> So if there are tons of Boltzmann brains then you should think you’re probably a Boltzmann brain. But if you think that then you shouldn’t trust your reasoning, because most Boltzmann brains have defective reasoning, so the belief is self-defeating. Additionally, Boltzmann brains die after a few seconds, so if you notice yourself not dying quickly then you know you’re not a Boltzmann brain.
This is a pretty terrible argument. You have no evidence that you're not a Boltzmann brain. You can't "notice yourself not dying" because every Boltzmann brain would also not notice dying, since noticing yourself dying instantly is impossible. So you have no evidence that distinguishes your current experience from such a brain.
Your argument just boils down to "I'm going to take as an article of faith that I'm not a Boltzmann brain, therefore god", which is not a serious logical argument.
Secondly, you are gerrymandering what counts as an arbitrary constraint. According to you God is not arbitrary because its "perfect goodness". A multiverse is arbitrary because it requires lots of randomness. You give no criteria as to what is "simpler". You also appear to be saying that *any* constraint is "arbitrary", which obviously favors god because once you define every constraint as arbitrary you essentially stipulate there must be an unconstrained being.
What is the source of these definitions of simplicity or arbitraryness? You have never provided one in all of your blog posts. The categories are simply made-up!
This was basically my take on this piece. We can more-or-less grant the soundness of the argument chain up until the multiverse bit (though I am not nearly well-versed enough in inflation cosmology to do readily), but then it starts to fall apart. If the process required for generating multiple universes is complex, then so is a process for generating a benevolent God. A claim that God somehow is immune from the questions that can at the same time be valid in regard to the multiverse is not a self-evident assumption at all to me. Why "a God is necessary", but not "the multiverse is necessary"? Better yet, why not grant a possibility for the thesis "the universe is necessary"? With expanding scientific knowledge of early cosmology, it may be that eventually this can be steel-manned it to find a sound objection for the fine-tuning problem. "God did it" historically hasn't served us that well when we lacked a scientific explanation for an elusive phenomenon.
Very well written. I've studied the science side of the fine-tuning argument a fair bit. I find it very impressive. You've helped me to see that the more conceptual-philosophical side is also impressive. I'm very glad that you've come to believe in God. It is not unlikely that a God that created a universe with creature like us in mind would want to communicate with us. You're getting closer to Christianity. I know there are still a lot of options and questions and issues. But you're getting closer.
My biggest issue with this is "If there is a God, then the universe being finely tuned makes sense." For the probability of us being fine-tuned to be high under theism, that deity must have a certain set of contingent desires. The set of possible worlds a deity could create is massive: it could create no universe, no life, a world that our brains can't make rational sense of, etc. The probability of being in a fine-tuned universe like this under theism seems just as improbable to me as under naturalism.
Now, if you want to claim these properties of your deity are not contingent, then idk why we can't just claim the same under naturalism. We can just claim this universe is necessary. Both seem just as ad hoc to me.
We also have very good independent reasons to believe God isn't the Triple O god. It seems to me the problem of evil brings the priors close to zero. If God was perfect, not only is the existence of things like child cancer and factory farming deeply implausible, but the very existence of humans indicates this type of God doesn't exist. If God was perfect, and this meant he wanted to create lots of life, he wouldn't create humans of all things - if anything, he would create an infinite number of beings that are the ontological equivalent of God (which I guess rules out monotheism). At the very least, he would create beings a lot better than humans. It's also implausible he created so much life lower than us, most of which experiences nothing but suffering and then death. And if we would expect him to create infinite life, why is the universe so devoid of life? Why did he tune a universe to be nearly entirely inhospitable to life? If perfect God existed, we would expect a much better "tuned" universe, with vastly more life in it. We would also expect a maximum amount of life - which we clearly don't have, just on earth we could have many more people. Why didn't we just start with the earth at the maximum capacity of humans? Do all possible people exist, just not at the same time? Does God maintain like a vat of souls with all possible people in it somewhere? Maybe, but this seems to just get increasingly and ridiculously ad hoc.
Anyway, if we can rule out a perfect deity, which I think we can, then we're left with a deity that we have no reason to believe would care about life, making universes, making life capable of understanding their reality, etc. In which case, its existence doesn't alter the probability of us existing.
For this objection to work you have to think the odds of a deity creating a finely tuned universe is on the order of 1/10^10^123. That seems uber implausible.
I agree evil is a problem but I think theists can explain it, and God is just much likelier than a limited God that we should take the probabilistic hit and still believe in it.
It seems uber implausible because you're not taking into account the various commitments of the Christian faith. When you go to heaven or hell, apparently all that's needed is your soul. God can create any number of universes with various metaphysical foundations - and the Christian faith already takes it as a given that there are alternate realms called heaven/hell with different laws than our own. Why wouldn't God place us in a non-finely tuned universe to show off his metaphysical prowess, or put us in a relatively structureless universe as the only composite beings, or in a world governed by any of the other infinite metaphysically possible laws he could instantiate? Christianity doesn't offer any convincing reason to expect God to create a finely tuned universe vs a non-finely tuned universe. (Also, God apparently blundered most of the physical space and timeline of this universe by making it uninhabitable to physical life. What are the odds of that happening again by an omnipotent being?)
“If anything, he would create an infinite number of beings that are the ontological equivalent of God (which I guess rules out monotheism).” Actually, this is pretty close to the classical Christian account. Cf. Hart, You Are Gods.
As a Catholic with a PhD in physics, I really want to be convinced by something like this. But I can't. My main comments are (apologies if this is a bit fragmented and may be circular at times; life forced me to type this in quick installments):
1) This is not a crucial point, but the cited examples of fine-tuning in initial conditions become completely unmysterious in the framework of inflation (currently the most mainstream early universe paradigm). Some versions of inflation (inspired by chaotic inflation) also give rise to multiverses of the type discussed here. I don't think there is one specific model which solves the global problem, but the existence of these classes of models should raise big questions about the likelihood of overestimating the strength of fine-tuning arguments.
2) I don't think you can dismiss the anthropic principle argument that easily. And I suppose a lot of this point is semantic since really the whole thing is taken a lot more seriously in the multiverse section. But really the multiverse arguments are a special case of the anthropic principle argument, and therefore are strictly weaker, not stronger. The key insight from the anthropic principle is that "the probability of design given fine-tuning", "the probability of design given the existence of life", and "the probability of design given fine-tuning and the existence of life" may be roughly the same. Crucially, they don't need to be a priori in all versions of the argument; but in the cases where they are the observation of fine-tuning doesn't add anything to the likelihood of design. It's not a refutation of design, but it can be a powerful negation of the evidence from fine-tuning. I guess unless you're using this to re-evaluate you estimate for one of these probabilities based on the rest.
3) The argument for Boltzmann brains cannot be so easily defeated by "noticing yourself not dying quickly". Brains have memories, and even Boltzmann brains could have information corresponding to coherent memories spanning decades. Logically, I think perhaps the way to dismiss Boltzmann brains lies in investigating the probability of such a brain having coherent memories (is this analogous to noticing you are dreaming? maybe!). Intuitively, it seems to me the same as dismissing classical solipsism.
4) The bits about discoverability seem either equivalent to the usual fine-tuning or even more vulnerable to anthropic-style reasoning (of course all the particles we've discovered are discoverable; it's not like anyone thinks we've discovered all the particles there are). Maybe the critical challenge here is describing what a fully undiscoverable (as opposed to partially discoverable, like ours may be) universe with intelligent life would look like so we can think about how much more likely one is than the other. Maybe there is something convincing once that work is done, but right now I don't buy it.
5) There is something to be said about the logic of "multiverse is more likely with design than without, therefore multiverse raises the probability of design" and then connecting that to the anthropic principle argument with multiverse. I still don't like it - both because it's not clear to me that multiverse solutions have to be that high a percentage of solutions to the fine-tuning problem (because of things like the point above about fine-tuning in initial conditions) and because it would seem like most multiverse solutions could also imply things like potentially infinite Boltzmann brains or local universes in constant suffering (which is more of an aesthetic argument than a logical one, I guess).
1) See the section about the possibility of more fundamental laws that fix the constants. That pushes up fine-tuning a level to the being finely-tuned to produce such laws!
2) I'm not really tracking what you're saying. Do you agree that your existence can be evidence for a hypothesis that makes your existence more likely?
3) But most Boltzmann brains would die quickly because they'd freeze in outer space.
4) It's not that the particles we've discovered are discoverable, it's that they're ideal for discoverability. We could still know about them but not use them to have a broader picture of reality. That's overwhelmingly more likely, as Collins argues.
>1) See the section about the possibility of more fundamental laws that fix the constants. That pushes up fine-tuning a level to the being finely-tuned to produce such laws!
That's the thing. Inflation is the perfect counter-example to this. It is _the_ mainstream formalism for the very early universe, it completely solves three or four (depending on how you define them) fine-tuning problems of the (now old) Standard Model of Cosmology, and the main problem it does have is the exact opposite of a fine-tuning problem: it is too easy to write an inflation model which solves all of these problems! Worse: the simplest imaginable class of inflation models (slow-roll inflation), complicated models involving fun things like brane-worlds and superstrings, and even models postulating macroscopic numbers of new scalar particles with randomly distributed masses are all consistent with all available observations (including high-precision Cosmic Microwave Background observations). This is definite proof that there exist ways to "fix" fine-tuning problems without pushing the problem up a level. (And in this case it's not even exactly a moe "fundamental" theory doing the heavy lifting: merely an earlier stage in cosmic history.)
>2) I'm not really tracking what you're saying. Do you agree that your existence can be evidence for a hypothesis that makes your existence more likely?
I do. But that is true without fine-tuning problems too. My point is that the anthropic principle argument tells you that the fine-tuning problems of modern physics aren't that much stronger evidence than St Thomas Aquinas' observation that he existed (and, in fact, should be seen as equivalent to biblical "arguments" around God ordaining seasons or day/night cycles or whatever at just the right timings). It is not surprising that most possible laws of nature make life impossible.
>3) But most Boltzmann brains would die quickly because they'd freeze in outer space.
Yes, but no Boltzmann brain would notice their death. The only thing they would experience is the illusion of remembering decades of experiences in a physical world - just like you and I do! Meaning we can't use "not noticing sudden death" as evidence that we are not Boltzmann brains.
>4) It's not that the particles we've discovered are discoverable, it's that they're ideal for discoverability. We could still know about them but not use them to have a broader picture of reality. That's overwhelmingly more likely, as Collins argues.
I find it hard to argue that neutrinos are in any way "ideal for discoverability". They interact only through half of the known fundamental forces and can only be detected by enormous detectors to make up for the tiny probability that they will interact with any given particle.
But also, we do have good reason to believe there are fundamental particles we haven't observed, and no guarantee all of them are even discoverable. Gravitons were the first thing that popped in my mind, but actually dark matter may be a plausible better candidate: it could very well be that dark matter is entirely composed of "gravitational neutrinos" which interact with regular matter only through gravity. Such a particle would be almost impossible to detect.
So I'd say even if you ignore neutrinos there are too many holes in the Standard Model to definitely make this claim.
But really maybe the most interesting ramification of this discussion is how much money we should be willing to bet on it. Currently there are discussions about building a "super-sized LHC" to try and learn more about physics beyond the Standard Model (which we know for a fact is incomplete because of things like its inability to account for the fact that neutrinos have masses). One big argument against it is that there is no reason a priori to expect that new physics to show itself at higher energies that humans can probe without converting the solar system into a particle accelerator. How much money should we be willing to invest into such a new accelerator if our only goal is to discover new particles/physical phenomena? (To be clear, I expect to answer here. I just thought of it and thought it would be interesting food for thought. It is not meant as a knockdown argument fo anything.)
(To be extra clear, if we did build that new LHC and that helped us formulate a consistent Grand Unification Theory that explains all fundamental particle physics phenomena known to man, then I'd very much buy this argument about discoverability. It's just the number of undiscovered stuff can be anything between ~1 and infinity.)
1) Inflation is not a counterexample it's just a description of how it occurs. Inflation may fix the parameters to be finely-tuned, but one could imagine all sorts of other theories that fix the parameters to be other things. So then it's surprising that we got inflation that is such to produce fine-tuning.
2) I'm including the surprisingness of the world containing life at all as part of fine-tuning. But the additional point that the laws are fragile favors theism--the odds of that are super low on naturalism, not necessarily super low on theism.
3) They would notice as they start to die, which would take a few seconds!
4) I'm not informed about the physics, but I'd recommend you look, in more detail, at Collins's work.
1) That is not how inflation (or its alternatives) work. The thing is that the key to solving these specific problems was to notice that the fine-tuned initial conditions we wanted to explain actually correspond to a natural "attractor" for where you expect the Universe to be after relatively simple types of evolution which can be easily realised by many different specific models. In very simplified terms, those fine-tuning problems can be summarised as "the Universe looks to be extraordinarily flat, yet relativity tells us that a Universe filled with radiation and/or regular matter should evolve away from flatness, so it would have needed to be astronomically close to perfect flatness in the "initial conditions"". Inflation basically just says "maybe one of a million possible things made the very early Universe expand super fast, so that even if it was super curved before the patch of the Universe we can observe is very close to flat (because any manifold is locally flat by definition), and also look when you make the Universe expand like that you generally turn tiny quantum fluctuations into macroscopic ones and that happens to match CMB observations". The fact this works so well for inflation is due to the fine-tuned values actually corresponding to a very specific geometric situation (flatness) which can arise in multiple different plausible ways. You could not do this with any random value for those initial conditions.
2) I may need to consider this point more carefully when I'm less sleepy. I'm always afraid this sort of argument is too sensitive to measurement or basis problems, if that makes sense to you.
