Rediscovering Ramanujan
Interview with Prof. Bruce C. Berndt.
Interviewer : Asha Krishnakumar
The academic lineage of most eminent scholars can be traced to famous
centres of learning, inspiring teachers or an intellectual milieu, but
Srinivasa Ramanujan, perhaps the greatest of Indian mathematicians, had
none of these advantages. He had just one
year of education in a small college; he was basically self-taught.
Working in isolation for most of his short life of 32 years, he had
little contact with other mathematicians.
"Many people falsely promulgate mystical powers to Ramanujan's
mathematical thinking. It is not true. He has meticulously recorded
every result in his three notebooks," says Dr. Bruce C. Berndt, Professor of Mathematics at the University of Illino
is, whose 20 years of research on the three notebooks has been compiled into five volumes.
Between 1903 and 1914, before Ramanujan went to Cambridge, he compiled
3,542 theorems in the notebooks. Most of the time Ramanujan provided
only the results and not the proof. Berndt says: "This is perhaps
because for him paper was unaffordable and so he
worked on a slate and recorded the results in his notebooks without the
proofs, and not because he got the results in a flash."
Berndt is the only person who has proved each of the 3,542 theorems. He
is convinced that nothing "came to" Ramanujan but every step was thought
or worked out and could in all probability be found in the notebooks.
Berndt recalls Ramanujan's well-known i
nteraction with G.H. Hardy. Visiting Ramanujan in a Cambridge hospital
where he was being treated for tuberculosis, Hardy said: "I rode here
today in a taxicab whose number was 1729. This is a dull number."
Ramanujan replied: "No, it is a very interestin
g number; it is the smallest number expressible as a sum of two cubes in
two different ways." Berndt believes that this was no flash of insight,
as is commonly thought. He says that Ramanujan had recorded this result
in one of his notebooks before he cam
e to Cambridge. He says that this instance demonstrated Ramanujan's love
for numbers and their properties.
Although Ramanujan's mathematics may seem archaic by today's standards,
in many respects he was far ahead of his time. While the thrust of 20th
century mathematics has been on building general theories, Ramanujan was
a master in finding particular result
s which are now recognised as providing the core for the theories. His
results opened up vistas for further research not only in mathematics
but in other disciplines such as physics, computer science and
statistics.
After Ramanujan's death in 1920, the three notebooks and a sheaf of
papers that he left behind were handed over to the University of Madras.
They were sent to G.N. Watson who, along with B.M. Wilson, edited
sections of the notebooks. After Watson's death
in 1965, the papers, which contained results compiled by Ramanujan
after his return to India from Cambridge in 1914, were handed over to
Trinity College, Cambridge. In 1976, G. E. Andrews of Pennsylvania State
University rediscovered the papers at the T
rinity College Library. Since then these papers have been called
Ramanujan's "lost notebook". According to Berndt, the lost notebook
caused as much stir in the mathematical world as Beethoven's Tenth
Symphony did in the world of Western classical music.
Berndt says that the "unique circumstances surrounding Ramanujan and his
mathematics" make it very difficult to assess his greatness among such
mathematical giants as Newton, Gauss, Euler and Reimann. According to
Berndt, Hardy had provided the following
assessment of his contemporary mathematicians on a scale of 0 to 100:
"On the basis of pure talent he gave himself a rating of 25, his
collaborator J. E. Littlewood 30, German mathematician D. Hilbert 80,
and Ramanujan 100." Berndt says that it is not R
amanujan's greatness but only its measure that is in doubt.
Besides the five volumes, Berndt has written over 100 papers on
Ramanujan's works. He has guided a number of research students in this
area. He now works on Ramanujan's "lost notebook" and on some other
manuscripts and fragments of notes. Recently in Che
nnai to give lectures on Ramanujan's works at the Indian Institute of
Technology, the Institute of Mathematical Sciences and the Ramanujan
Museum and Mathematical Centre, Berndt spoke to Asha Krishnakumar
on his work on Ramanujan's notebooks, the
broad areas in mathematics that Ramanujan had covered, the vistas his
work has opened up and the application of his work in physics,
statistics and communication.
