In an age of limitless information, organization is the burden that kills us. It’s hard to notice. Our phones automatically sort contacts by name. They categorize our shopping list into produce or dairy products. And when we really need something, we can use magical search engines that provide instantaneous results. To be sure, information has certainly become easier to find than in decades past, when you may have been screwed after misplacing your address book or vaccination records. Yet I have recently become increasingly anxious over how I organize the hundreds of books in my reading room and thousands of liked songs in my Spotify library. The pile just keeps growing with no end in sight.
My deep fear of forgetting knowledge began in college. I wrote down copious notes for every class in perfect prose and kept every reading assignment, no matter how small. Upon graduation, I had built a compendium of hundreds of pages of notes, thirty textbooks, and thousands of minutes of recorded lectures — all of which have been organized into a searchable website-like hub and tables of content for each course. I have always coped with my anxiety through intense, rigorous organization.
Nowhere does my anxiety flare up as much as corporate America. At just one organization, there are millions — billions — of datapoints managed by multiple owners across numerous files sometimes in an arbitrary database — or worse, a random employee’s hard drive. This is a universal challenge I have seen at early startups, legacy firms, and even (especially) the government. As a consultant, there is no task quite as laborious as retrieving data when clients suffer the plague of disorganization.
Companies and universities have certainly known about the challenge of organization. Information science has emerged as an interdisciplinary field which studies how information is created, managed, and stored. The Bureau of Labor Statistics (BLS) estimates that computer and information science degrees will increase by 23% over the next 10 years as organizations of all kinds continue to wrestle with data and data and data. And while the technological focus of the field may be more recent, the challenge of information management is not new. Imagine the management of one company’s information applied to an entire university’s, city’s, or world’s database of knowledge. Indeed, one of the original fields of study that information science draws upon is library science.
The hidden world of Library Science
Library science encompasses the study and management of libraries and information systems. A Master’s in Library Science (MLS) or Library and Information Science (MLIS) is often the standard credential for professional librarianship, though there are Ph.D programs and even some bachelor’s degrees. The coursework of a library science student ranges anywhere from developing catalogue systems to strategic planning.
Studying librarianship is far from common. According to the Department for Professional Employees, in 2022, there were approximately 165,000 librarians — 60% of which have attained a master’s degree or higher — employed in public libraries, schools, museums, private libraries, and government agencies. Less than 0.5% of the United States workforce have attained an MLS or equivalent degree. Of course, it may be because the path towards credentials is expensive.
Tuition for a Master’s of Library Science at University of North Carolina Chapel Hill can be $15,000 per year for North Carolina residents or more than $32,000 per year for non-residents. With a median wage of $49,000 — falling 23% short of the 2022 national average wage index of $64,000 — the outsized cost of a master’s degree puts librarianship out of reach for many low-income individuals.
Assuming one has the finances, scholarships, or assistantships to afford the degree, librarians are thrust into a multidimensional role that goes far beyond restocking books on shelves. Some activities are expected. Librarians may study collection development, which focuses on curating the most useful books, journals, and digital resources needed to support a group of patrons. They may learn to create systems that facilitate information retrieval, such as the design and maintenance of library catalogs or digital repositories. And they may hone their referential skills, learning how to dig for information needed to assist patrons’ questions and research projects.
However, there are other roles that the layman may not expect librarians to take on. There are librarians who focus on library management, which covers budgeting, personnel management, and strategic planning across a wide branch of libraries. There are librarians who act more like archivists and ensure the long-term preservation of historically relevant documents — such as a local town’s first set of periodicals or Thomas Paine’s manuscripts. And some librarians take on the role of a community advocate and educator, planning publicly available resources and events tailored to local challenges — such as career fairs for the unhoused or language learning sessions for non-native English speakers. Librarians can be trained to wear a lot of hats.
There is, unsurprisingly, one aspect of librarianship that has captured my attention for months and subsequently inspired this essay — cataloging and classification. Library science involves the creation of catalog records for each item in a collection. Cataloging involves describing the item, assigning subject headings, and classifying it according to a standardized system.
For example, American Prometheus: the Triumph and Tragedy of J. Robert Oppenheimer is a book. Its subject headings may include, among many others, “Oppenheimer, J. Robert” who is the primary subject, “Manhattan Project” or “atomic bomb” to describe critical topics of the book, and “biography” or “history” to describe the genre. A librarian may place the book next to other biographies regarding scientists or perhaps the cold war. The process is simple. That is, before you must repeat the process for every piece of media in existence and then be able to quickly find it amongst three floors of one hundred shelves each.
