Never Neutral, Never Alone

“Never Neutral, Never Alone.” Transforming LIS education for professionals in a global information world: digital inclusion, social inclusion and lifelong learning IFLA Satellite Conference. Vatican City (via video).

Speech Text: Read Speaker Script

Abstract: Library science is getting harder to teach. The variety in libraries of all types is increasing as more and more mold themselves to their communities rather than field-wide norms. How can library science education change to meet the new variety, and the variety in a post-neutrality world.

Audio:

[This is the script I used for my talk. I’ve also taken the opportunity to add some foot notes and links.]

Never Neutral, Never Alone

August 22, 2019

It is time to have a frank conversation about LIS education. The problems with how we prepare librarians are often phrased as a gap between theory and practice. The argument goes that library schools are not producing graduates with a real-world practical skills; instead focusing on generalities and theory. This is a perennial argument, and if there was a library school in ancient Greece, I’m sure Dewey’s Socratic equivalent would be criticized for not preparing students to argue effectively in a marble building as opposed to a brick one.

This theory/practice gap, however, is not the real problem. The real problem is that no one knows what new librarians need in the second year of their career, much less their 25th. There is no common entry point, because there are fewer and fewer commonalities between libraries. As libraries of all types are organizing themselves around the local needs of a community – be it a town or a university or a school or a hospital, the differences in working environments for librarians is changing not only quickly, but diversely. What once was applying a standard set of reference skills to an owned set of databases, or applying cataloging skills to local classes and codes, is now about community outreach librarians knowing the unique culture of a city, or a user-experience librarian learning the realities of undergraduates in a particular school at a particular time.

The libraries that we hold out as global exemplars like Dokk1 in Aarhus, or LocHal in Tilberg, or San Giorgio in Pistoia, or the libraries at University of Michigan or there at the Vatican with its petabyte data center and global digitization initiatives are as diverse as they are impressive. No one school can prepare all starting librarians for all libraries. This doesn’t even consider the inclusion of archives, special collections and research services that are not even connected to traditional library institutions.

The standards and competencies we develop will continue to become more general, and more focused on lifelong learning and community engagement areas. Where once we could define cataloging skills down to the standard, we now must recognize that information organization can take the form of MARC, RDA, FRBR, Dublin Core, or just general concepts of the semantic web. Theories of classification still apply, and still must be taught, but the specific skills that accompany these skills are now purely illustrative. Where once we taught reference as a series of genres like atlases, and encyclopedias, today we teach learning theory and pedagogy. These are important areas to teach, but they will never meet the mark of first year practical skill.

Before I jump into thoughts on addressing this situation, let me say these are good problems to have. The reason there is no canon of skills is that librarianship is a vital and dynamic profession. The reason there is so much diversity in the field is because the need for librarianship is growing. The communities we seek to serve are becoming more diverse and varied because we are at least attempting to go beyond real barriers of class and race. If all we were doing was preparing spare parts for a handful of libraries that hadn’t changed in decades, our stable and satisfying curriculum would be the surest sign of the impending death of libraries.

No, the answer is not to try and develop a single standard for all, but to create continuous systems of learning that are agile, connected, and embedded. The library education of tomorrow, and increasingly, today, must smash the divide between the “real world” and the “academic.” It must also break the idea that one degree at the outset of a career is sufficient preparation for an entire lifetime of serving a community. Lastly, it must also fully embrace that we are preparing librarians, not library workers. And accept that librarians are not neutral, and must develop skills that are as much about resilience and self-examination as they are about how to run an organization.

Let me take these ideas in turn. I’ll begin with agility. What is an agile system of library education? It is one that is constantly seeking out not only best practices in librarianship, but innovative ones. It develops a curriculum and means of delivering that curriculum that are flexible and can be deployed quickly. One example of this is in Norway where the Akershus University College of Applied Science’s Department of Archivistics, Library and Information Science holds a biannual conference for its alumni and other librarians. It is a chance to not only bring in the latest thinking from the field, but to connect and listen to graduates and what they need.

At the University of South Carolina, we are pairing every library science degree with a specialized certificate that documents areas of focus such as data science, health information and so on. However, we have structured the certificate so that the specialties can change from year to year. We see students getting certificates in artificial intelligence and librarianship, library construction and design, and service to refugee populations. The list of specialties will be long and change year to year, student to student, as the world these librarians seek to serve changes.

Which brings me to my second new “standard” for library science education – connected. I would love to say my faculty represented hundreds of specialists all expert in the latest develops in the field. They do not. They are scholars with specialties and a broad view of the field, with an ability to connect practice with larger concepts. However, our alumni and the institutions they work for, and that we partner with, do represent hundreds of specialists developing and deploying innovative services in communities across the globe. Library schools must be a part of creating a network of libraries directly engaged in the education of new librarians. 

This goes well beyond a set of adjuncts who teach a few classes, or internships, or field trips. We must develop a network of libraries that share both in the responsibilities of education and the funding of such systems. The library science school of tomorrow is truly a hub that delivers a core of library concepts and research skills, and then connects students with developing innovations in the field. Your faculty may be on the tenure track or working the reference desk. Your mentor may have the tile of professor, or librarian, or archivist, or programmer. The hub ensures rigor in the learning, but more importantly ensures cohesion in a student’s degree.

The dynamism in the library profession can be clearly seen in the enormous offerings of professional development. A librarian could spend a week just sitting in webinars and online workshops in just about any aspect of the library profession. Our library associations, our vendors, our universities, our publishers, our libraries are in the midst of an amazing creative rush of developing online education. However, there are no real attempts to coordinate and link all of these together into a coherent understanding of the field. Faculty in the library school of the future will spend as much, if not more time evaluating portfolios of these diverse online resources as they do teaching classes. The days when the expertise of a field was contained within a single library school are gone. The days when the totality of library expertise could be represented in a single faculty are gone. 

We must look to other models of how we prepare professionals, hence, “embedded.” That network of libraries and expertise we build must also be seen as places for residencies where we embed students for direct, contextualized learning. The advent of online education has made place irrelevant in many of our programs. You no longer have to move to Columbia to get our degree. However, in making this shift, we have also lost the power of place. We must now join the power of place with the flexibility of online. 

Students will no longer move to Columbia because that’s where the faculty are, they will move to Aarhus, and the Hague, and Taiwan, and Charleston because that’s where innovative practices are being formed. Taking a page from the medical residence, we are turning our network of partners into residency opportunities for our students. Libraries can use these residencies to attract the best new librarians to job openings, and the students gain authentic specialized knowledge on top of the core we provide. And hosting these residencies is an opportunity to expand the learning of the students to the learning of the whole organization.

In Charleston South Carolina, the local school district pays for 10 in-classroom teachers to get their master’s degrees and become school librarians. The funds for these cohorts are then re-invested in the school district. The tuition of the students pays for national speakers, onsite workshops, even open course development that are provided to the entire district. This creates a sustainable means of continuous library education well beyond the granting of a degree. By enrolling 10 teachers, the district enrolls the whole district in library school.

And what are these students learning in their residencies and in the network? They are learning to be librarians. Not people who work in a library, but a set of values, research skills, and a mission they will take with them to jobs in libraries, or the technology sector, or the banking sector, or government. They will be going into these libraries, and businesses and governments a point of view. They are not neutral deployers of skills, they are professionals on a quest to improve communities through learning. They will go not as parts of a system, but as advocates for inclusion, privacy, access, and openness.

