The Harris-Sullivan Debate: II

As the debate continues, Sam Harris responds to Andrew Sullivan’s contention that many of the fundamentalists he knows are compassionate, caring people who do many selfless good deeds for others:

For instance, you claim that many fundamentalists are tolerant of dissent and capable of friendship with you despite their dogmatic views about sex. You also remind me that many devoutly religious people do good things on the basis of their religious beliefs. I do not doubt either of these propositions. You could catalogue such facts until the end of time, and they would not begin to suggest that God actually exists, or that the Bible is his Word, or that his Son came to earth in the person of Jesus to redeem our sins.

I believe a well-articulated and properly understood atheism should consist of two claims: first, that religious beliefs are not true; second, that on balance they have done more harm than good. The first claim explains why we are atheists, the second explains why we speak out about it. Though I’ve never advocated belief in convenient falsehoods just because they are comforting, I’ll be the first to say that if religion had only ever been a force for good in this world, I probably wouldn’t have gone to the trouble of creating this website.

But we should be careful not to confuse these claims, or let theists confuse them for us. More importantly, we must take care not to suggest that religion is false because it is harmful. These are separate claims and should be defended separately.

Harris goes on to offer an incisive remark on Sullivan’s defense of religious morality:

As you may know, I’ve argued that religion gets people to do good things for bad reasons, when good reasons are actually available; I have also argued that it rather often gets people to do very bad things that they would not otherwise do.

This is exactly right, and this is a point we should be making more often. The problem with religious morality, even when it causes people to treat others well, is that religious morality is grounded in an idea – obeying the supposed will of God – that is orthogonal to human concerns. We should not be caring for the poor and the sick because it is what God wants; we should be caring for them because it is the right thing to do, and because these others are human beings who need our compassion. A morality not grounded in human concerns can still produce good results, when what the believer is taught to be God’s will lines up with what their fellow humans need; but it can also produce dreadful results when the alleged will of God does not align with human desires and needs, as my letter of advice to a Christian illustrates. And, as this story shows, these two seemingly contradictory behaviors can paradoxically exist within a single person. An atheist morality of compassion would produce good results consistently, not just by happenstance.

Andrew Sullivan says:

My response rests on an understanding of truth that is not exhausted by empiricism or materialism. I do not believe, in short, that all truth rests on scientific premises and can be ‘proven’ by empirical or scientific methods. I believe science is one, important, valuable and respectable mode of thinking about the whole. But there are truth questions it has not answered and cannot answer.

Sadly, he does not elaborate on what non-empirical – in other words, non-evidence-based – method he has in mind. (He does claim, incorrectly, that historical investigation is not an empirical method of truth-seeking because it’s not based on repeatable experiments, a common apologist mistake.) Most religious believers don’t, in my experience, which makes claims like these an exercise in vacuous special pleading. The usual assertion is that there are some beliefs that should be exempted from the tiresome necessity of evidence because, well, just because.

Sullivan goes on to address one of Harris’ arguments:

…the content of various, competing revelations renders them dangerous. They are dangerous because they logically contradict each other. And since their claims are the most profound that we can imagine, human beings will often be compelled to fight for them. For if these profound matters are not worth fighting for, what is?

Sullivan has grasped the central problem here and summed it up with aplomb. He goes on to admit that this is a “central problem” for religion and “will always be so”. However, his proposed solution is entirely inadequate:

We have Scripture; we have reason; we have religious authority; we have our own spiritual experiences of the divine. But there is still something we will never grasp, something we can never know – because God is beyond our human categories. And if God is beyond our categories, then God cannot be captured for certain. We cannot know with the kind of surety that allows us to proclaim truth with a capital T…

The answer is: humility and doubt. I may believe these things, but I am aware that others may not; and I respect their own existential decision to believe something else. I respect their decision because I respect my own, and realize it is indescribable to those who have not directly experienced it.

This solution is glib and superficial. As Sullivan is certainly aware, because he says he has dealt with them many times, there are a huge number of religious believers who do not share his commendable humility. There are a huge number of theists who feel that they do have surety, that they do know God’s will (because, they hasten to add, God has told them what it is), and even worse, that God wants them to be the agents of his will, imposing it on others who do not agree. What does Sullivan propose when his “non-empirical” methods of truth-seeking lead, as they inevitably will, to people adopting this theocratic viewpoint? What do we do then?

