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.
I suggest we try the following (it won’t hurt in any case):
we create a new web
titled “nLab (reviewed)” (or similar)
publically read-only
write-access by a passoword which is shared by the steering committee members
Then we use this as follows:
whenever (possibly never, but let’s see) we have some Lab entry
all whose genuine co-authors agree that it is reasonably stable
for which the steering committee has sought and found a referee who wrote a positive report, just as for a journal article
we put a copy of that article into the web “Lab (reviewed)” and leave at the top of the original article a remark saying
a stable and reviewed version of this article can be found here
(My nature would be to simply create such a web and confront you all with its existence, but maybe luckily for you I am not sure if I still know the password that I need to do that…)
Hi Urs,
I’m not much convinced this would be a good idea (on top of it, I’m unable to see which would be the advantage in having such a reviewed copy of the nLab). But there’s a variant of what you are proposing that I find interesting: nLab based papers. We could have a section “_n_otes from the Lab” or whatever, which could be organized as a real journal: one may write a paper using nLab hyperlinks, this would make one avoid all that part in paper writing which is recalling standard definitions and constructions and go directly to the heart of the mathematical question. Once the paper is written it can be submitted to the sterring comittee which will look for a referee. If the referee’s report is positive, then the paper can be made publically available (read-only) in the journal section of the nLab.
Sounds good to me!
i am not sure, though, how what you are suggesting is essentially different from what I was suggesting! :-)
is it that you make a sharp distinction between the content of a “paper” and an “Lab article”? I think in the end these are at most questions of formatting. Formatting is determined largely by the medium. The new medium will eventually find its own most natural formatting. In the end there is a web-page with well-exposed (if everything works out)scientific content. If we call that a “paper” or an “entry” or whatever seems to be secondary to me.
Would you agree with that? If not, maybe give me more details on what you would and what you would not want to have on nLab (review) .
But in any case, I’d be happy with saying ” nLab(review) is for publishing reviewed papers”, if that’s what it takes to find consensus.
Not sure if you are all following the blog discussion. But this thought here arose from my musings here.
The new medium will eventually find its own most natural formatting
I absolutely agree with this: what I intended to say in my post was that I’m unsure current nLab entries would suit nLab (review); but the new medium will have its peculiar entries (which I called “papers” since I imagined they could be something close to a journal paper, but this was just my imagination).
I’m unsure current nLab entries would suit nLab (review)
Yes. But as soon as we have the option to have genuinely reviewed pages, we all and others may feel a whole new motivation to write some entries so that they do qualify. I know a handful of entries of mine that I want to bring into such a form eventually. Of course the vast number of entries will not be of this kind and should not, but some do.
Anyway, I think we two agree.
I agree that this is a good idea. There are certainly some things on the nLab not found elsewhere, and it might make others more comfortable using the Lab as a citable resource if there is some element of review. Certainly this makes room for quite small ’papers’ that are heavy on prerequisites. Think of some of Brian Day’s papers, which are often only a few pages, but have great ideas in them. But unlike his papers, results can be hyperlinked to further developments.
Maybe call it the nJournal?
Certainly this makes room for quite small ’papers’ that are heavy on prerequisites.
Yes, it would open a whole range of new possibilities!
We should try it. It costs us nothing to set it up, the worst that can happen is that nothing much is uploaded to the reviewed web at all (in which case it will be no worse than the several of the personal webs already existing that people had requested, but which are essentially abandoned now)
This is an exciting idea! I think I would prefer “nProceedings” to “nJournal”, but I also take the point that we might want it to have a more traditional-looking name for the purpose of citations, like Proceedings of the nLab.
One question: this all arose (partly) out of the desire to “write papers in wiki-format.” In particular, that should mean that “a paper” would consist of more than one wiki page. In principle there is no obstacle to this, but do we want some formal way of keeping track, on the nProceedings web, of which pages belong to which “paper”? Instiki categories perhaps?
I also think we should consider a formal policy regarding the updating (and perhaps adding) of links to pages in the nProceedings. Some journals, like TAC, have a policy that if your paper cites unpublished work which is later published, you can update the bibliographic reference in the online published version. Analogously, if an nProceedings page links to something on the main nLab, which is later incorporated into a different nProceedings paper, it would make sense to update the link. But what if an nLab page on a subject didn’t exist when an nProceedings paper was published, but later such a page develops and becomes a valuable reference; would it be possible to add a link to it in the relevant places of the older nProceedings paper?
Also, I presume that nProceedings papers can and should link to the corresponding “dynamic” version on the main nLab, so that readers can easily find out about more recent developments?
It would also be nice, however, if the nLab server were a little more reliable, so that people looking for papers published in the nProceedings could be sure of finding them.
@Mike it seems to be back!