In order to prepare these librarians, we must develop a curriculum of self-reflection and analysis. We must address, in the curriculum, self-care, vocational awe, resiliency, and self-awareness. These are not soft skills, but techniques that allow our librarians to assess, engage, and adapt to community needs and realities. It is no longer acceptable that we send out librarians into communities prepared to answer reference questions, but unable to process the poverty they may find there. It is no longer acceptable to train academic librarians to recognize gaps in the collection, but not to recognize student homelessness. It is no longer acceptable to train archivists who do not understand the politics inherent in controlling the memory of a community.

Analysis cannot be limited to the individual and introspection, however. Methods of analysis – of research- are necessary. No matter the environment our new librarians find themselves, they will need to know how to understand a community, how to assess services, how to collect, analyze, and protect data. Participation is a goal, and we shall never know how well we are matching that goal without instruction in research methods – instruction that is embedded in real communities with real questions, and contextualized methodologies. 

And so these are my new metrics for evaluating the effectiveness of a library science program:

Agility – what ongoing methods are in place to identify, evaluate, and prepare students for developments in a rapidly changing profession?

Connectedness – who are the partners networked with the program and its faculty to ensure direct connection of the classroom to the field?

Embeddedness – what are the program’s ability to deliver authentic field experiences to students that allow them to contextualize theory and research methods?

Resiliency – how prepared are librarians to face, understand- that is analyze-and solve the problems in a community in line with the professional mission and values of librarianship?

Today the librarians we prepare are building makerspaces, they are crunching masses of data in civic redevelopment projects, they are saving tweets for posterity, and housing masses of research data. Our graduates are delivering knowledge and food to rural communities left behind in an information economy. They are supporting the research of Nobel laureates and citizen scientists fighting for clean drinking water. They are fighting for access to the world’s knowledge in developing economies and bring dignity to marginalized communities. They need a strong platform to prepare them for this work and then support them throughout that work. Library science programs can be that foundation, but not alone. We must connect the innovative librarian stifled in a large bureaucratic library with an innovative librarian revolutionizing a small town a continent away. And connect them both to scholars and the means for continuous learning.

Library schools are a vital part of the reinvigorated library profession. Yet, just as we have seen the road to success for libraries is in adapting to and including the community, so too must our schools become open platforms orchestrating participation and adapting to the community of our alumni.

Thank you.

The Knowledge School: or Why Teaching Library Science is Getting Harder

“The Knowledge School: or Why Teaching Library Science is Getting Harder.” École de bibliothéconomie et des sciences de l’information, Université de Montréal. Montreal, Canada.

Abstract: A school of thought represents a shared set of approaches, beliefs and values shared by a diverse set of players. A prime example is the Chicago School in architecture that wasn’t a department, but a shared vision of architects, engineers, and city planners. In this presentation, Lankes discusses the merging knowledge school and how it is shaping the field globally.

Slides: Slides in PDF

Neighbors Not Users, Members not Customers, Partners not Patrons

“Neighbors Not Users, Members not Customers, Partners not Patrons.” Congrès des professionnel.les. de l’information. Montreal, Canada.

Abstract: It would be easy to see the advent of open educational resources, open access publication, and repositories of data sets as a continuation of the traditional mission of a research library. Namely, providing access to the scholarly record including items studied as well as the results of study. It would also be easy to see this as happening in parallel to a pivot of libraries to more community centered models. In this presentation Lankes will show how these developments are deeply intertwined in how we conceptualize scholarly communication and the need for advocacy around data in all aspects of higher education.

Slides: Slides in PDF

Audio:

Data, Media, and Society

“Data, Media, and Society.” University of Maryland Libraries Future of the Research Library Series. College Park, MD.

Abstract: It would be easy to see the advent of open educational resources, open access publication, and repositories of data sets as a continuation of the traditional mission of a research library. Namely, providing access to the scholarly record including items studied as well as the results of study. It would also be easy to see this as happening in parallel to a pivot of libraries to more community centered models. In this presentation Lankes will show how these developments are deeply intertwined in how we conceptualize scholarly communication and the need for advocacy around data in all aspects of higher education.

Slides: Slides in PDF

Audio:

Library as Movement

“Library as Movement.” Victoria Libraries 2019 Planning Summit. Kalorama, Victoria, Australia. (via video conference)

Speech Text: Read Speaker Script
Abstract: Is the library a platform that allows communities to build knowledge? Is it a community center? A community hub? The concept of a library is evolving faster than our terms for describing the change. Today the library is being seen as a movement. Not a place, but a community-wide effort to improve the lives of community members through knowledge. This presentation talks about how the continued evolution of librarianship toward a proactive force for good in a community changes the skills of librarians, the responsibilities of the community, how we assess success, and how we determine common services while focusing on local realities.

[This is the script I used for my talk. I’ve also taken the opportunity to add some foot notes and links.]

Greetings! Thank you for having me, albeit via video. I very much wish I was with you in person. Today I’d like to talk about how our changing concept of what a library is directly effects our ability to plan for shared network services. In particular, how a shifting concept of the library as a community hub and emerging social movement requires a different approach to shared services. We have to take on the very real change of a library as a set of resources for a community to access, to the library as a supporter of building knowledge locally. 

While I could jump right into my recommendations for shared services for the next decade (hint – its more about people development than collection development), my recommendations might seem pretty arbitrary unless I set up why we need to change in the first place. I promise not to get too abstract here, but we need to start with some pretty basic assumptions. First, what do we mean by the word library anyway.

One could say I have a difficult relationship with words. In particular, I tend to get a bit obsessed on definitions. For example, in 2007 the then dean of Drexel’s ischool David Fenske and I were talking about libraries. He made a comment to the effect of “the field of librarianship will be held back until one can define a librarian without reference to a building.” In essence, what is a librarian without a library.

Four years later the direct result of this comment was the 400 plus page book the Atlas of New Librarianship. I think we can all agree, a bit of an overreaction. But here’s the thing about words like library and librarians. Once created, they rarely lead a stable life. Their definitions and usages evolve.

Fenske’s question about how we define what a librarian is began an intensive investigation about the term library that continues to this day. It began me thinking about how we define and conceptualize libraries and the work of librarians. You and I are on a journey of definition. That doesn’t really surprise you. After all, you keep having this meeting for the simple fact that what we do, and why we do it – our definition – changes. It changes for our communities as well, but they often aren’t aware of it. Our communities haven’t been deeply enmeshed in debates about the “library as a community hub,” or “members versus patrons,” and so the words we use to describe ourselves, our services, and even the terms we for them can be jarring for them, and it is dangerous for us if it is jarring for them.

To have a bit of fun with this let me quickly present the shifting ways of thinking about libraries as a sort of progression through evolutionary eras. To be clear, this is not a rigorous chronology, but rather a way of highlighting how we conceptualize libraries, librarianship, and the services we offer – and therefore the network services we need. For the purposes of this talk, these eras start out as more cartoon and stereotype.  However, I believe it is a helpful way of thinking about and planning for shared library service.

I would add some nifty Latin terms to these eras, but, well, I failed Latin. So we’ll begin with the Era of the Book Palace.

The start of the book palace epoch is a bit hard to nail down, but it arguably defined libraries in the western world for well over three centuries from the sixteen hundreds onward. But for our purposes we’ll pick on the Dewey era turn of the 20thcentury.