Excerpts from the interview:
How did you get interested in Ramanujan's notebooks?
After my Ph.D. at the University of Wisconsin, I took my first position
at the University of Glasgow (Scotland) in 1966-67. Prof. R. A. Rankin
was a leader in number theory at that time. I remember being in Rankin's
office in 1967 when he told me about Ramanujan's notebooks for the first time. He said: "I have a copy of the
notebooks published by the Tata Institute of Fundamental Research,
Bombay (Mumbai). Would you be interested in looking at it?" I said, "No,
I am not interested in it."
I did not think about the notebooks for some years until early 1974 when
I was on leave at the Institute for Advanced Studies in Princeton, U.S.
In February that year, I was reading two papers of Emil Grosswald in
which he proves some formulae from Ramanujan's notebooks. I realised I could prove these formulae as well by
using a theorem I proved two years ago. I did that and then I was
curious to find out whether there were other formulae in the notebooks
that I could prove using my methods. So, I went
to the Princeton University library and got hold of Ramanujan's
notebooks published by the TIFR. I was thrilled to find out that I could
actually prove some more formulae. But there were a few thousand others
I could not.
I was fascinated with the notebooks and in the next few years I wrote
papers around the formulae I had proved from the notebooks. The first
was a repository paper on Ramanujan's theta 2n+1 formula, for which I
did a lot of historical research on other pr
oofs of the formula. This I wrote for a special volume called Srinivasa
Ramanujan's Memorial Volume, published by Jupiter Press in Madras
(Chennai) in 1974. After that, wherever I went, I was all the time
working on, and proving, the various formulae of
Ramanujan's - to be precise - from Chapter 14 of the second notebook.
Then I wrote a sequel to this.
Let me jump ahead to May 1977, when I decided to try and prove all the
formulae in Chapter 14. I took this on as a challenge. There were in all
87 results in this chapter. I worked on this for the next one year. I
took the help of my first Ph.D. student,
Ron Evans.
After about a year of working on this, the famous mathematician George
E. Andrews visited Illinois and told me that he discovered in the spring
of 1976 Ramanujan's "lost notebook" along with G. N. Watson and B. M.
Wilson's edited volumes on Ramanujan's t
hree notebooks and some of their unpublished notes in the Trinity
College Library. I then got photocopies of Ramanujan's lost notebook and
all the notes of Watson and Wilson. And so I went to the beginning of
the second notebook.
What does the second notebook contain?
This is the main notebook because it is the revised and enlarged version
of the first. I went back to the beginning and went about working my
way through it using Watson and Wilson's notes when necessary.
How long did you work on the second notebook?
I really do not know how many years exactly. But some time in the early 1980s Walter Kaufmann-Buhler, the mathematics editor of Springer Verlag in New York, showed interest in my work and decided to publish it. That had not occurred to me till the
n. I agreed and signed a contract with Springers.
That was when I started preparing the results with a view to publishing
them. I finally came out with five volumes; I had thought it would be
three. It also took a much longer time than I had anticipated.
After I completed 21 chapters of the second notebook, the 100 pages of
unorganised material in the second notebook and the 33 pages in the
third had a lot more material. I also found more material in the first
which was not there in the second. So, I fou
nd a lot of new material. It was 20 years before I eventually completed
all the three notebooks.
Why did you start with the second notebook and not the first?
I knew that the second was the revised and enlarged edition of the
first. The first was in a rough form and the second, I was relatively
certain, had most of the things that were there in the first and a lot
more.
What did each notebook contain?
The new results that were in the second notebook were generally among
the unorganised pages of the first. And the third notebook was all
unorganised. A higher percentage of the results in the unorganised parts
of the second and the third were new. In oth
er words, you got a higher percentage of new results as you went into
the unorganised material.
