Saturday, February 15, 2014

Alvin Plantinga on atheism

In Is Atheism Irrational (mentioned by 3quarksdaily who appear to have a real beef against atheism), Gary Gutting interviews Christian philosopher Alvin Plantinga on the philosophical status of atheism. Plantinga's arguments for the irrationality of atheism, which he defines as "the belief that there is no such person as the God of the theistic religions." Plantinga's critiques of atheism, however, are at best weak and sometimes fallacious.

Plantinga first asserts that most philosophers, even believers, reject the traditional theistic arguments for the existence of God as unsound. Plantinga, however, claims that merely the failure of these arguments does not provide sufficient grounds for atheism. Gutting mentions Plantinga's argument that the question of theism is similar to the question as to whether there are an even or odd number of stars in the universe. Plantinga believes this similarity is better than Russell's Teapot because we have "a great deal of evidence against teapotism," albeit indirect, whereas we have no evidence for or against the evenness of the number of stars. Hence, the proper response is agnosticism rather than atheism. However, Plantinga fails on a number of counts here.

First, he misrepresents Russell's argument. Russell adds to his argument, "But if I were to go on to say that, since my assertion cannot be disproved [by direct observation], it is intolerable presumption on the part of human reason to doubt it, I should rightly be thought to be talking nonsense." Russell is not talking about an abstract question of epistemology, but addressing the assertion of social and cultural privilege on the basis of unfalsifiable propositions.

Second, the even-number-of-stars (ENS) analogy fails to capture the scope and character of religious belief. ENS is a very simple, prosaic, even trivial proposition that has no implications at all on the character of the universe; there is no indirect evidence we could use to adduce any probability other than 50%. Religion is not so simple nor prosaic. Does Plantinga really want to claim that religion is as simple, prosaic, and trivial as ENS, without any conceivable indirect consequences?

Finally, the strict distinction Plantinga makes between agnosticism and atheism is a straw man. As Russell elaborates, "I ought to call myself an agnostic; but, for all practical purposes, I am an atheist. I do not think the existence of the Christian God any more probable than the existence of the Gods of Olympus or Valhalla." Russell, like many other atheists, consider atheism not a philosophical position but a practical position. Atheism need only be philosophical agnosticism applied to positive claims made without evidence.

Plantinga goes on to say,
I don’t think arguments are needed for rational belief in God. In this regard belief in God is like belief in other minds, or belief in the past. Belief in God is grounded in experience, or in the sensus divinitatis, John Calvin’s term for an inborn inclination to form beliefs about God in a wide variety of circumstances.
Let's try to take this statement apart. Plantinga sharply distinguishes arguments on the one hand, and experience or sensation on the other hand. According to Plantinga, belief in God rests on the same foundation as "belief in other minds, or belief in the past," none of which require argument to believe.

Plantinga seems to be using "argument" in more restricted sense than the common definition as "a connected series of statements intended to establish a proposition." We can easily create arguments based on experience; science is nothing but arguments from experience. Plantinga seems also to distinguish direct sensory experience ("sensus divinitatis") from "experience," which is presumably indirect. But drawing an indirect conclusion from experience requires argument. Belief in other minds and belief in the past both require argument, based on experience, to rationally establish. Even establishing the existence of a sensus divinitatis (direct sensory apprehension of God) requires an argument to establish that it is indeed a sense rather than a purely internal mind (brain) phenomenon. Plantinga is just speaking nonsense here, presumably to insulate his position against any kind of argument.

Although Plantinga denies that arguments are unnecessary for belief in God, he offers a few anyway. (Why? I don't need to argue that we can all see the tree in my back yard; if you are in doubt, come and look.) He starts with the Fine Tuning argument: "[T]he universe seems to be fine-tuned, not just for life, but for intelligent life. This fine-tuning is vastly more likely given theism than given atheism." Plantinga does not seem to understand that at best, the Fine Tuning argument is controversial, and probably proves the opposite: given fine tuning, atheism (naturalism) is vastly more probable than theism, because a God could create intelligent life given any physics, whereas life can exist naturally only in a fine-tuned universe.

