- Misplaced burden of proof
- Privileging the hypothesis
- Proving a negative
"You can't prove that your house is not full of invisible landmines, so you'd better stay still in case you accidentally set one off."This is a good demonstration of why the negative side doesn't bear the burden of proof; it is for all intents impossible to demonstrate something is absolutely incapable of happening, and it would be impossible to live one's life in light of all the things that might be true. It's far easier, if there is proof of the positive, to demonstrate it. Put more simply, if someone has advanced no good reason to believe something is true, believing it is true anyway is unreasonable. This does not necessarily mean it is untrue (see Fallacy Fallacy), but it does mean it is not logical to believe it is true. Linguistic trickery can often be used to make a negative appear to be a positive (for example, claim that rejecting the existence of the invisible landmines is a positive claim that "I can explain everything in the universe right now"). It pays to be careful in evaluating which side the burden actually belongs on. One popular form of shifting the burden of proof is to demand your opponent ‘do their own research’. This places the onus for proving the point you're trying to make onto them. This particularly applies if the research they're supposed to perform is defined extremely vaguely, such as ‘take some classes’, ‘you can find dozens of examples’ or even ‘use common sense’.
- In Carl Barks' story "The Golden Helmet", a viking named Olaf the Blue left the titular helmet somewhere in the continent before the arrival of Christopher Columbus. Azure Blue, a (self-proclaimed) descendant of Olaf's, planned to use his (alleged) relation and the Code of Discovery to claim North America as inheritance from Olaf. Whenever somebody asked how Azure could prove he was a descendant of Olaf the Blue, Azure's Amoral Attorney countered by asking how they could prove he wasn't. Azure wasn't the only character trying to use this fallacy either.
- At the beginning of Plan 9 from Outer Space, the narrator claims it is a depiction of real events based on sworn testimony. At the end he asks, "Can you prove it didn't happen?" (however, note that the film was never actually billed as being Based on a True Story).
- At the end of the novel Cetaganda, the Cetagandan Emperor gives Miles, a traditional enemy of his empire, the highest award in his empire, then has him alone with three other members of the aristocracy, but asks him nothing. When Miles rejoins his superiors, the Spy Chief asks him what happened in there. Miles grins and says "Nothing," noting that there's no way to prove that the Emperor's goal was to make his own people think he might have been suborned or cut a secret deal, which was the Emperor's goal. "Let's see you try to prove a negative. I want to see you try."
- In the Rihannsu series the Romulan legal system works on the assumption the defendant is guilty until proven innocent... Except that the prosecution has to prove its case first to even get to the trial, and the defense can work by simply proving the prosecution missed an alibi.
- Early in Umineko: When They Cry, Battler relies on the "Devil's Proof" frequently. One example is the "mysterious person X" he posits, because you can't prove there isn't such a person on the island. Beatrice once shifts the burden back, magnified, with Hempel's Raven: since this "mysterious person X" can't be found, the only way to prove that's the culprit is to prove everyone else is innocent by watching all of them all the time!
- In Ava's Demon Ava is unconvinced by salvation promised by TITAN on these grounds.
Ava: Do they ever leave?Gil: Why would they want to leave? No one leaves Paradise. What a stupid question.Ava: I just thought the followers would come back and talk about it, or something. I guess you don't know if it really exists then, heh.
- In a legal system where the burden of proof is held by the prosecution, with the defendant regarded as innocent until proven guilty (applies to both common and civil law systems). The requirement is that the prosecution prove that the accused did do it; the defense doesn't have to prove he didn't. This can bite the prosecution badly, especially if the evidence is weak, or has been mishandled. It also doesn't necessarily follow the logical burden of proof, since the defence can demand the prosecution prove a negative. An infamous example of this is the O. J. Simpson trial, where O. J.'s defense lawyers demanded the prosecution prove an endless series of alternate scenarios were impossible.
- Meanwhile, since it's extremely hard for the defence to prove that someone definitely didn't commit a crime, when the burden of proof is on them in the media they often have to resort to proving that someone else did do it. That, and it's more exciting that way.
- It is common, at least in the US legal system, for the party bearing the burden of proof to be required to prove a negative. The most straightforward way to disprove A is to prove B, where B => ~A. That is, evidence inconsistent with the thing to be disproved is introduced. For example, in patent practice, nonreceipt of correspondence can be shown by producing the docket record into which the correspondence, if received, would have been entered.
- Note that in the US criminal law system, it isn't necessary for the prosecution to definitively prove guilt, merely "guilt beyond a reasonable doubt." That is, far-fetched alternative scenarios that also fit the evidence are not reason to acquit if they do not affect a "reasonable person's" belief in the defendant's guilt. As to which doubts are reasonable and which aren't, no one's really sure about that one. In US civil law however, the burden of proof is still on the plaintiff, but it's much lower. Jurors decide the case simply by a "preponderance of the evidence" (i.e. 51%), which is why O.J. Simpson was acquitted of murder but held liable for wrongful death. But if neither side is more likely than the other, then the defense wins.
