Comment Re:Does Digital _Really_ Last? (Score 1) 480
The historical method for preventing "decay" in documents has not been, though people often thought it would be, flawless copying. Instead, the best method has been cheap, numerous copying. For instance, originally works were copied by hand. The potential for mistakes with this method is high, and so the copyists had to be extremely dedicated and careful about their tasks. But mistakes did happen, and often went unnoticed because any given copy was likely the only copy in a very wide area, and few people could afford to be knowledgeable of single works, much less cross-reference them.
With the advent of comparatively cheap printing presses, the possibility of copyist mistake was not at all lessened--witness the "sinful Bible," a run of Bibles which, among other things, contained the erroneous Commandment "thou shalt commit adultery." If you were to look at the printing press with the intent of demanding as much precision as possible from copies, then it would fail miserably--the chances of mistakes were not only increased by semi-educated printers, but exacerbated by the massive printing runs which would propogate errors.
Of course, time has shown that the opposite is true; or at least, that the above is true but unimportant. The printing press' contribution to accuracy was not through inherent precision, but in the propogation of inexpensive copies. Errors--inevitable with any human effort--are much easier to find and correct when multiple sources, print editions, and publishers are available. Works can be compared and cross-referenced not only easily, but more importantly, cheaply, due to the radical drop in cost of acquiring collections of works.
That's where digital media can contribute to accuracy--not from their (admittedly superior) capability for precision, but because they (can) reduce the cost and labor involved in copying and distribution almost to nil. Copyist accuracy is reached through the ubiquity and ease of reference of works, not through the precision of the copy method.