Death By A Million Cuts

Individuals within the software development community are in one of two states. One believes that there is a deep rooted problem of sexism that constructs multiple barriers against females joining, and thus deprives this community of the majority of a talent pool that makes up half the population. The other, simply hasn’t realised it yet.


I used to think there was no problem. Most people in a position of power use deliberate, analytical thinking to evaluate people honestly and fairly regardless of any stereotypes that may exist. Job vacancies and course placements go to the most able worker or student, irrespective of gender. I just always assumed there were more men in the discipline of computer science because either men were, on average, better than women or, fewer women were interested in this line of work.


Thus I recall when I first heard a computer science lecturer in my research department talking about trying to “fix” the imbalance, I thought that was the thing that was sexist and wrong. Trying specifically to get more women into computer science? That’s positive discrimination which is as abhorrent to me as negative discrimination. I found the position insulting – I didn’t need specific help to allow me to become a computer scientist, why should females be given preferential treatment?


I consider myself lucky that while working as a lab demonstrator for introductory programming courses at university, I noticed certain patterns in my behaviour, and that of others, that made me rethink my beliefs.


In a learning environment where there was no structure imposed, it was interesting to see similar groups occurring year after year. A lot of students worked on their own, steadily progressing through that week’s programming assignment occasionally asking for help if truly stuck. Just as many though formed informal units – pairs or groups of as many as five – to help talk through the task and come to a common understanding of the problem. They would help each other with bits and pieces of syntax or perhaps new programming language structures. I recognised them because I used to do the same with my friends when I was an undergraduate.


Women though were less likely to have the support of such a group. This is owing to the fact that they need to either (a) create a group with other women or (b) join a mixed gender group and accept the resultant change of dynamic. The former is preferable but the lower takeup of women for computer science courses makes the choices more limited. The latter, though a valid alternative option, comes at a cost.


A group formally made up exclusively of male programmers changes when a female joins. The group with all males is relaxed. The conversation likely to jump from topics relating to the assignment to social activities or interests that are male dominated. The group with a woman added plays out a subtle competition for her attention. The importance of attracting this attention is amplified by the dual forces of the rarity of female programmers in general, and the likelihood that male programmers aren’t generally successful at socialising with women i.e. many programmers do not naturally socialise in activities with an even gender balance so it becomes important to take advantage of any situation that comes along.


I saw women with strong character letting the men know she was their simply to learn. I saw women who craved attention loving their authority within the dynamic. Some enjoyed it while remaining focused on the task of learning, some took advantage and allowed their team of coders to almost completely finish the assignments on their behalf. Most disappointing of all though is that I saw less confident women feel uncomfortable with all the unwanted attention.


When I’m asked what I look for most in a job, I’ll probably say hard, interesting problems with clever people using new technologies and best practice methods with a good salary. What I haven’t said is probably the most important detail and one that in some sense goes without saying. I want a comfortable, safe environment. No boss that shouts and swears, no Machiavellian colleagues or physically threatening CXOs. Often, the criteria that I implicitly expect as my right is unavailable to women. A female signing up for a Programming 101 course made to feel uncomfortable by a group of fellow students is one who will likely walk away from this discipline at the earliest opportunity. When there is such little gender diversity as it is, any action that pushes yet more females away is significantly damaging.


Like I said, many students worked through the assignment on their own and unaided and there were both men and women working this way. The strength of these informal groups is that they are a lifeline to students who are struggling, students for whom the logical statements and expressions don’t come naturally. A lifeline that exists more for men than women.


And then I also noticed behaviour in my own actions that was sexist. So often when dealing with students who had gotten stuck, I’d talk them through the compiler errors messages and programming element required. The difference was that while giving explanations to both, I’d take the keyboard and mouse from women. It was almost as if I was somehow impressing the women, whereas I only needed to instruct the men.


After a few years of helping out in the same lab and noticing this aspect of myself I started to look philosophically about how I behaved around men and women in the work environment. I wanted to make sure I didn’t commit a similar infraction again, however minor.


Sexism need not be conscious nor malicious


The main lesson I’ve learned from my observations and experiences is that, sexism is often not committed consciously nor with any intent of malice. From reading comments online in response to issues dealing with potential gender discrimination, I sense that some men do not welcome being told there’s a problem because they know in their hearts that they themselves aren’t sexist and neither are their male colleagues. The issues of prejudice that some may raise appear trivial to them and they react against a perceived feminist crusade.


Without needing to source any statistics, it’s clear that there is a gender disparity because the population split is 50/50 but the programmer split is, say, 10/90? 25/75? It will vary between company to company and college to college but it’s obvious there’s some reason women aren’t becoming programmers. Once that has been accepted we need to understand and appreciate that there’s no agenda – women who speak out against the status quo are reacting to the problems that they regularly endure.


The journey of a female from being technologically quizzical to an expert coder is a long process, and one that requires overcoming numerous obstacles. And by that I mean obstacles specifically relating to their gender – they have to overcome these in addition to all the other tests that males face. Facing the competition for attention that I noted from my time as a university tutor is just one such example. The problem starts much earlier.


