I’m going to suggest something that may seem kind of crazy on its face: that research libraries need to worry less about the objective quality of the books they acquire, and instead focus on another property, at once more concrete and more elusive: utility. The obvious and reasonable retort to such a suggestion would be “How useful can a low-quality book be?” Read on.
Let’s start by considering two books, both of which, one could reasonably argue, are objectively bad, though each in a different way. One is a book of popular punditry, the other a scholarly monograph.
The first is “Slander: Liberal Lies About the American Right,” by political commentator Ann Coulter. This book asserts that liberals “have been wrong about everything in the last century” and that conservatives “are the most tolerant (and long-suffering) people in the world.” According to Coulter, every — that’s every — “pernicious idea to come down the pike is instantly embraced by liberals.” On the other hand, all — that’s all — new ideas that have emerged in the last 20 years have “bubbled up from the right wing.” (Thank you, right wing, for the iPod, the tablet computer, and the open access movement.) Coulter also informs us that liberals, by definition, regard ethical principle as “nothing; winning is everything.” Her book teems with these kinds of breathtaking howlers, along with breezy factual misstatements, moral analogies that can charitably be characterized as bizarre, and gratuitous personal attacks on individuals with whose politics she disagrees. Even those who share her political views have pointed out the fundamental flaws in this book. One does not have to disagree with her politics to find “Slander” bad by any number of objective standards.
The second example is “Arming America: The Origins of a National Gun Culture,” by Michael Bellesiles. This book presents an argument as to the nature and extent of gun ownership in 17th– and 18th-century America. Bellesiles’ thesis is that a broad gun culture did not develop in the United States until after the Civil War; before then, gun ownership was relatively rare. The book was received warmly, particularly by advocates for stricter gun control, and Bellesiles was awarded the Bancroft Prize in 2001. However, significant elements of his research fell apart under critical examination. Nor were the book’s problems simply a matter of poorly formulated arguments or irresponsibly selective data: some of the documents Bellesiles cited did not exist, he misquoted critical documents that did exist, and his math was in several important instances inexcusably wrong. In the wake of these revelations the prize was revoked, Bellesiles was formally investigated by Emory University, where he was a full professor (he resigned), and he was called a “fraud,” “a liar,” and “a disgrace” by prominent reviewers who had previously championed his work.
So here’s the question:
Should a research library acquire these books?
And in the case of the Bellesiles book, having acquired it before its dishonesty was uncovered, should a research library keep it or withdraw it?
The answer is going to depend on what one believes is the fundamental purpose of a research library. If the library’s fundamental purpose is to offer its users a well-crafted repository of the world’s best thinking and research on topics of interest to its users, then you could make a very strong case that neither of these books should be included. Ann Coulter’s book, though it may offer some accurate critiques and maybe even a perceptive insight or two, is so padded with know-nothing bloviation and irresponsible name-calling that it completely undermines its own authority as a citable source. Bellesiles’ book also probably contains some good and useful information; unfortunately, there’s so much deliberate fabrication and misrepresentation in the book that it can’t be trusted as an independent source either. If the library exists to give people access to high-quality documents, then neither of these books belongs in a library.
However, I’m here to argue that it’s not the purpose of a library to give people access to high-quality documents, but rather that the purpose of the library is to give people access to useful documents — or, more to the point, to connect its users with the documents they need in order to do their work. And there is all kinds of very good work that can be supported by “Slander” and “Arming America,” as well as by other equally terrible books and articles that are bad for a whole variety of reasons. “Slander” may be a really lousy source of information about the moral and intellectual standing of the American left, but it’s an excellent example of a particular type of political discourse that has emerged in the late 20th and early 21st century. The roots of this discourse can be traced very far back in our nation’s history (and before), and there is a fascinating thread that can be traced from, for example, colonial American broadsides and cartoons up through the nativist movements of the 19th century, the early radio demagoguery of the 20th century, and the print and radio fulminations of figures like Coulter, Glenn Beck, and Rush Limbaugh today. I don’t see how a library could support research into modern political discourse without offering patrons access to books like “Slander.”
