Gender and Open Source

What follows is one of the articles I wrote for The Encyclopedia of Gender and Information Technology. It was accepted pending revision, but I decided to withdraw it. The deadline was just about to pass, and I suppose I could have asked for an extension, but the changes the reviewer requested ended up requiring more research on the history of women in computing than I have time to do, unfortunately (must spend time on dissertation only!). Plus, this is much more of a nonacademic position paper than an objective, informative academic encyclopedia article. I wrote two other articles for the encyclopedia anyway, which is probably enough. So, enjoy. Maybe you'll learn something you didn't know before; I hope so. Not being a software developer myself, I'm sure I'm wrong about some of the technical matters I discuss here, and I'd appreciate any corrections.

UPDATE: See also the playlist I created on this topic. It contains a few sources I found since writing this.

UPDATE: There's a thread at Linux Weekly News about an article in NewsForge titled "Getting in touch with the feminine side of open source."

UPDATE: Máirín has a thoughtful response you should read.

Introduction: Proprietary and Open Source Software

The software most people use every day – common applications like Microsoft® Word®, Adobe® PhotoShop®, and the like – is proprietary. That is to say, we pay for the use of it, and suppliers pay programmers to write the source code. Consumers purchase the software, install it, and use it. While consumers can use the software applications, because they are only sold the executable files, they cannot view or alter the source code. This means they cannot write new modules or tinker with existing features; for example, consumers can't make changes to a proprietary word processing program's bulleted list feature to add a new style of bullet or write a module that would make the program recognize a given file format it does not currently recognize. In proprietary software applications, the code is under traditional copyright, and because the corporations only sell the executable program file(s), the user cannot view or build upon the source code. Instead, the user must wait for the next version of the program, hope the company has added the features she wants, and pay for the upgrade. Programmers who work for software corporations must sign confidentiality documents agreeing not to share the source code with any unauthorized personnel.

Moreover, as Richard Stallman reasoned in the early 1980s, the wealth generated by the sale of software would quickly end up in the possession of a small minority unless the source code – the knowledge undergirding the product – were freely available (DiBona, Ochman, & Stone, 1999). In 1984, he started the GNU project, the goal of which was to create a free-of-charge and manipulable operating system. Other such projects soon followed, and the open source movement began. Free and open source software (FOSS) is software whose source code is made available under the General Public License (GPL), which is an alternative to traditional copyright. Open source software is often, but not always, available at no cost, but its code is always available for study and modification. Obviously, not everyone has the technical expertise to add to or change source code, but the free availability of the code – the knowledge – has significant implications for global social change, free and open education, and acquisition of wealth. Families and schools that lack the financial resources to buy Windows and Microsoft Office can download Linux and OpenOffice free of charge. Students in developing countries who want to learn programming can study and experiment with the code of Mozilla and WordPress. Badimo and Muthoni (2005) of LinuxChix Africa explain clearly and forcefully women's stake in the issue:

With FOSS, countries will no longer have to prioritise between poverty and the digital divide. Also since women are the ones mostly affected by poverty and HIV/AIDS, it is relevant that they be properly tooled and positioned to make that difference in their lives. Furthermore, ICT is still male dominated, moreso the Open Source technical environment, so Linuxchix Africa will play a role as a catalyst that will demystify FOSS to the people who stand to benefit the most from it.

FOSS furthers goals that could be construed as feminist, broadly defined. Access to the code is available for anyone who learns programming languages to write modules for open source applications. Anyone can release new versions of the applications under the GPL. Getting a job at a software corporation and signing confidentiality agreements not to give away company secrets is not necessary. Open source advocates argue that keeping source code closed and proprietary, in economic terms, creates a false scarcity, whereas open source is a gift economy based on cooperative peer production.

Gender and the Open Source Development Model

The practice of keeping source code open not only gives more people access to software for downloading, it also speeds up innovation. Raymond (1998) claims that “too often software developers spend their days grinding away for pay at programs they neither need nor love.” In contrast, the open source development model is a highly motivated, passionate, and collaborative process in which the principles are to provide early releases of programs and release often. Unlike proprietary software corporations that employ a relatively small number of developers, FOSS development communities are open, allowing far more people to participate, which results in more brainstorming and ideas. Developers who write new modules offer them to the community for testing and discussion. Users, who are considered co-developers in the open source model, offer feedback, and the developers use it to improve future versions of the software (Raymond, 1998). Tasks such as programming, documentation, and testing are distributed among members of loosely formed communities of volunteers.