Also, I think we should distinguish between the Steering Committee of the nLab and the Editorial Board of the nProceedings. There will certainly be overlap between the two, but I would guess that not all current members of the steering committee would want the second responsibility, and I can also easily imagine people who would be willing to do the second but not the first.
I have recently been involved with setting up the Journal of K-Theory. Once we had an editorial board, we then set up various guideline documents (advice to authors, advice to editors etc.) and a Wiki to keep track of submissions. This has made life a lot easier and should be easy enough to adapt.
but I also take the point that we might want it to have a more traditional-looking name for the purpose of citations, like Proceedings of the nLab .
Yes, let’s pick a traditional name, not to counteract the main purpose of the whole thing: that people want to and benefit from making genuine reviewed publications there.
do we want some formal way of keeping track, on the nProceedings web, of which pages belong to which “paper”? Instiki categories perhaps?
That sounds like a possibility. I am sure we can find some way. A floating TOC that brackets all pages corresponding to one publication comes to my mind :-)
I also think we should consider a formal policy regarding the updating
Yes, I think one big advantage of wiki-publications is that they are hyperlinked, and so it would be good to have a policy that allows hyperlinks to be up-to-date. I think we can easily find sensible rules here, along the lines you already indicate.
Also, I think we should distinguish between the Steering Committee of the nLab and the Editorial Board of the nProceedings.
Yes. The Steering Committee should set up the primordial Editorial Board.
It would also be nice, however, if the nLab server were a little more reliable,
You can say that again.
I also like “Proceedings of the nLab”. I can’t create webs at the moment, since it seems I forget the system password. Maybe you or somebody else can.
Also, I think we should distinguish between the Steering Committee of the nLab and the Editorial Board of the nProceedings.
I agree: we should try to have a quite vast and authoritative editorial board. precisely since at the beginning there is likely to be some skepticsm towards considering nJournal a true reviewed mathematics journal, we should set a quite high standard from the very beginning. A good idea could be to first have the list of those in the steering comittee who would like to be in the editorial board, too. Then to contact and invite those people who are not involved in the nLab but are so close that we can expect an interest and a positive answer (the first three names that come to my mind are John Baez, Jim Stasheff and Andre’ Joyal). Once this kernel is set, we can try to boarden the editorial board: the idea is that people we will contact in the second phase will be able to find the nJournal already set up in a preliminary form, and will not have to join something that is purely a project. For this second “call for editors” I guess each of us can invite those of his contacts one feels could be possibly interested (I, e.g., could try to involve Gabriele Vezzosi). With this second phase over we can have a look at who’s in and decide whether we want to attemp to call sme “big name” to join.
Apart from editorial board, we need a clear Aims and Scope to be written, in which we also specify that the nJournal is nLab-based, and a Sample paper to illustrate the format and the use of hyperlinks to contributors.
A lot of work is coming with 2011 :)
could try to involve Gabriele Vezzosi
That would be really great! If we could get to that point, I’d be quite pleased. Maybe we could even eventually make the Lab proceedings a reasonable choice for publishing derived geometry texts.
To get started, let’s make sure we continue taking steps so that we can feed Tom Leinster’s article through a Lab-proceedings refereeing process, as a first test case.
First, let’s create that darn web, so that we can start playing around with it. It seems to me that among those who voiced opinions so far (not many) the title
Proceedings of the Lab
has a slight majority.
I suggest we make the hard-coded name of the web be “Proceedings”, so that its URL is
http://ncatlaborg/proceedings
That would be sufficiently descriptive and still leave some room for how exactly we want to refer to this in full text.
(Somebody with the system passowrd please do!)
Once we have the web, we can play around with writing out Aims and Scope etc.
I would certainly endorse ’proceedings’ as a title.
I like “Proceedings of the Lab”, too.
I also like “Proceedings of the nLab”.
Oops! I just left a comment on the nCafe suggesting a post “Naming contest for a new higher category theory journal”, but it seems a consensus is forming around “Proceedings of the nLab”.
I also endorse Urs’ idea of having the link be http://ncatlab.org/proceedings
I’m happy to see this evolution occurring.
Here is the new web
I have added a little bit of first tentative introductory text, drawing on the discussion here.
(This is write-protected, as it must be. The passowrd is the system password.)
Maybe the next important step is: let’s agree on a primordial “editorial board”. I think as soon as volunteers are available, the steering committee should decide.
For practical purposes, I think it would be good to have a topos-theory expert on the board soon. So that this person can serve as the editor in charge of our first test-case submission: Tom Leinster’s topos-theory exposition.
I approve of the above.
Great! Instead of forcing authors to write their paper on the nLab, maybe we can setup a personal web for authors. If they submit a rough abstract to the Editorial Board, they can be provided a password to an proceedings author’s web. I can imagine some authors being shy about working on a draft in such a public forum that can be edited by anyone.
