Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
Did we ever resolve the question of whether publication in the nJournal is “orthogonal” to publication in other journals (thereby encouraging resubmission of material published elsewhere) or “counts as” a publication in the traditional sense?
Actually, I don’t see why we need to pick one only — why not allow both types of publication? There is already TAC and TAC Reprints; we could just treat them on a more equal footing.
I think that once we are guaranteed that material on the nJournal will be perpetually available on the nJournal (just as it happens for arXiv) then we should not be interested in whether an author wants to submit an article also to another journal he/she would consider more “official”. I guess in a few years, when the nJournal will be fully recognized as a “publication in a traditional sense”, this habit will cease by its own nonsense, but in the first period I can imagine an author fearing that submitting to nJournal would prevent that paper to be considered a “true publication” for ever. My feeling is we should just do not care about what an author does with his/her paper after the paper has been published in the nJournal as soon as the paper will not be removed. I think we should not even care about having the paper credited in citations: if an author publishes in Annals of nLab and then republishes in Annals of Mathematics, and then prefers to hide the fact that the paper had appeared in Annals of nLab in his bibliographies, why should we care? It is like someone not citing the fact that a paper had first been realized as an arXiv preprint. The quality of nJournal will have to speak by itself, it is not us who should impose it by saying “Hey, you have to properly cite this!”. We are stating clearly that “For science and for humanity, what counts is not the authorship of and fame and credit for a result. What counts is the result.”, so why bothering about “official credits”?
For the same reason I agree we shoul allow republishing in nJournal a paper already published elsewhere, if the author has the right to do so. But here we get into legal issues I have no idea how to deal with. The ideal would be that responsability for a submission is entirely the authors: we should not be worried if someone submits to nJournal something he/she has not the right to republish. However I see in this case the original publisher could ask us to remove the file, and this would be a big problem if other papers will be linking to that. So we should write some good disclaimer to tell an author that we are happy to republish, but we do not want this to end up in endless problems.
and then prefers to hide the fact that the paper had appeared in Annals of nLab in his bibliographies, why should we care
Well, it is helpful to be clear from the beginning and not bring people into a tricky situation. Judging by most institutions is based on the fact weather it is truly published or not, where the latter can not be repeated. Reviewer of somebody’s career should be able to decide what somebody’s publication record means. So one looks into the list, opens web page of the journal, or arxiv, and sees what a publication there means. If we make it ambigous, the referees of submitters careers may make arbitrary, uninformed or unconfirmed reports or decisions. That may be unpleasant. Similar doubts can be raised by societies or indexing services. Will MathSciNet want to include references to a journal which is not sure about such issues ?
We are stating clearly that “For science and for humanity, what counts is not the authorship of and fame and credit for a result. What counts is the result.”, so why bothering about “official credits”?
Do you care to have this in MathSciNet ? Do you care of being allowed to list it in official applications for EU grants ? Do you care of recommending other contributors to be able to take a credit in present unperfect system or you just tell them, look guys, we Labers live in an ideal world, so take the same criteria when submitting and tell them lasciare ogni speranza, voi chi entrate about their attachement to worldly career goods.
That would be extremely unpleasant. But we are already clear on what a publication in NJournal means: there is a peer review and a well recognized editorial board taking “scientific responsability” of what appears in the nJournal. Can we make this even clearer? yes, and we should, there should be no doubt to any reviewer of one’s career who want to be honest to him/herself of what being published in nJournal means. But we cannot ignore that many reviewers will just check against indexing services and sum up, and this means that for at least a few years publications in nJournal will count zero. Which will probably prevent many authors to submit to nJournal if they feel they can be published on a well established journal. So we should not prevent them from doing this. As nJournal will gain recognition from the community I guess no one will make use of the resubmission-to-other-journal issue, but leaving this possibility may make the difference in teh first times. Maybe we should write a “Resubmission to other journals” item in the journal home page.
To Domenico: If you change the rules after few years, then one has an awkward situation that the publications in some period radically differently count than those in another.
Mike 1
There is already TAC and TAC Reprints; we could just treat them on a more equal footing.
