Later On

A blog written for those whose interests more or less match mine.

Arianna Rosenbluth Changed the World Before Leaving Science Behind

leave a comment »

By and large, society during my generation treated women badly (and by “society” I mean “men” but also organizations (overwhelmingly managed by men) and social conventions). This article provides one example but there are many others. One example often offered is how Crick and Watson used Rosalind Franklin’s findings but failed to credit her. That particular story does not correspond to the facts but there are many others — for example, how long Wally Frank had to wait before she was finally able to take a suborbital flight yesterday.

Here’s an example that is true, which Anastasia Carrier recounts in the Harvard Gazette. She writes:

A few years ago, Jean Rosenbluth was visiting her mother at a nursing home in Pasadena. The occasion was a holiday party, and Jean and her husband were seated with her mother and another couple. It came up in conversation that the man sharing the table was a history of science professor, specializing in physics.

“Oh, my mother was a physicist,” Jean said as she introduced her mother. “This is Arianna Rosenbluth.”

The professor was stunned. “Wait, the Arianna Rosenbluth?” Arianna smiled shyly and kept eating her lemon meringue pie.

Arianna Wright Rosenbluth, who received a master’s degree in physics from Radcliffe College in 1947, was one of five scientists who created the revolutionary Metropolis algorithm—the first practical implementation of what are now known as the Markov Chain Monte Carlo methods, go-to tools for solving large, complex mathematical and engineering problems.

Over the years, these methods have been used to simulate both quantum physics and markets, predict genetic predisposition to certain illnesses, forecast the outcomes of political conflicts, and model the spread of infectious diseases. It was Rosenbluth who found a way to get early computers to use the Markov Chain method, creating a blueprint that others followed.

“Arianna’s impact would last for a long time,” says Xihong Lin, a professor of biostatistics at the Harvard T.H. Chan School of Public Health, who used Markov Chain Monte Carlo methods to analyze a large set of COVID-19 data from Wuhan and to calculate the infectiousness of the virus. The methods have also helped specialists evaluate the effectiveness of quarantine and stay-at-home measures.

“Without Rosenbluth, I don’t think the field of Markov Chain Monte Carlo would go that far,” says Lin, referring to the role of the Radcliffe-trained scientist in enabling wide use of the tool across disciplines. “Implementation is critically important. That’s why her contribution is a landmark and really should be emphasized—should be honored.”

The paper that Rosenbluth coauthored—along with her then-husband, Marshall Rosenbluth, Edward and Augusta Teller, and Nicholas Metropolis—was published in 1953, but the algorithm’s origin story remained a mystery for five decades. In 2003, Marshall shared his memory of the achievement during a conference celebrating its 50th anniversary. The researchers developed the tool to illuminate how atoms rearranged themselves as solids melted, he said. Marshall did most of the conceptual work, and Arianna translated their idea into a computer algorithm—a task that required a fundamental understanding of physics and computer science, and also creativity.

By all accounts, Rosenbluth, who died of COVID-19 complications in December at age 93, was brilliant. She earned her PhD in physics at Harvard at 21 and in her short career worked under two physicists who went on to earn Nobel Prizes. And yet she effectively quit science in her late 20s, leaving her job at the Los Alamos Scientific Laboratory to be a stay-at-home mother. She rarely spoke about her time in the lab—although she sometimes mentioned to her children how irritating it was that her ideas were overlooked because she was a woman trying to make it in a male-dominated field. Other times, she would lovingly describe MANIAC I—the Los Alamos machine that she used for computing the Metropolis algorithm.

“She was ahead of her time,” says Pierre E. Jacob, the John L. Loeb Associate Professor of the Natural Sciences and a professor of statistics in the Harvard Faculty of Arts and Sciences, whose work involves Markov chains and probability modeling. In his syllabus, he renamed the Metropolis algorithm the Rosenbluth algorithm after reading about Arianna’s death.

“Better late than never,” he says.

Star on the Rise

Growing up in Houston, Arianna Wright was a mystery to her parents.

“Her mom and dad had this genius child, and they kind of didn’t know what to do with her,” says Mary Rosenbluth, one of Arianna’s four children. Leffie (Woods) Wright was confused by her quiet and introspective daughter, who didn’t care for fashion and rules but loved reading, especially fantasy books like L. Frank Baum’s Wizard of Oz series. Mary recalls a newspaper article among her mother’s things that described Arianna as a child genius.

“It kind of struck me,” she says. “Here’s this girl growing up in suburban Houston, and she was just so different from everybody else.”

Arianna received a full-ride scholarship to Rice Institute (now Rice University) in Houston and took a bus to her classes. She earned her bachelor’s when she was 18, with honors in physics and mathematics. During her college days, she fenced against men as well as women, winning city and state championships. She qualified for the Summer Olympics in 1944, but World War II led to the cancellation of the games. She qualified again four years later but couldn’t afford to travel to London.

At Harvard, Arianna was rejected by one potential advisor because he didn’t take female PhD students, says Alan Rosenbluth, Arianna’s oldest child and a retired physicist. That was not uncommon. “Women were discouraged every step of the way,” says Margaret W. Rossiter, a Cornell historian of women in science. But Arianna forged ahead, in 1949 becoming just the fifth woman to earn a PhD in physics from Harvard.

She accepted a postdoctoral fellowship funded by the Atomic Energy Commission to study at Stanford University, where she . . .

Continue reading. There’s much more.

It strikes me that Arianna’s recognition of how to express the ideas of Markov chains using code in a computer algorithm bears a passing resemblance to how Zhi Bingyi’s recognition of how to express Chinese characters using the Latin alphabet on a keyboard (as described in the previous post).

Written by Leisureguy

21 July 2021 at 11:29 am

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: