Tuesday, December 30, 2008

Jayant Haritsa on enhancing research capabilities in our premier institutions

The Research track at the PanIIT Global Conference featured a session on IIT Research: Inputs and Outputs. My colleague and friend Prof. Jayant Haritsa was one of the panelists in this session; the slides for his talk are here.

One of his points, which should be familiar to readers of this blog, is about how our top institutions need to do more -- much, much more -- to set people up for success. Since there's a lot in his talk that I agree with, I'm producing his extended abstract below. Treat this as a guest post by Jayant.

* * *

Rethinking the Research Crisis @IITs

Jayant Haritsa
Indian Institute of Science

The following opinions are based on my experience of 8 years as a graduate student and researcher in the US, followed by over 15 years as a computer-science faculty at the Indian Institute of Science, Bangalore.


Here's a quote from one of the slides:

Cassius to Brutus: “the fault lies not in our stars, but in ourselves”

Similarly ... “the fault lies not in our inputs/outputs, but in how our institutions support and utilize these inputs/outputs”

* * *

The issue raised by the panel is how to improve the quality of the faculty and students that enter premier scientific institutions in India and how to improve the productivity and impact of their research. Certainly there are a variety of national stumbling blocks that do negatively impact the attractiveness of the IITs as scientific destinations, and these factors have been enumerated by the moderators. However, I would like to make the case that this is really the proverbial tip of the iceberg, and the more germane and urgent question to be asking is "How do the talented students and faculty that do join the IITs fare?". My point is - had most of these folks turned out to be success stories, then we would have seen a "domino effect" that would have automatically generated a rich stream of similar applicants - in fact, we wouldn't be having this panel discussion in the first place! In a nutshell, "institutional mechanisms must be consciously set up to mentor star faculty and students for visible success".

Unfortunately, the typical experience is otherwise. Specifically, we have had several cases of faculty with top-notch academic pedigrees who after returning to India have largely disappeared from the international research arena. For example, it would not be stretching the realms of credulity for a faculty to have produced more PhDs and publications in a few years as a junior faculty in a top US school as compared to a couple of decades spent at our institutions!

Similarly, we have also had several instances of our own best master's students who have subsequently gone on to join our Phd programs failing to either produce a thesis commensurate with their talents, or sometimes even to complete the program. In a recent instance, a star in-house student "timed-out" after spending no less than ten years in the PhD program!

The above anecdotal information serves only to highlight my point that stemming the perceived rot first requires introspection about the internal mechanisms of our institutions, before we start agonizing over the external interfaces. Specifically, what causes even highly talented individuals to unceremoniously fade away without giving full expression to their abilities?

My analysis is that the following reasons are primarily to blame for this unhappy state of affairs:

Firstly, the poor level of academic professionalism and administration. A typical situation faced by applicants is the wall of indifference with no response, either positive or otherwise, forthcoming to their applications for extended periods of time. Subsequently, their offices and accomodation are rarely set up adequately by the time they enter the academic portals. Then, during the initial years, they rarely get to meet with top administration officials and are often left to fend for themselves in figuring out the system. Overall, an atmosphere of benign indifference, often bordering on neglect, characterizes our treatment of new recruits.

Secondly, a particularly thorny issue that often plagues new researchers, while they still retain their idealism, is the institutional pressure to carry out "locally relevant" research, which is also highlighted in the panel agenda. While on the surface this would seem an unarguable objective, yet in practice it usually turns out to be the case that it merely provides a "holier-than-thou" fig leaf to justify poor quality research and third-rate publications. I would suggest that it would instead be far better for us to first attempt to keep our heads above water in international research, and if it coincidentally happens that the work is also of local relevance, so much the better. That is, the local relevance should be the icing on the cake, and not a moralistic justification for a moldy cake. Our first duty is to put India on a high pedestal in world research.

Further, the lip-service paid to locally relevant research is evident from the fact that most institutions talk only about the research objectives without using the same yardstick for the students or faculty performing the research. If one wants to be truly relevant to local society, there is a much simpler and immediate option - welcome the reserved category students and faculty with open arms, instead of the arms-length approach currently in vogue! That is, start with "locally relevant researchers" before we wax eloquent on the merits of locally relevant research.

Thirdly, the unwillingness to call a spade a spade when it comes to professional judgements of one's colleagues or their students. For fear of vitiating the collegial department atmosphere, we make the grave mistake of providing sugar-coated assessments and, in the long run, encourage researchers to fall prey to a false sense of accomplishment. Instead, we should clearly enunciate the performance metrics at the time of joining and, most importantly, stick with these metrics in the judgement process. At IISc now, quite a few departments, especially in the Electrical Sciences division, have authored documents outlining their expectations and modes of evaluation for both faculty and students. As a simple case in point, a recently implemented policy in the Computer Science department is that ME students can be considered for an S (outstanding) grade in their project work only if they have submitted a paper to a journal or conference that is internationally recognized to be of A+ or A calibre. While earlier about 60% of the students used to routinely obtain an S grade in the project, beginning last year, we have not only had the percentage come down steeply to around 20%, but more importantly, the departmental publications have shot up significantly. Further, grades are not only used as carrots but also as sticks. Last year, for the first time, a non-neglible number of students were forced to stay behind without graduating until they satisfied a baseline quality requirement. The basic point here is that once we enforce accountability criteria for both faculty and students, the genius of the Indian mind is such that it will automatically deliver!