Andrew Sullivan, after all, is not the problem. He does not want a theocracy. But there are millions of believers who do, fervently so, and are fighting to bring that outcome about. Sam Harris is right to point that out, and to argue that religion is a very dangerous thing because of it. Sullivan’s response that he isn’t one of those bad guys does not undermine the validity of this point, and though he gravely frowns over the matter and pronounces it a serious problem, he proposes no feasible way to deal with it.

There is only one workable solution to this problem, in fact, and that is that we need a method of truth-seeking that relies on evidence which we can all examine for ourselves. That is the only way we will ever be able to reach consensus. Relying on subjective claims of internal revelation and other “non-empirical” methods of belief-formation mean that when those methods lead to violent and tyrannical actions grounded in the actor’s belief in God, there is nothing for religious moderates to do except to protest that they believe something different – which is precisely what Sullivan does, and which is Harris’ point. What we need to do is to teach the supremacy of evidence and reason over faith, and teach people to think skeptically, critically and rationally. That, in the long run, is the only possible way to end religious bloodshed and jihad.

The fourth section contains, I think, one of the most revealing exchanges in the entire debate. Sam Harris ends his letter with an extremely important question, and Sullivan’s response is startling in its candor:

Let me close by asking you a simple question: What would constitute “proof” for you that your current beliefs about God are mistaken?

I have no ability to stop believing. Crises in my life – death of loved ones, diagnosis with a fatal illness, emotional loss – have never shaken this faith. In fact, they have all strengthened it. I know of no “proof” that could dissuade me of this, since no “proof” ever persuaded me of it.

…You will ask: how do I know this was Jesus? Could it not be that it was a force beyond one, specific Jewish rabbi who lived two millennia ago and was executed by the Roman authorities? Yes, and no. I have lived with the voice of Jesus read to me, read by me, and spoken all around me my entire life – and I heard it that day. If I had been born before Jesus’ birth, would I have realized this? Of course not. If I had been born in Thailand and raised a Buddhist, would I have interpreted this experience as a function of my Buddhist faith rather than Jesus? If I were a pilgrim right now in Iraq, would I attribute this epiphany to Allah? An honest answer has to be: almost certainly.

So, according to Sullivan’s own words, he is a believer because of a subjective, emotional experience he had, which he cannot explain and which did not come upon him in response to any evidence. He admits that he interpreted it as a confirmation of Christianity, rather than some other religion, only because that was the religion he was brought up with. And yet, this faith built on nothing is so persuasive in his eyes that he would refuse to reconsider it in response to any evidence.

I thank him for putting it so plainly, because I think something very similar is at the heart of most religious belief. A person experiences something that has a powerful emotional impact on them, they interpret it in the context of whichever religion is dominant in their society, and they decide through faith that they could not possibly be mistaken about the cause of this experience or what it signifies. Those three steps – emotion, tradition, faith – probably explain nine-tenths or more of all religious belief, notwithstanding the unconvincing after-the-fact apologetics offered by a minority of the faithful.

It should be obvious that this is an utterly insufficient basis for believing anything, much less anything of such extreme importance. Sullivan almost seems embarrassed by it himself, as he phrases his position not in terms of choice but compulsion, claiming that he is unable to believe differently. Even if that is the case, and I think he gives himself too little credit, it does not follow that he should want to continue believing in this way. After all, we all know the experience of having higher-order desires that contradict our lower-order desires – like the smoker who wants a cigarette, but also wants not to want cigarettes. Even if Sullivan cannot stop believing, he could still decry (higher-order) the emotional addiction to religion that has left him in that state. That he does not do so suggests that his problem is one of will, not ability.

Coming up: Thoughts on part 5 and 6 of the Harris-Sullivan debate.

Weekend Coffee: March 28
New on the Guardian: Beyond Debating God’s Existence
A Christian vs. an Atheist: On God and Government, Part 11
Why Atheism Is a Force for Good
About Adam Lee

Adam Lee is an atheist writer and speaker living in New York City. His new novel, City of Light, is available in paperback and e-book. Read his full bio, or follow him on Twitter.

