The Bumbling 1960s Data Scientists Who Anticipated Facebook and Google

And what a shop it was. Despite Lepore’s repeated references to the Simulmatics team as “the best and the brightest,” the group that Greenfield assembled was, well, not that. There was Bill McPhee, a manic-depressive mathematician who wrote some of Simulmatics’ early data analysis programs from a locked ward on Bellevue and had a habit of abusing and humiliating his wife in front of their 6-year-old daughter; Eugene Burdick, a political scientist / literary celebrity / Ballantine Ale pitchman whose writing, according to Lepore, was less subtle than a sledgehammer; and Ithiel de Sola Pool, “a numbers guy who taught at M.I.T. and walked the halls of the Pentagon” but who seemed to lack the basic competency and knowledge expected of a scientist. (Later in his career, Pool’s writings helped ensure the internet would be free of government regulation. In an interview with Lepore, Daniel Ellsberg, the Pentagon Papers leaker and a onetime friend of Pool’s, called him “the most corrupt social scientist I had ever met, without question.”)

In its decade of existence, Simulmatics helped this newspaper report early results of the 1962 midterm elections, developed strategies for “selling shampoo and dog food,” predicted where and when urban rioting would occur, analyzed communications in communist countries and helped direct the United States’ disastrously murderous counterinsurgency efforts in Vietnam. Or, rather, it tried to do those things; Lepore undermines her attempt to elide over the differences between Simulmatics’ ambitions and its accomplishments by quoting or summarizing post-mortems from the company’s clients: Its election-night collaboration with The Times was “a completely disorganized shambles”; its work in Vietnam was “dubious and its methods questionable”; the studies it put together for a national commission on civil disorders suffered from “poor design, lack of expertise and misrepresentation” and were “reprehensibly sloppy.” The coup de grâce comes courtesy of the Department of Defense’s Advanced Research Projects Agency during a period in which the department was responsible for 70 percent of Simulmatics’ annual revenue: “Simulmatics reflects discredit not only upon itself as an organization — it appears more a sham — but upon behavioral research in general.”

Over the last decade, Lepore, a Harvard history professor and New Yorker staff writer, has repeatedly shown herself to be an uncommonly astute and insightful interpreter of American history, and one of her many strengths is the moral clarity that infuses her writing. Lately, however, it feels almost as if she’s trying to stanch the flow of hatred, misinformation, racism and venality that threatens to overwhelm the country by the sheer volume of her work: “If Then” is her third book in three years, and that includes 2018’s “These Truths,” a monumental history of the United States. So far in 2020, she’s also released a 10-part podcast “about the history of truth” and published 12 articles in The New Yorker. (I’m overwhelmed thinking about consuming that much information, never mind producing it.)

This prolificness likely explains why her latest effort feels as if it was rushed out the door before it was ready. When she’s at her best, Lepore’s writing has a nimble fluency that can be exhilarating. Here, however, events are described out of order, crucial context is missing and stylistic tics become intrusive: McPhee “drank and he drank and he drank and he smoked and he smoked and he smoked”; his long-suffering wife was forever “demurring, demurring, demurring, demurring.” By the time Lepore describes Ed Greenfield’s hunger for McPhee’s “voting prediction machine” — he “wanted it, wanted it, wanted it” — I couldn’t resist penciling in the margin, “I get it, I get it, I get it.”

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: