Michelle Wu, associate dean for library services and professor of law at Georgetown Law, recently sat down with Jocelyn Kennedy, executive director of the Harvard Law School Library, for a conversation on leading law libraries in the 21st century. (Photo credit: Georgetown Law)
Jocelyn Kennedy: Michelle, thank you so much for doing this. We are very excited that the Harvard Law School (HLS) Center on the Legal Profession decided to do an issue of their digital magazine, The Practice, on legal information and law libraries. When Adam Ziegler here at the HLS Library asked me who I thought we should talk to in the library community about the intersection of legal information and technology, my first response was, “Well, we have to talk to Michelle.” The work you’ve been doing in the open-access space and with controlled digital lending (CDL) is just so interesting. To kick us off, could you talk a little about your background, how you got to law, and how you got to law libraries? Tell us about that journey.
I had the opportunity to hear the reference librarians interact with pro se patrons and see the transformational effect that had on the users.
Michelle Wu: There is nothing in my background that drew me to law. I more or less fell into it. I had attended the University of California San Diego with the intention of applying to medical school, not so much out of desire, necessarily, but more out of expectation. My focus shifted when one of my roommates started studying for the LSAT. As odd as it sounds, I found myself much more interested in the logic puzzles in the practice tests than in organic chemistry, so I ended up taking the LSAT and applying to law school. I was offered a full scholarship, so ultimately it was more the convenience of an opportunity than a conscious choice!
Once I was there, I was fascinated by the law and its nuances, but I knew pretty soon that I didn’t want to practice. The only practice that captured my interest while I was in law school was juvenile justice, and all it took was one externship to realize that most of the problems in that area couldn’t be addressed effectively by legal actions.
In the meantime, I’d been working at the San Diego County Law Library. While my job was administrative—shelving, collecting fines, and the like—I had the opportunity to hear the reference librarians interact with pro se patrons and see the transformational effect that had on the users. Just discovering there were resources that could help them make a difference changed their demeanor and their outlook. There were patrons who continued to come back to the law library long after their legal dispute ended simply because they found open minds, a welcoming space, and information they had never imagined would be in our collection. That’s what led me to librarianship. I wanted to be an agent of that type of change, so I went to the University of Washington’s law librarianship program and came into librarianship in 1995.
Kennedy: You entered libraries, so to speak, right as the internet became generally available in society—and right at that initial transformational moment for libraries regarding computer technology. Between 1995 and 2019, how would you assess this ongoing transformation? And how does that tie into what originally got you interested in librarianship—the notion of a library as a welcoming space that offers resources and connections to a community? Has this changed with technology? Does technology make it easier or better or worse?
Wu: It makes things both better and worse. In terms of how it makes things better, when I was in library school, Mosaic, the first graphical browser, debuted. That made the entry into librarianship fascinating because every year I was doing something different. What kept me in librarianship was just the amount of information and change that was available, particularly around issues of accessibility. The possibilities were endless. As much as technology can open up access, however, it can also be used to distort information. We’re seeing a lot of that already, and not just in libraries, but in the wider environment, whether it’s “fake news” or the manipulation of data.
But, weighing the good and the bad, I certainly think the positives outweigh the negatives. Technology made the open-access movement possible and created new avenues to get information. For example, basic reference questions can be answered and verified through a search engine and the links that it provides. If you’re looking for a president’s birthday, you just type in “George Washington’s birthday” and that information will come up. Another example is what we’re doing at Georgetown: we’ve started a controlled digital lending pilot, a project that allows us to share books digitally—not the books that we’ve licensed but books that we’ve taken from our physical collection and digitized. None of this would be possible without technology.
As much as technology can open up access, it can also be used to distort information.
Kennedy: Let me go back a bit, and this ties into the open-access question and the idea of digital lending libraries and that spark that you saw in the San Diego Law Library. What do you see as the role of the library in the open-access movement? And how does that connect to that spark for the user, the researcher, or those pro se litigants who are trying to get help? What does it mean for the physical library as the place they seek out for that information?
Wu: The pro se part is much harder to answer because there is a divide in terms of not just wealth but education and access to technology. We need to keep this in mind as more libraries move toward technology so that we don’t worsen those divides. But I can say that I’ve seen that type of spark with our researchers. They are able to find information that was not easily findable before. For example, I hear from a lot of historians who say, “You know, this project I’m undertaking now wouldn’t have been possible 10 or 20 years ago. It would have been too daunting. The resources were scattered, and I wouldn’t have been able to get to them.”
