oreilly.comSafari Books Online.Conferences.


Why Do People Write Free Documentation? Results of a Survey
Pages: 1, 2, 3, 4, 5

Do Online Help Forums Constitute Communities?

Community is an overused word, and claims to build online communities have been challenged by skeptics on many levels. So it's interesting to look at a bit of research on communities and community building.

Jürgen Habermas and Benjamin Barber provide starting points for many such discussions, but their work doesn't map well to the types of forums discussed in this article. For them, community is the culmination of a process rather than a starting point. Habermas uses community to understand truth, and Barber to create a vibrant democracy.

Certainly, the defining features of community in Habermas and Barber (tolerance, good listening, help for new members or those whose skills need strengthening) can teach a lot to people who want better technical forums. Such guidelines for productively managing developers and users online are offered by Karl Fogel in his book Producing Open Source Software: How to Run a Successful Free Software Project, which I edited and O'Reilly published.

More direct insights can be gained from The Great Good Place by Ray Oldenburg. One-third sociology, one-third nostalgia, and one-third exhortation, Oldenburg's popular text documents the gatherings of people who don't have any practical reason to be together but are drawn to bars, bistros, or even public parks by the chance to mingle and share thoughts.

Because gatherers are freed from the hierarchy and strict protocols of work on the one side, and from the judgments and emotional pressures of family life on the other, they can open up in this "third place" and be themselves. This in turn creates a sense of community that spills out of the cafe door into the greater town, thus civilizing it.

Oldenburg's third places are rooted in their physical communities, but online forums can create some of the same conditions. Participants are very aware that they are judged, as in a bar or social hall, by what they offer and how they behave, not by credentials and affiliations. "A place that is a leveler is, by its nature, an inclusive place. It is accessible to the general public and does not set formal criteria of membership and exclusion...Third places, however, serve to expand possibilities, whereas formal associations tend to narrow and restrict them."

Online forums act as third places in this regard. For instance, technical experts who are unemployed--and thus have no status in the work world--know they can be valued on an online forum for their contributions.

The free conversation in third places breeds independence and democratic spirit, to the point that tyrannical governments have regularly tried throughout history to shut down bars and cafes. One doesn't expect Jacobins to arise on a mailing list to discuss Ruby programming, but participants on all these forums are aware of their presence in a neutral zone not controlled by any vendor or authority, and exercise their rights to challenge and criticize accordingly.

Also like Oldenburg's third places, online forums are always active: "Third places that render the best and fullest service and those to which one may go alone at almost any time of the day or evening with assurance that acquaintances will be there." On a forum, you can log in at any hour and find both help and interesting conversation. Like a third place, you can come and go as you please.

One specific observation of Oldenburg is particularly relevant to technical forums: his lament that suburban environments are hostile to youth. The design of American cities around cars, and other social barriers to leaving the house, leave young people without any healthful place to congregate. Ironically, Oldenburg's cure for many problems of society consists of keeping kids on the streets.

I suspect that these all-too-familiar conditions are responsible for the tremendous technological fascination and online creativity shown by children and teenagers, whether it be shooting videos, synthesizing music, or coding up computer programs. For these young people, also, online forums provide the critical outlet for their gregarious nature. Places where they can share their work stimulate their creativity, and the results of the creatity encourage more sharing. Once again, community and mutual aid become reasons for a virtual something to emerge out of nothing. And in the places where people seek information to support their work with computers, such sharing can provide immediate solutions that would not have otherwise existed.

So What Do We Know Now?

Readers can be forgiven for complaining that my survey of the documentation landscape has left them in the fog. Despite receiving 354 responses, many with fascinating and meticulous testimonials, I've established nothing for certain. Where do we go from here?

Is There a Better Sextant for This Survey?

The tentative insights suggested by this survey's results make one yearn for more and bigger surveys. No one would enjoy doing another survey more than I, but it doesn't seem feasible to me.

It's hard to think of a way a survey could bring in more people and more diversity. What seems to get people's attention is an endorsement of the survey by someone they respect, and how could this be done better than by employing the formidable social capital of O'Reilly Media, as I did? To get the more responses would require buy-in from the leading individuals we already contacted and failed to interest.

Similarly, I feel the survey was about the right length, and that loading it up further would cause participation to decline. I would also be reluctant to ask for many personal details. To make the survey longer or deeper, we might have to offer some incentive--and what a perverse way to recruit respondents in a field driven by the willingness to give information for free! Any incentive we would offer risks attracting people whom we don't want and whose input we don't trust.

Finally, a more detailed and incisive survey would call for more professional control in everything from fashioning the questions and recruiting respondents to analyzing the results. I don't know who would be willing to pay for it.

Experiments to Move Forward

In lieu of another survey, I recommend launching some projects to see their impacts on the issues I laid out at the beginning:

Encourage community

Because community building emerged as the dominant motivation for contributing to documentation, let's start with that.

Technical forums already play multiple roles, and are already employed to do a lot of explicit community building: participants share information about conferences and user group meetings, gossip about developers and other project leaders, argue about news events, recruit people for testing and other useful functions, and even crack jokes in conformance with Ray Oldenburg's observations of "third places." All this could be encouraged.

I'd recommend, though, that different types of discussions be clearly identified through well-established techniques such as keywords on subject lines, so that people can approach each type of discussion in the right spirit. Separating topics into different lists, however, would be going in the wrong direction because it segregates the community-building activities.

Asking people to mark their subject lines and separate the threads of discussion would also allow the list to tolerate contentious posters who bring up controverial subjects. You want these people around, but you also want to shut them out of your consciousness sometimes.

While we want a place for debate and criticism, most forum participants agree that it's important to maintain respect, at least when dealing with newbies. Self-moderation usually works well, but at least some veterans of a forum should stay conscious of the need to remind people of their manners. This is a difficult topic because different cultures have widely varying attitudes toward debate and authority, but the worldwide reach of technical lists is one of the features that makes them appealing to contributors.

Offer more bang for the buck

Although forum participants love to help people, they also get tired of answering the same questions repeatedly and will probably contribute more if they feel their contributions go farther.

So when useful summaries exist, such as formal documentation, wikis, or FAQs, encourage their use at every opportunity. If you point a newbie to the wiki where the answer to his or her question can be found--instead of just tossing the answer into your reply--not only do you make it more likely for newbies to use the wiki, you also make it more likely for expert members of the forum to contribute to the wiki. They'll know that their effort will pay off.

Recognize that many people have trouble culling the information they need from sources such as wikis and FAQs. Teach them how to do so. Newbies benefit from training in how to train themselves, and perhaps expert list members can even get training in how to train others.

Allow reputations to shine forth

Although reputation seems to offer a smaller incentive than many people expect in drawing forth contributions, it's potentially powerful. Something as simple as a sidebar showing the most prolific posters can encourage participation. A few people may try to abuse the system by flooding the forum with irrelevant or low-quality postings, but eventually they'll almost always get flamed off the list.

Rating systems are heavyweight solutions that require too much effort for most forums, but one could try more informal reputation-building systems such as regular "recognition days" where people explicitly thank particular people who helped them, or vote for a roster of helpful list members.

The tools available for organizing documentation are still in their infancy. The tools available for organizing people into online communities offer even more potential. Both can be put to use to improve the education of computer users and to expand their sharing of that education.

Thanks to Bill Lebow for his advice on statistics and data display, and to several others who gave advice on this article.

Related Articles

Andy Oram is an editor for O'Reilly Media, specializing in Linux and free software books, and a member of Computer Professionals for Social Responsibility. His web site is

Return to

Sponsored by: