Search This Blog

Thursday, March 10, 2011

Language skills and JEE

Recently, there was a strange drama in the parliament. Due to some confusion, some Rajya Sabha MPs thought that JEE was conducted only in English. The Minister HRD added fuel to the fire by responding to the question in English, and maintaining that the decision on the language is best left to IIT Council, an autonomous body. Finally, the Rajya Sabha had to be adjourned for 15 minutes, to bring the situation under control.

Of course, JEE question papers are available in Hindi too. So the whole discussion was unnecessary. But just imagine the uproar that will happen if IITs were to decide that they will introduce a language test in JEE. The language test could be in multiple languages - it is only to test linguistic abilities of the candidate. But there will always be criticism that English test is simpler than non-English tests. And, therefore, forget about testing language in JEE.

A few years ago, I met a German student who was carrying out research on the success of Indian IT sector. He told me that he is shocked by whatever he had learnt about the Indian higher education system. One point that he made was that India was a unique country in the world (well, amongst the 50 largest countries anyway) in that it did not consider language skills for the purpose of admission to professional courses. There is a huge amount of research which shows that the language skills are the best predictor of being a successful professional, and that is why the entire world considers these skills as amongst the most important parameter for both under-graduate as well as post-graduate admissions. But India did not do it. (And as I said above, language skills are not synonymous with English skills. It should be possible to design a language test in multiple languages.)

Today, I met a professor of another institute, who has been doing a lot of traveling to various institutes in India, and he told me that he found BITS Pilani students to be amongst the best, and we started discussing this. We felt that the reason for this (perception?, how does one really check this?) could be that BITS Pilani is the only institute in India amongst the top rated institutes which considers language skills as important, and includes them in its admission test. (And also a section on logical reasoning.) It also happens to be the institute with perhaps the highest cutoff for the 12th class marks. (It requires 80 percent in aggregate of Physics, Chemistry and Maths.) And of course, there is a lot of research which shows that the performance in school has a better correlation with the performance in college compared to the correlation between entrance exam with performance in college.

As a result of different admission processes, while a lot of good students in schools (including those who are really good at at least one language) do manage to get through JEE and join IITs, but a lot of students who manage to get admission to IITs do not have good performance in school or do not have good language skills.

Is there a way to influence stake holders to include an exam component in JEE on a language (of the choice of the candidate), and also increase the cutoff marks in 12th from 60% to at least 65%.


Jayadevan said...

Sir, I strongly think that for science subjects, the medium of instruction from the primary school itself should be in English. But the text books shall be printed in both the local language as well as English. This will help the students in understanding the concepts and documentation of those concepts in English.
In India Hindi is spoken only in eight states (UP, BH, DL, RJ, HR, MP, CH, JH). If the JEE question paper is printed in Hindi, it has to be printed in other official regional language as well. Regional political parties can make an issue out of it..
I think that a student who wants to be in an IIT should at least understand the questions written in least….at least… Our politicians should understand least..

Srinivas said...

I wonder about this too. I do believe though that there are performers at the extreme tail of normally distributed populations who may have rather skewed skills; for example excelling in analytical or mathematical skills, but with modest or even poor language skills, who would then find it hard to get through the JEE if you were to rejig the JEE to be more in line with for example the admission criteria used at BITS(OK, I exagggerate a bit). But I'd like to submit that the kind of testing that the JEE gives, while ignoring all previous performance and all other aspects of an entrant, is in a way good, because it greatly levels the playing field for entry. I bet that the socio-economic and IQ levels of those entering BITS is much more narrrow than those entering the IITs via JEE. I think that in itself is a great strength of the IITs. It is true though that that can lead to greater problems for students that may be underprepared in languages or for the professors who must then address those deficiencies. On the third hand, of course, there's a moral hazard, given the importance of the JEE, that school students might tend to ignore the important linguistic component of their education "because it doesn't matter for admission to IITs". On the fourth hand (foot?), most of our "native" linguistic development is complete by the time we get to the tenth grade which is around when students typically start thinking about things like careers.) Education, testing and admissions in a diverse and pluralistic society was never an easy thing . The best we can do is to design tests that will still admit students with perhaps a skewed but otherwise brilliant profile, rather than asking for all skills to be uniformly above a relatively high threshold. (Skewed skill profiles are, I am told, on the increase, if the autism-spectrum-disorder epidemiologists are to be believed.)

Dheeraj Sanghi said...

@Srinivas, If you compare the student profile at BITS Pilani (all campuses combined) with the student profile of all IITs combined, it may be true that there is more diversity in IITs than BITS. If that is indeed true, my gut feeling will be that it would be due to higher brand name of IITs (I am sure many people in Delhi would consider DTU/NSIT as better than BITS, many people in East would consider Jadavpur as better than BITS, and so on), and much cheaper cost of education. BITS is low cost as the private sector goes, but cannot compete with Rs. 1500 crore subsidy from the government to IITs. Also, BITS does not have any form of reservation. And though BITS now has campuses in Goa and Hyderabad, they have not grown to have the same quality as Pilani campus. So BITS would attract less students from across the country (except AP).