The open source development model, while in many ways more open, egalitarian and democratic than proprietary software development, still has some of the problems encountered in the field of computing in general. For the purpose of this essay, I will examine the underrepresentation of women in open source development communities and the problems women face in them. Despite fewer barriers to entry (one need not have a degree in computer science or IT or a job in software development to participate in an open source development community), open source is still male-dominated, and women are still being driven out of these communities, most often by the unconscious sexism of well-intentioned men. The remainder of this essay will be devoted to describing two cases that illustrate the problems women face in these communities: women in Linux and the experiences of female hackers Raven Alder and Starla Pureheart.

Women and Men in Linux Development Communities

Before describing interactions between men and women in Linux development communities, it is necessary to clarify the meaning of the term “hacker.” Unlike the popular culture stereotype, the vast majority of hackers are not computer criminals. Wikipedia's (2005) definition sets forth three often overlapping uses of the term “hacker”: the “intruder and criminal,” the “brilliant programmer,” and the “security expert.” While not all network security hackers are involved in programming and not all programmer hackers are involved in network security, many hackers are interested in both programming and security. Eric Raymond's commonly-cited Jargon File definition of hacker also foregrounds programming expertise (2004). Alder (2004) shows the security/programming overlap:

I am a hacker. I run OpenBSD, FreeBSD, and five flavors of Linux. I have done custom kernel work on almost every box I own. I write documentation for Linux. I watch the intrusion detection systems for the Nessus servers. [. . .] I have found zero-day vulnerabilities in major security products. I code my own exploits and I run pen-tests for fun. I am a hacker.”

In this essay, the term “hacker” will be used to represent a programming expert in open source development.

Linux is perhaps the most successful example of an open source development model and product. There are Linux development communities all over the world, large conferences to discuss its development, and a magazine, Linux World, is devoted to Linux-related news. Linux communities are not as successful, however, at recruiting and keeping women. “Sexism is alive and well, and it is driving women out of Linux,” Val Henson argues. Henson's widely distributed essay “HOWTO support women in Linux” provides insight into the problems women experience in Linux development communities. Henson is an operating systems programmer for Sun Microsystems and a longtime Linux developer (Henson & Vesperman, 2004). Henson offers recommendations to men who would like to see more women get involved with Linux development (and use, which in open source is co-development). Henson argues that, first of all, “[w]omen severely underestimate their abilities in many areas, but especially with respect to computers,” and this lack of confidence can lead many women to feel intimidated by complex and unfamiliar programming languages and software applications. She claims that the injunction for girls to be quiet and ladylike and “avoid self-promotion,” to the extent that the woman internalizes the injunction, becomes “a handicap later on in life, when being loud and insistent is the only way to get your opinion heard – for example, on the linux-kernel mailing list.”

Women's lack of confidence can be attributed partially to gender socialization, which encourages modesty and social bonding and discourages competitiveness in girls. The same socially reproduced and perpetuated misconceptions about women and about computing that compromise women's ability and willingness to play active roles in computing in general also become a hindrance in open source development. Henson (2002) reviews several of these qualities, including the common misconception that computing is a solitary activity, predominately male representations of computing expertise in popular culture, e.g. television and movies, women's reluctance to pursue computing in a single-minded, obsessive manner (women desire a “life-work balance”), and the objectification of women in computer gaming.

Henson cites as particularly important the rarity of visible female role models in computing in general and in Linux in particular. She points out that “men usually tend to mentor and become friends with other men,” and “[w]hile women have male friends and mentors, it's often harder and more difficult for women to find a community and then to fit in with it” (Henson, 2002). What results, Henson claims, is a “feedback loop” in which women leave IT in higher numbers and enter it in lower numbers. Underrepresentation breeds underrepresentation. Similarly, when women are overrepresented in information technology or represented in equal numbers, far more women are drawn to it. Linux developer Pia Smith (2004) points out that Malaysia, for example, has “a huge proportion of women in IT, across the board,” and that an effort is being made to recruit men to IT.

Henson observes a perception that “Linux development is more competitive and fierce than most areas of programming” (2002). While those who contribute code might earn the respect and approval of the group, they are also at times ignored or flamed, and Henson points out that this environment is hostile to women, whose confidence in their abilities is already lower than men's, and whose socialization trains them to be nonconfrontational. In addition, Linux developers sometimesmake sexist jokes on email discussion lists. Henson writes, “Every time a woman sees a sexist joke or comment, she feels angry, left out, and belittled. Every time a woman sees a man stand up against this behavior, she feels included and valued.” Henson advises men to intervene when such jokes or comments are made, because when women protest the jokes, they are often dismissed as “too sensitive” or labeled as stereotypically angry feminists.

