This is the Best of All Possible Worlds . . . Really? (Part 2)

In the past few years while I was researching and thinking about a possible challenge to Perfect Being Theology, I stumbled across the argument I’m sharing here.[1] So far the other philosophers and theology minded people I’ve shared it with have all received it positively. I do think it’s interesting or perhaps ironic that I only came to these conclusions after Eden was diagnosed.

So there is a broad consensus in philosophy of religion that our world is not the best of all possible worlds and that there is no best of all possible worlds. I’ve only been able to find one dissenting voice to this consensus.[2] But here I argue that this is simply not the case. Given one prerequisite (which I do admit is a big one) it’s highly likely that we live in the best of all possible worlds. And if this is true, that should be able to provide us with some much needed comfort despite not being able to know why. Or to say it differently, if this is true the primary reason for suffering is that things would be even worse without it.

Now the reasons for the consensus opinion are pretty obvious. As Plantinga explains;

Just as there is no greatest prime number, so perhaps there is no best of all possible worlds. Perhaps for any world you mention, replete with dancing girls and deliriously happy sentient creatures, there is an even better world, containing even more dancing girls and deliriously happy sentient creatures.[3] 

And the same goes for islands. No matter how great an island is, no matter how many Nubian maidens and dancing girls adorn it, there could always be a greater one with twice as many, for example. The qualities that make for greatness in islands-number of palm trees, amount and quality of coconuts, for example-most of these qualities have no intrinsic maximum. That is, there is no degree of productivity or number of palm trees (or of dancing girls) such that it is impossible that an island display more of that quality.[4]

There are obviously many things wrong with this world that could be different. These range from things that are very serious such as wars and the malice and contempt humans show to each other to the very mundane such as that on the morning of September 13^th 2019, I had a very unsatisfying cup of coffee. The world would obviously be a much better place if there were no wars and if everyone treated each other in virtuous and kind ways. And it would also be a better place if my coffee on September 13^th had been satisfying, though the degree of that improvement would be dramatically less. From these observations and the standard assumptions of perfect being theology a valid argument for a weak type of atheism follows:

1.               God would create the best of all possible worlds.

2.               This is not the best of all possible worlds.

3.               Therefore God did not create this world.

And of course, an inference to draw from 3 is that God did not create this world because He does not exist. Conversely, Gotfried Leibniz reversed this argument to claim that we must live in the best of all possible worlds. Leibniz was already quite convinced that God existed and is omnibenevolent, omnipotent, and omniscient, so he argued that we must live in the best of all possible worlds, the prima facie reasons and evidence to the country withstanding.[5] 

1.               God would create the best of all possible worlds.

2.               God exists.

3.               Therefore this is the best of all possible worlds

This argument was mocked by Voltaire in Candide, as it seemed incredibly obvious that this is not the best of all possible worlds, but it’s still a better argument than it initially appears to be.[6] Leibniz was in a sense, saying that he trusted in who and what God is more than he trusted in the apparent facts about this world. Now while I think we can commend and praise the piety Leibniz was demonstrating there is an obvious problem. It is incredibly obvious that there are many many ways (perhaps an infinite set) that this world could be better, and these ways fall under both the broadly logically and the practical senses of possibilities. Hence, it seems that we must accept that our basic intuitions and perceptions about the world are way off to accept Leibniz’s argument. The degree of the error Leibniz’s argument demands of us is no small thing. Almost, if not all, of our basic intuitions and perceptions about how things could be, or what is and is not good, must be way off.

Both our modal and moral reasoning must be seriously misguided if Leibniz is right.[7] If this were indeed the case, we risk falling into the trap of having universal epistemic defeater (see the footnote if you are unfamiliar with this term) for all our perceptions and intuitions.[8] This is quite a high price to pay.

So there are three possible answers to the question of a best of all possible worlds. We can claim:

(1) there is a best of all possible worlds, we can claim

(2) there are multiple “best” worlds that are different but are also equal in value, or we can claim

(3) that there is no best of all possible worlds.

This third solution to this problem is what Plantinga argued for above. The trouble here is for every world we can conceive of there is always a way to make it better. Suppose everyone on Earth was a virtuous person who always did what was morally right and lived happy lives absent unnecessary suffering. This is conceivable as there is nothing prima facie contradictory about it and thus it is a possible world.[9] It would be a better world than the one we live in. So we might argue that such a world would be the best of all possible worlds. But suppose instead there was a world where there were two such Earths. As there would be more virtuous and happy people in this second world it would be better than the first. This would be the same for a world with three such Earths, a world with four such Earths, and so on.

