Dr. Noshir Contractor is an alumnus of the batch of 1983, Department of Electrical Engineering. He was awarded the Distinguished Alumni Award on this institute day. He is the Jane S. & William J. Professor of Behavioral Sciences at the Northwestern University, USA. In an interview with Chennai 36, he talks about his unconventional research on behavioral science and organization research and reminisces about his insti days.
How would you describe your time at IIT Madras? Are there any fond memories you would like to share?
Well, it was quite a few years ago. I graduated from here in 1983, so that’s 35 years ago. I arrived on the campus 40 years ago, 4 decades since I got here in 1978. At the time the Btech programme was 5 years. Some of the fondest memories I have of my days here was the fact we had an environment that was conducive towards being creative in every way we could. The faculty as well as students had a lot of freedom and autonomy and gave us a lot of encouragement to be creative in many different ways. And so one of the best memories I have is to be surrounded by the smartest people I ever was surrounded by because of the quality of students and the quality of the faculty here. And I think that it was clearly a very formative stage in my own life in terms of where I wanted to go and what I wanted to do. And the training is just something I am always going to be grateful for.
From Electrical Engineering to Communication and Behavioural Science. What prompted the change? Could you tell us a bit more about your journey post-IITM?
I will say that even before coming to IIT, I wasn’t sure that I wanted to be an engineer. But my father like many Indians said no if you have to get a degree then you should try and get a degree in engineering or in medicine. Given that I really couldn’t stand the sight of blood, the only option I had was engineering. And back when I was getting ready to graduate from school, my father said, “Oh he doesn’t want to go to engineering school because he won’t get into IIT.” And my mother begged me saying “Please prove your dad wrong. Write the entrance exam, see if you can get in just to prove it to your dad”. I did write the entrance exam, I was able to get in and so then my dad said, “If you have gotten in, you have to go”. And at the time I was in Lucknow. But he insisted I come to IIT Madras rather than go to IIT Kanpur which would have been right down the road from Lucknow. And his reasoning was that he felt that IIT Madras was a very serious place. At the time, the director was P.V Indiresan and he knew of P.V Indiresan from times when my dad was getting post-graduation degree at the Indian Institute of Science in Bangalore and so he felt that people in Madras knew what they were doing and were more serious. So he sent me here. And I have to say that I told him at the time that there was a good chance I may not stay in engineering. He said that it was okay as long as I had an engineering mindset, that’ll be good forever. And he was absolutely right. He is with me here today, with me on this trip, my sister and my dad. I am really grateful that I had this degree because frankly the rest of my life was shaped by the fact that I began to think like an engineer and that happened right here at IIT Madras. So here I had classes on things like communication networks and electrical power in transmission and distribution networks, etc. and that gave me a background. I’ll have to say that once you have taken those kinds of classes, the way that makes you think has analytically shaped my life following that. I ended up getting a Ph.D. in Communication as a social scientist. After that I have stayed sort of in the mix between social sciences and engineering throughout at the University of Illinois Urbana-Champaign, I was very involved with folks at the National Centre for Computing Applications where the Mosaic browser first came out of the University of Illinois Urbana-Champaign, in 1993, 25 years ago. So I have always been involved at the intersection of technologies and communication and how they are shaping societies and how society is shaping them. Today I continue to be interested in that in the context of what is called as web science. Once you had a computer you had computer sciences, so now when we have the web we should think of something called web science, which is looking at how the web is shaping the society but also how the society is shaping the web. One of the things I’m particularly thrilled about is that on this visit, in a couple of hours from now, we are going to officially inaugurate IIT Madras as the first Indian university that has been invited to become a member of the prestigious Web Science Trust Network of Laboratories. I must say that it is with great pride that I was able to, as a member of the board of trustees of the Web Science Trust, to nominate and have my fellow board members approve inviting IIT Madras to be a part of this elite network of laboratories.
With over 250 research papers, you have made a mark in the area of communication and organization research. Could you tell us something about your research?
My overarching interest is to understand the socio-technical factors that explain how networks form and for these factors explain how they perform. I think that one of the things that I am, again, very grateful for is that I have a large number of collaborators from a lot of different disciplines who have helped me able to publish what I have been able to do, etc. I am a big fan of interdisciplinary work because I think some of the problems we face today in society aren’t problems that can be solved by a single discipline. I would say that there is not a single project that I have done that did not involve somebody from another disciple. And most of the time it is engineering disciplines, including computer science of course, but also mechanical engineering, civil engineering, etc. The reason I do that is because many of the grand societal challenges that we face in life today, whether it is in the area of disaster response, global health issues, innovations, etc., and most recently whether it’s in the area of how we are going to successfully send a manned mission to Mars, in all of these areas the questions that we have to address at the end of the day have to do with how we organize. And how we organize is at the end of the day, a social science issue. It benefits a lot from technologies. I like to say that there is a distinction people make between the hard sciences and soft or social sciences. I think the hard sciences are really hard, but I think the social sciences are the harder sciences – not soft sciences. It’s going to take a long time for us to figure out how we can organize. Along the way, the benefits are that technologies are creating novel forms of organizing. The ways in which we can organize today were certainly not true when I was an undergraduate student right here at IIT Madras. Thanks to the web we have peer production networks like Wikipedia, where people come together to work etc., there is lots of interests in flash organizations, for instance, if you look at websites like Kaggle, where people come together to engage in software development. So, the new technologies are giving us new ways of organizing and that means we need to do more research to see how we can be more effective at organizing using these technologies. My vision is to help us not only understand but also enable novel forms of organizing.
