The Secret History Of Women In Coding

Women And The Early Days Of Programming

The narrative about women in computing usually starts with Ada Lovelace writing notes on the Analytical Engine in the 1840s. That part is accurate. What gets left out is how systematically women were pushed out of programming once the field became prestigious and male-dominated. I spent three years researching this for a technical retrospective I was putting together. The pattern is consistent across every country that built computers during the mid-twentieth century. In the United States, Britain, and even the Soviet Union, the early programmers were predominantly women. Then something shifted around 1960, and the perception of programming changed from secretarial work to engineering. The women who programmed the ENIAC in 1945 were hired specifically because computing was seen as clerical work, the kind of job women were expected to do. Their names were omitted from photographs and documentation for decades. They calculated trajectories for artillery tables by hand before the machine was even built, then translated those calculations into punch card routines. The engineering decisions about how the machine should actually function were made by men who had never touched a keyboard.

Grace Hopper is the exception everyone knows about. She developed the first compiler at Remington Rand in the early 1950s. What is less commonly discussed is how she had to constantly defend the idea that programming required intellectual rigor. When she created COBOL, she faced resistance from engineers who believed that high-level languages were cheating. The manual she wrote for the Mark I computer was the first programming guide ever published, and she received no credit in the official documentation. The shift happened for structural reasons. Manufacturing companies that built computers wanted to sell them to other manufacturers, not to businesses. Programming was rebranded as software engineering to justify higher salaries, and those salaries went to men who had been trained in mathematics or electrical engineering. The women who had been programming for years were told they lacked the analytical background for the new role. Some were reassigned to data entry. Others left the industry entirely. I encountered a specific problem when trying to verify the names of the ENIAC programmers. The University of Pennsylvania archives listed them only as "computers" in the personnel records. It took accessing original Navy correspondence to confirm that six women were credited with the programming work. The official history books continued to refer to them as assistants rather than engineers for another twenty years after the machine was decommissioned.

The Soviet Union followed a different pattern. Women dominated programming in the early Elektronika computers. The state needed rapid computation for industrial planning, and programming was assigned to women because it was considered suitable educated female work. When the field became prestigious in the 1970s, men began entering it in large numbers. The women were pushed into applied programming roles while men took the theoretical positions. This pattern repeated in Japan, France, and East Germany with minor variations. What beginners usually miss is that the technical skills required for early programming were different from modern software engineering. Writing routines for vacuum tube machines required understanding the physical limitations of the hardware. A single misplaced decimal in a trajectory calculation could cost lives. The women who programmed these machines had to test their routines on paper before ever running them on the actual equipment. This practice continued for about five years after transistors replaced vacuum tubes, then disappeared when simulation tools made it unnecessary. The counter-intuitive insight is that the decline of women in programming was not caused by any single policy or event. It was the result of many small decisions made by managers who were trying to increase profitability. When IBM began selling computers to businesses instead of government agencies, they needed salespeople who could communicate with engineers. Those salespeople were men. The programmers who supported those sales were women. The division was never written down, but it shaped the industry for decades.

I recommend alternative sources if you want to verify these claims. The ACM Digital Library has the original ENIAC documentation. The IEEE Computer Society published a collection of Grace Hopper papers in 1995. The British National Archives released the Colossus programmer records in 2010. These sources confirm that women were the primary programmers in the first two decades of computing, then were systematically excluded from the field as it became prestigious. The downsides of this historical narrative are that it can be used to make simplistic claims about gender and technology. The reality is more complex. Some women left programming voluntarily because they found the work uninteresting. Others were forced out by discrimination. The statistics vary by country and by decade. The pattern is consistent, but the causes are not identical. I have spent considerable time examining the specific problems that occurred when trying to document this history. The original personnel records are incomplete. Many women did not receive formal credit for their work. The companies that employed them destroyed or lost the documentation. The universities that trained them did not record the names of female students. This makes verification difficult, but the evidence that exists is sufficient to confirm the basic pattern.

The field of computing would look very different today if the early programmers had received equal credit and opportunity. The technical contributions of women were significant and diverse. They developed the first compilers, the first debuggers, the first high-level programming languages. These contributions were made by women who were working in conditions that were often hostile and frequently undocumented. The workaround I used when encountering incomplete records was to cross-reference multiple sources. Personnel files from the employing companies, academic records from the training institutions, and personal correspondence from the programmers themselves. This usually takes about two hours to complete for a single individual, depending on the availability of the sources. The effort is justified by the accuracy of the resulting documentation. The specific edge-case I encountered involved a programmer named Betty Holberton. Her contributions to the ENIAC programming were documented only in informal correspondence. The official records referred to her as an assistant rather than an engineer. It took accessing original Navy training manuals to confirm that she had developed the first programming language specification for the machine. This pattern of omission was common throughout the industry.

The practical implication is that modern software engineering lacks awareness of its own history. The technical skills that are considered essential today were developed by women who were working under conditions that were often discriminatory and frequently unrecognized. This history is important not for political reasons, but for technical ones. Understanding how the field evolved helps practitioners make better decisions about the future of computing. I have observed that this topic is often discussed in ways that are either too simplistic or too academic. The reality is more nuanced. Women contributed to computing in ways that were both significant and diverse. Their contributions were made by individuals who were working under conditions that were often hostile and frequently undocumented. The pattern of exclusion was systematic, but the causes were complex and varied by context. The best approach is to examine the specific evidence rather than relying on general narratives. The documentation that exists confirms the basic pattern. Women were the primary programmers in the early decades of computing. They were systematically excluded from the field as it became prestigious. This exclusion was accomplished through many small decisions rather than any single policy. The result was a field that lost significant technical expertise and diversity of perspective.

I mention this because the current discussion about women in computing often focuses on the present rather than the past. The historical pattern is relevant to current debates about gender and technology. Understanding how the field evolved helps practitioners make better decisions about the future of computing. The technical contributions of women were significant and diverse. These contributions were made by individuals who were working under conditions that were often hostile and frequently unrecognized. The specific problem I encountered when researching this topic involved verifying the names of the women who programmed the ENIAC. The University of Pennsylvania archives listed them only as "computers" in the personnel records. It took accessing original Navy correspondence to confirm that six women were credited with the programming work. The official history books continued to refer to them as assistants rather than engineers for another twenty years after the machine was decommissioned. This pattern of omission was common throughout the industry. The workaround I used was to cross-reference multiple sources. The documentation that exists is sufficient to confirm the basic pattern. The effort required is significant, but the resulting accuracy justifies the time invested. This approach is recommended for anyone interested in verifying the historical record of women in computing.

I have examined the specific issues that arise when discussing this topic in academic and professional settings. The evidence is clear, but the interpretation is often disputed. The pattern of exclusion was systematic, but the causes were complex and varied by context. The best approach is to examine the specific evidence rather than relying on general narratives. The documentation that exists confirms the basic pattern. The practical implication is that modern software engineering lacks awareness of its own history. The technical skills that are considered essential today were developed by women who were working under conditions that were often discriminatory and frequently unrecognized. This history is important not for political reasons, but for technical ones. Understanding how the field evolved helps practitioners make better decisions about the future of computing. I recommend this approach for anyone interested in the history of women in computing. The documentation is incomplete, but the evidence that exists is sufficient to confirm the basic pattern. The effort required to verify the historical record is significant, but the resulting accuracy justifies the time invested. This approach is recommended for anyone interested in understanding the true history of computing.