Week 2: So how do you find things?
color-coding, classification, could I possibly have written a longer post?
Part 1: Color-Coding
This week on Twitter, there was a whole lot of hand wringing because Amanda Gorman, the First! National! Youth! Poet! Laureate! apparently, possibly, potentially organizes her bookshelf by color. I encountered this while doomscrolling and was like “dear god, do we have to have this conversation again?” I realize that if you’re reading this, you most likely have never had a heated conversation about the moral implications of organizing your personal bookshelves by color (spoiler alert! It’s an amoral decision. No one dies if you decide to group all the red covers together). But I have. And I have been on both sides of the debate, though I am now firmly in the camp of “do what makes you happy.” My previous stance of “it makes no sense, this is somewhat mind-numbing” was snobbish and dumb and here’s why:
People can do what they want with their own books
There is historical precedent for organizing one’s books by color. I loved this article by Alexandra Alvis, which recounts the historical, social, and psychological contexts of organizing books by color (I really recommend reading it, it’s not technically dense at all). She points out that in the hand-press period (when books were made on hand-operated printing presses), book bindings were commissioned by their owners, which meant they were purposefully choosing what the exterior of their books looked like. This is how you can get a printing of a book from, say, 1745 that features an array of bindings (different leather colors or gilt designs, etc.). And contemporary intellectuals issued suggestions on coordinating color and subject in personal libraries so that their owners could more easily navigate them. Basically, wealthy readers were concerned with what their books looked like, and they thought a lot about how those books would hold up in their libraries.
There are other examples of color-coding books that you are probably more familiar with. Penguin Books famously color-coordinated their paperbacks according to genre in their original runs beginning in 1935: orange for fiction, green for crime/detective fiction, dark blue for autobiographies, and so on. (I have a lot of beef with Allen Lane, the founder of Penguin, but this is one good example for what we’re talking about here.)
The first ten Penguin books published 30 July, 1935. I don’t remember where I got this image. Please don’t sue me.
In short (though this will be long), organizing things by color is one system of classification, or the methods of deciding where an item belongs [on the shelf, in a database, in relationship with other materials]. Like the Library of Congress Classification System or the…Dewey Decimal System. (I’m coming for you, Dewey, you don’t get off that easy). Classification is one element of bibliographic control, or (and I am quoting directly from my professor’s powerpoint here) “the identification, description, analysis, and classification of books and other materials of communication so that they may be effectively organized, stored, retrieved, and used when needed.”
Bibliographic control is how everyone from me with my Excel spreadsheet to mega libraries documents, sorts, shelves, and locates the things they put on their shelves. There are a few steps and a few elements to bibliographic control, but today is all about classification, of which there are a few kinds:
Arbitrary (like alphabetical order. There’s no relationship in the order of the words in a dictionary. Though my friend Josh who possibly reads this newsletter has memorized every 2-8 letter word in the English Scrabble Dictionary so maybe there are relationships to that order? You’d have to ask him)
By subject or genre
Color-coded!
or something else entirely: the Warburg Institute in London, where I hung out a lot of the time because they had tea and cookies at 4pm every afternoon, organizes their works by Image, Word, Orientation, and Action, which is encoded in the classmarks on an item’s catalog record.
Different classifications work better for different kinds of information. And classifications are often paired: note that in the historical examples Alvis provided, the color-coding was paired with genre or subject of the text. The books were not just in rainbow order but rainbow and subject order. On my own shelves, I organize by genre and then internally by alphabetical order by author surname (I also have shelf codes but not classmarks, though maybe I’ll write out my own classification system when I’m procrastinating the next issue of this newsletter).
But pure color-coded classifications without additional sub-classifications do not translate well to institutional collections, and for a few reasons. Should Amanda Gorman’s personal library ever end up in an institution (or perhaps…a presidential library?? Ma’am, please hire me), her books will likely get scrambled under a new organization system because color won’t cut it:
There are plenty of instances wherein the same titles have different covers, either for the aforementioned custom binding situation or because new editions offer new cover artwork (I’m thinking of one blockbuster series about a boy wizard that has the “children’s” covers, “adult” covers, “anniversary” covers, “House” covers…seriously people, stop buying those books). If all of those editions were entered into a library that organized by color, you wouldn’t just split up different titles by the same author, you might also split up the same title by the same author.