  • James Bradbury

    He does claim, incorrectly, that historical investigation is not an empirical method of truth-seeking because it’s not based on repeatable experiments, a common apologist mistake.

    I’ve come across this kind of suggestion from apologists a number of times, in relation to for example, the formation of the Earth, which we obviously can’t repeat. I feel it is somehow in a different category to saying it was all made in seven days, because we have more than one source of evidence, but beyond that I’m at a loss to explain.

    Have you elaborated this point elsewhere that I’m not aware of or could someone offer a suggestion, perhaps an example related to this?

  • Ebonmuse

    Richard Carrier’s book Sense and Goodness Without God has a whole section on different levels of certainty and what methods of determining truth we should trust more than others. He says that the critical-historical method is less certain than empirical methods whose results we can observe directly, and cannot overrule such methods – so, for example, one could never use a historical account to establish the existence of a miracle when the occurrence of such an event is contradicted by repeated experiments in the present. I can certainly agree with this.

    However, just because the historical method does not give total scientific certainty does not mean it gives us nothing to go on at all. We can compare the testimonies of different writers and evaluate their objectivity and trustworthiness, and we can study the evidence left behind by historical events through methods like archaeology, to determine which historical claims are more or less plausible than others. And while we can’t absolutely prove any historical claim to be true, there are some claims we can definitely rule out.

  • BlackSun

    Every claim that certainty is impossible blithely ignores levels of probability, and thus presents a false dichotomy: either something is knowable for certain, or not at all. This is the typical slam on atheists, that they walk around being certain of everything. To which I respond, “not certain, just relying on better evidence than you are.”

    Perhaps it’s a function of the generally poor math skills of U.S. citizens, or the stunning fact that only 28% of the U.S. population understands science well enough to interpret news reports, that allows these misconceptions to flourish. It’s also the desire for neat answers and black vs. white opinions. Living in a scary world of shades of gray and nuance is sadly beyond the capacity of many.

    Sullivan’s quote:

    My response rests on an understanding of truth that is not exhausted by empiricism or materialism. I do not believe, in short, that all truth rests on scientific premises and can be ‘proven’ by empirical or scientific methods. I believe science is one, important, valuable and respectable mode of thinking about the whole. But there are truth questions it has not answered and cannot answer.

    Skeptico talks about this as the “appeal to other ways of knowing.” It is, of course a clear logical fallacy. It’s the scientific equivalent of the used-car-salesman trick of “I have to check with my boss.” Well you can never talk to the boss, and you can never examine the “other ways of knowing,” so this dodge becomes a neat way for believers to prevent any intellectual examination of their favorite brand of ‘woo’ that would force them to an uncomfortable re-evaluation.

    It’s getting tiring dealing with the same objections like this over and over again. Someday, as consciousness becomes more understood, and we begin to analyze direct brain function, we will expose these responses for what they are. These are not simply bad arguments: Since they are so widespread, it’s clear there must be some evolutionary basis for specific cognitive defense mechanisms which allow people to live with their stratospheric levels of denial.

  • Infidel753

    I think pretty much everything you say here is spot on, but I’d question the definition of atheism in your second paragraph. Anybody who doesn’t believe in the existence of God is an atheist, even if he believes religion is harmless or beneficial (it is theoretically possible, though rare, for a person to believe that it is good for people to believe things that are not true). Certainly very few people take the position that God does not exist but that it is still good for people to believe in God; nevertheless, this is still an atheistic position.

  • Jarrod

    Not to direct the conversation away from the debate itself – but I’m really curious about this willingness to abandon “other ways of knowing.” Forget God and religions. Is everything really knowable through scientific, “objective” evidence-driven, reductionist procedure? Why doesn’t experience count as a type of knowledge?

    I could have every possible piece of evidence (or what would be considered evidence) that a hot stove will burn my hand, but it seems to me that the moment I actually do burn my hand, I gain new information about hot stoves and hand burning. Or, if I read a poem, my knowledge about the poem would change once I actually read it, and my knowledge about the poem would change again if I found it meaningful in some way.