There are perhaps an infinite number of ways to express this problem. While there are some legitimate questions about what constitutes a better possible world there are enough clear and obvious examples of a better possibility to make this problem apparent. The “best of all possible worlds” seems to suffer from a problem of infinite progression as there is no upper limit on the ways and number of ways things could be better. For every best world we can conceive of there is always a better one as I could simply add one more good thing and thus create a better world. Even if we reduce the addition to something as mundane as me having a more satisfying cup of coffee on September 13^th the problem still emerges.  As there does not seem to be an easy way around this problem and the actual world conceivably could be better, we are left with (2) or (3).[10]

Now it’s important to clarify some distinctions here as there are different senses of “possible.” Something is broadly logically possible if it is simply not incoherent. That is there is the absence of a contradiction or contrary relationship within it. A square circle is impossible for this reason and thus is not broadly logically possible. But there are things that are broadly logically possible that we still call impossible. Plantinga once gave the useful example of Voltaire swimming the Atlantic.[11] Voltaire swimming the Atlantic is impossible, but it’s not incoherent and thus is broadly logically possible. So there are “impossible” things that are not impossible in the sense of square circles. This is what we usually mean in non-technical language when we say something is impossible. i.e. Will the U.S. Government ever operate efficiently? No, that’s impossible.

It’s clearly not impossible in the sense of being contradictory or incoherent, but there is something about it such that we can say no it’s impossible for Voltaire to swim the Atlantic or the U.S. Government to be efficient. Different terms are used for this idea, but the most common one is feasability. So while it’s not logically impossible for Voltaire to swim the Atlantic, we can say that such a thing is not feasible as we know enough about human beings to know that is not something they can’t do. So there is possibility in the broadly logical sense and possibility in this feasible sense. And if you want to get very technical, we can even talk about possibility in a metaphysical sense, but I don’t think that’s needed here.

So the reason this distinction matters is that it’s now pretty common to acknowledge that it is broadly logically possible for God to have created a better world, but it’s also common to push back that it simply wasn’t feasible (possible in this weaker sense) for Him to do so. So God could have created a world where cancer and similar things do not exist and cannot exist, but in order to do so, He would have had to change other things, like how basic genetics and biology work, how the laws of physics operate, or the ability of human beings to choose to smoke cigarettes. Since God wants genetics, biology, and the laws of physics to work a certain way, it’s then not feasible for Him to make a world without cancer and thus a world with cancer is the best feasible world.

Now this is a reasonable distinction that takes some of the teeth out of the more devastating critiques that were made of the Leibnizian claim. It just is the case that reality requires tradeoffs.[12] If we want A and A logically entails B, then we simply have to accept B when we get A. There is a large human tendency to attempt to rebel against this basic truth of life, but such rebellions have always ended in disaster. So modifying the claim to “the best of all feasible worlds” does it make it better.

However this does not completely overcome the problem as there are many things that are good, seem immediately and obviously feasible, and simply are not so. So suppose that a world without cancer is just not feasible, which is the direction this type of reasoning goes. It’s still the case that it seems perfectly feasible for my daughter to not have cancer. That is perhaps cancer as a category needs to exist for other reasons. Even granting that it seems both perfectly feasible and perfectly good that my daughter not be in that category. And, with the possibility of a few exceptions, it seems that all the other parents of children with cancer and all the other people who have cancer can say the same.[13] Essentially, “Sure, perhaps we just have to grant that world without cancer is not feasible, but surly it is feasible and good that I not have cancer.” Claims like this seem quite right to me, at least some of them certainly are (see footnote 13), and we can easily find the same problem in many other areas even with the extra distinction of feasibility. So while feasibility improves things it still leaves us with the no upper limit problem. For every good feasible world you give me, I can find another one the is at least prima facie also feasible and is better by just adding one more good thing (that is feasible).

A proponent of the use of feasibility here will need to argue that those other feasible worlds I can point to are not actually feasible, and that would do it. But this seems to be a task that no human being could actually accomplish as the infinite or near infinite complexities that arise from considering how the world would be different if say, “Eden did not have cancer” was true, are simply beyond human capacity to consider. It is broadly possible that there is an infeasibility, contradiction, or contrary deeply buried somewhere within how the world would change if “Eden did not have cancer,” but good luck dragging it out. So since there is no prima facie or detectable lower-level infeasibility with me declaring, “The world would be better if Eden didn’t have cancer,” then we still have a variant of the no upper limit problem. And of course you can run this example for numerous other things (not a near infinite, but still a very high number). I only used Eden’s cancer as an example because that’s what I’m writing about. I suspect the proponents of feasibility will have one more retort to turn to, but I’m going to deal with that below as I think that retort is on the right track.

