Wikipedia talk:Reliable sources/Cost
Isn't this essay conflating verifiability and reliability?
Moreover, does not some aggregate cost impact usability? An extreme example: an obscure work, in a foreign language, with incomplete information on publication. Perhaps I'm simply suggesting that the more obscure and potentially more difficult a source is to access, the more incumbent it should be on the editor to provide complete publication information. But ultimately, what happens when even a source's existence is almost impossible to verify? It would seem to me that there's a lot of ground between not requiring that any particular person at any given moment be capable of verifying a source and requiring only that some person somewhere possibly be able to verify a source. Assuming that the source actually exists and supports a statement that it is purported to support. I'm not suggesting that faking references is common practice on Wikipedia, but I've encountered a few editors who have offered as support sources that they clearly had not personally consulted.
A cost-related issue is the lack of a page number. Though it appears to me that WP policy does not require one, for practical purposes, a lengthy source without a page reference is essentially unverifiable to anyone who hasn't memorized that source. Moreover, without a page number, it's difficult to reasonably assume that the editor correctly understood (or perhaps remembered) the source material (or in some cases, even consulted it).
Extremely difficult to access works
Imagine a tablet sunk to the bottom of a trench in the Pacific Rim. Only accessible by a skilled and well-moneyed expedition of mariners. No other copies. Prior to sinking, the tablet was read and used as a source by several well-known Wikipedia editors. Still verifiable? --Atethnekos (Discussion, Contributions) 22:07, 11 July 2013 (UTC)