What’s love got to do with it? further thoughts on libraries and collections
Rick Anderson, in his Ithaka S+R “Issue Brief” paper, Can’t Buy Us Love: The Declining Importance of Library Books and the Rising Importance of Special Collections, proposes a radically different vision of the future of libraries in which libraries cede the organization, preservation, and curation of large areas of the information landscape to individuals and the private sector.
His paper has sparked a lively discussion. Barbara Fister (What Are Libraries, Anyway?), Chris Bourg (Looking for love in all the wrong places and Thing called Love: Further thoughts on #lovegate), and Wayne Bivens-Tatum (Opting In) have contributed thoughtful reflections on the paper and replies from Anderson and comments from others — including FGI — at those links and on twitter have also been thoughtful and sometimes heated as any meaningful debate ought to be.
The discussion is emerging as a fundamental contrast in values. The choices we make today will affect the preservation of information, the ability of users to get and use the information they require, and the future of libraries.
A modest proposal?
Anderson focuses on a single trend that he describes as an “existential threat” to some of the key roles of libraries. Although he recognizes the trend as being driven by our networked digital environment, he says that the digital shift itself is less important than what he calls a “gap” that has been created between “commodity and non-commodity documents.” Anderson’s thesis is that, because a lot of books are easy to find and cheap to buy, libraries need not buy as many because users can get their own — "The bottom line is that libraries are no longer needed in the way they once were to provide access to documents that are available in the commercial marketplace." His proposal is that libraries should therefore shift a portion of their budgets and staff resources away from these easy-to-find and cheap-to-buy books (and other “commodity documents”) and toward acquiring, digitizing and giving access to Special Collections- "non-commodity" materials that are often hard to find and hard to access.
If he were only suggesting that libraries should do a better job of balancing their finite resources by making minor allocation adjustments away from information that users can get easily and toward information that is difficult or even impossible for users to get, his proposal would not sound radical or controversial. When he says that his goal is to "enrich the scholarly environment with useful books and other documents that would otherwise remain hidden from scholars and students,” who would disagree? In fact, most librarians could easily list additional collections that have been neglected over the years and could benefit from additional resources in the same way that Special Collections could. Some examples of these include: audio recordings, slides and other art reproductions, research data, non-book materials, locally produced materials like ‘zines, government documents, and other collections that are unique-as-collections even if their individual components are not rare.
Indeed, libraries already regularly review their resource allocations and adjust them to accomodate changing resources, the changing needs of their communities, and the changing information environment. Choosing to refrain from purchasing some books available in the commodity marketplace is not new. Libraries have never been able to acquire everything that is in their scope. Libraries have always attempted to balance resources wisely by shifting allocations. For example, think of joint purchase programs and shared preservation initiatives including the Center for Research Libraries, the Western Regional Storage Trust (WEST) and the Law Library Microform Consortium (LLMC). Consider also how libraries increasingly rely on inter-library loan and how budget allocations have shifted from monographs to serials and among disciplines. If Anderson were only pointing out a way to do this more effectively, his paper would not have prompted the discussion and controversy it has.
If he were suggesting that libraries consider commercial availability of information as one of many factors in building a collection, his paper would be a useful elaboration of that single factor. If he acknowledged how libraries already help in making information easy to find and acquire, and in the provision of information distribution, and in the Open Access movement, it would be clear that libraries play a key role in the “transformative reality” he describes, rather than being “undermined” by it. If he acknowledged the important role of librarians in providing services for collections to users, his recommendations would at least seem less sparse and disconnected. But he does none of that.
A radical proposal
Instead, he offers a very narrow view of the roles of libraries. He defines curation as “ensuring that access endures” and as providing “storage.” Instead of acknowledging the importance of library collections built through the iterative and intellectual processes of selection, acquisition, curation, and preservation, he reduces all of this to a “broker” role of simply “buying access.” He expressed his vision of the library even more explicitly in an article in Scholarly Kitchen where he wrote, “[T]he purpose of the collection isn’t to be a great collection; it’s to connect patrons with exactly what they need.”