It would be great if http://ncatlab.org/proceedings took you directly to the read-only version without being asked for a password. The Editorial Board can find it easily enough without requiring a link.
Exciting stuff…
Is there a place for lab elves in the proceedings? There are simple typos I’m tempted to fix, but can’t.
I’m not suitable to be on any “board”, but I could potentially be an “editor”.
I like ’Proceedings..’ too. I would volunteer to be an ’primordial editor’, but only if others deem I am up to it. It may be handy to have an editor in this time-zone (John Baez is in this corner of the globe, but we’d have to check if he is at all interested, given his new career).
On second thought, a separate author’s web would kind of go against the name “Proceedings of the nLab”, which implies a close connection to the nLab. Maybe, at a mininum, we can create some categories, or flags, or standard notices, specifying that the page is under construction and intended for the proceedings. For example, if an author wants the lab elves to stay away, they can mark it
category: no edit
Similarly, if they want lab elves to begin editing right away, it could be marked
category: edit
Or maybe a combination:
category: draft, editable
Or something…
Noting Urs’ survey, differential cohomology in an (∞,1)-topos – survey, perhaps later on, once we have the proceedings off the ground some refereed n-Lab survey articles would be a good thing to add in to some on-line n-lab journal (n-Lab surveys). (I am just throwing this idea to the wind to see what you all think. I have not thought it through at all as to modalities etc.)
By the way, if I can help on the proc., count me in, more or less in any function, editorial board, referee, etc.
I have created a freely-editable copy of the Proceedings-text at Proceedings preparation (nlabmeta).
If any one of you feels like improving and/or adding text, please do so on that page and announce your changes here, for discussion. Once it stabilizes, we can copy it over to the write-proteced Proceedings-web.
A great job what you set up!
By the way, if I can help on the proc., count me in, more or less in any function, editorial board, referee, etc.
I subscribe!
@Toby and Domenico and others: okay, thanks for the reactions!
I have now added a section Difference to traditional publishing. Let me know what you think.
Fixed a minor typo
Let me know what you think.
Extremely clear, I like it very much. Maybe, in describing in what a web-based paper can differ from a traditional paper we could also emphasize that the use of links to nLab allows focusing the exposition of the main results, rather than in recalling definitions and basic examples, or fixing notations.
Maybe, in describing in what a web-based paper can differ from a traditional paper we could also emphasize that the use of links to nLab allows focusing the exposition of the main results, rather than in recalling definitions and basic examples, or fixing notations.
I did indicate this briefly at the beginning. Feel free to edit the text at Proceedings preparation (nlabmeta) to emphasize this more, as you see the need.
I now see that’s clearly stated in the Format section: no need to emphasize it more. Now let us focus on building the Editorial board: let us set up a basic kernel in a short time, so that immediately after this editorial board can write down a list of dreamed editors and can discuss who can be realistically invited to join.
Now let us focus on building the Editorial board:
Whatever we do, I hope that Mike volunteers to be on the board and volunteers to function as editor to which Tom submits his article.
Another thing: I would like to add a remark along the following lines to the description of how the Proceedings differs from traditional forms of peer-reviewed publication:
I would like that we allow referees to become part of the authoring process
To some extent this is inevitable for the Proceedings : if a referee finds wanting an nLab entry that has been submitted and cares about the nLab at all, he or she will want to improve the nLab entry anyway.
But there is something deeper here, that I would try to get to: what i find frustrating about the traditional peer review process is that after I, as a referee, have invested the energy and time to absorb an article and possibly find mistakes and possibly suggest their solution, I am effectively at the stage where I make research statements about the matter at hand. But the anonymity and the intermediate editor make an effective scientific exchange with the original author impossible. That’s a huge waste of potential energy, I think.
I have recounted several times my discussion with an editor of a succesfull online journal in the field atmospheric chemistry that I talked to last year when I was at that symposium on new means of electronic scientific interaciton. He told me about their online journal where articles are posted to a kind of blog, where referee reports are posted as kind of comments to these blog entries, with clear referee names if desired, and where the author may choose to reply on that blog.
He told us exciting things about the effects these had: he said the referee reports tended to be of exceptional quality if they were posted with clear name. He said they frequently added genuine value if not concrete results to the original article, and that several collaborations between an author and one of his referees resulted, that led to subsequent joint publication. I think this is easily imaginable.
I think this example shows that refereeing can be more part of the genuine scientific process than it tradtionally is. I would enjoy it if we could agree that we allow the possibility to have something like this. For instance allow the possibility for a referee to sign with his or her clear name and hence have his or her contribution (if any) to a published article be recognized. Refereeing time should not be time wasted for the referee, but time invested into his or her genuine scientific interest.
