Bentham’s Design Argument is Still Not Convincing

Physics is hard and full of terrors

Please note: the email version of this post will be clipped as it’s too long for email. Click on the title at the top to go to substack for the full post.

The most prolific theistic substacker is, without a doubt, Bentham’s Bulldog aka Matthew Adelstein.¹ Recently, he responded to a review of Ross Douthat’s lazy book on religious belief, saying some things that I thought were wrong.

I left a comment on that post with criticisms of my own, but upon reading the links of his post more thoroughly, I realized he addressed (or rather, thinks he addressed) those criticisms in another post. So, I decided to respond to his main post on design, giving it the proper treatment it deserved.

Unfortunately for my theist readers, my opinions on the design argument have not changed.

I’ll start by saying positive things about Adelstein: He is very good at provoking conversation and challenging people to conduct more research before asserting an argument. He is well read, extremely curious, extremely smart,² and knows how to write in a way that provokes a response.

Having said that, upon doing (rudimentary) research on design and physics in preparation of this post, I found that his original post omitted key information and made fallacious assumptions to the point where a line-by-line rebuttal just wasn’t worth the time and effort.

I am going to be relying on a pretty stellar YouTube documentary by atheist Phil Halper on the design argument, which I’ve posted below. Please note: I’m not relying fully on this documentary. Rather, I’ve found many of the problems I have with the design argument better expressed by people much smarter than me in the documentary, many of them mathematicians and physicists.

I don’t expect Adelstein to see, read, or respond to this post. His following is 200x mine. I get annoyed when random strangers comment on my post, misunderstand my arguments, and badger me to respond, as if I owe them.³ I similarly don’t want to inject negative “DISCOURSE WITH ME BRO” energy into Adelstein’s life.

Anyway, this post is for me and for folks who aren’t as smart as Adelstein. I needed to get my own thoughts on fine tuning written down somewhere. This will be the first of two parts; this post will be on my problems with Adelstein’s design argument, the next on my own general objections to the design argument.

I hope it will provide others the tools to spot when Adelstein oversells his case for design, or other stuff, like odd theodicy fanfic. On the off chance Adelstein sees this post, I hope he sees it not as a challenge to change his opinion, but to improve his craft and make his arguments better and with improved clarity.⁴

Unfortunately, I thought his original post on design failed on both measures.

A Quick Reason To Be Skeptical

I agree with Adelstein and his main source of citation, Dr. Luke Barnes, that the universe is fine-tuned, as defined that “the parameters of the universe…fall in a very narrow range needed for life.” I agree that the universe’s laws, constants, and initial conditions are fine-tuned because there are seemingly narrow conditions in which life could arise and we are living within those parameters. So far, so good.

The problem with Adelstein’s post is that he gives you the impression that the debate on possible universes is (or was) between Barnes and the late Dr. Victor Stenger.⁵ Stenger died in 2014 and Barnes had the last word, and as no one has rebutted Barnes, you’d have the impression that Barnes won.

The problem here is that Barnes’s co-author for his book A Fortunate Universe, Geraint F. Lewis, is a non-believer who disagreed with some of Barnes’ conclusions. Second, many scientists disagree with Barnes (see the documentary). So immediately, I’m skeptical of how Adelstein sells the science.

Scientism? How About Philosophyism? Or Theologyism?

What’s more, it’s important to point out that Adelstein’s argument is predicated on the Stanford Encyclopedia of Philosophy. That’s not a bad source! But many of these arguments are empirical, and not exclusively in the purview of philosophy.

I have noticed in my many years of observing religious debates (about the existence of God or otherwise), that many (theistic) philosophers are just ignorant about what science has discovered, or at least what the latest science can explain or demonstrate. And so, these philosophers sometimes come up with logical theories that are contradicted by what we do know about science. This isn’t just a modern phenomenon. Aristotle got gravity wrong, after all.

For a modern example of this: Plantinga said that under naturalism, we wouldn’t expect our brains to evolve to know the truth.⁶ This is silly and kind of like saying our eyes couldn’t evolve to see. Sure, the human eye has a limited range of sight and it is capable of malfunctioning, much as our brains have a limited range of cognition and is capable of malfunctioning. But it’s a fundamental assumption of philosophy that our minds at least somewhat reliably track with the facts and truth of reality.⁷

Another example is the constant insistence by theistic philosophers that dress up the idea (in some capacity) that we need an external agent (God) to “give” us morality, when there is abundant literature of how moral instincts in humans are evolved, observable in babies, and yes, in non-believers and non-religious societies.