It was a time when collecting books was vital because they were scarce. The great value of the library was in pulling collections together, and the vast majority of the books gathered were about the rest of the world, not the library’s service community. It was a time of grand architecture. It was also a time of the universalists and documentalists. That is, folks who believed that knowledge could be contained in the pages of a book, and that the knowledge of the world could be sorted into neat categories…ignoring that those categories were developed by and for a culture dominated by white guys. The king of the information world was the book, and libraries were an apex predator in the information ecosystem.

One of the ways you can see how we and others thought about libraries is in looking at the relationship we had with “them.” Them, as in, what did we call the folks we served? This was the golden era of the patron. Patrons supported the library, they received service from the library. They were also nearly anonymous. We didn’t spend a heck of a lot of time defining patrons because we were mostly the only players in town, and our value to our patrons was implicit.


Just as we look at the tools developed to demark epochs of human evolution – the stone age, the bronze age, and so on – we’ll do the same here. What were the defining tools of the book palace, particularly for networks of them? Union catalogs and interlibrary loan. With the dawn of the computer age we could connect our collections together. The work of consortia and library networks was around standardization and efficiency. We sought to use the same cataloging standards; we had closely aligned management structures. We were linking similar to similar. The role of the network was to create standards and to share materials.

So, what brought about the end of the Book Palace Epoch? Networks. Not computers – we’ve been putting up online catalogs since the 70s. No, it was that we could now network not only records, but digital files as well. We needed to get into the full-text business. We needed to get online, we needed to change.

So, libraries became information centers. We were no longer just a place with stuff, we were a gateway to the world. And that world was information. Gone was the quaint Victorian era concept of the patron, in was the modern user – a term freshly taken from computer scientists and drug dealers. Now it wasn’t about having it all, it was about getting to it all. 

And it if wasn’t online, then by God it better be. This was the era of mass digitization. Google and libraries hooked up to scan the world. I remember talking with a Harvard librarian at the time who said their main obstacle in digitizing materials was how to get trucks onto Harvard yard to take away the books and turn them into bits flying free. 

The scanner and the contract were the defining tools of the information center age. If we couldn’t scan it, we would license it. Database, ebooks, video services – the drive was to expand the collection with resources from around the world. In our drive to provide users access, we also transformed the very nature of collecting. Gone were the days of owned materials being ferried around the countryside in delivery vans (well, not gone, but let’s just say we didn’t put them on the postcards anymore), in were Dublin Core and metadata schemas to build towering virtual libraries.

Gone also were the days of budgets being strained to buy materials one time. Now we had to devote budgets to paying for access to a resource annually – a change that is now once again coming back to haunt us with terms of ebook lending. We also spent a LOT of money on public access computers. 

Our collaborative services? Digitization support; shared and state-wide licensing agreements; metadata schema development; and training to build the killer website.

A funny thing happens when you move from patrons to users, and from collecting to accessing. You tend to move from relationships to transactions. Instead of telling the story of the library in outcomes for our communities, we begin to quantify ourselves. Now instead of just counting the volumes in our buildings, we emphasized hits, circulated items, attendance, and of course gate counts. 

So what pushed us out of this era of libraries? Simple, we lost our monopoly.

Now to be clear, libraries haven’t been the sole source of information and access since, well, ever. Though we did have a lock in medieval Europe until Gutenberg went and screwed that up. But we at least had a large portion of mind share in our communities. With the advent of ubiquitous networks like the Internet, and the ability to monetize access, mostly through advertising, our portion of the mind share shrank.

We needed a new way of thinking about libraries and librarians and our value to communities. We didn’t just invent these out of thin air, rather we saw non-access and non-collection activities in a new light. We saw that the value we provide to the community was in the community itself. We became the 3rdSpace, and instead of users, we had citizens or members. 

Our focus wasn’t on collections alone, but on being a place where community members could come and think and work with, or without, those collections. Our newly emphasized focus was on civic improvement. We helped folks find jobs. We provide vital literacy services to youth and adults. We were a safe place to explore dangerous ideas. 

And what tool helped define this epoch? The Library Café. Yes, the café as a literal place to serve coffee, but also the numerous spaces where we pulled down the stacks, or never built them in the first place to allow folks to get together. We called them living rooms, or agora, or simply “the teen space.” Many cities rebuilt or refurbished central libraries to promote economic development. We began hosting co-working spaces

Our consortia still paid for licensed resources and we still shipped materials around. But now, our joint services began to go a bit adrift. How do we collectively support what is by definition a very local thing?

This is also the time when our communities began to get very confused. Sometimes that was phrased inelegantly as “why do we need libraries when we have Google?” It was when our communities began wondering, what is the difference between a library and a community center?

It was also a time when we got very good at posts on Instagram. Because while we had a hard time putting our contributions into words, we had no problem showing the growing number of diverse faces coming into our buildings. Our identity became more diffuse, and more local in nature. But it was the seeking for an identity that lead to our next era, though it’s more a later part of the 3rdspace era. But for now we’ll call it the era of the community hub.

We began to put words and concepts to the third space – but as often happens, we were better at saying what we weren’t as much as what we were. We weren’t a community center as in an open meeting room. We weren’t indoor parks with books. We were a learning center and community hub. Our members became learners, and our focus rested squarely on the community creating its own knowledge and identity. Our tool of preference? The Makerspace. 

No not just 3D printers in a room, but the idea that the community could come together and create in a library. For some librariesthe maker space is 3d printers and hand tools. For others it’s a wide-open living room for group chats. Still others it is the marked spike in programs where community members teach fellow community members.

In libraries across the globe video and audio studios began to pop up. Those scanners we once used to digitize the materials of the library were turned loose on family photo albums. Our walls were pulled down for workshops. We looked into the eyes of the Smart City and claimed the smart citizen turf. We loaned out baking pans with our books and even had cooking classes to boot. We not only paid for Kanopy, we created our own YouTube channels. We talked about great libraries building communities and the communities as the true collection of any library.

The Tilburg public library in the Netherlands took over a former train maintenance warehouse and built pop up libraries right next to incubators for new business start-ups. They used these pop ups as places of experimentation and play that ultimately lead to the impressive LocHal. The IP Centre at the British Library moved the business reference books to the side and retrained the librarians as business planning experts.

The era of the community hub was, well, is a reaction to the retreating human interface to government. Our members could no longer talk to a person with questions on taxes or social services. The face of health care went from a doctor or a nurse, to a patient portal. Into this vacuum stood librarians ready to help. And to support them, social workers. And to support them all artists and writers in residence. Instead of giving the books the best views from our new libraries of glass and steel, we created a destination. Our value was now in quality of life.

I would say many of us are living firmly in this Community Hub Epoch. We are, however, already starting to see the need for continued evolution in this approach. In the UK, for example, too many local councils have seen the community well integrated into the workings of the library and made the jump that the community itself can maintain the library in the form of volunteers. 

In the Netherlands there are no more library schools because community-centered librarianship is being defined as user experiences and customer service versus librarianship and its values and skills. In Florida they are having theme park experts design libraries as an experience, instead of librarians designing libraries as a service. Don’t get me wrong. We should be designing our libraries with the experiences of people in mind. We should be building organizations that serve. But in doing so we must recognize the unique value librarians bring to this endeavor.