I am inclined to ask Jim Stasheff if he might want to be an editor, but the category-theoretic emphasis might be an obstruction.
Is there any sensible way at all that one can allow editors to publish themselves in the journal? If somehow it is made very explicit that the editor in question is not in charge of his own submission? Probably tricky.
Refereeing time should not be time wasted for the referee, but time invested into his or her genuine scientific interest.
I could not agree more :)
I am inclined to ask Jim Stasheff if he might want to be an editor
I agree. I would also suggest John Baez and Andre’ Joyal: they both have been interested in nLab activities, and could be interested in the Proceedings. There’s also Kevin Costello who’s currently writing a paper with Owen Gwilliam using a collaborative wiki format.
Is there any sensible way at all that one can allow editors to publish themselves in the journal?
maybe we can decide that, apart from non being in charge of his own submission, the review and eventual acceptance of papers from editors will be publically available. i.e., that the optional clear name refereeing is not optional for members of the editorial board. maybe this is not a definitive solution, but it can be a starting point.
@domenico #38
I doubt John would want to do this. He is focused on his new project. BUT… this is the second time you presented your wishlist of editors and failed to mention Tim. I can’t stand for that :)
It was probably assumed. Anyway, it is a forgivable sin that we can attribute to holiday spirits… err… alcohol :)
maybe we can decide that, apart from non being in charge of his own submission, the review and eventual acceptance of papers from editors will be publically available. i.e., that the optional clear name refereeing is not optional for members of the editorial board. maybe this is not a definitive solution, but it can be a starting point.
Yes, that sounds good. I have this idea that the whole anonymity business about refereeing is going in the wrong direction anyway. In an better world, there would be an article and then below a list of names “These people have actively looked at this content and found it to be okay.”
Best of all, in our wiki journal, we can easily include errata. It could say beneath each submission:
submitted by xyz on DD/MM/YY
actively reviewed and found to be okay by abc1, abc2 and abc3
a remaining very subtle subtletly fixed by VeryAttentiveReader2011 on DD/MM/YY.
And the visible version would always be the best possible one that the community can come up with.
I am inclined to ask Jim Stasheff if he might want to be an editor, but the category-theoretic emphasis might be an obstruction.
I don’t understand why it would be an obstruction. My experience is that Jim has always been very supportive of category theory, and is so well-connected that he would not have a problem finding sympathetic referees for submissions. I think he should be approached.
I volunteer to be a referee.
I am inclined to ask Jim Stasheff if he might want to be an editor, but the category-theoretic emphasis might be an obstruction.
I don’t understand why it would be an obstruction. My experience is that Jim has always been very supportive of category theory,
When we write articles together, Jim usually asks me to translate “category speak” to homotopical language where possible. My feeling was that he might hesitate to be an editor of something advertized as being something like a higher version of TAC. This was just a feeling of mine. But I shouldn’t be publically speculating about his feeling but just ask him!
Concerning editors, I am thinking that we should try to change strategy:
“We” as a community of people active on the nForum currently should try not to find among each other a collection of editors. We would have a much stronger case for the whole project if we found an editorial board among people whose expertise is close to the topic of the Lab proceedings , but who are not (currently) among the Forum-community.
I think if we would actually succeed in getting, say, a triple of people such as for instance (drawing on names that had been mentioned) André Joyal, Jim Stasheff and Gabriele Vezzosi (which seems to be in the realm of possibility, even if maybe not too likely) that would be quite a powerful kick-start of the project.
Todd wrote
I volunteer to be a referee.
Great. I was sure hoping you would.
We will have to sort out how referees are being chosen. Usual practice is that an article is submitted to an editor, who then tries to dream up a person who might referee.
But we don’t need to follow that usual practice. If we do proceed with a refereeing practice as I tried to envision above, then it would make perfect sense to reverse this process and allow people to approach the editorial board and say: “I wish to provide a review of the content at [url] such as to promote it to a Proceedings-contribution.”
I agree with Urs #44 (except that I don’t actually know Vezzosi’s reputation myself, but I’m sure that you all know what you’re talking about). I think that we should still ask John Baez (maybe I should ask him?), even though he’ll probably say No.
“We” as a community of people active on the nForum currently should try not to find among each other a collection of editors.
If we manage to get this, that would be optimal. It would also be a solution to the problem “how can nLabers publish on nProoceeding without making it seem it is a private jornal we created for publishing our papers?” (giving such an impression would be the wort possible thing, and it has to be avoided at any cost). If there’s agreement on the triple André Joyal, Jim Stasheff and Gabriele Vezzosi to start with, I can contact Gabriele. Let me know.
