Correspondence with Cornell re the unsatisfactory management of the physics preprint archive

Disingenuous answers, not reasoned response, the norm in response to
concerns of arxiv's contributors

A sad day for science, a sad saga for Cornell!

Introduction

The physics preprint archive (arxiv.org) is a publicly funded facility with two functions.  It is both an archive, and a mechanism for facilitating communication within the physics community.  Registered users log on and upload preprints, which are moved to the publicly viewable area of the archive on a daily basis after moderation.  For the convenience of users, the archive is divided up into a number of sections and subsections.  Some regulation of content is clearly necessary, and so there is a moderation process.  Such a process can be done well or not so well.   The cases documented on our web site archivefreedom.org suggest that the moderators frequently step in on the basis of finding particlar topics unacceptable, rather than for any defensible reason.  Those who run the system are significantly less tolerant than what is the case with the scientific community as a whole; some papers that have been published in refereed journals have been barred from the archive, while endorsements by experts outside the system have been overruled by moderators.  Publication in a referreed journal indicates that papers have been vetted by experts while papers in the archive have subjected only to rudimentary checking, and this differential is the reverse of what one might expect.

Worse, the archive has no effective complaints mechanism.  In principle, one can send a complaint to a moderator address, but any such complaints are met by bland (or on occasion facetious) replies, sent anonymously under cover of an arxiv alias.  A typical, frequently employed, disingenuous response is 'you have available to you many other channels by which you can communicate your ideas' (disingenuous because none of these other channels is in any way comparable to the archive).

One can attempt to bypass the system by writing to Cornell library staff, since the archive is officially a service provided by the library.  The first time I tried this, some years ago, I had the misfortune to encounter an apparatchik and got nowhere.  More recently, irritated by having had a paper categorised as 'unsuitable for crosslisting'.  I tried again and was pleased to encounter what appeared to be a human being.  Now read on ...



A promising start, after I had written to a colleague to request endorsement for the quant-ph section of the archive:

from www-admin@arxiv.org, Aug. 24th at 11:08 (all times GMT)

You've just been endorsed to submit papers to the arXiv archive quant-ph
(Quantum Physics). Visit

http://arxiv.org/submit/

to submit papers. Endorsement is a necessary but not a sufficient
condition to have papers accepted in arXiv; arXiv reserves the right to
reject or reclassify any submission.

But then ...

from www-admin@arxiv.org, Aug. 24th at 20:15

 Your submission has been moved to the physics.gen-ph (General Physics) subject class on the advice of our moderators, who have determined it better suited to that subject class.

Any questions regarding moderation must be directed to moderation@arxiv.org.  (The www-admin@arxiv.org address is for technical queries only.)

crossposting barred

Here is the correspondence that followed when I complained about this:

To: moderation@arxiv.org, Aug. 25th. at 11:18, the channel for complaints (not answered):

Dear Moderator,

    Would someone please explain to me the grounds on which this paper (1108.4860) has been barred from cross-posting?  Anyone who in such a short time was able to comprehend the difficult ideas in that paper deserves my admiration, but I have reason to suppose that the real reason is that, Sarah Thomas and Jean Poland's denials notwithstanding, I am on a blacklist and whatever I had uploaded would have met with the same fate.

    One would have hoped that the archive's moderators would have learnt from the letter detailing the flaws of the archive that Nature highlighted some time back, deliberately waiting till it could be displayed as the leading letter of the week, but evidently not:

http://www.nature.com/nature/journal/v433/n7028/pdf/433800a.pdf

"ArXiv has become a vital communicative resource for the physics community. The moderators' attitude to any challenge to conventional thinking is likely to result in the loss to science of important innovative ideas. Radical changes are required in the way the archive is administered."

For your information, the following may clarify the essence of the paper. A physicist might compare the mass of the Earth and the mass of humanity and, having examined possible catalytic processes, conclude that the latter could not possibly significantly affect the former. And yet we know this is not the case.  Why?  The explanation lies in the 'interpretation cascade' discussed in the section 'From simplicity to complexity and back' of my paper which you are censoring.  This is a radically new idea as far as physics is concerned and you should not allow superficial reactions to cloud your thinking.  Please think again.  quant-ph may not be the best section for this (it happened to have an endorser who was not away at the time) but confining it to the 'dump section' gen-ph is inappropriate if anything is.