Similarly, “Arming America” is a terrible source for accurate information about gun ownership in early America. But it seems to me that it’s an indispensable resource for a researcher seeking to understand the rhetoric of gun control in the America of the 20th century. Terrible as it is, it’s a book that had a huge impact and significantly affected the conversation about gun ownership in this country, and to understand what has happened with that conversation in the years since, one would need access to that book (among many others, of course). “Arming America” is also potentially useful as a cautionary example for undergraduate journalism and history students: I can imagine a professor profitably assigning students to examine the arguments made in the book, chase down some of the mis-cited sources, and explain what they find. How did this author twist and misrepresent his findings? How would you have characterized the sources if you were writing this book? What are your legitimate options when faced with sources that directly contradict your thesis or your assumptions?
In other words, “Slander” and “Arming America” are not useful despite their badness; they’re useful precisely because of their badness, because their badness is of a particularly instructive type. None of this is to say that quality doesn’t matter or that it’s irrelevant to researchers — it’s to say that quality is only one of many factors that can make a book useful to people doing scholarly work, and that other factors may well outweigh it.
This, of course, begs a serious question: If you accept my position on the potential usefulness of bad books, what does that imply for the role of librarians? That will be the subject of my next post.
23 Thoughts on "No Such Thing As a Bad Book? Rethinking "Quality" in the Research Library"
Another interesting question for librarians, scholars, and members of the general public: What is the value (or usefulness) of quality utterances in this age of truthiness, info (and opinion) glut, and a good-enough-to-get-by attitude (unless we get caught or scrutinized)? What is quality, and what is its role in culture and society? It’s enough to make me want to go re-read Plato and Pirsig.
I heartily agree. The best way to appreciate the emptiness of the argument is to read the argument as present by its proponents. An intelligent person should ask “What it is that they know, or think they know, that I’m not getting?” For example, the bible is as an indispensable book for atheists (for actual reading) as it is for bible thumpers (who don’t read it). Reading the christian fundamentalist books about evolution is an eye-opener.
Reading both sides of the issue promotes critical thinking — which is the best argument for “Teaching The Controversy” in schools.
Rick, you talk about books as if they exist on two abstract axes –quality and utility– and then argue that librarians should be focused on the latter.
Would it be better to state that librarians should select and acquire content that is important, understanding that this notion of importance combines both quality and utility?
In that way, it leaves the decision of whether to acquire something a little more amorphous than setting up a dichotomy between factual accuracy (implied in your definition of “quality”) and influence (implied in your definition of “utility”).
Phil, I think you’re operating under multiple misreadings of my posting. First of all, I most certainly did not define “quality” as “factual accuracy.” I offered no explicit definition of “quality” at all, but in my critiques of the Coulter and Bellesiles books I implied a variety of quality criteria that include, among others, coherence of argumentation, appositeness of analogy, and reasonably charitable characterization of those who disagree. Factual accuracy is obviously important, but I didn’t (and wouldn’t) argue that factual accuracy is sufficient to define quality.
Nor do I define “utility” as equivalent to “influence.” I cite influence as one possible index of utility in the case of the Bellesiles book, but I also point out other factors that may make a book useful despite its low quality: for example, its applicability as an exemplar of a significant rhetorical or cultural trend and its usefulness as a case study in poor or dishonest journalism. These and other factors may make a book useful to researchers regardless of its wider influence.
I was also very careful _not_ to set up a dichotomy between “quality” and “utility.” In fact, I said quite specifically that quality is one factor among many that may make a book useful. Suggesting that utility ought to be the ultimate criterion for book selection is not to say that libraries should hold only useful books, and reject high-quality ones. It’s to say that utility ultimately matters more than quality. In my argument the two categories are overlapping, not dichotomous.
So, all of that said: no, I don’t think it would be better to state that librarians should select and aquire content that is “important,” partly because importance is an imperfect index of utility and partly because importance is a moving target. A book that is arguably unimportant may still be tremendously useful to the researchers my library supports, and a book that seems important today may turn out to be trivial later on. Am I saying that “importance” doesn’t matter? No. I’m saying that usefulness matters more. As it does with quality, usefulness is a criterion that overlaps with importance.
Rick, could you please define “usefulness” for me?
Essentially, you are relying on an abstract construct (like “quality” or “importance”) that is easy to talk about in the abstract, or to defend in the negative by what you are emphatically not talking about.
Second, how would you measure “usefulness”?
