Harvard’s Secret ‘Computers’

A woman sitting at an early IBM computer.

Source Getty

Why you should care

Because 80 women toiled away to catalog the night sky and got almost no credit.

Astronomer Edward Pickering dreamed of mapping the night sky. When he began his tenure as director of the Harvard College Observatory in 1881, technological advances brought him closer to his dream by making it easier than ever to image stars. He arrived to a galaxy of data in the form of hundreds of glass photonegatives of the star-filled skies. The only problem? There weren’t enough staff members to catalog it all. So Pickering turned to computers — well before even the punch-card using ones existed. Back then, the “computers” were women.

Since cataloging stars was considered women’s work — tedious and “secretarial” — Pickering hired his maid, Williamina Fleming. She proved so accurate and efficient that he employed more than 80 “female computers” until his death in 1919. But they were far from brainless machines. Many of them made important contributions in their own right and continue to inspire groundbreaking discoveries, especially now that Harvard researchers are digitizing their work.   

Black and white, tight headshot of Pickering in suit and looking very seriously into the camera

Edward Pickering.

The women were banned from using the observatory telescopes, but that didn’t stop them from making crucial discoveries.

Observatory astronomers photographed the night sky onto glass plates until 1990, when they transitioned to electronic imaging. In 2004, Harvard researchers began digitally scanning the plates so that anyone with Internet access can click on a star catalog and pull up valuable information showing how it has brightened or dimmed over time. The team released its second data set in October and hopes to digitize all 525,000 plates by 2017.

It was grueling work. Some computers classified stars by comparing the photographs with known catalogs, while others cataloged the photos themselves, noting the date of exposure and region in the sky directly on the plates. They then transcribed this information, as well as each star’s brightness and position, into tables in their notebooks.The stars appear on the glass plates as tens of thousands of black specks dotting a light gray sky. Under Fleming’s supervision, the Harvard computers — also referred to as “Pickering’s harem” — scrutinized the glass plates in the observatory basement six days a week for 25 to 50 cents per hour. Men doing the same job would have made twice as much.

The women were banned from using the observatory telescopes, but that didn’t stop them from making crucial discoveries. Antonia Maury, for example, identified the second binary star — two stars orbiting around a common point — shortly after Pickering discovered the first. She was also the first to calculate the path of these stars’ orbits and the time they took to complete them. Pickering published Maury’s findings in 1890, noting that she had merely made “a careful study of the results.” He made no further mention of her. Maury left Harvard soon after.

“I do not think it is fair that I should pass the work into other hands,” she wrote to Pickering. “I should have full credit for my theory.” She returned to Harvard in 1895 to develop a system of classifying stars based on temperature, but left again after Pickering disagreed with it. A decade later, Danish astronomer Ejnar Hertzsprung used Maury’s system in his method for identifying giant and dwarf stars.

Black and white image of Annie looking at the camera sitting behind her desk in a dress.

Cataloging phenom Annie Jump Cannon.

The star computer was Annie Jump Cannon, who cataloged 350,000 stars and greatly simplified Maury’s system. The International Astronomical Union adopted her method in 1922 as the official classification system for stars, which astronomers still use today. But they didn’t name it after her. Instead, it’s called the Harvard system of spectral classification. 

Today, the digitized catalogs are still yielding breakthroughs. Researchers at Harvard and other institutions used Henrietta Leavitt’s discovery of the Cepheid variable stars for measuring interstellar distances to identify the next-generation cosmic yardstick, called Type Ia supernovae. The faintness of these stellar explosions indicated they were farther away than expected — a sign that the expansion of the universe was accelerating, rather than slowing down as previously thought. The researchers won a Nobel prize for their discovery in 2011.

Few people mentioned Leavitt at the time, just as with the other computers. While they may have toiled away in a basement, that doesn’t mean we should keep them there.

Editor’s note: An earlier version of this article, first published Dec. 7, 2013, did not adequately credit a source, Smithsonian Magazine.


OZYTrue Stories

The intimate, the harrowing, the sweet, the surprising — the human.