The difference engine

Earlier this month, within the space of less than a day, two significant events occurred in the life of Donna Strickland, an assistant professor at the University of Waterloo. She won the Nobel Prize in Physics, and she finally got her own Wikipedia page. As the biologist and Wikipedia activist Dawn Bazely writes in an excellent opinion piece for the Washington Post:

The long delay was not for lack of trying. Last May, an editor had rejected a submitted entry on Strickland, saying the subject did not meet Wikipedia’s notability requirement. Strickland’s biography went up shortly after her award was announced. If you click on the “history” tab to view the page’s edits, you can replay the process of a woman scientist finally gaining widespread recognition, in real time.

And it isn’t an isolated problem, as Bazely points out: “According to the Wikimedia Foundation, as of 2016, only 17 percent of the reference project’s biographies were about women.” When Bazely asked some of her students to create articles on women in ecology or the sciences, she found that their efforts frequently ran headlong into Wikipedia’s editing culture: “Many of their contributions got reversed almost immediately, in what is known as a ‘drive-by deletion’…I made an entry for Kathy Martin, current president of the American Ornithological Society and a global authority on arctic and alpine grouse. Almost immediately after her page went live, a flag appeared over the top page: ‘Is this person notable enough?’”

Strickland’s case is an unusually glaring example, but it reflects a widespread issue that extends far beyond Wikipedia itself. In a blog post about the incident, Ed Erhart, a senior editorial associate at the Wikimedia foundation, notes that the original article on Strickland was rejected by an editor who stated that it lacked “published, reliable, secondary sources that are independent of the subject.” But he also raises a good point about the guidelines used to establish academic notability: “Academics may be writing many of the sources volunteer Wikipedia editors use to verify the information on Wikipedia, but they are only infrequently the subject of those same sources. And when it does occur, they usually feature men from developed nations—not women or other under-represented groups.” Bazely makes a similar observation:

We live in a world where women’s accomplishments are routinely discounted and dismissed. This occurs at every point in the academic pipeline…Across disciplines, men cite their own research more often than women do. Men give twice as many academic talks as women—engagements which give scholars a chance to publicize their work, find collaborators and build their resumes for potential promotions and job offers. Female academics tend to get less credit than males for their work on a team. Outside of academia, news outlets quote more male voices than female ones—another key venue for proving “notability” among Wikipedia editors. These structural biases have a ripple effect on our crowdsourced encyclopedia.

And this leads to an undeniable feedback effect, in which the existing sources used to establish notability are used to create Wikipedia articles, when serve as evidence of notability in the future.

Bazely argues that articles on male subjects don’t seem to be held to the same high standards as those for women, which reflects the implicit biases of its editors, the vast majority of whom are men. She’s right, but I also think that there’s a subtle historical element at play. Back during the wild west days of Wikipedia, when the community was still defining itself, the demographics of its most prolific editors were probably even less diverse than they are now. During those formative years, thousands of pages were generated under a looser set of standards, and much of that material has been grandfathered into the version that exists today. I should know, because I was a part of it. While I may not have been a member of the very first generation of Wikipedia editors—one of my friends still takes pride in the fact that he created the page for “knife”—I was there early enough to originate a number of articles that I thought were necessary. I created pages for such people as Darin Morgan and Julee Cruise, and when I realized that there wasn’t an entry for “mix tape,” I spent the better part of two days at work putting one together. By the standards of the time, I was diligent and conscientious, but very little of what I did would pass muster today. My citations were erratic, I included my own subjective commentary and evaluations along with verifiable facts, and I indulged in original research, which the site rightly discourages. Multiply this by a thousand, and you get a sense of the extent to which the foundations of Wikipedia were laid by exactly the kind of editor in his early twenties for whom writing a cultural history of the mix tape took priority over countless other deserving subjects. (It isn’t an accident that I had started thinking about mix tapes again because of Nick Hornby’s High Fidelity, which provides a scathing portrait of a certain personality type, not unlike my own, that I took for years at face value.)