Gutting simply accepts the strength of the Fine Tuning argument, but implies that a fine-tuning God "fall[s] far short of . . . an all-perfect God." In response, Plantinga defends Christianity with an obvious fallacy of the excluded middle: "[Y]our qauestion makes sense only if the best possible worlds contain no sin and suffering." The grammatical error aside ("best" is a superlative; there can be only one best possible world), the argument against an all-perfect God requires us only to say that this exact real world is not the best; while it's possible arguendo that the best possible world contains some sin and suffering, it could contain less sin and suffering than this one. An all-perfect God would, at best, create only the best possible world, even if He would create any imperfect world at all. If this is not the best possible world, an all-perfect God is either does not exist or the concept is vacuous.

Plantinga has an odd notion of what a good world is:
he first being of the universe, perfect in goodness, power and knowledge, creates free creatures. These free creatures turn their backs on him, rebel against him and get involved in sin and evil. Rather than treat them as some ancient potentate might — e.g., having them boiled in oil — God responds by sending his son into the world to suffer and die so that human beings might once more be in a right relationship to God. God himself undergoes the enormous suffering involved in seeing his son mocked, ridiculed, beaten and crucified. And all this for the sake of these sinful creatures.

I’d say a world in which this story is true would be a truly magnificent possible world. It would be so good that no world could be appreciably better.
Really. If this is the best possible world, I don't understand anything at all about good.

It's hard to see Plantinga as saying anything but that a world with the Holocaust, millennia of war, human-induced famine, genocide, rape, murder, torture is not only good but best; if those horrors had not happened, the world would have been worse, because God would not have suffered. At least Plantinga is clear: this is the sort of world that Christianity offers. I want none of it.

Gutting notes that atheists justify their disbelief by the assertion that God does not have explanatory power. Plantinga deflects this challenge by noting that lack of explanatory power does not justify disbelief. Plantinga argues that we are not justified in disbelieving in the existence of the Moon, for example, merely because it is no longer considered a good explanation for lunacy. But Plantinga seems to ignore that the existence of the Moon still retains considerable power to explain why we see a big round object in the sky that seems to correlate with the tides.

Plantinga then goes argues two ways in which God does have explanatory power. The first is religious experience. This argument would have considerably more force if religious believers all believed in the same God, rather than gods that always resemble their own cultures and justify their idiosyncratic social norms and political power structures.

Plantinga repeats Thomas Nagel's ad hominem speculation that atheists disbelieve in God because we don't want there to be a God. Plantinga asserts that theism poses a "serious limitation of autonomy." Plantinga is completely contradicting himself here; above, he argues that this is the best of all possible worlds because we have the maximum amount of autonomy, sufficient autonomy to inflict the worst sort of harm, suffering and pain on other human beings. Precisely what limitations on autonomy does theism impose? It is true, however, that atheists object to the limitations on our autonomy imposed by human beings who claim to be speaking for God, but I cannot imagine objecting to any limitations on my autonomy that a truly just and loving God would impose, any more than I object to the (rational and justifiable) limitations on my autonomy imposed by the state.

Gutting asserts (without mentioning, much less citing, any source) that materialism is a "primary motive" for atheism. I have no idea where he gets this from; having read a considerable amount of atheist philosophy, I have never seen this argument; it cannot possibly be "primary." I don't even know what Gutting means by "materialism"; in the 19th 19th century idea that nothing exists but atoms in motion, materialism has been long since debunked by science. Gutting here offers Plantinga an opportunity to expound on one of his favorite theories, that evolution is incompatible with materialism. Plantinga declares that he cannot give a full account of his argument in the article (or anywhere else; the full account is only in his book, which I might look at later) but his summary is so full of holes I can't imagine even the full treatment can fill.