- O.J. is a complex case, because his acquittal also reflected years of misdeeds by the LAPD, and the main police investigator was proven to lie on the stand while doubt was presented to how the DNA testing was done. Modern law school students, when studying the trial in depth have concluded that as presented to the jury, the evidence is pretty strong that O.J. was the victim of a botched frame-up job. He probably did it, but the framing muddied the evidence and increased the doubt.
- Another problem, as Vincent Bugliosi (who successfully prosecuted Charles Manson) and others pointed out, is that the prosecution failed to introduce evidence that indicated guilt — O.J. making a stop in Chicago where he may have dumped the bloody clothes/murder weapon, the police interrogation notes, etc. Why? They were probably afraid of looking for evidence and then finding nothing in Chicago, or introducing the notes where, even though O.J. made many incriminating statements, he still denied guilt.
- One of the most valid uses of the "Devil's Proof" comes in IP suits over black box technologies, such as software and patented manufacturing processes. If a company releases a product with similar enough function and appearance to a competitor's product, a plaintiff can't prove that their source code or production method has been infringed, and the court can't find conclusively that it wasn't without granting the plaintiff further rights of discovery or subpoenaing internal documentation from the defendant. Both of these remedies place the burden of proof on the defendant.
- A favorite of many groups; you'll often see demands to prove there is no global conspiracy or that aliens can't possibly have influenced the building of the pyramids.
- Invoked by both sides with regard to the Turing Test and the question of "strong AI": some people think that as long as you can't prove a machine isn't thinking, it makes compassionate sense to treat it as if it is, based on the fact that it would react in the same way as a human being (and in general, we do assume other humans are thinking creatures). Others counter that since passing a Turing Test only means failing to prove the machine is not thinking, there's still no reason to assume that it is (obviously at the moment the debate remains entirely academic.)
- The well-known quote "Extraordinary claims require extraordinary evidence" means that if you claim that something exists, you must provide evidence that it does. This manifests itself in many arguments, and results in a lot of shifting:
- A well-known deconstruction of this concept is the Invisible Pink Unicorn (May Her Hooves Never Be Shod). She is completely impossible to observe in any way, but you cannot prove she does not exist. Thus, anyone claiming that people must believe in X without a good reason NOT to do so must likewise believe in the Invisible Pink Unicorn. Russel's Teapot and the invisible dragon in Carl Sagan's garage are other well-known deconstructions of this fallacy.
- Theists frequently demand that atheists prove that there is no God (this is called presuppositionalist theism). It is impossible to disprove certain conceptions of God (a God who never does anything or interacts with the universe in any way is just as unobservable as the Invisible Pink Unicorn, and an omnipotent deity like the Invisible Spaghetti Monster who deliberately hides evidence of his own existence, up to the point of altering our brains so that we cannot observe it, is likewise by definition impossible to observe), making this an impossible task - but the burden falls on those making the assertions about the existence of their God, who frequently claim their deity does exert some influence on reality. This is complicated by the fact that atheism is traditionally defined as "the belief there is no God" but many self-described atheists don't hold this, only that God is not proven (which many theists agree with, plus of course agnostics). So whether or not they hold a burden of proof depends on how the term has been defined. Inadvertent strawman arguments can be made because of this. Atheist philosophers generally accept the traditional definition and do attempt to prove there is no God with various arguments. Agnosticism, meanwhile, is either the view God is not proven (compatible with one type of atheism and theism) or that it cannot be proven, the latter of which also bears a burden of proof.
- Similar arguments exist for aliens visiting earth, psychic powers, and magic.
- In an example of a Real Life Invoked Trope, Glenn Beck was asked to prove that he did not murder a young girl several years ago. This was nothing more than a parody of his standard operating procedure, where he asked leading questions about a potential conspiratorial coverup or political imbroglio. However, he was so incensed by the query that he drew attention to it via lawsuit. The Streisand Effect resulted.
- The scientific method uses a concept called a "null hypothesis" which is very similar: any scientific experiment will start with the assumption that the hypothesis can be proven false, and will attempt to do so (falsification through experimentation). If the experiment returns no positive proof of the assumption, the default negative is taken to be true. In science a theory is something which has been tested in this manner, while a scientific hypothesis is something which, while consistent with all existing data and violating no fundamental principles, has not been tested in this manner or is not necessitated by any observation.
- One example of this is called "Russell's Teapot", named after philosopher and scientist Bertrand Russell. The idea there is that, somewhere in the solar system, there is a teapot, perfectly shaped and formed, in an elliptical orbit around the sun. Because it is functionally impossible to prove that there is no such teapot, the assumption that one exists is by its nature seemingly logical, despite the unlikelihood. Whether or not this is a functional argument against applied theology is, of course, up to the reader.