Peer Review Anonymity

Alexander Berezin berezin at MCMAIL.CIS.MCMASTER.CA
Fri Nov 3 09:51:57 EST 1995

On Fri, 3 Nov 1995, Simon M. Brocklehurst wrote:

> I agree with the points the Graham has made in his posts on this
> subject.  I suspect that the silent majority would also agree
> with this position.

> Dr Berezin apparently believes that most published science
> is garbage (to quote him on one of his earlier posts).
> If this were true, then it might be worth trying a new system.

There is a huge range of opinions on this subject (what
share of the published science is "garbage"). Also, obviously,
you can find variations (also huge) about what constitute
garbage. There is a very big litereature of the subject.  

> The _fact_, however, is that there are many papers, published in
> good journals, that are of high quality.  

I don't deny this point. Even if 90 % of all published work
is garbage the remaining 10 % is still a very impressive output.
Scientific paper is published (world-wide) every 6 seconds.
That means a new "good" paper is published every minute. 
Not that bad. So, calm down on my estimate.

> Thus
> it seems that apart from any other arguments, the system
> of anonymous peer review works serves the scientific
> community quite well.

Here, I believe, you make logically unsupportbale inference.
You say this (good papers) are published BECAUSE anonymous
peer review works well. I can make exactly the opposite inference
from this "fact" [ which is "good papers ARE published": we
both agree that this "fact" is proven ], namely that:

"good papers are published IN-SPITE of anonymous peer review,
because 10 % somehow manage to slip through it" 

My inference in no way worse (or better) than yours.
> Dr. Berezin might like to note that we do in fact already have
> widely read tests of systems where peer review
> is not used as the main criteria for acceptance of manuscripts.
> For example when editorial teams (and a small number of
> referees) play a major role in the decisions (e.g. in Science and
> Nature), standards can often be lower than they should be (i.e.
> unoriginal work or flawed work is often published).  

THis is what I say: there are of course various ways around
peer review (in a standard journal APR meaning). This argument
though contributes rather little on pro-APR side. And we
all very well know that APR is NOT a gurantee against "flawed
data, unoriginal work, etc". An example recently was quoted
when the same paper published few years years ago (theoretical 
physics) was EXACTLY plagiarized by (somewhat "naive")
ambitious gradutate student and republished by another 
prestigious journal (the editor latter published an apology
to the original author). So, it did not occur to any of
anonymous peer reviewrs ("experts in the field" !) that 
paper was already published long ago. So good for APR.   

> This is 
> because these people are (not unexpectedly) simply unable to judge 
> what is novel and/or of interest in the wide range of fields that
> their journals cover.

Unfortunately, ability to judge (predict) what is really
novel or potentially importnat is largely a chimera.
Enourmous case literature on this. One quote (which I 
already gave earlier): John Ziman, Prometheus Bound,
Cambridge Univ.Press, 1994. 

> -- Simon

More information about the Bioforum mailing list