THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
— The GNU General Public License (GPL), Version 3

At the beginning of this year, First Monday (a longstanding, online-only academic journal) published Joseph Reagle’s article “‘Free as in sexist?’ Free culture and the gender gap”. The article is the only comprehensive study I’ve seen so far of online discourse drawn from free and open-source software and data communities that focuses on attitudes towards gender and sexism.

In what follows, I examine Reagle’s presentation of two major themes: how dominant definitions of “geek identity” serve to keep communities homogeneous; and how ideologies held by open-source workers sometimes serve to justify mistreating people in the name of freedom of speech. Finally, I suggest another reason for open-source communities’ problems with diversity and equality: an economic one. I’ll use the terms “geek culture”, “open source culture”, and “hacker culture” roughly interchangeably. Not all geeks or hackers work on open-source projects, but open-source communities represent, to me, a highly valued position in the hierarchy of value subscribed to by many people who identify as geeks and/or hackers.

Disenchantment

I had a visceral reaction to the “On being a geek” section of “Free as in sexist?” This section covers ground that is familiar to me: the obsessive, monomaniacal approach to programming that hacker culture valorizes; the relationship between this style of working and a confrontational, aggressive style of argumentation; and the relationship between geek identity and normative whiteness and maleness. (As I don’t have any special authority to speak about race or racism, I won’t discuss those issues in depth here; I recommend Mary Bucholtz’s paper “The Whiteness of Nerds: Superstandard English and Racial Markedness” [PDF link], in which she argues that American nerd culture consitutes an explicit rejection by certain white youths of those aspects of American popular culture that arise from Black Americans.) Even so, the section affected me on more than just an intellectual level. As I read the quotations Reagle chose from sources such as Richard Stallman’s and Joseph Weizenbaum’s writings, as well as interviews with women studying computer science, I felt afraid and disappointed. I felt ready to get out of this field myself as fast as possible. Before I could help myself, my subconscious was already rushing ahead and reviewing the plans I’ve turned over in my head about jobs and careers that I could do that wouldn’t require me to be either a Toxic Open Source Guy, or an enabler for one.

When I was 15, sleeping in a lab and working for 20 or 30 hours at a stretch appealed to me. I wanted to lose myself in code, stop noticing my physical body because I was too engrossed in turning over abstractions in my mind. I think some part of me thought that if I got to be a competent programmer, it wouldn’t matter that I didn’t know how to form connections with other people or that my body was the wrong shape for me. I know now that escaping into work is not a helpful coping mechanism for me. Nowadays, I’ve exercised agency to make my body more comfortable for me; I see a therapist; and I have friends. I want to do my job reasonably well for eight hours a day and go home. I don’t want to run away from life outside the screen by immersing myself in work. I know most of the guys who do the sleep-in-the-lab, work-twenty-hours thing aren’t running away from what I was running away from. (I wonder what they are running away from.)

In Reagle’s article, I read, “Bente Rasmussen and Tove Hâpnes found female CS students who did not want be associated with the dominate [sic] identity of “key-pressers”, i.e., those who were not able to talk about anything beyond computers.” I thought — that’s me, too! I don’t want that either. I don’t think I have to quit being an open-source programmer if I want to have an identity that isn’t just about computers. But sometimes when I’m around people who do seem more like key-pressers than I am, I feel like that’s the way I have to be in order to fit in and be accepted.

Then I try to imagine what it would be like for me if on top of all of this, I felt like I had to conform to a vaguely woman-ish gender role. I didn’t know I wasn’t female until I was 18, and didn’t know I was male until I was 26, but I never felt much pressure to be what girls or women were supposed to be. On the other hand, if I was a cis woman, or even more so, if I was a trans woman (since trans women get expected to conform to gender stereotypes for women even more so than cis women are when their trans status is known), working in the industry I work in, I would have an almost impossible set of constraints to solve. As Reagle shows, success and status in open-source (and even in non-technical “free culture” communities like Wikipedia editing) are correlated with adopting a (superficially) overconfident, aggressive, argumentative persona. Women get to choose between being socially stigmatized for violating gender norms, or being ignored or mocked for violating open-source cultural norms. It’s a double bind.

