Patron-Driven Acquisition: Rick Anderson Answers Your Questions

Due to the large number of questions that emerged from the first session of Rick Anderson's workshop Patron-Driven Acquisition: Radically Re-Thinking the Collection, we have posted Rick's responses as a separate blog post. Rick's answers are below--feel free to chime in via the comments area and join in the discussion!

What’s your definition of Patron-Driven Acquisition?

Fundamentally, it's any system whereby documents are acquired by the library in response to patrons' direct requests or selections, rather than in response to librarians' speculations about which specific documents patrons are going to need. At a basic level, this kind of acquisition has existed for as long as patrons have had "Request a Book" options. But in the current environment, PDA usually refers to much more ambitious and comprehensive programs that make the patron a central figure in the selection process.

I've been moving away from the term "Patron-Driven Acquisition" in favor of the term "Patron-Driven Access," because I'm becoming less convinced that permanent acquisition of materials by the library is always the right outcome of a PDA transaction. In many cases, the transaction might result in short-term access, or even in a personal copy for the patron. This, of course, calls into question the whole foundation of traditional librarianship, which is why I spent so much time discussing the history and philosophy of library collections during the first part of the workshop.

The presentation presumes that print is dead, which is debatable. The Gutenberg led to mass -produced books, therefore allowing multiple simultaneous users. Could you elaborate or give your perspective on the difference between book as abstract text and book as object?

My presentation does not actually presume that print is dead. What I emphasized was that print is a very bad format for distribution and research—it remains a very good format for extended linear reading, but that kind of use is only one of many that documents get in a research setting.  I also have to disagree that Gutenberg technology allows for multiple simultaneous users. It allows for multiple copies, but each copy can only realistically be used by one person at a time—and what a library almost always owns is a single copy.

My perspective on the difference between book as abstract text and book as object would be that in the general, circulating collection of a research library, the book-as-object matters not at all. It matters only as a container of intellectual content, and the important question is not "How do we make this physical object available?" but rather "How do we make this content available?" In some cases, the best way to make the content available will be by providing access to a physical object. But I would argue that in a research library, the best way to provide access is usually to make the content available online rather than to put a physical copy on a shelf. (There are exceptions, of course, but I'm proposing a general principle.) Some books, obviously, are valuable in and of themselves as physical artifacts. For the most part, those books do not belong in a circulating collection; they belong in Special Collections.

If librarians aren’t good at selecting, who is? Selection by librarians saves the time of the user. People want the best sources, but often settle for the easiest to find. Librarians’ selection skills give users accurate information. Regarding the concept of “at least one use,”  do we have research on whether what patrons acquire is what they need?

Being "good at selecting" means knowing what it is that patrons actually need in order to accomplish their work. Patrons don't know perfectly what they need, of course—but they know much better than librarians do. Selection by librarians only saves the time of the user if the librarian actually succeeds at guessing what the user will need, and all available research suggests that we do so with a very high failure rate (roughly 40%). I'm not sure exactly what the sentence "Librarians' selection skills give users accurate information" means—is it trying to say that librarians are better at discriminating between high-quality (i.e. more accurate) and low-quality sources? Or does it mean that when librarians select resources, we also provide accurate information about those sources? In any case, I'm not sure the assertion stands up to examination. For one thing, especially in a research library, our patrons often know the literature better than we do; for another, it's important to remember that useful documents are not always the "best" ones. For example, the Michael Bellesiles book "Arming America" is a very bad book—it's full of misrepresented data, fabricated sources, and fallacious arguments. So as a source for understanding the history of gun ownership in America, it's a very poor resource. However, as a resource for understanding the national conversation about gun ownership and gun control in modern America, it's an absolutely essential document. If you believe that the purpose of a library is to provide only high-quality resources, you might be tempted to withdraw that book. But if you believe the purpose of a library is to support the scholarly work of its patrons, you might be less tempted to do so. What all of this points up is the tremendous complexity of the idea of scholarly quality, and therefore of the librarian's role as an assessor of quality on the patron's behalf.

It seems there is a tension between libraries roles in archiving and giving access.For large research libraries, collecting for current users has to go hand-in-hand with preserving the scholarly record for unanticipated future uses. Circulation statistics can’t capture that role, yet it’s important.

Yes, the tension between archiving and access is real. In the print environment, the tension is primarily one of access versus control: you always pay for access with control (i.e. archival permanence), because the more access you grant to the physical item, the more its integrity is threatened; and you always pay for control with access, because the more control you impose on an item the harder it is for people to use it. In the online environment, where access doesn't tend to threaten directly the integrity of the document, access vs. archiving is more of a financial issue: how do we provide all the access our patrons need now while at the same time providing for future access to things they don't yet realize they need?

The answer, of course, is that we are only equipped to do one of those things well. We are in a position to meet current needs fairly well; we are in a very bad position to meet all possible future needs—partly because we can't possibly know what those needs will be, and partly because we can't afford to meet all of them even if we knew what they were going to be. But this is nothing new. No individual library has ever provided for all future needs, and even in the aggregate we have only done so marginally well.

Are libraries using PDA with media other than books and articles?

Not that I'm aware of—we haven't gotten there yet! It's still a new enough idea that we're all still wrestling with the implications for books and articles.

What are the considerations for the library's systems department in implementing PDA?

Good question. The implications can be considerable, depending on the PDA model used, but they don't have to be. I'll discuss this more during the second part of the workshop.