@Eric:
failed to mention Tim
of the two Tim’s that come to my mind, one I assumed was implicitely in, the other I would not dare dream of :)
I agree with Urs #44 (except that I don’t actually know Vezzosi’s reputation myself, but I’m sure that you all know what you’re talking about). I think that we should still ask John Baez (maybe I should ask him?), even though he’ll probably say No.
Vezzosi is with Toën one of the founders of “derived algebraic geometry” on whose work all the recent rise of interest in higher category theory among algebraic geometers and homotopy theorists is based.
I think that we should still ask John Baez (maybe I should ask him?), even though he’ll probably say No.
John has been publically repeating that he is no longer interested in what would be core topics of the Proceedings . The names on the editorial board will, more than anything else, convey to people what kind of journal the Proceedings is going to be. I am afraid that a name of somebody who emphasizes that he is all into climate science now would convey a wrong idea.
Concerning #43: I have changed the leading sentence at Proceedings prepatation (nlabmeta) to
[…] that are usefully discussed from the point of view of category theory (nlab) and homotopy theory/higher category theory.
in an attmept to bridge the terminology gap between the two communities.
At some point soon someone will have to take responsability of making a decision on the names of the people to be contacted to be on the editorial board (if interested). My feeling is that this decision should be taken by the steering committee. Once the committee has prepared such a list, we can decide on the forum who is going to contact each candidate editor.
It goes without saying that I wholly approve of this venture, and am wholly unqualified to act as an editor.
I’m particularly interested in what Urs describes in #35 about that atmospheric chemistry journal. I love the idea of person X refereeing Y’s paper, and Z then jumping in to comment on X’s review.
I have sent an email to Jim. Does anyone feel like formulating a letter to André Joyal, asking if he could at all imagine being involved in some way? I am in a rush right now, and I don’t want to write this in a rush.
Yowtch, lots to respond to, and at a bad time for me too. A few brief thoughts:
I think hyperlinks should be specifically mentioned under “differences to traditional publication,” even though they were already mentioned above in “format,” particularly since this is the sort of thing that Eugenia wished for that got this whole ball re-rolling. I added a rough attempt to the meta page.
I am worried about “Iterated resubmission” as it is described on the current version. I think one purpose+advantage of traditional publication, which is not mentioned, is permanence. When you refer to a published paper, you can be sure of what you are referring to and that anyone who goes to track down your reference will find what you intended them to find. Yes, yes, past revisions are always present in the history, but that would mean that references to an nProceedings paper would always have to include a version number. I think that would be at least a little offputting to people used to traditional publications, and also something that many people would forget to do no matter how much we emphasize it.
To brainstorm a little, what if maybe one is allowed to add new pages to a published article (= set of wiki-pages), but not to substantially change old ones (except for, e.g. adding links to the new ones)? I would be inclined to say that the new ones should specifically say at the top that they were added to the publication at a later date.
We should, however, encourage the fixing of typos and errors.
Regarding “how can nLabers publish on nProoceeding without making it seem it is a private jornal we created for publishing our papers?” and the related question of getting non-nLabers to publish here, perhaps “Proceedings of the nLab” is not so good a name after all? If this project is going to be successful, it will have to attract submissions from many, many, people who do not contribute to the nLab, and “Proceedings of the nLab” may make it sound as though those people are not welcome.
A general question which I am a bit embarrased to admit I do not know the answer to: is someone on the editorial board of a traditional journal never allowed to (or not expected to) publish in that journal? I guess there is potential for conflict of interest, but when there are fairly few journals in a given subject area, it would seem rather mean to the editors (who presumably are among the leading mathematicians in that area) for the journal they help to edit to be off-limits to them.
I would be happy to referee.
Hmm - stuck my foot in it there in my last post. After sleeping on it, I’m obviously not qualified as even a temporary editor. However, I’m happy to referee.
As per editors publishing in their own journals, Jim has published in JHRS, of which he is an editor. Also, TAC has a number of editors who I’m sure have published in it.
I think that open refereeing sounds like a good idea, but so does communism on paper. The problem is that reality sucks. The potential for less than objective reviews is high I think. You guys are all friends, for the most part. Can Mike really write a review of Urs’ paper saying that it sucks (hypothetically speaking) under this system?
Far fewer people participate on the nLab than should for a reason. Either they are shy or perfectionists or whatever. If these people are not already participating on the nCafe or nLab, they would be unlikely to want to write open reviews that would appear as something like a blog entry or wiki page. As sad as it is, some people will never take blogs and wikis seriously. Not saying you should cater to those people, but just stating a fact.
@Mike 53: The usual use of Addenda to a published paper, with corrections, slicker proofs etc, is not that far from what you suggest. Additional ’pages’ with links to them might be added but would need reviewing if still to be given the ’stable’ imprimatur of the journal. As you suggest they would have a statement at the top (I suggest the title Addendum be used)
I have been involved with setting up the Journal of K-Theory and there the editors have a detailed set of rules for handling submission by themselves, so there are definite models to avoid conflict of interest.