Considering its wide varied application, how are computational social sciences likely to evolve in the coming years?
It is very interesting you ask that question because in the next year, 2019, it will be 10 years since we published an article in Science magazine where we introduced the idea of computational social science. So we are coming up to a decade and it has prompted me to think about what have we accomplished in this decade and where do we go from here. I think that, again, Computational Social Science (CSS) is not going to replace other methods used in the social sciences, there’s going to be interviews, survey, ethnographies, other forms of archival research etc. and as well as experiments. But what computational science brings to the table is that we are now able to address challenges at scale because it relies very heavily on really large data being able to get access and bring both computational power not only to collect the data but to analyze the data, to do simulations and computational modeling based on that. To me it gives social science an opportunity to take an analogy we know from the natural sciences and the life sciences where we have a lot of work done in basic science and then we have fields like engineering that takes the basic sciences and makes something from it. I think one of the criticisms that has been made of the social sciences is that, “yeah you might tell us in retrospect how to understand something, but how can you help us make it better?” And I think, one of the advantages of CSS is that it gives us some more tools that will allow us not only to gain insights about how we are as a social system but how we can improve as a social system. An example is the work we are doing right now on trying to help NASA put together the best dream team crews to go to Mars or in the near term go back to the moon. One of the challenges is that when you take the situation like Mars and you take 6 people from as many as 6 different countries and send them on a mission to Mars, at present rocket propulsion pace, that is going to be 9 months to get to Mars, probably a year on Mars and 9 months back. And when you get closer to Mars, the latency in communication is going to be as much as 21 minutes. So this is not a team that has an exit strategy. You have to stay with them throughout. So how do you put together a team that you think will really get along well, that will have the chemistry to work in an isolated confined extreme environment of a space capsule going all the way to Mars? How do you make sure they don’t fall apart in terms of their interpersonal relationships? And if you suspect that something is going to happen, can you forecast it sufficiently in advance so that you can take mitigating steps to try to help reduce that level of tension that might build up?, etc. CSS is a way for us to analyze data upfront, where right now we are studying crews who are put into isolated, confined and controlled environment at NASA’s Johnson Space Centre in Houston, Texas to simulate long-distance space exploration where they are actually put. We get to, metaphorically, but also literally, poke and prod them while they are in these confined spaces for 45 days right now. There is one that is coming up that we are doing also in collaboration with the Russians where they will be in there for 4 months or so. But all of that data is being used by us to build computational models to predict what happens in these groups. We are collecting data from them every single day, sometimes many times a day and we are building models that will allow us to simulate what will happen in this context. So just like today it is commonplace if you look at the design of a Boeing 787 or the Airbus, all of that design is done through computational science. No one actually builds it until all the simulations are done. We hope the benefit of CSS, is that we would be able to do the same kinds of things when it comes to understanding designing teams, not taking chances on it, being able to anticipate what the problems are and taking it from there.
Despite its immense popularity, in recent times Recommender systems have been facing flak, especially in light of the Facebook-Cambridge Analytica controversy. What do you think are the major challenges in the application of Recommender systems and how can these be addressed?
I think that the Cambridge Analytica story has certainly caught the attention and in some cases the fear of the larger public. There are several ways in which I think those dilemmas have to be dealt with. One is, if you recall, the Cambridge Analytica data were originally collected for research by a faculty member who was trying to see if they could predict somebody’s personality characteristics based upon their actions, interactions and transactions within Facebook. From a social science point of view, that is a very interesting idea. If there are ways of being able to do that, surveys, which is how we collect data on personality variables today, are not very scalable. That is, they take a lot of time and effort. So one of the things that has been happening in the broader areas of CSS, but also Network Sciences as well, is that we are trying to find ways in which we can harvest data that will allow us to understand people’s behaviors, actions, attitudes, etc. their relationships with other people based on text analytics and other kinds of technologies that we are using. And I think those efforts are really very important but a story like Cambridge Analytica is going to jeopardize, undermine and make it more difficult for us as researchers to do that. That said, as a citizen, it scares me to see this level of privacy invasion. And frankly, there was more and more discussion even in the last couple of weeks, that while the story we are talking about revolves around Facebook, the amount of data Facebook has about us is actually quite small compared to what Google, for example, has about us. So there is a lot of privacy issues. It is pretty clear that self-regulation might be replaced in the not too distant future by some government regulation. I think Europe has been ahead of this already, in terms of the GDPR – general data protection regulation. Those kinds of issues are also likely to come up in the US Congress. Part of the challenge is, the US Congress needs to be better educated about the web before they can make regulations about the web.
