Yesterday, in the first session of EPIC2013, Martin Ortlieb asked a question that stopped me in my tracks. Plenty of people, he said, think that big data collection is creepy. Every day, as we interact in a computer-saturated world, bits of information about our personal selves are being collected and deployed by marketing companies to sell us stuff and spy on us. But are classic ethnographic methods really any different? Don’t people find it creepy when us anthropologists go poking around, asking questions?
One conference panelist responded that she didn’t think anthropological methods were creepy, because we spend so much time building relationships with people. And, in a sense, she is absolutely right. We don’t (at least in ideal practice) just run into the field, take what we need, and leave. We make friends–we are even adopted into families–and we take what people tell us seriously. They trust us to use their data appropriately and ethically. Clifford Geertz coined the term “thick description” to describe the care we take as ethnographers to get to the heart of what people do and why they do it.
But does this mean that what anthropologists do is not creepy? I have my doubts. Sure, we are trusted, we generally collect data about far fewer people, and it’s often difficult to identify the individual in what we do collect. But, realistically, our research participants have no better idea what we do with the information they give us than do social media users whose data is harvested en masse.
In fact, what happens to ethnogrpahic data may be nothing if not all the more obscure. At least, if a social media giant uses my data, I know that the purpose is to attempt to sell me something. Sometimes the results are hilarious. A popular way to make fun of Facebook’s marketing strategies is to share the most ridiculous product that the site has tried to sell you.
With ethnography, though, participants rarely see any results at all. We generally publish our findings in academic journal articles that are inaccessible in two ways. First, they are often behind pay walls (although this is changing as more people make versions of their papers publicly available).
Second, we write up our work in impenetrable language. By the time we’ve analyzed, theorized, and epistemologized our work, not to mention filled it up with obscure references, it is often entirely unintelligible to anyone but an anthropologist with a good decade or so of training up their sleeve.
Mind you, a move towards writing for the public is taking hold in anthropology. It is also, infamously, a running fight in the tech world, where open source has been a big issue for a long time.
The takeaway point I got out of this discussion is that what makes data creepy is the obfuscation of process. Building good relationships with the people during the collection phase can help, but it doesn’t ensure that data is handled in a way that benefits our participants.
Whether we’re dealing with big data or thick data, we could reduce the creepiness of our craft by putting at least some of what we do in the public realm. Co-creation, sharing sites, and writing for the public are just some ways we can reduce the obfuscation of data processing. And we may just bring more people into the conversation, which itself can have positive spin-offs for knowledge production.