Home 9 Against the Grain 9 v27 #5 Curating Collective Collections — A Forest for the Trees: A Response to Jacob Nadal’s “Silvaculture in the Stacks”

v27 #5 Curating Collective Collections — A Forest for the Trees: A Response to Jacob Nadal’s “Silvaculture in the Stacks”

by | Dec 7, 2015 | 0 comments

by Andrew Stauffer  (Associate Professor, English, University of Virginia; Director of NINES)

Column Editor:  Bob Kieft  (Retired, Occidental College, 688 Holly Avenue, Unit 4, St. Paul, MN  55104)

Column Editor’s Note:  My November 2014 column, “What Exactly Are We Retaining When We Retain That Book?,” asked, not at all rhetorically, questions that arise in the planning for shared print agreements  because of the physicality of books.  When libraries and their readers consider the consequences of joining a shared print agreement, particularly the potential such an agreement holds for reducing the number of copies held by the parties to it, people wonder about access to retained copies, the reliability of the partnership, the integrity of the copies designated for retention, and the role that variation or artifactual value among the copies in the collective might play in retention decisions.  When it comes to the book as a physical entity, librarians and readers ask when a given volume is a copy of another, that is, what do we mean by copy, when can a given volume stand in for other examples of that book, and what are the ways in which and purposes for which a volume is not just another instantiation of the same text.  Catalogers have developed answers to these questions, of course, but the answers can be more complex than the reasons for matching a volume to a WorldCat record make them out to be, and decisions based on those answers can be important for several kinds of scholarship.

“Curating Collective Collections” followed up that article with “Silvaculture in the Stacks; or, Lessons from Another Conservation Movement” by Jacob Nadal, Executive Director of ReCAP, in which he uses a forest conservation metaphor to help librarians frame the issues these questions raise as they design collective print management programs.  In this column, Andrew Stauffer responds to Jake’s article from the point of view of a historian of books and the cultures of reading.  Andy has been developing this argument for some years now in such venues as his Book Traces project (http://www.booktraces.org/), sessions at the Modern Language Association annual convention in January 2012 and 2013, the Print Archive Network Forum and RBMS Preconference at ALA Annual in June 2014, and a grant to the University of Virginia’s libraries under CLIR’ Hidden Collections program.1  He pursues here the questions of copy variation and uniqueness, understood both as physical attributes of volumes and, most important for him in his grant-funded work, as evidence readers have left of their interactions with the text and the volume itself.  As he points out, not every volume that meets the bibliographic qualifications for entry of a holdings symbol on a given WorldCat record is the same book.  Andy thus follows Jake into the dispute about “logging” that Ronald Reagan, when a candidate for the governorhip of California, participated in and during which he is famously paraphrased as having said, “When you’ve seen one redwood, you’ve seen them all.”  He in fact said something less vivid but more arboreally encompassing, “…you know, a tree is a tree, how many more do you need to look at?”2  Arborist Stauffer begs to remind us how different those trees can be and how important those differences are for the forest.

Andy is a colleague on the multi-organizational Future of the Print Record task force (https://printrecord.commons.mla.org/), and I am grateful to him for continuing this important thread in the discussion of print collection management and the roles of print books in scholarship.  Follow the URLs to a description of the grant and to vivid examples of the volumes that he is concerned be retained.  This discussion will continue in an upcoming double column by Mike Garabedian, Whittier College, in which Mike makes a case for shared print agreements’ considering “condition,” understood as proximity of a given copy to its as-published state, in determining which and how many copies to retain; in making the case, he will report on a survey of condition he performed as part of SCELC’s planning for a shared print agreement. — BK