In terms of recommender systems, in particular, one of the challenges of any recommender system is that it is always trying to create a balance between knowing enough about you, that’s where the privacy comes in, and being able to make a recommendation to you on the one hand, but on the other hand it also runs the risk, that if it tells you exactly what people like you do, like Amazon, for instance, “all these people bought these things, like you did, so you should buy another thing like them”, what it does is, it creates, a very conservative system, where there is not much likelihood of change. You are going to start living in what is sometimes called, a filter bubble. Everyone around you is exactly like you and that’s the only people you see. And you see what they see, read what they read, buy what they buy, listen to the songs they do, etc. and you start creating these conservative bubbles, if you may, and that’s not good for innovation. That is not good for dialogue. One of the ironies of every technology that has come down the pipeline is that we say “Oh this new technology is democratizing, it’ll help create a dialogue between people who have different points of view”. Instead what we see historically and repeatedly is that every technology gets used to basically reinforce an echo chamber where people live in a particular echo chamber just like them. And we have seen that in Facebook, we saw that in the elections in the US, in Brexit, that people were only surrounded by social media and networking platforms by people like themselves and as a result when the election results were different from what they thought, they were shocked because none of the people they were interacting with on Facebook could have possibly voted in the other direction. I think that one of the interesting ways in which we address these issues is to recognize that in recommender systems we have to break the conservative loop there. I don’t mean conservative as in the liberal-conservative. I mean conservative as in continuing to give you what you have and allow you to change a little. And I think there is a lot of movement both on the research side as well as commercial social media platforms where people are recognizing that it is important to be able to expose you to things that you may not normally see. There is some advantage to that. From a commercial point of view, they will say “Now if you show me things that I don’t like or I’m not likely to like, then I am not going to spend so much time on your platform”. So that is the tension they face. They have to inject just enough diversity and not too much so that you can continue to get the benefits of a recommender system. But in addition to getting the benefits of the recommender system, you will also have the opportunity to be exposed to something new and probably move in a different direction than where you are so far.
Drawing from Social Sciences and Industrial Engineering & Management Sciences, your work on social networks is a perfect example of interdisciplinary research. How do you see the current state and scope of interdisciplinary research in India?
I think it is going to be very important in the Indian context. I think it is already very important in the Indian context. As I said, all the major societal changes that we face today are not going to be solved by a single discipline. They need excellence across disciplines. So I think that one of the reasons why (given that a lot of these big societal challenges are in fact in the developing world) we need interdisciplinary research to come in.
The work that we see, for the most part, say in the US for example, is focused on problems that they see in the world that are not always exactly the same as the problems that we may have in India. An example of this is work that, I have concluded now, something we did in India on scaling up global health solutions. So the project, funded by the Bill and Melinda Gates Foundation, noted that in places like Bihar and largely in the developing world, not so much in the West, that there is a real problem in terms of neonatal mortality. The rate of neonatal mortality continues to be 12 times higher in India and the developing world as compared to developed world. The good news is we have examples, we have a lot of experiments, demonstrations, and testing that there are inexpensive solutions, literally a topical solution called chlorohexidine, which, if applied on the umbilical cord of a child at the time of birth can reduce the neonatal mortality by 20% or 25% in some studies that have been done. So the question then is how do we get people to use it ? How do we scale up its use? How is this an interdisciplinary issue? You had people in global health who were interested in finding the right solutions that were inexpensive, good etc. But if you want to actually make this happen, the interdisciplinary team has to include not only public health scientists who come up with these ideas but teaming up with social scientists who can tell you how you can take an idea like this and scale it up. It’s one thing to have a solution that can be used by health workers to reduce neonatal mortality. It is a somewhat different issue to get there and for the mother, in particular, to be able to accept this. And that is an issue that has to do with cultural norms in many places. Some of it is literacy, but not all of it is literacy. It very easy for a person to say “well if the people only knew the correct facts from science, then there would not be any problem”. Well, we know that’s not true. There are lots of people around the world, who will tell you “I don’t care what facts you tell me about climate change, I am still entitled to my beliefs”. And the same is true in areas of health etc. So having an interdisciplinary team where you combine people who come from different branches of science and engineering, with people who have more insights about social behavior. What makes people persuaded about something, how does network influence how you act? We know for a fact that our social networks determine what we do. We don’t live in an atomistic world where we act as individuals. Our networks are shaping us and we are shaping those networks. So all of this requires us to have an interdisciplinary approach. And the problems that we have sometimes in India are different from the West. So interdisciplinary research in India by Indians for Indian problems is definitely an important challenge we face.
