I’ve gotten a few emails and /msgs about this so I really wanted to post a clarification.
When I put the color survey together, I was mostly interested in making maps and tables of color names; the opening survey was almost an afterthought. Finn added a question about chromosomal sex, since it’s closely correlated with colorblindness (Finn is one of the rare people with two faulty X chromosomes for color vision).
We debated for a long time to find a wording of the question that would be answerable unambiguously by everyone, regardless of gender identification or any other issues. In response to a friend who was suggesting we were overcomplicating things, Finn said, “I *refuse* to word the question in a way that doesn’t have a good, clear answer available for transsexuals, intersex people, and people who already know they have chromosomal anomalies.” I felt the same way, and at the same time I didn’t want to assume everyone remembers what the hell chromosomes are. After hours of debate, everyone was happy with this:
Do you have a Y chromosome?
Don’t Know Yes No If unsure, select “Yes” if you are physically male and “No” if you are physically female. If you have had SRS, please respond for your sex at birth. This question is relevant to the genetics of colorblindness.
We didn’t add a question about gender identification, in part because I wasn’t really planning to do anything with the survey data beyond basic calibration and didn’t want to hassle people with more questions, and in part because gender is really complicated. We recently programmed Bucket, the IRC chat bot in #xkcd, to allow people set their gender so he can use pronouns for them. This ended up taking hundreds of lines of code, three pages of documentation, and six different sets of pronouns and variables, just to cover all the basic ways people in the channel with different gender identifications wanted to be referred to (even without invented pronouns like “xe”, which we vetoed). And that’s just to cover the pronouns. The role of gender in society is the most complicated thing I’ve ever spent a lot of time learning about, and I’ve spent a lot of time learning about quantum mechanics.
So when I wrote the survey, I really didn’t have anything in mind for the data. After it went up, I saw the DoghouseDiaries comic, and immediately wanted to investigate. I was really amazed by the results, particularly the top-five list of colors, which came as a complete surprise. Everyone agreed it was the most interesting of my results (at some point, my friends were sick of hearing me talk about hues and saturations) and I couldn’t resist publishing it somehow.
Originally, my post had a big wall of text discussing how all I had was chromosomal data, and that what the comic talked about was gender identification. I rewrote this post a bunch of times, and ended up with roughly the wording that’s there now:
[…] realized I could test it (as far as chromosomal sex goes, anyway, which we asked about because it’s tied to colorblindness).
I didn’t want to spend a long time boring people about sex and gender (I’ll talk forever if you let me), but I also wanted to clarify that this was something I cared about and was trying to pay proper attention to. I ran it by some friends before posting, and they approved; one specifically thanked me for adding that note. So I figured I’d found a good balance.
But a number of people were still offended or upset by my use of the chromosomal data in a conversation about gender. Now, there are always going to be people upset about anything; as Ford Prefect said, “Fuck ’em. You can’t care about every damn thing.” But this is an issue I really do care about, and one I spend a lot of time trying to get right—and I genuinely appreciate the guidance. If people were offended or feel I didn’t handle this right, I’m sorry, and it’s my fault. But it wasn’t for lack of caring.
And to anyone writing software that handles gender or sex information, it’s a good reminder that these questions are not always straightforward for everyone, and a little courtesy can do a lot to make someone feel respected.
xkcd 
In what way were people offended? That the question wasn’t a straight “male/female” thing, for example, or that they belong to some smaller group that they felt wasn’t represented?
LikeLike
I really appreciate how much effort you put into sex vs. gender. Thank you.
LikeLike
The offended people might have been intuitionists, who don’t accept the law of the excluded middle.
LikeLike
I, being trans, appreciated the distinction. Yes, it kinda sucks to identify as being genetically male, but the fact that you didn’t just put a Sex: M/F/U was awesome.
LikeLike
Lol, people offended by statistics.
Lol, my recaptcha contains “smuggest”.
I am offended by my recaptcha.
LikeLike
I’m really impressed and touched by how much thought you put into this topic and by how you acknowledge the complexities behind gender.
