About Rationally Speaking

Rationally Speaking is a blog maintained by Prof. Massimo Pigliucci, a philosopher at the City University of New York. The blog reflects the Enlightenment figure Marquis de Condorcet's idea of what a public intellectual (yes, we know, that's such a bad word) ought to be: someone who devotes himself to "the tracking down of prejudices in the hiding places where priests, the schools, the government, and all long-established institutions had gathered and protected them." You're welcome. Please notice that the contents of this blog can be reprinted under the standard Creative Commons license.

Monday, April 16, 2012

The Simulation Hypothesis and the problem of natural evil, part IV

by David Kyle Johnson

[This post belongs to a four-part series that Rationally Speaking is running with one of our podcast guests, Prof. David Kyle Johnson of the Department of Philosophy at King’s College in Wilkes-Barre, PA. Kyle's philosophical specializations include logic, metaphysics, epistemology and philosophy of religion. His publications include “God, Fatalism, and Temporal Ontology” (Religious Studies) and “Natural Evil and the Simulation Hypothesis” (Philo). He also teaches and has published extensively on the interaction between philosophy and popular culture, including a textbook (Introducing Philosophy Through Pop Culture) and two edited volumes in the Wiley/Blackwell series (on Christopher Nolan’s Inception and NBC’s Heroes). He maintains the blog Plato on Pop with William Irwin for Psychology Today, and hosts a podcast.] [go to part Ipart II, or part III]

So, finally, we come to it. Why are many theists committed to believing we live in a computer simulation? Because, even if we grant them knowledge of God’s existence, the only way to make sense of why our universe has natural evil as an inevitable consequence of the very laws that govern it, without violating God’s moral integrity, is to assume that someone else designed our universe — and the most likely alternate designer is a computer programmer.

To see why, let us return to the Caleb example. I know that Caleb cannot commit a morally heinous action, despite the fact that the evidence seems to suggest that he did. But there are two kinds of evidence that suggest that he committed a morally heinous act. One is the evidence that the cold blooded murder of the infant was carried out by Caleb; the other is the evidence that suggests that the cold blooded murder of an infant is morally heinous. Given that I know that it is impossible for Caleb to commit a morally heinous action, I must maintain that one of these pieces of evidence is faulty. Either someone else committed the crime, or the cold blooded murder of an infant is not actually a morally heinous action (i.e., it is not the kind of action that it is impossible for Caleb to commit). In deciding which piece of evidence to reject, I should clearly reject the one that is least intuitive. Thus, what we should conclude — in fact, what I am sure you had already concluded — is that someone else committed the crime. No matter how good the evidence is that Caleb committed the crime, it could never trump the reasons I have for thinking that the cold blooded killing of an infant is morally heinous.

But the theist who has been granted their claimed evidence and knowledge is in a similar situation in regard to natural evil and God’s existence. The theist has been granted knowledge that God exists, and is tri-omni; thus the theist knows that God is not capable of performing a morally heinous act. Yet, as the problem of natural evil shows, the evidence suggests that he did. Natural disasters are evil, yet they seem to have been woven into the very fabric of our universe by God’s design. However, there are two kinds of evidence here: the evidence that natural disasters are evil (and thus whoever wove them into the very fabric of our universe is not wholly good) and the evidence that God is the designer of our universe. Which piece of evidence do we have less reason to think is accurate; which notion should we reject? The answer is clear: we should reject the notion that God is the designer of our universe. Just like with Caleb, we shouldn’t conclude that the crime in question wasn’t really evil. We should conclude that someone else did it.

Why do we have more reason to reject the notion that a tri-omni God designed our universe than we do to reject the notion that natural disasters are evil? The belief that natural disasters are evil is about as ubiquitous as beliefs can get. That they are evil is why we used to think they were caused by demons. This is why we work tirelessly to prevent them and mitigate the damage they do — tornado detection, tsunami warnings, earthquake proof buildings, cures for diseases — the list goes on. Hell, just making a joke about a natural disaster can make you lose your job as the Aflac duck. If we can agree on one thing as a species, it is that natural disasters are evil! On the flip side, however, we have no reason at all to think that God is the designer of our universe. Even if we have reason to believe our universe is designed, and we know God exists, we have no reason to believe that it was God specifically that designed our universe. Even the best design arguments only point to “a designer.” In fact, the problem of natural evil gives us a specific reason to conclude that it wasn’t God.

But if I have granted the theist knowledge of God’s existence, what is the most likely scenario in which God exists but didn’t design our universe? I suppose it’s possible that, even if God exists, our physical universe still sprang from nothing — but that won’t help us solve the problem of natural evil, for it denies premise (3), another theistic conviction: that God designed the physical universe. So how could God have created the physical universe but be divorced from all moral culpability for the laws that govern ours? Simple: by our universe being a computer simulation — a simulation created by a moral agent within the physical universe (or another simulation within that universe). I’m sure we could think up some other scenarios (Descartes comes to mind), but given what I pointed out in my first blog entry, the computer simulation is the most likely one.

