Walter Sinnott-Armstrong v. William Lane Craig

This debate took place on April 1, 2000 at Wooddale Church in Eden Prairie, Minnesota. Atheist philosopher Walter Sinnott-Armstrong, professor at Duke University and author of numerous books on ethics and morality,1 argues in the affirmative on the question, “Does evil and suffering disprove God?” William Lane Craig, a well-known Christian philosopher and apologist who also has a lengthy list of books in his name,2 argues the negative. Prior to this, Bill and Walter had debated in 1999 at Dartmouth, and following this second round, the speakers worked some of their arguments from both debates into a book entitled, God?: A Debate between a Christian and an Atheist, which has received generally stellar reviews on Amazon.

The structure of the debate is a bit unusual, with each speaker giving a 15 minute opening speech, followed by 6 minutes of speaker discussion, 10 minutes of moderated discussion, an 18 minute Q&A session with the audience, and finally 5 minute closing speeches.

SA’s Opening Speech

Sinnott-Armstrong (henceforth abbreviated as “SA”) starts things off on a good note,  thanking the church for inviting him and putting on the event, before making a joke about feeling like a Christian in the Roman coliseum. From the outset, he explains that his desire is for an open and honest discussion where everyone benefits and learns something, including himself. What he doesn’t want is any kind of “high school debate” that prizes winning above all else, nor is it his goal to deprive anyone of their faith. Some people need religion to make sense of their lives, he says, but we can have the positive aspects of religion without all the complicated theological doctrines.

Before we get on to the argument, let me just note how impressive this introduction of sorts is. I have listened to an unholy amount of theist-atheist debates, and there is nothing else quite like it, with the possible exceptions being in Austin Dacey’s two debates with Craig. Walter is respectful, funny, clear, concise, and comes across as being genuinely concerned with having productive dialogue. Unfortunately, many atheists debate like the late Christopher Hitchens, showing vicious disdain for believers and their views from the very beginning, having a fiery and almost angry attitude, and displaying no real intent of civility, often while ironically missing or ignoring their opponent’s arguments. To me, such an approach is universally unproductive, serving only to circle the wagons on both sides. SA, however, keenly disarms the audience with humor and politeness, appeals to the search for truth that we all share, and puts his intentions in no uncertain terms. Whether or not these were “tactics” to earn favor is not the point, though. What matters is that they fostered the thoughtful, civil, and meaningful discussion that resulted.

Walter begins by defining god as all-powerful, all-good, and all-knowing. It is this conception of god, he claims, that is incompatible with the existence of evil. But rather than attempt to offer conclusive proof, which he doesn’t see as possible for either side in the case of a being defined with such attributes, SA states that he will try to give adequate evidence against the existence of this god.

The type of argument that Dr. Sinnott-Armstrong lays out is an evidential argument from evil. This means that rather than being deductive, or arguing from the premises to a definite conclusion, it inductively reasons from the premises to a probabilistic conclusion. To help illustrate his case, SA uses the example of a sick child who suffers and dies at a tragically young age. Of course, this is not the only example, and others like earthquakes, famines, and floods are raised throughout the debate, but Walter emphasizes that for his argument to work, there just needs to be one unnecessary evil. He formalizes the argument as follows:

1. If there were an all-powerful and all-good God, then there would not be any evil in the world unless that evil is logically necessary for some adequately compensating good.
2. There is evil in the world.
3. Some of that evil is not logically necessary for some adequately compensating good.
4. Therefore, there can’t be a God who is all-powerful and all-good.

SA defines evil as anything that all rational people avoid for themselves, unless they have an adequate reason not to avoid it. He uses the example of how we go to the dentist to have our teeth drilled on; although dental work may be painful, we have an adequate reason not to avoid it, because tooth decay is even more painful in the long run. Evil is justified, he says, only when there is no other way to avoid it, or no better way to avoid it. God is more skilled than we are, though, which means that he could conceivably prevent tooth decay with no pain, or save an infant from dying without causing it to suffer. If there is not a logically necessary reason for some evil existing, then it’s unlikely that an all-powerful and all-good god exists.

Next, Walter addresses some common theistic objections to the evidential argument from evil. The objections are noted by an O, whereas SA’s responses are noted by an R.

O: God uses evil as a punishment for sin.
R: Evil is not distributed as if it is in accordance with sin. There is no reason to think someone struck by lightning is more sinful than someone who isn’t struck by lightning.

O: Suffering in this life is repaid with eternity in heaven.
R: God could send a suffering child straight to heaven and avoid the suffering altogether.

O: Evil is the consequence of free will.
R: Free will cannot account for natural evils, i.e. disease, earthquakes, lightning, etc.

