Sunday 6 October 2013

The Computer Scientist and The Cleaner

Let me tell you a story
The computer scientist and the cleaner had a long and happy marriage. One of their few arguments was when she forgot their wedding anniversary.  But their marriage was strong and he forgave her.
It's not a great story, but it's a story.

Now let me ask you a question.

Who forgot the anniversary?  Was it the computer scientist or the cleaner?

The answer doesn't matter.  What matters is: when you read it did your brain make you think that the computer scientist was male and the cleaner was female? 

For an unbiased view, let's ask Google Images about computer scientists and cleaners. The answer is clear: computer scientists are men, and cleaners are women. 

We've got a problem: we, meaning everyone in computer science.  The problem is twofold. There's a major gender imbalance in computer science. And that women are often not treated right. It's a very widely known problem, and in fact I think it's the most important problem for computer science.  (Not the most important problem in computer science, but for it.)

I've been thinking about gender issues in computing because I am first year co-ordinator in the School of Computer Science at the University of St Andrews. It suddenly occurred to me that I have a responsibility to new students to tell them about the most important problem for computer science. 

So I decided I should prepare a talk to give to first year students, both men and women, to tell them about this major problem, and some small pointers on what to do about it.

A draft talk is below, but here are a few of the key points.
I know of only two good reasons we need more women in computer science
  • It’s right
  • Computer Science would be better
  • this point being that since some women don't get into CS or leave early, the average quality of CS would go up with more women in it. 
What should people in CS do about it?
1. Don't be a jerk to women 
2. Don't use sexist language 
3. Understand it's not you who decides if you are doing 1 or 2
We need males in CS to be Allies
Allies?  Basically an ally is a non-female who thinks that both men and women should be treated right in CS. See the geekfeminism wiki for some pointers. 
Don't get me wrong. 
I want all our male students to have fun and do brilliantly in their CS degrees and afterwards.  It's just that I want the same for female students and other disadvantaged groups.  And I want us to work together to reduce those disadvantages.
So here is my talk, embedded below or direct at SlideShare.  



If you want to comment, please do comment here, or by email to me, whatever you want.  I'm obviously especially interested in any ways I can improve the talk (apart from making it shorter, which I know it needs to be).  Given this topic can be sensitive, I will remove what I regard as offensive comments without hesitation.


Credits: This is my own work and I take full responsibility but helpful comments have been received from: Kate Cross (special credit for suggesting anniversary instead of a car crash in the story), Perdita Stevens, Vicky Larmour, Karen Petrie, Ursula Martin, Louise Dennis, Julian Bradfield, Dave Berry, Judith Underwood, Chris Jefferson, Jeremy Frank, Juliana Bowles.   (There are actually many others who have commented usefully, and I will add you here happily if I've omitted you.)

[Update: 6 Oct 2013: following comments from two friends, I have made some changes to this post and the draft talk.  This was removing some of the more personal material and changing the story from a car crash to an anniversary.

Update 2: 7 Oct 2013.  Changed "in computer science" to "for computer science"

Update 3: 7 Oct 2013.  To my horror realised that I was guilty of cissexism since the FLORIN slide implied gender was binary.  Have removed this.   There are also many other changes I want to make in this based on other comments but this was the most urgent.

Update 4: 9 Oct 2013.  Quite a number of changes.

Update 5: 12 Oct 2013. Added many slides about the Petrie Multiplier.

Update 6: 20 Oct 2013. Slides as delivered to students at St Andrews, with one or two minor deletions of local information.

As I am changing this talk and it's in the open, I thought I would make available earlier versions.  Here are version 1version 2, version 3version 4, version 5, and version 6 (the version delivered to students in Oct 2013.]

30 comments:

  1. This is beautiful, thank you.

    ReplyDelete
  2. Many thanks Anonymous, that is very kind and much appreciated.

    ReplyDelete
  3. I think an example or two could definitely help. Otherwise, you're likely to have a load of people nodding their heads and then say something in the labs they don't realise are offensive — which most of the time won't be raised as an issue by women in the labs (they are the minority, after all).

    ReplyDelete
    Replies
    1. I agree. I strongly estimate that just describing at an abstract level what the problems are will go right over the heads of people who don't realise that they are part of the problem. Nothing like an example to illuminate. Same goes for teaching programming...!

      Great that you are tackling this, Ian.