    I see how brain processes could theoretically account for experiential knowledge, especially as in the case of hand-burning on a stove. But, concerning meaning – why things and people “mean” something to me – science seems a longs way off from explaining that, and I wonder if it ever will be able to. If I have knowledge that an object, event, or person means something to me, any evidence I provide supporting that knowledge will be incomplete and will miss the point of my knowledge. In this case, my knowledge is completely about my internal experience (in relation to an external something).

    Sorry to be dragging on – one last attempt at lucidity. If I am “touched” by piece of music, you may explain my reaction to the music in terms of my personal history and brain processes: my brain did this-thing-here, because of this-occurrence-there, in light of all my own affinities and physical states (ignore the vagueness of “affinities”), and, thus, I felt “touched” by the music. I want to claim, however, that such terms seem insufficient for explaining my reaction to the music; the account explains how I came to be affected by the music, but the account doesn’t explain my experience of it all. And it seems that I, the person touched by the music, am the only one in a position to know what it meant for me to be touched by the piece of music.

    Perhaps the question is a philosophical one. I’ve never thought all true knowledge can be found by something like a scientific method, so I’m always surprised when that view comes up. Am I mis-characterizing positions, or deluded by romantic ideology? There’s implications for religions, here, sure, but I’d rather not leave the natural world quite yet.

  • James Bradbury


    To fully explain your reaction to the music, surely we’d have to know the exact state of your brain at the time you hear it. I think this possible in theory. Perhaps a person who knows you well would be able to guess what effect certain music might have on you. Such estimations feel inadequate as our ability to catalogue the state of the brain is very limited.

    As regards the hot stove, you have a theory and you test it and it matches the evidence you gain through that test. It may not seem very scientific, but you would be gaining knowledge through experience.

  • sharon wortman farnham

    I have a srong personal belief in God I believe good as well as evil exists in all of us . It is the evil that is telling you GOD and religion are bad for you people today don’t talk enough about the devil .They are stupid to believe he doesn’t even exist he is real he chases after every soul on Earth and tries to get us to do bad things . One cannot believe in GOD without believing in the Devil for some reason he doesn’t get time on T. V. any more probaly to busy out their doing bad to have time to show up

  • James Bradbury


    Please tell us more about the devil. For example, why is it that God saw fit for the devil to exist?

  • Jarrod


    So, as our ability to catalog brain-states improves, we should be able to explain reactions of meaning to objects. I suppose we should have faith that this will happen because neuroscience has already begun to explain other once-puzzling aspects of our cognitive lives. But I wonder where science will one day have to say “It’s just that way.”

    Our brain interprets certain light waves as color. Is it the same for experiences of meaning in general, that our brain just happens to interpret certain things as having meaning to us? Science could eventually step in and more thoroughly explain how and why our brain interprets certain things as having meaning. But what is meaning? Certain wavelengths equal, for us, the color red – there’s no more to it than that. Yet meaning has deeper implications; meaning, for us, touches the question of what is it to be a human person. Colors seem trivial, a mere physical interaction with the world. Things that have meaning seem less trivial, a bit more than a mere physical interaction with the world. But, I guess you’re right: a detailed-enough understanding of the brain could explain all cases of meaning. In that case, we’d just have to say “It’s just that way” about meaning, that some physical processes carry non-physical appearances (i.e. my brain only makes me think I’m having a special, meaningful reaction to music). So maybe it is a scientific question. But science’s answer seems disappointingly far off.

    Going back to the hot stove – I think you’re right that it is a test that involves evidence. But now you’ve raised the point I think I was initially trying to get at: the evidence and testing is available only to me. I cannot show anyone my theory of the stove’s hotness, nor can I show anyone the evidence (other than bodily marks) of my test of the stove’s hotness.

    You’ve said that my experience was a form of evidential testing. But how should others know that? Bringing it back now to the original discussion – how do we know that religious experiences aren’t genuine religious experiences? It seems that my experience of the hot stove would be accepted, but if I had a religious experience, it would not be. What’s the difference? And I’m trying to focus on experience itself. What “proof” is there, in fact, to directly support any experience? It seems that proof (as bounced around among people) only affects interpretations of experience. I think it’s an important nuance.