And I don’t even think that I was wrong. Wikipedia is naturally skewed in favor of the enthusiasms of its users, and articles that are fun to research, write, and discuss will inevitably get more attention. But the appeal of a subject to a minority of active editors isn’t synonymous with notability, and it takes a conscious effort to correct the result, especially when it comes to the older strata of contributions. While much of what I wrote fifteen years ago has been removed or revised by other hands, a lot of it still persists, because it’s easier to monitor new edits than to systematically check pages that have been around for years. And it leaves behind a residue of the same kinds of unconscious assumptions that I’ve identified elsewhere in other forms of canonization. Wikipedia is part of our cultural background now, invisible and omnipresent, and we tend to take it for granted. (Like Google, it can be hard to research it online because its name has become a synonym for information itself. Googling “Google,” or keywords associated with it, is a real headache, and looking for information about Wikipedia—as opposed to information presented in a Wikipedia article—presents many of the same challenges.) And nudging such a huge enterprise back on course, even by a few degrees, doesn’t happen by accident. One way is through the “edit-a-thons” that often occur on Ada Lovelace Day, which is named after the mathematician whose posthumous career incidentally illustrates how historical reputations can be shaped by whoever happens to be telling the story.  We think of Lovelace, who worked with Charles Babbage on the difference engine, as a feminist hero, but as recently as the early sixties, one writer could cite her as an example of genetic mediocrity: “Lord Byron’s surviving daughter, Ada, what did she produce in maturity? A system for betting on horse races that was a failure, and she died at thirty-six, shattered and deranged.” The writer was the popular novelist Irving Wallace, who is now deservedly forgotten. And the book was a bestseller about the Nobel Prize.

Quote of the Day

I’ve been letting the weather and my stomach muscles and a great chord change in a Pretenders single make up my mind for me, and I want to do it for myself.

—Nick Hornby, High Fidelity

More sights, more sounds

“What really matters is what you like, not what you are like,” John Cusack’s character notes in High Fidelity, and this sentiment goes a long way toward explaining why we find lists of all kinds so fascinating. As I’ve argued before, a list of one’s favorite books or movies is as close to an honest self-portrait as any of us will ever come, and this isn’t a recent convention: as far back as the Iliad, we encounter the ascending scale of affection, in which a hero defines himself by ranking what matters to him most. (Quick story: Back in college, soon after High Fidelity came out, I pointed out this similarity to one of my classics professors. Later that week, I went to see the movie a second time—and saw my professor sitting three rows in front of me. The following day, he entered the classroom and said: “Alec, you’re my pop culture hero.” And that was the high point of my career as a classical scholar.)

This is what makes the Sight & Sound poll so irresistible. Most of the coverage has revolved, understandably enough, around the displacement of Citizen Kane by Vertigo at the top of the list, but the real story lies further down, in the lists of individual critics, which were posted on the site this morning after a short delay. Reading a critic’s list gives us as accurate a thumbnail sketch as we can possibly have of a stranger’s personality, tastes, and idiosyncrasies: I don’t think there’s any way to learn more about a person in thirty words or less. When I look at the list of author Kim Newman, for instance, the fact that he named both A Canterbury Tale and Duck Amuck tells me more about him in five seconds than I’d probably learn from reading one of his books. The same goes with critic Mark Kermode, whose list includes Brazil, Don’t Look Now, and Mary Poppins.

Looking at the top 250 offers even more food for thought. If I’d been surprised earlier by the absence of Powell and Pressburger from the top fifty, the explanation is readily at hand: every single one of their great movies made the long list—The Red Shoes, yes, but also A Matter of Life and Death, The Life and Death of Colonel Blimp, A Canterbury Tale, Black Narcissus, and I Know Where I’m Going!—which suggests that without a consensus choice, all these classic films simply split the vote. (When we see the list of directors ranked by number of total votes, I expect that they’ll be in the top ten.) I was delighted to see that the second-highest Kubrick movie on the list, after 2001, is Barry Lyndon, and that Miyazaki is represented by both Totoro and Spirited Away. And the short list of movies from the past few years to make the list is a fascinating one: The Tree of Life, There Will Be Blood, WALL-E, Uncle Boonmee Who Can Recall His Past Lives, Melancholia.

As always, the list provides ample occasion for reflection, argument, and education. It tells me that the director whose work I need to seek out most urgently, along with Tarkovsky, is Apichatpong Weerasethakul. It’s a reminder that critical tastes can change radically over time, as we see in the critical ascent of such movies, overlooked at their first release, as Vertigo, Rio Bravo, Imitation of Life, and Singin’ in the Rain, not to mention The Texas Chainsaw Massacre. It tells me that I wasn’t entirely wrong, seven years ago, about the enduring reputation of Wong Kar-Wai’s 2046, which got votes from three critics. And it tells me that my own tastes lie more or less within the mainstream, with a few outliers: of my own recent top ten, the only two not to make the cut were L.A. Confidential and Star Trek II: The Wrath of Khan, neither of which received a single vote—which only confirms that in some respects, I’m still ahead of the curve.