But also, can the point about fragility not be made with medieval physics - the Earth-Sun distance is fine-tuned and your only naturalistic solution to this is postulating a "multiverse" of countless Suns and Earths out there and then invoke the anthropic principle? Are we actually adding any evidence for theism by considering particle physics?
3) How would they notice starting to die? Presumably most Boltzmann brains are more aggregates of information than anatomically correct nervous systems. They need not be connected to any external "sensors" that may trigger survival "alarm bells"?
4) I'd like to say I'll do that. Instead, I'll say I will open a tab and hope I get round to it!
My atheistic-ish answer to this kind of question, which I playfully but sincerely expound at https://outlandishclaims.substack.com/p/universes-arent-real , is that the multiverse requires no explanation. God does, though, doesn't He? Isn't his only advantage as a prime mover the way that, by convention, we write about Him like we can assume Him into existence? Even if you accepted the fine-tuning argument 90% of the way, God still seems like too specific of an answer to be reachable a priori. Why not an entity wanting to maximize suffering, rather than happiness? Why not a committee of 12 mystical entities with different preferences, of which our universe is the uneasy compromise?
You couldn't have posted this before I subscribed, I guess :(
I did say in my subscription message that I wanted to be challenged. I did not say that I wanted to get recycled Christian apologetics, a topic on which I have wasted far more than enough time on already (not with you, obviously).
I'm not aware (and I _did_ just perform a quick check) of any converts from cradle* atheism/agnosticism to any theistic religion (Islam, _theistic_ Judaism, Zoroastrianism :), Hinduism, Bahaism, cryptotheistic Buddhism, ...) _other than_ Christianity who
a) had some sort of fairly strong scientific background; and
b) after their theistic conversion/awakening/whatever you want to call it, take up apologetics and evangelism for their theistic worldview.
* I mention cradle agnostic/atheist in view of the example of people like C.S. Lewis, who was raised in lukewarm CoE Christianity, which experience I suspect implanted latent theistic epistemic viruses that later became virulent :(
Leah Libresco Sargeant is the best example convert from cradle atheism I can think of, more an embryonic moral philosopher, but she did at least understand coherentist epistemology in a Bayesian framework, etc. - I did buy and read her conversion experience book, which I considered quite honest, albeit poorly reasoned. I'll leave it at that.
G.E.M. Anscombe, apparently ?) raised in an agnostic household, while not any sort of scientist, converted to Catholicism in high school. Interestingly her capsule biography states that one of her brothers became a clergyman. I have not read anything about the backstory of her conversion, but given her fondness for teleology and Aristotelianism via Thomism, she probably wound up in the right place for her. It would have been interesting to hear a colloquy between her and a philosophy who strongly rejects Platonism and its works (I'm agnostic in a utilitiarian way re mathematical Platonism, and intensely, virulently :) anti-Platonist regarding ontology, epistemology, etc.).
There is Christian B. Anfinsen, shared 1972 Nobel in Chemistry, who publicly converted to Orthodox Judaism (coincident with his second marriage at age 73). Eight years into that marriage he wrote "my feelings about religion still very strongly reflect a fifty-year period of orthodox agnosticism." That fifty year period weakly suggests to me, perhaps incorrectly, that his childhood was spent in a Lutheran milieu (his parents were Norwegian immigrants who taught him the Norwegian language and culture); at any rate that he did not identify as agnostic until he was an adult.
My standard functional atheist answer to the strong anthropic principle (versus to the weak anthropic principle, for which see mostly 2) below) is:
1) It is at best a remarkably poor metacognitive strategy, at best actively dishonest (generally IME in the direction of self-deception) given known human biases and failings, to apply the label "God", with all its historical connotations and opportunities for motte and baileying, to some hypothetical low Kolmogorov complexity cause of the fine tuning.
On the very rare occasions when I have attempted to engage with actual theists (generally coming out of some sort of broadly neoThomist tradition, often Ed Feser fans of one sort or another, but that's very likely sampling bias at work on me; there's long explanation for that sampling bias that I will spare you) I suggest we use Ghod** - "generalized hypothetical organizing device" as a heuristic to remind us all to bracket the TBD degree of overlap between Ghod and the God who is typically hypothesized to be remarkably interested in the specific doings of some Iron Age pastoralists in the southwest Levant some three thousand years ago. (More extra fine tuning presumably to achieve that result :( )
** not my invented term, but I _am_ trying to signal boost it.
2) Given that we postulate some sort of Ghod ***, our evidence for the aforesaid Ghod being whatever level of fine tuning we can charitably consider to be "not special pleading", e.g. something about the value of the asymptotic fine structure constant, roughly 1/137, sure - we could call that fine tuning since we have relatively little reason to think that it has _not_ been the same everywhere and everywhen since the universe's temperature dropped below around 1 GeV [1] ; re Christian apologists, well, I've read a little of [multiple expletives deleted] Alexander Pruss' work and your merely citing him pushes my inclination to discount you to roughly as much as I would discount a convert to Trumpism;
I had never hear of Collins that I recall, a quick check of Collins' CV posted at, sure enough, Messiah University, tells me I'm not going to bother with him either. (interestingly, the link on his homepage to his "statement of faith" is broken, but his bibliography makes it clear that his advanced reasoning powers pretty much only operate when they lead him toward orthodox Christianity and its axioms).
*** I'm trying to phrase that hypothesis such that the posterior probability of that weakly defined Ghod is strictly greater the posterior probability of the God in which Pruss and Collins have faith.
Anyway, given a weak anthropic principle fine tuning Ghod along the lines of Einstein's
Einstein - G(h)od does not play dice with the universe.
Bohr - stop telling G(h)od what to do.
and allowing for a very weak hypothesis that Ghod (cf theism as opposed to a real watchmaker deity unconcerned with the doings of humans or other equivalently cognitively capable species elsewhere) has opinions regarding what we humans ought to be doing or not doing, how do you propose an implementable search strategy for a very finite human (e.g. yours truly with probably around 20-25 years life expectancy, though with the last ten of those probably experiencing significant cognitive decline)?
I propose to limit that search strategy to operate on evidence that was at last hypothetically available to extremely well equipped time travelers, with all of today's scientific knowledge and instruments, but none of today's theological inclinations, looking around say four million years ago (before the split of genus Homo from Hominina, so including the proto-australopithicenes, but well after the split of the ancestors of chimps and bonobos from our line).
At least _most_ of the fancy fine tuning that theists point to as leading to the pinnacle of all creation, namely humans born recently enough to know Jesus Christ, had occurred by five million years ago (as evidenced by anatomical signs of increased bipedalism; I'm being deliberately vague re causation of the fast increase in encephalization in genus Homo. but I *think* there's a reasonable consensus but not dogma that bipedalism's uptake came first - I haven't studied the literature of the last fifteen or so years in any detail).
I don't know whether Pruss, Collins, or any of the other theists (where their theism is based on pure reasoning from fine tuning) thought those distant ancestors of ours were equipped with the necessary material substrate in the brain to have the Sensus Divinitatis. I have not come across any statements either way (those statements would have to have been made in the last fifty odd years, when we really started getting a moderately filled in fossil record).
(I wonder if Collins studied directly under Plantinga at Notre Dame? The timing makes it entirely possible, but Collins' CV just says that he got his PhD in philosophy in 1996; Plantinga was still giving at least seminars at ND as recently as 1995).
[1] re the fine structure constant, yeah, I know in a broad sense about Webb et al 1999 inferences from some quasars' spectral hyperfine structure - and also Lamoreaux and Torgerson re isotope ratios from Oklo - to suggest that alpha might in fact be slowly decreasing over time. I'm agnostic on that one.
And the asymptotic value only comes meaningfully into play at moderate energy scales, e.g. < 1GeV.
I'm sorry that you haven't liked the content since your subscription :(. As for your objections, well this isn't Christian apologetics--I'm not a Christian. It's just an argument for God. Arguments like this are what have convinced me!
Your comment is a bit long, so I won't address everything, but just to address a bit: what's wrong with inferring God? I think God is very simple and explains lots of stuff. I also don't think that simplicity is calculated by Kolmogorov complexity.
Pruss is crazy smart and a good philosopher, though I disagree with him a lot.
No worries - a lot of the previous content was excellent (e.g. Cancel Culture Is A Tax On Being Interesting, Losing Faith In Contrarianism, There Aren't Naive Views) hence subscription. I'll start with the assumption that the fact that this was the very first thing you posted after I subscribed is approximately an accident (not truly random any more than a coin toss is, but the causal networks associated with your developing oeuvre at your substack interacting with my process of deciding it was time for me to start contributing $ is dense and hidden enough that we can treat it as a random variable with an unknown distribution).
Yes, you have made quite clear that you are not a Christian at this point (cr Steelmanning Christianity, which I have not read yet; it's on my reading list). Did not think you were. Just that your two primary sources (Pruss and Robin Collins) are not merely believing Christians (in e.g. the Francis Collins sense, and I mean no slight on F. Collins' faith thereby) but #Christians# or whatever typographical emphasis you prefer.
In response to Jacob's response to me (thanks for upvoting it), challenging me to read more of Pruss and some of Collins, I plan to give it a shot in the case of Pruss, who has some publications in math domains where I have enough knowledge to not be totally lost in the first few paragraphs. In "There Aren't Naive Views", you described Pruss' IQ as approximately one billion (greater than Ralph's? It's unclear whether Pruss fits under the rubric of a person you know in the sense that you know Ralph; I infer that he does not and is at least arguably smarter than Ralph based on your knowledge of the two).
What article of Pruss' (ideally public access in one form or another) would you recommend to me as best demonstrating his goodness as a philosopher (by your lights, obviously)? It's certainly possible that there is something significantly new under the sun in the neoThomian world, which I last (lightly!) perused more than a decade ago. His work would be more accessible to me in demonstrating his quality as a philosopher if it did not start with the assumption of the rightness of his interpretation of Catholicism (I have no reason to doubt that he does very well in arguments with other Catholics over the finer points or implications of various Catholic teaching where he and his interlocutor(s) concur on the basics; I'd be too ignorant of some of the terms of art in that discussion to detect misuse via equivocation, etc., although I'd probably be reasonably objective :) ).
I just found Pruss' very recent (2023 update) "ACCURACY, PROBABILISM AND BAYESIAN UPDATE IN INFINITE DOMAINS" which is going to require some mastication and cross reference checking, but it looks promising. No evident feedforward from theism to that paper, although I could see how he might use that result in some of his scaffolding.
The first like half of this reply is completely irrelevant. Why does it matter what happened to Leah Libresco Sargeant, among others, in a reply to BB’s carefully thought through argument for theism from fine-tuning? Matthew isn’t a Christian apologist, he’s an honest truth-seeker.
As far as I can tell, your first real criticism is that there is no reason to infer God from fine-tuning because it does not tell us about his perfection and other actions and characteristics, so we should only infer to “Ghod”. That’s very true, and would be a good criticism if BB hadn’t directly addressed this. He argues that a perfect God is a better explanation than some other arbitrarily limited explanation because of its higher prior probability due to its simplicity (it posits a single thing - perfection, and has high coherence, in Draper’s terminology). BB is not inferring directly to God from fine-tuning, he infers to Ghod and then argues that the particular form of Ghod that is the simplest and best explanation is a perfect God also because this explains other things, like psychophysical harmony and anthropic evidence.
Also, calling citing Alexander Pruss Trumpism is utterly strange. I trust Alexander Pruss way more than I trust your vague gesturing towards some objection or other, maybe especially because you hadn’t heard of one of the leading philosophers working on fine-tuning, Robin Collins, despite your claim to have wasted more than enough time on the topic - waste more! Worse still is discounting Robin Collins because of where he works instead of engaging with his argument.
Thanks for the reply, Jacob! - I look forward to seeing what content you eventually place at your substack (check shows just the placeholder post right now, but I thank you for the other two recommendations you offered that I don't already read, namely Emerson Green and Majesty of Reason). I assume you *don't* object to my subscribing in advance? (no problem at all if you would prefer me to stay away, and no need to explain to me why you don't want me - it's your substack)
A couple of things back at you - I hope you will not mind my taking your points somewhat out of order - I prefer to find some of the areas of agreement first, plus I wish to reassure you that I read your response multiple times and took it both seriously and literally :). If you don't have much time to waste on me, perhaps just skip to [4] at the end, which I think contains the meat of what I suspect is an unbridgeable methodological chasm between you and me.
1) Your point that I should waste more time on the fine tuning argument has some validity according to my epistemology and decidedly incoherent sense of intellectual honesty. Collins' first publication on the fine tuning argument was published in 2003 as part of Manson "GOD AND DESIGN The teleological argument and modern science", and (as my cites to Feser and Pruss might indicate), I have not spent any time to speak on the more recent forms of what look _so far_ to me like the pretty much the same arguments.
Hooray - I found a library loan e-copy of the book! I make no guarantees about *when* I will get to it, but Collins' section is only about 22 pages - looking at the book's TOC and seeing sections in there from Swinburne, William Lane Craig, Dembski, and Behe, where the subheadings to their sections suggest nothing new from any of those four people, where the leaves me immensely skeptical a priori (I note that the volume does include four pages of a skeptical view in the sixty pages devoted to "fine tuning" derived from physical cosmology, the broader section that includes Collins).
I failed on my first try (admittedly quick and shallow) to locate some of Collins' other content as listed on his CV - I will make another try at it. So thank you again for the challenge.
2) Re g(h)od and general arguments from fine tuning, strong anthropic principle, etc.