I also see it in terms of being able to share information that is otherwise inaccessible. For example, Georgetown has created a Foreign Intelligence Collection. There is no other database that provides as comprehensive a collection on foreign intelligence law and national security information. It was built because we had a faculty member here, Laura Donohue, who worked with one of our librarians, Jeremy McCabe, for years to collect information, including through Freedom of Information Act (FOIA) requests and tracking down materials held in nonlibrary collections. From that, we made a single database for all researchers and practitioners to access, and we see the spark both during the initial acquisition of the information as well as when it is used by those who otherwise would have no access.
Kennedy: Do law libraries have more of an imperative around open access than other kinds of libraries? Do we have a deeper obligation? For instance, open access comes up a lot in the context of social justice, and sometimes people look to law libraries because of this social justice component. Is there something different about law libraries and open access than maybe a public library, an art library, or another kind of special library?
Wu: Yes, and I think there are two reasons why there is a stronger imperative for law libraries. First, we’re dealing with government information, access to which is essential to a meaningful democracy and a fair justice system. If the only people who can access legal information are those who can afford to buy access, that skews the outcome. Second, a lot of law librarians—not all, but many—have law degrees. We have the ability to understand when laws are trying to restrict access and the knowledge to push back against them.
Kennedy: When we think about access to government information and the ways that people are using our collections as we make them more available, I start to think about privacy concerns. As people are using our collections online, the ways in which metadata attaches in your browser to the activities you’re engaging with or the way license agreements are structured around search tools, the information that vendors are collecting is tremendous. What role do you see the library playing in that space of privacy protection? Obviously libraries have always had a strong imperative to protect the privacy of our patrons. Can you talk a little bit about how that may be becoming more complex? Is it different in a digital environment than it is in a print environment?
We have the ability to build robust, broad databases of information where we could control data collection and protect privacy in a way that we cannot with licensed databases.
Wu: Privacy issues are much more challenging in a technology environment. The amount of information that can be gathered is a bit frightening when you think about it. Worse, because so much information is now licensed instead of owned, some libraries feel they have to sign away users’ rights in order to obtain access to unique information. Our role needs to be, first, pushing back and trying to get better licensing terms and then, second, to start building our own databases. This is where open access comes into play. If we are the creators of those databases, we get to control not only what data is gathered but who has access to it.
Kennedy: Do you see a will within the library community to create those databases?
Wu: Yes and no. If you’re saying would the library reproduce all of MOML, JSTOR, or any other existing database—probably not. Maybe there will come a day when the technology becomes so inexpensive, easy, accurate, fast, and automated that we can do it. Right now, we’re not there. But I do think the potential is there. Libraries could distribute the digitization of materials to avoid duplication of effort and then share the results such that complete databases could be built and accessed by all. We have the ability to build robust, broad databases of information where we could control data collection and protect privacy in a way that we cannot with licensed databases. The possibilities are there, and some of us are doing it independently—Harvard did it with its scanning of case law (for more on this, see “Making the Law Computable”). So we can do it. The possibility is there.
Kennedy: It’s a question of gathering the will and perhaps the network of a broader community—maybe people reading this article!—to force that kind of activity. Switching gears a bit, could you talk a little bit about what you perceive has changed about legal research with this deep reliance on electronic material? Do you have a sense of where you see legal research moving in the short, medium, and long term?
Wu: From my perspective, research has become both more careless and more thorough. In terms of the amount of data that’s out there, people who are interested will be able to dig deeper and longer into any particular topic than they have been able to before. At the same time, I also think that users are less proficient at searching, both because it’s so easy to search without understanding what’s being searched as well as because what is displayed to you is much harder to assess in terms of quality. If someone doesn’t know how to how to assess the quality of the information that they are getting, or the quality of the source that is providing it, their research risks becoming careless.
In the short term, there’s going to be a greater focus on how to verify sources and assess the legitimacy of any information obtained. With the rise of fake news, that’s where legal research instruction is going. In the medium and long term, one place where I see change is developing more effective ways to search across platforms and seemingly incompatible data sources. Search is complex, and many users don’t know where to search for what. We need to find a way to make it easier for people to get to the best database for their topic—something like an effective reference interview for the public. Either that or we have to identify the most complete list of quality sources as possible and make them simultaneously searchable, which is more challenging but would be much more useful to researchers. The second change is unquestionably the rise of artificial intelligence. It’s already in Westlaw and LexisNexis, but it’s going to become more sophisticated. We will start seeing it more as an active partner in research, particularly in areas of big data and data mining.
If you’re talking about strategy, including understanding human behavior, I don’t think that that’s a strength of AI right now.
Kennedy: I was just at a talk this morning about artificial intelligence and libraries. Do you see a role for AI in the regular function of research? I’m personally skeptical about the role of AI in some types of legal research, like practitioners preparing for litigation. But in the contractual and transactional space, I can see AI serving a positive function. There is also concern within the legal profession that lawyers will be replaced by AI. Some of our courts are even using AI for sentencing. Do you see AI as being a viable replacement for litigation research in the short- to medium-term versus transactional research?
