Fine-Tuning, Made By God

A simple proof that most rebuttals to the fine-tuning argument go wrong

Starting around 1980, physicists discovered something shocking—the most revolutionary find relevant to God’s existence in centuries. Specifically, they discovered that the universe was finely-tuned—that that values that are plugged into the equations of physics fall in a hugely narrow range needed to give rise to complex structures. The cosmological constant governs the expansion rate of the universe; if it had been a bit different stronger, everything would have immediately flown apart, while if it had been a bit weaker, everything would have rapidly collapsed in on itself.

The range of values the cosmological constant could take on, consistent with our basic picture of physics, is vast. In order for any complex structures to form, it has to fall in a tiny sliver of that range—about one part in 10^120 of the range (this is an oversimplification, see here for more details). This is a coincidence on the order of throwing a dart across the known universe and hitting any particular atom (actually, that’s about 40 orders of magnitude too low, but who’s counting?)

There are a lot of objections to the fine-tuning argument—I’ve addressed them pretty comprehensively in this article. Many of them require rather lengthy and technical expositions. But I think there’s one argument that can be made pretty quickly, that shows that arguments against fine-tuning prove too much. An obviously successful argument couldn’t be successful if many of the arguments against fine-tuning worked.

Suppose that physicists uncovered the initial state of the universe, and discovered that it was arranged to spell out the words “made by God, through love,” in every language. That would be pretty damn good evidence for the existence of God. Maybe it wouldn’t get us all the way to a perfect God—depending on one’s priors—but it would clearly make some sort of design hypothesis quite reasonable to believe.

However, most of the objections to fine-tuning are equally objections to thinking that a designer exists in such a world.

For example, one of the most common objections to fine-tuning is that the claim of improbability is bogus: we only call something improbable if it turns out a way that it usually does not. We call it improbable if a person gets a royal flush in poker, because royal flushes are rare. But then how can we say that the cosmological constant is rare—we only have one example of it taking on a value?

But by the same logic, we couldn’t say it’s improbable that the initial conditions would spell out “made by God almighty, through love,” in every language. This objection therefore proves too much—if it was right, the argument from design would be impotent even if the initial conditions spelled out made by God. Thus, there must be something wrong with the argument. And there is—it gets probability wrong, as I explain in more detail here!

In fact, by this logic, we couldn’t even have much evidence for continental drift. Much of the evidence that Africa and South America used to be conjoined is that the fossil record indicates that they had similar plant and animal life up until about 140 million years ago, when they started to diverge. But by the above logic, that can’t be declared improbable by chance, because it only happened once!

Here’s another common objection to fine-tuning: we can’t get evidence for God from fine-tuning because we didn’t predict it in advance. Evidence comes from experimentally discovering stuff that a theory predicts! But because no one used the God hypothesis to predict fine-tuning in physics before it was discovered, fine-tuning can’t be evidence for God.

Once again, I think this objection is based on a confused view of evidence, that would rule out our knowledge of continental drift, most of the evidence for evolution, and evidence for dark matter. But it can be rebutted more simply, just by noting that it would make the “made by God,” design argument illegitimate. If evidence needs to be in the form of correct predictions, then so long as no theists predicted the initial conditions spelling out “made by God, through love,” in every language, that couldn’t be evidence for God. But obviously, it could! In fact, it would be insanely strong evidence for God’s existence! Therefore, this objection proves too much.

A third objection is called the measure problem. How is it that we decide upon the fine-tuning range? Most of the time, the relevant values can take any value over an infinite range! Thus, picking out a finite range is arbitrary.

Now, this objection has the problem of stupendously misstating the physics behind much of fine-tuning. Regarding the cosmological constant, for instance, physicists don’t assume that each value of the constant is equally likely, but instead note that many different values affect the cosmological constant. These values are determined at small scales, but almost exactly cancel out at large scales, so that the universe doesn’t expand either too quickly or too slowly. This is a coincidence on the order of the number of seconds humans have spent happy minus the number of seconds humans have spent unhappy equaling just twelve seconds. The values going into the fine-tuning argument are about 10^120 times greater than the value of the cosmological constant, and when one allows these values to cancel out, they get a slight positive cosmological constant.

It also misstates probability. The value of the constants can only be calculated across scales consistent with the basic standard model, and only a tiny slice of values consistent with the standard model produce complex structures. This would be like, to use an analogy from Robin Collins, hitting a tiny red dot on a dartboard surrounded by a white area. Even if most of the dart board is shrouded in darkness, this is very improbable! It doesn’t matter the total percentage of the dartboard that is occupied by red, because the surrounding region is almost all white—hitting the tiny red dot is thus astronomically improbable.

One can also get their probabilities to nicely sum to 1 by taking a ~log prior across the values of the constants! When they do that, once again, they get a ridiculous degree of fine-tuning.

In addition, as Robin Collins notes, this is a very weird reason to reject the fine-tuning argument. On this picture, if the range of values that the constants could take on was smaller, then the fine-tuning argument would work. If the constants had to take on a value between, say, 500,000 and 1,000,000, then if their values could be any value from 0 to googol, there would be a fine-tuning argument. But, if their upper bound is infinite rather than googol, then the argument stops working. Surely raising the range of inhospitable values shouldn’t weaken the fine-tuning argument!

