Written: Nov. 20, 2015
As some of you may know, I believe that scientific ("peer-reviewed") journals (both print and on-line) do no longer accomplish their original purpose of transmitting knowledge. This is much better done via arxiv.org and one's personal website. As far as novelty and correctness, the arxiv is great. It is read by many more people than any journal, and if you post an article that has an error, or is a rediscovery, people tell you soon enough. Luckily, so far, I have not posted any erroneous articles, but I recently offered to donate 100 dollars to the OEIS foundation for a proof of a "conjecture" that turned out to have been already solved, see here, and once it was pointed out, after one day I posted another version, acknowledging that the conjecture has been solved before I posed it (and I still made the donation). Another example, is this gem, that turned out to be old hat, and for which we also posted a second version acknowledging that it was a rediscovery.
Unfortunately, peer-reviewed journals (including those published by notorious commercial sharks, like Elsevier), will not disappear as soon as one may hope. They serve the "important" sociological need of humans for inclusion and exclusion, just like lunch tables in high school. They are edited by (often mediocre) "editors-in-chief", or "managing-editors", and usually have a distinguished editorial board, who are rarely (and often never) consulted. The more "selective" ones often reject papers without even sending them to referees, just because, they claim that they are "of insufficient interest to the readers" (as though anyone actually reads their journal, their only use is for academic red tape.)
A case in point is the JournalNameRemoved. Recently, my collaborators (stupidly, in hindsight) submitted this masterpiece, and it was rejected outright, with the following rejection slip, signed by NameRemoved's henchwoman, one AnotherNameRemoved. Apparently, NameRemoved is too important to send rejection emails by himself, and leaves this chore to an editorial assistant. As a matter fact, I have some doubts whether AnotherNameRemoved is a real person, perhaps it is an automaton, who automatically emails a rejection letter. They are probably all the same, e.g. by googling I immediately found a blog with an almost identical rejection slip.
It is very sad that, even today, that an editor can outright reject such an interesting paper, that is not only computational, but has interesting theorems for humans, and that is not "computational" in the sense of "number-crunching", but develops symbolic-computational algorithms that can generate many theorems, far deeper, (and may I say, more interesting) than what any human (for example, NameRemoved), unaided by computer, can ever hope to find. For example, for any prime, p, the following statements are true:
These are not just "numerical data", but genuine deep theorems, valid for any prime p.
Of course, the sequence of Motzkin numbers is just one example. Analogous statmentes can be automatically discovered and automatically proved for a wide class of sequences. There are quite a few, humanly-generated, papers in JournalNameRemoved that only treat one specific sequence at a time.
But the rejected paper has also a beautiful general theorem (this part was done by Qing-Hu Hou, and I really admire it) proving finiteness, and is independent of any computers.
Number theorists, especially those that are editors of journals, should get rid of their computer-phobia, not unrelated to a previous opinion of mine.
It was not my idea to submit to JournalNameRemoved, but my collaborators', but after I found out about the rejection, I browsed some of the current papers that were accepted. I won't mention any specifics, but I am sure that the members of the distinguished editorial board (who probably never look at it, they just kindly lent their names), can easily spot quite a few accepted articles that are utterly trivial (in the usual, human sense of the word). That's another (lesser) reason that it is time to relieve NameRemoved from his editorial duties.
However, the main reason why NameRemoved should go, is that he is a prejudiced against symbolic-computational work, that he does not understand, and hence does not appreciate