Yet another example of this (that I’m admittedly less knowledgeable about) is the Aristotelian metaphysics of causality advanced by Catholic philosophers. Namely, it’s just not something physicists incorporate in their models of the universe. If you’re more of a Protestant, the Kalam argument, as advocated by William Lane Craig, may be incompatible with General Relativity.⁸

All to say, philosophers (especially theistic philosophers) like to wag their finger at reddist atheists and physicists who practice “scientism” - the idea that philosophy is useless and that the scientific method is all you need to determine truth (which is itself a philosophy).

I typically agree with these criticisms, but I also think these philosophers and theologians similarly encroach upon fields that they have no expertise in (physics, biology, psychology) and it’s similarly a bad practice. Scientists should not make false pronouncements about philosophy if they aren’t sufficiently read, and neither should theologians and philosophers make false pronouncements about science if they aren’t sufficiently read.

But because we don’t have a catchy word to call out these philosophers and theologians (Philosophyism! Theologyism!), they get to opine freely and advance false theories about science, and theists are credulous.

There’s been discourse on substack in the last few weeks about falsifiability and evidence. I am sympathetic to criticisms of internet atheism if only because saying so scores me points with my interlocutors on substack (please subscribe!).

But what these criticisms miss is that falsifiability and saying “actually this is incompatible with our best scientific theories, and so I reject this idea” are important epistemological antibodies to prevent us from being distracted by theories that don’t match reality.⁹

The Physics Is An Empirical Question

Anyway, I will punt on the specific argument of whether life could arise under different physical laws because I think it’s an empirical question and I’m not qualified, as I know fuck all about physics. I dropped physics 101 as a college freshman because I was awful at calculus.¹⁰

If you’re curious about these issues and have a better understanding of derivatives than I do, I implore you to watch the documentary where the smart, accomplished theoretical physicists like Alan Guth, Sean Carroll, Roger Penrose, and others explain different theories of how the universe came to be, how some of them invoke a multiverse, how the multiverse is actually a hypothesis that the math of theoretical physics points to (and is not an anti-theist cope), how some of these theories explain life arising, and how some even assert there’s no fine-tuning.

Do I understand any of this? Absolutely not! But the explanation is there, and the conversation is interesting, even if like me, you can’t quite understand it.

Counterfactuals vs Probabilities

One of the best sections of Adelestsin’s piece is titled “What if you change a bunch of constants at once?” The short way of summarizing this section is that Adelstein finds the possibility of other universes creating life irrelevant because it still doesn’t grapple with the fact that the universe we live in right now is improbable.

The non-physicist, non-mathematician atheist response (mine) is that if we’re a part of a multiverse and there’s an infinite number of universes with different laws of physics, an infinite number of them will look like ours, so the probability that this could happen isn’t infinitesimally small. Instead it’s p=1, because infinity makes things weird, and if you plug it into a bayesian equation, it will spit out p=1.

It could be that we just got lucky and live in one of those universes that support life, even though there are also an infinite number of universes that don’t support life. On the flip side, if we just have one universe, then we have no reason to think the laws of physics could have been different. Again, p=1.¹¹

The physicist answer - the better, more correct answer - is more blunt: We don’t know what the probability of what different universes look like.

We can imagine scenarios where universal constants were different and thus create different universes, but we don’t have a theory of how those constants could change, and thus we have no idea if those universes could happen.

Put another way: Just because we can imagine a vast array of counterfactual universes, does not mean those universes are more probable. Adelstein is confusing counterfactuals with probabilities.

But hey! Don’t take my word for it, listen to the smart people in the documentary:

“There are different cosmological theories about the origins of the universe, but none of them really posits a mechanism that can generate different universes with different possible values of those constants such that we can understand the probabilities of getting those constants falling within different ranges. So we don’t have the analog of the roulette wheel”

(8:11 - 8:34)

“The counterfactuals don’t give you the probabilities. If I’m playing craps and I lose, I can say if I’d rolled a 7, I would have won. That doesn’t tell me anything about how likely or unlikely it is of rolling a 7. So we can do the thought experiment and imagine a universe like ours with the same laws of physics and principles of physics but with different values for those physical constants. That might tell us what would have happened in such a universe. It doesn’t tell us whether those values of the constants were probable or improbable.”

(22:17-22:53)

Bayesian Fundamentalism Leads To False Conclusions

Confusing counterfactuals and probabilities is an understandable thing because probability theory is hard to intuit for most people. The problem here is that Adelstein is, um, how do I say this nicely: a Bayesian Fundamentalist. Or maybe a fanatic. I’m not trying to do ad hominem or be mean, it’s just that Adelstein seems to reject any reasoning or logic that isn’t Bayesian in nature. He doesn’t seem to care about conversations about what is possible or impossible, but what is improbable or probable.¹²

The problem with this is that what is possible or impossible are important conversations that frame our conceptualization of Bayesian calculus. As Nate Silver says in the Signal And The Noise: Bayes theorem can only tell you if God’s existence is probable, given God’s possibility. If you think God’s existence (or at least proving God’s existence) is impossible, Bayes theorem is useless.

All of that is to say that logic and critical thinking are important constraints on Bayesian probability. If you don’t have those constraints, you can come to absurd conclusions, and can be talked into believing anything. So, we need to have dull conversations about the nature of logic, whether the chair I’m sitting in exists, how we know these things, and the circumstances where we can disregard illogical conclusions, before we bring out Bayes theorem.

Throughout his piece, Adelstein tries to use metaphors and thought experiments as evidence of probabilistic conclusions. The problem is that many of his thought experiments are just bad¹³ and metaphors are by definition not 1:1 comparisons. We use metaphors to illustrate complex, counterintuitive reality, but metaphors differ from the things they compare to in key ways that make them disanalogous.

Adelstein’s inability to grasp this leads him to formulate false conclusions or oversell his conclusions as more certain than they are. This is the most puzzling thing he does because many of these conclusions on first glance don’t pass the smell test.

Take for example the section on the anthropic principle. He cites an instance in which 500 people shooting and missing one person being evidence of conspiracy. Is it now? What kind of guns are we using? How trained are the people using these guns? Did they fire at the same time? How far away is the target? Do some of them have moral objections to killing someone and intentionally miss? While others love killing but are terrible shots? There are so many explanations that can step in here, among them is the fact that improbable things happen all the time!

Or his examples in the same section about the circumstances of one’s conception. That someone being alive today is evidence that their parents had sex or didn’t use proper birth control techniques.¹⁴ It’s definitely a safe assumption that the overwhelming majority of people’s parents had sex with each other and didn’t use any birth control techniques. But that is different from saying that a person merely existing is evidence in favor of that event occurring. Some people are conceived through artificial insemination. Many people are conceived in spite of birth control.¹⁵ How can one’s existence be evidence for something that didn’t occur?

Post-edit: Sure, if we're using a PhD philosopher’s definition of evidence where it's any data point supporting one hypothesis over another, anything can be evidence for anything. But that doesn't mean it's *good* evidence. Me wearing shoes can be a data point in favor the hypothesis of me of doing any number of actions, and excluding others. But me doing or not doing certain actions may have no bearing if I was wearing shoes. It doesn't move the needle. In this way, some datapoints don't move the needle. The whole conversation is predicated on evaluating the data points, not assuming them.

In this way, Adelstein conflates “improbable” with “impossible” and “highly likely” with “by necessity.” In practical circumstances, that’s usually harmless, but when it comes to figuring out the truth of reality, this confusion can have you cling to false evidence and come to false conclusions.

Put bluntly: it is false that all people were conceived without birth control. The existence of an individual is not evidence that their parents didn’t use birth control. Sure, their existence is a data point you'd expect if they didn't, but that doesn't mean that they didn’t. If we want to understand all actual circumstances in which someone may be conceived, a model has to accommodate even improbable outcomes.

If you or your model of reality are saying that it is untrue that people can be conceived upon using birth control, you and your model of reality are wrong.

Now, as the saying goes, all models are wrong, but some models are useful; we can’t hope for a perfect model of the universe. But Adelstein’s model is not merely wrong in the normal ways that models are wrong. It’s not formulating a hypothesis based on internally consistent axioms to explain and make predictions about the universe that can be tested, like the many different models physicists have put forward in the documentary. Rather, it’s aiming to explain a specific occurrence, but without any specific testable predictions.