Notice that this solves, perfectly, the logical problem of natural evil; it is a story in which (1), (3) and (4) are all true together. God exists; he designed the physical universe; and natural disasters are a result of the laws that govern our universe. Yet since God didn’t design our (simulated) universe he can’t be morally blamed for that and can still be tri-omni. Of course, one might wonder why God would allow a computer programmer to get away with creating a universe such as ours, but since even some atheists have admitted that God could allow moral agents to perform morally heinous actions, this doesn’t present much of an obstacle. So, the theist can defend their theistic belief from the logical problem of natural evil, but the solution comes at the cost of embracing that we live in a computer simulation.

One might liken this is to granting Michael Behe his intelligent design arguments. We all know that they don’t work, but even if they did they would not provide good reason for thinking God exists. Since there are numerous flaws in our (and other species’) design, we can’t be the handiwork of a perfect being. A creative, but imperfect, alien is a much more likely explanation. Likewise, even if I grant the theist God’s existence, since there is natural evil embedded into the very fabric of our universe, it can’t be the handiwork of a perfect being. A computer programmer is a much more likely explanation.

Of course, the simulation hypothesis is not something the theist is going to want to be tethered to — and that is part of my point. The logical problem of natural evil has been ignored for a long time because of a general, unwarranted, assumption that Plantinga solved it. He did not. Theists need to return their attention to the logical problem of natural evil. Otherwise, they better start arguing that believing we live in a computer simulation isn’t crazy.


  1. But doesn't this beg the question of why would god allow the programmer to do something so extensive and immoral? Isn’t this omni-benevolent being ultimately responsible for anything evil he could prevent? It seems to me that this argument hangs on making the distinction between natural and moral evil that you made in your first post. I’m not certain there is much of a distinction for anybody on the receiving end. You are saying that we intuitively view natural disasters as evil - isn’t evil just evil in some elemental way, programmer or not?

    1. It replaces a natural evil resulting from natural laws by an evil caused by a being exercising free will. The latter is much easier to justify for a theist - because free will is such a good thing to have, we need to accept it may be misused, and so on.

    2. I may be missing the point, but I believe Al's point still stands.
      Giving an agent free will is definitely a good thing, but only when the recipient is capable of using it wisely. Otherwise it's plain irresponsible and, hence, either callous or foolish. The truly moral, smart deed is to grant each agent only as much freedom as they can handle.
      If you want the best for your children, you don't let them do anything they like. You give them increasing autonomy and responsibilities as they show signs of accountability.
      So if one of your creatures ends up creating a simulation with intrinsic hazards for sentient beings living it, you gave them too much free will. It's time to step in and fix it if you can, or accept you blew it if you can't.
      At the end of the day I guess that with or without a simulation, a moral, omnipotent creator of the universe is still contradictory.

    3. Rodolfo,

      You may be right here. But the point, I think, is that evil caused by a free agent is much easier to defend for a theist than natural evil occurring without reason. I am not sure the free-will defence works but, at least, it has some traction and there is a feel of reasonableness to it.

      So, we might say there is a “hard” problem of evil and an “easy” problem (caused by a free agent). What David's argument does is reducing the hard case to the easy one.

    4. Is that really the case though? Is there really that big a difference in the hard vs easy problem of evil? It seems to me that a complication here is that it seems unreasonable to think that the “natural” world could be set up to operate in a way that allows the benefits but prevents the catastrophes. In order for this to line of reasoning to work, the benefits to having the natural world function the way it does, with occasional natural disasters, can’t outweigh the costs in human suffering of natural disasters. But is that really true? The natural world supports billions of people and many of them live quite well. The natural “evil” of disasters, the human suffering, seems to me entangled with the benefits which are actually quite incredible.

    5. Al
      JP is basically right. I don’t think that the “free will” solution to the problem of moral evil is all that convincing either—it definitely has its problems. But theists think it is definitive. In fact, many atheists have “admitted” that the free will solution works. (Again, I disagree.) But I am not offering up a defense of theism. What I am trying to do is show that, even if we grant the theist all kinds of things that we normally wouldn’t grant them, they still land in a very precarious place.

      So, let’s grant them knowledge of God’s existence. Let’s grant them the free will solution to moral evil. The logical problem of natural evil still commits them to something they don’t want to be committed to--that our world is a simulation—for the reasons I have stated.

      Now, the simulation hypothesis does “solve” the logical problem of natural evil by turning natural evil into moral evil, but it does not fall prey to the same problem that Plantinga’s original argument did. Plantinga tried to solve the problem in that way, but failed because he misunderstood the problem; his solution was not a story in which 1, 3 and 4 were all true together. (His story denied 4.) But the simulation hypothesis is a story in which 1,3 and 4 are all true together. Thus, upon the assumption that the free will solution can explain why God would allow a being to create such a world, God’s existence is defended—but at the cost of maintaining that we live in a computer simulation. So the theist still has some work to do.

