Bertrand Russell illustrated the Burden of Proof with his Celestial Teapot analogy, but he also did the best possible job, in my opinion, of addressing the question of evidence that God does not exist.
This essay is the best I've found in terms of summarising Russell's position on this topic, and is from the excellent Bertrand Russell Society.
Was Bertrand Russell An Atheist or Was He Really an Agnostic?
By R. Perkins
It’s quite clear that up until he was sixteen or so, Bertrand Russell did believe in the existence of God, understood in the traditional Judeo Christian sense. It’s also clear that from roughly that time on he abandoned his belief in God on grounds of lack of evidence. But it’s less clear whether his lack of belief should be described as “agnostic” or “atheist”. Russell himself considers the question on several occasions, but his conclusions are not completely unambiguous.
We ought to understand right away that Russell was concerned with the traditional Judeo-Christian concept of God as a perfect being—all powerful, all knowing and all good (all PKG).
In his Autobiography he recounts his WW1 prison experience in which the warder asks his religion. He replied “agnostic.” The warder asked him how to spell it. When Russell did so, the warder wrote it down remarking, “Religions are many, but I suppose they all believe in the same God.” Russell writes that this exchange kept him cheerful for a week.
In his famous 1948 BBC debate with Father Copelston, he explicitly calls himself an agnostic. And in a 1953 paper “What Is An Agnostic” Russell says the atheist—as opposed to the agnostic-- claims that we can know there is no God; and he denies that he would make such a claim. So too in his 1962 Essays in Skepticism (pp. 83-4) and in his 1959 interview with Woodrow Wyatt (BR Speaks His Mind, p. 20), Russell says that he does not claim certainty (or to have a “conclusive” argument) for God’s nonexistence. These passages do strongly support the claim that Russell was really an agnostic, not an atheist.
But significantly, in the same period he also often says that he regards the epistemic status of God’s existence as “exactly” the same as that of the Olympic gods or the Norwegian gods. He says he can’t prove that they don’t exist either. As he puts it in his 1959 interview with Wyatt:
“I don’t think it certain that there is no such thing as God, no. I think it is on exactly the same level as the Olympic gods, or the Norwegian gods, they also may exist. I can’t prove that they don’t. I think they are a bare possibility.” (See handout item 1)
The Christian God, he thinks, is no more likely to exist than the gods of Homer, which he --and most us-- believe non-existent. So, in a practical, popular sense, his position could be better described as “atheist.” As he puts it in his 1962 Essays in Skepticim (Philosophical Library): [Therefore] If I am to convey the right impression to the ordinary man in the street I think I ought to say I’m an Atheist.... (pp. 83-84)
The point to take here is that Russell’s epistemic similarity between the existence of the Homeric gods and the Christian God makes it fairly obvious that he thinks that it is significantly more reasonable to believe that God does not exist than that He does. And in that sense it’s misleading to call him an agnostic rather than an atheist.
In a little known 1939 lecture at U of Mich, unpublished until 1996, Russell considers the existence of a limited deity along the lines of that argued for by H.G. Wells and others. Of such a God Russell says: : “That sort of God is, I think, not one that can be actually disproved, as I think the omnipotent and benevolent creator can.” Vol. 10, “The Existence and Nature of God”, p. 258. (See handout item 2)
Just what sort of proof Russell likely had in mind here is a question we shall address presently.
Let’s understand “atheist” as one who holds that there is justified, if not conclusive, reason to believe there is no God. We needn’t be concerned here with a detailed account of the epistemology of evidence beyond saying that when the evidence for a belief is warranted and makes the belief significantly more likely than not to be true, we can say the belief is justified, perhaps even “known”, though in a sense short of certainty. In this sense, it seems pretty clear that Russell did believe that God does not exist, and in that sense, was an atheist.
Now, the important question becomes: why did he believe this? What sort of evidence did he think warranted belief in God’s nonexistence?
Russell’s Evidence for the Nonexistence of God.
I contend that Russell’s evidence is developed in two cogent arguments that incorporate some of the same ideas that science uses when it goes about assessing empirical hypotheses. The main idea is the idea of testability, the idea that if a hypothesis or theory is true, it will make a difference in the world in some measurable, observable way. This is closely connected to the idea that hypotheses (under certain conditions and assumptions) make predictions about what we will observe. And if those predicted occurrences are indeed observed, then the theory has thereby been confirmed. But if not, if what the theory predicts is not observed to occur, the theory has been confuted. Russell uses this scientific method of assessment, especially the method of confutation, in his assessment of the God hypothesis.
But before we consider Russell’s atheistic arguments, let’s first look at an example of Russell’s that has become widely cited in recent years, but that could be misunderstood as showing that Russell thought that the God hypothesis could not be confuted and that agnosticism was the most that a scientific examination of such a hypothesis could warrant.