Corollary 1: I could see classifying books by cover design as useful for a collection about book making and/or graphic design…but I would still itch to split the items up by year or region or some relationship. Organizing a big institutional collection by color would be like telling a joke and never finishing the
Corollary 2: or maybe in a library or archive that records how people organized their libraries? This is a fascinating subject of study because it documents how people linked information in real time, in their personal spaces, for their private intellectual purposes. Bookcase Credibility kind of addresses this phenomenon, though the account is more satirical than analytical. And there are titles that people buy or place strategically on their bookshelves to convey things about them (have you ever read the gigantic copy of Merriam Webster’s Dictionary you got for high school graduation?). Books are texts and objects. They can and should be analyzed both individually and together. (Alvis put it more succinctly: “‘What is more meaningful,’ asks book artist Ulises Carión, ‘the book or the text it contains?’ What is more meaningful, I ask, the rainbow bookshelf or the quality of its owner?”)
Part 2: How to find things
In libraries, classifications are (often) encoded as coll numbers (the long string of letters and numbers that precede a book in a catalogue), and they’re also how you organize search databases. The Library of Congress Classification System (LCCS) organizes materials by assigning them to broad subject classes and more specific subclasses. These subject headings are extraordinarily useful for searching because they link materials that you might not otherwise think of as related or that you might not have keyword terms for. For example, if you have a very specific search term that begets few results, you can click the subject headings and find all the materials listed under that subject heading. I searched “Sarajevo Haggadah” in the Library of Congress catalog and found that it has 7 subject headings, a few of which are very specific (“Manuscripts, Hebrew – Facsimiles” will lead me to other facsimiles of Hebrew Manuscripts in LoC’s holdings).
Subject headings also link materials that might otherwise fall under the radar due to spelling or term differences. Last week I mentioned phenakistiscopes, which are pre-cinema animation artifacts. Basically, they’re discs with slightly different images around the border (like a flip book!) that fit onto a backing disc with slots along the borders. You face the disc to a mirror and spin it: looking through the slots breaks up the motion of the disc into frames, revealing an animated picture on a loop, like a Gif. They’re super cool.
(Here’s a weird video featuring a lot of different discs with very disorienting music that makes me feel like I’m in an episode of American Horror Story even though I’ve never once watched that show in my life).
(Like many things I mention, I was cataloguing a phenakistiscope last week for my bookshop boss, who bought it because they’re really super cool.)
(Finally: in my research on said phenakistiscope, I discovered the Tumblr of Richard Balzer, who unfortunately died in 2017 (so I could not message him) but collected all kinds of pre-cinema artifacts, proving once again that there is something for everybody. His wife donated his collection of over 9,000(!) objects to the Academy Museum of Motion Pictures (the museum for the organization that hands out the gold statues to mostly white actors and occasionally Bong Joon-ho.))
The term “phénakistiscope” (alternatively spelled “phenakistoscope” and meaning “to deceive eye”) was coined in a French newspaper right after the object’s simultaneous invention in Belgium and Austria. The Belgian inventor Joseph Plateau didn’t like the name (he referred to phenakistiscopes as the knock-off versions of his invention), preferring the term Phantasmascope. Other printers referred to the device as a Phantascopic, a Fantascope, a Pantinoscope, and so on. Without subject headings, these variant terms for an unusual item would render incomplete search results in aggregate catalogs like WorldCat, which is rectified by assigning a consistent classification to the object, regardless of the unique title. The Beinecke Library at Yale University’s copy of this item is entitled “The phenakistiscope, or, Magic disc” but is under the subject heading “Phenakistoscope” (note the variant spelling) and the genre “Toy and movable books,” linking it to two additional phenakistiscopes in the catalog and a whole slew of movable books. The Morgan Library’s copy, on the other hand, is only classified as “Toy.”
Basically, if you ever hit a wall with your library searching, scroll down and click on a subject heading. You might wind up with compelling results. Classification is not just the methods of deciding where an item belongs [on the shelf, in a database, in relationship with other materials], but the means of finding it again.