  • John P

    Having followed the Harris/Sullivan exchange, it seems to me that Sullivan is about this close to figuring out he’s an atheist. He certainly has the intellect for it, and sufficient doubt, so it’s just a matter of time, and thought, before he’s able to shake the hold his upbringing admittedly has on his ability to self-analyze. A good start would be to wait a few months, then go back and re-read what he wrote, what Harris’s responses were, and, for good measure, read the analysis here.

  • Ebonmuse

    In that case, we’d just have to say “It’s just that way” about meaning, that some physical processes carry non-physical appearances (i.e. my brain only makes me think I’m having a special, meaningful reaction to music).

    No, you are having a special and meaningful reaction to the music. That is obviously true and will be true regardless of what we discover about how the brain works. What neuroscientists seek to explain is the physical basis for that reaction. Just because we explain the cause of a mental phenomenon, that does not mean that that phenomenon does not exist; very much the contrary.

    I cannot show anyone my theory of the stove’s hotness, nor can I show anyone the evidence (other than bodily marks) of my test of the stove’s hotness.

    On the contrary: you can build a thermometer or some other measuring device that will reliably detect the stove’s temperature, and independent observers can use that device (or build one of their own based on your instructions) to obtain the same results when they repeat the test.

    And for that matter, why wouldn’t a burn or a blister be evidence? If you were so inclined, you could set up a bizarre experiment where some people touch the stove and others don’t, and an independent observer who wasn’t present to observe that would then try to determine which people had touched it by examining their fingers. It’s an event that leaves evidence available for any objective person to examine.

    You’ve said that my experience was a form of evidential testing. But how should others know that?

    Because your test is repeatable under specified conditions. When it comes to hot stoves, you don’t have to tell people that they’ll perceive the heat only if they have enough faith, or under rare and mysterious conditions that neither you nor anyone else can ascertain in advance. Turn the dial, let the stove heat up, and if you touch it you will be burned. None of this is applicable to religious experience.

  • Jarrod

    I think I see the difference you draw between the hot stove and religious experience – although I wonder whether the difference is only apparently easy – but I don’t feel that we’re talking about the hot stove in the same way.

    It seems that you’re viewing the hot stove as someone standing back and watching people approach and touch the stove. Call it an external point of view. I’m wondering about an internal point view: the perspective of an individual as he or she approaches the stove and touches it. (Ignore the presence of other people and, again, skin marks. Neither are required for experiencing or learning from experiencing the hotness of the stove.) James said that knowledge gained through personal experience still could be evidence based: i.e. a person has a theory about the stove and tests that theory by touching it. The person’s theory remains inaccessible to everyone but him or her, because, at the moment it’s only an internal mental construction (that’s what I meant by not being able to show the theory to anyone). The experience itself of the stove’s hotness is known only by the person having the experience. Sure, other people can see the effects of the stove, or watch the person touch the stove, or hear the person comment about the hotness of the stove – but only the person who actually touched the stove will have knowledge of his or her particular experience with the stove.

    The person touches the stove and interprets it as hotness – or, rather, the person interprets his or her experience as one wherein he or she touched a hot stove.

    Now, if a person has a religious experience, it seems that we can’t deny that an experience happened – the experience itself, after all, is knowable only to the person who had the experience. What’s debatable is the interpretation of the experience, i.e. whether the religious experience was an experience of divinity or of laced muffins. I don’t think any of this is controversial; it seems that this is just the nature of experience and experiential knowledge.

    When people talk about “proof,” they seem to want some externally verifiable sign of something. My point in bringing up the notion of experiential knowledge is that I’m not sure if externally verifiable signs exist for all there is to know. (The key word, I guess, is external.) Perhaps this is what Sullivan had in mind. Maybe he wasn’t convinced by “proof”; rather, maybe he was convinced by experience. Cases against his Christianity, then, must not only trump his arguments – they must also trump his experience.

    Experience is used to argue for both sides. What to make of it? Atheists say religious believers are deluded, and I’ve heard Christian presuppositionalists say the same about atheists. We can’t separate ourselves from experience, nor can we separate ourselves from even immediate interpretations of experience.

    No answers here. Just hopefully worthwhile thoughts.