O: Evil builds character. Our suffering can make us better persons, and suffering can generally influence and shape all of us.
R: God could form character by other means, such as through dreams or visions. It also is unjustified to let one person suffer to improve the character of another.

O: God uses evil to bring people closer to him.
R: God could bring people to him in lots of other ways. The idea of god causing a baby to suffer and die horribly to bring someone else closer to him also seems reprehensible.

O: God has a reason for allowing evil, but we’re unable to see it.
R: We wouldn’t normally accept this as excusing anyone else, so why accept it just for god? We have to go by the standard that we have.

O: The problem of evil is overridden by the arguments for god’s existence. If god exists, we can assume he has reasons for allowing evil.
R: The arguments for god are unconvincing, but even if they do work, most cannot establish that god is all-good from merely arguing that a creator exists.

SA concludes his opening speech by outlining his argument’s implications. In the face of unnecessary evils, Christians must either reject that god is all-good or reject that he is all-powerful. Yet this is the traditional conception of god defended by philosophers and theologians for centuries, and it is the only one forming the basis of their debate, Walter explains.

It’s hard to believe SA could fit so many great points into a 15 minute speech, but it shows why he’s a worthy opponent for Dr. Craig, who is a seasoned debater that’s well known for taking the shotgun approach in his own opening statements. Sinnott-Armstrong anticipates many of the objections Craig raises next, and it doesn’t seem like Craig ever really recovers.

Craig’s Opening Speech

Bill begins by summarizing SA’s argument, drawing attention to its focus on gratuitous, or unnecessary, evil. Because we are finite beings limited in our perception of time and space, we should not expect to see the reasons why god permits apparently pointless evils, he contends. It may be that the justification of some instance of evil or suffering will only appear in another place or another time, well outside the scope of our individual experience.

Craig lays down three Christian doctrines that he believes undermine the premise that any evil is gratuitous:

1. The purpose of life is not happiness, but to know God. There may be gratuitous evils if happiness is our focus, but these would not be gratuitous for producing the knowledge of God.
2. Mankind is in rebellion against god and his purpose, and so human evils are the consequence of our free will.
3. The promise of heaven outweighs any “slight momentary affliction”, as the apostle Paul called it.

One will notice that all of these objections are addressed by SA, but Craig dishonestly says in his speech that Walter “hasn’t even tried to [refute them]”. In fact, the 2nd doctrine might as well be entirely irrelevant to the debate, since SA concentrates pretty much exclusively on natural evils, and, as he said, free will cannot account for natural evils. The 3rd doctrine SA anticipated with the response that god could just send an infant straight to heaven rather than waiting until it’s suffered enough evil, then sending it there. If we recall the clarification in Walter’s opening speech, that evil is justified only when there is no way, or no better way, to avoid it, we can see the inadequacy of Craig’s 3rd doctrine against the suggested alternative. Of course, one might try and argue that it’s not known if god could really avoid letting an infant suffer a while before sending it to heaven, but this falls back on sheer possibility and Craig’s first argument about the limits of human understanding, which is a different reason than the promise of heaven.

Commenting on the first doctrine, Bill says that SA has to show that god could reduce the amount of evil in the world while still retaining the same amount of knowledge of god. First of all, as we’ve seen, SA said that there are other ways god brings people closer to him, and the same response to the character-building objection can apply here. God could bring us to know him deeper through dreams and visions, as the Bible reports him doing for people like Paul, Peter, Solomon, and Mary’s husband Joseph. God could bring us closer through scripture, as the Bible also reports him doing. It seems to me that the sole reply the apologist has to these alternatives is that god is constrained in some way to not be able to use these methods for all persons, such as a particular amount of evil being necessary for free will or to produce the ideal amount of knowledge of god. But again, these reasons are different from the first doctrine.

Secondly, the bigger issue is that these arguments can’t effectively grapple with the evidential argument from evil, because they all boil down to speculative possibilities. Walter’s argument is an inductive argument, though, which deals with probability. Craig has to do more than say what god could do, he has to make a case for why we should think god has done any of what he suggests. This looks to have been the intent behind the three doctrines he elaborates, but none of these are persuasively defended by Craig – not just because he gives no probabilistic arguments, but also because he fails to rebut SA’s responses. Now, you might complain that it’s unreasonable to expect that Craig would have answered those responses in his prepared opening speech, but I can’t be so sympathetic considering that Craig has expected as much from opponents at other times, and he’s been capable of forming such impromptu replies in other debates.