I suspect there are some pretty interesting (probabilistic) arguments from generally agreed upon facts in cosmology, physics, etc. It seems plausible that there are some people that don't proceed fairly quickly from those facts to theism (or at least a very turbocharged deism, cf Einstein) have written about their tentatively drawn conclusions, although how much operational utility those conclusions might have is another story (conclusions to Christian, Islamic, Zoroastrian, etc. theism, or to any form of theism that somehow implies that the theistic deity has opinions/values/preferences/whatever about what we lowly humans do in our lives and perhaps pre-lives, afterlives, ... are another matter and do at least point toward operational utility). But I have missed those non-theistic thinkers. I freely admit I have not looked very hard.
If you, presumably having read more deeply and *recently*, know of anyone who stopped at deistic find
I don't agree _at all_ with what I can gather of the stated reasoning from g(h)od to God, but I have nothing new to offer to the well hashed counterarguments to Matthew's steps, counterarguments which he asserts that he has considered and rejected (the whole perfection thing has been thrashed over and over, generally in the context of some flavor of the ontological argument; Plantinga's modal version of the ontological argument being a fairly recent that I know some non-theistic philosophers have engaged with, notably "Chinese room" Searle). My response was long enough already :) and I don't see any likelihood of any progress for anyone there.
3) My apologies to you and to Alexander Pruss (as noted in my comment, some of whose work I _have_ read, but nothing too recent - I intend to correct some of that; I do plan to give Collins a bit of a try as noted above); I have no idea whether Pruss is or is not a Trumpist, nor do I particularly care (I suppose I would mildly prefer that he not be in the interest of, err, diversity within his intellectual circles). I meant rather (and *thought* I said) that my prior probability on the expected value of (re-)engaging with Pruss' content (at second hand) is rather low, as it generally is regarding engaging (on different topics, of course, with Trumpists).
You did inspire me to look (via his CV) at some of Pruss' more recent work in math and math adjacent, nontheological philosophy (last time I did that was maybe 2008 or so?, and it was argued as I recall from very intensely Christian theistic premises). A couple of papers there look interesting (e.g. "“Independent tests and the log-likelihood-ratio measure of confirmation” and “Regular probability comparisons imply the Banach-Tarski Paradox”; maybe also "“Non-classical probabilities invariant under symmetries” although looking at the rest of his work in that field, he seems to be talking about probabilities that are hyperreals or orders rather than what I find rather more interesting and potentially useful, probabilities that are locally approximatable as tensors) although many of their titles, e.g. “The badness of being certain of a falsehood is at least 1/(log 4 – 1) times greater than the value of being certain of a truth” cause me to want to run very far away from someone who speaks in terms of truths and certainties (for that matter, of falsehoods as well, unless his certainties and truths and falsehoods are shorthands for probabalistic judgements).
4) One area, highly relevant to Pruss [based on my reading of him, again possibly out of date] and relevant to Collins [based on inferences from titles in aggregate bibliography as stated in his CV and a weak inference from where he works; I did make an admittedly brief attempt to locate a couple of his later works as open source on the Internet; I will try harder - Collins has one quite early paper that looks like it might be quite interesting], where I think we disagree (at least, so I infer: I request, if you are willing to spend the time, to hear the respects in which you believe my supposition regarding your view of my approach is substantially wrong)
I don't discount Collins simply because he works at Messiah; I discount him based on probabilistic inferences from the combination of the sum total of his CV titles and journals and the faith statement of Messiah as a whole, to which I rebutably assume Collins assents.
I disagree, quite strongly, with the idea that it is fallacious** to reason thus:
[start bracket
I have relatively high confidence that Person A reasoned poorly in domain A1 from fact set A2 and according to reasoning methods A3.
Therefore, it is valid, probably mandatory, to reduce my prior distribution on the expected validity and utility of person A's reasoning in domain B1 from fact set B2 according to reasoning methods B3, where the magnitude of the adjustment depends on the overlap between the respective domains etc.
For a working example to jump let's say 100 milligodwins into the argument in one bound, if someone asserts that the 2020 election was stolen from Trump via [insert preferred methodologies], yes, that person is going to have to possess incredibly strong pre-2015 credentials in virology, epidemiology, etc. for me to give *that person's arguments* (either pro or con) re the lab leak hypothesis any attention. Life is too short.
n.b. I consider the lab leak hypothesis plausible but probably wrong. If you have not seen the rootclaim/slatestarcodex analysis of the lab leak that published recently**, I highly recommend it as an example of applied Bayesian analysis on contentious, politically/emotionally loaded claims that are in principle falsifiable to high probability given easily imaginable facts not currently in evidence
in case links are disapproved here or seen as spamming, google "slate star codex rootclaim leak" should hit at #1
end bracket]
Does that specific approach, which is indeed fallacious in classical logic** have a specific name? It seems like it should have one, but a couple of quick scans have not found one. It's arguably a subset of the genetic fallacy or an interesting permutation of the appeal to false authority.
** I find non-probabilistic logic to be of very limited use in cases like the example of argument from fine tuning.
this has already gone far far too long, and assuming you read to the end, I thank you for your patience and indulgence.
If you wish to reply, please provide some general parameters of how you would like my reply to your second reply constrained (e.g. a maximum word count :) ).
Thank you very much for the reply! You seem to have put a lot of work into the response and also seem to overestimate me a bit 😅 - I was just replying as I thought Matthew might respond. I haven’t actually read Pruss or Collins or any of these people firsthand either, I just know what they argue.
I will do my best to respond to your points, probably tomorrow, assuming there is some disagreement left over after I read what you’ve written.
No worries. FTR I have read some of Pruss directly (also Swinburne, Craig, Dembski, and Behe), although as I noted not all that recently. Also, unlike cosmology, physics, and any *serious* acquaintance with some of the mathematical fields that Pruss has published in**, I do have (also sadly dated now :) ) formal training and domain knowledge in biochemistry, which makes it far easier for me to reject Dembski and Behe with extreme prejudice based on some of their earlier writings :)
i.e. Pruss could *easily* BS *me* on some of those math papers and I would be none the wiser - higher math (anything past say the second year of real analysis) makes my head hurt, and I admit defeat from my multiple attempts to grok it. I can of course look at duelling papers in the fields that Pruss is publishing in, but the ugly truth is that I'm way out of my league there and I know it.
Don’t worry you can say whatever you want in your replies it’s just that if they’re longer it will take me longer to get back to you :)
Okay so the first point is thank you! I don’t know if I’ll ever end up writing any content of my own, but if I do I’d be glad to hear your thoughts.
Emerson Green and Majesty of Reason are great, but Substack isn’t their main platforms. I HIGHLY recommend their YouTube channels, especially Majesty of Reason. MoR honestly levelled up my standard of philosophical thought and knowledge immediately he is a literal genius and he makes such amazing videos with inhuman amounts of effort put in.
2) As far as I can tell, two philosophers that defend deistic hypotheses on teleological argument grounds are Paul Draper (who defends aesthetic deism, I think) and Antony Flew (who famously converted to deism after being a leading atheistic philosopher).
Is your objection that perfection isn’t actually simple? Or that it’s impossible or something? I don’t really know specifically what criticisms to Plantinga you are alluding to.
Also, I think the reason there are not that many deists as a result of fine-tuning might just be because the existence of a powerful and good designer is strong evidence that one of the religions that exists is probably true because that would be a very strong correct prediction.
4) I agree that you can infer that someone having a bunch of false beliefs is evidence that another one of their beliefs will be false, but I still think it’s unfair to make the argument you made from Pruss’ flawed reasoning in one area to flawed reasoning in another area. This is because you haven’t actually read Pruss’ reasons for being a Catholic, so you don’t know if his reasoning is flawed. Additionally, there might be some reason for the statement of faith - maybe that’s just something their workplace required or something. This kind of inference is useful if you can’t understand a person’s arguments or don’t have the time to research them, but I just think that skipping the higher-order evidence and engaging with the philosophers’ work directly is just a better approach and it lowers the risk of bias.
“which experience I suspect implanted latent theistic epistemic viruses that later became virulent”
I like how you put this because this is exactly how jesus described the kingdom of heaven in the gospels!
“The kingdom of heaven is like unto a grain of mustard seed, which a man took, and sowed in his field: which indeed is less than all seeds; but when it is grown, it is greater than the herbs, and becometh a tree, so that the birds of the heaven come and lodge in the branches thereof.”
“Another parable spake he unto them; The kingdom of heaven is like unto leaven, which a woman took, and hid in three measures of meal, till it was all leavened.”
Well, if we are going to be quoting gospels at each other, I'll give you Luke 15:3-7 :) **
** YMMV, but I think that's a rather better version of the parable of the lost sheep than Matthew 18:12-14
It should be unsurprising that my favorite Scriptural passage is Job 38:4-7 (NIV)
4 “Where were you when I laid the earth’s foundation?
Tell me, if you understand.
5 Who marked off its dimensions? Surely you know!
Who stretched a measuring line across it?
6 On what were its footings set,
or who laid its cornerstone—
7 while the morning stars sang together
and all the angels shouted for joy?"
My conditional atheist view, that informs my unoriginal "latent theistic epistemic viruses" phrasing, is that the human brain is, if you will, predisposed aka overly sensitive toward detecting agency, which leads fairly easily to broad sense supernaturalism such as you see in shamanistic religions (Tengrism, for example), and thence, particularly with growing societal complexity, to X-theism, where X is initially poly- or heno-, with the monotheistic variant showing up in the wild somewhere between 4000 and 3000 years ago (some of the earliest layers of the Zoroastrian Gathas; my amateur read is that the evolution of the Judaean religion from henotheism to post-Exilic monotheism is less than 3000 years old, but:
1) I'm not deeply read in the sources (for which I would need to know several different early Iron Age West Semitic languages, as well as at least a couple of flavors of Aramaic and probably Akkadian); and
2) I have no particular religious or nationalistic attachment to any particular chronology re the United Kingdom (of Israel and Judah, not of Great Britain and Ireland :) ) and its predecessors.
I am very much interested in the (secular) evolution of post-Exilic Judaism and Christianity, but again to have anything useful to say there without simply outsourcing my thinking would require far more commitment to the languages, sources, and scholarship than I am prepared to commit - serious hats off to the scholars who do it.
Oh yeah those are all good verses from the bible! I too find biblical scholarship very remarkable, I recently watched this presentation which was very enlightening. I do recommend it. https://youtu.be/ay_Db4RwZ_M?si=59F_ua6rFp5P9Pvb
After your extremes swings of opinion, a meta position would be “I am agnostic”. Our understanding of reality as a consequence of the fundamental laws of physics is so developed that have reduced chemistry and biology to the point where we synthesize life.
In around 400 years we have pushed theism and teleology from the universe. Multiverse is a little bit more for Marvel comics than for philosophy.
Still, in my view the theistic hypothesis is quite likely. But more like 40 per cent likely than 99 per cent likely…
You shall have a distribution on that parameter between 0 and 1, not a point estimate. The mean of that distribution can be 60 to you and 40 to me, but in any case, uncertainty on that estimate is large, as you have learned by its recent change, from less than 1 per cent to around 60 pc.
For example for “probability of heads or tails” in almost any coin I have had 50 per cent more or less all my life. Your estimates of the probability of something shall not be only numbers.
Have you read Bostrom on the Anthropic bias? I suppose it is the classical reference on this debate. I give small relevance to the argument on the universal constants. The existence of consciousness and the simple fact that reality exists is reason enough to consider a conscious cause to the universe (God) plausible. We really don’t now anything about the “meta laws” of nature (the space state where the parameters are chosen)
I dislike the whole enterprise, because (the same as in conscience) we ask science to go beyond its borders: the reduction of reality to physics. While we have done this in a surprisingly effective and complete fashion, then we begin to ask about “why this laws of physics?”
You cannot ask science anything beyond what the Laplace demon can answer. Consciousness and multiverses are obviously beyond the “Laplace demon frontier”.
There is not separation between “life” and “non life”, so we treat cell machinery as machinery and re arrange it. The most important application so far are closed cell membranes used for drug delivery at cell level, but other cell machinery are included in synthetic life products.
Your argument here, to clarify, is that there is no difference between things that are alive and things that are dead?
I understood that you were claiming that we can, without any living tissue, from dead chemicals, and nothing else, create fully functioning living self replicating organisms. As someone who studied medicine this is mistaken and not possible.
Of course there not difference. Life is simply a state of matter, as solid, liquid, gas or coloild.
In practice I seriously doubt anybody cares about not using “living” tissues in synthetic biology. But what Venter does is live beings micro engineering. The cell membranes are made out of lipids, but regarding organelles I am not sure to what extent they are taken from other cells.
You wrote, "Our understanding of reality as a consequence of the fundamental laws of physics is so developed that have reduced chemistry and biology to the point where we synthesize life."
Craig Venter is the most famous person in the development of synthetic biology. The final proof of reduction of biology to physic and chemistry. No elan vital, only atoms.
I love Venter's work. He synthesized life in the form of taking an existing life form, and modifying it quite a bit. I just wanted to make sure that you did not imagine that any scientist today is capable of starting with material not from living cells and creating a new living being. Its amazing that quite a few people think that.
Venter claims, based on both theory and his experiments, to have produced not only a simpler (less DNA base pairs) form of life, but something approximating the simplest possible form of life (at least of anything resembling any type of life we have ever found). In fact, not long after Venter's findings were published, I wrote this blog post explaining how they provide evidence for an Intelligent Creator of life:
For the time being let’s put the past aside. What I claim is that life is mostly reduced to physical processes. We cut paste, change and manipulate it in a similar fashion that we manipulate machines. There is not trace of any supernatural force in the machinery of life.
I'm not denying that life (certainly simple life like bacteria - creatures with consciousness like we have are another matter) can be reduced to physical processes. Life is, as you alluded to, like machines. In fact, even the simplest life (like Venter's simplest cell) contains many molecular machines. Like our human-made machines, these machines in cells consist of many parts which must be of specific shapes and sizes and made of material with specific properties and put together in specific ways in order to perform the functions they perform (you can google and find animated videos of cellular machinery - it is fascinating!). In other words, they have a high level of complexity that is functional.