Right, there are also International Mathematical Research Notices and International Mathematical Research Letters. Handled equally, by the same committee, but the latter format allows for smaller number of pages. It is useful that they are labelled differently, although the procedure form the point of veiw of the editorial services is not any different until the very moment of publication. Similarly, Physical Review Letters is higher level publication than Physical Review A,B,C,D,E (higher criteria for acceptance and also smalelr size). If you submit to PRL and get good enough for Phys. Rev. B for example and not for PRL then you do not need to resubmit, the editor will decide to recommend to PRB and it will almost automatically end there, unless you complain for the reconsideration in PRL.
It seems to me that there is definitely a need for “unique publication,” at least as long as academics are being evaluated at least partially based on their number of publications (which is not likely to go away in the near future). We eventually want nJournal publications to be countable in one’s CV, and I don’t think we can get around the fact that in order for that to be reasonable, people reading the CV need to be able to be guaranteed that the publication is not a duplicate.
I would expect the norm initially would be to publish a nJournal version, await comments, additions etc, then to submit to a traditional journal. We should allow this, (even possibly encourage it), but require that an acknowledgement to the nJournal and the help that that has been (?) and correct references to the nJournal version should be included in the final traditionally published form. The publication in the nJournal would be novel as it would allow use of greater linkages, and also access to nLab material. The traditional version would be different. It should also be mentioned that the normal mode would be to include a single reference in publication lists to the paper with BOTH references in the same entry (and let all break loose on the bean-counters data bases!!!!! (evil grin!!!)
We can require that the author agrees to this as a prerequisite for publication. The conditions of acceptance need to be clear on all such things.
Not sure if I understand the problem. I think:
If somebody has already published a result and wants to put it in our wiki, he does so on the Lab the way we do it all the time, with a pointer to where the stable and peer-reviewed version is. There is no point in putting that material into the Annals of the Lab if it already is peer reviewed elsewhere.
If somebody gets his result peer-reviewed and accepted in the Annals of the Lab first and then tries to resubmit elsewhere that seems to be his or her problem and and not ours.
Urs, you’re looking at the matter from the wrong end. Consider the author’s point of view. Why should someone publish their work in the nJournal? There must be something in it for them. Why do we publish in ordinary journals? To be able to list the publication in our CVs and know that the bean-counters will count it. That’s the only reason: I know of no other that actually matches up with reality. And that’s the one thing that, at least in the early days, we cannot offer. So we need to make our journal attractive in other ways and actually add value to the article. The main thing being that as it is online, it can benefit from linking in to and from the nLab and other articles, thus making the literature of mathematics reflect the web of mathematics a little more closely. Now I can see that someone who sees the value of this might like more of their work put in this format. The easiest way is to take an article already written and convert it, rather than writing it afresh. So putting up articles that are already published elsewhere is certainly something we should allow for. On the other hand, since we can’t yet offer CV points, we shouldn’t stand in the way of authors trying to get the best of both worlds. In addition, since we are the new kids on the block, it is our responsibility to ensure that we “play nicely” with the big kids already there.
It’s that question: what can we offer? that needs answering and codifying into the “mission statement” that John was asking for. I would start with the following:
I think that those three are the main things that an author would like, apart from CV points. So if we can get them right, we’re in good shape.
If somebody gets his result peer-reviewed and accepted in the Annals of the nLab first and then tries to resubmit elsewhere that seems to be his or her problem and and not ours.
A good service to the community is to give guideliness to contributors. A contributor is the guy who gives a gift, publishers are the ones who help with public relations. Should nJournals be worse in that respect ? Being innovative, the questions are more pronounced.
There is no point in putting that material into the Annals of the nLab if it already is peer reviewed elsewhere.
While I agree with this, some flags that a material is being reviewed somewhere may be useful. Additional checking is also OK. I think that most referees nowdays do not carefully check the details.
There is a boost for nJournal publication, and that is that being a wiki we can provide a easier feedback from ’the community’. Someone mentioned this sort of publication leading to new collaborations as it is very easy to contact by adding something to a comments page. In designing things, that sort of idea might be taken into account. If someone provides a simpler proof for some lemma or similar in an nJ paper then it is like discussing the material with colleagues.
A good service to the community is to give guideliness to contributors.