Even that very limited vision of the role of libraries would at least have an internal consistency if Anderson analyzed its sustainability by examining the potential long-term effects of its implementation on a wide scale. But he does not do this. He provides no hint at how libraries will “ensure” access to information that they neither acquire nor control. Neither does he explain how the marketplace will provide adequate, affordable (or toll-free) access for all academic users to the information that libraries relinquish. And he does not describe how information will be adequately preserved by the private sector. He does, however, assume that an “efficient marketplace” allows libraries to leave users on their own to a greater extent than ever before.
If the various readings of Anderson’s paper as a radical proposal were over-reactions, he could have corrected our misreadings. But he has not yet done so. For, although he addresses the speed with which the shift should happen, he is vague about how extensive it should be. He says that the shift will be gradual (for “most” libraries) and the “trajectory” of the shift will vary from library to library and should be determined in an open and transparent way. But he never quantifies how great a shift of resources he imagines, though he does say that a sudden 50% shift would be “both irresponsible and unwise.” He does, however, use ominous rhetoric (the “important shift”, the “existential threat”, our failure to prepare for a new reality, the “fundamental undermining” of our key roles), which suggests the need for significant changes, not minor adjustments.
Anderson is, apparently, addressing academic libraries in general, not just telling us what he is doing at his library, so how we understand his suggestions has the potential for affecting the entire ecosystem of academic libraries in general. Is Anderson actually proposing such a radical shift in the role of libraries, or only minor adjustments to the way libraries already behave? Is he saying that libraries must make a shift because of a “transformative reality,” or that libraries have a lot of flexibility in how they adjust to a changing information landscape? Is he proposing a change in library priorities away from “commodity” collections and toward access-but-not-curation, or a small shift in the composition of library collections?
Understanding the difference
The key to understanding his proposal is to understand how extensive a shift Anderson is proposing. To the extent that Anderson is proposing only minor allocation shifts, his proposal itself is minor and should not be particularly controversial. On the other hand, if Anderson is proposing a more radical shift in resource allocations, librarians should have a good understanding of the possible effects of such re-allocations before “buying” his song.
Before examining the potential effects of Anderson’s proposal, it is necessary to provide one additional bit of context to it. In focusing on the “gap” between “commodity and non-commodity documents,” Anderson narrows his focus to only one part of the information landscape. There are vast quantities of information that are not available for sale at all but are neither rare nor unique nor noncommercial. Perhaps the biggest such category is information that is available for licensing or leasing, but not for sale. This includes most commercial database services and most e-books. This is information that can be “accessed” for a commercial fee, but that vendors and producers are reluctant to sell outright — either to individuals or libraries. Anderson does not specify if he would include renting information access as part of his definition of “commodity documents,” but his analysis definitely glosses over an essential issue for library access and collections. Indeed, Clifford Lynch recently wrote eloquently about ebooks in particular, calling them a “Faustian Bargain” and saying that they are a grave threat to “both the access and preservation functions of our libraries.” With this issue left unaddressed, it is difficult to know the precise effect of Anderson’s proposal, but it is impossible to predict a positive effect.
Other information that falls outside of the commodity and non-commodity groups includes: the information that is freely available on web sites (blogs, output from research institutes, NGOs, and much more), government publications and data, open access scholarly articles, research data, and more. This is information that can be acquired without paying a commercial fee. Such information also deserves the attention and resources of libraries. Again, with the status of such information left unaddressed, it is difficult to judge the effects of Anderson’s proposal. He does not say if he thinks resources should be shifted into or out of parts of the information landscape that fall outside the narrow scope of his paper. By leaving these large categories of information out of his analysis, Anderson over-simplifies a complex information landscape and offers an over-simplified proposal without adequate context.