That is also not to say that as professionals we are necessarily where we need to be for our communities. Within the library we have to look at ourselves. Are we best structured to serve as a community hub? If evolution happens to ensure survival of the fittest, are we ready? How much of our preparation for librarians cover event planning? How much focus do we put in skills development on tech and collection building, and how much on community engagement and cultural skills? How open are we to the entire community when all too often we organize ourselves in hierarchical management structures where some positions never have to interact with the public?

Which brings me to an emerging era and my attempt to answer what shared services and resources do we need in today’s library landscape. It is conceptualizing the Library as a Movement. It is taking all of this evolution to the next level. The focus isn’t on collections, or access, or places, it is on mobilizing a community for social action. Instead of calling folks patrons or users, or even my personal favorite members, we don’t have a name at all – because the walls between “them” and “us.” Begin to break down. 

Libraries bring together people of diverse, and even clashing perspectives to seek common ground. The greatest asset we have in this era is trust. In a world filled with a cacophony of perspectives, propaganda, and belief, we serve as vital social infrastructure and trusted facilitators working across community divisions to develop a new community narrative. And I know that last sentence borders on buzz word salad, but all it means is we help members of a community find meaning, and power in each other. And in the era of the library as movement, how this happens is going to be different in every library and every community.

In this new era we not only support reading because literacy is a vital skill in making change and democratic participation – we team with the primary schools and the local pizza restaurant to ensure we use common vocabularies and we create a whole culture of reading. In a project for the Hearst Foundation, we created a community literacy initiative. We found that classroom teachers and youth librarians would use the exact same words, with totally different definitions. Literacy for the teachers was skill development for the decoding and understanding of texts. For librarians? It was reading enrichment and the promotion of a love or culture of readers. Parents would take their children from school to library, hear the same words, and not understand the mixed messages they were receiving. 

A literacy researcher on the team was demonstrating how story times could be used to both enhance literacy skills and the love of reading. After the session a mother with her infant child came up to the researcher. “What did you say on that page?” she would ask. Then “and what about on this page.” It took the research a moment to realize that the parent couldn’t read. But instilling reading skills in her child was so important she was going to memorize this book so she could share it with her baby.

That story then led to not only local restaurants having library picked books available for kids, but the city council passing a resolution that every town sponsored event had to have a literacy component. The governor of the state not only signed a declaration about the importance of reding, he recorded a video to be used in all of the schools.

My point is that the community-the schools, the libraries, the businesses, the parents – came together to create change, to create a movement. And the library was part of that movement and could never have done it on its own. And here is the most important part. What worked in Columbia South Carolina will not work in your community. No matter how well we document it, or call it a best practice, or try and turn it into a downloadable toolkit, it won’t work. It is meant to guide, instruct, and inspire you. You, the librarian, your job is to see what will work in your community. That’s the difference from the era of the Book palace. Rather than trying to connect similar to similar – to make a suite of unified and undifferentiated services for all, the networks of today have to train librarians to adapt, not adopt. The network supports and inspires.

Of course, all of these phases of our evolution still co-exist together. 

Take for example The South Carolina Center for Community Literacy – which was the South Carolina Center for Children’s Books and Literacy – which started as collection of award winning children’s books. 

The Center has a collection of both award-winning children’s books and books that help teachers learn about including diverse and representative materials into their lessons. It also has digital services. However, in the past few years the movement aspect of SKILL -as we call it -has emerged. 

We have a bus full of books. No big surprise there. But this bus is also filled with university students from across the campus, and this guy. This is Cocky, the university mascot and something of a celebrity here in South Carolina. That bus? Cocky’s Reading Express, goes to the poorest schools in the state. Those college students? They read to the kids and demonstrate how vital reading is. And the books? Well, that’s where Cocky comes in. You see he gives them out. Over a 120,000 books given to kids where they promise Cocky, a symbol of sport as much as anything else, that they’ll read every day. 

And the center doesn’t stop there. We work with social safety net organizations to do one on one consulting with those in need to connect them to housing, food, and services folks need in their most desperate hours. We work with immigrant groups to advocate for bilingual education. Why isn’t this a community center? Why isn’t this just a bunch of volunteers with book shelves? Because of the answer to David Fenske’s question – it is because librarians with their skills, values, and mission are not simply delivering these services, they are shaping them. Librarians are ensuring privacy in a data driven world. Librarians are ensuring these services are both inclusive in their views, and equitable in their delivery.

And here we come to the meat of the issue: what kind of shared services do we need in this epoch? What do libraries need to support local movements? Awareness, continuous learning, mentorship, and a memory. 

Awareness in looking across libraries, cities, and industries for ideas that can help the communities we serve. What do libraries do in the face of artificial intelligence? How can we best advocate with our communities to put in place safeguards for personal data? How can we better welcome refugees? How can we build platforms for the community to connect people of shared interest? The union catalog of today is the foundation of a community learning management system of tomorrow. A system that directly links a persona with resources, experts, and tracks personal progress in mastering new skills and insights. 

Of course, awareness of an idea and the ability to adapt it to local needs is a very different thing. And I do mean adapt not adopt. Our staff needs to be constantly acquiring new skills. Technical skills, certainly, but facilitation skills, political skills, research skills to implement technologies, programs, and services that look like the community itself.

As librarians develop their new skills and seek to implement new ideas in a community, they need guidance and fellowship. Networks of libraries need to provide mentoring and coaching. We need to develop future leaders and build strong ties among the most remote colleagues.

Lastly there needs to be a shared memory. This memory ranges from classical archives of community development to identifying and highlighting innovation among the group. 

If this sounds a bit familiar…well, it is. The future shared library service is a university of the people. A function that engages librarians and the community players who are part of local movements in learning. Teaching members how to organize collective action. Bringing together different industries together with librarians to forge common goals.

And just as all ideas need to be adapted to local needs, this library university does not have to look like a classical institution of higher education. Don’t build lecture halls, but cooperative laboratories. Teach online, sure, but also learn together in pop up libraries in malls and beaches, and football stadiums. Eschew periodic diplomas for continuous portfolio building. And no class without the partners and community members making this real in their lives.

So there you go, my quick trip through the evolution of libraries. A trip that will never be complete, because we are a living and thriving profession. Our communities need us now more than ever. They need us because we are a trusted place to make sense of a world increasingly wracked with xenophoabia and a flavor of nationalism that seeks to define a nation only as people who think and look like us. Our communities need trusted professionals to ensure not only their rights, but to amplify their voice in the debates on the future.

So there you have it – from an unhealthy relationship with dictionaries to the people’s university. I hope this has been useful, or at least entertaining. I look forward to the conversation to come.

Decoding AI and Libraries

AI

“Decoding AI and Libraries.” KB College: AI en de Bibliotheek – de computer leest alles. The Hague, Netherlands. (via video conference)

Speech Text: Read Speaker Script
Abstract: How can we think about AI and the role of libraries in AI development?

[This is the script I used for my talk. I’ve also taken the opportunity to add some foot notes and links.]

In the 90s I was working on a project called AskERIC – a service that would answer questions of educators and policy makers online. It was early days of the web, well before Google, Facebook, or Amazon. Yet even then we would regularly get questions about artificial intelligence; “i.e., Can’t machines answer these questions?” My boss’ answer was great – “We’ll use natural intelligence until artificial intelligence catches up.” 

A quarter century later, artificial intelligence has done some significant catching up. From search engines to conversational digital assistance to machine learning embedded in photo apps to identify faces and places, the progress of A.I. is breathtaking.