Yours sincerely, Brian Josephson
-----

to the person who dreamed up the system, Sept. 15th. at 23:25:

Dear Professor Ginsparg,


    I am writing to ask if you accept that in defining my recent submission, viz. arxiv:1108.4860, as 'inappropriate for cross-posting', your moderators were in error (yes, I have followed official procedure by writing to moderation@arxiv.org, making similar points to those below.  Any competent moderator would, I suggest, have examined the paper, accepted the points made, and immediately have restored cross-posting privileges.  Despite the lapse of a number of weeks, there has been no response, which I (and others) consider unacceptable).

    First I comment that this paper has been reviewed by a number of people, both in connection with the conference and in regard to its prospective inclusion in the conference proceedings.  One of these reviewers was hep-th endorser David Finkelstein.  Nothing that has come up in connection with these reviews, and the ensuing discussions, has suggested any deficiency in the paper itself.

    The idea of cross-posting, as I understand it, is that a paper by someone outside some specified research community may touch on matters of relevance to that community.  Cross-posting provides a way to bring the attention of people in that community to the existence of that paper, as may have fruitful consequences.  Readers can get some hint of what the paper is about through the abstract, and can then decide whether to ignore the paper or download the full paper to investigate further.

    My abstract is perhaps not ideal for informing physicists in general what exactly the paper was about -- it was written for the benefit of the conference attendees, who were more familiar with the ideas involved.  The abstract does nevertheless contain the sentence "In such a world-view, Wheeler?s observer-participation and emergent law arise naturally, rather than having to be imposed artificially."  Many quantum theorists would understand 'emergent law', this being a characteristic feature of string theory, and quite a number will have heard of Wheeler's argument to the effect that the choice of what is observed affects the physics.  People who have read his 'law without law' article referred to in the title of the paper will in addition know that Wheeler hoped to be able to derive all of physics somehow from the fact of observation and the distinction between subject and object, but that he had little idea how to accomplish this.  A number of people would, I believe, be intrigued by the assertion of the abstract that there is a natural way to deal with these issues, and would want to look further.  The main point of the present, and ongoing work, is that taking into account ideas from semiotics (the theory of signs) provides a completely new perspective.

    I can understand that a quick glance by a moderator might well not suffice to appreciate all these things, but my subsequent email to the moderators should have brought a swift response.  I trust there is nothing personal, or dogma-ridden, behind the fact that my last two submissions were moved from quant-ph to phys-gen, and barred from crossposting.  You are, I gather, funded by Cornell University, with which as it happens I have a connection, having been a visiting Fellow there 30 years ago.  I do not think Cornell would be pleased at the arxiv's administration acting on the basis of anything other than strictly scientific grounds.

    I look forward to a response from you at your earliest convenience.

-----

to Cornell Library, Oct. 5th. 2011 at 12:06:

    Having failed to get a meaningful response from either the moderators of the physics preprint archive or Prof. Ginsparg himself, I am writing to you instead since the archive is officially in the charge of Cornell University Library, whose name appears at the top of the archive's pages.  This concerns an issue that a number of people have had with the archive, viz. that attempts to post a preprint to a particular section of the archive are blocked by the arxiv team, no justification being given for this other than a bare declaration of 'inappropriateness'.  The significance of this barring lies in the fact that people who have subscribed to a particular section get emailed the abstracts of all postings, so that they can follow up in detail any that they find of interest.  If cross-posting to that section is blocked, people in the area will not get to know about the paper, contradicting one of the main declared purposes of the archive.

    The preprint of particular concern here is arXiv:1108.4860, based on a paper I gave at a conference and which was accepted for the Proceedings after refereeing.  This summarises a new approach I have been developing to the question of the role of the observer in physics, and the factors determining which laws of nature will be observed, both of which are important issues.