Sorry, I didn’t mean to be rude. My idea is the following: assume there was no nLab. Then we could create an arXiv-based journal simply as follows: we set up a group of experts, whose list is publically knon (this is the editorial board) that consider a paper from the arXiv (because an author has asked them to be evalued, or beacause they decide a paper was interesting, ore because a third part pointed their attention to a paper), and then, after they ore someone they trust has gone through the details of that paper and has found it correct and interesting, say a very simple thing: they say: we found this paper correct and interesting.
Does this add any value to the paper? surely yes, since the judgement has been rigorous and the judges are recognized experts. Does this adds beans to the paper? surely not. Does this prevent the paper from being published in an “official journal” and get beans? surely not.
And creating a nice web page where stroring all this information, with transparent criteria, aims and scope, etc. the list of the we found this paper correct and interesting would be a peer reviewed jornal from any scientific point of view. No beans even if we were able to have only Fields Medals in the editorial board, though.
And this we could do even without nLab. With nLab we can do more: we can have hyperlinked papers, with a stable version and an indefinitely improvable version. There’s however something we cannot do with respect to the purely arXiv based journal: we cannot just give a judgement on a work which is out there, but we need the source file. So probably we need a submission from the author. I must say I totally ignore whether I can pick the latex source of “On the classification of TQFTs” from arxiv, convert that to instiki, put it on an nLab page, ask a few experts a judgement and then eventually write on the top of it These people found this paper correct and interesting. But I’d like if I could do that.
Anyway, this is the way I think the nJournal should work: a place for fully refereed preprints. In what does a refereed preprint differ from a publication? don’t ask me, ask to bean-counters :)
So putting up articles that are already published elsewhere is certainly something we should allow for.
Right, but these should be put on the Lab, as as already common practice.
(b) having a traditional numbering system,
What’s the progress with this? Did the process stall when I suggested to name Andrew as the publisher and Andrew voiced concern about legal issues?
Yes, it has stalled slightly. Whoever is the “official publisher”, we should ensure that the legal stuff is in order.
Also, a quick read of the ISSN website reveals that we should finalise a few decisions before we apply. The ISSN is linked to the title of the journal, so at the least that needs to be sorted out! If we’re applying before the first “issue” is issued, then we need to take into account the following:
This is called a pre-publication assignment. However the publisher should provide all the possible information together with his ISSN request. He should also send to the ISSN Centre concerned a sample issue or copy of the cover, editorial page (name and address of the publisher) and masthead, as appropriate, once the first issue has been published.
Yes, we should try to have two or three example entries on the Journal, soon. Not just for the ISSN, but also for us. I am trying to push this. Maybe we have two examples currently, nicely at both ends of the spectrum: a) Tom Leinster long and traditonal-style article. b) The brief lab entry on geometric realization.
Maybe we find a third example, somewhere in between the two. We should all work on bringing these examples on the web and polishing them and the process that they go through to get some feel for what it will all be like and to run into issues to be solved that discussion alone will not reveal.
Unfortunately I am very busy with something else at the moment. As I guess we all are and will always be. Which points to one open problem to be resolved: we may need some kind of staff.
How did TAC manage this? Perhaps it would be a good idea to ask Bob.
I can live with “Journal”. It strikes me as a bit misleading, because te result will be quite a bit different from what people are used to call a “Journal”, but my impression from comments here is that this is indended and regarded to help with the social aspects of it all. If so, then so be it.
We could have somebody else be publisher, if Andrew is worried. It had been just an idea of mine that Andrew might do it, I hadn’t asked him. I thought it would make sense, but I am not dogmatic about this.
If David Corfield still agrees to be publisher, then I am happy with that.
Should we have a final vote over the title?
Is the proposal for “nJournal” itself, or for “Journal of the nLab” (with “nJournal” as the nickname)?
Regarding publishers and legal issues, I really would like to get these sorted out regardless of who is named as the publisher. Even if I’m not the official publisher, I’m still the legal something of these sites (I hesitate to say “owner”) and everyone here has provided content. Establishing a journal takes our exposure level just that little bit higher so I think it worth sorting out what our responsibilities and rights are.
There’s also the issue that naming one single person as publisher leaves everyone else with no leg to stand on if something happens to that person. Given that it will be a little while before we’re actually read to announce publication of something, I think that we should spend the time finding out more about these issues.