In short, I don't think JEE has anything to do with it.

In fact, JEE is causing an uneven student population because it almost forces you to go through coaching, and for some reasons unknown to me, coaching centers have developed only in few cities. A study that I did a couple of years ago showed that the number of students that pass JEE from Tamilnadu, Karnataka, and Kerala are minuscule.

Anonymous said...

I agree. There was a time when I used to think that strong analytical skills were the only requirement to strike it big in engineering. But as I mature, I realize that only 15% of your success can be attributed to one's individual contribution. The rest 85% is based on the cooperation your receive from your circle of friends and colleagues. Language skills are crucial for being able to communicate and leverage this circle of well wishers. This blog entry is one such an example. If I was a reclusive savant, you would not even be aware of my existence. But now due to my English writing skills and the power of the Internet, my voice is also being heard.

WebMiner said...

Over the last decade I have "advised" over 40 IIT UGs and it has been worthwhile talking to about 5. At least half cannot write logically. Most of the rest cannot say anything interesting.

Vineet Pandey said...

Hello Prof. Sanghvi,

Would you justify creating 'successful professionals' with great language skills at institutes such as IITs at the cost of an entrance scheme which favours some sections of students more than others. Because for many of us an intuitive understanding of English comes from a lot of extrinsic factors (schooling, home environment, location) while hard work & motivation is a student's individual characteristic. At the same time, four years is a good time for someone who is 'unsmart' lingusitically to turn it around. I am in Korea currently and there are so many people who are brimming with ideas * in Korean * but can't get them out in English. I believe it would be unfair to judge them harshly.

Also, on a broader note, do you think we dropped the ball by making Hindi the national language (yes, little can be done now, but just wanted to know your take).

I agree with the coaching issue and Srinivasa (assumed to be a peer)'s analysis.

Dheeraj Sanghi said...

@The Third Eye,

Communication skills is obviously important, as you point out. What is not so obvious and what I am trying to focus on is that language skills are important even when you are just thinking of an idea alone.

Dheeraj Sanghi said...


I am not saying that we evaluate English skills. I am arguing that we evaluate language skills. Just like JEE is given in more than 1 language, we could have a language test in multiple languages. Of course, a bare minimum English capability to the level of understanding lectures and books should be there too, but that is not the point of this blog.

Vineet Pandey said...

I was confusing it with communication skills. So this language skill you mentioned is the clarity in thinking. Wouldn't a lot of our coaching educated student population score badly on that (just wondering).

Dheeraj Sanghi said...

@Vineet, I am looking for a test where a person is checked for his ability to have a good size vocabulary, ability to distinguish the subtle differences in the meanings of words with similar meaning, ability to comprehend sentences and relationship between those sentences. It is said that if you do not know the subtle differences in the words with similar meanings, it is difficult to think of slightly different shades of your idea.

And all this testing can be done in several languages, because it does not matter what language one uses during the thought process.

English is needed for external communication, and is important, but that is not the focus of this blog. In general I believe that if one knows one language very well, and knows sufficient English to follow lectures and read texts, four years is enough time to improve English communication.

As an aside, I think today's culture of focusing on entrance exams right from 9th grade, and entrance exams not including any testing of languages, has resulted in the culture of coaching. And this has contributed to lack of creativity on our campuses.

L said...

Very happy to hear more voices emphasising the importance of language skills. Most of the students I teach have very poor language skills since they are told from childhood that science and Math are the only important subjects to be studied. The children are discouraged(yes discouraged!) from reading or any activities that may improve their language skills.
The irony is that due to poor language skills they are not able to understand concepts in science.
I hope such debate brings awareness amongst parents.

WebMiner said...

It is by now well-established in linguistics and psychology that we do not use language because we are smart; we are smart because we use language to think. Communication is only a secondary purpose of language. People outside the current IITs (including the "original" ones) may be shocked to discover the level of (poor) understanding of quantifiers, clauses, conditional statements with negations, nesting, implications, etc., in even UG students in IITs. If they were good in any one language, this would not happen. The typical IITian is not proficient in any language, natural or programming. The IIT brand currently rides on perhaps 1 out of 10 at the level that an uninitiated outsider would expect as "IIT quality".

prochot said...

Testing for language skills in JEE is a horrible idea on multiple fronts.

1) The research about linguistic abilities being good gauge of professional success is about proficiency in native language, and probably in mono-linguistic environments. I grew up speaking Bhojpuri, my medium of education was Hindi up to class 10 before changing to English. I guess that's a typical Indian. That makes my skills weak in any one language compared to someone who has spent all his life in English, for example.