      Now, if the free will solution doesn’t work—the theist is in trouble, but by argument still stands. It’s still the case that the theist is committed to the simulation hypothesis—it’s the best way out of the problem of natural evil I have raised—but not only is it unpalatable (who wants to be committed to believing we live in a computer?), but it raises another new problem (why would God allow a computer programmer to create such a world?) for which the theist has no answer. So my main point—that theists need to deal with the problem of natural evil, instead of ignoring it because they think Plantinga solved it—is reinforced.

      If you are interested, I dealt with your objection directly in the original paper. You can find it here:


    6. I appreciate your response and don't want to belabor the point but it seems to me that your case for simulation hinges on this one assumption – “After all, he could have designed our universe in such a way that no physical conditions could ever give rise to such atrocities, leaving the amount of evil and suffering we endure solely up to us.”

      That is what I am questioning in my second post; I simple don’t think it’s a valid assumption. In your paper you cite heaven as an example of a place that theists believe in where there is no natural evil, no bad consequences to natural processes. But a belief in heaven and a belief in god are separate but related; there are many people who believe in god but not heaven. For many people who believe in god, this is the only world that exists. An what about people who believe in heaven but don’t see it as a physical place?

      It seems intuitive and self evident (to me at least) to think that the benefits of the natural world must be linked to the costs, the disasters, in just the same way that it is intuitive to view natural catastrophes as “evil.” A world inhabited by humans where natural processes lead to beneficial outcomes but negative consequences to humans are prevented seems too paternalistic a place to also allow the freedom of thought necessary to free will and morality. Such a world might be logically possible but one could argue that it would only be made workable by such a transparent application of divine intervention that it would result in removing the opportunity for development of free will so it would not give rise to the right conditions that make moral choice possible.

      I'm just not that clever, somebody must have developed a simlar ine of reasoning.

    7. Yeah, it's true that some theists don't believe in heaven--but none I have ever come across think that heaven is logically impossible. And it has to be for this to be an legit objection to my argument. Any any objection you apply to the possibility of a world with no natural evil is an objection to the possibility of heaven. "Such a world is not conducive to free will and moral behavior." Well, maybe...but if so, neither is heaven--and if free will and moral behavior are so important, then heaven is not a perfect place.

      I'm not even arguing that such a world is logically possible--I'm arguing that, given their own commitments, the theist can't deny that it is logically possible--but if it is, they are committed to the simulation hypothesis.

      I guess they could turn around and deny that it is logically possible but then (a) they have to face the reasons I have given that it is and (b) that move is an ad hoc rationalization that renders this position irrational. Heaven has been admitted for 2000 years to not only be a logical possibility, but to be a real place--to turn around now and deny that it is even possible to save yourself is clearly just an attempt to dodge a bullet--they don't really believe it. (Not to mention, I still think it's a consequence they wouldn't want to embrace--just like the simulation hypothesis.)

    8. It's not that heaven isn't poosible, it's that it is radically different; it doesn't serve the same purpose. For people who believe in it, heaven is a reward for a moral choice that has already been made. Isn’t it supposed it meant to be some sort of post-moral existence? Just like Eden (your other example) was supposedly pre-moral. Think of heaven’s opposite, hell. Hell is a post moral existence as well because it’s a punishment for a moral choice or choices already made. No need for moral evolution or development, moral choices no longer count in hell.

      Then it's the natural world that is the theatre of morality - those other places not so much. If morality does't count there what happens to your argument?

  2. I wondered that you could show me what theists must believe, as opposed to what they mistakenly believe already that makes more sense to a lot of them. As the thought of placing evil's blame on computers takes away the purposive nature of evil, and that goes against the human instincts from which their Gods arose. On the other hand they don't mind wielding determinism for a purpose, and have easily handled what to us (or me) is a logical contradiction of goals.
    So you had at it, but the result was disappointing. Your concept of evil seems quite different from the dictionary versions, which, if we're NOT considering evil as a force of God or mother nature, should be usable. Destructive acts of nature, if not intentional, are not evil. Arguably without them, we would not have evolved. Neither is life's habit of eating life evil, as again we would not have otherwise evolved. And theists can see this as well as the rest of us, even if a bit of rationalization is needed on their part.
    I won't go into the rest of your argument. It's just remarkably weak, considering that it started off quite well.
    But again, as I had asked earlier, - why is it that theists 'must' believe that we live in a computer simulation. Why wouldn't they instead believe that their God had computational powers in a deterministic universe? That question wasn't answered. There wasn't any 'must' involved for one thing.