But machines require designers. They require designers because just as the right rocks and minerals would never on their own happen to assemble together in the right forms to produce a working car engine or an iphone, neither is it reasonable to think they could come together to form a working cell. And cells can do something that no human machine can yet do. They can use material in their environment to reproduce themselves! All this points to an Intelligent Designer being involved.
Is there anything in existence you believe not to be designed by a designer? If not, then you're not successfully articulating a boundary we can use to determine what is designed and what is not designed. If so, why would life be something special that needs to be designed? We've produced a myriad of various works in the physical sciences that detail how existing biological stuff can successfully self replicate, and how the complex systems that biological stuff forms can interact with each other in mutually beneficial, mutually sustaining ways.
Also
>just as the right rocks and minerals would never on their own happen to assemble together in the right forms to produce a working car engine or an iphone
But they do assemble to form caves, rock formations, strata detailing hundreds of millions of years of geological history, planets, asteroids, mountains, and so on. These are a few samples of emergent features that previously people thought would need Gods to e.g. sculpt the mountains, that we now recognize can occur without any conscious agent intervening in the process, due to entropy, coarsegraining, statistical mechanics more generally. I don't think there's any epistemic space left for "And then God intervenes!" when we create models to explain these various features.
I think many things, from the fine-tuning of the universe for life, to the origin of life, indicates that an Intelligent Designer (whom I believe to be God) was needed in the past. Recognizing this will help us to have a better foundation for thinking about the present and the future.
I’ve never been fond of the multiverse hypothesis to “disprove” agentic fine-tuning because it’s fundamentally a faith position. It’s invoking an indefinite (infinite?) number of unobserved (unobservable?) universes to account for observed facts. Where are the receipts? Where are the observations that give us reason to think that this is how the world is, rather than how atheists would *like* things to be?
Why would a Creator God bother with all the others if our finely-tuned universe is the one desired? Was there a lot of trial and error? Was he incompetent or just dicking around?
I think a simple answer is multiple universes certainly require a lot fewer assumptions and assertions about the world than creating some extra natural entitity(s). Much less a specific one.
One the one hand it seems odd we are here. On the other hand...where else could we possibly be?
Yes this is unsatisfying. So lets grant that this particular bundle of rules is in some sense "unlikely".
(Though I am not sure there is really any evidence for such a claim because we don't really have any evidence that any of the values we believe are "fine tuned" could be different. That they could be different is just an assertion with no evidence.)
Ok so here we are in a setup with some unlikely rules. What are some explanations for that?
Well one obvious possible explanation is there are multiple setups. So that we are in an "unlikely one" is less surprising.
It does not seem to me at all obvious that a Christian or other god(s) is preferable to this explanation, involves less assumptions, or frankly really solves anything at all since you immediately have questions about them.
Even assuming we are all Sims in some alien's/god's little toy, that god presumably has all the same questions about metaphysics in one way or another. That alien/god found probably find their existence fine tuned as well.
As a final aside, the fine tuning argument sems less and less strong the longer Fermi's paradox holds up. The idea that the universe is in some way "fine tuned" for such a small peripheral phenomena as life, seems pretty silly when it is 100% FOR SURE, not the main thing going on. Most of the mass, and the vast majority of the energy is doing stuff almost entirely unrelated to life.
In fact the universe is in such a state we seem less like some being's carefully crafted and curated sims, than a giant procedural random simulation that was just fired up and is producing a tiny weird recursive bug that is about to crash the program.
There's widespread agreement among physicists that there is cosmological fine-tuning. But what one should infer from fine-tuning is a philosophical question, which physicists aren't especially geared to evaluate. If you look at most of what they say about why they don't believe in God, it's nonsense about, for instance, the god of the gaps.
> But what one should infer from fine-tuning is a philosophical question
No, it is a factual question. I infer from the fact that an object is hotter than the things around it, that it was heated. I infer from the fact that you wrote a bunch of words that you are a human. Whether God exists or not is not a philosophical question, it is a factual one. If God descended from the Heavens and said "Yo, I'm God, here's some proof of me being God [performs physically impossible miracles]," then we would need no philosophy to conclude that he was real. The fine-tuning argument is exactly this: if you look at the laws of physics, the logical inferrence from them is that God exists.
More than anyone else, a physicist would be inclined to be capable of evaluating if the inferrence "God exists" follows from "there is cosmological fine-tuning." And it appears they do not evaluate it that way.
If we scale the issue back to our more immediate predicament, like the ratio of oxygen in the atmosphere, the mass of the earth, accessible water, orbit around the sun etc, we might ask the same question and assume fine tuning must be involved. But then we look a little beyond this scope, to other world - all these planets, with their own solar systems, and all is barren.
The existence of many other planets and solar systems is highly analogous to the theory multiple universes, only that the former is certainly real. Among all of theses worlds, ours is the only one that struck gold with all the cosmic and planetary parameters necessary for life. Given so many worlds, it serves as a very storage evidence for us to be the result of a statistically-improbable-though-possible coincidence, instead of the creation by an intelligent entity.
I agree that a multiverse is quite a complex thing to exist, but then so is a god.
I would like you to separate your arguments more and put probabilities on them. To me you say why the multiverse is unlikely but not why it is impossible and to me a god looks pretty unlikely too. So the real comparison is between unlikely options.
>Fourth, even if a multiverse explains why we exist, it doesn’t explain fine-tuning for scientific discovery.
This argument fails because believing you've discovered something is a psychological property of conscious agents. I will demonstrate this by example: I think you have a mistaken view on scientific realism - you are most likely a scientific realist and think science discovers the objective structure of the world. I think that's totally wrong, and that there is always a disjunction of available methods, tools, and models, at our disposal when conducting scientific investigation, such that there is never a last word, a final investigation, or a point where we reach "the world" and say "That's it, we've reached our target!". We construct the scientific program, we don't discover it. This means I don't affirm that we ever make scientific discoveries in the realist sense.
Yet I have no trouble accounting for your belief that e.g. we discovered electrons. You are in an epistemic position such that you believe there are electrons capable of being discovered, and that we have discovered them. In other words, there's no need to posit, "And lo and behold, there are electrons, for otherwise Matthew would not come to believe in them." Your believing there are electrons is not necessarily grounded in there really being electrons - just in your memories of what scientific realists have said and in your ontic beliefs.
In the same way, discoverability isn't a property of the world, it's a property of what epistemic agents believe. And epistemic agents can be so constituted that they believe they discover something brand spanking new about the world when something as pitiful as e.g. an electron whizzes by, or time passes, or they recall or hallucinate a new memory. In other words, God doesn't have to set up the world so that it's "really" discoverable - he just has to set up the world so that epistemic agents eventually develop beliefs that they discovered something. (I'm still unclear on why God would even care about this though.)
>the objective chance of a coin coming up heads is .5 because as the number of coins you flip approaches infinity, the share of them that will come up heads will approach half.
This isn't true, there's a non-zero chance that the coin will quantum tunnel through the Earth and never land anywhere, flying through the rest of the universe unimpeded. So you might think, "Oh, well I'll just build that into the objective probability calculation!" Except you also need to build in that the coinflip happens in a sufficiently powerful gravitational field - flipping a coin off an asteroid could again result in it flying across the universe unimpeded, never landing heads nor tails. And you also have to build in that the coin is indestructible - since each time you flip it, you cause parts of its mass to flick off - but now you have to find some physically possible way to create an indestructible coin. You also have to account for conscious agents flipping the coin in order to get a desired result, which would bias the final probability, controlling initial conditions so the wind doesn't blow it away forever, creating a perfectly flat coin so it can't land on its side, and so on. Whatever you end up considering the objective probability of flipping a coin to be, it would seem to be something inscrutable. Consider this a metacritique of your metaphilosophy - you deploy concepts like "objective probability" in idealized contexts and assume that all the conceptual difficulties that come with applying them to nonidealized contexts will just spontaneously resolve themselves.
There isn't a coinflip where nothing fishy happens. There's all sorts of physical constraints that inform our theoretical modeling of coinflips, and all sorts of physical concerns that will make an accurate coinflipping model sensitive to contextual cues. If you want to make a purely theoretical model, then it will fail when you try to apply it, since it doesn't account for the nuance that the accurate coinflipping model builds in by being attentive to relevant conditions.
For example - I stipulate a rule that I am right about everything. What does that mean? Well, you say "That's right" whenever someone says something true, right? And you use it to affirm truths like in the previous sentence, and so on. So I just start babbling, 1+1=3, there are and there are not colorless green ideas that sleep furiously... Do you still understand what it means to say that I am right about everything? What do you mean no? Just interpret it as in the nonfishy cases where someone is right about something! Clearly, the propositions I babbled and what I told you to do while interpreting me will not just work themselves out into peaceful equilibrium.
I like Jonathan Weisberg's objection. We already knew a priori the laws were life-permitting from the fact of our own existence; thus, what we discovered is not that life has a small probability of existing on the laws given naturalism, but that the laws are stringent rather than lax with respect to small variations permitting life or not. But it's not more probable that God would prefer stringent over lax laws. So fine-tuning isn't evidence for God.
Weisberg goes on to argue that this reasoning does not depend on the chronology of our evidence, i.e., even if we'd hypothetically discovered the stringency of the laws before our own existences (a possibility which I think makes no sense, but ignore that), the objection would still go through. Because fine-tuning is only evidence for design insofar as it's evidence that life actually exists. Once you learn the latter, the former evidence is screened off.
Also, this does not depend on assuming the prior on our own existences is ~1, even though I think that should also be regarded as true, and the standard firing squad arguments against this fail.
PS. Although I don't think this is necessarily fatal for you since you can use other constants, I don't think the fine-structure constant is fine-tuned. To suggest as much relies on using a uniform probability distribution - perhaps even over an infinite interval, which is problematic in itself. There's more motivation, by contrast, to using a non-uniform maximum entropy distribution, and when you do this, IIRC the probability of the life-permitting range for the fine-structure constant is relatively high.
I think there's a not terrible probability God would make stringent laws--so that we could see his existence--while conditional on atheism the odds are super low. Therefore, stringency favors theism over and above mere existence.
It's not at all obvious that the probability of stringent laws on atheism is super low, relative to lax ones. Remember, we're not talking about "stringently life-permitting laws vs. any set of laws whatsoever, life-permitting or not," we're talking about "stringently life-permitting laws vs. lax-ly life-permitting laws." Most of the lax but still life-permitting laws are going to have obscenely high complexity (if you don't believe me, try thinking of some!), or else they'll maybe be low-complexity on their own but rely on obscenely highly complex initial conditions (this is really a semantic issue of whether you consider the initial conditions to count as part of the laws or not). So they'll have very low priors.
That might be right. But the probability of atheism+lax laws is super low, so then if the probability of atheism+stringent laws is similar, then the probability of life-permitting laws on atheism is super low!
Maybe P(atheism+lax) is super low. But if it's comparable to P(atheism+stringent), and P(God+lax) is comparable to P(God+stringent), then the laxity/stringency of the laws aren't strong evidence for either theism or atheism. So this isn't really about fine-tuning at all, it's just whether the existence of life-permitting laws of any sort are evidence for God.
Even if that's true, I think the presence of life-permitting laws are super strong evidence for God. But this is extra evidence because it can both serve to clarify how unlikely life-permitting laws are--surrounded by the sea of non-life-permitting laws--and because if theism has any decent chance of getting stringent laws--which it does, because God would want to show fine-tuning--that favors theism massively. You admit the odds of laxity+atheism are super low, so if the odds of laxity+theism are not super low then laxity+theism is massively confirmed.
>Even if that's true, I think the presence of life-permitting laws are super strong evidence for God.
It's possible, but now you can no longer rely on the crazy-high likelihood ratios that fine-tuning arguments are supposed to be able to deliver, as in the ~10^10^123 figure you've named in other comments. So you can't necessarily outweigh a low prior on God.
Personally, as I've argued elsewhere, I believe theism is a vastly more complicated hypothesis than anything remotely in the vicinity of contemporary physics or whatever. So something like "contemporary physics + atheism" will have a much higher prior than God, and it will also predict that the laws are life-permitting!
> So if there are tons of Boltzmann brains then you should think you’re probably a Boltzmann brain. But if you think that then you shouldn’t trust your reasoning, because most Boltzmann brains have defective reasoning, so the belief is self-defeating. Additionally, Boltzmann brains die after a few seconds, so if you notice yourself not dying quickly then you know you’re not a Boltzmann brain.
This is a pretty terrible argument. You have no evidence that you're not a Boltzmann brain. You can't "notice yourself not dying" because every Boltzmann brain would also not notice dying, since noticing yourself dying instantly is impossible. So you have no evidence that distinguishes your current experience from such a brain.
Your argument just boils down to "I'm going to take as an article of faith that I'm not a Boltzmann brain, therefore god", which is not a serious logical argument.
Secondly, you are gerrymandering what counts as an arbitrary constraint. According to you God is not arbitrary because its "perfect goodness". A multiverse is arbitrary because it requires lots of randomness. You give no criteria as to what is "simpler". You also appear to be saying that *any* constraint is "arbitrary", which obviously favors god because once you define every constraint as arbitrary you essentially stipulate there must be an unconstrained being.
What is the source of these definitions of simplicity or arbitraryness? You have never provided one in all of your blog posts. The categories are simply made-up!