Wu: There are three different responses for the three different categories of things that you’ve mentioned. For transactional, I do think it could work. I talked to someone at Westlaw a year or two after they had debuted WestlawNext, and they said that in their surveys of practitioners, lawyers with access to the new search interface no longer felt they needed to spend as much time researching. They were finding what was relevant—or at least what they thought was relevant—right away.
Litigation, however, is as much strategy as it is data. And if you’re talking about strategy, including understanding human behavior, I don’t think that that’s a strength of AI now, and I don’t see it becoming a strength anytime in the near future. It just requires a more complex understanding of the elements involved than a data-based machine structure can comprehend.
As far as judges using AI for sentencing, studies show that computers can be biased, too—they share the biases of their programmers. Unless there is a way to correct for that, AI will be unequal to the task. It has been made abundantly clear through various studies, including those on facial recognition software, that AI can actually reinforce biases.
Kennedy: Do libraries have a role to play in teaching and training artificial intelligence? Is there space for libraries in the development of artificial intelligence? Is there a space for law libraries in particular as AI becomes more prevalent in these massive commercial databases?
Wu: Absolutely, though as AI gets more advanced, it will become harder to predict. Currently, AI does not know what a quality source is. It might know that more users have clicked on a certain publication, but librarians are better able to actually assess the substantive text than a machine. For that reason, we can train any AI that is designed to respond to users’ needs. We can help with metadata or order preferences, and that help can produce better search results. For example, I always thought that Google Books should have partnered with librarians. They had partnered with libraries for the collections, but they did not partner with the information specialists who could have looked at and improved on the metadata they were attaching to those records. Many of the complaints they received afterward about search problems could have been avoided through that type of partnership.
Kennedy: As the director of the Georgetown Law Library, how do you talk about the importance of libraries and, perhaps even more important, the importance of library staff when you’re talking to various members of the legal community—your law students, your faculty, the scholars, the practitioners who are going to be reading this article, and beyond?
Information is power, and libraries stand as the safeguard to the loss of that power.
Wu: This is where it’s really hard. I could expound for hours on the importance of preservation and access to information and the dangers of losing either, but this is often too abstract a concept to make an impact for most populations. Depending on whom I’m speaking with at the time, I usually try to find a story about something our library has created or has helped to accomplish—a story where the library’s role made the impossible possible or the invisible visible.
One general example is what law libraries did after the earthquake damage in Haiti. The country lost everything, including its historical, political, and constitutional documents. Without help, they would not have been able to reassemble their own history, and through a cooperative effort of law libraries and LLMC Digital, libraries were able to provide Haiti with access to a comprehensive collection of their own country’s legal materials. No library would have been able to do it alone. It was really the collection, preservation, research, and organizational missions of libraries together that allow that collection to be rebuilt.
My broad position is this: Information is power, and libraries stand as the safeguard to the loss of that power. On the one hand, libraries are the world’s insurance that the information necessary to democracy and our justice system will become and will remain known, accessible, and instructive. On the other hand, libraries actively search for and make available information relevant to the work and interests of current researchers, practitioners, and society. The Foreign Intelligence Collection that I mentioned to you earlier is an example of this—those are materials that will enhance the practice of law in the area of national security.
Kennedy: The imperative is really for libraries to find their niche—the thing that they can help to collect and curate, like the Foreign Intelligence Collection that you have assembled, that provides that level up. I was at a talk just yesterday where someone likened all the content on the internet to your thumbnail while all of the world’s possible knowledge would take up your entire hand. And what I like about the story you just told is that there is more information outside of the internet than there is inside, even though we’re developing and creating content at a faster rate than we have at any other time in the history of recorded information.
Wu: I would add that what is available, whether online or not, may also not be organized in a way that is accessible or findable. Libraries are a bridge between raw information and usability. A number of the documents that are in our Foreign Intelligence Collection database would not have been tracked down if not for the work of the library. So it’s not just that we can build a database or that we can gather the documents—it’s that we have the staff with the knowledge to be able to track those materials down and make them visible.
Kennedy: In addition to leading a library, you are also a copyright expert. In thinking about open access and authors’ rights, how do you view the library’s role in that space, particularly in online space?
Wu: In terms of authors’ rights, the interests are diverse, depending on the category of authors and maybe even the individual author, because the purposes of writing change from person to person. In the academic sphere, most academic writers want their works to be read, they want them to be influential, and they want them to be accessible. In those cases, it is easy to balance the rights of the authors with open access because both share a similar intent. Where the author retains copyright and has the interests I have just described, libraries can facilitate open access without any friction with authors’ interests.
It’s very possible there will be more of a collective virtual law library instead of individual structures or collections.