A Weird Principle of Indifference

Adelstein says:

“The basic principle is simple: if there are a wide variety of different ways something can turn out, and none of them are special, if some theory naturally predicts it turning out the way it actually does, that theory is majorly supported.”

Put another way, if there are 100 different possible outcomes, and your theory “naturally predicts” the outcome, it is “majorly supported.” Again, if you assume Adelstein’s Bayesianism and understanding of physics, this argument makes sense. But if you poke at it, there are problems.

I tried to come up with a thought experiment to even formulate the implications of this principle, but I have a hard time doing so. It seems to only work for this specific fine tuning argument or tangential design arguments, like the one he uses about deducing the actions of long dead aliens.

To name just a few problems: What do we mean when we say a theory predicts an “outcome?” What is an outcome? A historical event? A score in a football game? The physics of the universe? The balance of power in global politics? This principle is vague, because outside of specific mathematical outcomes, it’s hard to see how this principle could apply to anything. And as we have seen, it doesn’t really apply to the design argument either.

Does the scientific theory that “God exists” or “the universe was designed” offer any sort of insight other than “we are here?” Does it say anything about what the universe should look like with quarks and quasars and what not? How does one logically, causally, or scientifically demonstrate this beyond saying “God did it?” What predictions does it make and test?

It seems that he’s stumbling on something resembling the (Christian apologist) principle of indifference, which is summarized in the documentary after explaining our ignorance of the true probabilities of the universe.

“…fine-tuning advocates often appeal to something called the principle of indifference…that says that if we’re indifferent from two hypotheses then we should assign the same probability to both of them. But there’s many cases in which that’s going to give us inconsistent results for the probability assignments that we make.”

(8:34-9:17)

So for example, you could wake up one day and not know what day it is. It could be Monday, or it could not be Monday. Logically, it’s 50% either way. But it could also be Tuesday or Wednesday or Thursday, etc. So the probability isn’t 50%. It’s more like ~14%. The principle of indifference in this way could inflate or deflate a probability.¹⁶

Heck, you could call this a partisan election model of reasoning. In 2008, the principle of indifference would say that there was a 50-50 chance that Barack Obama or John McCain would be elected president. The chances were not 50-50, as polling and economic indicators strongly favored Barack Obama. There were Democrats who thought Obama would win, and Republicans who thought McCain would win. Was either partisan model of the universe correct or “majorly supported?” Were the odds truly 50%?

Obviously not.¹⁷ But we wouldn’t apply the principle of indifference to evaluate this election. At the same time, we wouldn’t say that individual partisan Democrats had any special theory about politics.¹⁸ But Adelstein would!

Any Theory Is Better Than Agnosticism

I think this is absolutely silly. It elevates lucky biased people and demotes reasonable agnostics. We’re not actually evaluating the reasoning, evidence, and processes by which models of reality discern it and create predictions about it, but rolling a pair of (figurative) dice over and over again and assigning credibility to the lucky.

I’m all for pragmatism, and sometimes pragmatism entails luck, but this is just blindly following the lucky. Most objectionable is that, under this principle, Adelstein is saying that any belief is more probable than agnosticism, “randomness,” or just saying “we don’t know.”

Think of the most obnoxious New Atheist stereotype of superstition (the tooth fairy, the flying spaghetti monster, leprechauns, you get it), and it’s more “majorly supported” under this principle so long as you can come up with any explanation that “naturally predicts [the world] turning out the way it does.” If you can come up with some word game with how fairies want a universe that supports life, congratulations, that’s more plausible than atheism or agnosticism, per Adelstein.

Using scientifically rigorous processes, we could probably eliminate these obnoxiously mythical creatures, but again, that’s not what Adelstein is saying here. It doesn’t matter that we don’t have good evidence for fairies aside from this model! As long as you can come up with any theory that predicts a random outcome, it’s evidence in favor of the theory. He doesn’t place any sort of limitations on how that theory was formulated; he just says anything goes, relative to agnosticism or randomness.

In this way, I can’t get onboard with Adelstein’s method of accepting models of reality. Sometimes we just don’t know things and we should remain agnostic. Sometimes operating on “the best” evidence is still a bad idea because the evidence is too scarce or poor to create a good model.