2) Language is a means of communication by definition. I would rather hire someone brilliant at technicals and OK with language than someone brilliant with language and OK at technicals. As long as you are able to communicate effectively with your coworkers, it's OK.

3) One can hone ones language skills outside IIT. JEE should concentrate on finding people who would benefit most from its teachings (technical in nature, barring a few exceptions of HSS).

4) The means of technical education should continue to be English. IIT should help people learn English if they are not good (as it does currently). English mode of education is one of the BIG factors why India has done so well. Indians find it easy to work with people from all over the world. You should have told the German student that part of reason Germany is doing so badly (negative population growth, lack of IT jobs for example) is because of lack of their English skills.

Here is my proposal
3) Teach technical courses in English
2) Teach English to students who need help
1) Select those students who would benefit/learn/earn/shine most from technical education in English given that they are taught English

In short, don't change anything.

WebMiner said...

@prochot: "Language is a means of communication by definition." --- Wrong. Reread my post, and then Chomsky (i.e. his technical work) if you have the stomach.

In general it is good blog etiquette to give evidence that you have understood prior postings, and not put forth an opposition without building a counter-argument to prior postings.

prochot said...


Frankly, I didn't care what you wrote in your comments. I don't have time or willingness to read random comments by random people. My comments were directed at Prof Sanghi since I "knew" him at IITK.

But, since you specifically addressed me, here are my thoughts on your comments:

1) I agree with your assertion "we are smart because we have language".

However, all the evidence, that I have seen, point to the fact that language helps only because it serves as a bridge of communication among different people, and also among different generations.

This means that you only need to be able to communicate (with bit error rates close to 0 - a geeky reference to language skills) with others, you are skilled enough.

2) "At least half cannot write logically. Most of the rest cannot say anything interesting.": I think you had a VERY bad sample set.

3) "The typical IITian is not proficient in any language, natural or programming.": I spent 4 years at IIT and have spent many more years in industry since then, and have not met ONE single guy from any IIT (or any other engineering college from India) who was not proficient at natural language or at programming languages.

By proficiency in natural language, I mean able to read/understand any commonly available media (newspaper, books, tv etc.), and write well enough that others can understand.

By proficiency in programming language, I mean able to program his code once he had "algorithm" ready.

On the other hand I have met plenty of people who were good at both these languages but did not have any technical knowledge to earn their livelihood.

I may have experssed all I have to say about this issue, so don't get offended if I have don't contribute further to the discussion.

Dheeraj Sanghi said...

@prochot, you say the following in your first comment: "The research about linguistic abilities being good gauge of professional success is about proficiency in native language, and probably in mono-linguistic environments." The word "probably" indicates to me that it is your guess and not that you have gone through any research yourself. Also, if you have seen research in a specific context, it does not mean that no research exists outside that context.

In that sense, you are making very sweeping statements without any basis.

Have you read any specific research which says that for multi-lingual people, not knowing any language well is better (or at least as good) as knowing at least one language well. If yes, I would like to be educated.

You continue to make sweeping generalizations when you say Germany is not faring well because of lack of English. Can you explain why US is not doing well, and why China is doing well. Please don't trivialize serious issues.

When you make such sweeping generalizations, it shows that you are really not interested in a discussion. That you believe that you know everything, and you are only interested in making some illiterates aware of your view point. You make that impression stronger by admitting in the second comment that you don't have time or willingness to go through the discussion. I think when we participate in a discussion, we have to give respect to other participants, and make efforts to find out what they are saying, at least to ensure that you don't repeat and don't waste time of others. If you were only interested in responding to my post, and not interested in the discussion, then you could have sent me an email directly, rather than responding on the blog, which is what several people have done.

Most of us are busy persons. You may have noticed that I do not write sometimes for weeks together. It is not that I do not get time to write even 20-30 lines in those days. But I always see my calendar and guess whether I would have enough time to participate in the discussion that may ensue. I will learn from my readers only if I respect their time, and value their opinion enough to respond to them. It is not necessary that I agree with them, but if I disagree, I must explain why I disagree. All this takes time. If I do not have so much time, I shouldn't blog.

Frankly, to me the first comment showed so much arrogance that I decided that it did not deserve any response from my side. But the second comment was worse in terms of arrogance (though the content is explained better), and I thought I should request you to tone down your language.

L said...

I think it doesn't need research to decide that language is essential for thinking. Every thought we have is in some language. We can experience it for ourselves every time. Language skill is just the ability to formulate words for more and more complex thoughts... the more complex the thought you are able to formulate, better your language skill. Usually, given the compulsion, a person who has excellent verbal skill in his/her native language, will easily acquire a high level of competency in English or any other language since the framework is already present.

prochot said...

Dear Prof,

Thanks for your comments.