Reagle quotes a passage from Margolis and Fisher’s Unlocking the Clubhouse: “‘Scary’ and ‘afraid’ are words that recur again and again.” For me, these are emotions that recur again and again when dealing with open-source culture, and when recalling the memories that reading Reagle’s article brought to mind. What strikes me, though, is that I’m almost twice the age of some of the undergrad students who Margolis and Fisher describe. When I was those students’ age, CS culture seemed safe, not scary. It was the rest of the world that was scary to me. Now, something’s changed. I think part of it is that I’ve had too many conversations with colleagues about gender politics that leave me feeling angry, frustrated, and helpless. Those interactions leave me afraid of being dismissed, dehumanized, objectified, or belittled again if I speak up. I’m also afraid of the sinking feeling that, for me, comes from being silent when I witness something I know is wrong. After a while, just walking in the door to the office seems like an entire day’s work.

Another quotation (from a social psychology journal article by Sapna Cheryan and colleagues) that stood out to me was “The profoundness of this alienation is hinted at in a recent study that found even an ‘ambient environment’ of stereotypical geeky items in a room (e.g., science fiction memorabilia and junk food) depressed female undergraduate interest in computer science.” While looking for a new place to live near my workplace in Mountain View, Ca. recently, I was browsing through rentals on AirBnB, and found a post advertising a bunk in a “hacker fortress”. I think the feeling I had when imagining living in such a place might be akin to how the women in that study felt when they saw a roomful of Star Trek figurines and Mountain Dew Code Red bottles. At 15, the summer I was doing an unpaid programming internship and drinking Jolt in the mornings, living in a “hacker fortress” would have seemed like an exciting idea (never mind the potential rape and sexual harassment that someone who looked like I did at 15 would have experienced — I probably would have dismissed that risk at the time). Now, even contemplating having to live in a place with a name like that sends my stomach dropping through the floor.

This section of Reagle’s article is valuable for showing that what I and so many others have experienced is part of a pattern; it’s not a coincidence, nor is it due to some weakness of character that we all happen to share. Women who have been involved, or tried to be involved, in free culture encounter hostility, not as a universal rule but as a recurrent pattern. It’s certainly not that Joseph Reagle is the first person to point out that free culture is systematically hostile to women — women have been saying this for a long time. But the evidence he collects is one more persuasive tool to put in the toolbox for convincing the naïve that yes, geek culture has a sexism problem. In the long term, though, we won’t have made any progress if people in the dominant group only believe women’s experiences when a male academic documents them.

It’s not just women who have been saying it, either. What Reagle doesn’t mention is that queer, trans, and genderqueer people in open-source share many of the same experiences that women do. In my opinion, like most transphobia and homophobia, that’s collateral damage from a fundamental hatred of anyone perceived as departing from a constructed heterosexual, cis male ideal — and that includes cis and trans women, as well as queer men and genderqueer and gender-creative people. (The omission of queer and gender-non-conforming people’s experiences could be due to a lack of written sources documenting it; there are various reasons why people in gender, sexual and romantic minorities might talk about their stories in a forum that lacks a permanent record.)

What makes me sad about all of this is that I still want to be around intellectually curious and playful people who are passionate about learning and making things (though, ideally, ones who don’t limit their inquiry to a single narrow specialty). I still want to have peers who inspire me to be and do more. I still love nerd humor when it isn’t mixed up with brogrammer racism and sexism. But what keeps me out of spaces that attract people like this is that I’m tired of being erased, silenced, and talked over. When I say how uncomfortable I feel when someone is engaging in homophobic hate speech at my workplace, and I’m told that it’s not hate speech or that my reaction to it isn’t real or valid, that’s stressful for me. It makes me want to disengage from the whole community. I’m tired of my female friends and colleagues getting death threats. I’m tired of being told I have a victim complex if I talk back to the abuse that gets directed at me and my friends even if nineteen out of twenty times, I’m silent about it. (It’s actually when I’m acting the least like a victim — when I’m not passively accepting whatever abuse is directed my way — that other software people shame me for “playing the victim”.)