The analysis changes completely where copyright is not owned by the author or the author’s objective is profit. If an author’s goal is to make a profit, that runs against open-access principles. Open access generally means free access for everyone. In theory, where the library is the publisher, it could say that it would make a document available to everyone but find a different way to pay the author, but that is a challenging proposition for entities that are nonprofits.
Kennedy: That’s the distinction—whether open access is free access. The monetization of information in the digital space might be problematic, but it will be interesting to see if any of that shifts with the University of California’s decision to cancel Elsevier.
Wu: I would love libraries to move into the publishing sphere. I know that we don’t necessarily have the right personnel at this time, but if we’re talking about peer review and libraries managing the peer reviewers and the managing publication—for academic works, I actually think we could do a pretty good job.
Kennedy: Looking ahead, how do you see patrons—scholars, legal professionals, law students—interacting with the law library of the future?
Wu: I actually think they will look to us for most of the same things that they do now, whether it’s help in finding a resource, a reference interview to help them determine what it is that they need, or instruction. But that contact will probably be more virtual than in person, and the assistance or instruction needed might be slightly different—they might need help learning how to datamine instead of how to do an advanced Boolean search. It’s very possible there will be more of a collective virtual law library instead of individual structures or collections. How people access information may very well change. And while there is value to the actual artifacts of books and there will remain some spaces to preserve them, there may not be as much of a need to have as many separate physical spaces as we have now. I’m not talking about public community libraries, but in academia it’s possible that instruction goes virtual and that academic libraries will similarly go virtual and will serve the student or researcher wherever they sit.
Kennedy: We talked a little bit about the ways in which, with so much information, the searches people are doing are returning what they perceive to be relevant results, but those results actually depend greatly on the specificities of the particular search parameters and may risk creating a narrowness of understanding. This is a different experience than that of the lawyer who researched in print, had a sense of the scope of the subject area, and could in some way more affirmatively and confidently say, “I found all the answers!” What do libraries do about this information literacy challenge? That is, our newer users—our incoming law students and maybe even people who are currently in high school who will eventually be in law school—approach information in the context of the vastness of the availability of it.
We have the ability to collect and make available much broader collections to all of our users than any one of us can do by ourselves.
Wu: It’s very troubling to me, and I do think that the change is tied to a shortness of attention span and how information is displayed. When users researched only in print, they actually had to read materials all the way through. There were no shortcuts because they did not know if what they needed was on page 20 or on page 220. Searching online today, a user could be presented with the exact sentence that they want, without realizing that the book in which that sentence sits completely disproves that one sentence. What do we do about this? It’s a great question. I’m actually not sure that libraries can counter this, at least not by ourselves. Ultimately, it will come down to judges and what courts do, because if they are satisfied with and will rely on poor research, then there is no incentive for students or lawyers to do better. But if judges, or their clerks, start unearthing information that they think lawyers should have found, or if partners start seeing this same problem in their associates, then there’s a penalty for sloppy research. Until there is an incentive, libraries can teach the best research methods possible—the ones that will lead to the best answer—but behaviors will not change.
Kennedy: I know you teach students as well, and one of the most difficult things I have had to convey to my students is that they have to actually read the cases. Just because you found 10 results in those cases and you key cited them so you know it’s still good law—you still have to actually read the case. We’re already in that slippery slope phase. We need the judges and the lawyers who are practicing now and who remember the thoroughness and the depth with which you needed to look at information to pull some of our newer researchers back. Given all this—and all that we talked about—what are you focusing on right now that you’re most excited about in your library and in your work?
Wu: Our story about controlled digital lending (CDL) is just getting started. Essentially, CDL has two drivers. First, many of our collections are not available online and are not being found by our users. Second, as libraries either get cuts in their budgets or as users want more information than they can afford, there is a risk that all will cut the same materials from their collections, reducing the diversity of information acquired and preserved for future generations. Both of these translate to less information being available or usable to our communities, whether through the inability to find it or inability to acquire it. CDL has the potential to address both of these concerns, while respecting both authors’ rights and the public interest, through digitizing physical works and circulating the e-version in place of the original. Digitization obviously takes care of making materials visible to an increasingly digital-focused population. We meet the other interest through CDL and partnerships. By distributing collection development and digitization across many libraries instead of tackling it individually, we have the ability to collect and make available much broader collections to all of our users than any one of us can do by ourselves. Ultimately, the hope of CDL is that it can make everyone better by giving access to more information to all researchers.
Kennedy: That sounds absolutely fascinating, and we will certainly be looking out for it in the future. Michelle, thank you so much for taking the time to talk with us.
Michelle Wu is the associate dean for library services and a professor of law at Georgetown Law.
Jocelyn Kennedy is the executive director of the Harvard Law School Library.