The inceptive question at the core of this series of visualizations is a simple one that comes with an answer attached: Are my bookshelves (i.e. the material I’ve chosen to procure, curate, and consume) biased politically, temporally, geographically, or otherwise? Predictably, the response to this query, be it answered regarding my bookshelf or anyone else’s, is “yes, of course.” However, what my inquiry is intended to illuminate is the degree to which my book collection reinforces my preexisting politics or weltanschauung rather than challenging my biases or providing a broader scope of insight, experience, etc. While quantifying evidence of my biased procurement of texts is anything but shocking, it is a necessary and sobering reminder that “filter bubbles” don’t purely exist as automated algorithmic processes driving one’s digital behavior; they are equally a product of our creation through our elected patterns of consumption (e.g., there are 147 blatantly leftist books that I’ve purchased in my dataset – what is this if not hypocritical over-indulgence in my own politics via engagement with an economic framework that such texts are critical of?).
Ultimately, these visualizations are less about illustrating that which could already be assumed (i.e., “my books represent bias”) and more about producing a model of analysis that explicates these acknowledged biases and their, at times, inconspicuous reinforcement via the seemingly “bias-destroying” pedagogical practice that is expansive reading and library-building. For these reasons, the intended audience for these visualizations is anyone with a bookshelf. Language is inherently political and should be critically considered as such. The commodity fetishization of a text, as tends to be our default relationship to a book, disregards its connection to the historical material conditions that produced it and the ideology at work therein. Ideally, analytical practices such as that which this project aims to at least scratch the surface of will work to encourage the considered examination of one’s personal library and the self-concocted filter bubble that might exist in one’s color-coordinated collection of Crown Publishing Group celebrity politician memoirs or in a Funko-Pop bespeckled shelf showcasing each of Brandon Sanderson’s six-hundred books.
In what follows, I’ll provide a brief explanation of each data visualization:
Number of Books on Shelf by Country: This spatial representation plots each of the 417 books in my data set on a world map to illustrate the limited geography that my book collection covers. This is intended to demonstrate the “gaps” in my collection.
Number of Books on Shelf by Original Language: Similar to the map, this treemap is intended to highlight the anglo-centrism of my bookshelf and the potential for bias that such an unevenness of experience and world-view might present.
Number of Books on Shelf by Genre: This is a simple breakdown of my bookshelves by reductive genres. The intent of this (and its positioning early in my visualization story) is to give a brief overview of my bookshelf by subject.
Books on Shelf by Year of Publication: This area chart is meant to showcase the temporal biases inherent in my collection. As one can see, the majority of my books come from the post-war period, confining my exposure to experiences temporally located within the last 70 or so years.
Political Affiliation of Authors on Bookshelf: Another simple breakdown of my bookshelves by the reductive political affiliation of the authors present. While my initial aim was to include more diverse categories of political thought, I found that reducing the dataset to a primarily right vs. left binary allowed for easier, cleaner data visualization.
Political Affiliation of Authors by Genre: The aim of this visualization is to locate which genres are comprised of the most outwardly political authors. For example, the substantial difference in my reading of leftist accounts of history from that of my collection of right-leaning historical authors paints a clear picture of how my understanding of certain histories might be overwhelmingly colored by leftist ideologies and interpretations.
Political Affiliation of Most Prominent Authors on Bookshelf: This visualization simply showcases the most prominent authors on my bookshelves (by number of books) and their politics. One’s number of books per author is generally an indicator of one’s most revered writers so it could be assumed that the political biases represented hierarchically here are representative of one’s worldview, values, etc.
Political Affiliation of Books by Means of Procurement: Through this project, I became interested in the way in which I come to own books, especially because I’ve had such strange luck finding books I want on stoops and sidewalks. It is notable that in both the “Found” and “Purchased” categories leftist texts are overwhelmingly prevalent while the politics of the selection of books I’ve been gifted are more ambiguous.
Political Affiliation of Books Published by Most Prominent Publication Companies on Bookshelf: This visualization is meant to indicate the potential biases of publishing companies. However, seeing as I am the one curating what I purchased from each company, this information is likely not indicative of anything other than my purchasing habits.
Political Affiliations by Genre: This stacked bar visualization is meant to further reveal where the biases of my collection exist by genre. For example, the breakdown of the political philosophy genre highlights an absurd degree of bias toward leftist political thought.
The production of the dataset behind each of these visualizations was a laborious process of incessant Googling and micro-decision-making. Unlike our present moment of hyper-politicization in which anyone’s ideology can be easily located in a Twitter bio, authors active prior to the Digital Age (outside of genres such as philosophy) have been less forthcoming with their dogmas, opting instead for allowing their beliefs, at times complex and irreducible to a reductive political binary, to subtly shine through the stories they tell (novel, huh?). However, this lost art of nuance is not conducive to easily deducing an author’s politics in the production of a reductive dataset, thus requiring me to define many of the figures listed in my data as politically “Unknown.” Using “It’s Complicated” to fill in the gaps and represent thinkers whose politics either shifted or never settled throughout their lives, I attempted to cover as much territory as possible with what ultimately amounts to a simplistic right vs. left analysis. My decision to include the “Means of Procurement” was largely rooted in my desire to understand the way in which I engage with the market in the acquisition of my assemblage of personal propaganda. As noted, many of my books have come from circumstantial discovery on the street and I found the notion that New York City exists still as a “filter bubble” despite the chance-discovery of free texts to be interesting and worthy of exploration. Certain information still lingers in this dataset unused, such as page counts and condition of the text upon procurement (used vs. new), offering the potential of expanded analysis. However, to provide answers to the base question regarding bias, I didn’t feel as if these columns of data offered anything wildly illuminating.
In the future, I’d like to extend this dataset to include sub-genres of texts and more thorough information about each author. The most noticeable inadequacy of this data is undeniably the reductive nature of its content, largely stemming from our finite timeframe and the limited information available when trying to restrain each time segment of research on an author to under a minute. However, if extending the dataset appears as too arduous a task, I think reducing the book count to focus on a single genre (such as philosophy) might allow for a more thorough analysis of the complexities of the political biases therein (rather than right vs. left, actual political affiliations could be quantified and considered). All in all, I think (and hope) that the expansive collection of texts represented through my data (despite their monotonality) and the critical consideration represented through clean bar graphs and geographical visualizations communicate clearly my intentions and work to stimulate further reflection regarding the texts we choose to keep and their role in the production of our subjectivity.
Link to “A Bookshelf & its Biases”
