A few days ago, an article titled Why Women Shouldn’t Code, perhaps the ultimate click-bait headline, was posted on Medium. There were already great responses by Kelly Ellis and Jenn Schiffer by the time I read it, but I quickly started drafting up my own comments.
I wrote my response assuming that the author of Why Women Shouldn’t Code was just another “old white dude” telling women what to do. Embarrassingly, I discovered that in fact the author was a woman! Here I was, trying to write some words supporting gender equality, and I made the classic mistake of assuming everyone on the internet is a man. Oops.
It’s often said that no-one thinks they are part of the problem of sexism, and this was a good reminder that it really takes a lot of work to over come biases that are common in society. I’ll try to do better next time.
After all this I was really wondering if I should even publish anything about the original article. The internet is already full of men telling women what to do or think. Finally I decided that it is worth speaking up about, if nothing else because I want to make it clear, on my own blog, what my own opinions are. But also, Why Women Shouldn’t Code has some seriously wrong assumptions about the nature of coding, something I hold very near and dear. So let’s talk.
On the Innate Differences Between Men and Women
Sure, you don’t need to do a brain scan to discover there are biological differences between men and women.
But taking the leap from “the part of the brain that we think is somewhat responsible for extremely high level activity X is slightly bigger in men, on average” to “thing X should only be done by men” is simply not a viable line of thought.
Proving differences in aptitude of any high level skill is almost impossible to support with science. There is way too much complexity in the human experience to ever come to a meaningful conclusion.
Besides, we can’t even measure the differences in productivity between teams of software developers even if we try to make the composition of the team as identical as possible. Given the same requirements, different groups of developers won’t even build the same thing. One team wouldn’t even build the same thing given a second chance.
Part of what makes creating software hard is that we don’t conclusively know what makes a team good or bad at it, even though we’ve been doing it for at least 70 years1. Under those conditions, is it really wise to exclude 50% of the population from contributing?
Most importantly, even if we did some day conclusively prove that on average the biology of a woman makes her less adept at writing code, who cares? Men do things that women are supposedly better at all the time, and no one bats an eye. Let them code.
On the Definition of the Profession of “coding”
If a software job simply consists of writing a program to do exactly what someone else tells you it should do, then there’s no question why women don’t want that job: that job sucks. No one wants that job.
Fortunately, while probably more common than we’d like, that job isn’t something that’s really needed. What we do need are people who can not only write good code, but talk to other people who are experts in other things, and together build something that is both useful and possible to build in a reasonable amount of time, for a reasonable cost. This job is hard, it requires technical AND people skills, and many people find it extremely enjoyable.
By the way, you don’t even have to code as part of a job. Coding can simply be done for fun, on your own, or with some friends, or even random strangers from the internet.
Even after robots have taken over the world and humans just sit back on their space cruise ships, someone will surely still be writing code for fun (probably even code that a million other people have already written). No one needs to tell you what to code, and you don’t even need to ask permission. That’s why it’s awesome. It’s the pure joy of creation2.
On 16 “Data Points” Being Enough to Prove Something
When you go to a toy store, the boys section is filled with superheroes, trucks, trains and airplanes. The girls section is filled with dolls and fake kitchens. Do we make fake kitchen toys for girls because that’s what they like, or is that what they like because that’s what we make them?
These sort of questions of causation have confounded humanity for so long that the entire idea of science was created to help answer them. While still lacking, there are a number of tools that we have today to help. Sample sizes should be large, biases in selecting the samples should be taken into account, and the study itself should be administered as a double-blind trial, all to make sure we aren’t tricking ourselves.
What definitely ISN’T in that toolkit is cherry-picking a handful of stories from people who’s lives we personally have had a long and significant influence over, and using those stories to support a belief we’ve had since before those people existed.
- For a great collection of mostly sound scientific research into what we DO know about reading software, watch Greg Wilson’s fantastic talk What We Actually Know About About Software Development, and Why We Believe It’s True, and then read the resulting book: Making Software: What Really Works, and Why We Believe It. ⏎
- You could even say the joy of nurturing something. ⏎