I didn’t attend the panel about “Is Selection Dead? The Rise of Collection Management and the Twilight of Selection” at ALA Midwinter, but fortunately my head of collection development did and pointed me to the recent summary at Against the Grain (he provided “Response #2). The panel discussed the role of patron-driven acquisition (PDA) in academic libraries, with at least two of the panelists, from the Universities of Utah and Arizona, advocating a greater role for PDA in libraries. One said, “This isn’t to say that all libraries should immediately stop building traditional collections, only that we should be willing to rethink the universal appropriateness of such collecting, and willing to experiment (even aggressively) with new models.” If I understood correctly, and I may not have, the person at Utah is experimenting aggressively enough to have gotten rid of several subject specialists and relied increasingly on PDA. That’s aggressive, indeed.

Before even addressing the collection issues, we should note the importance of subject specialists in reference and instruction. One assumption seems to be that if libraries adopt more PDA that subject specialists aren’t necessary. We don’t need experts to choose appropriate material, because the only expert is the current patron. For most libraries, the days of musty bibliographers sitting in back offices selecting books is long gone. Subject specialists don’t just develop and manage collections; they also teach people how to use them and provide help for the difficult and high level research questions. Maybe there just wouldn’t be much high level research skills to teach anymore. Knowing how to search Google would be just as good as actually knowing something about a scholarly field. Reference as Yahoo Answers.

I have no problem at all with selective PDA, and think it would be a great way to explore possible gaps in a collection, but what would happen if it were taken to its logical, if also absurd, conclusion. Here i want to consider what would be the most aggressive form of PDA, and argue that it would ultimately eliminate any claim for a library to be a research library. To do this, I need to explain what I mean by the most aggressive form of PDA, which I will call Radical PDA (RPDA). I’ll also need to explain what I think research libraries are.

RPDA would eliminate all librarian input into the library collection. Approval plans and librarian-driven firm orders would disappear. Nothing would be purchased or licensed that hadn’t been directly and immediately wanted by a patron who was currently associated with the university in question. The entire library budget would go to support RPDA. There would have to be some form of rationing or allocation so that the collecting would continue year round. Perhaps the money could be allocated by week or by month, and when that money was gone, collection would stop until the next week or month. It would have to be rigged so that all the money wasn’t spent in the first couple of months or so. Subject specialists might be gone, but libraries still need people to deal with acquisitions and licensing and such, and you can’t just leave those people with nothing to do from October to July, unless you wanted to make them temp workers, which I suppose would the cheaper way to approach this. A university library could try this experiment to see what it’s effect would be.

I have a hypothesis, which could be confirmed or refuted by about 20 years of such a practice. That seems long enough to see the results, and research libraries have to think about the long term. If a library starts the experiment, I’ll probably still be working and can write about the results on whatever long-form writing platform exists in 20 years. A library using RPDA as its collection model would have an esoteric collection to say the least. Instead of any systematic collection in any area, let’s call them seas of information about a given subject, it would have streams and rivulets disconnected from each other, and connected only to a research project someone happened to have been working on in the past. It would be as if the library purchased only what was contained in the bibliographies to the articles and books that people happened to be working on. It would be eclectic beyond even that, though. Imagine a library collection driven by what 18-22 year olds are interested in. I don’t really know what they’re interested in, but I know it’s not usually scholarly research. The library would probably have a lot of books about sports. A lot of money would be wasted on material of little value, because patrons have no interest in spending money wisely. They just want stuff. It probably wouldn’t have much of a reference collection. A lot of useful and used databases and indexes wouldn’t be purchased, because people wouldn’t know they exist. They are usually tools librarians use to help people.

This would be okay, say the proponents of PDA, because a given collection doesn’t matter. Anything that we don’t have, we can just buy or borrow when it’s requested, after someone discovers its existence through Google or Amazon or Wikipedia. It’s true that, assuming it has the budget, a library can buy books and other materials, and a great deal is readily available, but will it always be? It’s refreshing to see the faith of some librarians in the future easy accessibility of material being published now. There are all sorts of scholarly books now that are hard to come by, and are available to some libraries because some other libraries purchased them. There’s also all the primary materials libraries collect that won’t be available in the future. Research libraries collect for the future, not just the present. The assumption that everything will be online and accessible is a shaky one to build your future on, but it’s the assumption of RPDA. Don’t get it now, because we can’t afford it. But we can always get it later. What if we can’t? Research libraries also have to consider other parts of the world, where the digital present is nonexistent and the digital future not so rosy. Imagine how useless an area studies collection would be after 20 years if it was patron-driven.

