In the first chapter of Library: An Unquiet History, Matthew Battles describes the vertigo that often attends initiates of the large research library. Even at much smaller institutions, it is jarring to consider the breadth of a collection of books and the extent of what else is out there, not yet and probably never to be acquired (much less read). The idea of a universal library often animates these concerns, and while such an ideal seemed to have been losing currency and increasingly identified with a bygone Enlightenment mentality, my sense is that it has been reasserting itself lately, primarily as a result of digital repositories, open access, and other means of providing a lot of information to a lot of people. Feeding into this situation is the unending stream of publication. Yesterday I was looking through the 2010 religion/theology subject catalogs of Ashgate and Cambridge, which can instigate the same overwhelming feeling of inadequacy-- and this is only one field of study, from only two academic publishers, publishing only English-language works.
Any researcher is forced to confront a lot of material and to live up to it in some sense- the peer-reviewer will point out a reference that should have been included, and the researcher will either need to include it or appeal to the inevitable limitations of scope that justify ignoring it. Anyone working at a library, however, is forced to confront a lot of material and simply cope with it as best as possible. There is no way to "live up to it" because, unlike the researcher's task of merely synthesizing material in a written project, the library is a place that collects these materials for use. It has to maintain the bibliographic manifold, rather than simply make use of it. It's the difference between Atlas himself and a mere cartographer.
While it can be instructive and humbling, this sort of nauseating reception of bibliographic sprawl strikes me as something that is juvenile and should eventually be overcome. At a certain point we need to come to grips with the incompleteness that accompanies any extensive literature, and we need to make ourselves at home as best we can. There's a certain virtue in provincialism that should be pursued in libraries, from the personal collection to the national library.
Of course, a library is always already provincial. To take one example (that won't apply to everyone), if your library doesn't have substantial collections written in non-roman alphabets, then it's quite regional in scope. What is more important than simply being provincial, therefore, is the recognition of the provincial scope that is already present. And in this recognition itself, we gain a more cosmopolitan glimpse of our own boundaries and context. When our particularity becomes explicit we can form reasons for living with it in good conscience. We no longer harbor totalizing conceptions of things based on our limited perspective. Because in the end, the universal library could hardly be perceived even if it were recognized as an ideal rather than a concrete possibility. We simply wouldn't possess the categories to express something like that.
We would do better to work within our means and pursue smaller intentions along those lines. We should stop dreaming about a universal library, and build a lot of small ones with more cooperative mentalities. We should welcome the unevenness of mixed paper and digital collections and various temperamental databases whose navigation is somewhat counter-intuitive. My sense is that these sorts of frictions are what develop cultures, and that the monotonous efficiency of a single complete storehouse of reference is what discourages them.