I’m all for mass digitization. The more information searchable the better, as far as I’m concerned. However, I suspect that if there ever is total digitization of print collections it will be decades and perhaps centuries in the future. More likely, I don’t think it will happen because of the sheer mass of print-only material already extant and increasing every year around the world.

This raises several issues. First, will it increasingly be the case that “everything is online”? We know that de jure this isn’t the case now, but de facto for many library users it is true, because whatever’s not online (or perhaps just very easily accessible) won’t be used. For the mass of information seekers, this doesn’t matter much, but for scholars at any level past novice it does matter. Will students become ever more reluctant to seek out the difficult data or the remote but appropriate archive? And if they do, in how many generations will we have an Idiocracy of scholarship?

Perhaps more important is the question of what obligation do librarians have to collect the human record past the popular mass of material? It’s becoming increasingly clear to me that electronic collections are mass collections. For most colleges and universities, this is a good thing. The rise of JSTOR and other large aggregate journal packages means that plenty of academic libraries that never had much of a collection can now have easy access to the benefits of this mass digitization. Even at Princeton I find that these journal collections sometimes provide easy access to journals in areas that were never collected, but that can be useful for some kinds of interdisciplinary research here. But still, the electronic collections for sale are still mass collections driven by the rule that if everyone doesn’t want it, nobody gets it. “Everyone” might be just a handful of the richest institutions, but the rule still stands. Commercial products are created to make money. Only so much ephemeral material can be included in a commercially viable package. Other digitization projects change this somewhat, but it’s not clear how Google Books or the Million Books Project or independent digitization projects will affect this. Even a combination of the largest of these projects contains only a portion of available material.

But what about those materials that aren’t digitized and never will be? Increasingly they will become niche items, even more so than they are now. Who will collect these? Who will make sure they’re available to the scholars of the future? In most libraries, there’s the assumption that the library can rely on some other library to collect the stuff that won’t be used often or may be of peripheral interest at most. In larger research libraries, especially at the richer private universities and the largest state universities, the librarians often have a different view, though. Some library has to be the library of last resort, and some librarians at the larger libraries understandably think, if we don’t collect this, who will? All libraries promote services and access to materials through borrowing, but ultimately some libraries actually have to buy those materials, and richer libraries have a obligation to collect the human record in a way that smaller and lesser funded libraries don’t, and a corollary obligation to make those materials known and available.

This obligation is sometimes difficult to justify, though oddly enough it fits in with the Princeton motto, “in the service of the nation and in the service of all nations.” (I’m not implying by this Princeton is in any unique position, since it’s not even in the top ten libraries by size in the country.) Though whatever means, individual or consortial, large research libraries have an obligation to collect everything they can. Consortial collection arrangements are tricky things, but ultimately they be necessary to avoid never having material that now seems ephemeral but may one day be important for scholarly research, if it’s available.

The phrase noblesse oblige seems elitist (and we know how some people bristle at any thing that seems “elite”), but in some respects it’s the only appropriate phrase. The largest and richest libraries should have a sense of noblesse oblige, a sense that ultimately they have a greater obligation to collect the human record. For a lot of people, including a lot of librarians, this mission doesn’t seem very important compared to other goals, but for historical and scholarly purposes it’s crucial. This ignores issues of digital divides, of rich and poor, of access to even minimal information, but it’s just as, if not more important for the future of human knowledge and understanding.