The overwhelming sea of knowledge requires someone or something that can tame those waters — this was the original goal of library science. Since the 19th century, librarians have formally sought to create systems that could capture an expanding, and increasingly diverse, proliferation of books and media. These systems needed to assign values to genres or topics, often numerical values. Numbers are convenient for database management. These systems must also balance the breadth needed to capture many disparate topics with the depth needed to pinpoint precise topics. Creating such a system was — is — an undertaking, to be sure. One of the oldest, most recognizable, and most used classification systems is the Dewey Decimal Classification (DDC).
Don’t do it, Dewey!
The Dewey Decimal Classification is a system that organizes knowledge into 10 parent groups, each of which are assigned 100 numbers as subcategories. Further subcategories can be added as the titular decimals. From Brittanica:
The 10 main groups are: 000–099, general works; 100–199, philosophy and psychology; 200–299, religion; 300–399, social sciences; 400–499, language; 500–599, natural sciences and mathematics; 600–699, technology; 700–799, the arts; 800–899, literature and rhetoric; and 900–999, history, biography, and geography. These 10 main groups are in turn subdivided again and again to provide more specific subject groups. Within each main group the principal subseries are divided by 10; e.g., the history of Europe is placed in the 940s. Further subdivisions eventually extend into decimal numbers; e.g., the history of England is placed under 942, the history of the Stuart period at 942.06, and the history of the English Commonwealth at 942.063.
The system was created by Melvin Dewey, a pioneering American librarian who was perhaps most responsible for the development of library science in the United States. Dewey was one of the founders of the American Library Association (ALA), and he established the first institution for training librarians in the United States. Unfortunately, he was also deplorable. Several women, including his daughter in-law, were uncomfortable around him and his sexual harassment tendencies. He also fervently denied Jewish and Black members and guests from his private club. Dewey’s sexism, racism, and anti-semitism led to censureship by the ALA and later resignation from his state library position. Despite his public condemnation from the librarian community, his biases are ripely embedded within the Dewey Decimal System.
The Dewey system colonizes our libraries, perverting our knowledge systems by prioritizing the interests of a white man from the 19th century. Language topics comprise 400 through 499, yet European languages make up 400 through 489, leaving a mere 10 subdivisions for “other” languages. German and English, for example, both have 8 subdivisions each. Meanwhile every language spoken within the continent of Africa must share a single subdivision in 496. Similarly, European literature takes up 800 through 889, leaving 10 subdivisions to house literature from any other part of the world. The same can be said for religion, where, once again, the first ninety subdivisions are reserved for Christianity (200 through 289). The diverse sets of knowledge relegated to such a small portion of the library have their visibility limited and their significance undermined.
The most heinous and infamous bias of the Dewey Decimal System is its homophobia. Historically, LGBTQ+ topics have been classified not as social science books in the 300s but under psychology in the 100s. Specifically, homosexuality was classified as “Mental Derangements” and “neurological disorders” for decades. When LGBTQ+ topics were moved to the 300s, they were initially classified under “Social Problems” near controversies, crime, and pornography books. The current guidelines — because the Dewey Decimal System has been revised more than twenty times to rectify its harmful biases — now catalogs these topics under section 306, housing culture and institutions. Still, no amount of revisions can retroactively remedy such a faulty foundation fraught with flagrant discrimination.
To be fair, the system is a product of its time being from the 1870s. As such, the system is grandfathered into many libraries, lest the recategorize every single book to a new system. It is also simple and easy to use, which is often why it populates the more-often-than-not under-resourced public libraries and K-12 libraries. The hierarchal structure makes it straightforward for patrons, students, and categorizers to navigate, as they follow a systematic arrangement of broad categories and progressively narrower subcategories. And if one wanted to decolonize the library, there is plenty of programming to enact that drive meaningful change and may take priority over what numbers are on a book’s spine — in other words, the juice may not be worth the squeeze.
Still, as authorship becomes increasingly diverse and research further globalized, the foundation of a library — that is, a system used to retrieve knowledge— is fundamentally important to progressing dialogue in the academy and our communities. Fortunately, people have been reading for a long time, and there are plenty of well established alternative systems.
The unclear path towards a “just” classification system
The world of classification systems is mighty. An electric amount of nerdy energy has been dedicated to creating these organizational systems. There are three broad types: universal schemes which cover all topics and are meant to be used by any country, specific schemes which have a narrower breadth in topics such as music or medicine but are still meant to accommodate any country, and national schemes which cover all topics but is built entirely for one country. Seeking an alternative to Dewey will require a universal scheme.