What do you mean by new results?
Results that have not been got earlier.
What is the percentage of new results in the notebooks?
Hardy estimated that over two-thirds of the work Ramanujan did in India
was rediscovered. That is much too high. I found that well over half is
new. It is difficult to say precisely. I would say that most results
were new because we also have to consider
that in the meantime, from 1920 until I started doing this work, other
people discovered these things. So, I would say that at least two-thirds
of the material was really new when Ramanujan died.
Ramanujan is popularly known as a number theorist. Would you give a
broad idea about the results in his notebooks? What areas of mathematics
do they cover?
You are right. To much of the mathematical world and to the public in
general, Ramanujan is known as a number theorist. Hardy was a number
theorist but he was also into analysis. When Ramanujan was at Cambridge
with Hardy, he was naturally influenced by
him (Hardy). And so most of the papers he published while he was in
England were in number theory. His real great discoveries are in
partition functions.
Along with Hardy, he found a new area in mathematics called
probabilistic number theory, which is still expanding. Ramanujan also
wrote sequels in highly composite numbers and arithmetical functions.
There are half a dozen or more of these papers that ma
de Ramanujan very famous. They are still very important papers in number
theory.
However, the notebooks do not contain much of number theory. It is,
broadly speaking, in analysis. I will try and break that down a little
bit. I would say that the area in which Ramanujan spent most of his
time, more than any other, is in elliptic funct
ions (theta functions), which have strong connections with number
theory. In particular, Chapters 16 to 21 of the second notebook and most
of the unorganised portions of the notebooks are on theta functions.
There is a certain type of theta functions ide
ntity which has applications in other areas of mathematics, particularly
in number theory, called modular equations. Ramanujan devoted an
enormous amount of effort on refining modular equations.
Ramanujan is also popular for his approximations to pie. Many of his
approximations came with his work on elliptic functions. Ramanujan
computed what are called class invariants. Even as he discovered them,
they were computed by a German mathematician, H
. Weber, in the late 19th and early 20th centuries. But Ramanujan was
unaware of this. He computed 116 of these invariants which are much more
complicated. These have applications not only in approximations to pie
but in many other areas as well.
Have you gone through every one of the 3,254 entries in the three
notebooks and proved each of them, including in the unorganised
material?
I have gone through every entry in the notebooks. If a result has
already been proved in the literature, then I just wrote the entry down
and said that proofs can be found in this literature and so on. But I
will also discuss the relevance in history of
the entry.
What are the applications of Ramanujan's discoveries in areas such as physics, communications and computer science?
This is a very difficult question to answer because of the way
mathematics and science work. Mathematics is discovered and it is then
there for others to use. And you do not always know who uses it. But I
have regular contact with some physicists who I know use Ramanujan's work. They find the results very useful in their own
application.
What are the areas in physics in which Ramanujan's work is used?
The most famous application in physics is in the area of statistical
mechanics. Among those who I know have used Ramanujan's mathematics
extensively is W. Backster, the well-known physicist from Australia. He
used the famous Rogers-Ramanujan identities in what is called the hard hexagon model to describe the molecular
structure of a thin film.
Many of Ramanujan's works are used but his asymptotic formulae have
found the most important application; I first wrote this in 1974 from
his notebook.
Then there is a particular formula of Ramanujan's involving the
exponential function which has been used many times in statistics and
probability.
Ramanujan had a number of conjectures in regard to this
formula and one is still unproven. He made this con
jecture in a problem he submitted to the Indian Mathematical Society.
The asymptotic formula is used, for instance, in the popular problem:
What is the minimum number of people you can have in a room so that the
probability that two share a common birthd
ay is more than half? I think it is 21, 22 or 23. Anyway, this problem
can be generalised to many other types of similar problems.
Have you looked at the lost notebook?
