, for example, in the quantum gravity models in which time itself is emergent and appears from some sort of “quantum soup”. However, this just pushes back the problem: how did this “quantum soup” exist?


No, I’m pointing to what I said. Coming into existence, and having a first moment of time, is not the same thing.

Not sure directly observed or indirectly observed would be so necessary to distinguish. I guess I kind of slipped that term in there in my explanation in my previous comment kind of unknowingly. I would say, however, another distinction could be made between unaided observation and instrumentally aided observation.

The former would generally be the more reliable, and the latter would have degrees of reliability depending on the type of instrumentation and how well it can be objectively confirmed. There should also be a distinction with detection methods as well regarding degrees of reliability.

OK. But my first question is, what is meant by direct inference? Would the meaning of a direct inference include an objectively confirmed unaided empirical observation as being a direct inference? Next, has any objectively confirmed unaided observation ever been overturned? And if so by what was it overturned? And certainly it seems in making a probability judgment an unaided objectively confirmed observation would be higher than a subjective inference.

And then in relation I would imagine the scale would go down for objectively confirmed instrumental observations relative to the degree of decreased reliability of the instrumentation. Detection would be similar although more limited in what is revealed by it than the latter.

And as for classical mechanics, it’s not clear to me whether or not you’re including GTR and STR here. You seemed to be including them in one of your previous comments. If you don’t include them, then as I understand it, there were only certain specific aspect of CM that needed to be changed. The rest still stands. Is that correct?

If so, I might somewhat agree with your point here with some qualification. However, as I mentioned in a previous comment, I’m mainly interested in empirical scientific facts, not so much objectively confirmed abstract theories and formulas.

Is it possible to make the case of a beginning without the BGV? Well, I would say it would be a significantly weaker argument. But speaking as a layperson, I think I still have to beg to differ with your assessment of BGV.

From what I’ve seen I think that it is sufficiently strong enough to be included. I think we disagree on certain subjective aspects in regards to how strong it is which, based on evidence and other factors presented to me, I believe I’m warranted to take an alternative position on. I think if you factor my position on those subjective aspects into it, there’s enough reason to rate it’s strength as at least sufficient, if not more than.

And since Bayesian inference, like all other inferences, is subjective in nature, I don’t see any problem coming to different conclusions based on the subjective aspects involved in each of our positions. And I think that since other scientists and scholars seem to differ with your assessment as well, it seems I’m at least be within the bounds of reason.

Well, I think what he was talking about is a model that can account for both being past eternal and evolving 13.8 billion years ago. That’s quite different than one that does only one or the other.

Again, I don’t think this is what’s being claimed by Craig. I think what he’s saying is, if quantum indeterminacy is assumed and going only by what is known in the objective sense at present, including the age of the universe, under those circumstances the only way a quantum regime could be without a beginning is if it did nothing from eternity past until 13.8 billion years ago, at which time it evolved into the universe.

So what he seems to be saying is, a quantum regime model would need to be formulated based on both of those conditions being met, i.e., past eternal doing nothing, and suddenly, at the initial physical event 13.8 b years ago, go from doing nothing to evolving into the universe. In his estimation, that would seem to be the only option available.

What is unaided observation? All observation is instrumentally aided, even your eyes are instruments.

Well it’s your words, what do you mean by direct inference? As I mentioned, in the sciences, it seems difficult to claim that anything is really “direct”, given that all observation in the end depends on auxiliary assumptions about our instruments.

Of course I am including GTR and STR. They are classical, as they are not quantized.

Before you claim that objectively confirmed = empirical, now you claim that you are not interested in objectively confirmed abstract theories and formulas…

I really don’t think so.

Here is the real difference between us: I have read the proof, you have not. It is really difficult for me to argue with you, it’s like we are arguing about the color of an apple in a painting, where I have seen the painting, and you have not.

Again, I am not even sure what you mean by subjective… Bayesian inference is just how scientists evaluate evidence, even those that are “direct”.

Within the bounds of reason = plausible. Did you miss when I say that I am fine with plausible? I am okay with plausible, though I think you should have a better argument than appeal to authority…

I don’t even understand how this is relevant to my counterargument…

This is nonsense physics.

If this is truly what he thinks, then his estimation is wrong, and to be quite honest, nonsensical physics. If you like appeals to authority, perhaps you should consider that Craig is not a physicist, so take his statements on physics with a grain of salt.

@Jim, I think it is clear from the tone of this post that I am tired of this argument. In the end, I think I will repeat:

You have not even read the proof, and you don’t have any background in physics. I can’t use the argument, for example, that past-incomplete timelike geodesics just means that

\frac{dx^a}{d \tau^2} + \Gamma^a_{\; bc} \frac{dx^b}{d\tau} \frac{dx^c}{d\tau} = 0

cannot be solved for all \tau \in (\infty,0], and thus the metric is at least C^2 inextendible, as those statements would mean nothing to you. Further, you do not seem to understand how modern scientists work in the framework of Bayesian inference.

Because of this, we are now going in circles and the conversation is no longer productive. This is not your fault but mine, I cannot explain these concepts to a layperson in a clear enough manner, and as someone who has a lot of interest in outreach, I take this failure personally.

I think I am going to pull out of this conversation. It was nice talking to you, @Jim, but I just don’t find it a productive use of my time and efforts anymore.

1 Like

I’m not sure that @Jim’s portrayal here is accurate or gives the full picture. Craig is referring to this paper in the debate with Carroll. I’m not saying Craig is right; what I’m saying is that in general he doesn’t make up physics by himself - he tends to try to pull from the existing literature.

I am aware of that paper (which by the way, is not actually quantum gravity, but semiclassical gravity), and I don’t actually think that Craig is making up physics by himself. I was just annoyed at @Jim’s portrayal of it when I was writing that post.

1 Like

OK. Sounds good. I got the impression at times that you were probably busy and sometimes weren’t able to fully focus on what I was saying from some of your comments. So I think that it’s a good time to end the discussion. I do appreciate all your time and patience with me. I know I can be a little annoying with my persistent questions and questioning. :slight_smile:

All the best!

All the best!

2 posts were split to a new topic: The Man’s Comments on the BGV theorem