LikeLike
I’d have to agree. Not that I have any kind of background in these kind of things, but… Taking this kind of data and extrapolating is bound to produce generalizations. Unless my biology classes were horribly wrong, most people with a Y chromosome are male, and visa-versa. So, I fail to see the problem with equating one with the other in such general results.
I’m sure that there were some Y-chromosome-holding people who named colors after (what I take to be) lipstick colors. And I’m sure there were some Y-chromosome-lacking people who were reminded of the male anatomy from some shades. But they were in the vast minority. Whether that minority is more or less significant than the whole gender-confusion thing is up for debate… but for statistical purposes, they’re both still minorities.
LikeLike
Thank god for simplicity, a man is male, a woman is female, if they have more complicated situations I’d just use whatever pronoun they prefer, the whole subject is in my opinion ridiculously overrated and overdebated.
Though I still really appreciate the effort you made in this case.
now back to studying Quantum Mechanics …
LikeLike
Personally I agree with Ford Prefect. The Hitchhiker’s Guide to the Galaxy is a great book filled with truths. You can’t please everyone so I don’t even try. I am myself and I’m not going to change the way I am just to save someone else’s feelings (not that I’m going to be an a$$ hole or anything, but I’m not going to walk on eggshells for anyone).
Either way, I think you did an adequate job for ambiguity and if that matters to you, then all the more power to ya.
LikeLike
Amen. Especially to that last line. Thank you for putting so much thought into the subject!
LikeLike
@ Ammar
You cannot state that a man is male and woman is female. Despite how a person is born based off of their biological chromosomes does NOT define the gender that they will chose for themselves. Our genetic sex is only one part of the complicated issue involving gender.
It’s also easy for you to say its ridiculously overrated and over-debated because I assume you fit into one of the two boxes that the majority fit into, male->man or female->women, Do you have any idea how difficult is for someone that is biologically born as a female but every part of their psyche is telling them they are a man? When they have to live in an entire society that is constructed to a binary gender system that doesn’t allow for cross-gender expressions?
It’s an issue that can’t be described in one blag comment post as I have been working for years to simply define for myself what it means to be a man who is also biologically male. It’s because of this that my campus has an entire department devoted to gender and sexuality studies. So please, take some time to think about the gender vs. sex issue a little more because it isn’t as simple as you may think.
Now I am back to studying QM too…
LikeLike
Yes! Sex, not gender, is what’s relevant here. Thank you for getting this right!
Also – colorblindness is not all the same. The commonest type, red-green, involves the X chromosome. However, the rarer kinds can be due to gene mutations on somatic chromosomes – eg the blue-yellow type.
LikeLike
I have to say, I really appreciated the wording of the original question. I’ve gotten far enough along transitioning that I can say “Not a woman after all, oops!” but not far enough that I feel comfortable identifying as anything else. I’ve spend the last three months flinching every time I fill out a form, no matter how stupid the form, because I can’t mark M or F. (Sometimes it makes me cry.)
But my chromosomes– I know those. That’s much easier.
And the fact that you had an actual reason to ask about my chromosomes (colorblindness)– that made it easier too. It drives me crazy how many websites and offices and so on ask for “sex.” Especially since the ones that aren’t “M or F?” are “Male, Female, or Transgender?” which, uh, isn’t really how it works. (At least let me check two of the boxes!)
Gender identity, genitals, and chromosomes are all different things, and I hate it when I don’t know which one the form is asking about. Especially because they probably don’t know either, and don’t have any valid reason for asking about gender at all, except that it’s some kind of habit for forms. I think if people had to think about which category of information it is that they need, they’d ask a lot less, because they’d realise that none of it is relevant to my Yelp profile.
So the think I really appreciate here is that you thought about whether you needed sex/gender info at all, concluded that chromosomal information was relevant, and gave a justification right in the form when you asked. It was the easiest gender question I’ve ever had to answer.
Publishing the results with gender info suggests that maybe there should have been a second question, “What is you gender identity?” but I felt that your parenthetical sufficiently established that you knew what assumptions you were making, and that you didn’t collect chromosomal data with the intent of calling it gender/sex data.