    1. Roy
      I’m sorry you were disappointed, but I think you might be a bit confused about my argument.

      First of all, “natural evil” is simply short hand—traditionally, it refers to the suffering natural disasters cause. Certainly, tornados, earthquakes, etc are not evil in and of themselves—they happen on lifeless planets, like Venus and Jupiter, all the time and are not evil. But the suffering they cause on this planet is—and that is what I mean by natural evil—the suffering caused by natural disasters. Puppy killing machines are not evil if they are installed in an empty house, but if I put puppy killing machines in my house and then make my puppy live there—I can’t be considered morally good. And that follows even if my puppy always happens to never be in the wrong place at the wrong time.

      Further, I’m pretty sure it’s false that we need natural disasters in order to evolve. We need death, and survival, etc.—but we can have all of that without hurricanes, and earthquakes, etc. It may be that, as a matter of historical fact, such things have shaped our evolution but that is irrelevant for my argument. Let me explain. What you are saying is that God can be morally justified in allowing natural disasters because they lead to a greater good: our evolution (and eventual existence). But, as everyone who studies these arguments will tell you, God can be morally justified in allowing an evil to produce a greater good if and only if that evil is logically necessary for the existence of that greater good. This is not the case when it comes to natural evil (again, suffering caused by calamities) and the greater good of our existence. Although calamities did shape our evolution, they didn’t have to. And, although our existence was a product of evolution, it didn’t have to be. It is most certainly logically possible that our existence was not a product of evolution. As I pointed out, it’s logically possible for God to have created a world in which we exist but natural disasters do not (such a logical possibility is built right into Christian dogma)—and the fact he didn’t in this world doesn’t change the fact that it is logically possible. He could have. It is logically possible, and everything that is logically possible is within God’s power. And since not doing that, when one could, is morally reprehensible—well, God can’t do something that is morally reprehensible and still be morally perfect, as God is supposed to be.

    2. I am greatly confused by this question however: “Why is it that theists 'must' believe that we live in a computer simulation. Why wouldn't they instead believe that their God had computational powers in a deterministic universe?” I don’t even understand what that is supposed to mean. God would have computational powers, if he existed. So, you are basically asking “why wouldn’t theists believe our universe is deterministic?” I am arguing that the theist is committed to the simulation hypothesis because it is the most viable option they have for resolving the seeming logical incompatibility of 1,3 and 4. How does our universe being deterministic solve that incompatibility? How is that a story in which 1,3 and 4 are all true together? How is God having computational powers in a deterministic universe a story in which 1,3 and 4 are all true together. And even if so, how is that a more viable hypothesis? I’d venture some guesses—but I don’t even know what “God had computational powers in a deterministic universe?” even means, much less how it applies to this problem.

      Are you perhaps asking why don’t theists believe that God is the computer programmer? That can’t be right, because that doesn’t address the problem at all. The problem is “why would god regulate our universe with laws that give rise to natural disasters that kill us randomly”—God being the programmer would just mean he regulated our universe with such laws in a different way (by programming instead of creating).

      Do you mean why wouldn’t they just believe that, when it comes to the laws that govern the universe, God’s hands were tied and he only had one choice—the laws that we have. That doesn’t work. We know these are not the only possible laws. Additionally, God can do whatever is logically possible, and the laws only define what is physically possible—so a different set of laws is logically possible and within God’s power. This would greatly limit God’s power, and make the existence of heaven impossible. So the theist is not going to go for that.

      So, again—I’d answer your question. But I literally don’t know what it means.

    3. Kyle, you wrote: "What you are saying is that God can be morally justified in allowing natural disasters because they lead to a greater good: our evolution (and eventual existence)."
      Not at all, because that would be an argument of those who believe in God, not mine.

      "I’m pretty sure it’s false that we need natural disasters in order to evolve."
      I'm an advocate for adaptive mutation as a theory, and believe that we evolved as a reaction to the problems of experience, to take advantage of accidents, etc. And without the occasional catastrophes, we would not have had to become humans at all.

      "Although calamities did shape our evolution, they didn’t have to. And, although our existence was a product of evolution, it didn’t have to be. It is most certainly logically possible that our existence was not a product of evolution."
      Again, it's my opinion that they had to. One point would be that we might otherwise have remained dinosaurs. And of course to exist as we are was precisely due to how we adapted to the particular natural problems faced. It's logically possible that our existence was not a product of evolution? If so, it's hardly plausible, since even if we were seeded from other worlds, we still had to evolve on this one.

      Re the second item, if theists were inclined, as they are, to believe a God was determining their behavior (and programming is certainly determining), and then judging them for it at the same time, they'd have no problem believing that any 'simulator' was a God, regardless.
      And the theists that I grew up with felt that God let evil exist for his purposes, Devil being there to take the hindmost.

