I have never published anything in the JCM and had never been involved

in its publication and therefore have no personal interest in its existence.

I wish to say a few words about the scope of the JCM. It seems to me

that Terry Lyons’ criticism of the Journal is based on an arbitrary

assumption that the Journal should publish quality papers from so-called

“experimental” mathematics and from numerical mathematics. But the JCM

is unlikely to be competitive in these areas are because there is plenty

of journals around already covering them.

But the JCM plays, and is in position to continue to play, a unique role

as the leading journal in specific, increasingly important,

cross section of pure mathematics: symbolic and/or absolute precision

arithmetic computations, frequently at the limits of capacity of modern

computers, in algebra, representation theory, combinatorics, number

theory, algebraic geometry, etc. The role of mathematical work of this

type will only increase in future.

Let us take look at a very recent paper which I picked on the very first

page of the JCM site hosted by CUP because the subject appeared familiar

to me:

John Ballantyne, Chris Bates and Peter Rowley, The maximal subgroups of

E_7(2), LMS J. Comput. Math. 18 (1) (2015) 323-371.

A few general words about the area : finite groups appear in mathematics as automorphism groups of mathematical structures. In particular, they **act** on sets. The so-called **primitive** actions are building blocks of actions of finite groups on sets. The so-called **simple** groups are building blocks of groups. To make Classification of Finite Simple Groups, one of the key results of 20th century, fully usable across mathematics, one needs to know primitive permutation representation representations of simple groups. It is an elementary fact that description of primitive permutation representations of a finite group is equivalent to description of **maximal subgroups** in the group. **E_7(2)** is one of finite simple groups; it is big, it contains contains about 10^40 elements. It is a very big haystack to search for the proverbial needle; 10^40 straws have the same mass as 1 million Suns. In the paper, a much more difficult problem is solved: in a haystack of that size, specific symmetric configurations of straws are identified, and fully described.

And now is my assessment of the paper:

- The result may appear to be narrow and specific, but it will stay

relevant for decades to come, because it is about a concrete, eternal,

and important mathematical object. - Methods are an intricate intertwining of subtle and difficult mathematics with hard but subtle computations.
- It is hard to find any other journal to publish this paper together with 51 MB (3.36 MB in compressed form) of electronic files supporting the calculations.
- Two younger authors of the paper are brilliant young mathematicians with
**proven ability to compute in areas where no mathematician has computed before.**In my opinion, this makes them pretty special, As it has already been said in this discussion about similar situations, peer-reviewed publication of their papers is crucially important for the future career of young mathematicians.

This small case study explains, I think, why the JCM is a unique and special journal and why closing it is not in the interests of mathematics and, in my humble opinion, not in the interests of the nation.

To conclude, a comment from Rob Wilson (placed here with his kind permission):

I am perhaps at the opposite extreme, having published ten papers in the

JCM over the past 14 years, with another one in press at the moment.

Some might say this is a ‘conflict of interest’ that disqualifies me

from speaking, but I fully endorse what Sasha has said about the type of

computation and mathematics that is represented in the JCM. It is a much

misunderstood and even despised area of mathematics, butin fact is hard

core computation on the edge of the impossible.

## Leave a Reply