It is easy to get frustrated and impatient with the Editors of scientific journals to which research articles are submitted for publication. We authors often develop some funny paranoias. Suddenly we become convinced that Editors are out to get us, intent on enacting a sordid conspiracy orchestrated by few “powers-that-be”, aimed at keeping our brilliant and unconventional ideas out of the science elite, thereby denying us the recognition that we so richly deserve…
Except for my own case, to which the above description applies to the letter, reality is of course very different. The job of Editor is difficult. One is called upon to make decisions on the suitability of an article for publication, typically based on scant information, ambiguous and hard to decipher to boot.
Manuscripts are very technical in nature, and Editors, albeit themselves scientists with a research background, cannot be realistically expected to be competent in every subject of a broad field such as, for instance, condensed matter physics. They necessarily have to seek out recommendations from other scientists (referees), whom they attempt to select based on their specific expertise, ideally enabling them to provide an informed, reliable opinion on the work described in the manuscript.
In order to make referees feel free to express as candid and objective an assessment, their identities will not be disclosed to the authors, in turn granting referees a relatively significant influence over the fate of that manuscript. Abuse is certainly possible. For any given subject of research, only relatively few such scientists exist, and the likelihood is high that they will be in direct scientific competition with the authors of the manuscript to review.
In the presence of such an inherent, and almost inevitable conflict of interest, the Editor must be able to assess the objectiveness and unbiasedness of the reports that referees provide, as these persons may have an axe to grind.
Identification of referees, acquisition of reports and decision on the manuscript ought to take place in a relatively short time, as the content of a paper submitted for publication owes much of its potential interest to its timeliness, and because claims of priority of results will be made based on the date of publication. Editors are required to try their best to make that happen, but the sheer number of articles submitted daily makes the task quite challenging .
Ideally, of course, none of this would matter. While publishing it on a more prestigious journal certainly will give the work greater exposure, with the ensuing augmented recognition for the authors, ultimately its real long-term impact must be assessed by measures such as the number of citations. In principle, a solid piece of work which many other scientists find illuminating and/or ground-breaking will be dug out no matter where it is published, and cited accordingly. In practice, in our highly imperfect world things do not quite work that way, and the initial publication venue often determines much of the impact of that article, mainly due to the simple difference in readership . Thus, many a potentially important manuscript simply languish ignored, buried in some obscure journal, their authors receiving little or no credit for their work.
If I were the Editor, I would be of course no better than anyone else; I am sure I would make the wrong decision many times, often based, more or less unconsciously, on my prejudices and misconceptions, and as a result a fair number of manuscripts would receive a less than objective and fair treatment. I am afraid that some of that is really inevitable, even though of course, in this profession too, there are good and bad ones.
Still, there are a few things that I believe I would purposefully try to watch out for and/or avoid:
Of course we are all busy, of course it takes some time to go through a manuscript carefully one or two times, and of course we all wish to receive useful feedback expressed in the form of a reasoned, informed and well-written report. Therefore, we have to be willing to wait. But the question is, for how long ? Two weeks seems a reasonably adequate time span.
It is simply not fair to the authors to allow lazy, uncooperative, or worse yet ill-meaning scientists to play dirty, stalling indefinitely publication of the work of a competitor. In order to minimize the likelihood that after two weeks the Editors find themselves with no reports, several avenues can be exploited, e.g., a) send the manuscript out to way more potential referees than the minimum required b) allow authors to include in their submission endorsing statements from scientists working on the same subject and “at arm’s length” from them  c) seek out independently informal assessments on the quality of the manuscripts (for example, endorsements of the manuscript uploaded on ArXiv).
We are all human, and often times referees will take advantage of anonymity, as well as of their temporary position of power, to push their own agenda, promote their own work, or simply lash out at a competitor. Examples of this behavior include gratuitous, disparaging comments about the work, possibly phrased using a sarcastic tone, disseminated all over the report. These comments hardly promote the type of reasoned, useful scientific exchange that Editors wish to see between authors and referees, which may ultimately lead to an improvement of the manuscript. Rather, they are likely to be followed by angry responses from the authors, in turn making it increasingly difficult for the Editor to make an informed decision, as the tone of the discussion quickly degenerates into a fight.
One way to avoid this is for the Editors to redact reports, eliminating sentences that do not add to the scientific content but are simply meant to antagonize the authors. It should be noted that Editors will often time request that the authors exercise restraint, after receiving from them a rebuttal to a referee report phrased in a language deemed hostile. If one is going to demand restrain of authors, one should a fortiori do so with anonymous referees. So, before hitting the “forward” button and inflict upon authors patronizing, derisive and unfair reports, why not read them and edit them ?
Also, Editors should be cautious with reports “suggesting” that the article be “improved” by adding citations to the referee’s own work. While such requests are often legitimate, sometimes anonymity paves the way to (shameless) self-promotion.
Often times Editors give authors the impression of essentially ignoring the actual content of a report, merely going instead by its overall recommendation (“publish” versus “reject”), even though the motivations adduced by the referees are nebulous, dubious, or downright bogus.
While it is certainly commendable that high standards be set for authors, who should be expected to state their case precisely and convincingly in the manuscript as well as in their rebuttal(s) to referees, it is simply unfair to allow referees to get away with reports consisting of little more than a one-liner, or unsubstantiated claims, e.g., stating without citing any relevant reference that the work performed by the authors overlaps too substantially with that previously done by others, and more generally with reviews ostensibly underlain by sloppy and uninterested reading of the manuscript. Such reports ought to be weighted appropriately by the Editors, who at some point should discard them altogether if they do not meet accepted, minimum standards of scientific discourse. This is especially the case if the tone of the report remains vague, scarcely quantitative and dismissive even after a first, detailed rebuttal by the authors. It is unfair to regard that as a mere “difference of opinions”; it is unfair to allow referees to get away with half-assed reports and expect authors to respond, spending hours and hours (re)stating their cases — opinions that are articulated more substantially and effectively should count more.
For, if ultimately the Editor only cares about whether the referee will go with “accept” or “reject”, and the final decision boils down to establishing whether the “yays” have a substantial majority over the “nays”, then there is no need to seek a report. It is far more efficient to and surely a huge time saver to request just that of referees — is it a “yay”, or a “nay” ?
 It would be desirable that articles written on specialized subjects be sent to specialized journals. Unfortunately, the relatively low Impact Factor of specialized journals renders them scarcely appealing to authors, who by and large prefer to submit their work to more general, widely read journals, whose Editors find themselves overloaded as a result.
 This is the main reason why some of us (well, maybe just one of us) would like to see all refereed journals simply go away, to be replaced by a single manuscript archive such as ArXiv, possibly with some minor modifications.
 For example, scientists never having collaborated with the authors and/or based at different institutions. This seems particularly appropriate when there is a chance that the person endorsing the work may conceivably be selected as a referee.