This was basically my take on this piece. We can more-or-less grant the soundness of the argument chain up until the multiverse bit (though I am not nearly well-versed enough in inflation cosmology to do readily), but then it starts to fall apart. If the process required for generating multiple universes is complex, then so is a process for generating a benevolent God. A claim that God somehow is immune from the questions that can at the same time be valid in regard to the multiverse is not a self-evident assumption at all to me. Why "a God is necessary", but not "the multiverse is necessary"? Better yet, why not grant a possibility for the thesis "the universe is necessary"? With expanding scientific knowledge of early cosmology, it may be that eventually this can be steel-manned it to find a sound objection for the fine-tuning problem. "God did it" historically hasn't served us that well when we lacked a scientific explanation for an elusive phenomenon.
"your argument just boils down to..." is a neat trick. 😉
Very well written. I've studied the science side of the fine-tuning argument a fair bit. I find it very impressive. You've helped me to see that the more conceptual-philosophical side is also impressive. I'm very glad that you've come to believe in God. It is not unlikely that a God that created a universe with creature like us in mind would want to communicate with us. You're getting closer to Christianity. I know there are still a lot of options and questions and issues. But you're getting closer.
:)
My biggest issue with this is "If there is a God, then the universe being finely tuned makes sense." For the probability of us being fine-tuned to be high under theism, that deity must have a certain set of contingent desires. The set of possible worlds a deity could create is massive: it could create no universe, no life, a world that our brains can't make rational sense of, etc. The probability of being in a fine-tuned universe like this under theism seems just as improbable to me as under naturalism.
Now, if you want to claim these properties of your deity are not contingent, then idk why we can't just claim the same under naturalism. We can just claim this universe is necessary. Both seem just as ad hoc to me.
We also have very good independent reasons to believe God isn't the Triple O god. It seems to me the problem of evil brings the priors close to zero. If God was perfect, not only is the existence of things like child cancer and factory farming deeply implausible, but the very existence of humans indicates this type of God doesn't exist. If God was perfect, and this meant he wanted to create lots of life, he wouldn't create humans of all things - if anything, he would create an infinite number of beings that are the ontological equivalent of God (which I guess rules out monotheism). At the very least, he would create beings a lot better than humans. It's also implausible he created so much life lower than us, most of which experiences nothing but suffering and then death. And if we would expect him to create infinite life, why is the universe so devoid of life? Why did he tune a universe to be nearly entirely inhospitable to life? If perfect God existed, we would expect a much better "tuned" universe, with vastly more life in it. We would also expect a maximum amount of life - which we clearly don't have, just on earth we could have many more people. Why didn't we just start with the earth at the maximum capacity of humans? Do all possible people exist, just not at the same time? Does God maintain like a vat of souls with all possible people in it somewhere? Maybe, but this seems to just get increasingly and ridiculously ad hoc.
Anyway, if we can rule out a perfect deity, which I think we can, then we're left with a deity that we have no reason to believe would care about life, making universes, making life capable of understanding their reality, etc. In which case, its existence doesn't alter the probability of us existing.
For this objection to work you have to think the odds of a deity creating a finely tuned universe is on the order of 1/10^10^123. That seems uber implausible.
A perfect being is likely if theism is true because it's simpler, less arbitrary, solves the psychophysical harmony problem https://benthams.substack.com/p/for-theism-part-1, and explains the anthropic data. https://benthams.substack.com/p/the-anthropic-argument-for-theism
I agree evil is a problem but I think theists can explain it, and God is just much likelier than a limited God that we should take the probabilistic hit and still believe in it.
It seems uber implausible because you're not taking into account the various commitments of the Christian faith. When you go to heaven or hell, apparently all that's needed is your soul. God can create any number of universes with various metaphysical foundations - and the Christian faith already takes it as a given that there are alternate realms called heaven/hell with different laws than our own. Why wouldn't God place us in a non-finely tuned universe to show off his metaphysical prowess, or put us in a relatively structureless universe as the only composite beings, or in a world governed by any of the other infinite metaphysically possible laws he could instantiate? Christianity doesn't offer any convincing reason to expect God to create a finely tuned universe vs a non-finely tuned universe. (Also, God apparently blundered most of the physical space and timeline of this universe by making it uninhabitable to physical life. What are the odds of that happening again by an omnipotent being?)
“If anything, he would create an infinite number of beings that are the ontological equivalent of God (which I guess rules out monotheism).” Actually, this is pretty close to the classical Christian account. Cf. Hart, You Are Gods.
How did you implement "we have no reason to believe"? It seems impossible in 2 ways to me.
As a Catholic with a PhD in physics, I really want to be convinced by something like this. But I can't. My main comments are (apologies if this is a bit fragmented and may be circular at times; life forced me to type this in quick installments):
1) This is not a crucial point, but the cited examples of fine-tuning in initial conditions become completely unmysterious in the framework of inflation (currently the most mainstream early universe paradigm). Some versions of inflation (inspired by chaotic inflation) also give rise to multiverses of the type discussed here. I don't think there is one specific model which solves the global problem, but the existence of these classes of models should raise big questions about the likelihood of overestimating the strength of fine-tuning arguments.
2) I don't think you can dismiss the anthropic principle argument that easily. And I suppose a lot of this point is semantic since really the whole thing is taken a lot more seriously in the multiverse section. But really the multiverse arguments are a special case of the anthropic principle argument, and therefore are strictly weaker, not stronger. The key insight from the anthropic principle is that "the probability of design given fine-tuning", "the probability of design given the existence of life", and "the probability of design given fine-tuning and the existence of life" may be roughly the same. Crucially, they don't need to be a priori in all versions of the argument; but in the cases where they are the observation of fine-tuning doesn't add anything to the likelihood of design. It's not a refutation of design, but it can be a powerful negation of the evidence from fine-tuning. I guess unless you're using this to re-evaluate you estimate for one of these probabilities based on the rest.
3) The argument for Boltzmann brains cannot be so easily defeated by "noticing yourself not dying quickly". Brains have memories, and even Boltzmann brains could have information corresponding to coherent memories spanning decades. Logically, I think perhaps the way to dismiss Boltzmann brains lies in investigating the probability of such a brain having coherent memories (is this analogous to noticing you are dreaming? maybe!). Intuitively, it seems to me the same as dismissing classical solipsism.
4) The bits about discoverability seem either equivalent to the usual fine-tuning or even more vulnerable to anthropic-style reasoning (of course all the particles we've discovered are discoverable; it's not like anyone thinks we've discovered all the particles there are). Maybe the critical challenge here is describing what a fully undiscoverable (as opposed to partially discoverable, like ours may be) universe with intelligent life would look like so we can think about how much more likely one is than the other. Maybe there is something convincing once that work is done, but right now I don't buy it.
5) There is something to be said about the logic of "multiverse is more likely with design than without, therefore multiverse raises the probability of design" and then connecting that to the anthropic principle argument with multiverse. I still don't like it - both because it's not clear to me that multiverse solutions have to be that high a percentage of solutions to the fine-tuning problem (because of things like the point above about fine-tuning in initial conditions) and because it would seem like most multiverse solutions could also imply things like potentially infinite Boltzmann brains or local universes in constant suffering (which is more of an aesthetic argument than a logical one, I guess).
1) See the section about the possibility of more fundamental laws that fix the constants. That pushes up fine-tuning a level to the being finely-tuned to produce such laws!
2) I'm not really tracking what you're saying. Do you agree that your existence can be evidence for a hypothesis that makes your existence more likely?
3) But most Boltzmann brains would die quickly because they'd freeze in outer space.
4) It's not that the particles we've discovered are discoverable, it's that they're ideal for discoverability. We could still know about them but not use them to have a broader picture of reality. That's overwhelmingly more likely, as Collins argues.
>1) See the section about the possibility of more fundamental laws that fix the constants. That pushes up fine-tuning a level to the being finely-tuned to produce such laws!
That's the thing. Inflation is the perfect counter-example to this. It is _the_ mainstream formalism for the very early universe, it completely solves three or four (depending on how you define them) fine-tuning problems of the (now old) Standard Model of Cosmology, and the main problem it does have is the exact opposite of a fine-tuning problem: it is too easy to write an inflation model which solves all of these problems! Worse: the simplest imaginable class of inflation models (slow-roll inflation), complicated models involving fun things like brane-worlds and superstrings, and even models postulating macroscopic numbers of new scalar particles with randomly distributed masses are all consistent with all available observations (including high-precision Cosmic Microwave Background observations). This is definite proof that there exist ways to "fix" fine-tuning problems without pushing the problem up a level. (And in this case it's not even exactly a moe "fundamental" theory doing the heavy lifting: merely an earlier stage in cosmic history.)
>2) I'm not really tracking what you're saying. Do you agree that your existence can be evidence for a hypothesis that makes your existence more likely?
I do. But that is true without fine-tuning problems too. My point is that the anthropic principle argument tells you that the fine-tuning problems of modern physics aren't that much stronger evidence than St Thomas Aquinas' observation that he existed (and, in fact, should be seen as equivalent to biblical "arguments" around God ordaining seasons or day/night cycles or whatever at just the right timings). It is not surprising that most possible laws of nature make life impossible.
>3) But most Boltzmann brains would die quickly because they'd freeze in outer space.
Yes, but no Boltzmann brain would notice their death. The only thing they would experience is the illusion of remembering decades of experiences in a physical world - just like you and I do! Meaning we can't use "not noticing sudden death" as evidence that we are not Boltzmann brains.
>4) It's not that the particles we've discovered are discoverable, it's that they're ideal for discoverability. We could still know about them but not use them to have a broader picture of reality. That's overwhelmingly more likely, as Collins argues.
I find it hard to argue that neutrinos are in any way "ideal for discoverability". They interact only through half of the known fundamental forces and can only be detected by enormous detectors to make up for the tiny probability that they will interact with any given particle.
But also, we do have good reason to believe there are fundamental particles we haven't observed, and no guarantee all of them are even discoverable. Gravitons were the first thing that popped in my mind, but actually dark matter may be a plausible better candidate: it could very well be that dark matter is entirely composed of "gravitational neutrinos" which interact with regular matter only through gravity. Such a particle would be almost impossible to detect.
So I'd say even if you ignore neutrinos there are too many holes in the Standard Model to definitely make this claim.
But really maybe the most interesting ramification of this discussion is how much money we should be willing to bet on it. Currently there are discussions about building a "super-sized LHC" to try and learn more about physics beyond the Standard Model (which we know for a fact is incomplete because of things like its inability to account for the fact that neutrinos have masses). One big argument against it is that there is no reason a priori to expect that new physics to show itself at higher energies that humans can probe without converting the solar system into a particle accelerator. How much money should we be willing to invest into such a new accelerator if our only goal is to discover new particles/physical phenomena? (To be clear, I expect to answer here. I just thought of it and thought it would be interesting food for thought. It is not meant as a knockdown argument fo anything.)
(To be extra clear, if we did build that new LHC and that helped us formulate a consistent Grand Unification Theory that explains all fundamental particle physics phenomena known to man, then I'd very much buy this argument about discoverability. It's just the number of undiscovered stuff can be anything between ~1 and infinity.)
1) Inflation is not a counterexample it's just a description of how it occurs. Inflation may fix the parameters to be finely-tuned, but one could imagine all sorts of other theories that fix the parameters to be other things. So then it's surprising that we got inflation that is such to produce fine-tuning.
2) I'm including the surprisingness of the world containing life at all as part of fine-tuning. But the additional point that the laws are fragile favors theism--the odds of that are super low on naturalism, not necessarily super low on theism.
3) They would notice as they start to die, which would take a few seconds!
4) I'm not informed about the physics, but I'd recommend you look, in more detail, at Collins's work.
1) That is not how inflation (or its alternatives) work. The thing is that the key to solving these specific problems was to notice that the fine-tuned initial conditions we wanted to explain actually correspond to a natural "attractor" for where you expect the Universe to be after relatively simple types of evolution which can be easily realised by many different specific models. In very simplified terms, those fine-tuning problems can be summarised as "the Universe looks to be extraordinarily flat, yet relativity tells us that a Universe filled with radiation and/or regular matter should evolve away from flatness, so it would have needed to be astronomically close to perfect flatness in the "initial conditions"". Inflation basically just says "maybe one of a million possible things made the very early Universe expand super fast, so that even if it was super curved before the patch of the Universe we can observe is very close to flat (because any manifold is locally flat by definition), and also look when you make the Universe expand like that you generally turn tiny quantum fluctuations into macroscopic ones and that happens to match CMB observations". The fact this works so well for inflation is due to the fine-tuned values actually corresponding to a very specific geometric situation (flatness) which can arise in multiple different plausible ways. You could not do this with any random value for those initial conditions.
2) I may need to consider this point more carefully when I'm less sleepy. I'm always afraid this sort of argument is too sensitive to measurement or basis problems, if that makes sense to you.
But also, can the point about fragility not be made with medieval physics - the Earth-Sun distance is fine-tuned and your only naturalistic solution to this is postulating a "multiverse" of countless Suns and Earths out there and then invoke the anthropic principle? Are we actually adding any evidence for theism by considering particle physics?
3) How would they notice starting to die? Presumably most Boltzmann brains are more aggregates of information than anatomically correct nervous systems. They need not be connected to any external "sensors" that may trigger survival "alarm bells"?
4) I'd like to say I'll do that. Instead, I'll say I will open a tab and hope I get round to it!
My atheistic-ish answer to this kind of question, which I playfully but sincerely expound at https://outlandishclaims.substack.com/p/universes-arent-real , is that the multiverse requires no explanation. God does, though, doesn't He? Isn't his only advantage as a prime mover the way that, by convention, we write about Him like we can assume Him into existence? Even if you accepted the fine-tuning argument 90% of the way, God still seems like too specific of an answer to be reachable a priori. Why not an entity wanting to maximize suffering, rather than happiness? Why not a committee of 12 mystical entities with different preferences, of which our universe is the uneasy compromise?
article is obviously correct and it will be fun to read the comments by triggered, deeply confused atheists on it.