Now there is an overwhelming tendency in analytic philosophy to try and put everything

into numbers or probability that can be broken down into numbers. And while this does work sometimes and the tendency is understandable, it does not always work. Here our tendency to do that with possible world semantics has led us astray. I think this is why the consensus about the no upper limit problem has emerged and also why it is wrong.

So in the broadest possible sense, is it possible for one thing to be of such great quality that it is better than two things, or is it possible for two things to be lesser than one? Yes and there are many examples of this if we stop thinking about things in purely mathematical ways. There can be an exceptionally good steak, that is well prepared, so it is significantly better than two, three, or even four steaks of much less quality. Or there is nearly unanimous assent among emotionally healthy, well-adjusted people that it is a good thing for a man to get married . . . at least most of the time.

We can certainly find exceptions, but we celebrate weddings because we think that they are a good things. But now suppose we increase the number of the man’s wives to two, three, or even more. Has this just made things better? No quite clearly not for many obvious and some not so obvious reasons. There is a large consensus that while having a wife is a good thing, simply adding another one does not make things better. Rather it makes things worse. So simply adding “one more good thing” does not always make things better. A wife is a good thing, and yet adding one more does not make things better.

What this illustrates is how thinking about things in a purely quantitative way can lead us astray. Take Plantinga’s example of the island with dancing girls. Does adding the existence of another such island necessarily make things better? No because the complications of adding the existence of another island could cause negative effects. The additional island could disrupt weather patterns so that both islands have a worse climate. It could be the case that the existence of more people strains existing resources and leads to a lower quality of life for everyone. If the first island depended on tourism, adding a second might split the market. Now it is still possible that a second island will make the world better. We cannot rule that out. But the what ifs do demonstrate that making things better is not as simple as just adding one more. It turns out that an entire possible world is an extremely complex thing, and the difficulty entailed in actually making it better is inadvertently cloaked or hidden by possible world semantics.

So would the world be a better place if on September 19th, 2019 my cup of coffee had been satisfying? If and only if everything else about the world remained equal then the answer is yes. Or to be a bit more precise, if everything else remained equal, if everything else improved, or if everything else was better on balance. But how can I or anyone else know that? Possible world semantics has hidden the assumption that all else will remain equal, but that’s almost certainly false. But even granting that the other two options are still big ifs, that neither I nor any other human being are in position to know. Given the complexity of the world, it is unlikely that even a change that small would leave everything else unaffected.

The world is just too complex and interconnected for us to add and take away things and leave everything else untouched. Further, we human beings are not in a position to know how such things will actually play out. The butterfly effect hits us hard here, and this seems to be the best and strongest point of skeptical theism. Possible world semantics, despite being a good philosophical tool for many things, have inadvertently led us astray here by making it appear as if adding “just one more good thing” is simple.[14] It’s not.

So we should reconsider the consensus on the best of all possible worlds. As it’s really not as simple as just adding “one more good thing,” then the idea that this is the best of all possible and feasible worlds becomes much more reasonable . . . if God is in the picture. If God is already a live possibility or accepted, then He is the only thing that would have the capacity and capability of accounting for the complexities of adding not just one more good thing, but as many good things as are possible and feasible. This is also the further retort that “the best feasible world” proponent can make that I think is on the right track. With this consideration Leibniz’s argument looks much more plausible. It’s just framed in an imprecise way, but the thrust of his point actually stands up.

While we can appreciate that it seems prima facia obvious that this world could be better in many ways big and small, this is to ignore just how complicated it is to add “just one more good thing.” So on balance I argue that if God exists, it’s quite likely that we do live in the best of all possible and feasible worlds. I am not in position to say that no extra good can be added, I just spent a lot of time arguing that, but given God’s existence and a consideration of how un-simple this problem is, it’s likely that this is the best that things can be. Leibniz was on the right track, we just need to be more precise than he was.

And this is not to say that we are wrong in being able to assess possibilities. We are not wrong in the modal sense. Many things really “could” be different. Rather we are wrong in that we are not in a position to completely and totally assess the value of those alternative possibilities due to the unforeseen, and in many cases unforeseeable, complexities that would arise from even the simplest of changes.

The world clearly would be better if my daughter didn't have cancer, or if the cancer was treatable, could be removed, or similar things. But it is impossible that this change would leave everything else the same. Necessarily other things would also change as a result of Eden not having cancer. And so, I am arguing, no human being has the knowledge, wisdom, or even basic brain power to determine what the overall value of those changes would be.