1) Arrogance: I wholeheartedly disagree that there was any arrogance in either of my posts, which comprised of some facts and some guesswork, properly identified with phrases like "probably", "I think", "I guess" etc. It may look arrogant when put beyond its context.

For example, I said "It's a horrible idea ". This is my objective conclusion of reality based on the facts I cited and I showed my derivation. There may be flaws in my derivation, but none that I know, and I welcome any such identification (that's another motivation behind sharing my thoughts, in addition to my primary reason - to convince you of the problems with testing for language skills in JEE).

I dont know - may be - "I would hire someone" can be seen as flaunting my power. The defining word is "would". It means if I had capacity to hire someone and I were to make a decision, this is what I would do.

"Frankly, I didn't care what you wrote in your comments." is not arrogant either - it's just an objective truth and describes what I do. I bet, you too do not have time to read/respond to anyone and everyone. You respond to the comments on this site, because in past you have found such activity to be (at least) partially insightful/interesting. I am not a big blogger. I read your post and responded as such. The sentence was my explanation to WebMiner about why I missed his perceived etiquette of blogging!

I would like to understand what else you found arrogant so I can fix/explain them. I may have been just expressing my opinions in clear words.

2) Research about multilingual environment: As you identified, it's just my guess based on the fact that I have not found anything contradictory in any of my linguistic courses (and yes, I have read Chomsky, which, by the way, is not beyond criticism. His cherished Universal Grammar has attracted new criticism similar to the ones encountered by Penrose's "Emperor's New Mind" -- AI can emulate humans better than what we thought.), but, of course, I may be wrong.

However, my main point was that any connection of language and skills (circa 'we are smart because we have language') has been based on long term studies of how humans evolved. Needless to say, all modern humans are skilled to continue passing their findings to next generations!

3) I am aghast by proliferation of coaching classes and their negative impact, for example, people who got good ranks due to coaching classes, may do badly afterward. Also, they tend to skew opportunities in the favor of cities.

However, testing for language skills is just wrong since it would not serve any purpose (urban students will do better at English than Tamil students will do at Tamil).

-- continued --

prochot said...

4) Now, one can argue that language test is not to equalize opportunities, but, rather needed to find best people suitable for IIT. I think this was the point Webminer was trying to make. This argument is equally flawed, as I explained in point 4 of my first comment.

5) Contrary to what you say, I have not made any unsubstantiated generalization.
For example, "English mode of education is one of the BIG factors why India has done so well": IT sector has played a huge role in India’s development in last 20 years from curtailing corruption to ensuring fair elections. The primary reason behind flourishing Bangalore is good work done by IIT alumnus in the Silicon Valley, which was enabled by good education as well as Indians being good at English. For example, I do not know any 1st generation Chinese entrepreneurs in USA (doesn't mean there are none, just fewer), but I know several such Indians. In summary, English mode of education was indeed a BIG factor.

Similarly, lack of English is indeed one of (note, I said one of, not the) the reasons Germany is doing badly. I know several people firsthand who are trying to migrate from Germany to elsewhere due to language problems.
To answer your questions: US is not faring well, in spite of English. China is doing so good, in spite of lack of English. English is not the defining feature of development, but it sure helps!

6) To answer your primary question about how to make JEE more equitable to all: I have not yet thought of a foolproof way of doing so. JEE could do relative grading of students based on their environment. For example, a farmer's son scoring 80/100 is more deserving than my son scoring 90/100 (who had all the help in the world). The challenge will be identifying such classes (location, parents, parent's income etc.). Also, this will not be palatable to our politicians.


Dheeraj Sanghi said...

@prochot, let us relook at your first set of comments.

Point number 1. You are not sure about it yourself. But to the extent you believe in this, let me suggest an analogy. Suppose I have read many books which say Hindi is great, and you have read many books which say Bengali is great. Can I conclude on the basis of my readings that Bengali is not great, and can you conclude from your readings that Hindi is not great. Obviously no. You can conclude Hindi is not great only if you explicitly do research or read some one's research concluding that Hindi is not great. Same way, you may have read a lot about importance of communication in different contexts, but your not reading about language being a tool for thought does not mean you can deny it.

Point Number 2: No one has denied it on this blog/comments. The point of the blog has been that besides external communication, language is important for thinking of ideas.

Point Number 3: No one has contested it either.

Point Number 4: No one has suggested that medium of instruction in technical fields should not be English. Of course, your example of Germany is very naive. I mean, for 50 years, the country has grown so fast. But if it is not doing well for 5 years, you will blame lack of English for it, when there are so many examples of various non-English countries growing well, and English speaking countries not doing well.

So, the way I look at it is that almost all of what you are saying is orthogonal to what is written in the blog, but based on such statements, you are willing to use strong words as "horrible idea."

So there is no premise, there is no logic, but there is a strong conclusion.