The last 10 years of progress is particularly impressive when you realize that A.I. has been a quest of computer scientists since before there was such a thing as computer science.

Today the larger conversations of A.I. tend to be either utopian

A.I. will improve medicine, reduce accidents, and decrease global energy use

or dystopian

It will destroy jobs, privacy, and freedom.

A.I. has also become a bit of a marketing term – soon, I fear we’ll be eating our cereals fortified with AI.

The hype and real progress have merged into a bit of a jumbled mess that overall can lead to a sort of awe and inaction.

Awe in that many of us in the library, and particularly public library community, may feel the details are over our head. A.I. is a game for Google. Inaction because the topic seems too big – what role is there for a library when these tools are being created by trillion-dollar industries?

True story, the same day my dean asked me about the possibility of creating a degree in data science and A.I., MIT announced a Billion Dollar plan to create an A.I. college[1]. I don’t think he appreciated me asking him if I would have a billion to work with as well. 

I’ve found these reactions – awe and inaction – are often a result of muddled vocabulary. So, for my contribution to today’s agenda, I’d like to briefly break the conversation down into more precise, and actionable concepts. My focus here will be in the contribution of public libraries, but I believe that the concepts are not only relevant to other public sector organizations but can only be truly implemented with partners of all types.

So rather than just think of A.I. as a big amorphous capability, I ask you to think about three interlocking layers: data, algorithms, and machine learning.

Ready access to masses of data has led to high-impact algorithms and increasingly to machine learning and black box “deep learning” systems. If we, librarians do not seek to have positive impacts at each of these levels – well, to be blunt – I would argue we are not doing our job and putting our communities in danger.

So, let’s begin with data.

The first thing that gets thrown into the A.I. bucket is the idea of data or big data. From data science to analytics there is a global uptick in generating and collecting data. With the advent of always connected digital network devices – read smart phones – in the pockets of global citizens, data has become a new type of raw resource.

And when I say global, I mean it. In 2010 the United Nations reported that there are far more people in the world that have access to a cell phone than to a toilet.

With this connectivity, most in society have simply accepted that one of the costs to being connected is sharing data. Sharing it with a carrier and sharing it with the company who wrote the phone’s software. Apple or Google probably know right now where you are, who you’re with, and if you use Siri or Google Assistant, they are primed to be listening to what you might be saying right now. No, I mean literally, right now.

The phone thing probably doesn’t surprise you. But what about the road you used to get here today? When governments build or repave a road, there is a good likelihood they are embedding sensors into it. Why? Well one reason is to save the environment and money in northern climates. How? In the winter rather than just lay down salt or chemicals on every mile of road, smart sensors can pinpoint where ice melt is needed, and reduce the application of costly chemicals.

Sensors are also used to determine the amount of traffic on the road, when to change signal lights, collect tolls, and check for wear and tear.

Add to this the data generated by cars on the road – digital radio, GPS, increasing autonomous driving – and the data begins to add up.

In fact, by one estimate in a few years each mile of highway in the U.S. will generate a gigabyte of data an hour. As there are 3.5 million miles of highways in the U.S., that would be 3.3 petabytes of data per hour, or 28 exabytes per year.

Just in case you are wondering, five exabytes is enough to hold all words ever spoken by humans from pre-history to about the year 1995. Now imagine over 5 times that a year, just on asphalt.

Now that may seem overwhelming, but at the data layer there is a lot of need, and space for libraries to participate. The questions to ask and develop answers to are familiar. Who has access to that data? How is that data stored and how do you find anything in that exabyte haystack? How do we make people aware of the data they may be sharing? How do we advocate for effective regulation to protect citizens?

I argue that public libraries should be steward of public data. Libraries have a VERY long history of data stewardship that includes respect for privacy and seeking equitable access to information. If we are going to allow our governments and our businesses to harvest data then we need to ensure our communities have a strong say in how that happens and trust in those that make the decisions. Right now, libraries have a stronger level of trust than Apple, Google, Facebook, and most elected governments[2].

The accumulation of data in and of itself is not particularly alarming. As libraries have shown over and over again having a bunch of stuff means nothing if you don’t have systems to find it and use it. This takes us to our second layer of concern in A.I.: algorithms.

Companies and governments alike are using massive computing power to sort through data, much of it identifiable to a single individual, and then these folks make some pretty astounding decisions. Decisions like which ad to show you, or what credit limit to set on your credit card, to what news you see, and even to what health care you receive. In our most liberal democracies software is used to influence elections, and who gets interviewed for jobs.

Charles Duhigg, author of “The Power of Habit[3],” tells the story of an angry father who storms into a department store to confront the store manager. It seems that the store had been sending his 16-year-old daughter a huge number of coupons for pregnancy related items: diapers, baby lotion and such. The father asks the manager if the store is trying to encourage the girl to get pregnant? The manager apologizes to the man and assures him the store will stop immediately. A few days later the manager calls the father, only to find that the daughter was indeed pregnant, and the store knew it before she told her father.

What’s remarkable is that the store knew about the pregnancy without the girl ever telling a soul. The store had determined her condition from looking at what products she was buying, activity on a store credit card, and in crunching through huge amounts of data. If we updated this story from a few years ago we could add her search history and online shopping habits, even her shopping at other physical stores. It is now common practice to use online tracking, WIFI connection history, and unique data identifiers to merge data across a person’s entire life and feed them into software algorithms that dictate the information and opportunities they are presented with.

In her book “Weapons of math destruction[4],” Cathy O’Neil documents story after story of data mining and algorithms that have massive effects in people’s lives, even when they show clear biases and faults. She describes investment algorithms that not only missed the coming financial crisis or 2008, but actually contributed to it. Models that increased student debt, prison time for minorities, and blackballed those with mental health challenges from jobs.

The recurring theme in her work is that these systems are normally put in place with the best of intentions.

And here we see the key issue in the use of software to crunch massive data to make decisions on commerce, health care, credit, even jail sentences. That issue lies in the assumptions that those who use the software make. Often very dubious, and downright dangerous assumptions. Assumptions such as algorithms are objective, and that data collection is somehow a neutral act. Or even, that everything can be represented in a quantitative way – including, by the way, culture[5]and the benefit a person makes to society[6]

What role is there for librarians, curators, and academics here? The answer on the surface is about the same as in our discussion of data. Education, awareness, a voice in regulation. However, we must be very aware of the nature of our voice. 

For too long librarians saw ourselves as neutral actors. We collected, described, and provided materials believing that these acts were either without bias, or that those biases were controlled.

In collecting we took it all…except for works that were self-published, or from sources we deemed of low quality. In cataloging we relied on literary warrant and the language of the community – often ignoring that we only saw the dominant voices of that community. Our services were for all – during our open hours for those who could travel to our buildings.

We as a profession are now waking up to the fact that we are a product of our cultures – good and bad. We understand that the choices we make in everything from collections to programs are just that – choices. Our choices may be guided by best practice, or even enforced by law, but ultimately, they are human choices in a material world where resource decisions must be made.

So as a library we are not asking to be neutral arbiters of data collection and uses. We are seeking to improve society through data and algorithms – that means we have a point of view. We have a definition of what improve means.

However, the biases we bring, or more precisely the principles, we bring to the Googles and Facebooks of the world is that a strong voice that advocates for transparency, privacy, the common good, and a need for a durable memory is important.

