Big Data: Can It Revive the OAC and Resuscitate Anthropology?

When I first became a member of the OAC, the posts and discussions here were lively.  What brought me here was Keith's economic anthropology.  His thoughts  on  human economy  were intellectually fresh and exciting. I assume a lot of people became members of the OAC because of Keith's human economy.

I am not saying Human Economy topics are now old and no longer exciting.  They are all over though, and they are already mainstream.  In college, our economic anthropology was focused on primitive, socialist, and capitalist economies. Now my old department is treating and  teaching economy as a whole, a cultural life, and not solely theoretical and statistical but primarily social, cultural, and human.  I like that.  With that, we can entertain ideas outside of primitivism, Marxism, and capitalism.

To revive the OAC, we need another intellectual advocacy as exciting and insightful as Human Economy.  Maybe it can invite and excite more people interested in Anthropology.  Maybe we will have the same excitement and liveliness as before when Human Economy was still fresh and not yet mainstream.  We need something new here at the OAC.

One of my favorite free channels to watch now since I have no more Cable TV is NHK World, a public broadcasting program of Japan for culture and tourism.  For weeks now, they have been showing a documentary about Big Data,  Wikipedia defines Big Data as:

"Big Data is a broad term for data sets so large or complex that traditional data processing applications are inadequate. Challenges include analysis, capture, curation, search, sharing, storage, transfer, visualization, and information privacy.  The term often refers simply to the use of predictive analytics or other certain advanced methods to extract value from data, and seldom to a particular size of data set."

So far, that is a correct definition or description of Big Data, but it is is evolving.  I first heard of Big Data in early 2000, but it was purely quantitative, computational, and statistical and used in physical, environmental, computer, and engineering sciences. A decade later, it became huge in management, marketing, economics, development, but still quantitative and statistical though.

What NHK World has been showing is different.  Big Data is applied in clinical medicine and community health.  One study is about a set of people with asthma in New York.  Their personal and medical data are taken.  Environmental data such as geography and landscape and weather and wind direction are also included.  They also record the daily life and routine of the research participants. I find the last one so ethnographic and anthropological, but not the kind done in exotic places by anthropologists from the West.  The Japanese scientists in the documentary find Big Data useful and important, but the American media reject it as a fad in informatics.  I guess it comes off as such to anti-Wall Street Americans because Big Data is somewhat known now in the US as a business tool primarily for marketing.    

By comparing the data gathered from all participants and relating them to other data like geography and landscape, they find the asthma triggers and the frequency of attacks.  One such trigger is the presence of a horse stable in the area.  They successfully find that out because they record the daily movement and mobility of the participants. Many participants have attacks that are attributable to the horse stable.  Such study is correlational, longitudinal, and qualitative.  Big Data has now evolved into looking for patterns to analyze and predict the rise or onset of something. I can't wait to read that New York asthma research.   

Culture is a set of patterns.  I do believe that because people in a certain culture generally conform and belong, and they live in the same cultural and behavioral patterns.  Those who go against the general patterns are rejected and excluded.  Even in subcultures, there are also conformity and belongingness. Imagine if there is a "Big Data" for all inner cities in America.  Maybe we can see patterns affecting, describing, or predicting crime rate, malnutrition, unemployment, epidemic, etc.

The question about Big Data that interests me is about data gathering--who will gather the qualitative or ethnographic data since most scientists in information technology and data analytics are usually  interested only in the quantitative parts--computational, mathematical, and statistical?  I think anthropologists should do the job if they want to resuscitate anthropology, which is becoming irrelevant in a world that revolves around technical communication and  information sharing.  Do anthropologists still need to go to Timbuktu to describe the culture there  when the people of Timbuktu are already sharing their culture online?  

Anyone can do ethnography, but studying and analyzing ethnographic data should still remain in Anthropology.  Anthropologists must move quick before sociologists claim the  qualitative parts of Big Data.  I think it is time for anthropologists to predict and solve sociocultural problems.  The issue of ISIS and why some Muslims are radicalized, for example, is sociocultural--involving economics, politics, psychology, and religion.  Where are those anthropologists who have strong backgrounds in economic and political anthropology, culture and personality, and anthropology of spirituality and religion?  Why are they silent?  Even the President of the United States has only Bush to blame.  Where are those anthropologists who should advise President Obama that the issue is not all geopolitical and  about foreign policy?        

