Lasting literature and Von Neumann

Kumar Vadaparty (kumarv@bellcore.com)
Mon, 15 Apr 1996 11:46:25 -0400

(Note: This mail been longer than intended; if I had more time,
I would have written shorter note (as Pascal once said)).

> Sitaramayya Ari says:

> Should good poetry or literature really last long?
> Can't good literature be short lived?
> Is long-lived literature necessarily good literature?

I have seen similar questions on this net several times.

What is interesting is that these questions
can be examined in the context of almost any creative art you can think of,
mathematics and ComputerScience being the ones I am most familiar with. And
furthermore, these questions HAVE BEEN examined in the context of these
arts, mathematics being the most examined (in my experience/exposure).
I think the solutions/answers
arrived in those quests definitely provide an answer or at least some
framework to our questions here.

In mathematics the questions are "Should good mathematics also be practical?"
which is very similar to Ari's qeustion of "social committment" of poetry. The
spectrum is very wide and the answers very different for different people.
Let me give two extreme viewpoints on mathematics, and two extreme view points
on telugu literature, and then on computer science.

On the whole, I tend to agree with Ari.

I think the best answer was given by Von Neumann, and we can apply his answer
even to Telugu poetry. (you will see).

1. Hardy (of hardy ramanujam fame) wrote a book called "A mathematician's
apology" in which he belabours on how un-necessary it is for mathematics to
be of use. Indeed, he almost makes it as if useful mathematics is rather
ugly (i.e., not aesthetic).

2. Von Neumann takes a somewhat middle road. I am now quoting him (source
William PoundStone's "Prisoner's Dilemma")
"... [ pure mathematics] becomes more and more purely aestheticising,
more and more purely l'art pour l'art. This need not be bad, if the
field is surrounded by correlated subjects, or if the field is
under the influence of men with an exceptionally well-developed taste.
But there is a grave danger that the subject will develop along the
lines of least resistence... the discipline will become a disorganized
mass of ...complexities...
At the beginning the style is usually classical [ with some emperical
source], [but, slowly becomes] baroque and the very high baroque"

I can not think of a better abstract description of the degeneation of
prabhandha kavitwam that has more discussions on boobs than people. Porno
per se is not bad (as in kamasutra) where the context warrants it, but a
down pour of stock phrases describing female body-parts independent of the
context is at best a sign of baroqueness.

To a large extent, I agree with Ranganaikamma (muppalla) who pointed out
that Laxmana, when grieving of Sita's disappearance, laments on her
beautiful Thighs (and, he is supposed not to have seen all but her feet).

Mathematics also goes through similar problems, and so does Computer Science.
Lasting value in computer science is often measured by "Triple Subsscripted
MATHEMATICAL expressions" however un-necessary or un-wanted they may be.
A number of completely irrelavent proofs are produced in many conferences
and journals (to give examples from my own-sub-sub field: functional dependency
theory, negation in logic programming).

Just as sanskrit plays the hallmark
of respect/classism in telugu poetry, triple-subscripted-mathematics plays
the same role in computer science Ph.D. theses (I served on several committees
of these), and papers (I reviewed journal papers and served in conference
committees). If you are interested in the abuse of the so-called
"mathematics" in Economics you may want to read "Imposters in the temples".

In the beginning of a field/subfield, the style is usually classical, as
Von Neumann says.
The presentation of Codd's theory on relational databases can be compared to
the nice poetry of Potana (simple, easy to understand, enjoyable, and perhaps
with practical use). Another example is the thesis of Shamos "Computational
Geometry". With one thesis he started an entire field. I'd kill to be
the author of such a Ph.D. thesis.

However, later on, so many people jumped on to the bandwagon to make it
really baroque, infertile, and plain boring.
For example, the prabhandha-kawitams (e.g., nandi timmana'(?)s
parijatapaharanam where he uses so many stock-phrases to describe
satyabhama, I really got sick of them ... the famous poem ..."ana vini
vetuvadda uragangana vale...")
The examples of baroqueness in Computational geometry and Databases similarly
are "musium keepers problems" and "functional dependency theory" and negation
in logic programming. These topics have been killed to death, I wish someone
put an RIP to them...

Thus, what I 'd like to say is that there are at least 2 metrics that are
orthogonal to each other:

1. Aesthetics
2. Practical use (or, social commitment).

I am not sure which is better than which. It's best to have both: like writing
a paper that has both theorems and practical use. Gurajada's kanya sulkam
is an example of this. He himself says this in the preface: he was inventing
a new style (providing respect to the so-called crass telugu) and trying to
curb a bad social practice. What can you want more as a poet? It's nirvana
to have both social committment and aesthetic value.

To judge aesthetic value is very very hard. Even for the scholars and experts.
For, it is very subjective. I can only say that I think Potana had the
best of aesthetic values in telugu poetry that I know of.

Kumar.