    4. kyle8425, If you say that a natural disaster is a natural evil, then surely an epidemic would be also (e.g. the medieval plagues, the early 20th century flu epidemic). But, as our knowledge of biochemistry increases, it has become apparent that both cancer, as well as natural aging, must be viewed as certain types of "preventable" natural evils: viz, the progression of cancer in a single individual is in many ways like an epidemic; the mechanisms of aging are due (among other things) dimerization or cross-linking of sacharides. That is, both disease and natural aging must also be placed in the category of "natural evil". (This, to me, knowing what I know of physics and chemistry, is as inescapable as any statements about earthquakes).

      So, when you say "we can have death, etc but without the hurricanes", I think that is an utterly inconsistent statement. You've already accepted physics as the cause of hurricanes; so it must be likewise with disease and aging. You can't label one as a natural evil without labeling the other likewise.

    5. Roy

      You don't have to assume that God exists to say "God can be morally justified in allowing natural disasters because they lead to a greater good: our evolution." I'm speaking in hypotheticals. What you are suggesting is that, if God exists, he would be permitted to allow natural disaster becuase it leads to a greater good: our evolution. And, so you argue, if that is true, theists could use that to defend against my argument. I am saying that move is not open to them, because allowing natural disasters is not necessary to produce the good of our existence. God need not even create us via evolution.

      So I think you are still missing my point. As I posted out to Al, all my argument assumes is that it is logically possible that God could have created a world, with us in it, that doesn't have natural evil. You're right--it's not plausible. But it's possible. Sure, even if we were seeded we still evolved. But it's logically possible that God could have just created us, as is, without evolution to live in a world with no natural disasters. God could have had us evolve, but simply "tweaked the process" now and again to get us to human form, instead of relying on catastrophes.

      It is a matter of fact that catastrophes contributed to our evolution; no meteor hit to wipe out the dinosaurs and we wouldn't be here. And, if there is no God, it's very unlikely that we would just spring out of nothing fully formed--evolution is the best (although likely not only) atheistic explanation for the origin of life. But this doesn't keep other scenarios from being logically possible--and since there are logically possible morally superior choices, God should have--must have--chosen them, if he is to be morally perfect.

      Basically, your objection only works if "evolution pushed by natural disasters" is the only logically possible explanation for our existence--and not only does evolution not logically require natural disasters, but evolution is not the only logically possible explanation for life. It's the best, certainly--but it's not the only logically possible one.

    6. As for the second item:

      It sounds like you grew up around staunch Calvinists, who think that God controls everything, right down to our own actions, even causing us to do evil things, and then judging us for doing them, even though he had us do them for his own purposes. I feel for you--I knew some too.

      I didn't include this one because no academic takes this position seriously anymore--not even ones from a Calvinist tradition. It's completely incompatible with any kind of view of morality, free will, moral responsibility, omni-benevolence, etc.

      This view of God does not solve the problem of natural evil at all becuase one still wonders how God could create a world (no matter whether it is physical or a simulation) that has "human killing machines" built right into it, and still be an all-good god. This Calvinist story is not a story in which 1,3 and 4 are true together--(1) is false! God is not all good.

      And not only becuase he causes natural evil--but because he causes all moral evil in the world as well, and then not only judges people for evil actions that he made them do, but makes them burn in hell forever for "doing it" (i.e., for him making them do it). This Calvinist god is worse that Satan. (Ever had a friend that made you murder someone, and then called the cops on you so you'd get the chair?)

      The only way out the Calvinist has is skeptical theism--"God has reasons for being worse than Satan that are beyond us." But I covered the illegitimacy of that response in the post.

    7. "my argument assumes is that it is logically possible that God could have created a world, with us in it, that doesn't have natural evil."
      Kyle, I find that argument incredibly weak, since without your version of natural evil, his "us" would not have been our "us."

      In any case, your earlier proposition was that "In my next entry I will fulfill my promise and show why such theists must believe that we live in a computer simulation."

      Remote possibilities are not musts.

    8. Linas

      Yeah, disease certainly is a natural disaster...and I include that in my original paper. I didn't here for brevity. I think aging would be too--but I don't think we logically need either for evolution to occur. Death might not even be logically needed for evolution. The ability to survive is a fit train only because death makes one unable to reproduce. But one could be unable to reproduce for other reasons. There is a logically possible world no one ages, and no one even dies, and yet evolution occurs--certain traits just make you more or less likely to reproduce (perhaps before you lose the ability/drive to do so). Such a world would either require a lot of space, or be crowded, but it's logically possible. Even if death is necessary for evolution, aging, disease, etc. (natural disasters) are not the only way we die. We can die in a large variety of unnatural ways. So, again, even if evolution requires death, natural disasters are not necessary for evolution. But certainly, disease and aging are not required for "fit traits to work to fixation."