Now I'm probably better able to guess how not having a good cup of coffee would change things as opposed to something big like WWII never happened. Likewise if Karl Marx had never existed so that horrific philosophy of Marxism never existed, or if Descartes and Kant hadn't sent philosophy in the completely wrong direction for so long. I’m quite committed to those being bad things, but would it actually make the world on balance better if they had not happened? I’m arguing that I can’t know and that you can’t either. But if God exists, He can and does.

So to bring it back to Eden, perhaps a young child seeing our whole church publicly pray for Eden, or much of our community hold a massive benefit for her, is inspired to grow up and found a research organization that develops treatments for cancer and saves the lives of hundreds, thousands, or millions of people. Maybe instead he or she founds a ministry that serves people and leads them to Christ. Maybe my family and I will be able to consistently demonstrate resilience and faith (The jury is still out on that. Greater men than I broken have been broken by much lesser things.) and that example can both point people to Christ and inspire them to virtue and thus lead to less SPS in the world.[15] Maybe the positive changes caused will not even manifest for decades or longer. Or maybe it's just a bad thing because my daughter has inoperable brain cancer. I don't know because I'm too limited and so are you.

It's possible the whole world and Eden's life would be better if she didn't have cancer, but it's also possible that on balance the world and Eden’s life would be worse.

So we are not wrong in assessing that the world could be other than it is. It is broadly possible that my cup of coffee could have been better or that my daughter could not have cancer. There is no prima facie contradiction in asserting those things could be different. It's still possible (and I would argue even likely) that some of those alternative possibilities that we see do contain buried contradictions and so some perceived alternative possibilities are in fact no such thing. For example, to assert that a person is allergic to tomatoes but not to ketchup is a contradiction, but if you were not aware of the fact that ketchup is made from tomatoes you would have difficulty seeing the contradiction.

The reality of the contradiction is buried with the concept of what ketchup is. And sometimes contradictions are buried much further down than this. So it is quite likely that some of the alternative possibilities we perceive do contain contraries and contradictions and thus are not actual possibilities, however it is equally likely that not all alternative possibilities are this way. Some things really could be different and to assert otherwise is an extremely strong statement that is probably impossible to justify.[16] 

So with all those considerations we can make an inductive argument for this being the best of all feasible worlds that I argue is stronger than Leibniz’s deductive version.

1.     It appears prima facia obvious that our world could be better than it is. (It is broadly logical possible that things could be different and some of those differences are prima facia obviously better in a moral/value sense)

2.     Changing anything in a possible world entails an immense amount of complexity and is highly likely to have unforeseen consequences.

3.     Human beings are not an epistemic position to assess all of these complexities.

4.     If He exists, God would be in such a position.

5.     If He exists, God would make any actual world the best it can be.

6.     Therefore if God exists, it is likely that our prima facia observations are wrong, and this is the best of all possible and feasible worlds.

While not deductive, unlike Leibniz’s version, this argument accounts for the reality that it does seem like things could be better. But as we are not in position to know that and we have good reason to think God would make things as good as they can be, we should conclude that our perceptions are mistaken, and this is as good as things can be. As Joseph says to his brothers in Genesis, “You intended to harm me, but God intended it for good to accomplish what is now being done, the saving of many lives.”[17] God also seems to say something like this to Job.

So the consensus on there being no best of all possible worlds is probably wrong. It’s very likely this is the best of all possible worlds. Leibniz was on the right track even though his actual argument was imperfect.  Contrary to determinism and some of the other reasons that have been proposed, this I argue, is a good reason to claim our perceptions about possible worlds are off. It’s not that our basic perceptions of possibility are wrong, rather it’s that the world is much too complicated for us to say that we can add one more good thing without causing a host of additional problems.

To conclude this section, it also seems to me that this is at least part of what it means to have faith in God. God says, more or less, “Trust me. I know what I am doing, even though it doesn’t seem that way to you.” Do you believe Him or not? Many of the great heroes of the faith that the scriptures praise practiced faith this way wherein they believed in God more than they trusted their own perceptions.[18] In some cases they literally trusted God more than their own eyes. If a human being or human institution suggests we should trust them more than we trust our own eyes we should immediately get very suspicious.[19] But if God is who He says He is, then there is good reason to trust Him over my own eyes.

 Notes

______________________________________________________________________________________

[1] Phil Kallberg, “Is a Perfect Being Possible? Using Possible World Semantics to Analyze the Concept of God,” Paper presented at the 2025 meeting of the Society for Pentecostal Studies at Northwest University, Kirkland, WA.