    I uploaded the paper initially to the quantum physics section, for which section I had received endorsement from a well-known expert in this area. Not only was the paper moved to the section General Physics (which I have no problem with), but also it was blocked from cross-posting to any other area (as per the screengrab at http://www.tcm.phy.cam.ac.uk/~bdj10/images/restriction.jpg).  I could see no reason for this, given that the paper has been reviewed by a number of people, and while there was considerable discussion, leading to some improvements being made, no errors or misconceptions were pointed out by the referees.

    Finding it regrettable that researchers in quantum physics would not get to know about these ideas through the standard emailed list mechanism, on Aug. 25th. I wrote to the moderation thus:

> Dear Moderator,
>
>
>     Would someone please explain to me the grounds on which this paper
> (1108.4860) has been barred from cross-posting?
> etc.

Having received no reply, on Sept. 15th. I wrote to Prof. Ginsparg thus:

> I am writing to ask if you accept that in defining my recent
> submission, viz. arxiv:1108.4860, as 'inappropriate for
> cross-posting', your moderators were in error.
> ...

Again, after a number of weeks have gone by, no reply, not even a simple acknowledgement of receipt.

    The archive is not a journal, where publication implies that some kind of standard has been reached.  The fact that the endorser was willing to endorse this paper should have been enough, and the positive reception at the conference and by the referees confirms this assessment.  I accept that many theorists these days have very narrow interests and would have no interest in this particular paper, but I cannot believe that all people posting in this area are so intellectually challenged.  The fact that many people in this subject area would have no interest in the paper cannot be considered a credible reason for refusing to have this single preprint added to the 20 or so submissions received per day in this area so that those who do have an interest would be emailed the abstract.

    I do not find this behaviour, blocking the paper from cross-posting in the first place, and the subsequent ignoring of emails about this, acceptable. A predecessor of yours talked having no problem with the archive 'maintaining Cornell's academic standards'.  It seems to me that the problem in this case is that the running of the archive is not coming up to any reasonable standard of what one might expect in an academic context. And this is by no means an isolated occurrence, the whistleblowing web site archivefreedom.org detailing many more.

    Therefore I should like to request that you contact Prof. Ginsparg, asking him either to provide credible justification for this paper not being cross-posted to sections such as quant-ph, or to remove the bar on crossposting.  I would also hope that he and his colleagues stop this silly behaviour in future, and allow experienced scientists to decide for themselves what areas are appropriate for cross-posting, perhaps subject to some limit in number.

---------

from Cornell Library, Oct. 6, 2011, 15:39

Dear Professor Josephson,


Your message below has been forwarded to me. After investigation, I believe that all arXiv moderation practices and policies were correctly followed with regard to your submission.

After you submitted your paper to arXiv, arXiv moderators responsible for quant-ph and gr-qc did not accept your paper for their categories, and it was therefore announced in gen-ph.

What was not explained to you, and I apologize for this, is that arXiv rarely allows cross-lists from "General Physics", as this subject category has relevance to all other areas of physics, and is therefore already viewable by interested readers.

The arXiv moderation process applies a uniform set of procedures to every arXiv submission. While authors suggest appropriate paper classifications, final decisions about classification, including cross listings, are in the hands of subject moderators. arXiv administrators at Cornell merely implement these decisions.

While I understand your frustration that you cannot completely control the classification of your paper, please be assured that your submission to arXiv has been treated in a manner consistent with all other arXiv submissions.



Best regards,

-------

To Cornell, Oct. 6th. 2011 at 19:06

    Thank you for your reply.  You are now the third representative of the archive who has failed to respond to my request for an explanation for why my paper was not accepted for these categories.  The natural conclusion many would draw from this failure is that the moderators based their decision merely on 'a cursory glance' at the paper (a phrase taken from Prof. Ginsparg's recent article in Nature), that they did read it but lacked the ability to follow the arguments and thereby come to a meaningful conclusion, or alternatively that it was a matter of whim, or personal hostility, on the part of the moderator concerned. No responsible journal editor would reject a paper in this way without giving a reason and it is interesting that the arxiv people have not come up with one.