I’m talking about Russell’s “china teapot” example which he compares to the problem of proof for God’s existence. (See handout item 3)
Many orthodox people speak as though it were the business of skeptics to disprove received dogmas rather than of dogmatists to prove them. This is, of course, a mistake. If I were to suggest that between Earth and Mars there is a china teapot revolving about the sun in an elliptical orbit, nobody would be able to disprove my assertion provided I were careful to add that the teapot is too small to be revealed even by our most powerful telescopes. But if I were to go on to say that, since my assertion cannot be disproved, it is an intolerable presumption … to doubt it, I should rightly be thought to be talking nonsense. (CPBR Vol. 11, “Is There A God?”, 1952, p. 548)
The implication, up to the last sentence, might seem to be that God, like the teapot, can’t be disproved, and the most we can do is to doubt--and thus, we can’t get beyond agnosticism. But what Russell is actually doing is taking the tea potters (and, by analogy, the theists) to task for taking lack of proof for the teapot’s nonexistence as good reason not to doubt its existence, and even as sufficient evidence to affirm its existence. This would clearly be the fallacy argumentum ad ignorantium, which takes the general form:
We don’t have sufficient evidence for p (not-p)
Therefore, we are justified in believing not-p (p)
One problem is that belief in God, as with celestial teapots, needs positive evidence--not merely the absence of evidence for the negative (the nonexistence of God or celestial teapots). In other words, the burden of proof is on the believer. Of course, agnostics aren’t believers, and they have no burden of proof. But atheists are, i.e., they believe in the nonexistence of God. And I don’t think Russell means to exclude them from the burden of proof. More on this in a minute.
Another problem with the above teapot fallacy is that our lack of evidence may be temporary. Tomorrow, for example, we may have more powerful telescopes capable of seeing teapot-sized objects tens of millions of miles away. And if we did, then we may be in a position to put the teapot hypothesis to a test, and thus confirm or confute it.
And this brings us back to the issue of Russell’s evidence for his belief that God does not exist. His reasons seem to come down to two sorts of arguments—one well known, the other less so. Of course these won’t be conclusive proofs in the sense of a valid argument with premises that are certain or unable to be doubted. Russell held that almost everything could be doubted. He more than once claimed that he couldn’t strictly prove that the world was more than five minutes old (e.g. Religion and Science, Conquest of Happiness, Analysis of Mind), or that there was a material world independent of our experiences (Problems, Ch 2). But he certainly thought that belief in these things was justified and more reasonable than non-belief.
Let’s look at two atheistic arguments both of which are suggested by Russell and which, I believe, he would have regarded as cogent.
The Argument from Evil
The problem of evil as Russell used it may be presented as a reductio ad absurdum of the hypothesis that God exists. The strategy is to show that the hypothesis has logical implications which contradict other statements that we know (or have good reason to believe) to be true. Russell gives the essence of the argument: (See handout item 4)
“It’s a most astonishing thing … that this world with all …. its defects, should be the best that omnipotence and omniscience have been able to produce in millions of years. Do you think that if you were granted omnipotence and omniscience and millions of years in which to perfect your world, you could produce nothing better than the KKK or the Fascists?” (from “Why I Am Not A Christian” 1927)
Russell never formally or fully developed the argument, but it could be succinctly put as a form of modus tollens:
Prem1: If there were an all powerful, all knowing, all good (PKG) being, then the universe would be at least as good as such a being could bring about.
Prem2: The universe is not that good (it contains too much evil, especially the pointless suffering of innocent creatures).
Concl: Therefore, there is no God.
Now this argument is obviously valid by virtue of its form (modus tollens). But is it sound or cogent? Yes, but only if the premises are true or warranted. It would be virtually impossible to conclusively prove them true, especially the second one. But they do seem reasonable, significantly more reasonable than their denials. And I think Russell would agree..
An Argument from “Lack of Evidence”? Yes (see handout item 5)
Here’s an anecdote on Russell’s 90th birthday party (_New Yorker_, 1963):
A London lady sat next to him at this party, and over the soup she suggested to him that he was not only the world's most famous atheist but, by this time, very probably the world's oldest atheist. "What will you do, Bertie, if it turns out you were wrong?" she asked. "I mean, what if--uh—when the time comes, you should meet Him? What will you say?" Russell was delighted with the question. His bright, birdlike eyes grew even
brighter as he contemplated this possible future dialogue, and then he pointed a finger upward and cried, “Why, I should say, 'God, you gave us insufficient evidence.”’
If we combine this “insufficient evidence” with something Russell said in a 1953 interview (“What is an Agnostic”), when asked what sort of evidence, if any, could convince him, we get a hint of another sort of atheistic argument he very likely would accept. He answered: “I think that if I heard a voice from the sky predicting all that was going to happen to me during the next 24 hours, including events that would have seemed highly improbable, and if all these events proceeded to happen, I might perhaps be convinced…. But as far as I know no such evidence exists.”
So BR thought there could be evidence for believing in God (at least, he says, a god of “super-human intelligence”).