In closing, I would like to reiterate my main point - echoing Cassius' injunction to Brutus, "the fault lies not in our stars but in ourselves", I would say "the fault lies not in our inputs but in how our systems utilize these inputs".


  1. Niket said...

    Well said.

    Lack of skilled technicians is another problem. A lot of faculty time is spent in firefighting and getting equipments working after some problem. In grad school, there were technical staff to do this.

    Talking about bureaucracy: I have initiated the process of ordering two PCs from Dell India in August 2008. Still waiting for the PCs.

    At least in our institute, we pay lip service to research. A faculty with 60 international publications is still an assistant prof.

  2. Anonymous said...

    Lack of technical staff problem, that the above commenter mentions, is probably the problem that arises when scientific/technical/vocational education at lower levels of educational hierarchy is neglected

  3. Anonymous said...

    >That is, the local relevance should be the icing on the >cake, and not a moralistic justification for a moldy cake. Our first duty is to put India on a high pedestal in >world research

    Well I agree with Jayant's statement and is somewhat opposite to what Nobelist Dr. A.Sen recommends in his talk at the conference.

    Institutes should be set up with clearly separate missions...so there should IITs/IISc for world-class research and there should *good* institutes for social/local/national missions
    The lack of instis in the second category puts burden on the first category to solve all the ills and problems of the society.

  4. Ashutosh said...

    Great points, all of them, especially the ones about sugar-coating assessments and about first and foremost doing first-class research even if it's not "local".

  5. Anonymous said...

    All this angst about the future of IITs is misplaced. In a society where per-capita availability of energy is rolling steadily downhill, technology simply cannot command any broad support base. IITs, as envisaged in the 50s, are going to be phased out and abandoned within 20--30 years, replaced by mediocre degree mills. However, it is also very unclear why Indian youth would covet an engineering degree 30 years from now. (Medicine, law, peace and conflict studies, and management should be in much more demand.) India's problems do not include lack of engineering know-how. All the engineering solutions that can improve Indians' quality of life are already in the public domain. Civilized living isn't really rocket science---many other societies have proved that already.

  6. Anonymous said...

    The future of IITs indeed looks bleak even in the short-term. Given the casual manner of the proposed expansion and the lack of incentives for highly qualified people to join as faculty - along with the increased mobility (as opposed to 70s) and global demand for talent - I just can't fathom how IITs are going to catch up with the rest of the world. All the talk of maintaining stardards will be mere hogwash. IITs will be finished in 5 years time - will be replaced by 'local research' centres - organizing copious third-rate conferences and symposia of 'local relevance' - in the end, achieving nothing except making the local organizers richer. Meanwhile, kids of our elite - right from PM down to petty bureaucrats - will continue to flock to US,UK,Canada,Australia etc. for 'phoren' education. BTW, that's the pakistan way of higher education - that we too are soon heading to. Great!

  7. rahul thathoo said...

    well said jayant. agree with all of his points.

  8. Vj said...

    I agree with Prof. Jayant's practical comments. I have seen some faculty in the institute who came from the US with 60 -100 publications and now have spent atleast 10 years and have published only a handful. I sometimes feel that these folks come to India with an idea of semi retiring and let their US publications do the talking. They treat the research students in their groups to do stuff the way a student in the US would but they would not guide the students the way we expect. I have been in the institute for a while and have seen some close examples of my friends who were really brilliant but could only come up with a mediocore Ph.D thesis and i strongly believe that his US returned, 100 publications guy did not bother to provide him the needed guidance as a research advisor.

  9. Anant said...

    One serious lacuna in the argument that I see is this: on the one hand Haritsa says that there is a lot of indifference, etc., and on the other hand he speaks of sugar-coated assessments. So what exactly is going on here?

  10. Anonymous said...

    Assuming Jayant is referring to engineering research - success, even if defined by publications or quality of Ph.D. thesis, depends on a whole host of favorable environment. In this category,
    apart from the often quoted "infrastructure" one has to taken into account that a lot of "good" engineering research in US are due to close academic-industry r&d and academy-defense r&d based projects
    that place a lot of emphasis on "experimental" research which can be translated into products even if futuristic. This makes a big difference since Indian academic R&D is either too theoretical or too computational and generally single PI oriented which excludes a whole gamut of "good" research that is possible in US. I would also venture out to say that may also be a problem with science research in India.

  11. Varun said...

    Completely agree with all points. Although I feel that implementing all of them at the same time, will be difficult. Perhaps implement them slowly over a period of time...
    But yes, otherwise a very sensible and good report.