6. Making Programming Masculine
In April 1967, an article appeared in Cosmopolitan Magazine entitled "The Computer Girls." While the language employed seems condescending and sexist to a modern reader, comparing computer programming to planning a dinner and speculating on the chances to meet young men in the computer field, it does highlight the confusing and often contradictory messages sent to young women about their proper role in the emerging computer fields at a time when the programming community in particular was actively trying to make itself more masculine. The Cosmo article is right on several counts.
First, there were exceptionally large numbers of women working in computer programming in the late 1960s, in fact the article actually underestimates the number. Within the incredibly broad category of computing there are many different occupations, high-status jobs such as programming and systems analysis as well as low-status ones such as keypunch operator, where women ended to congregate. Even within computer programming there were different roles available to men and women, but the author, Lois Mandel, was right that programming was unusually receptive to women when compared to other traditional professions. Women were not only filling entry-level positions, but some, like Grace Hopper, were able to reach the highest pinnacle of the profession.
The article also did a reasonable job of explaining the characteristics of the programming labor market, which was doubling in size every year or two and as a result could not afford to discriminate against women. There was a serious shortage of programming personnel by the 1960s, threatening to wreak havoc on the entire computing industry, and employers went to extraordinary measures in the face of this crisis. Some, to avoid hiring women except as a last resort, even went so far as to recruit from prisons. Although programming was considered a highly skilled profession in the late 1960s, the exact nature of the intellectual labor involved was not yet clearly defined. Many firms preferred to train programmers from within, and would give aptitude tests to all employees regardless of gender. This made it many women's best bet to get any job in a comuter department with the intention of getting on the programming-trainee list, preferable by far to paying for one's own programming training.
The article lastly notes that it is a lack of knowledge of what programming is and that women were discouraged from pursuing math and sciences that kept women from entering the field in higher numbers. This gender imbalance persists today, with womens' participation in the computer programming field beginning to decline in the 1980s through to today, leaving the field egregiously male-dominated. It was not until the 1990s that historian began to rediscover and acknowledge the contributions of women in the early history of computing, revealing that the lack of female participation in computing is not due to lack of interest or inclination, rather their participation has simply been systematically ignored and underreported. This chapter argues, in light of this history, that although the modern associations of coputer work, programming in particular, are strongly masculine, this work began as women's work and had to be made masculine. This chapter provides insight into the ways in which structures of a profession reflect and replicate the culture of its practicioners.
The most prominent case study of women in computing is coincidentally the earliest, that of the six women- Kathleen McNulty, Frances Bilas, Betty Jean Jennings, Betty Holberton, Ruth Lichterman, and Marlyn Wescoff- responsible for much of the development and progamming of the ENIAC machine (Electronic Numerical Integrator And Computer). ENIAC was one of the first and most famous early electronic computers, and the women were recuited to "set up" this general-purpose machine and to program it to solve real-world problems. Though it was initially assumed that setting up the ENIAC would be relatively trivial work, the women programming it developed such a deep understanding of the macine that they could "diagnose troubles almost down to the individual vacuum tube" according to Betty Jean Jennings, adding that they could do this better than the engineers at times. They also affected the design in significant ways, including Betty Holberton convincing John von Neumann to include a "stop instruction." Yet the work of these women was disregarded and forgotten by many, largely because they were women. It should also be noted that they were "software" workers (an anachronistic use of the term, which would not be invented until 1958) in a hardware-driven project, and these terms reveal a gendered distinction in computing work- the "hard" technical mastery was for men while the "soft" aspects of computing work were relegated to women.
The assumptions were that the women would simply adapt computations already in place for human computers (a feminized profession in the 1940s) for the machine, not anticipating that this would be difficult and require radical and innovative thinking. The telephone switchboard appearance of the ENIAC only served to reinforce the idea that this was a craft rather than a science, and made programming (or coding) a secondary task to hardware design. By the 1950s it was becoming clear that programming a computer required intelligence and quite a bit of ingenuity, and programming was beginning to acquire a reputation as being incomprehensible to all but the small pool of talented insiders. Despite the continuing perception of programming as an artform, the comparisons drawn were to traditionally masculine artisan or tradesman roles.
"The heady combination of mathematics, engineering "tinkering," and arcane technique attracted a certain kind of male to computer programming."
This perception was reinforced by the mid-1960s, when 80% of companies were using a series of aptitude tests that focused on innate abilities, many of which explicitly referenced chess playing, musical ability, and mathematics. While the reliance on innate ability would at first seem gender neutral and even friendly towards women, the tests embodied and privileged masculine traits like formal training in mathematics (already at that time declining in importance within the work of programming). The closelt related personality profiles were worse though, emphasizing a lack of interest in interpersonal interaction and other stereotypically masculine characteristics. By emphasizing these masculine traits in recruitment and formalizing them in hiring practices, it normalized the notion that those are and ought to be the inherent qualities of a good programmer. It would seem this happened through a combination of laziness, ambiguity, and traditional male privilege. It was already recognized in the 1960s that aptitude testing was inaccurate, unscientific, and a poor predictor of future performance, but they were still used because they were cheaper and easier to administer than having one-on-one interviews, for example.
The 1960s "software crisis" also saw attempts at the routinization, simplification, and overall denigration of programming occupations by managers. Whether these efforts were successful is questionable, since computer programmers are on the whole highly paid, highly valued, and largely autonomous workers today. The same crisis produced an opportunity for women to fill in the workforce, as was often the case in male-dominated professions facing labor shortages. Yet the recruiting efforts for women were somewhat backhanded, relying on an image of women as docile and content without career advancement. If women were married or engaged, they were seen as a benefit because they were unlikely to move geographic locations, yet the same literature from which this description comes also says that young, unmarried women because they were likely to be distracted by their social commitments. Young women were used in advertisements at the time as a shorthand for depicting low-skill, low-wage labor, many claiming programming was so easy that a girl with no expereince whatsoever could do it. It would seem from these mesages that companies were admitting to the rising cost of programming and implying the solution lay in feminizing the work and thus making programmers replaceable.
What is perhaps the most striking in the face of this sexist, mixed message to young women was the backlash that it provoked from men in programming who saw this as demeaning and threatening to the future of the field. The push for professionalization in the early 1970s was a means of addressing the "software crisis." This in itself would be a laudable goal, but the reforms instituted to make programming more professional disproportionately benefited men, such as requiring a college degree or imposing certification or licensing requirements. More subtly, professionalization requires segmentation and stratification, elevating the work of some by distancing it from other "lower" work. This once again leaves women clustered at the bottom ranks of the field with little possibility to move upwards. The culture of computing became masculine through the appeals to chess and mathematics in recruitment, later formalized in aptitude tests. This in turn replicates a structure in which masculine ideals are seen as professional and feminine traits seen as undesirable.