That image has haunted the women who worked on Eniac and has contributed to the misconception that they were little more than low-level technicians who set up the machine according to the instructions of others.

To understand their true role, a brief review of the Eniac's history and of developments just prior to World War II is necessary.

If you ever sweated over a differential equation, consider what it would feel like if one could save your life, not just your GPA. During both World War I and World War II, ballistic trajectories were calculated by hand. The results were an attempt to mathematically model every possible field condition while taking into account the weight and shape of the shells and their propellant charge. The data as compiled into firing tables, which field gunners would look up often while under fire-to aim their weapons.

That worked fine when "the targets were being pulled by mule s," observed Lt. Col. Herman H. Goldstine of the Ballistics Research Laboratory, who was a first lieutenant in World War I. But by World War II, the targets were airplanes, and the United States was creating guns that could fire at supersonic speeds.

To keep up, the army sought college-educated women to calculate trajectories. "Able-bodied men were all needed for the war; even Pres[per Eckert, co-inventor of the Eniac] had to get an exemption for essential work from the Secretary of War," said Kay Mauchly Antonelli, who was the second wife of John Mauchly, the Eniac's other co-inventor. Mauchly Antonelli-born Kathleen McNulty-was recruited through "an ad that looked like a 'wanted' poster. It said, 'Wanted: Women With Degrees in Mathematics.' " The starting annual salary of $1,620 was very good for the time, she said.

"We had to calculate where the bullet would be every tenth of a second, and just to find out where it w as after a tenth of a second required over six multiplications, a couple of square roots and a couple of divisions and subtractions," said Mauchly Antonelli. Given those requirements, it could take up to three months to produce a complete firing table. The need for "computers"-as the women were called-grew so great that the Army looked beyond the few female math majors of the time. Lt. Col. Goldstine's wife, Adele; John Mauchly's first wife, Mary; and Mildred Kramer were hired to teach sophisticated mathematics to the women, who ranged from recent high school graduates to senior citizens. But even with the new recruits, it was impossible to keep up. "The number of tables for which work has not been started. . . far exceeds the number in progress," Goldstine wrote in 1944.

Like society pages
Goldstine is credited with convincing the Army to invest in a digital machine that could automate the calculations. But the Eniac was not completed until after the end of the war. Ironically, in Goldstine 's technical history, The Computer from Pascal to von Neumann , his mention of the women who programmed the Eniac reads like the society pages: "They were the Misses Kathleen McNulty, who subsequently married [John] Mauchly; Frances Bilas, who married Spence; Elizabeth [Betty Jean] Jennings, who married a Moore School engineer; Elizabeth Snyder, who married [John] Holberton; Ruth Lichterman; and Marilyn [Marlyn] Wescoff."

This glancing mention has helped add to the misunderstanding of the women's contributions. As Jean Bartik (née Betty Jean Jennings) described it, "He got about 90 percent of the facts right. One thing he got wrong was us."

Contacted recently, Goldstine said he was "trying to get the book done" and regretted the oversight, especially in the case of Betty Holberton (born Elizabeth Snyder), who later served on the standards committees for Cobol and Fortran.

A manual is born
Because the Eniac was not yet finished, the women had to learn its operations fro m wiring diagrams and block diagrams, and figure out for themselves what sequences worked best. What seemed at first to be a liability turned into an asset. A manual was not composed until 1946 and was written by Adele Goldstine. "I was astonished, after a year of working on the Eniac, to find out Adele had written this manual," said Mauchly Antonelli, proving again that the more things change in computing, the more they stay the same.

Imagine the Altair from Hell. The Eniac was a gigantic beast of cables and wires, and programs were run by directly setting its dials and switches. That is what the women are doing in the famous, misleading photograph. What was required before you could touch those switches was the knowledge to break down differential equations into difference equations, and a further knowledge of the sequence in which those operations needed to be set on the machine. In addition, the buses were separate, so one had to consider conflicts in the instruction paths and data paths. "We were dow n at the system level," said Bartik, "a physicalization of if/then statements."