Effects on Library Collections
Let’s look at the potential effects on a library’s own collections. What would be the effects implementing Anderson’s proposal to shift resources from “commodity documents” to “non-commodity documents”? To examine the effects, I assume first that a library would have to identify “commodity” information that is in its scope, which it would normally purchase (call that “X”), and then refrain from purchasing some percentage of X (call that “Y”). In each case below, a small shift (a small Y) would have less effect, but a large shift (a large Y) could have a major effect. I am trying to avoid predictions and focus instead on objective effects that follow Anderson’s expressed preferences (i.e., “commodity” over “non-commodity” documents, access over collections, etc.).
- It would decrease the quantity of scholarly information in the library collection. In proposing a shift in allocations away from X, Anderson is proposing that libraries decrease the quantity of one category (X) of information in their collections that is relevant to its user communities. We do not know if this would be a 1% decrease or a 99% decrease of X. If this were a decrease in resources “that make relatively little contribution to [the scholarly] environment,” it would not matter so much. But Anderson’s own examples of “commodity documents” include everything from inexpensive used books, to large print-run trade books, to scholarly journals, to anything in the commercial (“toll-access”) marketplace. As libraries decrease their acquisitions of information relevant to their user communities, they decrease the quantity of information critical to the scholarly environment that they control, curate, or preserve. The greater the shift in resources, the greater the long-term effect.
- It would have an effect on the quality of collections. To follow Anderson’s logic, a library would target for exclusion the information that is most important to users. His argument rests on the assumption that individuals are increasingly willing to spend their own time and financial resources to acquire information. Presumably individuals would only spend their personal resources on information that is most important to them. If a library must shift resources away from this category of information, it will be targeting for exclusion from its collection the very information that is, by definition, most important to users. The more the library proscribes, the greater the loss of quality to the collection available to the community as a whole.
- It would affect the overall value of the collections. Libraries will be under pressure to acquire more commodity information that is academically “popular” and acquire less commodity information that is more esoteric — even if academically valuable. Since libraries will be purchasing less commercially available information, they will want their purchases to reach as many users as possible and thus slow or reverse the “massive decline in the use of their general collections” that Anderson describes in introducing his thesis. Since all users will not be able or willing to afford to purchase everything they want or need, libraries will still purchase some “commodity documents.” The more people who want a given “commodity document,” the more likely it will be that some portion of those people will not be able or willing to buy their own copy; this will create a user community need for those items — which are the ones wanted by the most people. Information that is more esoteric, less well known, less visible, part of a small sub-discipline or specialized area will, by definition, be wanted by fewer people. It will be harder for libraries to justify spending their (shrinking) resources for such “commodity documents” under Anderson’s recommended approach. By focusing on “access” instead of on collection-building, and by focusing on “commodity/non-commodity” rather than on other criteria, libraries would be more likely to build low-quality, popularized academic general collections. The greater the shift, the greater the long-term effect on the value of these collections.
- It would have an effect on the functionality of collections. Anderson says that it is the commodity/non-commodity “gap” that is of real importance to libraries — not analog vs. digital, or purchase vs. lease, or open usability vs. contractually or technologically limited usability. But paying more attention to one set of criteria means paying less attention to others. The focus on “commodity documents” would therefore affect the choices available to the library and the resulting choices would have effects on the functionality of the collections chosen. For example, by reducing the allocations for first-sale documents, libraries might choose to lease temporary “access” to fill in gaps; they might more readily accept information restricted by Digital Rights Management (DRM) technologies; they might accept prohibitions on text and data mining; and so forth. Each such decision would (in the short term) fulfill Anderson’s desire for “access” but would reduce functionality and, over the long term, harm collections.