In a recent Against the Grain essay, “Silvaculture in the Stacks,” Jacob Nadal draws upon a forestry metaphor to discuss library print collections management.3  Warning against a rush to “clear-cut logging” (i.e., the rash withdrawal of books) to save space and money, Nadal makes the case for the measured, collective management of collections, recommending statistical models to identify preservation and withdrawal candidates across the library system (70).  He argues that we need to develop an overall collaborative holdings level for titles and then work within those guidelines, thinning and felling only according to broader knowledge of the forest.  In this essay and elsewhere, Nadal draws on the mathematical framework for biodiversity developed by Martin Weitzman, suggesting that we think of North American libraries as a total ecosystem, one in which we want to optimize holdings strategies to preserve as much variety as we can afford even as we draw down the print collections.4  This aspect of librarianship is essentially about statistics and risk management: how many copies do we need, where do they need to be stored and how accessed, to ensure the greatest chance that the information will remain available in, say, one hundred years?

Although he never uses the word, Nadal is talking about something colloquially called “weeding” — but you don’t weed a forest: the scale of shared print management is so large (regional, national, continental) that the garden metaphor seems less fitting.  Forestry offers a more capacious way of thinking about the process.  “Done properly, our shared print programs will clear away the cruft,” he concludes, removing the overstock of unnecessary redundancy in the North American libraries’ collective collection of perhaps a billion books (71).  Good forestry involves cutting down some trees, he reminds us, just as withdrawing books has always been part of librarianship.  Now, his logic runs, we have the data and the tools to do it better.  The metaphor strains here a bit, since the reason you cut down trees is mostly to make room for new ones, whereas the draw-down of print is now general and ongoing: the overall library acreage devoted to books is declining with some speed.  Withdrawals will make room for some new books coming in, but the systemic reduction of print collections is primarily to allow for other kinds of new growth: digital resources, common study areas, and other non-bookish aspects of the twenty-first century library.5  Nevertheless, Nadal’s “silvaculture” metaphor reminds us of the need to preserve the richness of our old-growth collections, and to be responsible stewards of that inheritance.

The forestry metaphor comes readily to hand, of course, since most of those billion library books are literally made from trees: “tree flakes encased in dead cow,” as William J. Mitchell memorably describes printed books.6 “I don’t read dead tree books anymore,” a friend recently told me, in praising the virtues of the Kindle.  Behind this rhetoric is a salient fact of book history: the transition from linen-rag to wood-pulp paper in the middle-nineteenth century, amidst the industrialization of the press and the massive expansion of book publishing that, in the event, would produce much of our non-rare legacy print collections.  Books became cheaper and more plentiful in an age of pulp.  In other words, trees enabled the library conditions that Nadal means to address via his silvaculture metaphor: we have to manage the forest of books that trees fed.  Moreover, the chemical composition of much of that wood-pulp paper means that many of the older books are now tanned and brittle, preservation candidates that complicate the retention agreement process.  In some ways then, trees are the basis of our general collections and are at the root (sorry) of the challenges facing library collections management today.

Insofar as the “collective collections” movement grows out of the bibliographic condition of nineteenth- and early-twentieth-century (i.e., non-rare and out-of-copyright) books, it needs to take into account the specific qualities of that condition.  Otherwise, the biodiversity model breaks down: you can’t take steps to preserve an ecosystem without a science that investigates the nature of organisms you are trying to preserve in relation to one another.  For books, that science is bibliography.  And my concern is that library print collections management programs are being driven by bibliographically impoverished metadata, catalog records that provide an insufficiently detailed picture of the books themselves.  What look like mere instances or copies from a distance turn out to be variant species when examined.  Preserving true bibliodiversity in the stacks depends upon a more nuanced view of the “copy,” even (perhaps especially) for books produced in the age of the stereotype plate and the industrial printing press.