    The 'arXiv moderation practices and policies' to which you refer do, unfortunately, allow a person, under cover of anonymity, to conduct a vendetta against individuals, and there is evidence, including what appears to be a blackmail letter from an arxiv official that has been reposted on the web, to suggest that this sometimes happens.  I am not satisfied that there was a good scientific reason for barring cross-posting in the present case.   I have a fair amount of experience lecturing to physicists on matters relating to this preprint, and the discussions following these lectures lead me to presume that there is considerable interest in the ideas.  I therefore request that the bar be lifted.

> What was not explained to you, and I apologize for this, is that
> arXiv rarely allows cross-lists from "General Physics", as this
> subject category has relevance to all other areas of physics, and is
> therefore already viewable by interested readers.

    I am afraid I don't follow your logic.  Viewability is not the issue but actual views (of the abstract at least), as my original letter should have made clear.  What proportion of quant-ph subscribers routinely look at the abstracts in gen-ph?  If you can convince me that a substantial proportion of quant-ph subscribers also scan the gen-ph abstracts to see if there is anything of interest to them then I will withdraw my request, but my guess is that the majority are too busy to do this, and look only at the sections that they subscribe to.

    I look forward to a satisfactory conclusion to my request.

-----

From Cornell Library, Oct. 6th., 22:35

Your paper was not accepted to quant-ph and gr-qc because the moderators responsible for those categories rejected it. They did not feel it was suitable for, or of sufficient interest to, those areas.

As you have noted, the arXiv is not a journal, and we do not provide editorial comments on the content of papers. If you wish such feedback, we urge you to submit your paper to an appropriate journal.

-----

To Cornell, Oct. 7th. 10:02

    Thank you again for your quick response.  I do indeed plan to write up this work for a journal in some form in due course, but it will be some time before that happens, and one of the aims of the archive as I understand it is to speed up the communication process, so that people can become aware of current developments in their field prior to their appearance in a journal.  It so happens also that I cannot submit this particular paper to a journal since it will be appearing eventually in the conference proceedings, and dual publication would not be permitted.  Let me comment also that it is not 'areas' that read preprints but people, and my experience in lecturing on this subject contradicts the view of the moderator.  I might add further that once upon a time someone did a test that showed clearly that he was being targeted by the archive; if he changed his email address so that the system did not recognise him then the archive stopped blocking his submissions.  This contradicts the view that it is the content of a preprint, rather than some extraneous considerations such as who is submitting, that determines whether or not the moderators undertake blocking action.

    But let me leave such issues aside and simply ask, just to clarify the situation, whether I would be correct in presuming that the moderators would not shift their position even were a number of prominent scientists to write in in support of my request?

    I should be grateful, also, if you could explain to me exactly what the problem is with cross-posting to an area where not many people might be interested in the paper concerned.  With a paper journal space is limited, and so readership interest is clearly a relevant consideration.  With an electronic repository on the other hand, disc space costs very little and so it is of less relevance.  And in the case of cross-posting, no disc space is involved at all; no matter how many sections a preprint is listed in, only one copy needs to be kept on disc.  I do not, quite honestly, see why it cannot be left to the judgement of authors as to which areas a preprint that is not obviously wrong should be crossposted to, while allowing the moderators to determine the primary area (which will ensure that unthinking posting to an genuinely inappropriate area can be overridden, since positive action has to be taken to cross-post a submission later on).  I accept the principle that crossposting should be numerically limited.

    The only relevant consideration might be the possibility of overloading a section with irrelevant material.  Let us then look at the statistics, to see if this is a real problem or just one in the minds of the administrators.  Last week, there were 21 preprints in the gen-ph section.   On checking this out, I was interested to see that, contrary to your earlier statement 'arXiv rarely allows cross-lists from "General Physics"', as many as 5 of the 21 were cross-listed, leaving 16 that were not.  In the worst possible case, where each of these were cross-posted to quant-ph by their authors, this would mean 16 extra postings to quant-ph in the last week in addition to the 106 actually posted.  A more realistic figure for crosspostings, taking into account the subject matter of the preprints concerned, would come to at most 2 additional postings per day.  Most would not see this as significant overloading, given that normally it takes only a couple of seconds for someone to look at a preprint's title and decide not to look further.  I do not think therefore that overloading is a serious objection.