Anderson might well claim that the increase in resources to Special Collections would mitigate some of the above problems and even provide a better balance to the collections. But this would only be a reasonable argument if those making collection allocation decisions had the freedom to ensure against the kind of damage described above and the freedom to actively manage the balance of the collections. Instead of this approach, however, Anderson recommends a more rigid approach to collection management based on a “gap” that he sees, but which he does not demonstrate will actually have any mitigating or balancing effects. When asked about the role of subject specialists, Anderson even said, dismissively, that they were out of the “scope” of his paper. This is yet another indication of Anderson’s focus on short-term access over long-term curation.
The more that libraries rely on Anderson’s “gap” approach to “access” rather than a balanced approach to collection-building, the greater the negative effects. The effect is unequivocal: as libraries reduce their collections of “commodity documents” they cede the organization, preservation, and curation of those documents to the private sector. To mitigate the effects, libraries should rely less on Anderson’s approach.
In each of the above examples, the key point is that the degree of harm to libraries would be relative in direct proportion to the implementation of Anderson’s proposal. The greater the shift in resources, the greater the harmful effect. The more we buy Anderson’s argument, the less flexibility we have to manage our collections in the best way for our user communities.
Second order effects
In addition to the above first-order effects (those that will directly affect library collections), there are a number of second-order effects. One might call these “unintended consequences” if they were not closer to certainties than to possibilities.
- There will be a shift from community resources to personal resources. The library’s user communities will have to rely on their own personal resources for access to whatever the library fails to acquire. The extent to which libraries follow Anderson’s recommendation will directly affect the extent to which users will have to rely on their own resources, rather than community resources, to get the information they need. The effects of Anderson’s proposal would thus fall inequitably on different user communities with different personal resources. This would also have a particularly negative effect on academic researchers who rely on large quantities of information that they do not necessarily want to add to their own personal collections. As individuals are increasingly forced to rely on their own resources for information they require, the role of the library as a resource for the community will be weakened.
- Information will be lost in the long-run. Short-term availability (easy-to-find and cheap-to-buy today) is no predictor of long-term availability. There is no reason to believe that the commercial sector will curate and preserve and make available at an affordable price all information that libraries would. This means that, over time, some information will be lost as it leaves the marketplace entirely, and some will not be within affordable reach to many users. The information that the private sector is most likely to keep available for the long term is the information that is most profitable. Ironically, this is is the very information (popular, used by many, in demand) that libraries will most likely acquire (see above). Tragically, the information that the private sector is least likely to keep available (the esoteric, little used, less popular) is the very information that libraries will be least likely to acquire.
- Access to information will be restricted, or limited, or expensive, or all three. By definition, all the information in the commercial marketplace will be available only for a fee. This is, in fact, a key pillar of Anderson’s thesis. Individuals paying for toll-access to information is an acceptable, even essential, part of Anderson’s vision. Anderson assumes that libraries will not need to “broker” (i.e., buy books) for users because it is less difficult for users to buy their own. (I must mention that this assumption reveals a deep flaw in Anderson’s argument. It only tells one of the reasons that libraries buy books and other “commodity documents.” Libraries also buy books to provide toll-free use of information and to build coherent collections for the long-term. Toll-free use is particularly important to academic communities because researchers do not buy every book, and journal, and magazine, and encyclopedia, and reference book that they would ever want to read, skim, refer to, or consult. Neither do they want to build personal collections that contain every bit of information that they will ever want. Nor do users necessarily have the skills, resources, or inclination to preserve and curate their own personal collections. Libraries collections are also more than “storage” used to facilitate access. They are selected and organized and preserved to reflect bodies of knowledge that their user communities can browse in, refer to, read, consult, and use in ways that are not as easy or as possible with masses of undifferentiated commercial information, which is often organized more by provenance or publisher or vendor than by the interests of the library’s user communities.) The marketplace creates barriers in order to sell a product. Libraries remove barriers in order to encourage and facilite the use of information and creation of knowledge. Libraries do this for academic communities. The marketplace does it for individual “customers” — one sale at a time.
- Use of information will be restricted. Inevitably, some of the information that libraries fail to acquire will be digital. The current marketplace for digital information relentlessly (and often arbitrarily) applies use-restrictions of various kinds. These can be technological (e.g., DRM technologies) or contractual. It is distressing, in a digital age when digital information has the potential for being used and reused and repurposed in new and useful ways, that a librarian would advocate relying on the private sector, which restricts such use.
As above, the extent of these second-order effects will be proportional to the degree of implementation of Anderson’s recommendation. The more that libraries shift resources away from “commodity documents,” the greater will be the impact of the above effects.
Relying on the private sector
As libraries shift resources away from “commodity documents” their user communities will have to rely increasingly on the private sector for access to information they require — both for short-term access and long-term preservation, curation, and access.
The primary mission of private sector companies is to make a profit, not preserve information. Even if information “access” is part of a company’s business model, preservation is, at best, only a byproduct of its primary mission of making a profit. And private sector companies, even successful ones, do not always last. For example, Marty Kelly at IBM pointed out that, of the top 25 industrial corporations in the United States in 1900, only two remained on that list at the start of the 1960s, and of the top 25 companies on the Fortune 500 in 1961, only six remained there in 2011. ITworld did a recent feature about ten prominent tech companies that have failed (Fruhlinger). An information ecosystem made up of hundreds of libraries with as many funding streams assures that, even if one library fails, the whole system remains functional for the long term.
I understand Anderson’s analysis of the current information landscape, but I disagree with his assumptions about that landscape and believe that his proposed path forward will more likely damage libraries and negatively affect our user communities. If many libraries followed Anderson’s advice to a large extent, the result could even affect the information landscape itself. Libraries have a symbiotic relationship with the market. As they withdraw or shift their resources, the market will respond. I do not think any of us can accurately predict the effect on the information landscape, but that unpredictability is another reason to not rely on the private sector for long-term access.
A clear, long-term vision
It is important to realize and understand that libraries working within the current information landscape do not work in a vacuum. The network of libraries, created over time by awareness and cooperation and serendipity, establishes a nexus of information access and curation. The shift to digital formats may make it more difficult and costly for libraries to fulfill their traditional roles of broker, curator, and organizer but they do not erode those roles or lessen their importance. On the contrary, the networked digital information landscape creates an environment that makes libraries’ traditional roles more important to their user communities than ever.
My FGI colleagues and I have over the last ten years attempted to map out a more positive, expanding, and expansive future for libraries in general and government information collections in particular, based on the assumption that there is a critical need for what libraries have traditionally been good at in concert — the collection, description, and preservation of, and provision of access to and services for, the world’s information and knowledge output.
I feel strongly that the solution is not to retreat from libraries’ traditional roles, but build on those roles. I think libraries can do more of what they have always done and do it better by using the digital tools available to us now. Anderson is right about one thing: Libraries do need to make a choice because of changes in the information landscape.
On the one hand, libraries could choose to accept user-fees for access to essential information, and choose to rely the private sector to select what information will be preserved. Libraries could choose to accept private sector control of an increasing body of important information, choose to let the private sector decide how that information will be organized and presented and delivered, and choose to rely on the private sector to decide how that information can be used — and not used. Libraries could relinquish control of information to the private sector and accept shrinking the scope of libraries to the few rare and unique and special things that we still control.
Or, on the other hand, libraries could apply digital tools, leadership, and activism to those traditional roles to build a better information landscape for the future. Libraries could continue to wrest information control out of the hands of the private sector, and not surrender selection, access, organization, and preservation to it.
Anderson sees vulnerability and wants to get out of the way; I see vulnerability and want to actively work to turn those vulnerabilities to libraries’ and, more importantly, the public’s advantage.
