Labyrinths of Reason by William Poundstone Page A

one popular example:
    H YPOTHESIS : Clairvoyance exists
and
it’s possible because there’s a lot that physicists don’t know about cause and effect.
    E VIDENCE : Bell inequality experiments, which seem to show instantaneous communication between subatomic particles.
    C ONCLUSION : Bell inequality experiments confirm the hypothesis, so they support the existence of clairvoyance!
Ockam’s Razor
    There is an aesthetic to science. The “beauty” of a theory is measured largely by its simplicity. A simple theory that explains a lot is preferred to a complicated theory that explains little—even though, on the face of it, there may be no particular reason to believe that the complicated theory is any less right than the simple one.
    This important principle is called “Ockam’s razor.” The name comes from William of Ockam (the name is also spelled Occam and Ockham), a Franciscan monk born about 1285. (Very similar doctrines were propounded earlier by Duns Scotus and Odo Rigaldus.) A controversial figure embroiled in disputes with popes and anti-popes, Ockam was one of the most influential of medieval thinkers. He died, probably of the plague, in 1349.
    Ockam is best known for something he may never have said:
Entia non sunt multiplicanda sine necessitate
, or “Entities are not to be multiplied beyond necessity.” The sentiment, if not those words, is his. He meant that you should not resort to new assumptions or hypotheses (entities) except when necessary. If a footprint in the snow
might
be explained by a bear, and
might
be explained by a previously undiscovered manlike creature, the bear hypothesis is favored.
    The principle can be misunderstood. It is not a matter of choosing the less sensational explanation. One favors bears over abominable snowmen only when the evidence (such as a half-melted footprint) is so deficient that both the bear and the yeti theory account for it equally well.
    Ockam’s razor is fallible. It has often favored a
wrong
hypothesis. Is the earth round? Do tiny living creatures cause disease? We now know that these hypotheses account for observations very well, but at some point the Ockam’s razor principle rejected them. A notorious case of misplaced skepticism (often cited by proponents of ghosts, UFOs, and other currently unaccepted beliefs) is the French Academy’s prolonged rejection of the reality of meteorites. On the finest scientific advice, dozens of meteorites in European museums were thrown out as superstitious relics.
    Here we come to one of the most troublesome points of confirmation theory. In every scientific discovery, there is a stage where two competing theories account for observations about equally well. There is often a simpler hypothesis, A, which everyone has been believing all along, and a new hypothesis, B, which postulates some new “entity,” in Ockam’s words. Theory A could be the belief that the earth is the center of the universe, and B could be Copernicus’s heliocentric theory. Or to take an example less obviously stacked in favor of B, A could be that there are no UFOs, and B could be that UFOs exist. When does the evidence justify the new entity?
    It is difficult to give a hard-and-fast answer, for we all believe many things on the basis of slight evidence. If you glance at the cover of a tabloid in a supermarket and read that a prominent actress has eloped, you probably take it for a fact. If the same tabloid the next week says in equal-sized print that UFOs abducted a woman in Arizona, you probably don’t believe it. As astronomer Carl Sagan points out, there is a significant yet usually unconscious rule of confirmation at work here: The more outrageous the hypothesis, the more evidence is needed to confirm it.
    The rationale is that a prosaic hypothesis is partially confirmed by all our prior knowledge of similar occurrences. An incredible hypothesis is not. This, however, raises the possibility of being tricked into believing a series of wrong