      But, just to be clear: "we can have death, etc but without the hurricanes" is not an inconsistent statement. It doesn't contradict itself. In fact, not only is it consistent, it's true. Since people die for non-hurricane related reasons, we can have death without hurricanes. I think what you meant to say is that it is an incomplete statement--you thought I was leaving out natural evils, like diseases and aging, that you think are necessary for evolution. But I have argued they are not necessary for evolution.

    9. Kyle, it's not only calvinists that see gods as deterministic (and you'll note that there's a difference between this and predetermiistic). There'd have been no need to envision gods at all if not as the determinant cause of everything.

  3. David,

    Quite an interesting series of posts!

    But, having nicely solved the problem of natural evil for theists (for which they should be grateful), can't we go one step further and get rid of theism altogether?

    Once a theist is committed to the computer simulation solution, then much else goes away: Plantinga's “sensus divinatis” is just part of the programming, as are intuitions about God, and so on. No reason to believe in the God of theism seems to survive the simulation hypothesis.

    If so, there is no problem of natural evil and, hence, no reason to commit to the simulation solution. We then end up nicely with a real, purposeless, godless, universe. Do you think this argument can be successful?

    1. JP

      This is a REALLY good question. Here is what I think the theist will say:

      "There is no reason that God cannot relate to simulated people, as he would physical people--he could still reveal himself (sensus divinitatis), cause religious experience, have a loving relationship, etc. In fact, simulated people may not be that different than physical people in the eyes of God--they are both unlike him, living in a "world" that he does not. I don't see the difficulty here."

      Now, we might push back and say, "Yeah, but you think that the sensus divinitatis is a faculty we have, but all our faculties are a result of the computer programers design, not God's design, so it can't get you knowledge of God--the computer programer can't design you with a 'God sense'."

      Theist's retort? "Why not? The sensus divinitatis, as a faculty, would be a function of our brain--the computer programer could create beings with such a faculty by simply copying how his brain (which presumably God designed with the sensus divinitatis) functions. Besides, God's God--he can do what he wants. He can reveal himself to simulated beings even if they don't have a sensus divinitatis."

      That last move might sacrifice a lot of the externalism/reliablism arguments for warrant/justification that Plantinga gives, and make one wonder why they bothered with the sensus divinitatis stuff in the first place, but I think it might still defend basic theistic belief--at least render in possible. I don't know--what do you think?

    2. I think an equally interesting question is--can we make sense of simulated Jesus? Could God send his one and only son to die on the cross for our sins, as a person program in a computer simulated world? Would that interfere with the free will of the programer? (Would giving the inhabitants of a computer simulation religious experiences interfere? If so, we might be able to make the your previous atheism argument work. What do you think?)

    3. It doesn't interfere with the programmer's free will - God could either have asked the guy to send Jesus in (with the option to say "no"), or the programmer could have included the I/O capability in the simulation design from the start.

      As to making sense, it makes about as much sense as the plot of Tron and how much "dying for our sins" makes at all. If Kevin Flynn could do it, God certainly can ;-)

      As to interfering with our free will - again no, it's part of the program. Actual free will for Sims might be stretching it, anyway... ;-)

    4. Yeah, I think that chbieck is right--there are ways to make it work.

    5. David (or is it Kyle?)

      Thanks for your answer.

      There is this: in order for simulated characters to be considered “persons”, they must have more than a superficial relationship with persons in the Real world. They must somehow be something more than simple simulations, isn't it? Perhaps this leaves room for the theist to sneak in a sensus divinatis or whatever it is they would try. In any case, how can we define these minimum requirements for a simulated person to be considered genuine?

      And if a computer simulation by an evil programmer does not quite make it, perhaps we can turn this backward and imagine that our universe is in fact essentially as it seems to be, except that it is the creation of an evil engineer from a world totally unlike ours - in which, maybe, there is not even a notion of personhood or consciousness (the latter perhaps problematic with computer simulations). I'm not sure this is very relevant to your argument, however.

  4. Wouldn't that just shift the problem to L0, i.e. the unsimulated real world? Even if (accepting the argument in part I) it is vastly more likely that we are living in an ancestor simulation universe than in the real universe, for the problem of natural evil it doesn't matter which level we are in, since all of them have natural evil. Or are you assuming changing natural laws (actually: simulation rules) as you go to outer levels? That would change the premise from an ancestor simulation to a monster simulation, though. (Which for a variety of reasons I think makes much more sense anyway.)

    1. Yeah, good question. I am just talking about a "computer simulation," not an ancestor simulation that replicates the laws of the original universe. Perhaps an ancestor simulation is a little more likely than a mere computer simulation (perhaps there is more motivation for doing a simulation that replicates the laws of the original physical universe, than not)--but a computer simulation is still more likely than, say, the brain-in-a-vat hypothesis. And that is what I need for the argument to work.

      (I'm not so sure ancestor simulations would be more likely, however. I think such simulations would most likely be games and people would "mod" them in many different ways.)

