Our industry sucks for diversity and conferences are only showing up the fact
fberriman.com
fberriman.com
> I think at this point we should start asking questions as to why we feel the need to poke hard at conferences, why are we positively discriminating, and what should we really be fixing for lasting change. Why are our starting numbers so low?
Discriminating. Positive or negative. Is wrong. Just because you are XYZ doesn't mean you somehow could expect to be treated differently. No. You are equal. You are just like anybody else. You are not special. On the other hand people in minorities whine(in many cases rightly so) that they are treated differently, while people in non-minorities call for positive discrimination which further promotes inequal treatment.
You're a woman/gay/trans/whatever in tech? Cool, now get back into line just like everybody else. Nobody cares -- hopefully.
The reason for reinforcing positive discrimination is that you are not aware of your own bias towards negative discrimination.
And honestly we aren't. We cannot judge equal, because we are extremely biased. Artificial skewing to confront people with more and more of XYZ is a good thing as it help estimating objectively.
I caught myself questioning a young woman who said she was working in machine learning several times, to get to the root of what she was doing really (turns out the real stuff). I wouldn't have done it in the same way with a guy (i assume).
So in other words you think it is perfectly fine to treat people inequally based on what they are? Some people just happen to be "more equal" than others?
> And honestly we aren't. We cannot judge equal, because we are extremely biased. Artificial skewing to confront people with more and more of XYZ is a good thing as it help estimating objectively.
Who decides how much "artificial skewing" is enough and on what grounds? There can be no correctness or authority over this, and as such any kind of skewing is wrong.
> I caught myself questioning a young woman who said she was working in machine learning several times, to get to the root of what she was doing really (turns out the real stuff). I wouldn't have done it in the same way with a guy (i assume).
Just like we dislike that guy with long hair. Or that guy with pink bread. Or that terribly tall guy. Or that bald woman. Or that .... whatever doesn't fit the "mainstream" what would be expected from a person in that field. This is a symptom of people not understanding that we're all people, and we're all equal and should be treated as such, regardless of our differences. It's first and foremost a mentality issue, whether that can be solved or not is another thing. Adding more discrimination to "balance" things is like giving opposing forces more weapons to "balance" the odds and thus avoid the slaughter. No need to mention where it leads to.
For people who are female, gay or black it's not a choice.
A Star Trek fan with a comment on Captain Picard: "Surely by the 24th century, they would have found a cure for male pattern baldness."
Gene Roddenberry replied, "No, by the 24th century, no one will care."
(Though I think Roddenberry at first did care.) This is the attitude zxcdw is taking. It doesn't matter if something is a choice or not, if it's irrelevant to the context under consideration then it's irrelevant. Long hair, baldness, gender, sexuality, race, all seem irrelevant under the context of tech jobs and conferences. If there is discrimination based on irrelevant details, that's a problem with who is doing the discriminating, and giving an artificial leg-up to the discriminatee is a less effective solution than finding a suitable punishment (fired, stern talking to, etc.) to the discriminator. I'm in favor of giving incentives for the behavior of not caring about these things instead of giving incentives to discriminatees to be over-sensitive.
Where I may be in disagreement with zxcdw (I don't know) is that I have no problem with discrimination on relevant details. If someone is dumb, I wouldn't want to hire them for any technical job. (Similarly I wish we'd get rid of minority grants and artificial leg-ups like effectively adding SAT points to a new university student applicant. RyanZAG's comment on this page could have replaced 'conferences' with 'universities' and 'speakers' with 'students' and still be a coherent argument.) If it just so happens that all humans are not equal (something I believe), there may indeed be clusters of people that have a high correlation with one relevant detail (such as fitness for a particular line of work) and another supposedly irrelevant detail (such as their gender or race or sexuality). One can notice such clusters but it would still be improper to start caring about the irrelevant details just as much as the relevant ones.
Edit: mgkimsal made a good comment above on how things such as gender and cultural background can be relevant to contexts like conference speakers, where one might otherwise assume them to be irrelevant.
As far this discussion goes my main concern was that conflating chosen traits with inherited ones is derailing the discussion. They are very different with completely different forces are at play since 'chosen traits' are by definition acts of social signaling while 'inherited traits' are just that: inherited.
