A letter to the editors of Mathematical Reviews

Dear Editors,

This is to notify you that I will not write any more reviews for you.

I have been a reviewer for 12 years and contributed more than 100 items. Each time I prepared my reviews carefully, trying to render the best service to the community by highlighting the papers' most important ideas and often pointing out interesting statements hidden in minor remarks.

In return, almost all of my publications were reviewed by incompetent or careless reviewers. One of my best papers which prompted a considerable amount of later research hasn't even been reviewed, you just selected a portion of the introduction which does not contain the main results. Wouldn't have it been easier to ask the author?

And now I see that my most recent book which is already used as a major reference in the subject did not merit more than a few lines from a reviewer who obviously did not read more than the blurb. Wasn't there an editor to say this is insufficient? How can you expect devoted work from your reviewers if they are treated as authors in such a way?

You may decide to send me a reply explaining that you are dealing with tens of thousands of reviews at a time and cannot possibly pay attention to everything. This is true, but still I can't understand how you can be so careless. If I as a user know which are the informative reviews and which are not, you should have at least a faint idea of it and ensure that important pieces don't get in the hands of good-for-nothings.

I am sorry to part from you with such a sour note but you must understand my disappointment about the way you are "honouring" honest work.


Tamas Szamuely


A comment: Actually, I was a bit unfair to my reviewers; initially they were quite correct, it is only in the past five years or so that I observe a serious decline. And alas, I think it is a phenomenon that concerns not only myself. I am surprised to see that in recent years many groundbreaking papers and major monographs have been reviewed by people who have neither the required competence nor the required care to do so. True, serious people have less and less time to write reviews as they are overwhelmed (like me) with obligations to write other reports, project proposals etc. But the problem is that MR reviews are read by thousands of people as opposed to the original papers, so they then necessarily form an opinion of somebody's work on the basis of a few lines by a second- or third-rate colleague. This was not so in the old times when many first-rate experts contributed reviews which were themselves often excellent pieces in mathematical writing. There was an attempt to commission featured reviews for a selected few of papers of major importance but the experience was halted after a few years.

Or am I mistaken? It is true that I am using MathSciNet more and more as a search engine, without reading the actual reviews. So maybe they should stop publishing reviews at all; publication lists with abstracts and links to the original papers will suffice. After all, it is easy to find most papers online nowadays -- this is less so with books. Then someone someday may realize that there is a need for succinct summaries and perhaps also criticism of important and difficult papers and books by genuine experts, and such a review journal or online resource will be established.