Also, I did learn my colors when I was a child… a girl child, as far as anyone knew. So while I would have strongly preferred to have my “butter yellow” and “peaches and cream” colors helping out your dude column, I picked up the habit from by chick upbringing.
tl;dr version – I’m trans and I see a lot of cis people screw this stuff up. This is the first time I’ve seen a cis person grappling with trans issues such that I think, “You know, he basically got it right.” Thanks for giving it all so much thought.
LikeLike
> a little courtesy can do a lot to make someone feel respected
You’d think…although the negative responses you got seem to better support the hypothesis “a little courtesy is largely a waste of time”!
LikeLike
i understand that political correctness is very popular right now, but if we just go around trying to avoid stepping on everyones toes, we will surely have an extremely boring society. anyway, isnt it fun to be a little incorrect now and then? anyways, we dont always have to refer to every permutation of a form ever created, we’d have so much space just for name, date, age, sex etc.
LikeLike
A still better question have been “how many X chromosomes do you have in your retinal cells?”, though. As it stands, people with Klinefelter’s and certain chimerical people might have thrown off your data. (If, you know, there were enough of them taking this survey to be statistically significant, which was probably not the case.)
LikeLike
@Boxy
“You cannot state that a man is male and woman is female”
That’s the natural conclusion unless they have other situations, That’s exactly why I said I’d refer to them as they prefer, I don’t need to check with everyone, it’s always their choice and I respect it.
I’m really sorry for the ambiguous reference, the issue that I meant was overrated was the wording of online forms, as we all know it’s very rare that the choice actually matters for the service provided, the only common exceptions that come to mind right now are social sites (just choose what corresponds to your preferred pronoun) and medical services or similar sites (in which case I love the ‘Y chromosome’ approach).
LikeLike
Karl:
We talked about that, in fact; did you know there are people with up to *five* X chromosomes? Apparently there’s on average a 10- to 15-point IQ drop for each extra X, plus lots of other problems. But no, fortunately this isn’t enough to skew the data, since the trend was pretty strong—in fact, if all the trans people reported the opposite answer on the chromosome question, I’m fairly sure the top list would stay unchanged (given how many people would have to switch to alter it), and I suspect the percentages would probably get stronger (although that’s just my hunch; I’ve got no data to back it up).
LikeLike
Marc, you wrote, anyway, isnt it fun to be a little incorrect now and then? That certainly is easy to say when you’re not going to have your life erased from the conversation, isn’t it?
Anyway, colorblind people tend to know that they’re colorblind, don’t they? If you were only asking about sex chromosomes for the purpose of knowing when responders were colorblind, couldn’t you just have written, Are you colorblind y/n?
LikeLike
Which we did, actually—”Do you know that you are colorblind?”
Elizabeth added the sex question, I think to just see whether the colorblindness percentages reported in the literature for each sex held up (since she’s that rare colorblind XX) and also to get a sample of people we were more sure *weren’t* colorblind (those without a Y chromosome) to see (among other potential things) whether we could spot colorblindness among people who had it but didn’t know it.
LikeLike
Allow me to join in the thanks for addressing this. It’s nice to know that we gender outlaws have allies everywhere 🙂
LikeLike
Karl: We considered asking about X chromosomes instead, but after doing a bunch of reading about chromosomal anomalies (which I hadn’t known much about before starting this project), I concluded it was more likely that someone might have an extra X chromosome and not know it than have an extra Y chromosome and not know it. Plus, having “three” as a possible response seems like it would’ve made more peoples’ eyes glaze over and increase the likelihood they’d skip the question.
All: The thing I was interested in when I added the question had nothing to do with gender–there wasn’t a clear picture at all yet of the ways Randall would be categorizing the results. What I wanted to know was how accurate the “Are you colorblind?” question was. Knowing some basic data about the chromosomes of the participants and about the genetics of colorblindness would tell us roughly how many people were “supposed” to be colorblind, which we could then compare to the set of people who knew that they were. If the difference was drastic, it might be worth noting as an effect on the results.