Phil, you’re asking two questions, one of which I think is a good one, and one of which I think is a red herring. The good question is how to define usefulness. In a broad sense I think that’s pretty easy: a book is useful if it enables its user to accomplish a desired task. To define it more specifically is harder, because a book can be useful in an almost infinite number of ways. It may be useful because it contains relevant and accurate information, or because it provides a good example of a particular kind of rhetoric (regardless of the honesty or intrinsic accuracy of that rhetoric), or because it contains information that is factually inaccurate but instructive regardless, or for any number of other reasons particular to the user of the book.
The red herring question is how to measure usefulness. I think it’s a red herring because the usefulness of a book (or a hammer or a sandwich or a yardstick) can’t be measured objectively; it’s a purely relative measure, and can only be evaluated by the person trying to use the book. Any book could be useful to a particular patron, depending on the type of research the patron is doing. This fact obviously has serious and potentially disturbing implications for librarianship (especially if I’m saying that librarians ought to concern themselves with usefulness while also saying it can only meaningfully be measured by patrons). That’s exactly the issue I’ll be addressing in my next posting.
One more quick note: concepts like “quality” and “importance” are, of course, abstract by their nature, and can only be discussed in the abstract. That doesn’t reduce the importance of discussing them. Quality, relevance, completeness, accuracy, importance — these are all abstract constructs, but they way we think about them has concrete effects on the ways we carry out our tasks in libraries. If we’re going to serve our patrons well, librarians need to think clearly about the ways we apply those abstract constructs in the course of our day-to-day work. So I’m trying to suggest a couple of ways of doing that.
I would think that those books that are worthy of study as social objects, but not as scholarly works, fall under the category of special collections. A more extreme example might be a collection of KKK hate literature. Such collections are not shelved with the general scholarly works but are available for scholarly study.
David, I think you misunderstand the purpose of special collections in a research library. The general purpose of SC is not to segregate books by topic (keeping those that are bad but nevetherless worthy of study as social objects away from those that are intrinsically good), but rather to preserve and curate books that are of value as physical artifacts. Books generally go into SC primarily because they need some kind of special protection–very often they’re rare or even unique, or they’re part of a defined subcollection that needs special care for some reason. A collection of KKK literature may be located in SC because of its nature as a distinct collection, and it may also be there because its content runs the risk of being stolen or defaced. However, it would be unusual for a library to house such material in SC simply because the content itself is objectionable or of low quality.
Mein Kampf is a great example of a book that is of very low quality by virtually any standard, but is nevertheless indispensable for researchers. Most research libraries would keep Mein Kampf in the general circulating collection; however, a rare or unusual edition of Mein Kampf might be kept in SC because of its value as a physical artifact.
There is a fundamental difference between primary source material (KKK literature and other opinion-based propaganda a la Anne Coulter) and scholarly research.
Great post! It is indeed refreshing to read an argument about what we *should* be collecting rather than what we *shouldn’t* be collecting.
I don’t think many people will challenge the notion that utility in some broad sense should be an important consideration in library acquisitions, and that quality is not a sine qua non for addition to a library collection. (I would add The Protocols of the Elders of Zion as another example of a useful work of low quality.) But then having admitted as much, the more difficult question becomes “useful to whom, for how long, and for what purpose(s)”? Clearly, a very wide range of books will be useful to at least some people on campus, and no library can afford to meet all needs for useful books. So, then, how does one rank some people’s needs above others? Do faculty needs rank higher than students because, generally, they will be at the university longer? Do graduate student needs outweigh undergraduate needs? Should needs for entertainment be ranked lower than needs for classroom use or research? Should a book that is useful for only a short period of time (say, because it relates to some event that is currently in the news, but not for very long) have equal priority with a book that is useful over a much longer period of time? And how does one weight these various types of utility against each other? I fear that any assessment based on utility can get very complicated when broken down in this way. No doubt librarians constantly struggle with such complexities in making their acquisition decisions. Or maybe they just pass the buck with PDA?
