July 21, 2011
The logic of science-7: The burden of proof in science
(For previous posts in this series, see here.)
The logic used in arriving at scientific conclusions closely tracks the legal maxim that 'the burden of proof rests on who asserts'. It should be noted that the word proof used here does not correspond the way it is used in mathematics, but more along the lines used in law. As commenter Eric pointed out in response to the previous post in this series, in the legal arena there are two standards for proof. In criminal cases, there is the higher bar of proving beyond a reasonable doubt, but in civil cases the standard is one based on the preponderance of evidence. So if the preponderance of evidence is in favor of one position, it is assumed to be true even if it has not been proven beyond a reasonable doubt. Scientific propositions are judged to be true not because they have been proven to be logically and incontrovertibly true (which is impossible to do) or because they have been established by knowledgeable judges to be beyond a reasonable doubt (which is not impossible but is too high a bar to result in productive science), but because the preponderance of evidence favors them. Evidence plays a crucial role here as it does in legal cases.
Scientific claims can be both existence claims and universal claims, and these two types of propositions are proved in different ways. In science the burden of proof in existence claims lies, as in legal claims, with those who make the claim. If they cannot meet the standard of proof, the claim is presumed to be false. With universal claims however (once at least some positive evidence has been provided in support of existence), the burden of proof lies with those trying to show that it is false. In the legal context, a witness who swears to tell the truth is assumed to be always telling the truth, a universal claim. A lawyer who wishes to make the point that a witness is not truthful is the one who is assumed to making an assertion and thus has the burden of proof to show that the witness has lied.
For an example of proof of an existence claim in science, the claim that an entity called an electron exists has to be supported by evidence that shows that an entity with the postulated properties of an electron (such as its mass and charge) has been, or at least can be, detected in experiments. The reason that I say 'can be' is that in some cases if there is strong circumstantial evidence in favor of the existence of an entity, a provisional verdict in favor of existence may be granted, pending more direct confirmation. The most famous case of this may the 'ether', which was postulated to exist on the basis of circumstantial evidence that it should exist, until it was shown that the theory of relativity undermined all that evidence in its favor and its existence was rejected. The neutrino is example of something that was granted provisional existence and was later directly detected.
The reason for these rules about how to judge the truth of existence and universal claims is simply because without them science would be unworkable. In most cases of scientific interest, it is impossible to prove that an existence claim is false and without these rules we would be swamped with existence claims for non-existent entities. The film Avatar, for example, postulated the existence of a valuable mineral called Unobtainium on another planet called Pandora somewhere in the universe. How could one possibly prove that such a mineral (or even the planet) does not exist? One cannot. Thus originates the scientific rule that to establish that a proposition of existence is true, one has to provide positive evidence in support of it. In the absence of such evidence, a perfectly justifiable scientific conclusion is that the proposition is false and that it does not exist.
This rule is hardly controversial. It is used in everyday life by everyone because would be impossible to live otherwise. To not have such a rule is to open oneself to an infinite number of mythical entities. To allow for the existence of something in the absence of a preponderance of evidence in support of its existence means believing in the existence of unicorns, leprechauns, pixies, dragons, centaurs, mermaids, fairies, demons, vampires, and werewolves.
This is why it is perfectly valid to conclude that there is no god. 'There is a god' is an existence claim and the burden of proof lies with those making the claim. Since no one has produced a preponderance of evidence in support of it, the claim is not to be taken seriously. Religious apologists who try to argue that god exists using logic alone without producing a preponderance of evidence in its favor are not being scientific and have entered the evidence-free realm of theology, in which one starts with whatever one wants to believe and then manufactures reasons for believing in it, even if that same reasoning is not applied to any other sphere of life.
Religious 'logic' is beautifully illustrated by this cartoon.
Next in the series: The power of universal claims in science