Many people start with the idea that an autograph is authentic and look further only if they are suspicious. This is backwards. You must begin with the premise that an autograph is not authentic and make it prove itself. In assessing this proof, we disregard representations of sellers and ignore sales pitches. Only an autograph that can prove itself outside seller claims can be said to meet this burden.
As for provenance: Look at the first six letters; they spell “proven.” Provenance means just that – proving where the autograph came from. And as strange as it may seem, this is very often the most important factor in authenticating an autograph. Knowing that a letter came from a reliable collection or dealer, and that it has an identifiable history, can be crucial to establishing a level of comfort. Not only have the pieces been vetted by them but were of sufficient interest to attract these giants of yesteryear to acquire them in the first place. If a piece comes from a famous collection of the past, such as the Philip Sang, Malcomb Forbes or Jerome Kern collections, all the better. As an illustration of how this can matter, some time back I bought a very rare document in the hand of the Puritan apostle to the Indians, John Eliot. It was dated 1645 and was docketed as approved and signed on the front by John Winthrop. The paper and ink were right and the writing checked out, but the incredibly early date and unique combination of signatures caused someone to question the document. When research showed that it came from the Sang collection, which was put together over half a century ago and carefully authenticated by the greatest experts of the time, everyone took that as sufficient evidence that it was authentic and the debate ended then and there.
But not every autograph can come from an important collection, so it is necessary to establish a more generally applicable rule for what provenance is sufficient. The existence of a letter or notation claiming to describe the circumstances under which an autograph was obtained is just a claim, not a provenance. So a statement that an autograph comes from the noted Joe Schmo collection is just a claim, absent tangible, independently verifiable evidence that: a) such a collection existed; b) there is evidence of its reliability; and c) this item was part of it. Every forgery seems to come with a story, and forgers use the techniques of providing cooked up “provenance” letters and inventing collections to deceive the unwary.
Explanatory letters that come with autographs are unreliable because, even when the writers are honest, they can make innocent but very misleading errors. I once had a situation where a girl at a 1963 Beatles concert gave a program to their road manager to have autographed. She observed the Beatles enter their dressing room, saw him take it in and then return with it signed ten minutes later. Her long and honest letter about her attendance at the concert and getting the program signed left out what she never knew – that their road manager, Neil Aspinall, had expertly forged the Beatles’ signatures behind the closed dressing room door.
Provenance is less significant in items that have a lot of handwriting and/or would be difficult or very challenging to forge (such as an ALS of Andrew Jackson).