All Coding is Feminine in Love and War
To figure out how to get out of a tricky spot, we should look at the steps we took to get us in this mess, take note, and march forward.
For this reason, I wanted to take some time during Women’s History Month to acknowledge the not well known “feminine” beginnings of programming, and how it got from being seen as clerical work for “feeble-minded women” to the male-dominated industry it is now.
Coding has its start as being simple and therefore feminine
In 1945, the Electronic Numerical Integrator Computer (“ENIAC”) was completed. The first general-purpose electronic computer was used to solve wartime ballistic problems “faster than a bullet” (Fritz, 20). The coding for these machines was done by the women and considered to be merely mechanical while the engineering done by men was regarded as the real science. Engineering was considered the real science, while coding was considered mere machine operating. It was men who engineered it, but it was women who coded it (Ensmenger, “Making Programming Masculine” 9). These women were referred to as the ENIAC girls (Ensmenger, “Making Programming Masculine” 8).
Later on, the programming process was outlined by John von Neumann, one of the leading scientists of ENIAC. He wrote that programming consisted of six steps. The first five were related to setting up and writing the algorithm and the last step was to code it into the machine. The first five were done by a planner that was usually male, and the last step was done by a “coder” who was usually female. The last step was considered nothing more than the simple setup of the machine (Ensmenger, “Making Programming Masculine” 9–10).
Hold on! Coding is Actually Kind of Difficult
As any programmer knows, the division of planning and coding is a blurred line, often one that goes back and forth. The coders would often have to revisit and revise the algorithm; this made it clear that the coder’s duties could not be neatly separated from the planners as von Neumann initially thought. Not only did the ENIAC girls have to understand and revise the algorithms, but they also needed to develop clever optimization methods to get every bit of speed out and to fit their programs into the limited memory space of the primitive computer. In-depth knowledge of the hardware was needed as well so they could differentiate hardware issues from software issues (Ensmenger, “‘Computer Boys’ Take Over” 158). Despite the amount of skill and knowledge required to code the machine, the profession of “coder” was recognized as being simply clerical until the 1950s (Ensmenger, “Beards and Sandals” 47).
So what changed in the 1950s? Well, businesses began to use computers as hardware costs went down, and demand for programmers drastically increased. Men started to join the field and soon outnumbered the female programmers. However, the programming profession was still considered clerical and feminine. This perception of programming work began to change as the software began to be used for more complicated issues. Sure, using ENIAC to calculate the trajectory of a missile is hard, but implementing a payroll system is even harder (read with sarcasm).
Time and money were poured into developing business applications, but they weren’t always successful (Ensmenger, “Beards and Sandals” 46). Companies had to start thinking about what made a good programmer (Ensmenger, “Beards and Sandals” 47). With programming being an undefined discipline with no neat category to fall under, such as science or mathematics, it began to be seen as a form of arcane tinkering. People began to believe that the people who were good at programming were simply born with the right stuff. So, how do you find the people with the right stuff? Well, through methods that solidified that programming is masculine (Ensmenger, “Making Programming Masculine” 13).
Tests are objective so it can’t possibly be discriminatory…right?
If you’re hiring a programmer, but have no idea what makes a good one, what can you turn to? Well, one thing was aptitude tests. One might think that they would have been perfectly gender-neutral, but they were not. Despite the consensus that advanced mathematical skills are not good predictors of someone’s programming abilities, mathematical problems that required university-level mathematics were included on the aptitude tests anyway. Therefore, these tests prevented women from entering the field since women at the time often did not attend university. Thus, beginning the process of making programming masculine (Ensmenger, “Making Programming Masculine” 14–15).
In 1967 psychologists Dallis Perry and William Cannon published a series of papers in well-recognized academic journals such as the journal of psychology, providing an in-depth profile of “vocational interests of programmers.” They claimed that programmers had a striking characteristic of being disinterested in people. The basis of their publication was the strong vocational interest blank or SVIB for short. It was a test that finds the interests of an individual and compares them to the interests of professionals in different careers (Ensmenger, “Making Programming Masculine” 15–16). By this time, aptitude tests have already filled the programming profession with males and thus these personality tests drew conclusions based on a biased data set.
“programming was officially assigned a gender when a 1970 psychometric literature reported that programmers received high masculinity and low femininity scores and concluded that programmers consistently have a masculine self-identity”
The information from the SVIB and the information from these papers helped form stereotypes that live on in the modern-day. These stereotypes were further perpetuated by an influential industry analyst Richard Brandon, who claimed that programmers are “excessively independent” to the point of mild paranoia and that the programmer type is “often egocentric, slightly neurotic, and he borders upon limited schizophrenia. The incidence of beards, sandals, and other symptoms of rugged individualism or nonconformity are notably greater among this demographic group” (Ensmenger, “Making Programming Masculine” 15–16). Later on, programming was officially assigned a gender when a 1970 psychometric literature reported that programmers received high masculinity and low femininity scores and concluded that programmers consistently have a masculine self-identity (Ensmenger, “Making Programming Masculine” 17). It was these personality tests and the psychometric literature which constructed the stereotypical programmer image as an antisocial male.
The perception of programmers changed throughout history. The stereotypical image of a programmer transformed from one of a woman to one of a man. The primitive computer’s appearance led people to believe programming was clerical work and therefore female work. This changed when demand for programmers went up and the development of business applications made people realize it was anything but clerical work. This led to personality tests and aptitude tests to be developed that preferred masculine traits and attributed women computer pioneers’ successes to their “masculine” traits.
Onwards and Upwards!
So now that we know what steps were taken to get into this predicament what can we do to get out of it? Ask yourself what possible unconscious gender-discriminating positive feedback loops you’re unknowingly involved in. Ask yourself what you think makes a good programmer and what possible biases you might have. Ask yourself how your hiring practices might be discriminating against women. Do they involve investing a lot of time that makes it inaccessible for women who are more likely to do most of the household chores and the women who are more likely to spend more time than their male counterparts on childcare?
Take note, and let us march forward towards a world where all hands can finally truly be on deck in this fantastic world of software.
An Aside to Address the Biology Argument
An argument I often hear is that men simply have a biological advantage when it comes to the reasoning skills needed for programming. However, the study that people often cite, which suggested that men have better spatial reasoning has been recently disproven by a meta-analysis that suggests that it is actually nurture not nature that produced those results. Societies that think boys have better spatial reasoning produce boys with better spatial reasoning.
This perception that males are biologically more inclined to be better programmers causes destructive self-fulfilling prophecies. As a former teaching assistant for Cognitive Science and Introduction to Computer Science, I’ve seen some female students believe that they “just aren’t made for programming”, an oppressive belief enforced by societal norms. Obstacles are often seen as proof that they are just biologically not capable, instead of the truth — that there is a missing knowledge gap and that they need practice.
References:
Making Programming Masculine by Nathan Ensmenger
“Beards, Sandals, and Other Signs of Rugged Individualism” by Nathan Ensmenger