Going back to the topic, how can Big Data revive the intellectual interest, liveliness, and enthusiasm here at the OAC?  I think we can make our own mini-Big Data here by encouraging anthropologists all over the world to submit or post their ethnographic data or the data produced in their communities.  Imagine if we have different data about the spread of Islam and the movement and mobility of radical Islamic preachers. Maybe we can see patterns of fundamentalism,  radicalization, or even violence.  Imagine also if we have different ethnographic data about rape coming from different parts of the world.  Maybe we can see patterns that are not only related to patriarchal power and gender politics feminists have been focusing on.  Big Data is, indeed, promising in the field of anthropology.  Does anthropological informatics or computational anthropology sound good?

My final question: am I dreaming?  


Views: 193

Reply to This

Replies to This Discussion

Hi M, that's an interesting idea, but you would have to show how Big Data might play a similar role to Human Economy. To take the latter first, I was prominent and continuously active here at the OAC; had started a large programme in that name; people didn't know what it meant, but they liked how it sounded; there was an obvious link to anthropology. As a matter of fact,it is only now that I get invitations to speak about it, when HE published output is getting to be known. So how does Big Data figure as a possible replacement? 1. It needs protagonists with clout and persistence 2. There has to be external reinforcement of the idea in anthropology 3. People don't know what it is, beyond something the internet moguls are into, and the name is hardly seductive 4. It addresses the opposite end of the social spectrum, large abstractions. Do you know any practitioners in this field? I know some, but they are sociologists. I am not trying to protect my old concept. The OAC served as an incubator for it and it has now gone public. I enjoyed reading your projection and am glad to have you here imagining our collective future. Thanks.

"how can Big Data revive the intellectual interest, liveliness, and enthusiasm here at the OAC?"

M., how about doing what Amazon does and tracking who participates in which kinds of discussion? The output would be a message to individuals with shared interests, "People who participated in this discussion also found these threads of interest...." The technical problem would be that the current input is so free-form that tracking interests at an individual level will be pretty hairy. Either a very smart AI or a huge amount of manual coding will be required. One quick and dirty solution might be to only allow new posts which are properly tagged using, let's say for example, three or more tags from a pre-defined list available to the poster in a drop-down menu. This would make it easy to identify posts on similar topics and rank them by degree of overlap, 1, 2.3 ....n. If individuals were willing to allow it, another application might go beyond suggesting other similar posts to introducing individuals with similar interests to each other. . . .

Of course, this sort of technical application of Big Data technique may not be what you have in mind. An entirely different approach would be to review existing theory and examine how Big Data could be used to improve it.... This might lead us to Michael's suggestions re quantum physics, with wave functions to predict the probabilities of individual social actors winding up in one or another specified state....

Just brainstorming. Could you tell us a bit more about what you had in mind?

Keith and John, can I  just respond to these in the seminar to avoid repeating myself in two different threads.  Big Data to me is (w)holistic and general.  Anthropologists are trained early to be generalists.   They should be included in this big data collection.    

Hi M, because of the time lag, John is alseep. But in my view any hope he had of keeping participants to a disciplined reading of Ingold's paper has already gone up in smoke. For you to tell us what you think big data is good for at this stage would imo wreck any sequence he may still have in mind. So maybe later.

Okay.  That's even better.

Those interested in Big Data might want to check out DataCamp. I am currently working my way through the first of the free courses on programming in R. It is very well done, indeed. All you need to understand it is two things.

  1. Arithmetic
  2. Awareness that computers are useful idiots — the code must be written in precisely the form that R requires. The good news is that the program will tell you when you've got it wrong.

R is a programming language developed specifically for doing statistics and creating visual displays of results. It's easy to get started with but very deep once you get beyond the basics. It is free, open source and widely used by data scientists in all sorts of fields.

M, to the extent that what you're calling for is larger, more systematically coordinated anthropological research projects set to generate joint databases, there is some stuff happening. As to my knowledge, where such thinking is most in vogue these days is among cognitive anthros. You might for instance keep an eye on the NSF-funded project "Cultural Models of Nature across Cultures: Space, Causality, and Primary Food Producers" (though still at the data collection stage).

Reply to Discussion



OAC Press



© 2017   Created by Keith Hart.   Powered by

Badges  |  Report an Issue  |  Terms of Service