I dont think we should spend too much time splitting hairs about whether the books research libraries select should be “important,” or “useful,” or “high quality.” The more important point to me is that they should be doing everything in their power to minimize “selection” because it’s too damn expensive (as in, “the rent is too damn high.”). I don’t think we can intelligently parse–on any continuum– the pros and cons of best selling non-fiction, be it written by Coulter, Chris Mathews, Bill O’Reilly, Bill Clinton or Glen Beck. The better course of action would be for a library to decide that it will buy best-selling non-fiction (and maybe fiction, but that’s not touched upon in your post) because these books– along with TV, blogs, Facebook posts, magazines, etc.– are influential in informing or misinforming the populace at large. If an average–or slightly better than average– research library spends $5 million per year on monographs, it could pick up all the non-fiction bestsellers for about $10,000 a year or 0.2% of its monograph spending. That’s a lot cheaper than paying librarians to decide whether a book is high quality, or useful, or important– or all or none of the above. Decisions about popular trade books have almost no discernible budget impact on research libraries, but passing over the books that educated people have actually heard about– lousy as they might be– probably ends up costing libraries a hundred times their MSRP as users come looking for them year after year. Better to buy ’em and move on– one administrative decision as opposed to hundreds of tiny “expert” decisions.
Mark, you too are anticipating some of the stuff I’ll say in my next posting. Geez, you guys. I guess I need to pick up the pace…
Sandy, these are good questions — you’re anticipating some of the issues I’ll be addressing in my follow-up post.
I guess implicit in this argument is the point that citation counts are unreliable guides to quality – because they also reflect usefulness.
Yes, I think that’s safe to say. Citations may be made in the context of either approval or disapproval — and in either case, the article arguably proved useful to the person who cited it.
“…the purpose of the library is to give people access to useful documents — or, more to the point, to connect its users with the documents they need in order to do their work.”
In my opinion, this is the most important statement in this posting. While librarians have long desired quality in their collections, what our users tend to want more than anything is access to content that they need. They will be the judges of what constitutes quality.
I review books to be added to the permanent collection that come in from titles rejected by a local newspaper for review. In that group are a lot of self help books, that are outside our normal collection parameters. However we have in the past put all those “rejects” into a current reading collection for a period of time, usually no longer than a year (depending on whether they get used or not) And honestly the self help books get used more than most of the other titles in this collection. I don’t consider that to mean we should be adding self help books to the purchased collection. Utility well, yes, because they get used. Purchases for an academic library collection, not so much.
Although I agree with the suggestion that utility needs to be weighted heavier in the selection process, the immediate problem would then be how to label WHY certain low quality but high utility items were included. I believe that most librarians (and I hope most library users) are aware of our attempt to include all and only those resources that are of high quality given the limitations of our budgets. It may be obvious what books are on the shelf for their meta-value, such as the Coulter work, but without surrounding documentation, an undergrad student will not know that the Bellesiles is similarly useful. Adding such information to the catalog would not do it given the general lack of awareness and interest in the valuable metadata already there. Do we then put warning labels on our meta-useful print books and click-through warnings on our eresources? I think that’s the most important hurdle to overcome for this to happen.
I’d suggest that the question “utility for what?” be answered. If a self-help book or novel or some other type of work is being used by a student for purely personal reasons, unrelated in any way to the student’s education in college, I’d argue that it is not a really important function of academic libraries to service such needs. that is why we have public libraries and bookstores.
Matthew, that’s an interesting question. On the one hand, it would clearly make sense to add some kind of information to, for example, the Bellesiles book in order to give readers a heads-up that much of the purportedly factual data it contains was actually fabricated. (I think such a notification would be less necessary in the case of a book like Coulter’s, since her book makes no pretense of being anything other than a partisan screed.) The problem is, where would we draw the line? There will surely be many books in any research library’s collection that also contain factual errors (whether intentional or not), some of them serious — at what point would we decide that a librarian’s intercession is needed to bring those errors to the user’s attention? And what about the greyer areas — a book that is considered egregiously misleading or wrongheaded by one librarian, but maybe not by another?
I actually think the most important hurdles to overcome are two: first, the assumption that it is the library’s duty to provide “all and only those resources that are of high quality”; I argue against that assumption in my posting. Second, our fear that library patrons are incapable of discriminating effectively between appropriate and inappropriate books. Granted, many library patrons don’t discriminate as well as we would like them to, and end up using books that serve them less well than others would. But I don’t see a solution to that problem that wouldn’t be worse than the problem. At some point, we simply need to give our patrons the freedom to use resources in they ways they see fit. (Note that I’m talking specifically about research libraries here; the librarian’s responsibility may well be different in, for example, a grade-school library.)