Outside of the United States, a number of countries have drawn upon the Dewey Decimal Classification to create their own universal system, including Japan’s Nippon Decimal Classification and Korea’s Korean Decimal Classification. There are also major countries’ universal inventions, such as the Chinese Library Classification and Regensburger Verbundklassifikation in China and Germany, respectively. Knowing that we need a universal alternative, we must then look at how the systems are structured.
Functionally, there are also three broad strokes of classifications. Enumerative systems list every subject heading alphabetically, then assign numbers to each heading in alphabetical order. You may imagine a glossary at the end of a book but applied to an entire system of books.
Secondly, there are hierarchical systems which divide subjects into general categories. Those divisions have subdivisions which are more specific, and each subdivision is further cut into even more specific sub-subdivisions. The Dewey Decimal System is hierarchical, along with many universal schemes.
Finally, there are faceted systems, which seek to create a system based on semantic categories — facets — to describe each book. Rather than using letters or numbers, books may be categorized by text values for general facets — time, place, and form, for example. In faceted systems, a book may be look like “open heart surgery:Indonesia:1982”, which may be more flexible and intuitive than “595.789/BROC.”
With a set of principles behind us, we can look to our first alternative to Dewey: the Library of Congress Classification (LCC). The system was also developed in the late 19th century, when the Library of Congress needed to update their classification from Thomas Jefferson’s bespoke method which had set the precedent since 1800. One century’s worth of new books (about 1.5 million volumes) had proved too cumbersome for Jefferson’s system, and other systems (e.g., Dewey, Cutter’s Expansive Classification, German Halle Schema) were deemed inadequate for Congress’s needs. Since then, LCC has become one of the most popular classifications around the world and can be found at most college campuses in America (and many beyond).
The LCC system is a hierarchal system that uses a combination of letters and numbers to classify materials. There are 21 broad classes signified by a letter ranging from General Works (A) to Social Sciences (H) to Library Science (Z). Under the classes are sub-classes, then divisions and sections — each of which use a combination of letters and numbers called Cutter numbers (e.g., A15). By starting with more classes in the first place, and using Cutter numbers to atomize the sub-classes and divisions, LCC enables a far more precise system than DDC.
Of course, the system has its quirks. Despite being a widely used system around the world both in and out of the academy, the LCC system is very much built for America and her interests. The History of the Americas takes up classes E and F, while the rest of the continents share only one class (D). Philosphy, Psychology, and Religion share one class (B) while the Military Science and Naval Science are split into two distinct classes (U and V, respectively). Congress has found that warfare on land is so disparate than in the sea, that they should be in different sections of the library.
We have since talked about disproportionate sizes of classes and subdivisions, but there is something I’ve yet to mention that plagues both the Library of Congress Classification and the Dewey Decimal Classification — subject heading bias. In March 2021, the Library of Congress reported nearly 300,000 subject headings in use, making it not only precise but unrivaled in comprehensiveness. However, the cost of complete precision is that information discovery becomes a game of finding the right keywords. As Claire Woodcock writes in LitHub,
Imagine you are looking for information about astronauts, but don’t know where to start. So you type in the keyword “Astronauts” and the system spits out 10,000 results. You scroll through the top five, and they’re all about Neil Armstrong. You would have had to go through dozens of pages before finding anything on Mae C. Jemison. In order for books about Jemison to appear, you would have needed to type in “African American women astronauts.”
In other words, the precision afforded as a means to provide every topic its own individual space has in fact reinforced the prominence of those who have historically been written about — old, white (American) men. And while this may be quite ideal for academic institutions where precision is paramount to accuracy, it is not sustainable for local and K-12 libraries. An alternative to Dewey will require more than the adoption of its sister system. We require innovation.
A selection of neat, fun classifications
One Dewey-free classification system that was developed quite recently is the Metis Classification. Implemented in 2011, a coalition of four librarians (Sue Giffard, Tali Balas Kaplan, Jennifer Still, and Andrea Dolloff) at the Ethical Culture School in New York City created a system tailored for children. Metis centers the thinking, interests, and information discovery needs of elementary and middle school children.