The Mythical Manarchist-Month

While “On being a geek” was an appreciated summary of ground familiar to me, I found the “Openness” section more novel. I was pleased to see that Reagle opened the section by referring to Jo Freeman’s “The Tyranny of Structurelessness”, because Freeman’s article resonates with me strongly in light of last year’s troubles at Mozilla.

In my opinion, though, Reagle leaves a few dots unconnected in his discussion of “‘bad apple[s]’ and ‘poisonous people'”. If it’s really a minority of the community that (quoting our own Terri Oda) “actively hinder women’s participation by trying to derail discussions, make contributions significantly more time-consuming, or send inappropriate or even violent private messages to contributors”, then why are they allowed to effectively dominate the community by putting pressure on women to leave whenever they feel like doing so? I think it would be doing a disservice to everyone to ignore the role of the majority of male contributors in the community, who stand back and watch, who fail to exercise effective moderation in discussion forums, who lack the courage to confront other men who are being actively sexist. It is also a disservice to everyone to ignore microaggressions. The ultimate effect of death threats or a constant stream of little reminders that no one feels obligated to include you (like co-workers addressing a mixed-gender group as “guys”) is to make out-group members feel like they’re just not wanted. “Good” people (people who think of themselves as tolerant, polite, and considerate), not just toxic “bad apples”, can engage in microaggressions. And even “good” people often get unnecessarily defensive when called on behavior they weren’t aware was a problem. There’s a fine line between recognizing the disproportionate power of a small number of belligerent people in the open-source community, and using that an excuse for other people to do nothing in response.

The section titled “Ideology” is perhaps the most challenging one to the cherished beliefs of open-source participants about themselves and their role in the political economy — Reagle tallies up a damning list of open-source idols (Stallman, Raymond, Wales…) and their Randian beliefs that would be amusing if we weren’t talking about men who so many people take seriously. Reagle’s insights on how an anarcho-libertarian ideology lends itself neatly to justifying the rightness of the existing gendered power structure are sorely needed. But again, I think he could have gone a bit further. The thing about freedom, at least the way it manifests today in open-source communities, is that it looks a lot like freedom from accountability, without freedom from the very real constraints that burden the many. It’s free as in freedom, not free as in beer, but I’ve started to hear “free as in free from consequences” when I hear open-source people use “free speech” as a reason to be abusive. It’s customary in both open-source and closed-source programming to use the legal mechanisms of licensing and copyright to absolve oneself of all consequences resulting from bugs in one’s software, as per the quotation from the GPL that I opened with. This is not where I want to debate the merits of that approach to the profession of engineering — I do want to ask what happens, though, when a programmer extends that approach to licensing into his personal life. What happens to a community when many of the individuals in it assert their right to “free speech” and thereby claim entitlement to shift responsibility for the consequences of their actions? Typically, when people feel entitled to make others pay the cost for their choices, the people who end up paying are people who the underlying social power structure places as subordinate. I’m using the pronoun “his” because people who are not socially recognized as men (specifically, white men) simply lack the power to do this.

One example of this freedom from consequences is the refrain that so many of us who speak out have heard, over and over, from our colleagues: “Have I offended you? Then the problem is that you’re so easily offended. Your feelings are your responsibility, and I have no obligation to not offend you. No one has the right to not be offended, and anyway, I’m an equal-opportunity offender, so if other people can take the heat, why can’t you? It must be because there’s something wrong with you. You really ought to lighten up, take a joke, get a sense of humor, not let those words have so much power over you, be less sensitive.” (The routine has become so standardized that Derailing for Dummies, as well as the Geek Feminism Wiki, catalog it.) How can these incantations of emotional blame-shifting be unrelated to the disclaimer of responsibility that appears in the GPL and other software licenses? If what characterizes the professional culture of software engineering is our refusal to own our work, what characterizes the after-hours culture of programmers is a refusal to own our words. It’s a culture of solipsism that makes minority group members into objects, designating people in the out-group as dumping grounds for the majority’s animus and need to mock the less powerful. Demanding that another person “be less sensitive” is rude, yet gets treated as polite. And because already-privileged people who make such demands get rewarded further (beginning with social acceptance), there’s little incentive for them to practice empathy.

The egocentrism I’m talking about isn’t just about dynamics between men and women. For example, Linus Torvalds’ public verbal abuse of Linux kernel contributors is an example of how open source culture also tolerates abuse directed by men at other men. (Sometimes it doesn’t just tolerate it, but even encourages it, as when bystanders comment “well, assholes get things done.”) Social hierarchies and displays of dominance are certainly alive and well in how men interact with each other; and because hackers often define themselves as beings of pure rationality and logic, they rationalize these hierarchies as “necessary” for “getting things done”. (I think we could also “get things done” if we recognized and accepted that as humans, we frequently act for emotional rather than purely “logical” reasons — and maybe even if we accepted that the dichotomy between emotion and reason is a false one.) That, however, does not mean that verbal abuse between men is just as intense for the recipient as verbal abuse directed at women by men. It doesn’t mean that verbal abuse between men gets excused as easily as abuse directed as women. And it doesn’t mean that there as just as many opportunities for a man to exploit another man’s vulnerabilities as for men to put women in their place. It could hardly be otherwise, given the wealth of experiences that women bring to interactions with men, of internalized messages that (even for those women who have worked hard to unlearn them) tell them that they deserve whatever abuse they get, that they really had it coming. It’s not that abuse is ever acceptable when directed at anyone of any gender. Rather, it’s that being punched in the face feels more intense than being tapped on the shoulder.

Ultimately, we have to ask whether the freedom to abuse people is one of the freedoms we value. Richard Stallman himself identified four freedoms when it comes to software: “the freedom to run the software, for any purpose”; “The freedom to study how the program works, and change it so it does your computing as you wish”; “The freedom to redistribute copies so you can help your neighbor”; “The freedom to distribute copies of your modified versions to others. By doing this you can give the whole community a chance to benefit from your changes.” (He notes that for the second and fourth freedoms, access to the source code is a prerequisite.) The freedom to be an asshole does not appear on this list. Rather, these values point to inclusivity (the freedom to run the software, as in: to be included in the community of people who get to use it) and altruism (helping your neighbor; helping the community by distributing a better version). (Perhaps the inclusivity part is a bit of a stretch — the freedom to participate does not explicitly appear, which may say something about what Stallman took for granted.)

Decades before, Franklin D. Roosevelt spoke about another set of four freedoms: freedom of speech and expression, freedom of worship, freedom from want, and freedom from fear. How often do you hear stereotypically privileged open-source guys talk about freedom from fear? As I’ve discussed, much of the dialogue that happens when hacker culture talks about diversity and inclusion is about laughing off the idea that anyone else’s fears might be reasonable. Likewise, techno-libertarianism has very little room for a discussion about freedom from want. There isn’t much time and space in hacker culture for freedom of worship, either — especially when you take a broad view of what “freedom of worship” means and interpret as freedom to believe in things that can’t be proven with logical rules from empirical facts (like the dignity and worth of each human being), without being punished for it through ostracism or in any other way.

In either case, “freedom to treat other people as if they don’t have feelings, or as if their feelings don’t matter” is not on the list. (Thanks to Leigh Honeywell for pointing out Stallman’s and Roosevelt’s four freedoms, and the parallels between them, to me.)

Diversity as Devaluation

I want to ask a question outright that Reagle at best hints at: Is the very nature of open-source, its fundamental ideologies and values, inherently bound up with the insulation of oneself from the collaborative social project of making progress towards equality?

Maybe the whole system by which people produce free and open-source software is designed to provide the same sort of cozy lifestyle that one can find by being a programmer writing proprietary software, but without all those nagging structures of accountability that one finds in the corporate world. Like policies against harassment and discrimination. It’s true that companies adopt those to protect themselves against lawsuits, not to be morally correct, but they do protect people. And open source is a world without that protection. Maybe comparing open-source and corporate proprietary software is the best experiment one can do to determine what measures attract or repel participation by women. We know that open-source projects have an even more lopsided gender balance, as a general rule, than proprietary projects mainly composed of people being paid by a corporation to work on them. Can that really be a coincidence?

