My presentation at ASAP/9 was blighted by an unfortunate PPT error. While this provided me ample opportunity to practice describing visual networks (“It looks kinda like a firework?”), I thought I’d share the second half of the talk as-written, complete with visual aids.
[Intro: Today I want to make a modest proposal: that we return some agency to the literary agent. My goal is to think seriously about the powerful role that the literary agent plays in contemporary literary production, from the form and content of fiction to the development of international literary trends. My talk is self-consciously descriptive: this project began with ethnography—that is, interviews that I conducted with leading agents over the course of six months. … I also employ computational network analysis to demonstrate that contemporary fiction is agented fiction: that the agent is a co-producer of a manuscript. That generic trends can be traced to agents’ tastes. That their influence is international.]
Part Two: Influence, Networked
What would happen if we restored the agent to a place of centrality in the literary field? What networks of reading might open up—what familial resemblances explored, what portrait of gatekeeping uncovered— when we acknowledge the centrality of the agent and her labor?
To begin to visualize the agent’s role as tastemaker, I want to turn to network analysis, a computational method that allows us to analyze relationships at the scale of the field as a whole.
This network graph shows the connections between all of the authors shortlisted the Man Booker, the National Book Award, and the Pulitzer Prize for Fiction since 2000 and the agents who represent them. The central premise of network analysis is that relationships, affiliations, and dynamics can be represented by a series of interconnected points in two-dimensional space. Agents are the central nodes (red), where authors (blue) cluster around the agent; a line—called an edge—connects the two. The more nominated authors an agent represents, the larger the agent node. Edges are weighted according to the number of times an author has been nominated for a prize—the thicker the line, the more nominations an author has received. 
There are, of course, some major caveats here. A network of bestselling authors would look pretty different. Likewise, this represents just three, English-language prizes. I am happy to get into the methodological and statistical weeds during Q&A. Even still, the yield of a network is the unforeseen relationships and connections it reveals.
So what does this show us? First, the majority of pairs are single author/single agent. Very few agents represent more than one nominated author, and those who have usually represent three or more. Despite how diffuse the field might appear, clusters have formed around some of the more influential agents—Andrew Wylie, Eric Simonoff, Peter Strauss, Amanda Urban. For now, it is less the density of particular clusters than the connections that interest me.
The edges between agent and author—and the implied connections between authors—reveal the contours of an agent’s taste. On the surface, books like Marilynne Robinson’s Home and Marlon James’s A Brief History of Seven Killings bear very little formal or topical similarity. Yet, they are united by virtue of their shared representation by Ellen Levine; each is, in its own way, in tune with Levine’s literary taste and sensibility. Placing these texts in conversation with one another may not do much for our understanding of either Robinson or James, but it might throw into relief the mechanics of at least one of the dominant tastes in the literary marketplace, beginning to clarify that which is often dismissed as unquantifiable and elusive.
The case of Nicole Aragi particularly instructive.
Representing Colson Whitehead, Junot Díaz, Julie Otsuka, Aleksander Hemon and Rabih Alameddine, Aragi is responsible for the representing and promoting many prominent authors of color, who write explicitly about issues of race and ethnicity in the United States, and who share an interest in manipulating literary form while remaining commercially accessible. While not all of Aragi’s clients fit this description, the resemblance between this group is striking. In her decision to represent certain novelists and to pass on others—and, in particular, in her choices to represent texts that discuss the experience of immigrants and first-generation Americans (Díaz’s The Brief and Wondrous Life of Oscar Wao and Hemon’s The Lazarus Project) or novels that chronicle atrocities perpetrated against people of color in the US (Whitehead’s Underground Railroad and Otsuka’s When the Emperor Was Divine)— we can and should read Aragi making a pointed intervention in the very white world of corporate publishing. These choices are not incidental, and not exclusively profit-driven.
While this network helps us better understand the creative agency exercised in a backlist, it shows us less about an agent’s influence in the literary field. Let’s shift the focus from the backlist to the prizes.
Here, you can see colored nodes at the three vertices of this triangle: the National Book Award in red, the Man Booker in blue, and the Pulitzer Prize for Fiction in green. The edges connecting agent and prize indicate that an agent’s client has been nominated for that award; the agent nodes are sized in proportion to the number of nominations.
This network is significant in terms of its spatialization of influence. This graph demonstrates just how far a single agent’s tastes can reach. Here, on the periphery, are agents whose authors have been nominated for just one award. Here, in the center of the triangle, are the agents whose authors have been nominated for multiple awards: these are the most decorated—and, by extension, most influential—agents of the 21st century.
There is so much more we could talk about here, that I do not have time to get into. What I want to point out, though, is that Nicole Aragi is located at the centermost point of our network. Aragi’s tastes are not only well-defined, as we saw in our last network; her literary critical acumen has resulted in widespread prestige and influence—recognition on both sides of the Atlantic, spurring conversations about and critiques of the whiteness of the Big Five and contemporary American literature on an international scale. The case of Nicole Aragi—as with Andrew Wylie or Peter Strauss—demonstrates the ways in which a single agent’s distinct taste has the potential to influence a large segment of the market.
[NB: Just for fun (i.e., cut from the talk for time), I made an additional network that included authors. It’s a little convoluted and challenging to read, but it offers a lot more detail.]
So. Where does this leave us? It is high time that we restore creative agency to the literary agent, acknowledging the role that they play in the creative process, their representation of work they believe in, and their contribution to the literary community. Agents are shaping the field of contemporary literary production—and our understandings of what it means to create, to collaborate, and to exercise one’s personal taste or make claims to autonomy within it—in ways we have yet to fully acknowledge. Prominent cultural producers, gatekeepers, and tastemakers, agents play a central role in today’s high-risk, conglomerated literary field; their influence will only continue to grow, shaping the forms and fictions of literature in the contemporary.
 This graph was made using the open-source platform Cytoscape and the plugin Allegro Layout. Raw data is available in a GitHub repository at http://github.com/lbmcgrath.
 All authors represented on this graph have been at least nominated, and their edges are weighted at a default 1. 1 point is added to the edge weight for each nomination (i.e, two nominations result in an edge weight = 2. I do not distinguish mathematically between winning a prize and being nominated for a prize (i.e., one nomination and one win result in an edge weight = 2, two wins also result in an edge weight = 2).