One example of these kinds of comments is found in Eric Raymond's essay “Sex Tips for Geeks” (2001). Directed mainly toward men, he airily adds that “[f]emale hackers can find most of what they need to know in the pages of teens and womens' magazines. Don't scorn that stuff; it works.” Another case that illustrates these jokes and comments is found in an email posted to a listserv addressed to hacker Starla Pureheart:

Dear Starla,

At Bugtraq Security Systems we pride ourselfs [sic] in having a keen eye for
young and upcoming talent. Specifically, core members of our Research and
Development team have expressed continously growing feelings of affection
towards your person. Or, as one of our researchers put it: "hehehe I'd
like to research and develop that _all_ night long :D:D:D". What it boils
down to is that we at BSS (Bugtraq Security Systems) would like to
offer you.. Starla Pureheart.. prolific warrior princess of the great Nuke Wars
of 1997... a position at Bugtraq Security Systems. Preferably on your
knees.

Love,

Bugtraq Security Systems [a pseudonym for a group of hackers]

PS: All kidding aside, get in touch with us Anna - we need a token femme

Hacker Raven Alder spoke out against this kind of behavior on her web site, explaining that she too had experienced propositions and jokes in hacker communities and later wrote that she got “hundreds” of messages from female hackers who claimed that they experienced similar treatment.

It is important to note that there are many men in Linux and open source in general who not only do not engage in sexist behavior but make active attempts to eradicate it. One “clueful male geek” urges female hackers to:

“Keep up the struggle. The best thing you can do is what you're already doing: lead by example. It takes a lot of time, but eventually you'll succeed in removing the male-only stigma from this profession. I think a large part of the problem is that female hackers tend to spend their time actually hacking, not crowing about it on the net. Because of this, there is a disproportionate representation of male voices” (qtd. in Alder, 2004)

These men want to do what they can to get more women into Linux development, but they do not always know how to recruit women or keep them from leaving. Henson (2002) recommends that men who want to see more women involved with Linux treat women basically like they would treat anyone else in the community. This especially means not making sexual advances, not doing too much for the women so that they don't learn on their own, not treating women as “tokens,” complimenting women on their technical abilities, and not criticizing so much that the women are discouraged. She reminds men that “absolutely no one was born knowing how to compile a kernel” and “at one point, you didn't know anything about Linux, either.” Henson also recommends that Linux developers make it a point to invite women to speak at Linux user group meetings and conferences.

Some women in Linux prefer to participate in Linux communities that are overtly women-friendly such as LinuxChix, which was founded by Deb Richardson in 1999 and taken over by Jenn Vesperman in 2001. Although it isn't women-only, LinuxChix is female-dominated with a focus on supporting women who want to use Linux and contribute to the Linux project. Community is the goal, and to achieve that, Henson and Vesperman delegate tasks in order to give many members of the community a sense of ownership of the work, and they make it a point to give credit to and reward hardworking contributors and to parcel out the work equally so that volunteers don't get burned out from too much work (Henson & Vesperman, 2004). LinuxChix is not really that different from any other open source development community, except that women are better represented, and it is a guaranteed safe space for women; e.g. sexist jokes or comments are not tolerated. Some feminists might argue that despite men's being allowed to participate, LinuxChix is a somewhat separatist organization and that it ultimately reifies the duality and hierarchy of gender, but in practice, many women benefit from participating in such communities.

Conclusion: Future Research on Gender and Open Source

Right now, little research is being done on gender in open source communities in Gender Studies and Computer-Mediated Communication. Such research is difficult to undertake given its unique challenges. It is harder to do research on distributed communities like those formed in open source development. Small case studies of specific discussion lists or participant observation of local LUGs can be done, but it would be more costly to travel to observe other LUGs or to attend Linux, Mozilla, or Drupal conferences, especially when it is unclear whether or not the number of women in attendance accurately represents the number of women in open source development. In addition, women's experiences may vary depending on the software tool; women in Linux communities might have different experiences from women in Drupal or WordPress communities. Despite its challenges, though, research on gender in open source is needed, especially because open source is spreading globally, and women can benefit from participating in its development.

References

Alder, R. (2004). feminist poet crypto geek @ raven.oneeyedcrow.net. Retrieved April 4, 2005, from http://www.oneeyedcrow.net/securitygeekfemme.html

Alder, R. (2004). feminist poet crypto geek @ raven.oneeyedcrow.net. Retrieved April 4, 2005, from http://www.oneeyedcrow.net/bss-log.html

Badimo, A., & Muthoni, D. (2005, Jan. 26). LinuxChix Africa press release No. 1/2005. Retrieved April 27, 2005, from http://www.africalinuxchix.org/?q=node/22

Bugtraq Security Systems. (2004, Mar. 15). Request: Starla Pureheart PE. Retrieved April 27, 2005, from http://seclists.org/lists/fulldisclosure/2004/Mar/0613.html

Burke, S. (2003). Feminism and the authorial subject. In S. Burke, (Ed.) Authorship: From Plato to the postmodern. Edinburgh: Edinburgh University Press.