One problem for contributors to the nLab may be that with the increase in exercises of the form of the UK’s Research Assessment Exercise, to ’evaluate’ research activity, there is pressure (mostly implicit, not explicit) only to contribute to things which will be ’counted’ by the bean-counters of the RAE. This damps down willingness to do refereeing (anonymous) as you cannot claim on a RAE submission something that is not verifiable by said bean-counters.
I think it is problematic to call the proof-checked nLab contributions “proceedings”. It is OK for the exposition of standard material but problematic for really new. Because most of those who will bring some form here first want to publish it in a regular journal (hopefully with a noncommercial publisher like AMS). If you raise it to “proceedings” then you can be subject to the rule of duplicate publication; some institutitions punish this severely, and will hardly look at details that it is actually a prepublication of a sort.
@Mike, concerning #53, resubisssions:
good point, we need to ensure that citations are stable. But I think we can have the advantages both of resubmissions and of stable citations:
if an accepted resubmission affects no previous citations (say if just one more theorem is added) then it replaces the previous page. If it would affect any previous citations (say if assumptions to theomems are being changed) then it gets its own pages and will exist in parallel to the previous version.
So then those following a citation to a specific result still find the previous version. Whereas those looking at the article as such have the option to go to the improved version.
@Zoran: I think we are working now on making the Proceedings of the Lab a place for genuine publications, not for “prepublications”.
Domenico tells me that Gabriele Vezzosi says he would be interested in being a co-editor, once the project gets off the ground. I am still waiting for a reply from Jim Stasheff. If nobody else already did it (did anyone?), i will contact André Joyal about this. If we get names as these in the editorial board and handle things sensibly, then there is, I think, fair reason to hope that this can work out.
Vezzosi is with Toën one of the founders of “derived algebraic geometry”
Simpson started this direction in derived geometry few years earlier (partly in collaboration with Hirschowitz), the polished version of his vision then appeared in the works of his school, including Vezzosi, Vaquie and Toën. Of course, the derived algebraic geometry had its other and much earlier versions in works of Kapranov, Bondal and Kontsevich.
on making the Proceedings of the nLab a place for genuine publications, not for “prepublications”
Then nJournal is much better name than nProceedings. When one talks to physicists, chemists, biologists and so on, then the name “proceedings” associates to a non-important publication; and if you are taking a non-standard format then every additional sign of being non-important makes it harder to make it respected, get listed in various impact factor lists and so on.
On the other hand, the title of tis discussion is about reviewed stuff in nLab. There was a discussion before of reviewed proofs etc. not associated to journals. One should also warn that contruibuting to a non-established journal is quite a sacrifice. I mean if you apply for major grant, like FP7 stuff then they ask about the impact factors and inclusion in various lists of journals. And to get those it is difficult, especially in category theory and even more with unusual formats of publication.
We can still change the title. How about Annals of the Lab ? :-)
Or just Als ? ;-)
Pubications de l’nLaboratoire? :)
Joking, but Zoran has a strong point here: establishing the journal as a recognized one will not be easy and we’ll have to work very hard. A carefully chosen editorial board will be just the first step toward this. While calling it Proceedings can really make everything harder. It is not a nice thing that a journal is classified by its name rather than by its contents, but we must face reality. I must say I liked Proceedings of the nLab a lot, but we should think to alternatives.
I do not know your precise intentions. Depending on your wishes it may or may not emphasise its connection to Lab. I mean it may be ujnderstood more inclusive if you call it Annals (or journal) of higher categorical mathematics or alike; with nLab in title it may look like the “extracts from nLab” or slightly better connotation “Annals of nBourbaki”. One should discuss with people in higher categorical community who are not strictly associated to nLab, which title and format they would feel inspired contributing to.
Proc AMS
yes, but they also have a Journal of AMS, and the Proceedings are less important than the Journal there. I think Zoran is referring precisely to this kind of associations.
Whew! What a lot’s been happening whilst I’ve been “away” (or at least, not paying much attention). Reading through the above, quite a few thoughts occur to me (not least a celebration that Urs has finally learnt how to find the system password!), but I’ll concentrate on the most important ones.
I think that the most important one is the Principle of Selfishness which (for me) underlies the success of the nLab. That is that although we try to work for the benefit of the wider community, our time is so full with other stuff that unless there is some direct benefit to ourselves in a given activity, it keeps dropping down our list of “things to do”. I would quite like to see this taken into account in the proposal. Particularly in light of my next point.
