Who Reads Your Papers?: Achievement vs Success, Part I

December 2, 2009

Who Reads Your Papers?: Achievement vs Success, Part I

By Michael D. Fried

Andy Magid reviewed his career as AMS editor by recounting  large editorial changes in the Notices, starting with the January 1995 issue [3].

Magid’s recounting – contrary to what one might have expected, for the changes were strikingly positive – shows almost all of today’s format was in place in that first issue. He also reminds readers that the Notices has always welcomed submissions from readers. My articles [1] were examples.

Another AMS change

Hugo Rossi spearheaded these changes. Magid didn’t mention that Rossi also championed, relevant to our theme, discontinuing the Research Announcements portion of the Bulletin. Roger Howe, its chief editor, instituted changes in the years proceeding. There was something lost and something gained in discontinuing it.

Lost (to the community) was a direct and successful approach to timely refereeing on technical, but promising, papers.

Gained (by the AMS) was no longer having to account for what appeared in the Research Announcements. Especially that it was unclear if the Announcements constituted an authoritative reference: it didn’t demand complete proofs.

Howe’s method was to encourage the editorial staff to do the brunt of the refereeing. This never precluded using specialists in a particular paper’s area. Still, that was to be corroborating evidence. Each paper was assigned two editors, known – I think – only to Howe.

Refereeing is usually a hidden process, theoretically to protect opinions from coercion. Yet, it also protects poor refereeing.

Among a collection of changes

I recall these two AMS changes as pieces of three overwhelming university pressures that started in the late ’70s.

1. Universities raised their standards on what they considered adequate/exemplar publication rates. Deans more stringently oversaw the evaluation process to enforce these.

2. Libraries archived – out of sight – large portions of previous research publications. Schools felt pressure to drop publications used only rarely.

3. Departments saved staff money by dropping faculty typing services. The TeX revolution influenced every publication aspect between 1988 and 1995.

These changes continue. Steven Krantz, see below, was early to speak on items #2 and #3, with concern.

How those institutional changes affected individuals

Many of my age group – my PhD was in 1967 – exhibited certain characteristics in responding to these changes.

While many from an older generation did type, most from my generation did not in our early careers. Advisers’ grants or personal funds paid the typists of our theses. It was painful to correct papers by hand, and then wait for ages to get back our changes, often with new mistakes.

Even my early electronic publications (from 1988) – first in Microsoft Word, then in now unacceptable TeX formats –  don’t generate an acceptable pdf file. A  considerably cited first paper is not on JStor. Even if it was, that is not the accessibility of the arxiv.

My earliest regenerable papers from extant typing appeared in 1991. Like some, I’ve taken seriously putting pdf scanned versions of our work on-line at my website. Papers between 1968 and 1991 (half my volume) are otherwise realistically unavailable.

Does it matter? After all, it is old work!

Doris Lessing, an acknowledged public intellectual, lamented in the 1970s:

“One shame of the 20th century will be all the unused research remaining on the shelves.”

Are results lost in long, difficult papers? That doesn’t explain how little the community shares knowledge of work. Funding panels often know an author’s reputation disproportionately better than their work. This is even after they have been charged with reading reports: another referee problem.

Even comments on paper shortcomings – when relevant – ought to be acknowledged. They can’t be as worthless as the silence they receive suggests. For one, there is how often new papers attempt to reprove what it turns out are old results.

Economic ideas gained force during the transition time we are discussing. Legislatures told universities to be more like businesses. So, the university also told us faculty. That increased research competition;  even the lip-service of our shared burden to the (mathematical, I presume) community vanished.

The repository problem

Yet, it is not the community health that concerns me. Rather, it is the health of the papers – each a claim to achievement – once they are accepted in print.

Nothing makes for healthy – successful – papers than that they are read. Maybe not in entirety, but read significantly in parts. We should all care that there is a repository, beyond the author, of – at least – partial knowledge of most papers in print.

Such a repository would start with referees, reviewers, adherents of the areas, and extend amorphously to those who have heard the results at colloquiums and seminars. Do Math Reviews and the promise of Google suffice? Check them yourself, they don’t yet.

Most authors work excruciatingly hard to get their papers in print. Yes, those papers may manifest varying breadth, depth, problem standards, and originality in technique and aspiration.

Some of these correlate to authorial mental and physical energy, and some  correlates to luck. One piece of luck is getting a good referee; someone who contributes momentum to the endeavor.

The theme of [1] is that a good referee is hard to find. Is this in question? I hear it from editor after editor. I also see that editors are hardly demanding about the quality of refereeing.

As in [1], my experience is that there are gaping errors, or failures to properly document, in about half the papers I have refereed. Many correctable to lesser results, but … Often the results are already in print, maybe unknown to the author, but … A nontrivial portion seem so easy that they could be rediscovered by anyone forced to address the issue.

Also, many times it takes an astute referee to figure that some authors aren’t quoting their references significantly. Many times they aren’t quoting them at all, but rather using them to prime the editors for referee sources.

This list doesn’t hit all places that could use a disinterested referee. As a referee I’ve started a fix to many difficulties, just by pointing to them.

Journal changes

The AMS wasn’t the only organization to institute striking journal changes in the early ’90s. First, journals proliferated, as did AMS publications. The proliferation may directly result from authors able to do their own typing.

The work of creating each separate paper is simplified, though instituting technical changes in formats adds work and calls for developing new talents.

Another journal development, though, leaves me puzzled. That is the growing list of editors on many journals. One driver must be the need for faculty to accrue accolades; one is being an editor.

The problem, however, is that the harder task is refereeing: that remains inglorious.

[1] suggestions that exacting technical expertise is rarely a necessity, nor even the key, to good refereeing. Still, desiring to be a good referee, one who can get into work outside one’s own exact papers, is. It is a neat attribute. It should be encouraged.

One encouragement would be to make it a part of the editor’s job, as I mentioned about for the now defunct Research Announcements.

That was an experiment, and it worked. It shortened a backlog, and it broadened the papers represented in that section. Running such an editorial staff requires an active Chief Editor. Yet it calls for less running after referees. As I said in [1], there is something askew with huge editorial staffs that do little editing.

Elaine Kehoe [2] introduces our new Notices editor. Since Steven Krantz is quoted more broadly, I use his concluding sentences from [2].

“I think that one of the keys to success in life – not just in academics … – is to find the means to periodically reinvent yourself. If you don’t, then you atrophy, you want for ideas, and your creativity dries up.”

To me that says one must continue throughout a career to take risks, going outside your present comforts. Even if few would/or should take great risks, most err on a side so conservative that Krantz’s prophecy seems manifest.

There are precedents in a heightened referee role. Promising new directions start from inspirations gained from being a serious referee.

Many claim that serious refereeing is a task for those who can’t do research. That is, anyone can do it. It would be consistent if the same people believe that actually reading, and using substantially, papers well is a similarly easy task. (It is after all, akin to refereeing.)

Part II of this topic will counter this in support of Krantz’s statement. I won’t plead to all mathematicians to referee better. Rather, I will suggest changes through journal policies to encourage entrepreneurship.


[1] Michael D. Fried, Should Journals compensate Referees?, May 2007 Notices of the AMS, Vol. 54 (2007), No. 6, p.585.  Also, The Uneasy Relation Between Referees and Editors, October 2007 Notices of the AMS, Vol. 54 (2007), No. 10.

[2] Elaine Kehoe, Steven G. Krantz Appointed Notices Editor, December 2009, Notices of the AMS, p. 1445.

[3] Andy Magid, Anniversaries, December 2009, Notices of the AMS, p. 1407.