Briefly, Plantinga argues that the content of a belief is distinct from its neurological properties, and that content has no causal effect. Two different neurological beliefs could have identical causal effects but different content. Because content has no causal effect, evolution cannot select for content. Therefore, the content of our belief in evolution was not selected for, therefore we have no reason to believe the content of our belief in evolution is true.

This argument, however, has a number of flaws, not only that no one is really a "materialist." What, precisely, does Plantinga mean by the content of a belief? If the content of a belief has no causal effect, then a naturalist would claim then that by definition we cannot determine what the content of a belief is, much less determine whether it is true. Indeed, the naturalist would say that "existence" cannot properly be applied to anything with no causal effect: the existence of a thing with no causal effect is indistinguishable from its nonexistence.

But perhaps "content" is epiphenomenal or emergent, in the same sense that the roundness of an inflated basketball is emergent or epiphenomenal: it is not the roundness per se that causes a basketball to bounce regularly, it is the atoms in the basketball that causes it to bounce regularly. However, the roundness of a basketball is fully determined by the arrangement of the basketball's atoms; change their arrangement to something not round, like an (American) football, and the ball will no longer bounce regularly. Thus, if the neurological properties of a belief determine the content of a belief (e.g. we cannot say that two identical neurological beliefs have different content), and evolution can select on the causal properties of a belief, then evolution does affect the content of belief, albeit indirectly. So Plantinga's argument simply fails.

I've been following Plantinga's evolutionary argument for many years, and he has not, to my knowledge, substantively altered his summary to address any criticism; the summary he offers to Gutting is substantively identical to summaries I read years ago. This is the best sophisticated philosophy that the theists have, and it's nothing but a mass of sloppy reasoning that an economist can demolish in post.


  1. I've never understood why Plantinga thought that his evolution argument(which I've, admittedly, only read in basic, truncated form) implied anything surprising at all. The message of his argument, AFAI can remember, is that we have no way of explaining human rationality given the way natural selection works, right? Eg.(and I think this is his example, albeit lacking in much detail) the genes of a child who runs away from a tiger will be selected for but whether that child believes the tiger to be trying to kill and eat them, or merely playing with them in an unthreatening, even friendly, fashion, is irrelevant to evolution.
    First of all, this example involves ignoring an almost infinite number of ancillary factors that ARE 'visible' to N.S. - off the top of my head, a child who believes the tiger is antagonistic is more likely to avoid them in the future, and have higher survival chances. Those genes will proliferate. The assumption one has to make for Plantinga's argument to work is that having a worldview which is relatively consistent, and in which you can draw lessons from one part and apply them to another, confers no survival advantage.
    It also requires us to believe that the tiger thought experiment, and in particular its ad-hoc, needlessly complicated 'friendly-tiger' interpretation, is a typical example of the kind of situation upon which N.S. operates - it's not, and the more one's beliefs track reality, the more possible it is for us to draw comparisons which equip us to deal with unfamiliar problems.
    Let's say the child continues to believe tigers are friendly yet, implausibly, survives to hunting age - a rational hunter with reasonable beliefs would be able to USE their beliefs in other spheres - eg. how best to approach and kill a ferocious tiger. The hunter who continues to believe tigers are harmless is incapable of applying that belief in a constructive fashion when hunting. Moreover, holding ridiculous beliefs requires ever more complex 'saving of the appearances' in order to integrate them - the way you respond and deal with the sight of a tiger eviscerating a close friend depends on whether it comports with what you already believe. If you believe tigers are friendly how does this new information tally?
    In Plantinga's confabulated toy-world his example is not just unrealistic, it is isolated, and, crucially, one of the two beliefs is revealed to be of little use when applied elsewhere. The reality-tracking belief, OTOH, has what David Deutsch calls 'reach'.
    The final, most obvious point: unless Plantinga is saying N.S. would NEVER select for rational beliefs, unless he's saying rational beliefs would NEVER confer a Darwinian advantage, all he's really saying is that we should expect people to frequently behave irrationally and believe nonsense on a pretty regular basis. This is, I would contend, a less than startling conclusion.