We recognize that bias exists even if we can’t always identify it, and so we require diversity and inclusive voices in our work. In this act we are not simply advocates, we are activists. A missionary corps of professionals equipping our communities to fight for their interests.

And this brings us to our last layer. The layer that most purists would say is true artificial intelligence development. The use of software techniques to enable machine learning, and especially the more specific deep learning.

That is, software that allows the creation of algorithms and procedures without human intervention. With techniques like neural nets, Bayesian predictors, Markov models, and deep adversarial networks software sorts through piles and piles of data seeking patterns and predictive power.

An example of machine learning systems in action would be feeding a system a number of prepared examples, say hundreds of MRI scans that are coded for signs of breast cancer. The software builds models over and over again until they can reproduce the results without the prepared examples. The trained system is then set upon vast piles of data using their new internally developed models.

With the wide availability of massive data, newer deep learning techniques do away with the coding, and go straight to iterative learning. Where machine learning used hundreds of coded examples, deep learning sets software free on millions and millions of examples with no coded examples – potentially improving the results and eliminating the labor-intensive teaching phase.

When this works well, it can be more accurate than humans doing the same tasks. Billions of operations per second finding pixel by pixel details humans could never see. And they can do it millions and billions of times never tiring, never getting distracted.

In these A.I. systems there are two issues that librarians need to respond to. The first is that these machine-generated algorithms are only as good as the data they are fed. MRIs are one thing, credit risks are quite another. Just as with our human generated algorithms, these systems are very sensitive to the data they work with.

For example, a maker of bathroom fixtures sold an AI-enhanced soap dispenser. The new dispenser reduced waste because it was extremely accurate at knowing if humans hands were put under the dispenser or say a suitcase at an airport. Extremely accurate, so long as the hands belonged to a white person. The system could not recognize darker skin tones[7]. Why? Was the machine racist? Well, not on its own. It turns out it had been trained on only images of Caucasian hands. 

We see example after example of machine learning systems that exhibit the worst of our unconscious biases. Chat bots that can be hijacked by racists through Twitter, job screening software that kicks out non-western names. Image classifiers labeling images of black people as gorillas[8].

However, bad data ruining a system is nothing new. If you’ve had about 10 seconds of work migrating integrated library systems, you know that all too well.

The real issue here is that the models developed through deep learning are impenetrable. That MRI example looking for breast cancer? The programmers can tell you if the system detected cancer, even the confidence the software has in its prediction. The programmer can’t tell you how it arrived at that decision. That’s a problem. All of those weapons of math destruction Cathy O’Neil described, can be audited. We can pick apart the results and look for biases and error. In deep learning, everything works until, well, an airplane crashes to the ground or an autonomous car goes off the road.

And so what are we to do? This is tricky. There can be no doubt that data analytics, algorithms taking advantage of massive data, and A.I. have provided librarians and society great advantages. Look no further than how Google has become one of a librarian’s greatest tools because it provides not only the ability to search through trillions of web pages in milliseconds, but often serves as a digital document delivery service undreamed of 25 years ago when I was working on AskERIC.

And yet, we still need that natural intelligence my boss, Mike Eisenberg, talked about.

Our communities, and our society, needs a voice to ensure the data being used is representative of all of a community, not just the dominant voice, or the most monetizable. Our communities need support, understanding, and organizing to ensure that the true societal costs of A.I. are evaluated, not simply the benefits.

That may sound like our job is the be the critic or even the luddite, holding back progress. But that’s not what we need. Librarians need to become well versed in these technologies, and participate in their development, not simply dismiss them or hamper them. We must not only demonstrate flaws where they exist but be ready to offer up solutions. Solutions grounded in our values and in the communities we serve.

We need to know the difference between facial identification systems, and facial identification systems that are used to track refugees. We need to know the difference between systems that filter through terabytes of data, and systems that create filter bubbles that reinforce prejudice and extremism.

And today is a great first step to honoring that responsibility.

Thank you, and I look forward to the conversations to come.


[1]https://www.technologyreview.com/f/612293/mit-has-just-announced-a-1-billion-plan-to-create-a-new-college-for-ai/

[2]https://www.youtube.com/watch?v=Tvt-lHZBUwU

[3]http://www.worldcat.org/oclc/881631924

[4]http://www.worldcat.org/oclc/1039545320

[5]his article certainly doesn’t claim that all of cultural heritage can be represented quantitatively. Rather I include the citation because it is a good introduction to the use of quantitative analysis of some cultural material and because it includes the very cool term Culturomics, “Culturomics is the application of high-throughput data collection and analysis to the study of human culture.” https://www.librarian.net/wp-content/uploads/science-googlelabs.pdf

[6]https://www.businessinsider.com/china-social-credit-system-punishments-and-rewards-explained-2018-4

[7]https://www.iflscience.com/technology/this-racist-soap-dispenser-reveals-why-diversity-in-tech-is-muchneeded/

[8]https://www.theguardian.com/technology/2015/jul/01/google-sorry-racist-auto-tag-photo-app

Librarianship in an Era of Big Data: The Vital Human Touch

“Librarianship in an Era of Big Data: The Vital Human Touch.” Conference of European National Librarians. Mo i Rana, Norway. (via video conference)

Speech Text: Read Speaker Script
Abstract: In a world of AI and Big Data the values, skills, and mission of librarians is increasingly vital. How do we prepare our professionals to guide and support communities across the EU? How do we ensure the smart citizen is at the center, and in control of the smart city? It is crucial that librarians advocate for issues of privacy and the common good in the midst of a growing market place that transforms users into products. How can librarians increase their value, work with technology giants to shape services, and in the end help our communities make smarter decisions and community members find meaning in their lives.

[This is the script I used for my talk. I’ve also taken the opportunity to add some foot notes and links.]

First, I must begin with an apology to Cecile and the organizers. For weeks they have been asking for slides. I have struggled to put into a concrete form what I want to say to such an important and, frankly, intimidating crowd. I have also spent a year wrestling with cancer and a bone marrow transplant. I don’t say this to get sympathy (maybe a little), but I have found that the experience has given me a new perspective on things that I am still trying to integrate. At once, I am grateful for just about everything, but tend to be less patient with things I feel need changed. And to be clear, as I look out at my home country and across the Atlantic, I see some need for change.

This crystalized for me when preparing for the talk and kept coming into the phrase “memory organizations[1].” If indeed Santayana was right that “Those who cannot remember the past are condemned to repeat it[2],” then what obligations does that place upon institutions who are charged as memory keepers of nations? I would argue that as stewards of cultural heritage, we are also stewards of society. I would also argue that we must stop serving communities, and start building them.

Our communities, our societies, our cultures are too important to sit on the sidelines and simply observe or collect their output. We must recognize, without the haze of nostalgia, that we are actors in this world and accept a responsibility to work directly with communities of all types to shape a better tomorrow. 

Now this could, and should, take many forms. Confronting growing economic disparities, alerting the world to the dangers of xenophobia mixed with nationalism, or confronting the realities of climate crisis. I could talk about the continuous marginalization of whole segments of the population because of race, or class, or sexual preference. Marginalization by society, and indeed, all too often by librarians and the libraries they manage.

Today, however, I would like to use just one societal issue to support a call for national libraries to directly build communities. This incredibly consequential issue often goes unnoticed. It is the dangerous aspects of an increased societal reliance on data-driven algorithms generated by artificial intelligence and machine learning methods. 