    Cornell is a very highly regarded institution (I myself regarded it so highly that in 1971 I spent 5 months there as Visiting Fellow in the Solid State section (LASSP)).  It should not be giving its blessing to a system that many would consider seriously flawed and counterproductive.  It may interest you to know that one well endowed Cornell alumnus has written the university out of his will on account of his irrational treatment by the archive.

-----

To Cornell Library, Oct. 9th., 14:10

    To take up these issues further, I gather Cornell provides much of the funding for the arxiv, and therefore does have an interest in how well it is being run.  The basic objection to the current arrangements is captured by the subtitle that Nature's editorial staff provided for a letter by myself (see attached) that they published in 2005:

"Putting control in the hands of a few can enforce orthodoxy and stifle innovative ideas"

In detail, I observed that:

> ArXiv has become a vital communicative resource for the physics
> community. The moderators' attitude to any challenge to
> conventional thinking is likely to result in the loss to science of
> important innovative ideas.

Apart from the introduction of the endorsement system, little has changed since that date.  The archive is still run in a way that leads to the stifling of innovative ideas: this is pretty inevitable, as I discuss below, given the way moderation is carried out, as enhanced by their inadequate responses to appeals.  Here, from my Nature letter, are some prize examples of the latter, from occasions when moderation did deign to respond:

> For example, having stated that a very distinguished physicist’s
[Cornell's Hans Bethe if I recall correctly]
> strong support of a submission carried no weight because this
> physicist "was not intimately familiar with the work in
> question", the moderators simply ignored subsequent support from an
> endorser with publications on the same subject.

(and they still ignore endorsement -- my paper had just been endorsed for quant-ph by a prominent worker in the field).

and again:

> ... in another example, the moderators' response to the information
> that more than one eminent physicist had an interest in a subject
> that they wished to bar was: "We are always thrilled to hear when
> people find an avocation that keeps them off the streets and out of
> trouble."

a style of dealing with complaints that might very well be characterised as a perversion of normal practice.  'Putting control in the hands of a few', indeed (and a rather unpleasant few, to judge by the above examples).

    Ginsparg has described the problematic moderation process thus: "Incoming abstracts are given a cursory glance by volunteer external moderators for appropriateness to their subject areas".  You will note that the volunteers look at the abstract, not the paper itself, and whether an abstract is recognised as being relevant or not is a function of the knowledge of the moderator*.  Both title and abstract of my paper referred to 'Wheeler's observer-participation', which should have caused a competent moderator to recognise the paper as being relevant to quantum physics.  Again, the 'law without law' idea mentioned in the title, of which my paper is a critique, comes from a chapter devoted to this concept in a specialist book entitled 'Quantum Theory and Measurement', published by Princeton University Press (my ref. 2).  The moderator concerned seems to have had insufficient knowledge of these matters, to the detriment of his assessment.

    Of course, any system is liable to error.  The problem here is that the moderators will not admit to error.  I asked them "Would someone please explain to me the grounds on which this paper (1108.4860) has been barred from cross-posting?" and there was no response.  They did not answer because there is no credible answer they could give, only that, contrary to the opinion of my endorser, they judged it 'inappropriate'.

    This 'cursory glance' approach does discriminate against innovative research for the following reason: innovative work is likely to follow different lines to typical papers and hence look inappropriate if one does not follow the detail.  It is also likely to get flagged as problematic by Prof. Ginsparg's ingenious automated filters, which perhaps do examine the whole of the text of a paper.