In a community with no formal governing structures, it’s far easier for people to take advantage of whatever privilege and power they inherit from the underlying society. One form this power takes on is that of speech acts that dehumanize and objectify people, and appeals to “freedom of speech” to immunize the speaker from the consequences of their speech.

I think that the desire to make boob jokes with impunity is not the only reason why male open-source programmers would want to keep women out, though. After all, the sexist jokes and comments that tend to engage the “free speech” defense the most are rarely funny or interesting. I think sexist jokes and comments are actually a means to an end, not an end in themselves. We know that male-dominated professions tend to be more socially prestigious and more highly paid than female-dominated or even gender-balanced professions. This can’t be an accident; men’s social over-valuation and their disproportionate participation in work that people think of as important form a positive feedback loop. For example, consider doctors and nurses: no doubt, women originally got tracked into nursing since medicine wasn’t considered an appropriate profession for a woman (gotta keep that power out of the hands of women). But even now that women have been allowed to study medicine for quite some time, nursing continues to be a lower-paid and less-praised profession, in large part (as far as I can tell) due to the significant presence of women in it.

The thing about prestige-as-male-domination is that it’s fractal. For example, within medicine, it’s common knowledge that primary care providers are likely to be women, while doctors who work in the most prestigious and highly compensated specialties (e.g. neurosurgery) are more likely to be men. Likewise, within computer science and software engineering, both of which are male-dominated as a rule, it’s harder for women to gain entry into some fields than others. Anecdotally, those fields (within academic CS) are theoretical computer science, programming languages, and operating systems. Among non-academic programmers, open-source programming (especially systems programming) occupies the role that theory, PL, and systems do within academe: looked up to and highly valued. By contrast, self-styled expert programmers tend to disdain jobs in areas like Web development and quality assurance — that’s “women’s work”, to the extent that any software jobs are. Technical writing, as an occupation, is even more looked down on and even more open to women. Perhaps that devaluation is part of a more general distaste among programmers for documentation, which could allow outsiders to glean the in-group’s secrets. Writing documentation is also a form of teaching, which is also a traditionally female-coded profession, and also a profession that’s frequently looked down on. So that’s why it’s so important for men in the high-status subdisciplines to maintain their status by making sure women don’t enter and devalue their field. Keeping women out means keeping salaries high.

(Statistics backing up what I just claimed about medicine — at least for the US — are available from the Association of American Medical Colleges (PDF link): see table 3 on page 13, “Number and Percentage of Active Physicians by Sex and Specialty, 2007”. The only specialty that’s majority-women is pediatrics; cardiovascular disease, neurological surgery, orthopedic surgery, and a few other specialties are over 90% men. I don’t know of any similar reports about gender distribution (and salary distribution) within different areas of the software industry, so I don’t claim to be speaking any more than informally, based on what I’ve heard over the years.)

“It’s amazing the things women did to advance computing before it advanced to the point that we learned women don’t like computing.”
— Garann Means

Before computers were machines, computers were women. Most of us know that part of the story. What I know less about, personally, was the specifics of the process by which men drove women out of the profession of computing as it, well, professionalized. I can guess that white middle-class dudes saw an easy desk job that potentially would pay well, and the rest is history. Without evidence (at least not any that I have handy right now), I claim that none of this was an accident. Expelling women from computing was essential to the historical process of the professionalization of software and hardware engineering. (I know that that’s roughly how it went down with the profession of medicine, as documented by Kristin Luker in her book Abortion and the Politics of Motherhood: as “scientific” medicine arose, mostly-male doctors needed a way to push mostly-female midwives off the scene, and one of the ways they did that was by inventing the supposed immorality of abortion as a wedge issue.) For many men, a job just doesn’t have as much value if it’s a job that many women do too. And numbers don’t lie: jobs in male-dominated professions literally do have more financial value than jobs in more equal or female-dominated professions.