Before I begin, let me throw out a few caveats and clarifications. The use of data, when appropriately gathered and analyzed is incredibly powerful. Indeed, it underlies most of science. The rise of artificial intelligence, machine learning, and indeed Big Data has unquestionably brought massive benefits to many disciplines. The ability to search through trillions of pages in milliseconds, search across massive number of images, and the ability to automate complex processes have directly benefited librarians. The issue, as I see it, is when we believe that data gathering, analysis, and encoding into algorithms are somehow neutral acts without social costs[3].

In terms of clarifications; I will use the term community a lot. That is often seen as synonymous with towns, or citizens of a city. I use the term more broadly. When I say communities, I mean a group of people joined by some known variable and that share a means of allocating limited resources. A town is indeed a community sharing a common location, and a system of governance that allocates land, taxes, and other resources. A university is a community of scholars, staff, students, and administrators. A community can be a law firm, or a hospital, or a national library.

The other caveat is on the topic of national libraries. I have done some work with national libraries, but much of my relevant experience is working with state libraries here in the US. There is an old joke that once you know one state library, you know…one state library. I don’t pretend to be an expert on European National Libraries. However, from what I’ve seen, this is true of your institutions as well.

Some of you are active in networks with public and academic libraries. In some countries, there are multiple national library agencies. Some actively seek to support business communities[4], others focus on scholarly research. Bottom line – no one set of issues or models will reflect your great variety.

That said, I am reminded of the saying, “every idea is a good idea in libraries, just not in my library.” It is very easy to focus on what differentiates us now and believe that prevents collective action in the future. I hope I can successfully persuade you otherwise

So, with the caveats and clarifications out of the way, my purpose today is to recruit all of you to build a network of proactive librarians around European. I am calling on you to directly support, train, and empower librarians from those working in the most rural public library to those in the most prestigious university. What’s more is that I am asking you to engage in community building to shape a better future. Why? Well, let’s start with a story.

Charles Duhigg, author of “The Power of Habit[5],” tells the story of an angry father who storms into a department store to confront the store manager. It seems that the store had been sending his 16-year-old daughter a huge number of coupons for pregnancy related items: diapers, baby lotion and such. The father asks the manager if the store is trying to encourage the girl to get pregnant? The manager apologizes to the man and assures him the store will stop immediately. A few days later the manager calls the father, only to find that the daughter was indeed pregnant, and the store knew it before she told her father.

What’s remarkable is that the store knew about the pregnancy without the girl ever telling a soul. The store had determined her condition from looking at what products she was buying, activity on a store credit card, and in crunching through huge amounts of data. If we updated this story from a few years ago we could add her search history and online shopping, even her shopping at other physical stores. It is now common practice to use online tracking, wifi connection history, and unique data identifiers to merge data across a person’s entire life.

I am hardly telling you anything new here. Facebook is only the latest business to dominate the headlines with privacy breaches, and hidden data gathering. Most citizens of the EU and the US now live two lives: their own, and one created, often without their knowledge, from the digital debris created through our devices. Add to this increased requirements by governments and businesses alike to be online – to apply for a job, to vote, to receive health care, to listen to music – and we see a world that is moving faster than regulation, and faster than realization by those we seek to serve.

In Toronto, Sidewalk Labs, a subsidiary of Alphabet, Google’s parent company, is working with town planners to redevelop Toronto’s Eastern Waterfront. The story of transforming old industrial areas into gentrified multiuse spaces is nothing new. However, a large part of the controversy in this case comes in the plan to make the new neighborhood a data generator. The plan, according to The Intercept, “includes a centralized identity management system, through which ‘each resident accesses public services’ such as library cards and health care[6].” There has been a large debate over who owns and controls the data generated by that system, and who can profit from it.

Many librarians might look at these examples and claim a sort of ethical high ground. After all, as a profession, we explicitly value privacy. In the US we count it as a core value of the field, and yet we often undermine it. We tell our online patrons that we don’t track their work. And yet their internet provider can indeed track every click they make. Therefore, we are often misleading that patron and giving them a false sense of security. How many libraries set up TOR servers[7]or anonymizing VPN services[8]for our service populations? How often in our licensing of databases or other software do we explicitly forbid the aggregation of user data or the selling of that data? How often do we check on those terms?

Then there is the question of how all of this data is used.

In her book “Weapons of math destruction[9],” Cathy O’Neil documents story after story of data mining and algorithms that have massive effects in people’s lives, even when they show clear biases and faults. For example, an algorithm that led to outstanding teachers being fired. How? O’Neil writes about an outstanding teacher who had proven positive effect on under-performing students– raising their performance and grades significantly. As a reward, the teacher is given a year with honors classes filled with the brightest students in the school. However, the impact a teacher can have on honors students is not nearly as evident as those with students needing a lot of help. After all, top students receiving top marks can’t get better than, well, top marks. So the algorithm saw a teacher that was no longer effective in a classroom, and recommended the teacher be fired. Recommended by a piece of software using criteria that was hidden from teachers, and was assumed to be objective.

Algorithms are now used here in the states to determine health care cost and availability; access to credit for home ownership; suitability of a candidate for a job,; and even how long a person should be in jail.

Yuval Harari refers to this reliance on collectable data and algorithms as Dataism[10]. It is the result of computing power combined with machine learning and the wide availability of constantly connected devices like our phones. It is the belief that if you gather enough data on a person or situation, you can accurately represent that person or situation and predict an outcome.

It often also comes with some very dubious, and downright dangerous assumptions. Assumptions such as algorithms are objective, and that data collection is somehow a neutral act. Or even, that everything can be represented in a quantitative way – including, by the way, culture[11]. And before I make you wonder what any of this has to do with the work of libraries, or think I’m letting our profession off the hook, I have to say that librarians have suffered from some of the same dubious assumptions.

For too long librarians and library science educators saw ourselves as neutral actors. We collected, described, and provided materials believing that these acts were either without bias, or that those biases were controlled. In collecting we took it all…except for works that were self-published, or from sources we deemed predatory or of low quality. In cataloging we relied on literary warrant and the language of the community – often ignoring that we only saw the dominant narrative and voices. Our services were for all – from 9 to 5 with a researcher’s card who could travel.

We as a profession are now waking up to the fact that we are a product of our cultures – good and bad. We understand that the choices we make in everything from classification to exhibits are just that – choices. They may be guided by best practice, or enforced by law, but ultimately, they are human choices in a material world where resource decisions must be made. We can speed up digitization with newer machines, but we still have to pick a starting point. We can expand those we serve on the web, but still must acknowledge that there are people with no broadband or connectivity.

Now it would seem like this may turn to a call to redouble our efforts in neutrality. A call to wipe away the biases in ourselves so that we can confront the cost to society of skewed machine learning efforts. But it is not. In fact, we must embrace that libraries, and the librarians that build and manage them, are biased[12]. What’s more, it is only by seeing libraries as biased that we prove our value in the world of massive scale data.

First, we must realize that it is impossible to be neutral. Putting a book on a shelf or in a vault is a choice. Every day in archives and special collections we make professional determinations of how accessible an item is versus how protected it is. We can seek out many voices and yes, gather data, to make those decisions, but in the end, they are decisions with consequences. Pretending we are neutral doesn’t change the consequences, it only allows us to pretend they are not the result of our action.

Now, I keep calling them biases, but a better word would be principles. Principles are an explicit statement of belief. They should be transparent and, most importantly, able to be assessed. Are we following our principles? 

And make no mistake principles are not neutral. Seeking to serve all equitably takes effort and resources. Choosing to provide images for fee or free is a choice. Fighting censorship is a decision. If you don’t think so, try balancing it against issues of hate speech and threats to marginalized communities.

It is in our decisions and our transparency in making those decisions that we build trust with our communities. Our scholars, and entrepreneurs, and citizens don’t trust librarians because we are neutral, but because they agree with our principles and see them consistently applied. The days when libraries had the monopoly on access to large collections is well over. Yet libraries in most places the world are not only in use, but in growing use – public, academic, school, and national libraries alike.

Where library use (not necessarily support, but use) is growing it is because we are seen as accessible, equitable, and trusted. Yes, the collections we hold are valuable. The fact that we hold unique resources that either haven’t been, or can’t be digitized is important. But it is only important if those who seek out these resources trust us to be honest stewards of these resources.

It is our embrace of our humanity – our human touch in an increasingly automated system that underlies our value. This is not a luddite’s call against technology, AI, or machine learning. Rather it is a belief that human connection – community- is more important than ever when the face of government and business alike become web pages and bots under the banner of austerity or efficiency.

The future of libraries is ultimately not set by which technologies are developed or deployed. It is not in a value that was defined a century ago. It is in our very human ability to build trust with our communities. It is upon that trust that we build support. It is upon that trust that we build use. It is upon that trust that we find and confirm our necessity.

It is with that trust that we must reach out to the computer science community, the online industry, and the governments collecting data and deploying algorithms. We must advocate for a seat at the table and represent the voices of those without a seat. We must use the hard lessons we learned and are still learning in issues of diversity, equity, and inclusion to help guide these technologies. We must be trusted by our communities to speak truth to power and to give those communities power to speak for themselves. We must follow our principles in actively shaping, with our communities, the policies, regulations, and laws that talk about data. 

National libraries must play a large role in civic data stewardship. National libraries must not just safe guard the heritage of cultures, but the privacy and intellectual safety of citizens. You need to be a memory organization in the realization that effective memory is both about remembering, and, forgetting.

How do I resolve the paradox that I just advocated for a common role in institutions that I also acknowledge are so diverse? If indeed, you each represent unique institutions, does this preclude collective action? Of course not. Because in effect you are what all libraries must become. Every library- public, academic, school- should be shaped to the communities they serve. Then, as librarians, we become the connective tissue that seeks the best of all libraries and shape those innovations to local needs. Gone are the days when every library looked alike or supported some cannon of common services. Gone are the days when best practices extend to all libraries of a given size or type. Throw away the toolkits and instead build a toolbox[13].

We must prepare our librarians, regardless of title, or training, or location to be a missionary force proactively engaging in the well-being of our cultures and communities. We must build national peer networks that rapidly and effectively spread ideas and help librarians effectively shape them when they meet local needs. These networks discard best practice and industrial standardization for conversation, learning, and adaptation. We must connect the best thinkers together regardless of status or institutional boundaries.

How do important national institutions do this?

We must create platforms for continuous engagement of librarians where they can share, learn, teach, mentor and support each other. This may be built upon and with national and regional associations, but the focus is on individuals, not institutions.

We must create a system to formally recognize participants within and beyond this platform. Work with library science programs where they exist, but also extend the recognition beyond formal degrees to continuous learning.

We must recognize Lighthouse Libraries[14]that embody innovation and serve as inspirations, not blueprints, for other libraries.

We must proactively engage this network of change agents to transform libraries, associations, institutions, and ultimately communities globally. Members of our peer networks, our communities of practice, must encounter daily new ideas from across the globe.

Think of a library as movement, not a place or an institution[15]. It is a movement of people committed to improving society. Librarians, certainly. But also, scholars, politicians, entrepreneurs, programmers, and authors. Discard terms like users that reinforce the idea that our communities are consumers, and our only value is in the utility we provide to a demand. We have members and citizens; neighbors and scholars that all own and shape the library.

Most importantly, this will not happen in one hour of a conference. It will take more engagement, more experimentation, and more investment. That is why I support the PL2030[16]project. Building off of the Public Libraries 2020 project, it is a group of librarians from across the continent seeking to transform public libraries across Europe one librarian at a time. It advocates for libraries and builds connections between elected representatives and library innovators. But it needs help. 

PL2030 and the work of its members represent the need for a new vital link between the cultural heritage mission of National Libraries and public libraries. Libraries are transforming from access points, collections, and information providers into community hubs across Europe. From Manchester to Cologne to the amazing Dokk1 in Aarhus to Delft and Tilburg in the Netherlands and Pistoia and Perugia in Italy public libraries are the places communities come to learn, create and dream together. Here, by the work of innovative librarians, libraries have gone from quiet places of retreat to loud places of engagement. The true collection of a great public library is now the community itself. Blacksmiths and bakers host conversations. Librarians lend out books and musical instruments and recording studios. Rather than bringing the world to the community, these libraries have become loudspeakers broadcasting the community to the world. These public libraries have become the cradle of cultural creation.

As institutions charged in part with preserving and supporting the cultural heritage of a people, you need to preserve and support the work of these institutions. Not simply as a backup or for posterity, but as part of the living and breathing centers of community conversation. In a connected world – connected through technology certainly, but also in trade, in governance, and in preserving the earth itself – There is no more front-line service and library of last resort. We librarians are obligated to serve all, and in your nations now is a network of libraries eager for your partnership.

I thank you for your time, and I look forward to the conversation to come.


[1]Or “Memory Institutions” like the CENL Strategic plan https://www.cenl.org/wp-content/uploads/CENL-Strategy-2018-2022_final-1.pdf

[2]https://en.wikiquote.org/wiki/George_Santayana

[3]I love Chris Bourg’s take on the use of “societal cost” in discussing AI versus ethics.

[4]I’m a big fan of the British Library’s https://www.bl.uk/business-and-ip-centre

[5]http://www.worldcat.org/oclc/881631924

[6]https://theintercept.com/2018/11/13/google-quayside-toronto-smart-city/

[7]https://www.torproject.org/

[8]There are plenty of good articles explaining VPN. Here’s one that actually compares VPNs vs Tor: https://www.cloudwards.net/vpn-vs-proxy-vs-tor/

[9]http://www.worldcat.org/oclc/1039545320

[10]http://www.worldcat.org/oclc/1060991037

[11]This article certainly doesn’t claim that all of cultural heritage can be represented quantitatively. Rather I include the citation because it is a good introduction to the use of quantitative analysis of some cultural material and because it includes the very cool term Culturomics, “Culturomics is the application of high-throughput data collection and analysis to the study of human culture.” https://www.librarian.net/wp-content/uploads/science-googlelabs.pdf

[12]Here’s a good place to start on the discussion of libraries, librarians and neutrality:https://americanlibrariesmagazine.org/2018/06/01/are-libraries-neutral/In particular check out Emily Drabinski’s take.

[13]Pithy phrase, but in case it is not clear I mean stop sending out assembled ready to implement toolkits and focus on librarians gaining the tools to develop their own programs and/or create local application of programs customized to their communities.

[14]https://publiclibraries2030.eu/projects/lighthouse-libraries/

[15]Stole this idea from the amazing Marie Østergaard: https://podcasts.apple.com/us/podcast/princh-library-lounge-ep-3-building-global-networks/id1451326347?i=1000437039135

[16]https://publiclibraries2030.eu/