Observing Communities

Great post from Humans of New York: “I don’t enjoy observing people as much as I used to.”

Creating great communities requires understanding what makes great communities great, and what precipitates failures within communities. Neither of these are settled knowledge. While people can certainly help good communities to emerge, based on a hodgepodge of patterns, cases, and received knowledge, there isn’t a single recipe that always works. As a result, the best community managers are those who know how to do their own research into community dynamics and press our understanding forward.

The idea that social research and applied research are worlds apart doesn’t really hold water. It is true that social research is intended, in many cases, to be generalizable in a way that applied research may not be. That can mean a couple of things.

First, it means that those studying communities are looking for a solution not just for the community they are studying at the moment, but also some rules of thumb that might apply in parallel situations. In other words, they are looking for patterns, in the sense that Christopher Alexander popularized. Alexander’s ideas around “pattern languages” attracted a strong following among architects and urban planners, but have probably gained an even greater toehold among designers of computer interfaces. The idea is simple: a set of rules of thumb for solving common design problems. Those problems arise when multiple values or desires come into conflict. So, for example, the placement of windows in an edifice (and especially a home) can be challenging. How do you make sure that there is an abundance of natural light, but maintain good heating or cooling, as well as the privacy of the residents? A pattern helps the designer to not reinvent the wheel. It provides a case-based rule that is linked to the context in which the pattern is encountered, and serves as a guide to future designers who might encounter something similar.

That is, perhaps, a tactical outcome of applied research: a set of rules of thumb that help guide future design. But the social scientist seeks a more strategic understanding of not just what works but why it works. Especially in business, the “why” doesn’t get a lot of attention. Why do people tend to purchase more when they encounter certain colors in the background?

As long as you make the purchase, the “why” doesn’t really matter to many sellers, but it does to the social scientist. Indeed, in many ways the “what” is only important to the degree that it gets you to the “why.” Why is it that communities work the way they do.

The Great Qual/Quant Divide

In 1959, CP Snow gave a now famous lecture on “The Two Cultures,” and argued that scientists were often pretty illiterate when it came to the humanities, and likewise, humanities scholars often lacked a good grounding in the sciences. Some might suggest that there exists a similar divide in the social sciences between those who prefer qualitative methods and those who prefer quantitative. That may be drawing the distinction a little to starkly. After all, if you ask someone which they are, they are likely to hedge and talk about mixed methods. Or they might say they are approaching the problem they are interested in from a qualitative perspective but hope to include some numbers, or something similar about hoping to bring a little explanation to their statistical analysis. Too often it never gets beyond that hope.

Even worse, in many cases people begin with their methodological stance and then apply it like a hammer to whichever problem they find. In practice, different methods are more or less effective at answering different kinds of problems, and entering with a set of approaches that either are purely about “measurement” or purely about “meaning” is a trap difficult to escape. What’s worse, adherents to both approaches who do not have broad exposure, are likely to accuse the other of being atheoretical.

There isn’t enough space here to really address this question fully, even if we limit the discussion to how it influences the research of communities. Very broadly, however, when people do used mixed methods, often it is to get a better understanding at different scopes and at different depths. One pattern for this kind of research would be to begin by doing an exploratory qualitative analysis of a community, and this might give you a hunch that could then be pursued through a system of measurements. The opposite is also a possibility: metrics might provide you with a relatively small number of either typical or atypical cases that the researcher might be able to investigate in more depth. In either case, the approach taken should be dictated by the problem to solve, rather than by familiarity or comfort with a particular method. Of course, that is far easier to say than it is to put into practice.

Thick Exploration

Even when not using grounded theory, one of the central ideas around ethnographic approaches, and often of other qualitative approaches, is to more fully understand the context in which social structures emerge. Particularly when communities represent such a diversity of lived experience, such explorations, when done well, can open up spaces for much better understanding of how a community works. The degree to which that might be extended to other communities is complicated. For some, the deep description of a single community, or of a set of people within that community, is enough; no general rule that might apply to other communities is necessary.

The term “thick description” was developed by Clifford Geertz to counter forms of description that simply relayed observations, interviews, and other data and failed to interpret this process and provide the tools to contextualize it. For Geertz, it wasn’t enough to simply say what was happening — the anthropologist had to explain why it mattered. By necessity, such a description had to be microscopic, boring down on a small group or a small set of practices to better understand how the people engaged with them used these processes to make meaning.