[2] The late John C Wingard Jr. It’s likely that there are others that I am unaware of, but the fact that I could only find one dissenting voice indicates how strong the consensus is.

[3] Plantinga, God, Freedom, and Evil, 61.

[4] Plantinga, God, Freedom, and Evil, 90-91.

[5] Gottfried Wilhelm Leibniz, Theodicy. Trans. E. M. Huggard. (Charleston, SC: BiblioBazaar, 2007).

[6] Voltaire, Candide and Other Stories. Trans. Roger Pearson. (New York: Oxford University Press, 2006).

[7] Modal reasoning being our reasoning about possibility, necessity, and impossibility. i.e. if I think something could be different, but that is not the case, then my modal reasoning has gone wrong or at least I have made a mistake.

[8] Universal means all, epistemic refers to things that we perceive, and a defeater is a belief that invalidates/defeats another belief. So this just means a defeater that would undermine all our most basic perceptions. I have illustrated this by telling my students that I trust my basic perceptions, but if one of them said to me, “Prof. Kallberg, we sneaked some LSD into your coffee before you came to class,” at that point I have good reason to no longer trust my perceptions. The LSD would function as universal epistemic defeater until it worked it’s way out of my system.

[9] There are objections that such a world truly is possible such as Plantinga’s own free will defense, but I’m leaving this aside for the sake of brevity. Plantinga, God, Freedom, and Evil, 7-64.

[10] (2) does not seem to be a widely considered option as it faces two additional problems. First simplicity dictates that in the absence of a good reason to do so we shouldn’t posit two or more things when one will suffice. So why posit two best of all possible worlds(BPWs) when one will suffice? Second, it seems difficult to maintain the idea that multiple BPWs can be different and have equal value. For example, ten $100 dollar bills are equal in value to a check made out for $1000. There is both a clear difference and an equality. However in some contexts a check will be less valuable as it must be cashed or deposited for the money to be obtained. Conversely a check is also, in theory, more secure as only the person it is made out to can legally cash it. So while the monetary value of the cash and the check are equal the practical value of them will still differ based on the surrounding context. And this will hold for BPWs that are supposed to have equal value. It seems very likely that upon deeper analysis some difference in value will always emerge when we are comparing proposed BPWs. Hence (2) is unstable as given any BPWs of supposed equal value a winner will almost certainly always emerge on a deeper analysis. The lone dissenting voice I could find, John C Wingard Jr, did argue for (2) but the meat of his argument seemed to be that it seemed right to him. While it is nice that he was honest, this is not much of an argument.

[11] Plantinga, God, Freedom, and Evil, 35. 

[12] This seems to be something that many of my fellow Americans have forgotten, but that’s a whole other thing.

[13] Exceptions we can imagine would be things like if I get lung cancer after smoking 12 cigars a day for 20 years. There’s an element there that is tragic, but we can also argue that there is a sense of justice as I would be reaping the rewards of my behavior. Real world examples like this will quickly get very complicated, but it’s certainly plausible that some things are good in the sense of being the natural rewards of our own behavior and if that behavior is bad than there is a sense in which it is good for the rewards to be bad. So while I don’t wish cancer on anyone, regardless of personal behavior, there is a sense in which these “bad” results can be “good” as they are just.

[14] When you look at how analytic philosophers like to write about and represent possible worlds this is quickly apparent. They will do things like compare possible world 1 (PW1) with possible world 2 (PW2) and then represent adding good to PW2 by making it into PW2*, PW2¹, or something similar. This makes it easy to miss how complicated it actually is to “just add one more good thing.” It’s broadly possible that the “one more good thing” will introduce a contradiction to PW2 and thus PW2 is incoherent, but we will never be able to figure that out as long as the modified PW2 is represented by just adding another symbol. This makes it seem like things are easy and simple. A little reflection shows that they are not.

[15] If more people were virtuous, there would be less suffering in the world and thus less seemingly pointless suffering in the world. This is just a truism.

[16] Meaning that determinism in the sense of it is impossible that things could be otherwise, is almost certainly false. So when I argue here that this is the best of all feasible worlds, I do not mean determinism. Rather I mean that this is both the best possible world in the sense of modality, and it is also the best feasible world in the sense of value and/or morality.

[17] Genesis 50:20

[18] See Hebrews 11.

[19] Of course there are possible exceptions. If I’m blind, I should trust other people over my own eyes. If I know that I’m hallucinating I shouldn’t trust my own perceptions, and so on.