Although Randall accepted responsibility for the trouble this has caused, it’s just as true to say I started it by putting the question there in the first place. To all the people who felt hurt, angry, or excluded by it, I am deeply sorry. I never stopped thinking about you when deciding how to word the survey, and I wish that had been more clear. If you have the time, I’d like to hear how I could have made you more comfortable. (Rather than having a discussion with me in someone else’s blog comments, please use the contact info on my website to reach me privately.)
Randy: I’m sorry you took so much flak for supposedly not caring about something I know you actually care very much about.
And to the transitioning anon who commented above, it’s a relief to know that some people did understand what I intended to convey. Thank you for posting to say so.
LikeLike
xkcd: Uh, we did.
Haha, I… honestly have no idea how I managed not to notice that you’d done that. ok thx n/m
LikeLike
Imogen: Not necessarily! I didn’t know that I was colorblind until last year. It’s very mild (I am, I believe, pseudoanomalous, which is like red-green colorblind but less so) and if I’m looking at almost anything except one of those dot test plates nothing seems amiss.
LikeLike
Interesting discussion! It sucks that people were offended or upset, but I can’t imagine there’s a way to phrase things that will guarantee that nobody will get offended or upset… it’s just human nature. There’s the question(s), which you can control, and then what people bring to the question(s), and there’s just no way to control that.
It’s obvious you’ve made a serious effort to do the right thing regarding gender identity; I’ll defer to people more familiar with the topic as to whether there would have been even “righter” ways to handle it, but the question as you phrased it did stand out as something coming from someone taking this seriously and trying to be respectful of everyone.
LikeLike
Having spent a huge amount of time studying QM and recently done a bunch of work on medical databases, I feel your pain and agree with your assessment of the complexity of the issue.
There are two reactions to intersex people:
1) “Sex and gender is too complicated!”
2) “Man/woman classifications are vastly over-simplified!”
I think the latter response is the right one, and asking “Do you have a Y chromosome” is not a bad way of dealing with the complexity in this case. Unfortunately in other cases, including the one I was dealing with, what really mattered were hormones, so in a longitudinal database covering potentially decades dealing with intersex people was tricky, and in the end I punted and settled on “Male/Female/Other” to at least capture the cases where sex needed to be dealt with more subtly. It was a bit of a battle getting the PI to accept that much.
What we need I think is to dispense with binary gender entirely, and have something more akin to the visible spectrum as a way of specifying both sex and gender: it’s multidimensional, and there is a sufficiently rich range of names to provide for a usefully fine degree of differentiation.
LikeLike
I don’t understand being offended by chromosomal data in this case. It really isn’t a conversation about gender- unless I’m missing something- it was a conversation about color blindness which is related to chromosomes (sex, not gender). There’s a long list of things I’d’ve added that I’m interested in, but that seems pretty basic.
You didn’t even ask “do you have male chromosomes?”. You just asked if the person has a ‘Y’ chromosome. (although I wonder how being XXY/XYY/X0/etc would effect those statistics)
But also grateful that you put so much thought into this.
@Ammar “I’m really sorry for the ambiguous reference, the issue that I meant was overrated was the wording of online forms, as we all know it’s very rare that the choice actually matters for the service provided, the only common exceptions that come to mind right now are social sites (just choose what corresponds to your preferred pronoun) and medical services or similar sites (in which case I love the ‘Y chromosome’ approach).”
Even then it’s touchy. For example, medical services, an AIS woman (phenotypically female, doesn’t respond to testosterone, genetically XY) whould mark ‘I have a Y chromosome’ and would be assumed to be transsexual- when in reality she’d have different needs.
I think it’d be better to have an option of “it’s complicated, I’d be more comfortable explaining it to the doctor so he knows the whole situation” or a space to fill info in for people who have an unusual sex with different needs. There are a large variety of intersexed conditions with and without a Y chromosome (or 3)- and they can’t be treated the same way.