      Delete
  4. Ben, many thanks. I completely see the point... the danger I see is that one ends up using "good" examples, i.e. terrible ones. Which could be counterproductive.

    ReplyDelete
  5. There's an old joke/story/conundrum which has a very similar basis. Here it is:
    A man and his son are involved in a car accident. The main is unharmed, but the son is injured and rushed to hospital. In the operating theatre, the surgeon says "I can't operate on him--he is my son". Who is the surgeon?
    In my household, my two sons don't get this conundrum at all. Actually, they fail to understand it on two levels. They don't understand *why* the conundrum even exists as such and they also don't understand why other people would not immediately get the answer [which they both got straight away].
    The 'answer' to the 'conundrum', of course is that the surgeon is the boy's mother. And why do my sons completely not understand the conundrum? Because their mother is a neurosurgeon, and it seems entirely natural to them that women can be physicians and surgeons. In fact, until they were about ten years old, they probably thought most surgeons were female. All their lives they have met surgeons, engineers and barristers drawn from the same pool of extremely smart, self-confident, gifted women for whom it has been entirely natural to excel. The problem certainly wasn't with the women...

    ReplyDelete
  6. There's an old joke/story/conundrum which has a very similar basis. Here it is:
    A man and his son are involved in a car accident. The main is unharmed, but the son is injured and rushed to hospital. In the operating theatre, the surgeon says "I can't operate on him--he is my son". Who is the surgeon?
    In my household, my two sons don't get this conundrum at all. Actually, they fail to understand it on two levels. They don't understand *why* the conundrum even exists as such and they also don't understand why other people would not immediately get the answer [which they both got straight away].
    The 'answer' to the 'conundrum', of course is that the surgeon is the boy's mother. And why do my sons completely not understand the conundrum? Because their mother is a neurosurgeon, and it seems entirely natural to them that women can be physicians and surgeons. In fact, until they were about ten years old, they probably thought most surgeons were female. All their lives they have met surgeons, engineers and barristers drawn from the same pool of extremely smart, self-confident, gifted women for whom it has been entirely natural to excel. The problem certainly wasn't with the women...

    ReplyDelete
    Replies
    1. That's great! Yes the surgeon story was in my mind when I put that together.

      My 10 year old daughter said that she didn't think it was either way round - maybe because both her parents are computer scientists.

      Delete
  7. Very nicely put, and great to see it being said.

    One minor point - when your slides talk about how 'we' behave towards women, that also implies that 'we' are all men, no? Not sure how to rephrase, but it's worth thinking about

    ReplyDelete
    Replies
    1. Thanks Francine. Yes that is something that needs to be revised in a later version.

      Delete
  8. Great draft, hope that it will be presented in St Andrews. Not only to CS1002 students.

    ReplyDelete
  9. Much of this is cool. There is one key thing that I am not clear about when you say "you don't get to decide" I think this is fine and I agree. However, it is important that you understand that the person who takes or doesn't take offence at what you say does so so as a matter of choice. What I am trying to say is that just because the other person takes offence it doesn't mean that whatever was uttered was bad (or good). There are people who take offence at for example the word 'Chairman'. But I don't. Thing is you don't get to decide whether I am sexist or not. You can of course call me sexist but that doesn't make me one.

    ReplyDelete
    Replies
    1. If somebody tells you they're offended, you've always got the option I laid out, I.e. say that you don't agree and you will carry on using that language/whatever.

      So I would do that if I was using say "manual" for "manual labour" and I was challenged on that - since it's not etymologically related to male.

      But if I was using the word Chairman and I was challenged on it, absolutely I would change.

      The more important point is not about actual offence, but the more subtle things. So if e.g. 3 men and 3 women meet to elect a Chairman, will there be an uncosnscious bias towards electing a man? Maybe.

      One of the main points of this piece is that non inclusive language can affect people subconsciously. For example, obviously both "computer scientist" and "cleaner" are gender-neutral words. Even so they can come with gender preconceptions (for some, not all).

      Delete
  10. If someone is inclined to vote for a man then i am not convinced that they would change that position because the position they are voting for is called 'chair'.

    I don't associate "chairman" with gender. I have lived in America and I don't want to see here a mirror of the Political Correctness that so dominates life there. This political correctness results (IMHO) a very polar society.

    But, yes, absolutely, we should look to making our environments friendly and comfortable for all our students to achieve the best that they can (so for example sexual banter is a no no).

