Professor Ananda Mitra is Chair and Professor in the Department of Communication, Wake Forest University, North Carolina, USA. Recently he was at Centre for Culture, Media and Governance, Jamia Millia Islamia, where he gave a lecture and also found time to speak to Arshad Amanullah on various issues related to social media, digital discourse, Big Data, cyber bullying, etc. Excerpts:
ARSHAD AMANULLAH: How do you define NarB [Narrative Bit]?
ANANDA MITRA: Every time somebody is doing a digital discourse, for example a Tweet, which is a tiny discourse, to Facebook or social media status update is slightly larger discourse, than to a blog or even a webpage; each one of these things when done by an institution or by individual, it tells a little story, and if you can collect those stories together, then you could have a good sense of life story or the story of the institution or that individual. Before digital became easily accessible to so many people, most of these were institutional stories. For example, if there was an event that happened, good or bad; a news worthy event. You would hear it as a story, it is still called a news story. But that news story would come from an institution; let be in old days, in India could be Doordarshan or Times TV, The Times of India, etc., so these are all the institutions which are telling the story.
What has changed and where I became interested is individuals; [they] can be the agent of the narrative. So, you could have watched the same thing, you could have put the status update on social media or you could have twitted on it. Those are tiny narratives or bit of narratives, that’s why the word Narrative Bits. That I have abbreviated as NarB. The objective is to collect these NarBs; systematically collect these NarBs; systematically analyse the NarBs and then get a sense of the individual narrative, or the narrative of the collective or institutional narrative. Those narratives can give you some predictive power. So you can decide, okay!
At individual level it is called profiling. There is negative part to all of this, because what you are saying is, you are losing privacy. So that’s the idea of NarB. It is based theoretically on the work of Walter Fischer. Walter Fischer has made the argument long time ago, (about 25 years ago) that, “Human Beings are narrative beings”. We are not only rational being or aesthetic being, we are narrative being. We make sense of the world with narrative and we express ourselves through narratives and if we can understand those narratives well, then we might understand what makes human beings who they are!
ARSHAD : What are the conceptual tools you are using to understand this? You have spent enough time on it, your book is ready in the press?
ANANDA : It has already come out in August of 2014. It is called “Digital DNA: Managing your NarBs”.
ARSHAD : What are your conceptual tools; one thinker is Walter Fischer. What you are drawing from Walter Fischer?
ANANDA : I think, I am not necessarily drawing from him. Basically what’s going on NarBs did not matter, till many people started to do digital discourse, digital update. It is only when we encounter Big Data. The idea of Big Data is very unclear. There is large amount of digital data that is available for analysis. The conceptual tool is analytic tool. There is large amount of digital data available for analysis tool, often called Natural Language Processing or Latent Semantic Analysis. What it does is, it takes huge amount of data and does the similar as what we used to do with content analysis in old days that was the part of Media Research. This is the concept which I am applying to large amount of Big Data which is narrative data.
Big Data is broken into two parts that you see in literature all over. One is called ‘Structured Big Data’. Structured Big Data could be demographics, could be measurable thing; that could be quantified. But there is unstructured data also, the unstructured Big Data is the challenge. That challenge is analytic.
So there are two conceptual approaches as in any research process, one is what theory I am drawing upon to make the argument that narrative matters, and that theory is Fischer’s theory, primarily I am also using some of the stuff that people like Stuart Hall have talked. He has made the argument, Identities are often narrated, so identity narratives. I have been working on that idea for quite some time.
- So it’s only that the location of the narrative has changed, the location is no longer in the analogue, the location has now gone off to digital, that’s one major change.
- Two, what has also happened is the amount of narratives have changed, there is much more narrative now than ever used to be. That’s the Big Data component.
- And third thing that has changed, which you said earlier, the actor has changed; the agency has changed. It is now the individual narrators, so when you have that condition, then you have to apply a conceptual analytic tool, that is the semantic analysis.
Changing Location of Identity Formation
ARSHAD : As you brought in the idea that the location of identity formation or the identity narratives have changed from analogue to digital platform. It provides me an opportunity to ask you, which developments in technological sphere made it possible to translate the production of NarBs in the realms of possibility.
ANANDA : I am going to talk little bit with western scenario, which is where I am located and doing most of my research. It is changing across the entire world. There have been two major technological changes that have made all this possible. The whole question of NarBs becomes important with two changes.
- One is the speed of access to the delivery system which is the internet, has gone up rapidly. So if you consider America for a minute; many families and homes have broadband that has been a huge change, it almost becomes natural that if you are in an institution or university you automatically have broadband access.
- Second major technological change is the fact that the device with which you are accessing has also changed. As earlier one, you have to be in front of you computer system at your desk to be narrating, setting up a web page, to be blogging. Today you can do that from your hand held device and that is the case all over the world. Anybody with access to smart phone and some kind of internet can create NarB.
So, those two technological changes suddenly made it possible for the individual actor to gain voice. Now the voice is available. It is now possible to analyse.
- And the third technological change that allows for this kind of analysis and work is increasing leaps in the language processing or machine processing of language or making sense of language. Because like I said, if you are doing content analysis of media; you would read, transcribe a television show, you get a sample of the show, code it and people would read it and write their analysis. Seven hundred million people are on one social media data site. The volume of data is amazingly large; it is not possible to do it by humans any more. So the third thing has been the better language processing tools.
ARSHAD : Is there any software which is available to conduct these kind of analysis?
ANANDA : Absolutely, there are software; three kind of software you get for language processing, the analytics part of it.
- One is what we called as propriety software for example a university or research group might have developed software which allow to by their group or collaborator, there are many of those from different institutions.
- The second is teams, which are doing this will actually develop their own software, they would have sufficient number of computer scientist and so forth, who can actually develop software themselves and those would be closed boxes. If you take Major Corporation like IBM, Facebook, Google they have their own software and you will never have an access to that. That’s their propriety software.
- And the third kind which I am using, which is sold off the shelves, that you can go on the internet and buy the license for! So there are choices much like you having choice between Photoshop and something equivalent. I am using that to demonstrate, that you might never need the first of two kinds. There is now possibility of going online and I am going to pay the money, (it is little bit expensive) to get the software. That’s I am going to use.
ARSHAD : Would you like to give some examples?
ANANDA : The one that I am using is made by IBM, called IBM SPSS Text Analytics or Text Analysis Tool. I don’t remember the other names but that’s the one I am finding to be most efficient for the kind of questions I am most interested in.