Even with transgendered people- the “do you have an ‘Y’ chromosome?” is lacking because a post-op trans woman will have different medical needs than a non-op, person who’s only on hormones/anti-androgens, etc.
LikeLike
I understand that there are many complications on how individuals describe their sex and their gender, phenotypically, genetically, and mentally. The personal nature of such definitions and identifications exacerbates the issue as well. Keeping this in mind, I completely understand why Randall asked for advice from lots of people. On the other hand, I don’t believe that this was intended to be a particularly rigorous scientific survey. The self-selection bias, among other aspects, likely affected the results in unpredictable ways.
In many ways, I liken ambiguously worded questions on sex and gender to questions in political polls. The question, “Do you support the ability of a woman to have a legal abortion?” means very different things to different people (depending upon how they interpret the restrictions or lack thereof in the question). A pollster could generate questions that conform to all the different combinations of ways individuals could support or not support a given issue, but that is usually beyond the reach that the poll is trying to study.
The most we poor subjects can do, if we want to participate, is to give the best answer for questions that are ambiguously worded.
LikeLike
Dos Ocho: I absolutely agree that one can control one’s own communication output but not the experiences and inferences that other people will use to interpret it. It is, however, possible to consider what those might be and do one’s best to allow for them. It’s easier when talking to one person face-to-face, of course, but even when broadcasting a message to a lot of people a lot can be done. This is something I think a lot about–communication between humans is a fascinating topic to me, especially what happens when it goes wrong and how to prevent that.
LikeLike
Z: I think people were upset not because the survey asked about chromosomes, but because the survey asked about chromosomes and then gender was discussed in the post. It wasn’t at all obvious that those two things were coincidental. If I hadn’t been involved in making it, I could easily have misunderstood and believed that he was equating sex and gender, in which case I would’ve been pretty upset too.
LikeLike
“The role of gender in society is the most complicated thing I’ve ever spent a lot of time learning about, and I’ve spent a lot of time learning about quantum mechanics.”
I wish that was short enough to tweet.
LikeLike
would you consider a sex-vs-gender-survey?
number of X and Y chromosomes, and number values for identities (just throw in all keywords, have people rate how much these describe them. like: 50% butch, 30% femme, 30% androgynous) maybe split between self-image and outside-image…
LikeLike
I thought it was strange that you had asked if the participant had a Y chromosone and assumed it was just a witty way of asking about sex. It is not unusual for a form to simply state male or female in the slightest, so to take offense at somebody asking from a scientific standpoint if you have a certain chromosome seems beyond comprehension. Then again, this is the internet after all.
The fact that you put so much thought and consideration into how you phrased the question is heart warming and more people (probably including myself sometimes) should be as thoughtful of others as you have been.
LikeLike
Thank you; you are awesome.
LikeLike
I can see why people were troubled, and could have used a slightly longer note. (Now, of course, it has a slightly longer note pointing here, so that’s covered!)
I deeply respected your wording on the survey, and your justification, and the fact that you at least briefly called attention to the distinction again before launching into your analysis. Thank you for your care and attention to this.
LikeLike
As a transwoman, I deeply appreciated both the way you worded the question on the survey (and clearly stated your purpose in putting it there) and how you handled it in the resulting post. It seemed clear to me that you had put some thought into the subject, and were doing your best not to propagate harmful memes. 🙂
LikeLike
Wait a minute, Bucket’s a bot?
Mind blown.
LikeLike
I’m an XY lesbian, and as a result, I can say from first-hand experience that gender can be very, very, very complicated. I was touched by how much thought was put into the wording of the “Y chromosome” question. Thank you for making me feel comfortable while doing science. ^^
On the survey itself, I find this data extremely interesting. I’ve been considering a similar project that would ask people to choose a color for a given word rather than a word for a given color, and this could provide a good starting point.
What is the color of “caboose” anyway?
reCAPTCHA: “milkmaids Fellow”
LikeLike
wait…. It took you “hours” to figure out how to ask whether or not someone is male or female… ? all personal sexuality bullcrap aside. -.-
LikeLike
I actually was taken aback at the sensitivity of the original wording. Not that I expected much different from xkcd, but it’s a side-effect of being consistently disappointed. My initial reaction was that in a colour survey, asking about sex, specifically as defined by chromosomes, was not only pertinent but arguably necessary. The elaboration is still greatly appreciated.