So arguably, open source is not just a different way to produce software. It’s also an experiment in building an alternative economy for status and peer review. At the same time as for-profit companies began to look harder at how to diversify themselves, how to create policies that would protect workers from sexual harassment and various forms of discrimination, the open source movement gained more and more momentum as a way to recreate all of the good bits of being a software engineer in industry (high social status, freedom, and money) without those annoying parts like human resources departments, processes, accountability, and rules (mostly rules to protect less-powerful members of the community). I don’t think that’s a coincidence.

There’s one misinterpretation of this section that I’d like to head off before it starts. I’m not suggesting that some nefarious group of patriarchs got together, had a meeting about how to exclude women, and disseminated the memo in a lockstep, hierarchical fashion. That’s not how it works. There is no intelligent designer or invisible hand that makes sexist decisions — rather, sexism is an emergent and self-reinforcing pattern that arises from the choices of many individuals. Just as organisms in nature behave in predictable ways without there being any central evolution planning committee, people who study societies have observed that groups of humans often act out predictable patterns too. Of course, sociology and anthropology have different methods and different standards of evidence than biology and physics do, but the social sciences are the only tool we have for rigorously analyzing how groups of people operate. It would be silly and anti-intellectual to discard these disciplines in favor of nothing just because they aren’t like physics.

Finally, a note if you’re asking “where does the money come from in open source?”: more than a few businesses pay engineers (often quite well) to work on open-source software for either part or all of their working hours. (I work for one of them!) In addition, open-source work is frequently a gateway to lucrative jobs and to the kind of social connections that make it possible to found startups. “Free as in freedom” doesn’t mean people work for free, and seemingly more often than not, they do anything but.

Conclusions

Reagle ends his meticulously researched piece with a conclusion that appeals to me as an intersectional feminist: he says that to achieve the goals of openness of diversity, we can’t just focus on openness and diversity as goals (any more, I might add, than an individual can live a happy life by resolving to strive for happiness); we can’t make things better by focusing on a single axis. Just as severe gender imbalances are a symptom of a broken community, addressing root causes will increase diversity as a side effect. But we can’t ignore gender (or race, class, sexuality, or ability), either. Responding to Kat Walsh’s writing about Wikipedia, he says, “the language of being ‘more open and diverse in general’ is problematic. Seemingly, there is no ‘in general’ yet when it comes to notions such as ‘geekiness’, ‘openness’ and ‘freedom'”. I agree — during last year’s code of conduct discussions at Mozilla, some people protested the idea of what they saw as a bureaucratic document codifying standards of behavior with “Can’t we just all be nice to each other?” But being nice, as many people construe it, includes subtly undermining the value and place in society of women and people experiencing a variety of other intersecting oppressions. Likewise, the concepts of “geekiness”, “openness”, and “freedom” will not magically lose their gendered connotations — we have to actively work at it. We can’t build a world where gender doesn’t matter by pretending we’re already there.

Hacker culture is a personal topic for me, so my own conclusions can only be personal. When I was 16, I saw geek culture as something I had to become a part of because I didn’t know any other way to be the person I needed to be. Now that I’m 32, I’m increasingly afraid that it’s something I have to leave in order to be the person I need to be. I know now what I didn’t know when I was 16: that I can be free from constant misgendering, no matter what job I do. I also know what I didn’t know then: I need to be somebody who is kind, patient, willing to admit he’s wrong, and able to make space for other people to join in. I’m not sure if that’s compatible with being in the open-source community, while also having self-respect, dignity, and a place at the table.

Where this is more than just my personal dilemma, though, is that once, I wanted passionately to write open-source code, and now it’s a struggle for me to keep going; not because the nature of the work has changed (on the contrary, it’s gotten more fun as my understanding of it has deepened and my confidence has grown), but because either the culture has changed or I’ve become more aware of its shortcomings (or both). Wouldn’t you want to know about it if you were driving away potential contributors — or forcing them into impossible trade-offs? I don’t think anyone should have to choose between doing good work they love and feeling valued and respected as a human being.

Thanks to GF contributors Leigh, Skud, Sumana and Shiny; as well as Graydon Hoare, for their comments. Thanks to Debra “Teacake” for linking me to the statistics on gender distribution in medical specialties.

ETA Wed. Feb. 6th: Joseph Reagle posted a response to the responses, which is also worth reading.