No problem, right? There’s always ILL! After 20 years of this, the library collection would be such a hodgepodge that systematic research would be impossible. Scholars wanting older material that hadn’t been purchased because 20 years before nobody happened to be researching that topic will begin turning to ILL in droves, assuming that with DRM ILL is even still possible. But would they be allowed to? What if other libraries started noticing that the RPDA library was borrowing extensively, but not loaning much, that it was free riding on the more systematic collections of libraries that were formerly its peers? They might stop loaning to that library. If enough libraries adopted RPDA, the collections would be so eclectic that ILL would be useless anyway. The PDA agenda seems also to be driven by something called webscale discovery, but such discovery in scholarship is only possible because so many libraries build research collections. You can’t discover what was never collected.

Research libraries as a group attempt to collect as much as possible of the human record in every format possible. No single research library, no matter how wealthy, can afford to buy everything, but the hope is that with some cooperation and some luck, the system as a whole will provide scholars with the support they need. Even scholars at colleges and non-research universities benefit from this endeavor because of the robust system of ILL that most academic libraries participate in, and the archives available both in the library and online. The goal of a research university is the creation and discovery of new knowledge. That goal has trickled down into the faculty of small colleges around the country that can’t possibly support it, but that’s okay because they can depend on research libraries for help. Small colleges aren’t free riders because they’re not pretending to be research universities and they still build coherent collections for their users, and research libraries help because research is an international exercise with every library playing its smaller or larger part.

One of the complaints PDA enthusiasts have about research libraries is that not enough of the books are used. Libraries buy books that are never used, and they get dusty, and then we have to take care of them forever, etc. They think the way to make sure every book is used is only to buy books of immediate use. That’s probably true, though it would also lead to the buying of a lot of books that might be used only once, books that before might have been borrowed through ILL. I’m assuming this is mostly an economic argument, as in, a library just can’t afford to collect systematically anymore, so it’s gambling on desperate measures. Sometimes I wonder if the argument would be there even if the money was, too. One speaker argued, “Thirty years ago it was easy to justify buying a book just in case someone might want it in the future — but what is our justification for doing so now? The purpose of a collection is not to be a wonderful collection; the purpose of a collection is to meet the information needs of library users.” I would agree completely with that last bit, but I’d interpret it differently. Who are the users of a research library? Only the people currently on campus? What about scholars at nearby colleges who depend on public research libraries, but who won’t have the benefit of any PDA? Or scholars at other universities whose library also didn’t collect systematically and now has enormous gaps? What about users of the future? Shouldn’t they have a say? Can we poll the newborn to make sure we can guarantee that their scholarly needs 40 years hence will be met?

I might disagree that one purpose of a collection isn’t to be a wonderful collection. That’s exactly what it should be, and the more of them we have the better chance of covering everything. All the “just in time, not just in case” reasoning is fine if you’re not trying to support a research university. The complaint might then be that libraries just shouldn’t be buying stuff that will get little or no use, period. It’s more a moral complaint against “useless” research, similar to the claims addressed in this CHE article (subscription required). The author responds to those who think there should be less scholarly publishing because so much of the scholarship is bad, or useless, or never used. Only the good stuff should be published, and for librarians with this mindset, only the good stuff should be purchased. The problem presented by the author is that one can’t always tell the good stuff from the bad without time passing. We might measure the impact of journal articles a decade after their publication and see that 90% of them were never cited, but at the time of publication nobody would have been able to say which 90% of them would go into scholarly oblivion and which 10% would be the most cited. In the humanities, the time scale would have to be increased significantly. How much history has been lost because it seemed too ephemeral at the time for libraries to collect? How many authors have come into style that can be studied now only because some research libraries in the past had the foresight to collect for the future?

We have to save as much as we can now, because we don’t know what will be of interest in half a century or more and we can’t guarantee that we’ll be able to acquire it then. That’s why research libraries have traditionally done what must be anathema to the PDAcolytes, collect to strength. Research libraries collect to strength not because scholars who currently happen to be using one library will use that collection, but because research is an international endeavor all research libraries participate in, and the hope, often fulfilled, is that someone, somewhere, someday, will want to use it. Libraries that don’t participate in this communal effort are no longer research libraries. It’s one thing to say, we just can’t afford to be a research library anymore, but another if anyone wants to make a virtue of this necessity. I know some readers might be thinking that this is easy for me to say, working for a relatively well funded private university library. However, I don’t think like this because I work at Princeton; I work at Princeton because I think like this. I also know how much scholars at all universities, including my own, benefit from having a network of research libraries that think of the future and not just the immediate present. Without those libraries, the future of research would be bleak indeed.