There were a couple of interesting (anonymous) comments in response to my post on what constitute rational and irrational beliefs. The writer said that I was overstepping the line that divided science from philosophy when I argued that religious beliefs were irrational. The arguments took a familiar form and went something like this:
1. We cannot prove that god does not exist.
2. Hence it is rational to believe that god exists.
3. Scientists should stick to the world of data and not venture to question god’s existence since that enters the realm of philosophy, not science. The author states that if a scientist is asked: ‘In your scientific opinion, does God exist?’ the proper answer should always be, ‘I don’t know. I don’t have any data on the subject.’
But the other two statements do not follow from the first. Just because we cannot prove, using data, the negation of some entity does not mean that it is reasonable to believe in that entity. Scientists constantly make judgments in the absence of data and act on those judgments. In fact, it is essential that they do so, as science could not proceed otherwise.
The only time that you can prove a negative is if you have the ability to do an exhaustive examination of every possible situation. As an example, I can prove to everyone’s satisfaction that no unicorns exist in my office because I can search every nook and cranny and show that none are there. But I cannot similarly prove that no unicorns exist anywhere on the Earth or elsewhere in the universe.
I also cannot prove the non-existence of magic unicorns in my office, that only materialize when I am not present and are capable of hiding all evidence of their visits before they disappear again. It seems to me that arguments for the existence of god are of this nature.
But there is another point about the word ‘proof’ that needs to be emphasized. When scientists use the word ‘proof’ they use it in a slightly differently way from the way mathematicians use it. In mathematics, a proof is a construct based on an agreed set of axioms and rules of logic. If someone challenges the validity of any of the axioms or one of the rules, then the proof is also called into question. But since the axioms are usually few in number and do not necessarily have to be based on data, mathematicians can agree on the validity of more things as working hypotheses than scientists can.
Scientific ‘proofs’ do not have the same level of rigor as a mathematical proofs because the axioms themselves are not simply assumptions but are also expected to justified based on evidence. Also there are far more explicit assumptions that go into scientific conclusions than go into mathematical proofs, thus opening them up to far more challenges. This greater degree of challenge that scientific assumptions receive makes scientific ‘proofs’ different from mathematical proofs. So although I and other scientists use the word proof frequently, we do understand that it is being used in a slightly different sense than a mathematical proof. The word proof is used to signify a reasoned judgment based on the merits of the evidence.
But just because scientific proofs do not have the same status as mathematical proofs does not mean that scientific conclusions cannot be extremely robust. Let me give an example. Most people readily accept that there are just two kinds of electric charge, positive and negative. This is about as well-established a ‘fact’ as one is likely to find in science. This is one of the most firmly held beliefs in all of science and the entire modern world is constructed on the basis of this two-charge model. No one even thinks of questioning this fact. (Note that ‘positive’ and ‘negative’ are just labels and the charges could just as well have been called things like ‘green’ and ‘blue’.)
The interesting question is how it is that we are so certain that there are just two kinds of charges that we base our entire society on it. Do we have certain proof that there are only two kinds of charges? Do we have direct data that no more charges exist? Have we looked everywhere and convinced ourselves of this? The answer to all three questions is no. So how is it that we are so sure that only two kinds of charges exist? It is because of the absence of certain kinds of data.
Here’s how that argument works. Suppose you have three charged objects A, B, and C. What scientists find is that if the charges are such that A and B attract each other and A and C attract each other, then it is always found that B and C repel each other. This set of three observations can be explained by (1) postulating that there exist just two kinds of charges, and (2) adopting a rule that says that like charges repel and unlike charges attract. No data has ever been seen that contradicts the consequences of these two assumptions.
Because of the absence of any data that contradicts any predictions based on those two statements , scientists will say that they are extremely confident that there are only two kinds of charges and this is all the ‘proof’ they need. But note that haven’t actually proved it in a mathematical sense. It is just a powerful inference based on the absence of certain kinds of data, but it is sufficient proof to convince scientists.
Notice though that even this ‘proof’ can be challenged. After all, we have done such experiments with just a few sets of charges. We have not exhaustively repeated them with every single charge that exists in the universe because it would be impossible to do so. As a result, someone can come along and say that scientists are wrong, that there does exist a third kind of charge but that either it has not been found yet or that it does not interfere with the experiments that scientists do. There is no way that scientists can prove this person wrong. How could they? But what they will do is ignore this argument as not worth responding to because that kind of argument has the same standing as magical unicorns in my office or a god who is determined to avoid leaving evidence of his/her existence.
A belief that has no observable consequences is of no use to scientists and they will work on the assumption that this third charge does not exist and that would be perfectly rational behavior. A person who clings to the belief in a mysterious third charge that has no observable consequences will be treated as somewhat eccentric.
Historians and philosophers of science have long pointed out that there is no proposition in science, however idiotic, that cannot be made immune from refutation by the addition of a protective belt of auxiliary hypotheses to shield its weaknesses. But if you want to convince scientists that something like a third kind of charge exists, you will have to provide positive evidence, some actual data that cannot be explained by a two-charge theory. For scientists, the absence of such evidence or data is taken as evidence of absence.
It seems to me that the arguments put forward by believers for the existence of god are of the same kind as those that might be put forward for a third charge: It exists but its effects cannot be observed. But just as scientists are perfectly justified in rejecting as irrational that kind of hypothesis when applied to a third charge and confidently proceeding on the basis that it is false, so it is that we can confidently reject the arguments currently given for the existence of god.
So although you may not be able to prove exhaustively that god does not exist, you cannot obtain a stronger scientific proof than what we currently have.
So if someone should ask me ‘In your scientific opinion, does God exist?’, I would answer ‘No’ with the same degree of confidence that I would say ‘No’ to the question as to whether a third type of electric charge exists.
POST SCRIPT: More lists of famous atheists
Although it should be obvious, I should add that the mere fact that someone famous is an atheist is not being offered as an argument in favor of atheism. Lists of this kind are simply to identify the members of an affinity group. One could do the same thing with lists of vegetarians or Bassett hound owners.