But once again, we don’t need to bother with all that. The objection simply proves too much. If it were right, then the initial conditions spelling out “made by God, with love” in every language wouldn’t be evidence. After all, there are an infinite number of shapes that the initial conditions could conceivably take on. In fact, the math behind the actual fine-tuning is considerably less messy than the math behind figuring out the odds of the initial conditions spelling out “made by God, with love,” as there are probably infinite different ways the initial conditions could spell that out, and no particularly elegant measured over the values that would render that improbable.

Another common objection to the fine-tuning argument is that the apparent fine-tuning might be explained by deeper physics. Perhaps, once we have the deepest theory of everything, all the apparent fine-tuning will go away! But once again, this objection could be made in response to the initial conditions spelling out made by God through love, in every language. Yes, deeper physics could set the initial conditions to that, but they could also conceivably set it to any other value. It’s thus super unlikely that deeper physics would set the initial conditions to spell out “made by God,” rather than to some other value. In other words, this just pushes fine-tuning back a level: why is the deeper physics finely-tuned to produce a life-permitting value of the cosmological constant?

(This is especially so because the deeper physics takes place at small scales, and absent teleology, the actions at small scales aren’t affected by what they do at large scales).

Lastly, people frequently raise the electrons in love objection. On this view, theism doesn’t explain fine-tuning, because God could make life no matter what the cosmological constant was. He could either create life via miracle or by designing the psychophysical laws so that the non-finely-tuned atomic soup gives rise to valuable consciousness.

Now, I think this objection might be successful at establishing the God isn’t guaranteed to make a finely-tuned universe, but it’s totally unsuccessful in establishing that theism doesn’t predict a finely-tuned universe with higher probability than single universe naturalism. God basically has four ways of creating agents:

1. Creating disembodied minds.

2. Creating a non-finely-tuned universe with psychophysical laws that give rise to create agents.

3. Creating a non-finely-tuned universe, but then making agents by miracle.

4. Creating a finely-tuned universe that naturally produces agents.

On single-universe naturalism, as described, the odds of fine-tuning are on the order of 1/10^120. In contrast, the odds of God choosing option four are obviously way more than that—if only because he might do all four in different universes! As Scott Alexander says (talking about other absurd probability claims):

Well, actually, we do know. It’s probably not the 10^-66 one, because nothing is ever 10^-66 and you should never use that number.

Let me try to justify this.

Consider which of the following seems intuitively more likely:

[Some scenario that he is denying is roughly 10^-66 probability]

Or second, that despite our best efforts, a research institute completes an unfriendly superintelligence. They are seconds away from running the program for the first time when, just as the lead researcher’s finger hovers over the ENTER key, a tornado roars into the laboratory. The researcher is sucked high into the air. There he is struck by a meteorite hurtling through the upper atmosphere, which knocks him onto the rooftop of a nearby building. He survives the landing, but unfortunately at precisely that moment the building is blown up by Al Qaeda. His charred corpse is flung into the street nearby. As the rubble settles, his face is covered by a stray sheet of newspaper; the headline reads 2016 PRESIDENTIAL ELECTION ENDS WITH TRUMP AND SANDERS IN PERFECT TIE. In small print near the bottom it also lists the winning Powerball numbers, which perfectly match those on a lottery ticket in the researcher’s pocket. Which is actually kind of funny, because he just won the same lottery last week.

Well, the per-second probability of getting sucked into the air by a tornado is 10^-12; that of being struck by a meteorite 10^-16; that of being blown up by a terrorist 10^-15. The chance of the next election being Sanders vs. Trump is 10^-4, and the chance of an election ending in an electoral tie about 10^-2. The chance of winning the Powerball is 10^-8 so winning it twice in a row is 10^-16. Chain all of those together, and you get 10^-65. On the other hand, Matthews thinks it’s perfectly reasonable to throw out numbers like 10^-66 when talking about the effect of x-risk donations. To take that number seriously is to assert that the second scenario is ten times more likely than the first!

(Note: Scott is only talking about a scenario with odds of 10^-65, which is about 55 order of magnitude more than the fine-tuning probabilities).

But once again, we don’t need to go through this entire rigamarole. One could make the same electrons in love response to the initial conditions spelling out “made by God, with love” in every language. After all, God has lots of ways of providing a message: he could design the psychophysical laws so that those words flash in our conscious field of view once a day. He could have the heavens declare it by miracle. Thus, one could similarly argue: God providing the message that he made the world could be done no matter what the initial conditions were, so theism assigns an even probability distribution across the values.

This is clearly wrong. The electrons in love response to fine-tuning is wrong for the same reason. Yes, maybe the odds of God making a finely-tuned universe aren’t that high, but they’re obviously not improbable on the order of 1/googol. While perhaps the electrons in love objection can work in tandem with a positive explanation of fine-tuning, it is clearly not enough to single-handedly defang fine-tuning.

This is a helpful exercise when approaching objections to fine-tuning. There are lots and lots of things people say. But for most of them, think to yourself: “would this disqualify the design argument if the initial conditions spelled out ‘made by God, with love’ in every language?” If it would, then the argument has gone badly wrong somewhere. Indeed, it seems that most of the objections to fine-tuning fail this basic challenge.