Just because Jamal is black doesn't mean he's any worse/better/special than Henry who happens to be white. Just because Sally is a lesbian doesn't mean she's any worse/better/special than Cindy who is straight. You get the point.
Stop categorizing people. There are no categories. There are only people. All equal regardless of age, sex, sexual preference, "race", origin, religious beliefs, ... Only people. Equal people.
Yes, affirmative action makes a big assumption when it posits that a lack of diversity is caused by prejudice. But might that not be the case, that often a lack of diversity is caused by prejudice that can most easily be countered by a directly oppositional force? To be fair, it is this assumption that must be challenged when you challenge affirmative action.
Also, from what I understand in the FLOSS world, the number of contributors that are women are once again a smaller fraction than the number of software developers in general.
What I am saying is that due to reasons such as social selection, bias, expectations of what women should do, discrimination etc less women are interested in software.
As you said, women are less interested in software, they also in general are less interested in science and technical tools than men are, this is why we can observe so few women in tech. I have no idea why this is, it could be "expectations of what women should do" or maybe and I know that this sounds bad but it can't be ignored just because it doesn't sound good, a natural aversion in many women (not all!) for scientific matters or maybe a combination of both and more. I have absolutely no idea as neither reasons have been proven scientifically so I won't venture into any wild guess. But blaming it on white males imposed discrimination seems ridiculous and misses the point.
Development looks like a boys club from the outside. Changing that perception will require active effort from the inside. No one puts in the effort because "We aren't sexists! We love women! (We love women so much! [Please, women, come talk to us!])" Apparently liking women forgives being completely incapable of creating an inviting environ.
None of this comes as a big surprise to anyone who knows developers. Right leaning people who are overwhelmingly from privileged backgrounds and spend most of their time absorbed in a subculture that places pride in technical accomplishment above manners. This is not a group to expect a warm welcome from, especially if it means that they might have to tell less crude jokes, or take down some objectifying imagery.
But sure, it could be a natural aversion. That's a studied hypothesis with thousands of hours of research behind it showing...something, somewhere...right? The truth must be somewhere in the middle, I guess.
This is not a simple matter of people being actively or even passively discriminated against. This is the result of the environments, decisions, and treatment people receive from their earliest elementary school days; there is almost nothing that can be done by the time people are in college. I have watched EE and CS departments do everything imaginable to try to increase diversity, yet they have made extremely modest improvements.
Somewhere between the beginning of middle school and the end of high school, girls seem to stop being interested in math, science, and things that are relevant to engineering and CS. Somehow, my female classmates who had high scores on the AP calculus test all wound up somewhere other than engineering -- psychology, accounting, management, etc. Despite every possible opportunity being made available to women, despite departments running the gamut of things that could attract women, despite women representing a majority of students at many colleges, women continue to avoid engineering.
So don't say that women are not getting the chance to be engineers or computer scientists. Women have more chances than anyone else; they are choosing not to take advantage of those opportunities. Figure out why that is the case, and maybe you can actually do something to solve the problem.
Wow. Such bold words. Except it's meaningless (or worse). Every person who submits a topic (PHP? jQuery?) to a conference that doesn't fit what the organizer(s) wants is 'discriminated' against. Every topic that is selected is 'discriminated' for.
Uh. Oh. That. Is. Wrong.
Sounds so much better when you. use. lots. of. periods. right?
FWIW, I organize what has so far been a 3-track 1 day tech conference with 21 slots. I go out of my way to make sure there are female speakers there (IIRC we had 5 or 6 this year).
Why? It's a factor that helps bring more audience in. Based on surveys and f2f chats, women have told me they're more likely to attend because they'll be more comfortable knowing women are presenting. More potential audience means more potential sales means lower losses or perhaps even a small profit.
There's other reasons I do this. The notion of saying that the organizers can pick "the best" presenters and topics is, at best, a guess, but is more often delusional. I've been to plenty of conferences where some 'big name' presenters were utter crap and the lesser known people were a mixed bag.
Getting a more diverse audience to attend - made easier by a more diverse speaker panel - will help improve networking opportunities, as you'll have people from more walks of life and with more diverse networks and opportunities of their own mixing and meeting. If everyone thinks the speakers and topics are the sole reason to attend tech conferences, you've either only been to one or two, or you're asocial, or perhaps trying to justify the non-diverse speaker range after the fact (yes, there's more options, but those spring to mind, especially the first one).
