Responsible Librarianship: Library Policies for Unreliable Systems
By Thomas Mann
There is a kind of “code word” situation that has developed in the library profession in recent decades; it is manifested in an appeal to a set of beliefs that, while largely unarticulated, is nonetheless socially endorsed without a perceived need for argument or evidence. The evidence is assumed to be there; after all, when enough people share the same assumptions that support networks can be appealed to, those social networks functionally take the place of what, in other situations, would require considerable explicit justification. Were the “code words” actually based on the “science” part of “library science,” then their adherents would have to realistically consider the possibility of falsifying evidence, of counter-examples, of whole bodies of literature to the contrary, and of the possible–perhaps radical–incoherence of their beliefs when situated in larger contexts of other beliefs known to be true via other tests. What matters with the “code word” mind set is not whether one examines possible falsifying considerations; what matters is simply whether one “gets it” or not. Fashion replaces argumentation.
Perhaps the most insidious of the “code” beliefs in the library profession today is the oft-repeated statement “We should not let the perfect stand in the way of the good”; or “The perfect is the enemy of the good.” This assertion has various implications in practice. Most frequently it means that, particularly for library catalogers, “throughput time” or “speed” in turning out records is now to be considered “the gold standard” of quality. Despite a mountain of evidence to the contrary (Cataloging and Classification Quarterly, 23 [3/4], 1997), it is further assumed that “indexer consistency studies” demonstrate that trained catalogers don’t agree with each other in any event, in the supposedly-perfect products they have been (vainly) trying to produce up to now. And therefore, it follows, consistency–i.e., standardization, categorization, authority work, cross-referencing, etc.–is not a realistic goal to pursue in the first place; instead, we just need more transcribed or harvested keywords taken from books themselves, that can be relevance-ranked (not standardized) by machine algorithms; or we just need more “tagged” keywords added to records by the general public, whose collective folk wisdom can replace (not supplement) subject experts who have actual knowledge, not just of the book (or other work) in hand, but of the larger context of its subject relationships to other works that are within, or related to, its own field. Neither relevance-ranking nor democratic tagging by non-librarians, of course, is expensive; neither requires thinking by library personnel. Neither requires any expensive professional work. Whether the needs of the library’s users–particularly academic and scholarly users–are met by such processing procedures is irrelevant, because “the code” also assumes (without argumentation) that the very goal of cataloging is no longer to show “what the library has”–that is much too parochial a focus when there is an entire Internet out there with billions of things in it–but is, rather, to provide “something quickly”–and provide it especially to remote users outside library walls who are further assumed not to need any training or education in how to go about finding what they need. (The “under the hood” software manipulations of whatever keywords they type into “a single search box” will handle those problems for them.) Traditional library-access mechanisms will obviously not “scale up” to dealing with billions of records; so it follows that they must be simply abandoned (rather than having them continue to deal with a much more manageable subset of all informational records, such as the set of pesky books that keep being published each year.)
So: where do we go from here? I suspect many librarians will have visceral feelings that something is wrong with “the code” they hear so frequently repeated. Perhaps instead of endlessly repeating these assertions we should actually look at the evidence, either in their favor or falsifying them. That’s where David Bade comes in. Bade, a cataloger at the University of Chicago’s Regenstein Library, is a genuine scholar in the library profession. And he has done something that is rarely seen in library literature: he has read widely enough to examine library science as a whole in the context of related disciplines. He has immersed himself in the literature of high reliability organizations, human error studies, ergonomics, reliability engineering, and joint cognitive systems. He brings to bear a knowledge of philosophy, history–-and even farming!-–in his considerations of what actually works in libraries, and for what purposes. What he offers in this book is a coherent integration of what is, demonstrably, established knowledge from a wide variety of relevant fields, weighted not by machine algorithms but rather by a fine professional discrimination based on decades of actual experience in doing the work of librarianship. He combines the perspectives of a 30,000 foot overview with the necessary corrections that must be made, extensively and routinely, at ground level. A sure sign of a real scholar is his or her ability to provide concrete examples from that “ground level” experience, with an extended analysis of their further implications, not just out-of-context individual sentences cherry-picked from user surveys devised by people lacking that experience, who may have therefore failed to ask the right questions to begin with. Bade’s extensive quotations from the literature he has researched–N.B.: extended quotations from, not just superficial footnote citations to–provide the rest of us with a coherent patterning and integration of knowledge that the library field so desperately needs at present. Bade offers not the “snippets” of information that quick and dirty searching offers, but a deep understanding that comes from his rare combination of very wide reading and very extensive personal experience, not just with the intentions, but with the results of the systems he’s talking about.
“The perfect is the enemy of the good”? Perhaps, after reading this timely and much-needed study by David Bade, the library profession might actually consider a counter-proposition: “The even greater enemy of the good is the slipshod, the incompetent, the superficial, the incomplete, and the demonstrably incorrect.”