Finally, Dr. Craig attempts to turn SA’s argument from evil around on him, claiming that if his first premise is taken along with the premise that “god exists”, then the conclusion will follow that gratuitous evil does not exist. In support of his second premise, Craig offers the cosmological argument, the fine-tuning argument, and the moral argument (he additionally suggests that the argument from evil then becomes a theistic argument – which is a point I will address in a moment). Once more, though, Bill makes an argument already anticipated by Walter, who explained that even if the arguments for a creator are assumed persuasive, they cannot establish that this creator is all-good. At best, they could perhaps show that he is very good, but there is nothing implicit in the cosmological argument or the fine-tuning argument that argues for god being all-good. Thus, two of Craig’s arguments are out.

What about the moral argument? On the surface, it may appear that the moral argument can establish a perfectly good god, but we run into problems the moment we scrutinize the first premise. “If God does not exist, objective moral values do not exist.” For this premise to be sensible, we must have some reason to think that god can serve as a meaningful foundation of objective moral values. Recently on my blog I asked Does William Lane Craig Actually Believe in Evil? A few statements Bill made in a debate with Michael Tooley raise the question of just what significance and meaning evil can really have if it’s ‘grounded’ in god’s nature. If god only permits what is right, and history is full of countless horrors god has apparently permitted, then what could even be wrong for god to permit? As a Divine Command Theorist, Craig believes that god’s commands flow from his nature, and yet if his nature is consistent with allowing even the worst rapes, murders, genocides, tortures, and child abuses, then what reason would we have for thinking god would never command anything most of us would intuitively identify as evil?

We have cause to be suspicious about whether even the moral argument can establish a perfectly good god. Craig proposes a reworking of the moral argument which he calls the argument from evil, where the second premise is simply replaced by “evil exists”. I don’t think this is actually defensible as a separate argument, because the implication from the first premise is still that evil is an objective moral value, and thus to say that evil exists is only to state one “half” of the premise that objective moral values exist. Furthermore, since I attacked the first premise of the moral argument, which is identical to that of the “argument from evil”,  deconstructing the moral argument will deconstruct Craig’s argument from evil, too.

Speaker Discussion

SA corrects Craig’s remark about his not addressing the theistic objections, but Craig still maintains that SA did not address the three doctrines he mentioned (which is not really true, as already noted).

Bill brings up the possibility of the theist having a combination of “solutions” that could cover all the bases and account for apparently gratuitous evils. Walter responds by reiterating that god allegedly has other ways of bringing people to knowledge of him. There are (according to Craig) proofs of god’s existence, too, so why should suffering be necessary? Bill tries to spin things back to Walter by asking who bears the burden of proof, but Walter at last calls out Bill for his constant retreat to the possible.

You can always defend a doctrine by saying, ‘might might might might might’, but I see no reason in this case to think that someone 50 years from now will come to knowledge of god by this child’s suffering.

If only SA would have also pointed out the insufficiency of countering a probabilistic argument with a claim of mere possibility, his point would have been even stronger. Walter not only made an inductive argument, but he rebutted numerous theistic objections to his argument, and Craig’s reaction was to simply speculate on what reasons god might have for allowing apparently unnecessary evils – failing to deal with Walter’s rebuttals to several of them – and then kick the burden of proof back to Walter as if he should suddenly be expected to argue the logical impossibility of Craig’s purely speculative answers.

Moderated Discussion

The moderator begins the next segment of the debate by following up on SA’s comments, asking if the theist should be expected to say “exactly” what purpose god might have for evil. SA rightly recognizes this as a jump to an extreme and returns the conversation to what justifies us in our beliefs. “We have to work from the evidence that we have,” he states. SA uses the example of witnessing a neighbor who lets their child suffer horrendously. While we could speculate about the reasons the neighbor might have for letting the child suffer, we wouldn’t likely accept those speculations as justifying the allowance of suffering.

Craig proposes that the reason we don’t accept the neighbor’s treatment of the child is because the neighbor is not all-knowing. SA observes that the neighbor is still at least in a position to know more than he knows, and so if he suspects that the neighbor might have a reason for allowing the child to suffer, he will go to the neighbor and ask for an explanation. However, he points out, all Craig has provided is speculation. If god has good reasons for allowing suffering, why doesn’t he tell us?

Bill uses a favorite line of his that’s made it into several debates I’ve heard. “That turns the universe into a sort of haunted house,” he claims. I’ve yet to hear Craig make an actual point out of this bit of strange rhetoric, but it certainly isn’t an argument as it stands. Craig asks if it’s reasonable to demand such a “hotline to god”, to which Walter brilliantly responds that he would be happy to use the hotline even once. “If god is omnipotent and omniscient and omnibenevolent,” Walter continues, “then my ignorance would be an evil that he should try to minimize, and I would like to see it happen.”