Mixing/mingling/networking at conferences - a place to let your hair down and meet others with common interests but different backgrounds - is (for me and several others I know) one of the prime reasons to attend conferences now. Many conferences put up speaker videos; many times the sessions are the 5th time a topic has been presented, or you've seen it before, or it's from the speaker's book; ask him (it's almost always a him, right?) about something in the hallway.
Hallway chats and networking sessions are never put up on video sites afterwards, yet are probably the most important parts of face to face conferences. These are the events where we can be human with a range of people. When that range is 200 white men under 30, it's less valuable overall, and when those people all hit 40, and are still seeing conferences that are all featuring under 30s white men, they'll probably start to understand the 'discrimination' thing.
Sexist women will choose conferences based on the gender of the speaker. Due to greed, and because these women are so desirable to have at conferences (why?), we should therefore discriminate in speaker choice.
Have I missed anything?
So let the market fix it, no need for affirmative action. It's just targeted advertising to move your product, at the end of the day.
This is an extremely bad idea. Firstly, it reduces the value of the conference itself as a means to share the best possible information. Secondly, it lowers the esteem and respect of the speakers by allowing the audience room to question their abilities. Lastly, and most importantly, it allows politics into the selection of speakers, and moves the system further towards a popularity contest from a meritocracy, where people argue for speakers based on increasingly superfluous factors.
These are technical conferences, and you guys need to keep them this way. The selection of speakers need to be done on merit.
On the contrary :
Conferences are simply showing up the diversity problem
in a particularly acute way – we will never fix it by
pretending the industry is more evenly distributed than
it actually is, and later blaming organisers for it.
In practice, however, while one can often tell a really bad speaker and a really good speaker apart there are plenty of bordeline cases and the variance is big, so using minority-biased criteria to "break ties" will not necessarily lead to worse speakers on average, as long as these criteria are uncorrelated with actual skills.
(and I'm not sure they are, as the mean quality of a woman-presented talk in conferences I've attended has been consistently higher than the mean man-presented talk)
All it does is give you statistical noise, making it easier to bury your head in the sand.
1) Diversity is valuable when it represents diversity of life experience and problem-solving methods, not when it represents some external characteristic like skin color or gender. To substitute these terms out is to be prejudiced, that is, to believe that we can look at someone externally and make informed decisions about their internal makeup. We cannot.
Two, you do not manage complex systems by simple statistics. If you could, then all the billions being spent in root cause analysis and so forth would be unneeded. Complex systems, such as those involving people, are, well, complex. For example, if the inputs to the system are heavily skewed, that is, if applicants for speakers come in at a heavily skewed ratio, then the process itself can't be responsible for the outcome. Of course, if the process only serves to facilitate an outcome, then that's fine, but it's no longer a process optimized for finding the best speaker or whatever. You can't have it both ways.
I can provide more examples of how trying to reason about these systems in a simplistic fashion is not only a waste of time, but can actually lead to counter-productive results.
I refuse to be sexist or racist and I won't tolerate those around me who are sexist or racist. But I also refuse to continue to consume material that promotes this kind of mushy-headed way of looking at something as important as diversity. If you don't understand the problem and don't understand how to deal with it even if you could manage to identify it, then any resulting conversation or action is not going to be very productive.
First of all, gender is not exactly an "external characteristic". Secondly the life experiences of men and women are inherently diverse.
Hopefully, someday the colour of your skin won't influence your life experience, but that is not the case in any human society that I am aware of.
Software isn't one of those fields, so bringing up gender parity in nursing in a discussion about the software industry is a bit of a misnomer.
[1] http://www.minoritynurse.com/men-nursing/recruiting-men-nurs... [2] http://www.degreesfinder.com/online/articles/should-there-be...
In either case, shouldn't organizers have the right to pick whatever they think is most interesting and then let their auditory choose, whether to participate or not in their event? By the very same logic, shouldn't editors have their right to choose what content they publish, with their readership voting for that choice using their wallets? I think in the end, if it is really important enough, underrepresented parts of community are still able organize their own conference, with all the right speakers.
If so, please tell us!