To answer Tom’s explicit question: I’ve been trying a few google searches, but the phrase “needle in a haystack” comes to mind! I do have a couple of lawyer friends, but this isn’t particularly their speciality. That’s why I’d really like to hear from someone who’s been in this situation before.
Tom, do go ahead. I don’t have any specific question - which is part of the problem! I don’t know what questions to ask! Basically, what should we be aware of, and what protection/framework/… should we have in place.
Anyway, this is the way I think the nJournal should work: a place for fully refereed preprints. In what does a refereed preprint differ from a publication? don’t ask me, ask to bean-counters :)
Domenico, with the present low quality of refereeing in mathematics, except for a small niche of very top editorial practices, there is almost no difference between refereed and nonrefereed. A colleague of mine who does not publish in journals any more but only posts things to some archives says: why would somebody tell me what to read ? I know whom I appereciate and I can judge myself in my area, so I decide what to read. Most of people are too busy and do not invest much time in refereeing, but just referee from general impression, what is often not of much worth. One of the incentives is to help people. So if somebody whom I suspect to be a good author sends an article to a journal and it will count to him, he needs some career confirmation, I will spend time carefully reading and suggesting improvements and rejecting if the paper is a mistake. If I know in advance that it will not count much to the guy (for example it is just a preprint like here, or it is in bad “open access” journal where author pay for publication) then I will spend less effort. Now if you tell me that this is just a refereed preprint, than, I do not want to spend time refereeing, unless I bump into something very useful to read, because I see less benefit for the author. I referee to some journals almost exclusively as a service to the authors who are good and as a preventor of bad science to appear on the other side of the spectrum.
I said this before: we should allow for refereeing that genuinely serves the scientific purpose, and not just the academic machinery. We should
allow referees to sign by name and make substantial contributions to the entry, if desired. That is the single best reason to do good refereeing: make it part of the referee’s genuine scientific interest.
Indicate in detail for each submission what kind of refereeing it did receive,.Not just “this is published”, which means essentially nothing, as Zoran remarks above. It should say something like
This entry was refereed by anonymous referees chosen and assumed to be expert by the editorial board;
The following people say that they read the entry and think that it is okay: abc, xyd, pfq.
The person rst started refereeing the article but ended up reworking it and adding to it substantially. The resulting new version can be found here.
Where “here” may be back on the unrefereed Lab or again in the Journal, with its own list of people who looked at it. And ever so on and on.
Something like this would be of genuine value. Not just formal procedure.
Like for 29 :) nice ideas!
Absolutely agree with Urs in #29 (and it’s one of the things I suggested in the original thread on this). Refereeing should be transparent in that everyone should know exactly what it means to say “This has been refereed”. I suggest that we think of some statements, like the ones that Urs says, which could be used. Exactly as Urs says, the purpose should be that the statements would be of maximum use to someone reading the article.
Of course, for articles exactly in my field then I ignore the refereed/not-refereed status because I’m fully able to judge the quality of the article myself and don’t need anyone else saying what’s good or not. But once I get out of my own area of expertise - which I do quite often - then I do, at least initially, need to have some guidance on what is good and what not. It might be something like, “This has errors, but is very good on explaining the ideas” or “This is very precise, but very hard to read”, stuff like that would make it easier for a newcomer, or even a fringe-dweller, to get to know the field more quickly.
Okay, I put something like comment #29 into Proceedings preparation (nlabmeta). In the new section Transparent refereeing. Feel free to expand and improve.
Tom,
don’t worry, I think you misunderstood what I or other people said. Your article is submitted in the version that you specify, and if that’s supposed to be the original version with nobody else’s edits, then that’s how it will be done.
But if somebody else feels inspired later on to write, based on the Lab version of your article, a 200 page opus A very detailed introduction to topos theory , he may then also ask the Journal editors for peer-review of his additions.
Right, we should maybe say this more explicitly. I think the keyword here is “branching”, as in open source software development: everybody is working on his personal version, but if anyone feels like joining in without actually collaborating he can “branch off” a copy and work on that independently.
Andrew will know more about how this is handled in detail in the IT community.
Interesting that you mention the IT side of things. It’s something I’ve thought about for a while: that mathematics is very like open-source computing. In the immortal words of the Stone Soup Group: “Don’t want money: got money. Want fame.” Apart from the “got money”, that’s how I feel about my mathematics. I don’t want to protect it, I want it out there! If someone wants to develop it further, please do! I’d like to know about it, and to be given the chance to work alongside, but I don’t want to hinder someone taking my ideas and developing them further.