The moderator next turns the conversation to justifying moral values. Craig commits the modo hoc fallacy, claiming that on atheism humans are “just” animals, and therefore we have no reason not to act like other animals. There are many obvious objections to this naive and simplistic reductio ad absurdum of the naturalistic worldview. It would not follow that because we are a type of animal, that we are no different from other animals, nor would it follow that we should behave like other animals. We are also not “just” animals, like a cat is not “just” an animal, but can also be a pet, offspring, a parent, a predator, and so forth. Human beings are unique compositions of matter with consciousness and self-awareness that allows us to reflect, reason, and do many of the things that make us moral agents. Sadly, SA is another atheist opponent in a long line of atheist opponents who didn’t identify Craig’s commission of this fallacy during the debate.

Before turning to Q&A, SA and Craig briefly touch on free will and determinism and how they each would give advice to someone on her deathbed.

Q&A Session

The Q&A time doesn’t offer much back and forth between the speakers, unfortunately, and so there isn’t much in it that I wish to comment on. Bill and Walter argue over the Ethiopian famine and god’s responsibility (or lack thereof) in it. Walter claims that god could allow humans to make their choices, but render them ineffectual. Craig responds that this would make moral decisions trivial, like turning bullets into rubber. However, his comment presumes that every evil action would be rendered ineffectual. Perhaps god would only occasionally intervene to show us he cares and to keep us on our toes. In fact, the Bible is full of stories of god intervening to save people from evil, like instructing Noah to build the Ark to avoid the flood, or guiding Moses to lead the Israelites out of slavery in Egypt. Did god’s intervention against those evils trivialize the moral decisions of Noah, Moses, or anyone else involved? Once again, Craig over-simplifies things and too hastily dismisses them.

Bill makes one of the most astounding statements of the debate when asked about why god doesn’t reveal his reasons for permitting evil to us. Invoking the “haunted house” imagery again, he proceeds to argue that it could well be plausible that god can’t reveal those reasons to us because they could foul up his providential plan. Imagine that! An all-powerful and all-knowing being that could somehow have its perfect, divine plan spoiled by its imperfect, finite creations… Craig entertains such speculation purely on grounds of preserving free will and vainly attempting to bolster his retreats to the possible. But this strikes me as a very bizarre position for a Christian theist to adopt. What’s next – are we going to find out we could actually be capable of somehow sending god to burn in hell?

Lastly, Craig plays a blatant bait-and-switch with the burden of proof, arguing that he doesn’t have to prove god has any morally sufficient reason for allowing evil, but the atheist claiming that god “probably” doesn’t have a morally sufficient reason has to prove that god “does not or cannot” have one. Suddenly we shift from probability to possibility again. Granted, SA didn’t frame his initial argument from evil in probabilistic terminology, but being that it’s an evidential form, and since he draws the conversation into probability and rational justification numerous times, I can’t let off Craig so easily. Considering SA’s argument and his stated aim, it should suffice to show that there is probably no morally sufficient reason for god allowing gratuitous evil. Bill has neglected to provide probable reasons, and in the absence of those reasons, we truly have no reason to believe.

Closing Statements

In their closing speeches, Dr. Sinnott-Armstrong and Dr. Craig basically reiterate and summarize the points and arguments they’ve previously made. Walter stresses that we should only believe what we have good reason to believe, while Bill persists in speculating in defense of his god. Craig ends his final speech by reading a letter from a parent who lost a daughter. Tragic as such a loss must be, the letter offers nothing but further speculation and emotional appeals.

Concluding Thoughts

In my opinion, this is one of just a handful of debates that pits a worthy atheist opponent against William Lane Craig. Walter Sinnott-Armstrong was prepared, organized, well spoken, familiar with the arguments, and thoroughly engaged Craig throughout the entire duration of the debate. Add to this a sense of humor, a courteous attitude, and a respect for the topic, and it’s not difficult to see why even Christian listeners have been largely impressed by SA’s performance.3 I also enjoyed hearing some divergence from Craig’s usual line-up of theistic arguments, and it did seem like Bill engaged on a deeper level with Walter than he has in other debates involving the problem of evil.

Even so, this debate was not without its flaws. I feel SA could’ve responded better to the moral argument and missed a major opportunity to call out the modo hoc fallacy, which Craig has committed in every single debate I’ve heard him in. Likewise, Craig seems to have mostly ignored the distinction between probability and possibility in this debate, and some of his responses don’t even constitute arguments.

These things aside, SA and Craig put on a very entertaining, thoughtful, and respectful discussion on subjects that are important and challenging for all of us. This is certainly among my top five theist-atheist debates.



1. Walter Sinnott-Armstrong’s Home Page, Bio, Duke University. Retrieved Aug. 19, 2013.
2. William Lane Craig, Biographical Sketch, Retrieved Aug. 19, 2013.
3. William Lane Craig vs Walter Sinnott-Armstrong: evil, suffering, and God’s existence, Wintery Knight (2013). Retrieved Aug. 19, 2013.