It was a task made awesomely complex by the Eniac's original ability to perform in parallel. But there was a problem in that different operations took different amounts of time. "It had been designed to have as many accumulators as there were columns in the sheets of paper on which we did the trajectories by hand," Mauchly Antonelli explained. "It was like a giant spreadsheet, with all the sums designed to go in at the same instant. But multiplication took longer; if you needed the answer to a multiplication, you timed it so you
could get the answer before the addition you needed." The timing problems became so complex, it was finally decided to change the Eniac's operations from parallel to serial.

Adele Goldstine died of cancer in 1964. Although all of the surviving members of the team remember her fondly as "fun-loving" and "brilliant and hard-working," her role is now confused. She did teach; she did write the Enia c manual. But how much she actually contributed to the transformation of the Eniac into a stored-program computer (and for that matter, how much John von Neumann actually contributed) is disputed by Herman Goldstine and Jean Bartik.

"Adele consulted to Los Alamos," Lt. Col. Goldstine said. "Von Neumann set her the task of changing it to a stored-program computer. She gave it to Dick Clippinger [of the Windtunnel Group of the BRL], who then changed it and simplified it some."

Bartik remembers the story differently. "Everybody claims it was Johnny von Neumann because he published. But John [Mauchly] and Pres [Eckert] had already thought about it," she said. "Von Neumann was basically a consultant on the instruction set. Dick Clippinger's problems were too big to go on the machine, so he was really the force that provided the money to do all this.

"Dick got me assigned to him, to teach him how to program. We'd hide down in the furnace room, to get some time alone to go over it. We did go to Princeto n to consult with Adele and von Neumann, but it's ridiculous to say 'he gave it' to Adele to do."

As word of the Eniac got out, brilliant mathematicians and physicists from around the world came to seek solutions for what were then known as "hundred-year problems"-problems thought too complex to be solved within a human lifetime. The women programmers worked with them to make their problems processable by the Eniac. Here, their work began to resemble programming as we know it today-as a separate skill set necessary to interface with a machine that must be learned in addition to, and separately from, higher math.

On March 15, Betty Holberton received the Augusta Ada Lovelace Award in Arlington, Va. The Lovelace Award is presented by the National Association for Women in Computing to individuals who have excelled by making a contribution on behalf of women in computing. Holberton is amused and delighted by all the public recognition that has come so late in her life. "I received all these cards from gir ls at Microsoft," she said on March 7, her 80th birthday.

Those birthday cards have another significance. The new recognition of the women of Eniac may offset some of the over-emphasis on Admiral Grace Hopper. The problem with that emphasis is that it risked leaving the next generation with the sense that only one woman aside from Ada, Countess Lovelace, ever succeeded in the field. "We don't need an icon," said Kathryn Kleiman, a telecommunications attorney with Fletcher, Heald and Hildreth (Rosslyn, Va.). "We need role models."

To that end, Kleiman has been researching the history of the Eniac women. As a Harvard undergrad, she discovered a reference to them in a footnote. Recently, as the result of an article in The Wall Street Journal , she has received the funds for an oral history documentary. Kleiman, a programmer herself, will ask the women to detail their technical contributions.

It is hoped that the work of Kleiman and others who are investigating the roles women have played in th e history of computer science will help convince younger women that Barbie was wrong: Math isn't hard. There are historical precedents for women in technical careers.

"I never had any feeling from the engineers we worked with-all of whom were men-that we were less capable," said Mauchly Antonelli. "The problem was with the Civil Service. [Originally,] only men could receive professional ratings."

Bartik also praised the engineers and succinctly summed up why their respect was mutual. "They found out we knew what the hell we were doing," she said.

Back to

Back to

Free Subscription to EE Times First Name Last Name Company Name Title Email address   Click here for your Free Subscription to EETimes Europe