There are 26 classes, one for each letter of the alphabet. Classes like “Scary” and “Making Stuff” replace more adult-centric topics like Religion or Social Science. “Pets” and “Animals” have distinct classes from one another, as does “Ourselves” and “Community”. And for every class is a visual element, such as a sticker of a book or house to help guide children that otherwise may have yet to learn about decimals. While some have critiqued the system for infantilizing the library, librarians have found an increase in circulation of extracurricular activities, such as an 87% increase in arts and crafts, AKA “Making Stuff”.
Another Dewey-free classification is the one that bookstores use — the Book Industry Standards and Communications or BISAC system. Books are assigned no more than three categories, many of which you are probably familiar with. For example, a book may be catalogued under Fiction, perhaps being Fiction: Fantasy or Fiction: Historical. Another may be catalogued under Biography and Autobiographies or possibly Business. These classes allow for the average lay person to take on a more exploratory mindset when looking for books.
Moreover, the system is built to be nimble based on market needs. Subject headings are pushed in and out of prominence based on people’s discovery habits; for example, the science section may center Black botanists for Black History Month. Classifications can also quickly adapt to emerging and diverse topics, unlike other systems which require a new edition to be published.
These strengths enable bookstores to create hubs of content where patrons discover books that appear disparate at first but could be tied by a through-line. A collection for women in STEM may lure in readers looking for a novel, like newly televised Love in Chemistry, and have them walk away with The Life of Lise Meitner, a Jewish physicist on the Manhattan Project. One of my favorite pairings is that when searching for Founding America: Documents from the Revolution the the Bill of Rights, Barnes and Noble places Howard Zinn’s A People’s History of the United States not too far away, which helps center America from the perspectives of the oppressed rather than the oppressor. Some public libraries across the United States have already found varying levels of success switching to BISAC because of its intuitiveness and familiarity.
Finally, a more traditional alternative could be the Universal Decimal Classification (UDC), which is popular in Europe. UDC explicitly intends to capture the entirety of human knowledge. Like the Dewey Decimal Classification, there is a numerical hierarchy of classes and subclasses. Unlike DDC, the Universal Decimal System does not have a fixed number of classes and it integrates facet analysis with the numerical hierarchy. Its four facets are: form, matter, energy, and language. By marrying hierarchy and facet, UDC is known for unparalleled flexibility that can accommodate for interdisciplinary connections that DDC and LCC functionally cannot.
Connecting signs are used as a mechanic for clarifying interdisciplinary topics. For example, plus signs (+) can be used to denote the coordination of two topics, such as a book on music composition and painting, and colons (:) can be used to denote relations, such as a book comparing the relationship of film to liberalism. These auxiliary features allow for UDC to specify subject headings without being boxed into one class or sub-class. Thus, UDC attempts to combat subject heading bias, the likes of which we found in LCC, by allowing subjects to live in multiple spaces.
Clearly, Universal Decimal Classification is precise. This strength is particularly useful for journal articles. UDC is used in national bibliographies of around 30 countries, including NEBIS (The Network of Libraries and Information Centers in Switzerland) which indexes 2.6 million records and COBIB.SI (Slovenian National Union Catalogue) which indexes 3.5 million records. Not only is UDC flexible and precise but it is capable of massive amounts of data.
Epilogue
There may not be a perfect solution to classification. While the Dewey System is clearly flawed, its precedent allows for libraries to function well when under resourced. Other systems clearly have benefits, whether it be better precision, greater shelf-space equity, or flexible groupings. Still, the perfect approach will need to balance ease of use with discoverability, all while amplifying equity. It’s a grand question within the invisible world of organizing books.
Furthermore, the ideal classification depends on the strategic goals of the library and its subsequent priorities. Some libraries must prioritize precision over flexibility, while others may want to encourage discoverability. Each library faces an ontological question of what the library is and whom the library serves. They confront questions about the selection the curate and how vast of offerings are necessary (such as every piece of media written about the holocaust vs non-Nazi-affiliated literature). And they reckon with the needs of their communities every day, whether they be children or the unhoused. Ultimately, the best classification depends on three things: the resources available, the community its serving, and the library’s vision for the future.
P.S.
Tasked with inventing a new classification tailored to my whims and desires for my local public library, I would optimize for two goals: discoverability and approachability. I have two universities within a twenty minute drive and two more community colleges. Precision is not my biggest priority. Rather, I want a library that can create innovative hubs of topics that call in diverse voices and narratives. I also like stickers and more playful classifications like “machines” and “tales”.
So, if I were to create my own perfect library, I would imagine a hybrid approach of BISAC and Metis classifications. Nothing would make me happier than walking out with more books than I came for and a handful of stickers.