Now, it seems reasonable to expect that much good would likely ensue if God did give such evidence: many, including many who would otherwise be non-believers, would believe--and without intellectual dishonesty-- and thus escape eternal damnation (of course they wouldn’t get to appear in Warren Allen Smith’s Who’s Who in Hell either, but let’s ignore that); many potential criminals would refrain from crime; fewer of the faithful would lose their faith, sink into despair and forfeit their chance of eternal bliss., etc. Hence, one would expect that if God existed, such evidence would be given, i.e. would be easily attainable. And this suggests a new atheistic argument easily constructed from Russell’s hypothetical Judgment Day complaint God’s giving us “insufficient evidence”—an argument, paradoxically enough, from lack of evidence, but not a fallacious one:
Prem1: If there were a God (all PKG being) then we would have ample evidence of His existence.
Prem2: We do not have such evidence
Concl: Therefore, There is no God.
Valid? Yes. Again, by modus tollens. Sound or cogent? I think so. And I think Russell would think so too; he apparently did believe the premises true.
Russell did, of course, claim that the traditional arguments for God’s existence were all non-conclusive; indeed, he often said they were invalid. But he also allowed, so I’ve argued, that there were reasons for going beyond agnosticism, and that the case for belief in God’s nonexistence, though not conclusive, was at least warranted. Whether he was right about that is of course a matter for ongoing discussion. But I think we’re justified in describing Russell himself on the question of God’s existence as an atheist rather than an agnostic.
HANDOUT
Bertrand Russell: Atheist or Agnostic? BRS AM 2012
R. Perkins
1. 1959 Wyatt Interview:
“I don’t think it certain that there is no such thing as God, no. I think it is on exactly the same level as the Olympic gods, or the Norwegian gods, they also may exist. I can’t prove that they don’t. I think they are a bare possibility.” BR Speaks His Mind
2. 1939 Little known U of Michigan Lecture on Religion:
“That sort of God [a limited deity a la H.G. Wells et al.] is one, I think, which cannot actually be disproved, as I think the omnipotent and benevolent creator can.”
CPBR Vol. 10, “The Existence and Nature of God”, p. 258 (my emphasis, rp)
3. Russell’s celestial china teapot example:
Many orthodox people speak as though it were the business of skeptics to disprove received dogmas rather than of dogmatists to prove them. This is, of course, a mistake. If I were to suggest that between Earth and Mars there is a china teapot revolving about the sun in an elliptical orbit, nobody would be able to disprove my assertion provided I were careful to add that the teapot is too small to be revealed even by our most powerful telescopes. But if I were to go on to say that, since my assertion cannot be disproved, it is an intolerable presumption … to doubt it, I should rightly be thought to be talking nonsense. (CPBR Vol. 11, “Is There A God?”, 1952, p. 548)
4. Argument from Evil
“It’s a most astonishing thing … that this world with all …. its defects, should be the best that omnipotence and omniscience have been able to produce in millions of years. Do you think that if you were granted omnipotence and omniscience and millions of years in which to perfect your world, you could produce nothing better than the KKK or the Fascists?” (from “Why I Am Not A Christian” 1927)
Also similar remarks regarding God’s benevolence: “I do not think myself that its logical to maintain that evil can have been created by a creator that was completely good. If you coiuld imagine yourself in a position to create the world, having the power to create such a world as you would like, you would realize that to create this world you would have to be a fiend beyond imagination.” (CPBR, Vol. 10, p. 257)
More formally developed (a la modus tollens):
P1: If there were an all PKG being, then the universe would be at least as good as an all PKG being could bring about.
P2: The universe is not that good (it contains too much evil, especially the
pointless suffering of innocent creatures).
C: There is no god.
5. An Argument from “Lack of Evidence”? Yes
a. Anecdote from BR’s 90th birthday party (The New Yorker, 1963):
A London lady sat next to him at this party, and over the soup she suggested to him that he was not only the world's most famous atheist but, by this time, very probably the world's oldest atheist. "What will you do, Bertie, if it turns out you were wrong?" she asked. "I mean, what if--uh—when the time comes, you should meet Him? What will you say?" Russell was delighted with the question. His bright, birdlike eyes grew even
brighter as he contemplated this possible future dialogue, and then he pointed a finger upward and cried, “Why, I should say, 'God, you gave us insufficient evidence.”’
b. BR says he can imagine evidence for God (“What Is An Agnostic?” 1953):
“I think that if I heard a voice from the sky predicting all that was going to happen to me during the next 24 hours, including events that would have seemed highly improbable, and if all these events proceeded to happen, I might perhaps be convinced.... But as far as I know no such evidence exists.”
c. Suggested reconstruction (a la modus tollens) for non-fallacious argument from lack of evidence:
P1: If there were a God (all PKG being) then we would have ample evidence of His existence.
P2: We do not have such evidence
C: There is no god.