Thanks :). Check dms!
You couldn't have posted this before I subscribed, I guess :(
I did say in my subscription message that I wanted to be challenged. I did not say that I wanted to get recycled Christian apologetics, a topic on which I have wasted far more than enough time on already (not with you, obviously).
I'm not aware (and I _did_ just perform a quick check) of any converts from cradle* atheism/agnosticism to any theistic religion (Islam, _theistic_ Judaism, Zoroastrianism :), Hinduism, Bahaism, cryptotheistic Buddhism, ...) _other than_ Christianity who
a) had some sort of fairly strong scientific background; and
b) after their theistic conversion/awakening/whatever you want to call it, take up apologetics and evangelism for their theistic worldview.
* I mention cradle agnostic/atheist in view of the example of people like C.S. Lewis, who was raised in lukewarm CoE Christianity, which experience I suspect implanted latent theistic epistemic viruses that later became virulent :(
Leah Libresco Sargeant is the best example convert from cradle atheism I can think of, more an embryonic moral philosopher, but she did at least understand coherentist epistemology in a Bayesian framework, etc. - I did buy and read her conversion experience book, which I considered quite honest, albeit poorly reasoned. I'll leave it at that.
G.E.M. Anscombe, apparently ?) raised in an agnostic household, while not any sort of scientist, converted to Catholicism in high school. Interestingly her capsule biography states that one of her brothers became a clergyman. I have not read anything about the backstory of her conversion, but given her fondness for teleology and Aristotelianism via Thomism, she probably wound up in the right place for her. It would have been interesting to hear a colloquy between her and a philosophy who strongly rejects Platonism and its works (I'm agnostic in a utilitiarian way re mathematical Platonism, and intensely, virulently :) anti-Platonist regarding ontology, epistemology, etc.).
There is Christian B. Anfinsen, shared 1972 Nobel in Chemistry, who publicly converted to Orthodox Judaism (coincident with his second marriage at age 73). Eight years into that marriage he wrote "my feelings about religion still very strongly reflect a fifty-year period of orthodox agnosticism." That fifty year period weakly suggests to me, perhaps incorrectly, that his childhood was spent in a Lutheran milieu (his parents were Norwegian immigrants who taught him the Norwegian language and culture); at any rate that he did not identify as agnostic until he was an adult.
My standard functional atheist answer to the strong anthropic principle (versus to the weak anthropic principle, for which see mostly 2) below) is:
1) It is at best a remarkably poor metacognitive strategy, at best actively dishonest (generally IME in the direction of self-deception) given known human biases and failings, to apply the label "God", with all its historical connotations and opportunities for motte and baileying, to some hypothetical low Kolmogorov complexity cause of the fine tuning.
On the very rare occasions when I have attempted to engage with actual theists (generally coming out of some sort of broadly neoThomist tradition, often Ed Feser fans of one sort or another, but that's very likely sampling bias at work on me; there's long explanation for that sampling bias that I will spare you) I suggest we use Ghod** - "generalized hypothetical organizing device" as a heuristic to remind us all to bracket the TBD degree of overlap between Ghod and the God who is typically hypothesized to be remarkably interested in the specific doings of some Iron Age pastoralists in the southwest Levant some three thousand years ago. (More extra fine tuning presumably to achieve that result :( )
** not my invented term, but I _am_ trying to signal boost it.
2) Given that we postulate some sort of Ghod ***, our evidence for the aforesaid Ghod being whatever level of fine tuning we can charitably consider to be "not special pleading", e.g. something about the value of the asymptotic fine structure constant, roughly 1/137, sure - we could call that fine tuning since we have relatively little reason to think that it has _not_ been the same everywhere and everywhen since the universe's temperature dropped below around 1 GeV [1] ; re Christian apologists, well, I've read a little of [multiple expletives deleted] Alexander Pruss' work and your merely citing him pushes my inclination to discount you to roughly as much as I would discount a convert to Trumpism;
I had never hear of Collins that I recall, a quick check of Collins' CV posted at, sure enough, Messiah University, tells me I'm not going to bother with him either. (interestingly, the link on his homepage to his "statement of faith" is broken, but his bibliography makes it clear that his advanced reasoning powers pretty much only operate when they lead him toward orthodox Christianity and its axioms).
*** I'm trying to phrase that hypothesis such that the posterior probability of that weakly defined Ghod is strictly greater the posterior probability of the God in which Pruss and Collins have faith.
Anyway, given a weak anthropic principle fine tuning Ghod along the lines of Einstein's
Einstein - G(h)od does not play dice with the universe.
Bohr - stop telling G(h)od what to do.
and allowing for a very weak hypothesis that Ghod (cf theism as opposed to a real watchmaker deity unconcerned with the doings of humans or other equivalently cognitively capable species elsewhere) has opinions regarding what we humans ought to be doing or not doing, how do you propose an implementable search strategy for a very finite human (e.g. yours truly with probably around 20-25 years life expectancy, though with the last ten of those probably experiencing significant cognitive decline)?
I propose to limit that search strategy to operate on evidence that was at last hypothetically available to extremely well equipped time travelers, with all of today's scientific knowledge and instruments, but none of today's theological inclinations, looking around say four million years ago (before the split of genus Homo from Hominina, so including the proto-australopithicenes, but well after the split of the ancestors of chimps and bonobos from our line).
At least _most_ of the fancy fine tuning that theists point to as leading to the pinnacle of all creation, namely humans born recently enough to know Jesus Christ, had occurred by five million years ago (as evidenced by anatomical signs of increased bipedalism; I'm being deliberately vague re causation of the fast increase in encephalization in genus Homo. but I *think* there's a reasonable consensus but not dogma that bipedalism's uptake came first - I haven't studied the literature of the last fifteen or so years in any detail).
I don't know whether Pruss, Collins, or any of the other theists (where their theism is based on pure reasoning from fine tuning) thought those distant ancestors of ours were equipped with the necessary material substrate in the brain to have the Sensus Divinitatis. I have not come across any statements either way (those statements would have to have been made in the last fifty odd years, when we really started getting a moderately filled in fossil record).
(I wonder if Collins studied directly under Plantinga at Notre Dame? The timing makes it entirely possible, but Collins' CV just says that he got his PhD in philosophy in 1996; Plantinga was still giving at least seminars at ND as recently as 1995).
[1] re the fine structure constant, yeah, I know in a broad sense about Webb et al 1999 inferences from some quasars' spectral hyperfine structure - and also Lamoreaux and Torgerson re isotope ratios from Oklo - to suggest that alpha might in fact be slowly decreasing over time. I'm agnostic on that one.
And the asymptotic value only comes meaningfully into play at moderate energy scales, e.g. < 1GeV.
I'm sorry that you haven't liked the content since your subscription :(. As for your objections, well this isn't Christian apologetics--I'm not a Christian. It's just an argument for God. Arguments like this are what have convinced me!
Your comment is a bit long, so I won't address everything, but just to address a bit: what's wrong with inferring God? I think God is very simple and explains lots of stuff. I also don't think that simplicity is calculated by Kolmogorov complexity.
Pruss is crazy smart and a good philosopher, though I disagree with him a lot.
No worries - a lot of the previous content was excellent (e.g. Cancel Culture Is A Tax On Being Interesting, Losing Faith In Contrarianism, There Aren't Naive Views) hence subscription. I'll start with the assumption that the fact that this was the very first thing you posted after I subscribed is approximately an accident (not truly random any more than a coin toss is, but the causal networks associated with your developing oeuvre at your substack interacting with my process of deciding it was time for me to start contributing $ is dense and hidden enough that we can treat it as a random variable with an unknown distribution).
Yes, you have made quite clear that you are not a Christian at this point (cr Steelmanning Christianity, which I have not read yet; it's on my reading list). Did not think you were. Just that your two primary sources (Pruss and Robin Collins) are not merely believing Christians (in e.g. the Francis Collins sense, and I mean no slight on F. Collins' faith thereby) but #Christians# or whatever typographical emphasis you prefer.
In response to Jacob's response to me (thanks for upvoting it), challenging me to read more of Pruss and some of Collins, I plan to give it a shot in the case of Pruss, who has some publications in math domains where I have enough knowledge to not be totally lost in the first few paragraphs. In "There Aren't Naive Views", you described Pruss' IQ as approximately one billion (greater than Ralph's? It's unclear whether Pruss fits under the rubric of a person you know in the sense that you know Ralph; I infer that he does not and is at least arguably smarter than Ralph based on your knowledge of the two).
What article of Pruss' (ideally public access in one form or another) would you recommend to me as best demonstrating his goodness as a philosopher (by your lights, obviously)? It's certainly possible that there is something significantly new under the sun in the neoThomian world, which I last (lightly!) perused more than a decade ago. His work would be more accessible to me in demonstrating his quality as a philosopher if it did not start with the assumption of the rightness of his interpretation of Catholicism (I have no reason to doubt that he does very well in arguments with other Catholics over the finer points or implications of various Catholic teaching where he and his interlocutor(s) concur on the basics; I'd be too ignorant of some of the terms of art in that discussion to detect misuse via equivocation, etc., although I'd probably be reasonably objective :) ).
I just found Pruss' very recent (2023 update) "ACCURACY, PROBABILISM AND BAYESIAN UPDATE IN INFINITE DOMAINS" which is going to require some mastication and cross reference checking, but it looks promising. No evident feedforward from theism to that paper, although I could see how he might use that result in some of his scaffolding.
I find just reading his blog to be helpful!
OK, adding to the reading list now. Quick scan looks promising. Thanks!
The first like half of this reply is completely irrelevant. Why does it matter what happened to Leah Libresco Sargeant, among others, in a reply to BB’s carefully thought through argument for theism from fine-tuning? Matthew isn’t a Christian apologist, he’s an honest truth-seeker.
As far as I can tell, your first real criticism is that there is no reason to infer God from fine-tuning because it does not tell us about his perfection and other actions and characteristics, so we should only infer to “Ghod”. That’s very true, and would be a good criticism if BB hadn’t directly addressed this. He argues that a perfect God is a better explanation than some other arbitrarily limited explanation because of its higher prior probability due to its simplicity (it posits a single thing - perfection, and has high coherence, in Draper’s terminology). BB is not inferring directly to God from fine-tuning, he infers to Ghod and then argues that the particular form of Ghod that is the simplest and best explanation is a perfect God also because this explains other things, like psychophysical harmony and anthropic evidence.
Also, calling citing Alexander Pruss Trumpism is utterly strange. I trust Alexander Pruss way more than I trust your vague gesturing towards some objection or other, maybe especially because you hadn’t heard of one of the leading philosophers working on fine-tuning, Robin Collins, despite your claim to have wasted more than enough time on the topic - waste more! Worse still is discounting Robin Collins because of where he works instead of engaging with his argument.
Thanks for the reply, Jacob! - I look forward to seeing what content you eventually place at your substack (check shows just the placeholder post right now, but I thank you for the other two recommendations you offered that I don't already read, namely Emerson Green and Majesty of Reason). I assume you *don't* object to my subscribing in advance? (no problem at all if you would prefer me to stay away, and no need to explain to me why you don't want me - it's your substack)
A couple of things back at you - I hope you will not mind my taking your points somewhat out of order - I prefer to find some of the areas of agreement first, plus I wish to reassure you that I read your response multiple times and took it both seriously and literally :). If you don't have much time to waste on me, perhaps just skip to [4] at the end, which I think contains the meat of what I suspect is an unbridgeable methodological chasm between you and me.
1) Your point that I should waste more time on the fine tuning argument has some validity according to my epistemology and decidedly incoherent sense of intellectual honesty. Collins' first publication on the fine tuning argument was published in 2003 as part of Manson "GOD AND DESIGN The teleological argument and modern science", and (as my cites to Feser and Pruss might indicate), I have not spent any time to speak on the more recent forms of what look _so far_ to me like the pretty much the same arguments.
Hooray - I found a library loan e-copy of the book! I make no guarantees about *when* I will get to it, but Collins' section is only about 22 pages - looking at the book's TOC and seeing sections in there from Swinburne, William Lane Craig, Dembski, and Behe, where the subheadings to their sections suggest nothing new from any of those four people, where the leaves me immensely skeptical a priori (I note that the volume does include four pages of a skeptical view in the sixty pages devoted to "fine tuning" derived from physical cosmology, the broader section that includes Collins).
I failed on my first try (admittedly quick and shallow) to locate some of Collins' other content as listed on his CV - I will make another try at it. So thank you again for the challenge.
2) Re g(h)od and general arguments from fine tuning, strong anthropic principle, etc.
I suspect there are some pretty interesting (probabilistic) arguments from generally agreed upon facts in cosmology, physics, etc. It seems plausible that there are some people that don't proceed fairly quickly from those facts to theism (or at least a very turbocharged deism, cf Einstein) have written about their tentatively drawn conclusions, although how much operational utility those conclusions might have is another story (conclusions to Christian, Islamic, Zoroastrian, etc. theism, or to any form of theism that somehow implies that the theistic deity has opinions/values/preferences/whatever about what we lowly humans do in our lives and perhaps pre-lives, afterlives, ... are another matter and do at least point toward operational utility). But I have missed those non-theistic thinkers. I freely admit I have not looked very hard.
If you, presumably having read more deeply and *recently*, know of anyone who stopped at deistic find
I don't agree _at all_ with what I can gather of the stated reasoning from g(h)od to God, but I have nothing new to offer to the well hashed counterarguments to Matthew's steps, counterarguments which he asserts that he has considered and rejected (the whole perfection thing has been thrashed over and over, generally in the context of some flavor of the ontological argument; Plantinga's modal version of the ontological argument being a fairly recent that I know some non-theistic philosophers have engaged with, notably "Chinese room" Searle). My response was long enough already :) and I don't see any likelihood of any progress for anyone there.