    Why in any case does the archive go to all this trouble to block papers from particular areas?  It is wrong to suggest, as Ginsparg does in his Nature article: "arXiv unintentionally becomes an accrediting agency for researchers, much as the Science Citation Index became an accrediting agency for journals, by formulating criteria for their inclusion." It is well known that arxiv is self-archiving and that any filtering is very rudimentary; no journal would consider being in the archive as a substitute for proper refereeing, and no minimally knowledgeable employer would consider appearance in the arxiv as a substitute for proper publication.  Some journals do ask authors to post their preprints on the archive before sending them out for review, but this is just to simplify the review process, and barring by the archive actually prevents the authors concerned sending in papers to journals that impose this requirement (incidentally, the arxiv has sometimes blocked published papers, again setting itself up as the arbiter of value rather than more generally recognised forms of assessment).  Ginsparg's lumping together arxiv with the citation index, as in the above quote from his article, is disingenuous: if a paper is cited by a number of other people, that is a genuine indication of its merit, in a way that inclusion in the archive is not.

    I find it extremely objectionable, and offensive, that some inexperienced moderator can, on the basis of a superficial examination of a paper, prevent the communication of my ideas to the quantum physics community, some of whom I am sure would be interested in my approach; and that they refuse to enter into dialogue about it.  The archive is supposed to be a service to research workers, and instead appears to be a disservice to people who have novel approaches to problems in science.

    I see, if I log on from our university network, that the archive is partly sponsored by Cambridge University (though it does not yet feature in the online sponsors list), and I am sure those who agreed to this would not be pleased at their funding being abused in this way. May I ask that these comments be passed on to those who oversee the system for their consideration.


Yours sincerely,


    Brian Josephson

PS: regarding your

> If you wish such feedback, we urge you to submit your paper to an
> appropriate journal.

while it was kind of you to make this suggestion, I have already had feedback regarding the paper from a number of reviewers, and the wish to get feedback was not the point of my communication.

* possibly the moderator concerned did look at the paper as well, but it uses generally unfamiliar arguments and I would not expect the average reviewer to understand what the arguments without putting in a fair amount of effort.

-----

From Cornell, Oct. 11th, 02:40

Thank you for sharing your critique of arXiv with us. I will see that it is passed on to the various people and groups that contribute to arXiv's operation.

Researchers at Cambridge University are among the heaviest users of arXiv worldwide, in terms of both submissions and downloads. We appreciate their use of arXiv and their support. Please do discuss your concerns with your local colleagues and librarians.

------

To Cornell, Oct. 11th, 08:15


> Thank you for sharing your critique of arXiv with us. I will see that
> it is passed on to the various people and groups that contribute to
> arXiv's operation.

    Thank you very much for doing this.  Since writing to you, one of my contacts has told me of a rather dramatic case where someone with 200 publications to his name published a paper in Physics Letters that 'showed that the work of some big shots [was] wrong'.  Soon after this, the author found himself barred from posting to the archive, believed to be the consequence of pressure from the 'big shots'.  I have not investigated the details, but can well believe the story because once, after I had disputed certain archive policies, I found myself being confronted, when I tried to upload a preprint, with a message that I was 'not permitted to post to the archive' (they thought better of this after I had written suggesting that they 'fix their system error').

----

To Cornell, Oct. 29th., 21:23

> Thank you for sharing your critique of arXiv with us. I will see that
> it is passed on to the various people and groups that contribute to
> arXiv's operation.

    I wonder if there has been any response from these people?  I suspect not; past experience suggests that they respond only to their own interests, ignoring those of the wider community, often not deigning to respond to an enquiry if producing a credible response to it is seen to be problematic.

    I take it that the issue of 'enforcing orthodoxy', thereby blocking the transmission of innovative ideas, is one on which Cornell has views, and may I suggest therefore that the people responsible be asked when they propose to respond to the email that you passed on to them. If/when they do respond, the adequacy of that response can become a matter to be discussed between us.

----

[On Oct. 30th. as background, details were sent of the case of Paul Violette, summarised thus:

I would like to bring you up to date on recent developments of my attempt to stop arXiv.org from blocking me in uploading my paper to the gr-qc section.  I recently discovered that just after I was successfully endorsed to upload to gr-qc, they continued to block me from uploading and then revoked the endorsement priveleges of my endorser who was eminently qualified.  He had uploaded 10 papers to the gr-qc section in the past 5 years.]