  2. The message of [Plantinga's] argument . . .First of all, this example involves ignoring . . .

    Plantinga would, I think, reject this rebuttal as missing the heart of his argument. How, he would ask, would the content of a belief, "tigers are antagonistic," have any causal effect on the child's behavior. Only the neurological, physiological, and physical properties of the neurons have actual causal effect, not the subjective interpretation we place on those physical events.

    The assumption one has to make . . .

    That is indeed Plantinga's argument. I think Plantinga would argue that things like "worldviews" are applied post hoc to physically evolved neural structures.

    It also requires us to believe . . .

    Much as I love to beat up on Plantinga, I think your criticism is a little unfair. Every philosopher uses these kinds of unrealistic thought experiments, and there is some value in looking at unusual edge cases.

    Let's say the child continues to believe tigers are friendly . . .

    Again, I think Plantinga's argument would be that all of these things, cross-utility, coherence, comportment, etc., are after the fact; they are just subjective interpretations we are arbitrarily imposing on already-evolved beliefs. We happen to value cross-utility, he might argue, just because we subjectively like cross-utility, not because cross-utility was an essential component of truth.

    Remember, in Plantinga's view, one does not apply a belief in any meaningful; the content of a belief, Plantinga asserts, has no causal power over one's action. The materialist view entails, in Plantinga's argument, that our neurons force us to behave the way we do, and our consciousness is simply sitting back and applying a story to story-less neuronal behavior.

    The final, most obvious point . . .

    Heh. But no, I suspect Plantinga believes that there is far more rational content to our belief than can be accounted for by materialism.

    1. How does he address the state of neurons in one brain being transmitted to the state of neurons in another brain so that the neurons in the second brain causally effect that second brain to do different things? As in, "Hey Joe, Phil wants to kill you. Maybe you should run."

      Is his argument really that the semantic meanings of Joe, Phil, kill, and run are purely post-hoc and are just as likely functioning components of a "run in the right circumstances for social creatures" brains as some kind of post-hoc gibberish of 'false' beliefs? Seriously? I mean, I get that thinking about brains and minds are hard for people but isn't the word "belief" just shorthand for some complex neural state like "hurricane" is for a complex atmospheric state? With a different design perhaps we could just plug our brain into our fellow's brains and dump the low level neural state, but given that we are limited by grunts and noises, we have to compress. This compression, or shorthand, while not the actual thing, the neural states, must maintain a wide ranging isomorphism with those neural states, and with the external reality, to be generally useful.

      What is his opinion of the "beliefs" of something like the Deep Blue the chess playing machine? If I say Deep Blue "believes his opponent would sacrifice a rook and so he's going to attack the knight" I am in some sense projecting onto Deep Blue an abstract property of belief for what is, in fine detail, maybe a set of confidences assigned to a wide range of board positions in some lookup table, and that itself is really just some bits whizzing around in a CPU so that even talk of a confidences or lookup tables is an abstraction. But to say that the content of Deep Blue's "beliefs" don't effect it's moves because, at base, it's just bits pushing bits around is to completely miss what the word "belief" means in this case, which is a shorthand summary of those bit states, a compression of those bits that loses as little information as possible. If Deep Blue has to transmit it's understanding of the chess situation to another computer but is restricted to only a hundred bits it'll be forced into the same kind of state compression that language forces us into. It'll end up telling the other computer something like, "I believe he'd sacrifice his rook so attack his knight." And the other computer, while losing some of the information in Deep Blue's state, will be able to proceed similarly nonetheless.

      As for toy examples, his argument would be more convincing if he'd do the hard work of constructing a more extended example. Sure, a toy example could be constructed to outline some interesting corner case, but I don't think his example does that. His example seems rather designed precisely to wave away all the really important detail about what he is describing.