LikeLike
It strikes me that, whether you want to be identified as male or female, the phrase ‘Are you, biologically, (male/female) ?’ is pretty clear and concise. Personal preference be damned; if you’re a ‘blank’ trapped in a ‘blank’s’ body that doesn’t make you any less of what you actually are.
Granted, there are anomalies, and so I do agree with the wording of your question. And, incidentally, I’m aware that not everyone (particularly in the scientific community) can afford to be as abraisive as I am about the subject, but damn, for the sake of science you’d think transgendered people or people who have undergone SRS (or whatever the circumstance may be) would be able to answer the question ‘What gender were you born?’ without any ambiguity.
LikeLike
I’m not an expert on gender; but in my humble opinion both your survey and blog post of results were fine, and anyone getting upset about their gender being misrepresented in there is just being oversensitive about it. — However, I’m not saying that there isn’t room for criticism, I’m just saying there is nothing there worth getting upset about.
LikeLike
@Mike:
“The role of gender in society is the most complicated thing I’ve ever spent a lot of time learning about, including quantum mechanics.”
LikeLike
I think the way it was on the test was perfect. Anything more and it would’ve been too much. I’m inclined to agree with Ford on anyone who thought the data was used incorrectly after that, but I don’t know anyone who is TG, and can see why it would be a bigger issue to someone who does. Slight tangent, I was outraged that facebook requires you to pick between M and F to make an account. I have no gender or privacy issues myself, but really don’t see why it’s so important.
LikeLike
> But no, fortunately this isn’t enough to skew the data,
> since the trend was pretty strong—in fact, if all the trans
> people reported the opposite answer on the chromosome
> question, I’m fairly sure the top list would stay unchanged
so, you are saying, however you asked that question, the results would pretty much end up the same. and you probably knew this even before asking it (given you investigated the issue, and given your understanding of statistics).
couldn’t you just ask the simpler question about gender (which “refers to the personal inner-sex identity” according to wikipedia). those who don’t care would simply answer male/female, but those who do could answer as they see appropriate.
LikeLike
something just occurred to me, you should have both questions the Y chromosome question and the male/female question, cause there are guys out there with two X’s who just might be colorblind. Just to clear up any lose ends. xD
LikeLike
Exposition: I’m trans. Hi! I haven’t read all these comments ’cause it looks like some of them would make me angry.
BUT!
1) It felt a little awkward for me to mark the box–not because of the question itself, but because of the clarification for people who don’t know, which kind of equated having a Y chromosome with maleness. Still, I get why it needed to be there.
2) I can’t express enough how thankful I am that you’re engaging with this stuff, and I think you’re right on. I’ve had a soft spot for Randall the person (and not just Randall the cartoonist) ever since you mentioned a discomfort with portraying heteronormativity at the Cambridge meetup, but it’s so nice to have you not utterly fail at trans issues like many folks do.
3) One problematic thing, though, is the way the data was used. Your definition of that variable was very clear. But in the interpretation, suddenly we’re assuming that chromosomal sex _is_ gender. (Yes, there’s a disclaimer sentence, but we’re still doing it.) Investigators often include gender in analyses even when it’s not relevant to the hypothesis, because they think of it as an easy, free piece of information to do a fishing expedition with (for fun, try convincing someone that their form/survey/etc. _doesn’t_ need to ask gender sometime). And I think what we see here is the result of a post-hoc fishing expedition, using one variable as a proxy for an unmeasured one. On a contentious topic like sex/gender this can get you into trouble, though :-.
Um, that seems like more negative things than positive things, but I am also trying to say you’re awesome and thank you. So. 🙂
LikeLike
I really feel for you. When I talk about cyan and blue, my friends start threatening me.
And I haven’t even talked to them about the whole pink/brown/orange thing, yet.
LikeLike