----

From Cornell, Nov. 11th., 04:57

I have discussed your critique of arXiv with those involved in its administration. You have raised provocative and interesting questions. I am afraid that I cannot provide an extended response, but I will make these few comments.

arXiv now receives over 6,000 submissions per month (nearly 7,000 in Oct.). It operates with very lean staffing and depends on the volunteer efforts of moderators worldwide. We do not ask moderators to peer-review articles. We ask them to assess whether articles are plausibly interesting to their communities. This is certainly a subjective judgment, and the boundaries of any scientific community are not clearly defined. Nevertheless, we believe that the arXiv moderation process is extremely accurate in its assessments, and that it is the most cost-effective and efficient way of dealing with the volume of submissions coming into arXiv. I can also emphasize that we are confident in the knowledge and experience of arXiv moderators.

Finally, we do not believe that arXiv is blocking the transmission of any scholarship or stifling innovative ideas, primarily because arXiv is not by any measure the sole means for communicating scientific ideas. You have available to you many other channels by which you can communicate your ideas to the quantum physics community.

-----

The following emails were sent to Cornell Library in response:

response 1, Nov. 13th., 18:30

    On 11 November 2011 04:57:55 +0000 you wrote among other things:

> we believe that the arXiv moderation process is extremely accurate in
> its assessments

Before I respond to the rest of your email, may I ask what is the basis of this particular assertion, which I find puzzling?  A fair proportion of papers that have been blocked (either completely or from a particular area) have been published in refereed journals.  Are you saying that even in these cases, where there is an apparent difference in opinion between moderators and referees, the moderators got it right and the peer-reviewers got it wrong?  That would be surprising, given that referees typically give papers more detailed attention than one gathers your moderators do.  Can you clarify, please?

It is unclear to me in any case how one would measure the accuracy of the moderation process, since a statement of accuracy requires some way of deciding what the correct result of a given assessment would be.  If however all you mean is that there is a high level of consistency between the views of different moderators, then I am afraid I have to disillusion you, since in the past views consistently held by 'experts' have often been wrong.  Here is a famous example [source: Wikipedia]:

> Semmelweis postulated the theory of washing with chlorinated lime
> solutions in 1847[1] while working in Vienna General Hospital's First
> Obstetrical Clinic, where doctors' wards had three times the
> mortality of midwives' wards. ... Despite various publications of
> results where hand-washing reduced mortality to below 1%,
> Semmelweis's observations conflicted with the established scientific
> and medical opinions of the time and his ideas were rejected by the
> medical community.

The existence of factors tending to make all members of a social community think in similar ways is, I am afraid, liable to give anyone who equates consistency with truth a false impression.  I, personally, would derive from the facts cited a different conclusion, viz.: "these observations suggest that moderators asked to judge the merits of a paper quickly have their judgments swayed by superficial cues, leading to their views being less reliable than those of referees who give a paper more detailed attention".  I am open to being persuaded that your own view is the correct one, however.

------

response 2: Nov. 15th., 19:34

    There is a further point that needs to be added to my previous comment which was to the effect that:
> ... moderators asked to judge the merits of a paper quickly [may]
> have their judgments swayed by superficial cues.
Namely, however expert your moderators may be in their own field, quite often it is the author who is the person most expert in the subject matter of a given paper, particularly if this contains novel ideas that would take a referee or a moderator quite some time to come to terms with, or if it incorporates ideas unfamiliar to most workers in the area concerned (as is arguably the case for those papers of my own that were moved to gen-ph by a moderator).  The handicap that the moderators may be under in such circumstances provides a strong argument in favour of the author's views having priority.

    I want to concentrate however on the following critical extract from your own letter:

> ... we do not believe that arXiv is blocking the transmission of
> any scholarship or stifling innovative ideas, primarily because arXiv
> is not by any measure the sole means for communicating scientific
> ideas. You have available to you many other channels by which you can
> communicate your ideas to the quantum physics community.

