Earlier in the week I wrote a piece about a frustrating blog post on the Guardian books blog by the rare books dealer and writer Rick Gekoski, which I argued was complacent and fell into a narrow view of how complicated published texts can be. In spite of my frustration, I still like the guy, and picked up his book Tolkien’s Gown (2004), which, though I haven’t had time to read much of it, has some great stories about rare books and the people who buy and sell them. The story about Gekoski getting drunk with Graham Greene in the process of purchasing his copy of Lolita, inscribed to Greene by Nabokov, and then, hungover, selling the book along to Bernie Taupin – Elton John’s writing partner – is pretty great, and made funnier by the somewhat bizarre constellation of cultural figures who come together in it.

But still, in the context of his article that questioned the use of literary archives and early revisions to texts, this passage in the Introduction to Tolkien’s Gown gave rise to further frustration:

…In a lifetime spent with books, what has often fascinated me is how easily even the most famous of books could have been other than it is. Authors rewrite compulsively, and are never sure when a manuscript has found its final form. Editors and publishers, even friends, often have an immeasurable impact on the final form of a text. So a published book, frequently, is a collaborative effort, to which only the name of the author is eventually attached. Books have biographies, and the study of their geneses and later lives is frequently instructive.

Minds change, contradictory views are held over time or even simultaneously, but still the inconsistency can be frustrating.

(The quotation is from p. xiii of Tolkien’s Gown, London: Constable & Robinson Ltd, 2004).