Everest Fallacy: a logical error, the confusion of the extreme and the normal.
The Everest Fallacy can best be explained with an example. In the second century BCE, there was a rule in Rome that soldiers were not to serve for more than six campaigns. However, there was a tendency that men had to be under arms more often. Ancient historians often illustrate this development by referring to a speech that was, according to the Roman historian Livy, delivered by one Spurius Ligustinus in 171 BCE.note[History of Rome since its Foundation 42.34] If we are to believe Livy -and there is no reason not to- this warrior had been serving in twenty-two campaigns, and now demanded not to be demoted.
As an illustration, this is impressive, but it is poor logic. The incident has been recorded in Livy's sources because it was exceptional. As a consequence, it can not be used as evidence to prove that soldiers were increasingly often forced to take part in more than six campaigns. You can not use extreme incidents to illustrate standard procedures.
This mistake is often made, for example in descriptions of the city of Rome as a place full of violence, corruption, and sexual licentiousness. This image can be fully based on written sources that are factually true, and yet, the overall picture may be wrong, because these sources are often written by Christian or Pagan moralists, who had reasons to mention the extremes.
This fallacy, which can be qualified as a special case of the fallacy of eliminating qualification, was called "Everest Fallacy" by the British ancient historian Keith Hopkins (1934-2004):
... they fall foul of what we can call the Everest fallacy, that is a tendency to illustrate a category by an example which is exceptional. The exceptional nature of the illustration is not made clear, and the illustration veils rather than reveals the normal. For example, Mount Everest is a "typical" mountain, Cicero is a "typical" new man, M. Aemilius Lepidus becomes a "typical" noble.
It is against this background that we thought it worthwhile to resort to statistics, because summary statistics transform individual fragments of data into usable aggregates.
It is almost impossible to proceed without sooner or later making this error:
- The tragedies of Aeschylus, Sophocles, and Euripides can not be taken as representative of Athenian drama; they have survived because they were considered to be a class of their own. Yet, writing a history of ancient drama without the big three is impossible.
- The Bible can not be used as description of everyday Jewish beliefs; the texts address the concerns of the literate class. Yet, writing a history of Judaism without the Bible is impossible.
- Cicero and Pliny the Younger were not average senators; they were exceptionally gifted orators, whose political carreers may have benefited from their gift. Yet, writing about the Senate without quoting this couple, is impossible.
The principle is well-known to journalists, who will (or should) immediately recognize that in the line "Christians reject Darwin's evolutionary theory" an extreme literalist belief is presented as typical for all Christians. The problem for journalists and historians is the same: our sources -press releases or ancient authors- usually deal with the exceptional and the extreme. The normal, the standard, and the typical need not be recorded. In other words: if a statistical analysis is impossible, we must suspect our written sources, precisely because they were written.
Keith Hopkins, Death and Renewal (1983), pp.41-42.