    2. Is his argument really that the semantic meanings of Joe, Phil, kill, and run are purely post-hoc and are just as likely functioning components of a "run in the right circumstances for social creatures" brains as some kind of post-hoc gibberish of 'false' beliefs?

      Yes, basically.

  3. Ray Freeman-Lynde2/24/14, 2:35 PM

    "These free creatures turn their backs on him, rebel against him and get involved in sin and evil. Rather than treat them as some ancient potentate might — e.g., having them boiled in oil ..."
    Isn't that sort of what the old testament god did?

  4. Fine-tuning required to allow life... If the laws of physics were different, who's to say that a different type of life and intelligence.wouldn't exist.

    I like the "many universes, each w/ different laws of physics".hypothesis. If the odds against life form is 1 in a trillion, and there are a million trillion universes, then you're going to get life somewhere..
    If there is an infinite number of universe, then anything can happen.

    Remember, in our galaxy alone there are 400 billion stars. And there's maybe 100 billion galaxies in our universe. The slight chance that life can form becomes a sure thing.

    This post is not anonymous.

    Dave Lerner

    1. David, a thought on probability of life. It's quite easy to get to a probability where life is unlikely in our observable universe. Maybe we'll be able to put realistic numbers on this in a few decades. The detection of life in other star systems is a possibility. We may also find that our universe is infinite or just zillions of times bigger than what is observable. This would shift the odds dramatically towards life arising somewhere. Our solar system may be that somewhere.

  5. "G.G.: O.K., but in any case, isn’t the theist on thin ice in suggesting the need for God as an explanation of the universe? There’s always the possibility that we’ll find a scientific account that explains what we claimed only God could explain. After all, that’s what happened when Darwin developed his theory of evolution. In fact, isn’t a major support for atheism the very fact that we no longer need God to explain the world?

    A.P.: Some atheists seem to think that a sufficient reason for atheism is the fact (as they say) that we no longer need God to explain natural phenomena — lightning and thunder for example. We now have science.

    As a justification of atheism, this is pretty lame. We no longer need the moon to explain or account for lunacy; it hardly follows that belief in the nonexistence of the moon (a-moonism?) is justified."

    What's pretty lame here is that Plantinga made a God of the Gaps argument (fine tuning) and claims a few paragraphs later that the position that God is not necessary to explain the physical world is a poor reason for rejecting belief.

    Where are these atheists who think that not needing God to explain natural phenomena by itself is "sufficient" to justify atheism? No, the lack of necessity is sufficient to demolish the argument from fine tuning. Thus, it supports atheism, and Plantinga's reply is non-responsive in addition to being a straw man and a failure to connect what he's saying now with what he's just said.

    1. Yeah. Plantinga is usually fractally wrong.

    2. (re anonymous posting: I am George Locke )

    3. You don't have to use a real identity. As long as you consistently use a definite pseudonym (i.e. "flies"), you're fine.


Please pick a handle or moniker for your comment. It's much easier to address someone by a name or pseudonym than simply "hey you". I have the option of requiring a "hard" identity, but I don't want to turn that on... yet.

With few exceptions, I will not respond or reply to anonymous comments, and I may delete them. I keep a copy of all comments; if you want the text of your comment to repost with something vaguely resembling an identity, email me.

No spam, pr0n, commercial advertising, insanity, lies, repetition or off-topic comments. Creationists, Global Warming deniers, anti-vaxers, Randians, and Libertarians are automatically presumed to be idiots; Christians and Muslims might get the benefit of the doubt, if I'm in a good mood.

See the Debate Flowchart for some basic rules.

Sourced factual corrections are always published and acknowledged.

I will respond or not respond to comments as the mood takes me. See my latest comment policy for details. I am not a pseudonomous-American: my real name is Larry.

Comments may be moderated from time to time. When I do moderate comments, anonymous comments are far more likely to be rejected.

I've already answered some typical comments.

I have jqMath enabled for the blog. If you have a dollar sign (\$) in your comment, put a \\ in front of it: \\\$, unless you want to include a formula in your comment.