DiBona, C., Ockman, S., & Stone, M. (1999). Introduction. In DiBona, C., Ockman, S., & Stone, M., ed. Open sources: Voices from the open source revolution. Retrieved May 1, 2005, from http://www.oreilly.com/catalog/opensources/book/intro.html

Free Software Foundation. (2005). Frequently asked questions about the GNU GPL. Retrieved May 1, 2005, from http://www.fsf.org/licensing/licenses/gpl-faq.html

Henson, V. (2002). HOWTO encourage women in Linux. Retrieved May 1, 2005, from http://infohost.nmt.edu/~val/howto.html

Henson, V., & Vesperman, J. (2004). Building and maintaining an international volunteer Linux community. Retrieved April 4, 2004, from http://infohost.nmt.edu/~val/uselinux_slides.pdf

Laboratory for Computer Science and the Artificial Intelligence Laboratory at MIT. (1983). Barriers to Equality in Academia: Women in Computer Science at MIT. Retrieved May 2, 2005, from http://www.cs.washington.edu/homes/lazowska/mit/

LinuxChix (2005). Retrieved April 29, 2005, from http://www.linuxchix.org/

LinuxOnline (n.d.). Interview with Deb Richardson, founder of LinuxChix. Retrieved April 29, 2005, from http://www.linux.org/people/richardson.html

O'Reilly. (1999). Open sources: Voices from the open source revolution. Retrieved May 1, 2005, from http://www.oreilly.com/catalog/opensources/book/toc.html

Perens, B. (1999). The open source definition. In Open sources: Voices from the open source revolution. Retrieved May 1, 2005, from http://www.oreilly.com/catalog/opensources/book/perens.html

Raymond, E. (2001, Apr. 3). Sex tips for geeks. Retrieved May 1, 2005, from http://www.catb.org/~esr/writings/sextips/intro.html

Raymond, E. (1998). The cathedral and the bazaar. First Monday, 3. Retrieved May 2, 2005, from http://www.firstmonday.org/issues/issue3_3/raymond/index.html

Raymond, E. (2003). Jargon File Resources. Retrieved May 2, 2005, from http://www.catb.org/~esr/jargon/

Smith, P. (2004, July 28). Women in open source. Retrieved April 29, 2005, from
http://lists.debian.org/debian-women/2004/07/msg00310.html

Spertus, E. (1991). Why are there so few female computer scientists? Retrieved May 2, 2005, from http://people.mills.edu/spertus/Gender/pap/pap.html

Wagner, R.P. (1999). The medium is the mistake: The law of software for the first amendment. Stanford Law Review, 51, pp. 387-408.

Wikipedia. (2005). Hacker. Retrieved May 2, 2005, from http://en.wikipedia.org/wiki/Hacker

Wikipedia. (2005). Linux. Retrieved May 2, 2005, from http://en.wikipedia.org/wiki/Linux

Terms and Definitions

Free and Open Source Software (FOSS): Software whose source code is made available to anyone who wants to study and modify it. FOSS is often, but not always, available free of charge.

GNU: Stands for “GNUs Not Unix.” GNU was developed by Richard Stallman and intended to be a free operating system.

GPL: Stands for General Public License. Written especially for the GNU in 198? by Richard Stallman, the GPL is intended to be an alternative to traditional copyright. It allows authors of software code (or any other kind of content) to keep the copyright to the version of the software they write, but the terms of the license give users permission to copy, distribute, and modify the content, for commercial or noncommercial purposes, provided they also license their derivative works under the GPL. The GPL has been both criticized as a “viral license” because of this “share alike” clause and lauded because the GPL keeps GPL-licensed content and its derivatives open for the public's use.

Kernel: The code that enables programs to execute.

Linux: Perhaps the most economically successful example of open source software and the open source development model, Linux is a kernel on which many open source operating systems and software distributions are based: Knoppix, RedHat, SUSE, Debian, and more. Linus Torvalds wrote the code for the kernel in 1991 and licensed it under the GPL.

LUG: Stands for “Linux user group.”

Open source development model: A collaborative model based on inclusiveness and constant revision, explained in detail in Eric Raymond's “The Cathedral and the Bazaar.”

Shell: A kind of program that allows users to interact with the kernel.