Of course, I don’t want someone plagiarising my text words inappropriately, and I’d rather my mathematical ideas were linked back to their source, but on the whole I’d rather they were used unattributed than not at all! To quote perhaps the most famous mathematician of all time: “I knew I’d written a true folk song when [The Irish Ballard] appeared in a collection of such songs attributed to ’anonymous’.”.
But that’s slightly orthogonal (can something be slightly orthogonal?) to why Urs invoked OSS. Certainly branching occurs in OSS, and is encouraged. But the author of the main trunk always has control of that trunk and can decide to accept or reject others’ modifications.
To be really radical, I’d propose that the “author-publisher contract” do the following:
I know that the fourth is understood within the mathematical community, but I think that if we make it explicit then it would really show up as nonsense the whole issue of copyrighting mathematical works for their protection. We could even go one step further and develop the Maths Public License, modelled on the GNU Public License, that would require anyone who developed the ideas further to release them under the same type of license. But perhaps this is a pipe dream …
Let me try to sum up the consensus conclusion to my original question in this thread, and see if everyone agrees. Publication in the nJournal does not preclude publication elsewhere. Initially, at least, we expect that publications in the nJournal will also be published elsewhere (either before or after) for bean-counter points, and we rely on the author to count them as only “one publication” on their CV with both places of publication listed. The advantages to the author of also publishing in the nJournal include transparent refereeing, community input and potential branching, wide exposure, and availability of hypertext. Eventually, we hope that the nJournal will become respected enough that articles published there will be counted by bean-counters without the need for republication elsewhere. Does that accord with what people are thinking?
Sounds good to me.
Perfect.
I like it too.
I got an incredibly helpful and detailed reply from Susan Hezlet. Here are some excerpts:
the pitfalls are all about claiming ownership, which is effectively what happens when you declare this body of work to be a journal
in stating that some parts of the site are now a journal, you are making a better defined statement that you are responsible for the content and you have taken some trouble to check that you have the right to publish this stuff, the authors are genuine originators etc, and the content isn't just a rip-off of someone else's IP
I would say that the guy who is hosting, assuming he is also responsible for editorial choices, is already legally responsible and effectively the publisher of the wiki
(It should be added that she isn't AFAIK a lawyer, etc etc... but she's very experienced.)
Separately, she also wrote:
Might anyone go?
MSRI: Workshop on Mathematics Journals - February 14, 2011 to February 16, 2011 (Berkeley)
… The workshop will discuss what is important and unique to the publishing of mathematical research articles and how we can best ensure that publishing practices support peer reviewed research in the long term. Much of the current discussion is taking place between funders and publishers, including learned societies, but not directly with mathematicians. A second goal is to see if we can find a consensus of opinion on what is important about journal publishing to mathematicians; i.e. where the balance lies between the desire for profits from publishing and the broader dissemination of research.
There have been some interesting discussion on the journal practices during the ICM meetings in Bangalore preceeding the congress in Hyderabad. A couple of links to ICM conclusions are at my page citations (zoranskoda).
So Andrew, given what Susan says, any more thoughts on whether you're willing to be named as publisher? Short of getting advice from a Norwegian lawyer, we're probably about as well-informed now as we're likely to get. On the other hand, there's no hurry: apart from anything else, we can't apply for an ISSN until we've decided on the publication's name.
Yes, given that I am the de facto publisher, I may as well become the de jure one as well! As you say, we still need to finalise the name. Also, it looks as though we need to mock up the main page. We could try styling it a little to make it look more attractive!
I think that we should seriously think about setting up a proper group to manage all the various stuff, but that needn’t hold us up now.
Wish I could go to that workshop …
@Andrew #50: Excellent! That's one more piece in place.
Oh, sad, I’d love to go to that workshop but I don’t think my schedule can handle it.
Usually MSRI puts videos of lectures online after some time anyway, usually a couple of months later (as far as I remember from about 2001, they had a system manager who was a very kind and sympathetic person, but unusually slow and laisy in updating things any near to the real time).
1 to 53 of 53