A clear, long-term approach
Instead of concentrating on an artificial “gap” that diminishes the scope of libraries, there are several better things we can do. We need to be advocates for our communities, not just for our libraries. We need to lead our campuses, not “sidestep” issues that are critical to the information landscape of the future:
- Librarians should recognize that the market is not a solution to information access, it is a barrier to access. In the digital age, producers, vendors, and distributors are vying for more control over information than they have ever had. Librarians should understand and explain to their university administrations clearly, concisely, and compellingly, that the gap we should worry about is the one between the market and libraries as information providers, not the gap between “commodity” and “non-commodity” material.
- Librarians should explain clearly, concisely, and compellingly the difference between decisions made by information professionals for their communities of users and those made by corporations for their stock holders. Where corporations look after their “assets” (i.e, information), a library looks after its community.
- Librarians should explain clearly, concisely, and compellingly that the gap that matters most is the one between between Constitutionally protected fair-use and use that is limited by commercial contracts; between first sale and pay-per-view; between Open Access and toll-access — even “freemium” access (which restricts use and usability); and between light archives and dark archives.
- In the digital age it is particularly important for libraries to advocate, promote, and protect the advantages of using, re-using, and repurposing information; librarians should fight against attempts to impose technological and contractual restrictions on use and re-use of information, not accept those restrictions passively. Libraries should lead and advocate for their communities, not bow to the demands of corporations that benefit from restricting access.
- Librarians should embrace the importance of their role in the preservation and curation of all information that their users need regardless of format and reject attempts (like Anderson’s) to minimize that importance for information that is for sale — or rent.
- Librarians should devote resources to collections and services — especially to those that do not fall into Anderson’s “commodity” / “non-commodity” dichotomy. Librarians that are keen to learn from the commercial information industry should note that successful information businesses provide services for collections that they acquire and control.
- Librarians and libraries need to develop and support those policies and infrastructures that enable libraries to gain sufficient control of information content to ensure its preservation, access, use, re-use, and repurposing. This includes supporting things like digital deposit — the concept whereby information producers deposit digital content, such as government publications and data, into systems under library control. They should also support local Institutional Repositories for Open Access, and distributed digital preservation systems built, like LOCKSS, on the Open Archival Information System (OAIS) framework.
- Librarians should also commit to learning technological skills and to using those skills to build, maintain, and share open-source technology tools and infrastructures. Rather that simply relying on commercial tools, librarians should collaboratively develop their own smart tools and infrastructures with community- and library values (such as privacy-protection, and transparency, and metadata standards) built-in. In an environment in which individuals are increasingly distrustful of commercial information services, there is a clear opportunity for libraries to build on their traditional trustworthiness and community-focus to build tools that put the information needs and expectations of users first.
- Library managers that struggle to articulate the need for and value of libraries should follow the lead of successful libraries and information projects like LOCKSS, the MetaArchive, the California Digital Library, the Ontario Council of University Libraries’ Scholars Portal, the Internet Archive, HathiTrust, GPO, the University of North Texas, the Douglas County Library ebook platform, etc. — when I started looking, I found too many to list here! Large and small, these are projects that have aimed higher rather than lower. They have used traditional library values, examined the information landscape and the technological tools available, and made a new role for themselves. Library managers that worry about small budgets should be developing new preservation and access and use projects (which campuses very much need!) and asking for more funding. Libraries should be leading their campuses to address the big issues and find solutions that will work for the long-term. They should not be “opting out” or “sidestepping” the issues or shrinking the scope of collections, or minimizing the roles of libraries.
Librarianship is a collaborative long-term endeavor. It takes libraries of all sizes and abilities to ensure the sustainability of our intellectual and informational ecosystem. Every library has its role. If Anderson wants the University of Utah library to concentrate on toll-access and Special Collections, that is the prerogative of the library and the campus. But for academic libraries in general to follow that model would be a mistake. Instead of opting out or arguing that libraries can no longer compete in the networked digital environment/market, more libraries need to work together to leverage non-library infrastructures/ecosystems to continue to do our critical work.
Let’s seed the cloud not cede the cloud.