That is to echo the point made by Tim and reiterated by Zoran, namely why should someone “publish” their article in the n-Proceedings? What’s in it for them? I know nothing about the business, but I would imagine that it would take some time before a new journal got in to the “rankings” whereby it would be accepted as evidence of research output by all those official bodies we loathe so much but nonetheless have to abide by. For a new journal to take off, therefore, it has to fulfil another need and provide a place for good quality articles that would not otherwise easily find a home. No evidence of such a list of articles waiting to be published has yet been produced. At the moment, we seem to have Tom’s article and the vague idea of promoting various nLab pages along similar lines to the Manifold Atlas.
In summary, what’s in it for the workers (the editors and referees) and what’s in it for the authors?
I have a slightly alternative proposal, based on something Domenico said:
It would also be a solution to the problem “how can nLabers publish on nProoceeding without making it seem it is a private jornal we created for publishing our papers?” (giving such an impression would be the wort possible thing, and it has to be avoided at any cost).
On the contrary! I think that a “private journal we created for publicising our papers” is a fantastic idea. Note my slight change of wording.
Why do I think this? I think that journals have two main reasons for existence. One is internal to the community: publishing an article in a journal raises it up slightly from the crowd and says, “Here, read this. It’s worth the time.”. The other is external to the community: the number of articles published in reputable journals reflects on the community’s opinion of the author and so provides a measure to those who cannot otherwise assess a researcher’s status in the field. This second external reason is beyond our control, comes about over a relatively long time scale, and cannot anyway occur without the first being addressed.
So I’m proposing that this Proceedings be a way for us to say to each other (and to the wider mathematical community), “Here, read this. It’s worth the time.”. In that respect, it really would be the “Proceedings of the nGroup” since one can imagine it much like an updated and “digitised” version of a society which meets for its members to read out papers that they’ve encountered and think others in the society would be interested in. In that regard, it’s an extension and enlargement of the posts “4001 things you may be interested in” on the nCafe and follows on in the tradition of those journals that still use the language “Communicated by X”.
Here’s some concrete proposals:
Publication in the Proceedings of the nLab is completely orthogonal to publication in “official journals”. We adopt a license a bit like the arXiv’s in that regard. That way, there’s no loss to an author in submitting their work for our perusal. Ideally, there should be some gain as an article published here would go through a round of (proper!) peer-review and so be more polished before being submitted to a Proper Journal. On the other side of the coin, if there’s an article that I think everyone here should read, why should the fact that it’s been published elsewhere stop me?
Publication is not just author-submitted. Editors (or rather, members of the nGroup/nSociety) should themselves go out into the wilderness and bring in articles that they think worth publicising. For the best, a wikified commentary can be provided (somewhat along the lines that Urs has done with the articles he mentioned above).
We divorce the refereeing process from the copy-editing process. It shouldn’t be a referee’s job to correct spelling errors and so forth. Moreover, with editor-submitted articles copy-editing may not be possible, and with wiki-style articles, the wiki-format makes it easy for anyone to do this correction.
The refereeing process should be fully transparent. This does not mean that referees should not be anonymous (though they can choose to waive such) but that it should be extremely clear what it means to say that an article has been refereed. One way to do this is to ask referees to select a suitable statement from a predefined list. The list should be drawn up according to the principle of “What would it be most useful for a reader to know about this paper?”.
@Tom
wait a bit to have my application approved by a human being.
It looks like Andrew Stacey is on vacation or otherwise distracted by more important things. He is the only one administrating the Forum (it is his software and his server and everything).
Regarding stable, permanent versions: I think this is really important. The Manifold Atlas people have a colour system: the static (approved) version gets a blue banner, and the evolving version a green banner. This seems a good system, at first glance at least. Why not copy it?
Done: the Lab was green already, now the
is blue.
I would like to keep the “Lab” in the name.
Sure, the project will have to build up reputation. But if we do it right, it will. And why not have that reputation be manifestly shared with the Lab? After all, the whole point of having this new journal is not that it is just another journal, but one that is combined with a wiki system. That is a central point in case. we should not be shy in saying this openly. On the contrary: what would really badly raise people’s eyebrows is if we behaved as if this were just a run-of-the.mill journal as all the others, and as soon as they click on it they find that it has all sorts of hyperlinks and stuff. No, we should highlight this, not make it a secret.
The system builds up reputation by the quality and usefulness of the material that it publishes. This is the kind of reputation we need to be after. But if we are afraid of saying it’s a wiki, then we shouldn’t do the whole project in the first place.
the whole point of having this new journal is not that it is just another journal, but one that is combined with a wiki system. That is a central point in case.