    ReplyDelete
  11. Your University most likely has a code of practice on bullying, harassment and the like, and it probably says something along the lines that if the person thinks it is harassment it should be treated as such Ursula (not anon but can't be bothered to create and account etc)

    ReplyDelete
    Replies
    1. Oh thanks very much! I wish I'd thought to look for that. Yes there is

      http://www.st-andrews.ac.uk/staff/policy/hr/harassmentandbullyingatworkandstudy/

      And very sorry for demoting you from CBE to MBE (will be fixed.)

      Delete
  12. When I first visited the Burn in my 3rd year, not a single woman joined us, even though there were some in our year. During a general discussion I lamented this fact, and was met with lots of eye rolling and even some outright hostility to the idea that this could be changed for the better. This was disheartening, but I let it drop. The next year, if memory serves, two women joined us, one of whom was in a relationship with the man who had been most outspoken against me the year before. She assured me that he's not really like that in private, a statement I met with skepticism. I brought up the topic again, and this time got a less hostile response, but still very little discussion.

    So, what I'd recommend is encouraging both faculty and students to have these discussions, to seek solutions to what is a varied and complex problem, and not to shoot down even the possibility that we can make a difference.

    On the up side, I will say that this is one of the very few unfortunate experiences I had at St Andrews, nor do I think any of my male colleagues wanted to exclude women, but as you point out, we need more allies who actively work for equality for all students.

    ReplyDelete
  13. Hi Ian,
    Thanks for posting this. I think you have a great start at introducing and addressing the very real, very knotty, problem of gender bias in computer science. I think it is superb that you are making this lecture, and attacking the problem, head on.

    I have a number of comments and suggestions. Please take all of them with a grain of salt, and as constructive criticisms.

    Parts of the presentation I thought worked very well:
    Slides 10-15: The introductory story is perfect. It reveals inherent bias very well. It did for me.
    Slides 16-21: The history lesson is a perfect follow-on. It should shock your listeners. It shocked me. They should be thinking now.
    Slide 24 is great.
    Slide 25 is great.
    Slide 31 is great.

    Parts of the presentation I think could be improved:
    Slide 9: DON'T APOLOGIZE for the fact that this topic isn't an algorithm, or math, or a proof, or a snippet of clever code. Either you think this is part of something your students need to learn, and you are going to teach it to them, or you don't. You weaken the teaching if you apologize.
    Slide 22: THE MOST IMPORTANT PROBLEM FOR COMPUTING? Really? OK, really. But I think some follow-up or elaboration is needed. Gender bias and gender imbalance are important problems for society. Computing is a society of its own, and part of a broader society, and therefore it is important. Women and men use computers, build computers, and (yes) program computers, and therefore gender imbalance (probably) lessens computers and computing as a discipline. My little beef with this is that gender bias and imbalance is not just a problem in computing; your use of this as a 'contentious proposition' is ok but could distract from your actual point if you are not cautious.
    Slide 23: Your 'overpoweringly' good reasons: it's 'right' and 'computer science would be better'. Slide 24 is a good explanation of the pernicious impact of gender bias. Reducing the gender imbalance would help redress and eliminate the gender bias, and 'right' the 'wrong' shown on slide 24. However, if that is what you mean by 'it's right', this isn't explained very well. Making this case is useful. With respect to 'computer science would be better', I think your case is unmade, and also weaker. As I mentioned before, computing and computer science is a society, and for this reason fixing the society of computing is worthwhile. But will this lead to better algorithms? Better proofs? Better code? Better computing hardware? Should it matter if that is the outcome? These are questions I have after reading this slide, and I think those questions might distract from your objective if not addressed, at least to first order.
    Slide 27: I think you should neither assume this is easy to understand (because otherwise why do so many people do it?) nor fail to provide some examples (one obvious one, and one subtle one). Just as you do with your (excellent) banter slide (number 31) you can probably find some excellent references on sexual harrassment to refer to.
    Slide 34: I think this slide is stronger if you eliminate the parenthetical (straight white...). Your main topic is gender bias. That you are now talking to 'the boys' is enough. You made your point about other forms of harassment on slide 33 (also excellent). The parenthetical, again, introduces the possibility of distracting attention because you forgot to identify all the other colors, didn't cover transgender, etc...

    Thank you for posting your talk and providing the opportunity to give feedback. I hope the talks go well!

    ReplyDelete
    Replies
    1. Wow! Thanks Jeremy. I already have a long list of people who have helped me with this and now I can add a rocket scientist! That's great

      I'm currently revising and trying to clarify "it's right" and "better for CS".