      But you are right--if natural evil is not absent from the one real physical world, suggesting we are in a computer simulation just backs up the problem. But since our world, as a computer simulation, need not mimic the physical world, my argument still stands.

      I perhaps should be a bit more specific: the theist is committed to believing that we live in a computer simulation, that may or may not be run in another simulation, but is ultimately rooted in a physical world without natural evil.

    2. Understood. Two quibbles with that:

      1. If I understand it correctly, Bostrom's argument does seem to rest on ancestor simulations. Not that big a deal, except that you argue from Bostrom in part I. If we are talking monster simulation (i.e. one run by beings in a higher-order world, e.g. a seven dimensional one), then "computer" is just an analogy. (I wrote a short story on that stuff a few years ago, like probably a large percentage of aspiring SF writers ;-) )

      2. I am not clear how stating that the theist must believe we are a simulation rooted in a world without natural evil is any better than saying "God works in mysterious ways". Natural evil is just a fancy name for the randomness of nature, and that is the result of the laws of physics. It is hard to see how a world without some randomness could actually work and and allow autonomous agents. If it is physically impossible, does it even make sense to say it is logically possible? (Can any physicists comment on the physical possibility?)

      [NB: I stand by the contention made in the comments to post II: a theist who tries to argue this stuff logically with you does not actually believe his deity is omnipotent, even though he might say he does. ;-) ]

      If I recall correctly, the Cathars believed something that sounds suspiciously like your simulation argument: we actually live in Hell, and it's all just a test...

    3. Bostrom deals with the possibility that our world is monumentally different than the world of the computer programer--like, if it's "aliens" (being biologically different than us). He does this in the FAQ section of his site on the simulation argument. I think the same thing he says there to defend his argument defends mine here.


      I don't think it is accurate to say that natural evil is a result of the randomness of nature--if that were true, there would not be uniformity in where hurricanes and tornadoes form, where earthquakes happen, how diseases transmit, etc. They are a result of the regularities--the laws--that govern our universe. They are not a result of randomness (although, exactly where they occur--what path they take--might be the result of a random quantum event...but I doubt it. In any cause, their existence is not.) But I don't see how a world without randomness is incompatible with autonomous agents. Free will requires alternate possibility (at least on the libertarian definition), not randomness. In fact, one of the common objections to their being free will is that the only reason that the world is not deterministic (and thus absent of free will) is because of quantum randomness--but random actions aren't free either. So randomness does nothing to save free will...at least, not that I can see.

      Others, in response to this problem, have argued that physical and logical possibility are one in the same thing--and so God hand's really are tied. He could do nothing but create a universe with the laws we have; but such arguments don't work. On all accounts, the laws of physics don't limit the laws of logic.

      (For this argument, See “Science and the Problem of Evil: Suffering as a By-Product of a Finely Tuned Cosmos” in Murphy, Nancy., Russell, Robert John., and Stoeger, William R. (eds.). (2007) Physics and Cosmology: Scientific Perspectives of the Problem of Natural Evil. Vatican Observatory Foundation(2007), pp. 148-150.)

      I will grant, however, that any theist who makes that move--he doesn't really believe God is omnipotent. If God is limited by the laws of physics, he is not all powerful.

      And God bless those Cathars.

      Thanks for the comments!

  5. Here's another curious possibility.

    Someone already argued, in another posting, that God surely does not have the power to to alter 2+2=4, to change the laws of logic, to alter the digits of pi, etc. That is, God cannot alter the Platonic conception of mathematical reality. So let us accept this as a starting point, and let us also accept Platonism. Well, one of the inhabitants of the Platonic universe is this idea of "Kolomogorov Complexity", which can be interpreted to say that there are fundamental, mathematical limits to omniscience. No matter how smart one is, even if one is infinitely smart, there are things you just cannot know. Furthermore, there's reason to believe that Kolomogorov complexity is somewhere in the range of 200 to 1K bytes. Whatever.

    So assuming 1) Platonism, and 2) God's inability to alter Platonic reality, then we conclude 3) even God cannot possibly be omniscient.

    This now leads to the curious possibility: perhaps God simply does not know of any laws of physics that do not result in a universe with evil in it. As I mentioned in a different post: air is described by fluid dynamics, which are just a set of differential equations. We know that these equations have solutions that have tornadoes in them. To not have tornadoes, we would have to live in a universe with a different set of equations for fluid dynamics.

    Now, there simply are not that many different kinds of differential equations. Yes, we know of thousands, and these fall into classes, some of which are properly called "infinite", but realistically, there are not that many. What does this mean, what am I saying? Well, to write down an equation on a piece of paper, one uses symbols. There are only a finite number of these. There are only so many different ways in which these symbols can be re-arranged so that they fit on a piece of paper, and still make sense. Its a finite number-- large, but finite.