    This is, on the face of it, a disingenuous argument.  Alternative communication channels exist certainly, but the issue is how effective they are compared with arxiv in regard to communication with potential readers (a matter glossed over in the above).  The outlets I am aware of are uniformly inferior to arxiv in one or more of the following respects: number of persons browsing the content, delays due to the refereeing process etc., and the proportion of high-quality papers that are accepted (for example, journals such as Phys. Rev. Letters have room for only a fraction of the papers of this kind that they receive, and so would not generally come up for consideration).

    Arxiv is a significant factor in promoting scientific advance, as evidenced by the number of emails one gets telling one about new results that people have become aware of through the archive.  If arxiv's administrators really believe that its moderators' actions do not have an adverse effect on the transmission of innovative ideas they need to make a reasoned case for this, not merely give bland, politician-like reassurances.  In this connection, it is worth noting the words of a correspondent:

> As a US citizen, I see the arXiv as being paid by the US Government
> (taxpayers) to provide a service.

Barring a paper from crosslisting is a disservice, not a service.

    The statement copied above (regarding the blocking of the transmission of scholarship) seems to me, and to others who have seen it, to be in essence a political statement and one that would crumble under close examination.  If then an author, who may very well understand the implications of his or her work far better than can some moderator who has had only brief acquaintance with the ideas in the paper, believes that his or her work is more likely to be discovered by researchers who will find the work of interest should it be listed in area X, then he or she should be allowed to have it listed in area X, always assuming the paper has no serious defects (real defects, I must emphasise, not ones that are purely reflections of moderator prejudice).

    Again, your assertion

> [arxiv] operates with very lean staffing and depends on the volunteer
> efforts of moderators worldwide

is, I am sorry to say, a red herring, as nothing needs to be done to implement the desirable changes other than a change in the instructions to moderators, as I will show.  My proposal is that the system whereby a moderator can move an upload from the area designated by the author to one that the moderator considers more appropriate should remain, since an author may in fact not have made a good choice of primary area.  However, if an author then considers the moderator's judgment to have been at fault, that author should be able to crosslist to the original area (and also, subject to suitable limits as to number, to other designated areas), simply by using the existing crosslisting mechanism.  In other words, the ability of a moderator to bar crosslisting should be removed, a change that would involve no extra work for those who run the archive.  Particular people may object to particular choices made possible as a result, but that should be considered their own problem, to be dealt with by their developing tolerance, or alternatively attending anger management classes.

    In conclusion, I think that most people seeing this correspondence will conclude that the people running the arxiv have been very busy inventing excuses to avoid changing a policy that is, ultimately, unjustifiable.  Cornell does not come out well out of this.

------

response 3, Nov. 30th, 2011 17:39

    I see there is no response to my 'response 2' of Nov. 15th. with its analysis of the 'disingenuous arguments' provided by the arxiv administrators in response to criticisms.  I can well imagine that a person of your integrity would not wish to endorse further the dubious defences of its procedures that the administrators have drummed up.  This business is clearly related to personal characteristics of those involved, disrespect being a prominent feature of the way arxiv deals with those who write to it, as well as its practice of ducking key issues.

    As with other situations, power corrupts.  Arxiv could have been run in a proper, civil manner, but it proved tempting for some people to take advantage of the absolute power that controllers of the archive have, and to abuse that power in the service of personal dislikes.  Thus we see the blocking of papers which had already demonstrated their admissibility to the archive by having passed the refereeing test, and the overruling of experienced endorsers (in which connection, I have previously pointed out the irrelevance of the degree of experience of the moderators, such experience being of little value when they have to come to quick conclusions in regard to new approaches).

    Once proper behaviour has stopped being the norm, behaving improperly starts to become the usual thing to do.  A sad day for science, a sad saga for Cornell!

------

The current situation (December 20th. 2011) is as follows:
A response to these suggestions is awaited.

Posted by Brian Josephson on December 20th., 2011

Comments to bdj10-at-cam-dot-ac-dot-uk