3) My apologies to you and to Alexander Pruss (as noted in my comment, some of whose work I _have_ read, but nothing too recent - I intend to correct some of that; I do plan to give Collins a bit of a try as noted above); I have no idea whether Pruss is or is not a Trumpist, nor do I particularly care (I suppose I would mildly prefer that he not be in the interest of, err, diversity within his intellectual circles). I meant rather (and *thought* I said) that my prior probability on the expected value of (re-)engaging with Pruss' content (at second hand) is rather low, as it generally is regarding engaging (on different topics, of course, with Trumpists).
You did inspire me to look (via his CV) at some of Pruss' more recent work in math and math adjacent, nontheological philosophy (last time I did that was maybe 2008 or so?, and it was argued as I recall from very intensely Christian theistic premises). A couple of papers there look interesting (e.g. "“Independent tests and the log-likelihood-ratio measure of confirmation” and “Regular probability comparisons imply the Banach-Tarski Paradox”; maybe also "“Non-classical probabilities invariant under symmetries” although looking at the rest of his work in that field, he seems to be talking about probabilities that are hyperreals or orders rather than what I find rather more interesting and potentially useful, probabilities that are locally approximatable as tensors) although many of their titles, e.g. “The badness of being certain of a falsehood is at least 1/(log 4 – 1) times greater than the value of being certain of a truth” cause me to want to run very far away from someone who speaks in terms of truths and certainties (for that matter, of falsehoods as well, unless his certainties and truths and falsehoods are shorthands for probabalistic judgements).
4) One area, highly relevant to Pruss [based on my reading of him, again possibly out of date] and relevant to Collins [based on inferences from titles in aggregate bibliography as stated in his CV and a weak inference from where he works; I did make an admittedly brief attempt to locate a couple of his later works as open source on the Internet; I will try harder - Collins has one quite early paper that looks like it might be quite interesting], where I think we disagree (at least, so I infer: I request, if you are willing to spend the time, to hear the respects in which you believe my supposition regarding your view of my approach is substantially wrong)
I don't discount Collins simply because he works at Messiah; I discount him based on probabilistic inferences from the combination of the sum total of his CV titles and journals and the faith statement of Messiah as a whole, to which I rebutably assume Collins assents.
I disagree, quite strongly, with the idea that it is fallacious** to reason thus:
[start bracket
I have relatively high confidence that Person A reasoned poorly in domain A1 from fact set A2 and according to reasoning methods A3.
Therefore, it is valid, probably mandatory, to reduce my prior distribution on the expected validity and utility of person A's reasoning in domain B1 from fact set B2 according to reasoning methods B3, where the magnitude of the adjustment depends on the overlap between the respective domains etc.
For a working example to jump let's say 100 milligodwins into the argument in one bound, if someone asserts that the 2020 election was stolen from Trump via [insert preferred methodologies], yes, that person is going to have to possess incredibly strong pre-2015 credentials in virology, epidemiology, etc. for me to give *that person's arguments* (either pro or con) re the lab leak hypothesis any attention. Life is too short.
n.b. I consider the lab leak hypothesis plausible but probably wrong. If you have not seen the rootclaim/slatestarcodex analysis of the lab leak that published recently**, I highly recommend it as an example of applied Bayesian analysis on contentious, politically/emotionally loaded claims that are in principle falsifiable to high probability given easily imaginable facts not currently in evidence
in case links are disapproved here or seen as spamming, google "slate star codex rootclaim leak" should hit at #1
end bracket]
Does that specific approach, which is indeed fallacious in classical logic** have a specific name? It seems like it should have one, but a couple of quick scans have not found one. It's arguably a subset of the genetic fallacy or an interesting permutation of the appeal to false authority.
** I find non-probabilistic logic to be of very limited use in cases like the example of argument from fine tuning.
this has already gone far far too long, and assuming you read to the end, I thank you for your patience and indulgence.
If you wish to reply, please provide some general parameters of how you would like my reply to your second reply constrained (e.g. a maximum word count :) ).
Thank you very much for the reply! You seem to have put a lot of work into the response and also seem to overestimate me a bit 😅 - I was just replying as I thought Matthew might respond. I haven’t actually read Pruss or Collins or any of these people firsthand either, I just know what they argue.
I will do my best to respond to your points, probably tomorrow, assuming there is some disagreement left over after I read what you’ve written.
No worries. FTR I have read some of Pruss directly (also Swinburne, Craig, Dembski, and Behe), although as I noted not all that recently. Also, unlike cosmology, physics, and any *serious* acquaintance with some of the mathematical fields that Pruss has published in**, I do have (also sadly dated now :) ) formal training and domain knowledge in biochemistry, which makes it far easier for me to reject Dembski and Behe with extreme prejudice based on some of their earlier writings :)
i.e. Pruss could *easily* BS *me* on some of those math papers and I would be none the wiser - higher math (anything past say the second year of real analysis) makes my head hurt, and I admit defeat from my multiple attempts to grok it. I can of course look at duelling papers in the fields that Pruss is publishing in, but the ugly truth is that I'm way out of my league there and I know it.
Don’t worry you can say whatever you want in your replies it’s just that if they’re longer it will take me longer to get back to you :)
Okay so the first point is thank you! I don’t know if I’ll ever end up writing any content of my own, but if I do I’d be glad to hear your thoughts.
Emerson Green and Majesty of Reason are great, but Substack isn’t their main platforms. I HIGHLY recommend their YouTube channels, especially Majesty of Reason. MoR honestly levelled up my standard of philosophical thought and knowledge immediately he is a literal genius and he makes such amazing videos with inhuman amounts of effort put in.
2) As far as I can tell, two philosophers that defend deistic hypotheses on teleological argument grounds are Paul Draper (who defends aesthetic deism, I think) and Antony Flew (who famously converted to deism after being a leading atheistic philosopher).
Is your objection that perfection isn’t actually simple? Or that it’s impossible or something? I don’t really know specifically what criticisms to Plantinga you are alluding to.
Also, I think the reason there are not that many deists as a result of fine-tuning might just be because the existence of a powerful and good designer is strong evidence that one of the religions that exists is probably true because that would be a very strong correct prediction.
4) I agree that you can infer that someone having a bunch of false beliefs is evidence that another one of their beliefs will be false, but I still think it’s unfair to make the argument you made from Pruss’ flawed reasoning in one area to flawed reasoning in another area. This is because you haven’t actually read Pruss’ reasons for being a Catholic, so you don’t know if his reasoning is flawed. Additionally, there might be some reason for the statement of faith - maybe that’s just something their workplace required or something. This kind of inference is useful if you can’t understand a person’s arguments or don’t have the time to research them, but I just think that skipping the higher-order evidence and engaging with the philosophers’ work directly is just a better approach and it lowers the risk of bias.
“which experience I suspect implanted latent theistic epistemic viruses that later became virulent”
I like how you put this because this is exactly how jesus described the kingdom of heaven in the gospels!
“The kingdom of heaven is like unto a grain of mustard seed, which a man took, and sowed in his field: which indeed is less than all seeds; but when it is grown, it is greater than the herbs, and becometh a tree, so that the birds of the heaven come and lodge in the branches thereof.”
“Another parable spake he unto them; The kingdom of heaven is like unto leaven, which a woman took, and hid in three measures of meal, till it was all leavened.”
Well, if we are going to be quoting gospels at each other, I'll give you Luke 15:3-7 :) **
** YMMV, but I think that's a rather better version of the parable of the lost sheep than Matthew 18:12-14
It should be unsurprising that my favorite Scriptural passage is Job 38:4-7 (NIV)
4 “Where were you when I laid the earth’s foundation?
Tell me, if you understand.
5 Who marked off its dimensions? Surely you know!
Who stretched a measuring line across it?
6 On what were its footings set,
or who laid its cornerstone—
7 while the morning stars sang together
and all the angels shouted for joy?"
My conditional atheist view, that informs my unoriginal "latent theistic epistemic viruses" phrasing, is that the human brain is, if you will, predisposed aka overly sensitive toward detecting agency, which leads fairly easily to broad sense supernaturalism such as you see in shamanistic religions (Tengrism, for example), and thence, particularly with growing societal complexity, to X-theism, where X is initially poly- or heno-, with the monotheistic variant showing up in the wild somewhere between 4000 and 3000 years ago (some of the earliest layers of the Zoroastrian Gathas; my amateur read is that the evolution of the Judaean religion from henotheism to post-Exilic monotheism is less than 3000 years old, but:
1) I'm not deeply read in the sources (for which I would need to know several different early Iron Age West Semitic languages, as well as at least a couple of flavors of Aramaic and probably Akkadian); and
2) I have no particular religious or nationalistic attachment to any particular chronology re the United Kingdom (of Israel and Judah, not of Great Britain and Ireland :) ) and its predecessors.
I am very much interested in the (secular) evolution of post-Exilic Judaism and Christianity, but again to have anything useful to say there without simply outsourcing my thinking would require far more commitment to the languages, sources, and scholarship than I am prepared to commit - serious hats off to the scholars who do it.
Oh yeah those are all good verses from the bible! I too find biblical scholarship very remarkable, I recently watched this presentation which was very enlightening. I do recommend it. https://youtu.be/ay_Db4RwZ_M?si=59F_ua6rFp5P9Pvb
After your extremes swings of opinion, a meta position would be “I am agnostic”. Our understanding of reality as a consequence of the fundamental laws of physics is so developed that have reduced chemistry and biology to the point where we synthesize life.
In around 400 years we have pushed theism and teleology from the universe. Multiverse is a little bit more for Marvel comics than for philosophy.
Still, in my view the theistic hypothesis is quite likely. But more like 40 per cent likely than 99 per cent likely…
I'm nearer to 60%.
Everything between 10 and 90 is more or less the same, so we agree.
You shall have a distribution on that parameter between 0 and 1, not a point estimate. The mean of that distribution can be 60 to you and 40 to me, but in any case, uncertainty on that estimate is large, as you have learned by its recent change, from less than 1 per cent to around 60 pc.
For example for “probability of heads or tails” in almost any coin I have had 50 per cent more or less all my life. Your estimates of the probability of something shall not be only numbers.
Have you read Bostrom on the Anthropic bias? I suppose it is the classical reference on this debate. I give small relevance to the argument on the universal constants. The existence of consciousness and the simple fact that reality exists is reason enough to consider a conscious cause to the universe (God) plausible. We really don’t now anything about the “meta laws” of nature (the space state where the parameters are chosen)
Yes, Bostrom seems to be in favor of the fine tuning argument! See my section where I talk about probabilistic reasoning surrounding the constants.
He is not very clear:
https://anthropic-principle.com/preprints/fin/Fine-Tuning%20Arguments%20in%20Cosmology.doc
I dislike the whole enterprise, because (the same as in conscience) we ask science to go beyond its borders: the reduction of reality to physics. While we have done this in a surprisingly effective and complete fashion, then we begin to ask about “why this laws of physics?”
You cannot ask science anything beyond what the Laplace demon can answer. Consciousness and multiverses are obviously beyond the “Laplace demon frontier”.
https://www.lesswrong.com/posts/nY7oAdy5odfGqE7mQ/freedom-under-naturalistic-dualism
We synthesise life? From non-life? Please I would like to see the evidence
There is not separation between “life” and “non life”, so we treat cell machinery as machinery and re arrange it. The most important application so far are closed cell membranes used for drug delivery at cell level, but other cell machinery are included in synthetic life products.
Your argument here, to clarify, is that there is no difference between things that are alive and things that are dead?
I understood that you were claiming that we can, without any living tissue, from dead chemicals, and nothing else, create fully functioning living self replicating organisms. As someone who studied medicine this is mistaken and not possible.
Of course there not difference. Life is simply a state of matter, as solid, liquid, gas or coloild.
In practice I seriously doubt anybody cares about not using “living” tissues in synthetic biology. But what Venter does is live beings micro engineering. The cell membranes are made out of lipids, but regarding organelles I am not sure to what extent they are taken from other cells.
Can you cite some science on there being no difference?
Between life and non life? Well, before you have a review on synthetic biology. The whole agenda of molecular biology is reductionism.
You wrote, "Our understanding of reality as a consequence of the fundamental laws of physics is so developed that have reduced chemistry and biology to the point where we synthesize life."
What do you mean by "synthesize life"?
I mean this:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8578974/
Craig Venter is the most famous person in the development of synthetic biology. The final proof of reduction of biology to physic and chemistry. No elan vital, only atoms.
Well, now there are self replicating artificial cells:
https://www.jcvi.org/media-center/first-self-replicating-synthetic-bacterial-cell-constructed-j%C2%A0craig-venter-institute#:~:text=The%20synthetic%20cell%20is%20called%20Mycoplasma%20mycoides%20JCVI%2Dsyn1.
Additionally, this is the opinion of Venter on natural selection:
“ Our findings demonstrate that natural selection can rapidly increase the fitness of one of the simplest autonomously growing organisms.”
https://www.jcvi.org/publications/evolution-minimal-cell
I love Venter's work. He synthesized life in the form of taking an existing life form, and modifying it quite a bit. I just wanted to make sure that you did not imagine that any scientist today is capable of starting with material not from living cells and creating a new living being. Its amazing that quite a few people think that.
Venter claims, based on both theory and his experiments, to have produced not only a simpler (less DNA base pairs) form of life, but something approximating the simplest possible form of life (at least of anything resembling any type of life we have ever found). In fact, not long after Venter's findings were published, I wrote this blog post explaining how they provide evidence for an Intelligent Creator of life:
https://parresiazomai.blogspot.com/2016/03/excited-both-as-science-geek-and-as-god.html
For the time being let’s put the past aside. What I claim is that life is mostly reduced to physical processes. We cut paste, change and manipulate it in a similar fashion that we manipulate machines. There is not trace of any supernatural force in the machinery of life.