It is very easy to do bad ethnography — and perhaps even easier still to do bad autoethnography. But there have been some outstanding ethnographers of digital spaces who manage to immerse themselves in cultures they wish to better describe and understand. Hine, Baym, Boellstorff, Beaulieu and a dozen others have provided the kinds of deep, contextualized interpretation of online community, all of which can serve as a model of practice. And a number of ethnographers have not just been reflexive in their practice, but have explicitly described that reflexive process.

That reflexivity often extends to participation in a community. Can you interpret the culture of a community without participating? Yes, though it is harder. Isn’t it difficult to observe a community when you become personally invested in it? Yes, but you can mitigate this by including yourself among your observations. Of course, even when you are observing from the outside your perspective is affected by the interaction with the observed. But particularly when you come to participate in the culture in which you are studying, you become a part of what you need to more fully understand. As Annette Markham has described in moving from interviewer to participant she realized that she was what was missing from her analysis.

From Virtual to Networked Communities

One of the advantages of that exploratory perspective is that you need not worry nearly as much about the boundaries of the object of study. It seems that one of the chief definitional issues for those wishing to study communities is drawing a line around what “counts.” Artificially constraining the borders often leads to a strange reflection of people’s lived experiences online. Even that term — “online” — has gotten in the way of better understanding how communities have evolved over time. We were never just “online” or just “offline,” despite what some early attempts to define “cyberspace” as a separate space. Interactions were always social, whether they were mediated by a foot or two of air or an extensive packet-switched network.

And as much of our sociality has moved from the wild Web 1.0 to the walled gardens of Web 2.0, there is the danger of focusing too heavily on the platform, or some of the structural walls of that platform, as the community. Reddit may refer to its users as members of a community, or a subreddit might do that with a bit more credibility due to scale, but it would be wrong to assume that these very explicit limits have particularly strong social effects. We wouldn’t limit a neighborhood to a strict set of city blocks, or a church community to the actual church buildings. Well, we might. We might stipulate that boundary as a matter of convenience. But when we do, we risk missing what is important to the members of that community.

Network of works produced in the ccMixter community.

The trick is that we have moved to networks-of-choice — that the borders of a given community have less to do with rules or even consensus, and more to do with the separate perspective of each of its constituent members. Limiting an analysis of a community to the online realm, to the most visible interactions within that realm, or to a particular platform or structure within a platform, risks missing what is most important to any given “networked individual,” who might best be described not in terms of a circumscribed social space, but rather by her relationship to other people and machines in her environment.

Once you take that perspective — that of a cloud of more-or-less connected individual networks — a study can suddenly feel overwhelmingly large. But if the people that you are studying don’t feel that way about their interactions, neither should the researcher. The researchers needs to be a particularly aware member of these messing networked communities, but may better understand how they support communities by becoming a meandering presence within them.

If you seek a boundary, that boundary may be time. But artificially restricting an investigation to certain people, certain artifacts, or certain kinds of discourse makes it unlikely you will understand the context in which people interact within these social milieux. Instead, those boundaries should be determined by the energy needed to collect and understand these experiences, and by the careful examination of how you may be affecting and being affected by your interactions with the people in it.

The Cyberflâneur

The flâneur is a figure that has, thanks is some part to Walther Benjamin, taken on greater meaning over time. The direct translation is “loafer” — someone who wanders around, usually, a city, without a particular aim or specific purpose. But thanks to Benjamin’s interpretation (via Baudelaire) of the flâneur as a kind of idealized urbanite, an observant writer or artist, the term has taken on a great deal more meaning. While he may not be doing something traditionally productive, the flâneur watches and understands the heartbeat of the city.

From The Flaneur Society

We might imagine a kind of cyberflâneur; a person who is able to observe without looking, to wander intensely, and to be alone while in the crowd. The experience of the cyberflâneur is in some sense singular, the world is interpreted from her unique perspective. But she also shapes her own observational capabilities, refining them and turning them reflexively on her own context in the online world.

It is also worth noting that a community manager — whether formal or informal — has a particular perspective that differs quite a bit from that of the flâneur. On one hand, a person in a leadership position has a unique set of tools and unique kinds of access (to people and other artifacts) that may not be available to someone who is acting as a participant observer elsewhere in the community. While a leader in the community may be able exploit that access to reveal useful connections, or to manipulate interactions to see what effect they have, they are inevitably blinded by their relational position. Both kinds of researchers are needed, but in many cases the ideal is a non-manager who is able to draw on the access and expertise of the community manager, but also stand apart from her.