A deep dive into racial inequality in the literary world

A few years back, Richard Jean So began noticing discussions about racial inequality in the arts growing online. First, #OscarsSoWhite was trending. The literary world followed soon after: #PublishingSoWhite and #PublishingPaidMe exploded on Twitter, with authors publicly disclosing their advances in order to highlight stark racial pay disparities in publishing.

“I felt that writers and editors and artists understood what was happening, in some ways, better than scholars,” says So, an assistant professor of English and cultural analytics at McGill. “There was a discourse in the public that seemed really important, and I wanted to provide some scholarly research-based study to help understand how we got to this point.”

So’s new book Redlining Culture: A Data History of Racial Inequality and Postwar Fiction, brings together data science, literary history and close reading to examine the publishing industry’s longstanding bias towards white authors. 

The term “redlining” refers to the practice of arbitrarily refusing or limiting services to specific neighbourhoods where residents are poor or people of colour. In his research, So takes up this term to explore a practice of “cultural redlining” that shaped American literary publishing between 1950 and 2000. 

Literary scholars typically think of the postwar period, and especially the eighties and nineties, as one of increased multiculturalism in American literature. Indeed, these years saw the rise of many brilliant writers of colour, such as Ralph Ellison, James Baldwin and Toni Morrison. 

But despite the successes of a small number of authors of colour, So’s research finds there was “a thick line that walled off nonwhite writers from the coveted resources of not only lucrative book contracts but also book reviews, literary awards, and bestsellerdom.” 

In Redlining Culture, he claims this process of systematic exclusion led to a “consolidation of whiteness across the entire literary industry.” So argues that this determined both the types of authors that were published by mainstream presses and how stories were told. 

“Cultural redlining in the postwar period invented a model that made possible and has helped sustain racial inequality in the arts and literature today,” he writes. “To defeat it, we need to know its history.”

Drawing on massive datasets to paint a picture of the literary industry at scale, So’s findings are striking. Between 1950 and 2000, 97 per cent of authors published by the literary giant Random House were white. White novelists represented 98 per cent of the authors who appeared on bestseller lists, 91 per cent of those who won major literary awards, and 90 per cent of those who had books reviewed in prestigious literary magazines. (For context, in 2000, white people made up 75.1 per cent of the overall U.S. population.)

So’s research also reveals that literary institutions’ commitments to supporting nonwhite writers have always been short-lived, rather than ushering in long-term change. For instance, the number of authors of colour published by Random House increased significantly when Toni Morrison joined the publishing house as editor in 1967. But as soon as she stepped down from the role in 1983, that number dropped sharply. 

When So expanded his research until the present day for a recent op-ed in The New York Times, he found that racial inequality in the industry lives on. Analyzing data from major publishing houses between 1950 and 2018, 95 per cent of books were written by white people. In 2018, the figure stood at 89 per cent.

It’s not just a matter of numbers, either. In Redlining Culture, So analyzes texts at the language level to see how the dominance of white writers has shaped the way race gets narrativized in stories. 

The machine learning methods he uses can provide complex representations of literary texts, analyzing a novel’s themes or how characters are portrayed based on their race. (Because of his background in humanities and critical race studies, So developed a model that accounted for the fact that white characters are often not identified as such on the page—something many data scientists would have missed.) 

His text mining found that white characters from the period’s novels are generally portrayed in flattering terms, while Black characters are mainly rendered through negative racial stereotypes. White characters also tend to evolve throughout stories, moving through a range of subjectivities, while Black characters mostly remain stable or unchanging. 

Of course, there are exceptions to all patterns. Computer scientists typically remove outliers from datasets, since they don’t impact the overall trend. But So is quick to point out that in his research, outliers can represent works of literary genius. 

“I think what’s special about literature, art and culture is that individual writers or artists can make a huge impact,” he says. In his book, he conducts close readings of these outliers, like Octavia Butler’s Parable of the Sower, to show “racial minority authors meaningfully subverting and transforming the systems of power that seek to contain them.” 

In the last five years, authors of colour have dominated the winner’s circle for the National Book Awards in the U.S. – the top prizes in the fiction category have gone to books written by Colson Whitehead, Jesmyn Ward and other nonwhite writers. Could this mark a turning point?

So addresses the question in his book’s conclusion. 

“We’ve been here before,” he writes, pointing to “an uptick in the number of minority authors winning prizes in the late 1980s, an increase that proves relatively ephemeral and nonlasting.” 

It is one thing “to give a few more awards to nonwhite authors,” suggests So. “It’s far harder to change the broader idea of what the literary establishment defines as ‘prizeworthy’ based on what’s on the page.”