      I'm really reluctant to give examples of being a jerk. But maybe a subtle one would be good. Or as you say I can find something online (it's not hard, sadly)

      It's an excellent point about it being inherent in society. I'll reflect on that.

      I do actually believe this is the most important problem for computer science. I could of course be wrong. As I said I'm trying not to argue whether it's more or less important than P=NP or writing effective programs with tolerable numbers of bugs - since those are problems within computer science, while this is a meta-level problem.

      Good luck getting back to work sometime!

      Delete
    2. P.s. yes I think the history of St Andrews is pretty shocking (and it's not massively behind other places either).

      Delete
    3. w.r.t. gender imbalance being a societal issue: Since you are presenting a societal issue in the context of a technical subject, I think it is important to try to explain why this issue impacts this technical subject. Opening the topic up to your students might be risky, but could reveal some very interesting answers!

      w.r.t. examples of being a jerk: I think, like your opening thought experiment, it is important to provoke those who will say 'oh, I know jerks, I'm not like that' with some examples to make them second guess themselves and increase their awareness of what being a jerk could mean. If you relate it to the point you make later (you don't get to decide, they do) it'll help.

      Delete
  14. HI Ian,
    I am here to your blog through the link that Jeremy Frank posted on his FB page. I think that your talk has a lot of strengths, and I applaud what you're working on! thank you for undertaking ally work; it's not easy!
    You might find this post at the blog Shakesville helpful regarding the use of "offended" or "taking offence" (I see that we already have some discussion of that in the comments to your post!): http://www.shakesville.com/2013/10/liss-says-stuff-4-im-not-offended-im.html (the post has video, with a transcript below).
    One of Melissa McEwan's central points is that when conversation about social justice gets sidetracked into whether or not someone is "allowed" to be offended is a classic silencing technique to get out of engaging in the work of social justice. Her technique (YMMV, of course) is to center the beginning of a discussion or observation about how the world works in racist/sexist/classist etc. ways is to communicate her contempt for the statement (not the speaker!), rather than engage in a discussion of whether something is or is not offensive. You might find the whole thing useful. Best wishes to you for this talk.

    ReplyDelete
  15. Bravo, Ian! As an engineer of MANY years, I have witnessed CS and in a broader sense STEM in general going from a completely male dominated area to one that is improving. When I was a young undergrad, (1970 - 1975) there were NO female students at all. There were a few brilliant exceptions, but it was certainly not the rule. I even got in trouble at my job for advocating the baby step of promoting qualified female "rework-girls" to full technician status -- I know they were qualified because I trained them myself. This was in 1972 -- not so long ago.

    It is sad that we are still battling gender -- shall we say "discouragement" -- in the present day.

    As an interesting side-note: our University's STEM and WISE programs are located just down the hall from my office. The are doing amazing work in gender-neutral outreach. I applaud them and you for all your hard work addressing this persistent issue.

    ReplyDelete
  16. Hi Ian,
    I like many of the changes you've made. This is going in a great direction. I have three minor comments on this version:
    - Slide 25: I suggest removing the first 2 bullets; they lessen the impact of the slide.
    - Slides 31-37: these are great. Keep them! But, perhaps now you want another 'title' slide, making it clear up-front that the issue is about gender imbalance and fair dealing with everyone in computing?
    - Slide 49: I suggest replacing the first line of this slide with: "No Matter Who You Are:"

    ReplyDelete
  17. Too good. Too informative. Liked it a lot. :)

    ReplyDelete
  18. Actually what popped out at me was that SHE forgot the anniversary. I couldn't answer your question about who was what because I had focused on the SHE part. Go figure...

    ReplyDelete
    Replies
    1. Hi Jane. My wife and I are both computer scientists and we usually BOTH forget our anniversary. When my father-in-law rings up to congratulate us we usually look at each other and shrug! Though this year my daughter got into our anniversary so we remembered ok.

      Delete
  19. Hi, Ian. Thanks very much for this. Here is an article I wrote on this topic for ACM's Interactions: http://interactions.acm.org/blog/view/the-present-past-and-future-of-women-in-stem

    ReplyDelete
    Replies
    1. Thanks Ashley, that's excellent. I especially agree with your points about thinking about the small things all the time.

      Delete
  20. This comment has been removed by a blog administrator.

    ReplyDelete