The terms, the line between them, and the famous cases with the numbers checked.
Most of the headline figures do not survive contact with the evidence, and that turns out to be the interesting part.
There is no committee handing out these titles, so the definitions below are the ones in common use rather than anything official. Where a threshold is disputed, it says so.
None of these say anything about level. Someone with four languages at B1 and someone with four at C2 both get called a polyglot, and the word hides the difference.
Every list of famous polyglots gives different figures for the same people, and the reason is not sloppiness. Nobody has agreed what counting means.
This site takes the same view of its own numbers. The coverage figures on every language page come from a named corpus, and where a well-known claim does not survive checking it gets marked rather than repeated.
Ordered by how well the claim holds, not by the size of the number.
Interpreter and translator, and the most useful of the lot because she wrote down her method instead of being mysterious about it. She learned largely by reading novels she barely understood and pushing through, and insisted she had no special gift. Her line worth keeping: language learning is the one thing worth doing badly.
One of the few whose ability is on video in most of the languages he claims, which is a low bar that surprisingly few clear. He has studied more than fifty at some point and is careful about the difference between studied and speaks.
Learned all of them without living abroad, which quietly dismantles the most common excuse. Adds roughly one every two years and is explicit that maintenance, not acquisition, is the real cost.
Notable for refusing to give a hard number, which is itself the mark of someone counting honestly. Attributes it to thousands of logged hours rather than talent, and keeps the log.
A diplomat who reportedly worked in 68 languages and studied over a hundred more, often learning one through another rather than through German. His brain was preserved after death and studied for structural differences. The figure comes from his own era and was never tested by anyone applying a standard.
A cardinal credited with as many as a hundred, though roughly half of those were dialects of languages already on the list. Michael Erard went to Bologna to examine the archive for his book on the subject and came back unable to settle it either way.
Listed by Guinness until 1998 as speaking 58 languages. On live Chilean television he was asked basic questions by native speakers in several of them and could not answer. He may well have known them once and lost them, which is a real phenomenon, but the claim as stated did not survive the test.
The pattern is hard to miss. The people whose numbers hold up are the ones who quote the smallest figures and describe their method in the most detail. Everyone claiming a number above forty is being counted by somebody else.
Nothing on this page is a target. The useful finding buried in it is that the verifiable polyglots all did the same unremarkable things: they read far above their level, they counted their hours, and they accepted that maintaining a language costs more over a lifetime than learning it did.