That is what I am working on now with Andrews. It contains about 630
results. About 60 per cent of these are of interest to Andrews. He has
proved most of these results. The other 40 per cent are of great
interest to me as most of them were a continuatio
n of what Ramanujan considered in his other notebooks. So, I began
working on them.
What are your experiences of working on Ramanujan's notebooks? Do
you think Ramanujan was a freak or a genius or he had the necessary
motivation to write the notebooks?
I think one has to be really motivated to do the kind of mathematics he
was doing, through either teachers or books. We understand from
Ramanujan's biographers that he was motivated in particular by two
books: S. L. Loney's Plane Trigonometry and Carr's
Synopsis of Elementary Results in Pure Mathematics (which was a
compilation of 5,000 theorems with a few proofs) at the age of 12. How
much his teachers motivated him, we really do not know as nothing about
it has been recorded. Reading these book
s and going through the problems must have aroused the curiosity that he
had and inspired him.
He is particularly amazing because he took off from the little bit he
knew and extended it so much in so many directions, leading to so many
new and beautiful results.
Did you find any results difficult to decipher in any of Ramanujan's notbooks?
Oh yes. I get stuck all the time. At times I have no idea where these
formulae are coming from. Earlier, Ron Evans, whom I have already
mentioned as having worked on Chapter 14, helped me out a number of
times. There are times I would think of a formula
over for about six months or even a year, not getting anywhere. Even now
there are times when we wonder how Ramanujan was ever led to the
formulae. There has to be some chain of reasoning to lead him to think
that there might be a theorem there. But ofte
n this is missing. To begin with, the formulae look strange but over
time we understand where they fit in and how important they are than
they were previously thought to be.
Did you find any serious errors in Ramanujan's notebooks?
There are a number of misprints. I did not count the number of serious
mistakes but it is an extremely small number - maybe five or ten out of
over 3,000 results. Considering that Ramanujan did not have any rigorous
training, it is really amazing that he
made so few mistakes.
Are the methods of mathematics teaching today motivating enough to produce geniuses like Ramanujan?
Some like G. E. Andrews think that much of the reforms have come about
because students do not study as much. This, along with the advent of
computers, has changed things. A lot of mathematics which can be done by
computations, manipulations and by doing
exercises in high school are now being done using calculators and
computers. And the computer, I do not think, gives any motivation.
The books on calculus reform (that is now introduced in the U.S.)
include sections on using a computer. To calculate the limit of a
sequence given by a formula, the book says press these numbers, x, y and
z... Then there appears a string of numbers that
get smaller and smaller and then you can see that is tends to zero. But
that does not lead to any understanding as to why they are tending to
zero. So, this reasoning, motivation and understanding of why the
sequence tends to zero is not being taught. I
think that is wrong.
There seem to be two schools of thought: one which thinks that the
development of concepts and ideas is important and the other, like that
in India, which thinks that development of skills is important in
teaching mathematics. Which do you think is more important?
I think you cannot have one without the other. Both must be taught. The
tendency in the U.S. is to move away from skills and rely on computers. I
do not think this is correct because if you have the skills and
understanding, then you can see if you have
made an error in punching in the computers. Andrews and I have the
experience of students putting down results that are totally ridiculous
because they have not understood what is going on. They do not even
realise that they made mistakes while punching
in the computers. So, developing skills is absolutely necessary. But on
the other hand if you just go on with the skills and have no
understanding of why you are doing this, you lose the motivation and it
becomes just a mechanical exercise.
However, even now there is a possibility that geniuses like Ramanujan
will emerge. It is important that once you identify such children, books
and material should be found for them specially. The greatest thing
about number theory in which Ramanujan work
ed is that you can give it to people of all ages to stimulate them.
Number theory has problems that are challenging, that are not too easy,
but yet they are durable and motivating. A foremost mathematician (Atle
Selberg) and a great physicist (Freeman Dy
son) of this century have said that they were motivated by Ramanujan's
number theory when they were in their early teens.