Saturday, October 04, 2014

The mess we are in with scientific publishing

Next post in this series

I am a little out of the Italian academic scene since my retirement, but if I understood the recent rounds of hiring in Italy they occurred in the following way. It was possible to apply for  "abilitazione", that is a judgment was made if you were at an appropriate level to hold a post. This was made in most cases on the grounds of purely numerical indicators, and was supposed to be a threshold. It was not a competition. A deeper analysis, actually looking at the papers or asking experts who had read the papers, was clearly impossible since, for example, in Computer Science a committee of 5 had to judge in a year the qualities of approximately 900 applicants. (There were even complaints when more than numbers were used, since that gave power to the "barons".)
After this judgment a great number of the abilitati were given permanent posts, thus filling up vacancies for some time.

The pressure this type of thing puts on scientific publishing is enormous. Referees, while trying to make judgments on papers, now have to consider that they are deciding the careers of young people, the grants for older people, that the prestige of the journal will affect jobs and grants. The reason the pressure is so enormous is that the job and granting committees don't look at the papers, just the numbers.
The job of a referee has become impossible, at the same time that there is more and more need for referees since scientists are being forced to publish more and more.

Scientific publishing must free itself from these pressures.

Thinking about this situation brought back to mind some thoughts of Bernhard Neumann.

He started a journal, The Bulletin of the Australian Mathematical Society, in 1969 with very specific ideas. I found an account of his ideas in an obituary by Cowling, Jones, Morris, Oates-Williams, part of which I reproduce here:

"... in his editorial Bernhard made it clear what his future policy would be. He was unhappy not only at the length of time taken by many journals to publish a paper, but also at the way some authors seemed to expect the referee to do half the work for them. He decided that there was a place
for a journal that would guarantee swift publishing; and would do this by making rapid decisions on each paper, and not accepting more papers that would fit into the next issue. And, most importantly, not accepting any papers that were not in publishable form; he felt that a paper that required more than half a day's work from a referee was not ready to be published."

His idea was that the responsibility for the paper was in the hands of the authors. Referees made only a judgment that the paper was scientifically appropriate - not even that it was correct - that was the authors' problem. It was not the position of referees to put personal manufactured obstacles in the way of  publication, as happens so often today.

Of course this project failed. The journal soon became a normal journal, but it seems to me that if ever the scientific community manages to escape from the stranglehold of  Government-promoted  numerical evaluation, then Bernhard's ideas, arising from his experience of publishing in his youth, might be incorporated in a new system of scientific publishing. Bernhard's proposal was made before internet and to some extent arxive now plays the role of the journal he had in mind but without the " half a day's work from a referee" for each paper.

Update 11 October 2014 I have just discovered from David Roberts' Google+ blog that there is a journal with very similar aims to Bernhard's. It is called PLOS ONE and, commencing in 2006, it has become the largest scientific journal.
Update 18 October 2014 From a more recent post on David Roberts' blog it seems that there is very little chance for publishing mathematics in PLOS ONE.

Labels: ,


Post a Comment

<< Home