Which brings me to Joyce Kilmer’s Trees and Other Poems, published by George H. Doran in New York City in 1914.7  I recently started buying up copies of this once-popular, now-neglected edition, which now can be found in quantity on the used book market; and WorldCat lists 585 copies in its members libraries’ collections.8 All of the thirty-plus copies in my collection conform to the same metadata and would be listed in WorldCat along with these as copies.  But no two are identical.  They vary either because of differences at the point of production — different bindings (brown boards or grey, blind-stamped or not), pasted labels (green ink or black, thick font or thin), number of free end-papers (from one to four), treatment of the top-edge (gilt or not) — or because of their condition and evidence of use: dust-jacket or not, gift inscriptions, inserts, marginalia, etc.  To map the true history of this edition and the ways it was read, one needs as much of this evidence as possible.  My point is not that the 1914 Kilmer’s Trees is particularly multiform: virtually all books from this era and earlier vary from their peers in more or less significant ways that are not being captured by WorldCat metadata.  That’s what humanities scholarship is for: to disambiguate and explicate a given textual / cultural scene.  But it can’t be done without the evidence found only in the historical record itself.

Comparison across multiple copies is the foundation of bibliography in its various forms: enumerative, descriptive, analytical, and textual.9  Moreover, book history, literary studies, and other humanities disciplines also rest upon attention to the specific objects and interfaces produced and used by a culture.  We all know this, but such knowledge frequently gets submerged in conversations about the collective collection, even by professionals such as Nadal who are working to preserve bibliodiversity in libraries.  The data-driven de-selection movement in the library profession is currently proceeding along distorted — because oversimplified — lines.  Until we pay closer attention to individual copies in their relations to one another, we will be operating in a darkness all the more confounding because it looks like light.  The acquisition of Sustainable Collections Services by OCLC in January of this year lends a sense of urgency to the situation, since SCS is a leading force in the field of data-driven de-selection.10  We need more nuanced conversations about so-called duplicates of pre-1923 monographs, which embody an irreproducible archive in their aggregate force and individual evidentiary weight.

Joyce Kilmer’s most famous poem, “Trees,” was once memorized by children and adults across the nation: “I think that I shall never see, a poem lovely as a tree.”  After several stanzas of evocative description, Kilmer concludes, “Poems are made by fools like me, but only God can make a tree.”  Even in a more secular age, we can see that the books constituting our legacy print collections are like Kilmer’s trees: they can only have been made by the forces of agency and accident that we call history.  Any “fool” can make a surrogate of a single copy and proclaim that the book has been “digitized.”  Any “fool” can look at a spreadsheet of 500+ identical pieces of metadata and call the books they reference “duplicates.”  But once we manage down Nadal’s “old growth wilderness of the stacks,” those trees won’t come back: we will be permanently winnowing a multiform internationally-distributed collection whose significant variations lie hidden behind blandly aggregated metadata.  Libraries have long been the custodians of the bibliographic plenitude that enables scholarship, and that should continue as a defining mission.  We need a forest for the Trees.


  1. http://news.virginia.edu/content/grant-boosts-efforts-catalog-secrets-hidden-old-library-books.
  2. http://www.snopes.com/quotes/reagan/redwoods.asp
  3. “Silvaculture in the Stacks; or, Lessons from Another Conservation Movement,” Against the Grain 27:1 (February 2015), 70-1.
  4. http://scholar.harvard.edu/files/weitzman/files/noahs_ark_problem.pdf
  5. See, for example, Scott Seaman, “Collaborative Collection Management in a High-density Storage Facility,” College & Research Libraries 66:1 (January 2005), 20-27: “today’s library planners and architects choose to reduce footprints of traditional shelving in order to open vast areas of library floor space for collaborative study space and information technology” (20), an observation that goes back to the 1999 ARL SPEC Kit 244.
  6. City of Bits: space, place, and the Infobahn (Cambridge, MA: MIT Press, 1995), 56.
  7. On this edition, see Bibliography of American Literature 11104.
  8. http://www.worldcat.org/title/trees-and-other-poems/oclc/284968
  9. See Fredson Bowers, “Four Faces of Bibliography,” Essays in Bibliography, Text, and Editing (Charlottesville: Bibliographical Society of Virginia, 1975), 94-108.
  10. https://www.oclc.org/news/releases/2015/201502dublin.en.html


Submit a Comment

Your email address will not be published.

Share This