    Now, in the real world, all of our differential equations that describe our physical universe are all remarkably small. They simply don't take up that much room on a piece of paper.

    Perhaps there is some large, very long, very complex differential equation that does not have tornadoes in it? Err, well, no. There almost surely isn't.

    The other thing to notice is that, in the real world, most all of our equations have "tornado-like" solutions: they have either chaotic regimes, or outright turbulent ones.

    So, even if you could invent and create brand-new laws of physics, in the form of differential equations, you would find it extremely difficult to find equations that did not have tornadoes and hurricanes. Given the Kolomogorov complexity of representing a differential equation, vs. that of predicting the outcome of it's solution, well, one can and should argue that perhaps an even infinitely powerful God simply cannot know, is simply unable to design a universe that does not have natural evil in it.

    So what does God do? The same thing as any poor shmuck with a hard problem does: run a simulation, and find out. Although maybe he can't figure it out, well, perhaps by running enough simulations, he can stumble upon a set of laws of physics that result in universe without natural evil in it.

    Anyway, that's *my* personal take on natural evil. Given what we know about mathematics, there may be very real and very finite limits to what even an infinitely smart God can know, and do. And natural evil may simply be on the wrong side of the complexity boundary.

    1. If I understand you right, you want to argue that "a world with no natural evil" is a mathematical impossibility, and since what is mathematically impossible is logically impossible, and God can't do what is logically impossible, God can't create a world without natural evil, even though he is infinitely powerful. Interesting!

      As I mention in my conversation chbieck, there have been those that have tried to "tie God's hands" in similar ways, and say a world without natural evil is logically impossible (I gave a citation above). But they usually do it by trying to equate physical possibility with logical possibility. That fails. But you are doing it by trying to argue that it is mathematically impossible.

      I think this is very interesting; a full blown argument would be neat to see. You seem to think it requires Platonism about math, and that might be a problem. But I'm not even sure it requires this. If there are certain things it is impossible to know with even infinite knowledge, and that's one of them...maybe this works. I don't know--I'd have to see the full argument. (My original paper is published in Philo--maybe you could get a publication by responding to it.)

      One problem may be...well, you say, "Perhaps there is some large, very long, very complex differential equation that does not have tornadoes in it? Err, well, no. There almost surely isn't." Wouldn't we have to prove there isn't for this to work. If we can't prove there isn't, should we say it's logically possible there is...and thus my conclusion still follows?

      There is also this problem: This might get you out of tornadoes, but does it get you out of diseases, earthquakes, meteors, etc? Can you make all of these things "mathematical necessities"?

      At any rate, this all would I think make heaven an impossibility--there will be natural disasters in heaven if they are mathematical necessities. And I doubt many theists are going to accept that consequence.

      Thanks for the comments!

    2. Kyle, I think you've already hit on the problem with this argument. If we require that this knowledge be inaccessible even with infinite knowledge, then this argument fails. The actual result that I think Linas is referring to is the result that Kolmogorov complexity -- the function mapping strings to lengths of minimal descriptions of those strings -- is uncomputable. But there are caveats; the proof of uncomputability presented on the Wikipedia page applies only to finite programs.

      A finite program capable of computing Kolmogorov complexities would create a paradox, because it would enable you to create another finite program p that, given an arbitrary string of minimum complexity k, would find a corresponding algorithm and run it, generating the string. But if k > len(p), this produces a paradox, because now p can generate our string, and therefore the complexity of the string is not k but len(p), violating our initial assumption.

      This is somewhat analogous to the following paradox: "N is the smallest integer that cannot be described in fewer than twenty words." If n exists, then it can be described in fewer than twenty words, and so does not satisfy the predicate.

      But if our Kolgomorov-complexity-computing program can be infinitely long, then it can compute the complexity of any string that can be generated by a finite program, because its size is greater than all such programs, and the above paradox does not hold. (However, it cannot, in turn, calculate the complexity of strings that can only be generated by programs longer than itself.)

      The upshot is that an infinite program could compute the K-complexities of all finite programs.

      ...I think.

    3. Whoops, that last line should read "an infinite program could compute the K-complexities of all strings that can be generated by finite programs."

  6. Kyle, I also have a very simple question. Aren't you basically talking about the demiurge here (at least according to some traditions)?

  7. One thing not pointed out here, and there seems to be a lack of theological understanding is that death isn't "evil" at least in Christian theology. It is objectionable but it is a penalty so God allowing justified death isn't evil but actual murder, unjustified killing is. I don't know many Christians who think that the weather is direct punishment by God by the end result that people die is the penalty. Don't shoot he messenger I'm simply pointing out theology.
    Now you may say that a child dieing can't possibly be a penalty that is justifiable but you can only really say that God allows it, not directly causes it personally only that things were set into motion by him and by that you are casting your own moral judgement that you believe God to be committing evil acts.


Note: Only a member of this blog may post a comment.