For example, let’s say we live in the year 3000 BC and we’re trying to figure out why the sky is blue - pretty much all available theories are going to be wrong (and bad, and unfalsifiable) because we lack the scientific understanding to formulate a proper model. Agnosticism is the best model to adopt, until we have a better understanding of how the world works.

And so, when it comes to the actual mathematics and probability of the design argument, I will join the army of non-theists in saying it’s probably the best argument against non-theism. But, I still don’t think it a good argument for the existence of God, or the Christian God. That will be the subject of my next post.

PS: Here’s a great follow up discussion about the documentary, responding to criticisms of the original documentary.

1. In a previous post, I called him Mr. Bulldog, which was a little goofy, but I think he deserves to be known by his real name. If he hasn’t run into the CosmicSkeptic/Destiny problem (of people knowing him by his username) he will soon!

2. I suspect his IQ is higher than mine and I feel no shame in admitting that. There are many more-read philosophy blogs on substack. Their ability to suspend disbelief is both a superpower and handicap in that it allows them to explore interesting topics, but also come to conclusions that are in my opinion a little silly. I think their ability to do this is because they’re smarter than me.

3. A good illustration of this was, a few weeks ago, I wrote a post on how stupid and bigoted a lot of Christian Apologists are. Someone commented something along the lines of “If no one would know about it, what would prevents you from stealing from a billionaire.” I gave them one reason why I wouldn’t (it would hurt my conscience), while also pointing out the original question was loaded: They took consequences off the table, which is a pretty big constraint. It’s totally valid to say what keeps me from not stealing is the government punishing me, my friends shaming me, or being worried about those things happening! I overall made a good faith effort to be positive in that interaction. They responded “so you have no reasons.” I blocked them, obviously.

4. He recently went on a YouTube channel/podcast and did a much better job of presenting some of the ideas he published in the post, which makes me think some of the problems with the post was that he didn’t have an editor or, being the keyboard savant he was, pounded them out in a couple sittings. Nothing wrong with that! But I will be critical of the ideas as they are presented; if anything, to challenge him to communicate them better.

5. Someone I typically cited in these conversations.

6. Oversimplifying as this post is long. Check out this YouTube video by Alex O’Conner and Joe Folley, who discuss it

7. Okay, this is definitely a controversial claim, but I recommend reading this great intro on Skepticism by Duncan Prichard. He puts forward a Wittgensteinian response to the problem of skepticism with the idea that our reasoning or that the assumption that we aren’t a brain in a vat is arational - we can’t really do any philosophy without assuming it is the case. So, for instance, you can’t bring up the brain in the vat as a valid problem or thought experiment without presupposing or using our reasoning capabilities in the first place.

8. I’m saying “may” here to hedge, as I’m not very physics literate. I need to read more on this issue. The point is, there are lots of non-scientific things that theistic philosophers assert as true when they are indeed false or incompatible with our best scientific understanding of the universe.

9. This section was a complete digression from the point of this post, and added at a late stage of drafting. I would typically edit it out for relevance (Adelstein isn’t making these claims), but I think it’s somewhat relevant to substack discourse (I see a lot of great substackers and students who were definitely smarter than me take these ideas more seriously) and so I’ll keep it in.

10. And believe me, this was 2012, I was a New Atheist shitlord, all jacked up on Carl Sagan and Richard Feynmen memes. I wanted to be a physicist. But I couldn’t. To pretend that I can answer this complicated empirical/theoretical physics question would make me a charlatan, so I won’t.

11. I don’t think Adelstein either explains, understands, or engages with this version of the argument properly, as he seems to just get mad and call people ignorant when they point that out.

12. One way of expressing this is that Adelstein, and many Christian Apologists, want to have the conversation about probability given evidence. Meanwhile, atheists want to have the conversation about what qualifies as good evidence. Having chatted with Adelstein today about this in passing, I think this is an understated point of disagreement.

13. I’m not the only one to point this out

14. I know it’s a cringe thing that woke people do, where they object to broadly true stereotypes (“not all men are stronger than all women”); I want to be clear that that’s not what I’m doing here

15. 33% of my parents’ children fall under this category

16. Notice! We didn’t use Bayesianism to falsify this.

17. Nate Silver isn’t a partisan Democrat

18. Unless they predicted, say, the specific vote totals for each candidate? Outcome is a very broad category outside of math!