Scholarly quality: all ye know and all ye need to know?

In the ongoing conversation about the current and future health of the marketplace for scholarly monographs, there tends to be a lot of discussion around the issue of quality.

This is understandable. Quality is a very comfortable topic for all concerned, since (with relatively few exceptions) scholarly books don’t generally get all the way through the editorial gauntlet of serious publishers unless they’re works of pretty high quality. Publishers, authors, readers, and librarians would all agree on that, I think.

For some commentators, this seems to be pretty much all one needs to know: if the books are of high scholarly quality, then this means that a) they should be published and b) they should be bought—if not by individual readers (the taste and discernment of individuals being notoriously inconsistent) then at least by libraries, whose job, these commentators believe, is to discriminate between high-quality and low-quality books, and to buy the high-quality ones—thus providing essential support to good authors and publishers, building rich and coherent collections, and ensuring that the libraries’ patrons (both current and future) will have access to quality resources as they pursue their scholarly work.

Quality vs. resources

There’s a problem, though. (Isn’t there always?) It’s the problem of limited resources. There may be some academic libraries out there that can afford to buy, shelve, and permanently curate every high-quality scholarly book published, but if there are, then they must be vanishingly few. Most—virtually all—academic libraries have to make tough choices about which high-quality scholarly books they’ll buy and which ones they won’t. This means that they have to make choices based on some variable other than quality. That variable is—and has always been, though we haven’t always said so very explicitly—relevance.

What does “relevance” mean in this context? It’s simple: relevance is the measure of a particular book’s ability to meet real-time, real-life scholarly needs at the purchasing institution. The quality of a particular book doesn’t vary at all from library to library; a great book is a great book, no matter who owns it. The relevance of a particular book, however, varies greatly from place to place. And this is where the conversation gets less comfortable, because while we can pretty much all agree that scholarly monographs tend to be of high quality, it is also very often true that they are quite narrow in focus. And this, in turn, means that if libraries apply relevance as a criterion in their book selecting, they will actively decide against purchasing some scholarly books, despite those books’ high quality. This seemed less problematic to publishers when librarians (whose selections were driven by considerations of both quality and of potential and future relevance) were selecting books; now that books are being increasing purchased as a result of patron behavior (which tends to be driven by actual and present relevance), there is a very real risk that fewer high-quality books are going to sell.

A matrix model

Lately I’ve found myself starting to think in terms of two-dimensional matrices that illustrate the confluence and interaction of multiple variables. (In the November issue of Against the Grain, I’ll publish a two dimensional model of what I call “librarian depth perception.” Some Scholarly Kitchen readers might find it interesting; it will be publicly available and I’ll plug a link in here when it’s published.) Here’s an example of one such matrix that illustrates the interaction of quality and relevance in the context of scholarly book purchasing:

In the past, when libraries were relatively richly funded (and when there were fewer journals, and journal prices were lower) libraries were able to buy a much higher percentage of the books that fell solidly into the high-Q/high-R quadrant of this matrix—and we also had the luxury of buying quite a few books the relevance of which was less clear, but which were nevertheless of high quality. During this period, our purchases were more broadly distributed across two quadrants: up and down the Relevance vector, and around the “High” end of the Quality vector. Today, I think, our purchases would be more tightly clustered around the “High” end of the Relevance vector, though still somewhat concentrated at the “High” end of the Quality vector.

Relevance trumps quality

It’s worth noting, however, that all research libraries contain at least some low-quality books—and this is not only because librarians failed to notice that they were of poor quality. It’s also because a book can hold value to students and researchers that goes beyond its intrinsic scholarly qualities. Poorly-written books filled with bad argumentation and even evil intent (e.g. Mein Kampf) may nevertheless provide an indispensable window on the thinking of important and influential figures in history; books containing lies masquerading as biography (A Million Little Pieces) may be important sources of information on sociological or cultural trends; deliberately fraudulent scholarship (Arming America) may still valuably illustrate strains of thinking and rhetoric in public discourse on important issues. Similarly, the greatest book in the world in a discipline not served by a particular academic library is likely to be a poor purchase, whereas a more mediocre book that supports the institutional mission very directly may be a wise purchase. The bottom line here is that it’s a myth that the library exists to curate and showcase the best in scholarship; in reality, the library exists to facilitate new scholarship—and new scholarship requires access to more than just what is good and true.

What this all means is that no good research library will limit itself strictly to high-quality books. Relevance is not only a restrictive factor on which high-quality books we choose to buy with our limited funds; it’s also an inclusive factor when it comes to considering mediocre or low-quality books.

As I said above, when we turn from the issue of quality (which makes everyone feel good) to the issue of relevance (which makes us more nervous), the conversation becomes less comfortable. Scholarly publishers don’t like the idea of libraries buying mediocre books, because that’s not what they sell; and they don’t like the idea of libraries declining to purchase niche books, because that is very often what they do sell.

Rick’s Law of Potential Relevance

All of this begs an important question, though. How can you really judge the relevance of a book? A book that seems niche-y or only peripherally important today may turn out to be highly relevant (and even prescient) tomorrow—and if not tomorrow, then maybe ten years from now, or twenty.

This is a true statement, but it’s not a good point. Why? Here I will introduce Rick’s Law of Potential Relevance, which states: There is no document about which one cannot say “Someday this document may be essential.” Here’s the corollary to that law: Since potential future relevance is functionally unlimited, potential future relevance is an inappropriate criterion to apply when allocating limited resources.

Here’s another way of thinking about it: the further one looks into the future, the more possible scenarios there are. And the more possible scenarios there are, the more potential there is for any given book to be relevant to one of them. This means that if you try to build a collection based on potential future relevance, you had better have an unlimited book budget and an infinitely expandable library—because that’s the only way you’ll be able to do it.

The importance of being patron-driven

What’s the alternative? The alternative—which is really the only fiscally responsible and sustainable approach—is to build the collection that makes the most sense in terms of real, current, demonstrable needs, recognizing that your collection-building strategy will need to be adjusted over time. And right now, one of the most effective and least wasteful ways to do that is by putting real, current, academically active library patrons in the driver’s seat when it comes to selecting books. Given that this represents a radical departure from centuries-old library practice (and from the assumptions that have guided publishing decisions for centuries as well), we can expect that such a move will continue to be controversial and difficult—for librarians and publishers, anyway. Maybe less so for researchers and students.