Part 3: Failure to File
This isn’t to say that classification systems are without flaws, which brings us to a new segment I like to call
DUNKING ON DEWEY
Classifications are systems of knowledge organization, and those systems are skewed based on how you (or me, or in this case, Dewey) think knowledge works. One of my professors described subject headings as a “textual map of the universe of knowledge.” I would amend this to say that they are a “textual map of a universe of knowledge.” Subject headings are selected, not inherent (just note the differences between the Yale and Morgan records for the same item), and those selections were made with biases. Biases that actively prioritize a white hegemony and exclude non-Western modes of knowledge organization.
Two examples of such biases from the presentation “Bias and Inclusivity in Metadata: Awareness and Approaches” (that I watched for class), by Julie Hardesty, a metadata analyst at IU Libraries:
In LCCS, there is one (1) subject heading for “Indians of North America” (subheadings: Indian wars and Indian tribes and cultures) which is so reductionist I want to fold myself into a box and drop into the Atlantic Ocean. There are 574 federally recognized Indigenous nations in the United States. This heading is too broad, too imprecise, uses language that these people would not use to describe themselves, and obscures the information of the items they describe.
Hardesty searched through LCCS subject headings assigned to United States Poets. Poets like Robert Frost and Allen Ginsberg were just labeled “US Poet.” Sylvia Plath got an extra “women poets” heading. Maya Angelou was also listed under “African American Author” and “African American woman author.”
Now you might be saying, “Shira, that’s a good thing, they’re being specific about who those people are.” Yes, true, they are being specific about Maya Angelou, but they’re being unspecific about Robert Frost. They presume (and we internalize) that “American Poet” means “white, male, European, Christian, middle-class” poet. Under this system, we differentiate Maya Angelou from that default instead of differentiating everyone from “poet.”
Update, November 26, 2021: this disambiguation of American/African-American, poet/woman poet has since changed.
And two more examples from my User Services and Tools class:
The Dewey Decimal System was created in 1876 by Melvil Dewey, a founding member of the American Library Association (who has since been excoriated for his discriminatory practices). He split his subjects into 10 classes, from 000-900, with subclasses denoted by the tens digit and sub-subclasses denoted by the ones digit, and so on. Class 200 is religion. Subclasses 200-289 are related to Christianity. 290 is “Other religions.” We finally get to Judaism at 296, and Islam, Babism, and Bahai are crammed into 297. This means every subdivision of any items in these headings are further denoted by extremely long strings of digits after the decimal point. What could be more symbolic of shunting anything “Other” to the wayside?
Another, even more stomach-roiling example: To Dewey’s mind, all of what happened and was ever going to happen regarding Indigenous Americans had happened in the past, so the category of “Native Americans” or “Indians” is a subsection of 970: “History of North America.” This means that even in 2019, Nina O’Leary’s book Native Enough was classified in the DDC under “History of North America.” Sit with that a moment. In this system, we still erase Indigenous people from our present.
What is there to do about this? The sheer volume of material that has been classified in LCCS or DDC makes any sort of overhaul utterly suffocating. One option is to use alternative controlled vocabulary classifications (think drop-down menus with pre-selected terms) at institutions and closed collections with specific focuses. The University of British Columbia Library uses the Brian Deer Classification System for Xwi7xwa materials (pronounced “weh-wah,” meaning “echo” in Squamish), rather than subsuming them into colonial LCCS subject headings. The Atria Institute on Gender Equality and Women’s History uses the European Women’s Thesaurus that include synonyms for broad subject headings. Homosaurus.org similarly uses linked data vocab as a companion to broader subject vocabularies.
Changes to classification systems are incremental. They involve a lot of technical stuff about linked metadata that I don’t totally understand yet, so I won’t try to explain and get it wrong and then spread misinformation (what a novel concept!)
The TL;DR: classification is how we find things, and different systems work better for different collections. If you recall books based on their covers and not their titles, organize your bookshelves in rainbow order. Amanda Gorman does it, you can too.
House keeping:
Some quotations from Alvis’s article that hit the spot:
“There is a tendency to equate a valuing of aesthetics with shallowness, as if appreciating something for its beauty automatically lowers the IQ.”
“Sure, some organization methods may not be your cup of tea, but that doesn’t make it wrong”
Update on the Case Study: I turned it in.
Also: I have learned my lesson. Writing this shindig the day I want to send it out is a rookie mistake.
Please subscribe for potentially shorter posts.
More later.