[bloak]

It depends on the book. In the case of Kafka's "The trial", for example, the manuscript is "the original" because the author never finished the work and all editions were made after his death based on that manuscript. However, with most books what happened is that the author sent the manuscript to the printer/publisher and there was then a bit of back-and-forth during which changes suggested by proofreaders were accepted or rejected by the author, other last-minute changes were made by the author, and so on, so in that case the manuscript is best thought of as "a draft" rather than "the original".

Usually the last edition produced during the author's life and under the author's supervision gives you the most reliable indication of the author's intention, what they call the "copy-text" in textual criticism, I think, but there are some interesting exceptions. For example, many people think the first edition of Mary Shellely's "Frankenstein" is better than a later edition, in which the older author appears to have wanted to make certain passages more respectable, less shocking, and thus arguably spoilt the work she produced in her youth. But that's presumably a matter of opinion. For scientific textual criticism we need all the editions, all the manuscripts, any manually corrected proofs or published copies that the author has written on, ...

[tialaramex]

The existence of distinct variant works is mildly interesting, and so there's (diminishing with volume) value in preserving variation, Big Wave (神奈川沖浪裏)† is a woodblock print, like many modern low scale printing processes for art it produced definitely different items each time, if you've seen Big Wave in a local city gallery and can see another somewhere they're certainly not just the same thing, the way every student's identical print of a famous band poster is.

But on the other hand, most of the value is in the core thing, not the variation. Modulo special cases like the "Wicked Bible" (which has a misprinting in the Ten Commandments) we care about that and not the variation.

It's easy to say that Blade Runner is an important movie. Hard up for preservation capacity if we're choosing obviously Blade Runner beats Ghostbusters 2. Easy. Ghostbusters 2 wasn't a terrible movie but it's no Blade Runner. OK, how about the "Director's Cut"? Sure OK, two copies of Blade Runner. The "Final Cut"? Ugh. Fine, OK, let's have three versions of Blade Runner rather than kick one out for a sub-par sequel to Ghostbusters. There are four more commercially released versions of the movie. That's too many. At some point we should cut our losses and say no, another Blade Runner is not worth it, we'll have Ghostbusters 2 instead thanks.

† That's deliberately not an accurate translation, but we all know what I'm talking about.