Toys. Toy companies promote certain toys for boys, and others for girls. They obsess over gender divisions in cases where there is no need. Computer games are the domain of males despite the fact there is no reason for this to be the case. Girls enjoy playing games just as much as boys, and yet the message is sent out that this is a medium for boys. My first experience with programming was as a result of computer games – I was astounded that lines of letters and numbers could create the magic moving shapes and sounds coming out of the television. If an interest in computer games leads a certain proportion of gamers to look into programming, then companies that needlessly promote unisex products exclusively for the male gender, are pushing girls out. Females who choose to play regardless are exposed to blatant sexism online from male gamers which also pushes girls out. The existing disproportionality of gender in technical disciplines creates a self-fulfilling societal bias – a divide that young children don’t always have the confidence or empowerment to cross.


On their own these are small, trivial examples, and ones that you may agree or disagree exist individually. But becoming an expert is a long process so we start out as extreme generalists. We learn the basics in so many subjects at school, dismissing many as we get older and concentrating on the ones that fill us with the most satisfaction. Every stage that introduces a element of trivial, non-malicious, subconscious gender discrimination create another collection of girls saying, “No thanks, it’s not as rewarding as I’d hoped.” as they specialise in another more welcoming field of study.


Look around at the few female programmers you know who persisted through to the end. They really wanted it.


Benefits of a universal, barrierless entry


If you accept my argument that there is a kind of death by a million cuts, that filters exist discouraging more females out of the discipline than males, you may still not see the problem. Certainly, it’s unfair, but life has never been fair – what’s the big deal?


The fundamental issue is to do with the talent pool available to computer science and the software development industry. It only grows in size and quality when we have the best people from all walks of life wanting to join. A filter of any kind that discriminates against any aspect of a person other than their skills, is one that incorrectly discards quality candidates. We want the best and most talented people to become programmers. They will be more efficient, they will write usable tools and performant libraries, they will help and inspire us to become better coders ourselves. Only the lazy should fear competition.


So when you think about the problem of a shortfall of good quality programmers, imagine how reduced the problem would be if there were as many female programmers as male. There would be more programmers and the standard of the average one would be increased.


GitHub and knowing what not to say at work


Even though this is an old debate and an issue that I’ve barely seen improve since I started programming in the 90s, I was prompted to write this after the recent fallout at GitHub (a company I have a lot of time for), or more specifically the reaction from the community to the fallout. There are many sides to every story and anyone who attempts to summarise the situation is ultimately doing so from a position of ignorance. Basically, it’s not healthy to jump to conclusions. And yet that’s what many attempted to do.


For all the “He said, she said” that has come out of the tweets and official statements, one thing in particular caught my eye. Certain interactions can be open to interpretation but some are more binary i.e. they either happened or they didn’t. One of the issues in particular was that a developer declared his love to a female colleague. I personally find this behaviour wholly inappropriate and at odds with a healthy workplace environment. It surprised and disappointed me not to find more people registering their concerns at how unacceptable this is.


Of all the other issues and interpersonal communications (not to mention the hula-hooping), the leaders of GitHub will discuss the issue internally and propose a resolution. But more important than what the bosses of one company decide is what we as individuals in a community decide is acceptable behaviour or not. The tension surrounding acceptance and rejection that comes from bringing sex into the workplace or the classroom for one is at odds with creating a safe, inclusive environment. Date your colleagues by all means but save those conversations for purely social occasions e.g. if it’s a work outing to a bar, it’s still not a social occasion. How can you feel safe in a work environment when someone declares romantic intentions? And this isn’t just a problem for a company but a college lab, a tech meetup, a hackathon. We need to recognise behaviour that is another barrier for women and, as a community, reject it.


How we can make things better


We all need to play a part in making things better if we want that stronger, larger and ultimately, fairer development community. We need to take responsibility for doing it ourselves. I challenge everyone to undertake the following.


  1. Recognise that there is a problem, one that exists to the detriment of our industry.
  2. Talk to others about this issue. Many don’t recognise the problem because they haven’t considered all the viewpoints. Only by prompting discussion can we look to change things.
  3. Meta-thinking. It took me a long time to realise that I had a sexist prejudice in my behaviour. Think about the way you behave in your work or class situations. Imagine swapping the gender of the people with whom you’ve interacted. Would you have acted differently? If so, think clearly if your behaviour was appropriate.


My main hope is that I’ve reached a few people who, like me, didn’t previously think we had a real problem. Women were some of the original pioneers in computer science but now I struggle to think of someone as prolific as Grace Hopper in recent times. It shouldn’t be this way. The imbalance won’t be fixed by monitoring and attempting to manipulate the gender ratio or waiting for the results of one companies’ internal investigation – it’s not about numbers but rather opportunities. When a 5 year old girl wants to learn how to write a computer program, we need to see the doors of opportunity as plentiful for her as any boy her age, and crucially, for them to open just as easily.