I'm not denying that life (certainly simple life like bacteria - creatures with consciousness like we have are another matter) can be reduced to physical processes. Life is, as you alluded to, like machines. In fact, even the simplest life (like Venter's simplest cell) contains many molecular machines. Like our human-made machines, these machines in cells consist of many parts which must be of specific shapes and sizes and made of material with specific properties and put together in specific ways in order to perform the functions they perform (you can google and find animated videos of cellular machinery - it is fascinating!). In other words, they have a high level of complexity that is functional.
But machines require designers. They require designers because just as the right rocks and minerals would never on their own happen to assemble together in the right forms to produce a working car engine or an iphone, neither is it reasonable to think they could come together to form a working cell. And cells can do something that no human machine can yet do. They can use material in their environment to reproduce themselves! All this points to an Intelligent Designer being involved.
Is there anything in existence you believe not to be designed by a designer? If not, then you're not successfully articulating a boundary we can use to determine what is designed and what is not designed. If so, why would life be something special that needs to be designed? We've produced a myriad of various works in the physical sciences that detail how existing biological stuff can successfully self replicate, and how the complex systems that biological stuff forms can interact with each other in mutually beneficial, mutually sustaining ways.
Also
>just as the right rocks and minerals would never on their own happen to assemble together in the right forms to produce a working car engine or an iphone
But they do assemble to form caves, rock formations, strata detailing hundreds of millions of years of geological history, planets, asteroids, mountains, and so on. These are a few samples of emergent features that previously people thought would need Gods to e.g. sculpt the mountains, that we now recognize can occur without any conscious agent intervening in the process, due to entropy, coarsegraining, statistical mechanics more generally. I don't think there's any epistemic space left for "And then God intervenes!" when we create models to explain these various features.
Perhaps we need God in the past (while everything indicates the opposite) but in the present, we need not elan vital.
I think many things, from the fine-tuning of the universe for life, to the origin of life, indicates that an Intelligent Designer (whom I believe to be God) was needed in the past. Recognizing this will help us to have a better foundation for thinking about the present and the future.
How do you know nothing that is currently undetectable is not involved?
I’ve never been fond of the multiverse hypothesis to “disprove” agentic fine-tuning because it’s fundamentally a faith position. It’s invoking an indefinite (infinite?) number of unobserved (unobservable?) universes to account for observed facts. Where are the receipts? Where are the observations that give us reason to think that this is how the world is, rather than how atheists would *like* things to be?
The fact that there's a fine-tuned universe is the data that supports the multiverse existing.
Why would a Creator God bother with all the others if our finely-tuned universe is the one desired? Was there a lot of trial and error? Was he incompetent or just dicking around?
I think a simple answer is multiple universes certainly require a lot fewer assumptions and assertions about the world than creating some extra natural entitity(s). Much less a specific one.
One the one hand it seems odd we are here. On the other hand...where else could we possibly be?
Yes this is unsatisfying. So lets grant that this particular bundle of rules is in some sense "unlikely".
(Though I am not sure there is really any evidence for such a claim because we don't really have any evidence that any of the values we believe are "fine tuned" could be different. That they could be different is just an assertion with no evidence.)
Ok so here we are in a setup with some unlikely rules. What are some explanations for that?
Well one obvious possible explanation is there are multiple setups. So that we are in an "unlikely one" is less surprising.
It does not seem to me at all obvious that a Christian or other god(s) is preferable to this explanation, involves less assumptions, or frankly really solves anything at all since you immediately have questions about them.
Even assuming we are all Sims in some alien's/god's little toy, that god presumably has all the same questions about metaphysics in one way or another. That alien/god found probably find their existence fine tuned as well.
As a final aside, the fine tuning argument sems less and less strong the longer Fermi's paradox holds up. The idea that the universe is in some way "fine tuned" for such a small peripheral phenomena as life, seems pretty silly when it is 100% FOR SURE, not the main thing going on. Most of the mass, and the vast majority of the energy is doing stuff almost entirely unrelated to life.
In fact the universe is in such a state we seem less like some being's carefully crafted and curated sims, than a giant procedural random simulation that was just fired up and is producing a tiny weird recursive bug that is about to crash the program.
If the fine-tuning argument is so good and true, why do physicists - the ones most capable of understanding it in detail and evaluating if it is true - the least likely group of scientists to believe in God? https://www.pewresearch.org/religion/2009/11/05/scientists-and-belief/
There's widespread agreement among physicists that there is cosmological fine-tuning. But what one should infer from fine-tuning is a philosophical question, which physicists aren't especially geared to evaluate. If you look at most of what they say about why they don't believe in God, it's nonsense about, for instance, the god of the gaps.
> But what one should infer from fine-tuning is a philosophical question
No, it is a factual question. I infer from the fact that an object is hotter than the things around it, that it was heated. I infer from the fact that you wrote a bunch of words that you are a human. Whether God exists or not is not a philosophical question, it is a factual one. If God descended from the Heavens and said "Yo, I'm God, here's some proof of me being God [performs physically impossible miracles]," then we would need no philosophy to conclude that he was real. The fine-tuning argument is exactly this: if you look at the laws of physics, the logical inferrence from them is that God exists.
More than anyone else, a physicist would be inclined to be capable of evaluating if the inferrence "God exists" follows from "there is cosmological fine-tuning." And it appears they do not evaluate it that way.
"Look!", sayd the puddle, "How perfectly the ground fits me. It must have been made just for me! What glory!!!"
The universe is not fine-tuned for us, we are fine-tuned to it.
If we scale the issue back to our more immediate predicament, like the ratio of oxygen in the atmosphere, the mass of the earth, accessible water, orbit around the sun etc, we might ask the same question and assume fine tuning must be involved. But then we look a little beyond this scope, to other world - all these planets, with their own solar systems, and all is barren.
The existence of many other planets and solar systems is highly analogous to the theory multiple universes, only that the former is certainly real. Among all of theses worlds, ours is the only one that struck gold with all the cosmic and planetary parameters necessary for life. Given so many worlds, it serves as a very storage evidence for us to be the result of a statistically-improbable-though-possible coincidence, instead of the creation by an intelligent entity.
I agree that a multiverse is quite a complex thing to exist, but then so is a god.
I would like you to separate your arguments more and put probabilities on them. To me you say why the multiverse is unlikely but not why it is impossible and to me a god looks pretty unlikely too. So the real comparison is between unlikely options.
I don't think God is complex https://benthams.substack.com/p/10-ways-god-can-be-simple?utm_source=publication-search
>Fourth, even if a multiverse explains why we exist, it doesn’t explain fine-tuning for scientific discovery.
This argument fails because believing you've discovered something is a psychological property of conscious agents. I will demonstrate this by example: I think you have a mistaken view on scientific realism - you are most likely a scientific realist and think science discovers the objective structure of the world. I think that's totally wrong, and that there is always a disjunction of available methods, tools, and models, at our disposal when conducting scientific investigation, such that there is never a last word, a final investigation, or a point where we reach "the world" and say "That's it, we've reached our target!". We construct the scientific program, we don't discover it. This means I don't affirm that we ever make scientific discoveries in the realist sense.
Yet I have no trouble accounting for your belief that e.g. we discovered electrons. You are in an epistemic position such that you believe there are electrons capable of being discovered, and that we have discovered them. In other words, there's no need to posit, "And lo and behold, there are electrons, for otherwise Matthew would not come to believe in them." Your believing there are electrons is not necessarily grounded in there really being electrons - just in your memories of what scientific realists have said and in your ontic beliefs.
In the same way, discoverability isn't a property of the world, it's a property of what epistemic agents believe. And epistemic agents can be so constituted that they believe they discover something brand spanking new about the world when something as pitiful as e.g. an electron whizzes by, or time passes, or they recall or hallucinate a new memory. In other words, God doesn't have to set up the world so that it's "really" discoverable - he just has to set up the world so that epistemic agents eventually develop beliefs that they discovered something. (I'm still unclear on why God would even care about this though.)
>the objective chance of a coin coming up heads is .5 because as the number of coins you flip approaches infinity, the share of them that will come up heads will approach half.
This isn't true, there's a non-zero chance that the coin will quantum tunnel through the Earth and never land anywhere, flying through the rest of the universe unimpeded. So you might think, "Oh, well I'll just build that into the objective probability calculation!" Except you also need to build in that the coinflip happens in a sufficiently powerful gravitational field - flipping a coin off an asteroid could again result in it flying across the universe unimpeded, never landing heads nor tails. And you also have to build in that the coin is indestructible - since each time you flip it, you cause parts of its mass to flick off - but now you have to find some physically possible way to create an indestructible coin. You also have to account for conscious agents flipping the coin in order to get a desired result, which would bias the final probability, controlling initial conditions so the wind doesn't blow it away forever, creating a perfectly flat coin so it can't land on its side, and so on. Whatever you end up considering the objective probability of flipping a coin to be, it would seem to be something inscrutable. Consider this a metacritique of your metaphilosophy - you deploy concepts like "objective probability" in idealized contexts and assume that all the conceptual difficulties that come with applying them to nonidealized contexts will just spontaneously resolve themselves.
I don't see why this is a problem. We just imagine a coin where nothing fishy happens!
There isn't a coinflip where nothing fishy happens. There's all sorts of physical constraints that inform our theoretical modeling of coinflips, and all sorts of physical concerns that will make an accurate coinflipping model sensitive to contextual cues. If you want to make a purely theoretical model, then it will fail when you try to apply it, since it doesn't account for the nuance that the accurate coinflipping model builds in by being attentive to relevant conditions.
For example - I stipulate a rule that I am right about everything. What does that mean? Well, you say "That's right" whenever someone says something true, right? And you use it to affirm truths like in the previous sentence, and so on. So I just start babbling, 1+1=3, there are and there are not colorless green ideas that sleep furiously... Do you still understand what it means to say that I am right about everything? What do you mean no? Just interpret it as in the nonfishy cases where someone is right about something! Clearly, the propositions I babbled and what I told you to do while interpreting me will not just work themselves out into peaceful equilibrium.
I like Jonathan Weisberg's objection. We already knew a priori the laws were life-permitting from the fact of our own existence; thus, what we discovered is not that life has a small probability of existing on the laws given naturalism, but that the laws are stringent rather than lax with respect to small variations permitting life or not. But it's not more probable that God would prefer stringent over lax laws. So fine-tuning isn't evidence for God.
Weisberg goes on to argue that this reasoning does not depend on the chronology of our evidence, i.e., even if we'd hypothetically discovered the stringency of the laws before our own existences (a possibility which I think makes no sense, but ignore that), the objection would still go through. Because fine-tuning is only evidence for design insofar as it's evidence that life actually exists. Once you learn the latter, the former evidence is screened off.
Also, this does not depend on assuming the prior on our own existences is ~1, even though I think that should also be regarded as true, and the standard firing squad arguments against this fail.
PS. Although I don't think this is necessarily fatal for you since you can use other constants, I don't think the fine-structure constant is fine-tuned. To suggest as much relies on using a uniform probability distribution - perhaps even over an infinite interval, which is problematic in itself. There's more motivation, by contrast, to using a non-uniform maximum entropy distribution, and when you do this, IIRC the probability of the life-permitting range for the fine-structure constant is relatively high.
I have a section where I address that argument!
I think there's a not terrible probability God would make stringent laws--so that we could see his existence--while conditional on atheism the odds are super low. Therefore, stringency favors theism over and above mere existence.
OK, sorry I missed it.
It's not at all obvious that the probability of stringent laws on atheism is super low, relative to lax ones. Remember, we're not talking about "stringently life-permitting laws vs. any set of laws whatsoever, life-permitting or not," we're talking about "stringently life-permitting laws vs. lax-ly life-permitting laws." Most of the lax but still life-permitting laws are going to have obscenely high complexity (if you don't believe me, try thinking of some!), or else they'll maybe be low-complexity on their own but rely on obscenely highly complex initial conditions (this is really a semantic issue of whether you consider the initial conditions to count as part of the laws or not). So they'll have very low priors.
That might be right. But the probability of atheism+lax laws is super low, so then if the probability of atheism+stringent laws is similar, then the probability of life-permitting laws on atheism is super low!
Maybe P(atheism+lax) is super low. But if it's comparable to P(atheism+stringent), and P(God+lax) is comparable to P(God+stringent), then the laxity/stringency of the laws aren't strong evidence for either theism or atheism. So this isn't really about fine-tuning at all, it's just whether the existence of life-permitting laws of any sort are evidence for God.
Even if that's true, I think the presence of life-permitting laws are super strong evidence for God. But this is extra evidence because it can both serve to clarify how unlikely life-permitting laws are--surrounded by the sea of non-life-permitting laws--and because if theism has any decent chance of getting stringent laws--which it does, because God would want to show fine-tuning--that favors theism massively. You admit the odds of laxity+atheism are super low, so if the odds of laxity+theism are not super low then laxity+theism is massively confirmed.
>Even if that's true, I think the presence of life-permitting laws are super strong evidence for God.
It's possible, but now you can no longer rely on the crazy-high likelihood ratios that fine-tuning arguments are supposed to be able to deliver, as in the ~10^10^123 figure you've named in other comments. So you can't necessarily outweigh a low prior on God.
Personally, as I've argued elsewhere, I believe theism is a vastly more complicated hypothesis than anything remotely in the vicinity of contemporary physics or whatever. So something like "contemporary physics + atheism" will have a much higher prior than God, and it will also predict that the laws are life-permitting!