Taking off from that, how do you see the outlook towards research in general and interdisciplinary research in particular, here and in the US?
I think that in India there has always been this idea that people at the top of the academic ladder will be interested in going into research. That may be changing… It is certainly not true in the West. In Europe, US, etc. the students who rank the highest from an undergraduate background often go into professions rather than going into research. In India one of the things, due to which we actually have suffered in the US is, as the Indian economy is doing better, more and more students who are graduating from the top IITs like IIT Madras are not necessarily thinking of coming to the US as we did, to get PhDs. They are looking to get a job here or get into the IIMs and not necessarily go into an academic career. So as much as I’m really happy that graduates from IIT Madras are contributing directly to the Indian economy, my academic hat makes me unhappy about that because we are not getting some of the best students that we did from places like India.
To answer your question, we know obviously that more needs to be done to incentivize the top Indians to be able to stay in research. When I was a high school student, the families always joked that if you want to be happily married you should go into the IAS or get into IIT. But at the time it was at that transition point. I think that one of the things that India would need to compete very seriously against China in particular, is that, I’m not sure that I have seen India make as many investments in research, both at the undergraduate and graduate level in attracting the talent that’s necessary to do the kind of research that, for example, China is doing. I see that if you look at the two big powers that are around the corner, this is one area where I believe, based on my knowledge, that India is not investing in basic research at the level at which China is. And in the decades to come, that’s going to be a big comparative disadvantage as I see it for India.
More than 3 decades have passed since you graduated from IITM. How do you see the transformation of IITM over the years?
Some things haven’t changed and some things have changed. I think it’s fair to say that the stature of IITM as a top-tier institution in this country and the world remains unchanged. That continues to be the case. I was gratified just to go up to my guest house room right now and see a calendar that showed IITM ranked as the number 1 engineering school and I took a picture of that as it continues to strengthen me and give me pride that I was a student at this institution. Of course, a lot has changed. The talent and the skills of the students who are here are much higher than what it was during my time. I had dinner a few weeks ago with a colleague of mine at Northwestern University, who a few years ago, got a distinguished alumnus award from IIT Powai. He and I were chatting at lunch and we both agreed that neither of us would get into IIT today because the students that are now getting into IITs are so smart that you would never be able to compete with them. So we were grateful that we got in when we did because the talent is so much higher now. In terms of the physical layout, the campus still brings back the old memories. There’s still the Gajendra Circle and all the beautiful landscaping that I remember here, except that there are a lot more buildings and a lot more students here since I was here. But on the other hand, the facilities are great. I was here a few years ago with my batch when we were here for our 25th-anniversary reunion and we had an opportunity to see many of the new facilities and I think those were great signs that IIT Madras is keeping up with everything that they need to do, to be competitive.
I would like to say that in the US, probably today the best-known Indian brand, is not a corporation. It is IIT. So IIT in my mind has the best and the highest global brand equity of any Indian entity. I can go to any place there and if I were to say that I was a student at IIT, everyone recognizes me. Not just within universities but also in industry, in Silicon Valley. I think that IIT Madras is a big part of the development of that brand equity and I don’t see that changing anytime soon.
What would be your message for the students and scholars of IITM?
My message to IIT Madras students is to continue to be creative, to take the initiative to think outside the boundaries, to make sure that you are able to leverage and able to build on all the resources that a place like IIT provides, both physical resources and intellectual resources in terms of the excellent faculty that you have here. I would say, continue to push the boundaries. Look for other ways in which you can be the next generation’s scholar. One idea that I have been pushing is the notion of interdisciplinary work. Working with people from other areas, it doesn’t mean that you have to change and become something else. For me, the path was that I took an undergraduate degree in electrical engineering and then switched to it. That’s not necessarily what I would advocate. But whatever is your passion, in whichever areas of engineering or humanities, etc. that you are getting your undergraduate degree in, think about how what you are doing will benefit by collaborating with people across different disciplines to address the big societal challenges. Because as I said, most of those challenges don’t have single disciplinary solutions. So be creative about it and enjoy yourself. I think that one part that I really remember fondly from my days in Madras was how much fun we had. It was not necessarily a place where we had comfort resources. There was no AC, no mobile phones. But we had a lot of fun. We could go and spend the nights outside Taramani, the village out there. And they had these little coffee shops that stayed open all night. We would study for our periodical exams that were on Monday, Wednesday and Friday mornings, every week. We would go out there and have fun. That bond that was built there continues to be very valuable. So my message to students is to have fun, be bold, be creative but most importantly, enjoy it.
Author: Parth Doshi (BT-BE ’21)