  • James Bradbury


    Thanks for continuing to discuss this with an open mind.

    In your example, the stove unquestionably exists in the real world. Perhaps we can’t be sure (without a CAT scan) that the experience of it exists in someone else’s mind, but when we can see them touch it and touch it ourselves, it is reasonable to infer that they have a similar experience to us. If they touched it and didn’t recoil in pain we might come to the conclusion that they were having a very different experience.

    With a religious experience, not everyone can/does experience it and many would debate whether it really happened. Meanwhile for those who have had a religious experience it will seem utterly real even if they know such things to be improbable. They may feel like a fool to admit that their mind was playing tricks on them and they believed it, but anyone’s mind can create convincing illusions. I don’t think it’s fair to expect people to be convinced by the illusions going on in someone else’s mind.

    My point in bringing up the notion of experiential knowledge is that I’m not sure if externally verifiable signs exist for all there is to know.

    I’m not sure the stove experience is an example of this and I can’t think of any.

    Re: Proof. Doesn’t really exist in the real world. Some say proof is for mathematics and alcohol. All we have are theories which are consistent with observed evidence.

  • Jon

    I don’t think Sullivan will ever come out as an atheist. He’s a writer, and as such, he relies heavily on an interesting and contradictory persona: liberalish conservative gay catholic. If you take out the Catholic part, he becomes a little less interesting. That’s why this debate between Harris and Sullivan bores me. It’s not just because Harris is 100x more convincing, it’s that I suspect Sullivan’s catholicism is part of his marketing. Just a hunch. Love the site by the way.

  • Otto Henderson

    Dear Adam,

    Just found this site and am very impressed. Like anything else in life, one has to whip through a lot of chaff to find the good grain…

    That said, I would like to submit that perhaps we need to see people like Sullivan in a different light. You said:

    “Andrew Sullivan, after all, is not the problem. He does not want a theocracy.” You then said he has a problem with those that do, but doesn’t have any answers.

    I submit that Andrew Sullivan is as guilty as any pederast priest, any adulterous pentacostalist, and any hypocritical homosexual behind the pulpit. As I have gathered from Harris’ rather strident stand, it is the middle of the roaders, the quiet believers who do not publically and actively police their fringe freaks who are ultimately responsible for letting this go on.

    Sullivan allows that a book of poorly plagiarized fairy tales is good enough for him, so it should be okay to let the hate it promotes go unprotested. Sullivan, Fallwell, ANY person who puts stock in an invisible guy in the sky IS part of the problem and they need to be set straight and exposed for the magical thinkers they are.

    I have always enjoyed Sullivan’s columns but I am glad Harris is showing us that even ‘benign believers’ need to be held up to the same standard as any fringe freak.
    To think that there are people not only walking around on this planet but breeding as well who allow their emotional feelings about an invisible guy in the sky to dictate their entire approach to life makes this human ashamed to be on the same tree…

    Anyway, congrats on a beautiful, thoughtful, and hopefully long-lived site.

    Yours in freethought,


  • Jarrod


    I’ve enjoyed this little exchange and have learned a thing or two. Unfortunately, I’ll have to stop after this post. (Let’s hear it for mid-term assignments.)

    I’m still struck by how religious experience doesn’t seem all that different from typical experiences. I acknowledge that there are differences – it seems at least true, as you said, that “With a religious experience, not everyone can/does experience it and many would debate whether it really happened.” But, nevertheless, there are similarities that make discounting religious experiences less easy.

    Both types of experiences involve a personal knowledge that is externally unverifiable (except with something like a CAT scan). Both types of experiences involve comparing a personal knowledge with the similar experiences of others: a religious person would say that your statement “when we can see them touch it and touch it ourselves, it is reasonable to infer that they have a similar experience to us” applies to religious experience as well as typical experience (hence the existence of religious congregations). And both types of experiences can occur even in the face of improbability – the phrase “I wouldn’t believe it if I hadn’t seen it” is used often in reference to typical, not religious, experiences. (To use music as an example, again – there’s plenty of music out there I would’ve never believed existed if I hadn’t actually heard it.)

    Might there not be a difference, then, between typical experiences and religious experiences (barring, of course, some future revelation from brain-scanning research)?