Yes. And this is also my answer to the question “why should someone “publish” their article in the n-Proceedings?” raised above: because it is a place for fully reviewed wiki-based math hypertexts, and at least now there seems not to be others around :)
The idea should be that an article fon the Lab journal has to be something specifical for this medium, not just an on-line version of a paper which could be published elsewhere. It will be hard to get reputation, but reputation will come from quality of the papers: it will not be a from night-to-day process, it will take time and a lot of work to see the Lab journal be recognized. But if we do everything well, step by step, with no rush, I’m confident we’ll do it in the end.
that an article fon the nLab journal has to be something specifical for this medium, not just an on-line version of a paper which could be published elsewhere.
I would say it may well be based on a paper that could be published elsewhere, but when published in the Als it will be fully hyperlinked and everything.
“why should someone “publish” their article in the n-Proceedings?” raised above: because it is a place for fully reviewed wiki-based math hypertexts, and at least now there seems not to be others around :)
If you are at an average institution hosting not only math, than this argument does not buy you a penny and does not count. Most institutions and many funding agencies add up impact factors. Theory and application of categories was not in SCI list until this year and it did not count. It is yet not in CC and I will loose my present job if I ocntinue sending to such journals. My best paper is the one in Banach Center Publications, it is a summary (with most proofs omitted) of my thesis research with very complete and somewhat unexpected result on existence of locally trivial quantum fibration in a very strict sense. It does not count, while worse papers of mine count. My 93 page survey in London Math. Soc. Lec. Note Series does not cpunt either. At the time it was one and half years of work, I read parts of about 100 references to assemble that survey and it had 3 new independent results which were separately publishable.
Thus if you want to advertise in nice hyperlinked way something here, one can think of some review/pre or republication work format. If one is about to have an independent publication forever here without credit somewhere else than it will be sacrifice for most of the people. I mean, Urs is creating a famous project and probably it makes sense for him: if you are class above you are not measured by standard tools. Lurie had 2 undergrad papers in journals when he got his professorship, and after his book he got associate professor. But everybody knew important and genius of his work there. But for the rest of us mortals and at less noticable institutions people count impact factors, citations in Thomson-Reuters journals and so on. If one wants toget funding within FP7 or similar big financing scheme this matters. Proc. Nat Acad of Sciences count at least 10 times more than TAC after its inclusion into indexes.
I would say it may well be based on a paper that could be published elsewhere, but when published in the nAls it will be fully hyperlinked and everything.
LaTeX from nLab device for LaTeX is hyperlinked. If one sends a hyperlinked version to arXiv it retains hyperlinks.
Zoran,
I am not sure what you are trying to suggest. The problems with citation counts and impact factors that you mention are severe, for sure, but they would apply to any kind of new journal that we’d try to set up, be it related to the Lab or not. As you mention, even TAC has been struggling with this for a long time.
Of course we should try what we can to design the Als such as to ensure that eventually it will be a respectable place to publish. But this we will have to do anyway. I just don’t think it makes sense to hide the fact that it is related to the Lab, if that’s what you are suggesting. if it were not related to the Lab and be in wiki-format, why would we bother with it?
This may not be the place for you to post the publications you need for your employment. And even though you suggest so, I very much doubt that it will be the place for me to get the publications that I need for my employment in the mid-term future.
But it may be the right place right away to post
little facts that we work out on the nLab all the time anayway, and which we might want to have stamped for approval;
articles that do not have to score on Thomson’s index.
Zoran,
maybe for clarification, because I am getting the impression you are maybe assuming the contrary:
I am not planning to submit my “diff-cohomology”-writeup to the Als, and this is not part of my motivation to set this up. On the contrary, as we speak here, I am busy with the mind-numbing and endlessly tedious task of converting all that material into LateX form and prepare it for traditional publication. If you see me commenting and being active here at all, it is just me procrastinating that task.
74:
but they would apply to any kind of new journal that we’d try to set up, be it related to the nLab or not.
This is not true, though it looks plausible. Noncommutative geometry is a new journal, but it is setup in careful traditional way and it had being included in all counts right from the start. I do not say this is right way to go (just taking Springer or alike), though I think it should be considered to have some mainstream higher category journal (my way: to publish, promote it and administer via one of the major societies, say Amereican, European or London Math. Society, in electronic form).
articles that do not have to score on Thomson’s index
The problem is that many in category community already are frustrated. Eugenia Cheng has written some post how even in MathSciNet many of her articles do not count as even MathSciNet chooses to count citation only FROM some journals. This also makes us aware that not only weather the article is cited but also weather the citations in the article are counted for the others. Thus if you cite somebody else whom you appreciate, you may care if your careful citation had the effect.
I do not quite understand the motivation for digression made in 75, I mean why would a particular publication of yours, have anything to do with our discussion on the prospects of some future editorial business. Of course, I hope you will get Thomson credits in this stage and I like the smart decision which you explain, to channel your central work into mainstream, as this may make more prospects for you to get some good European grant (edit: to get it already in near future; whatever you do you will get such in later future, I am sure!).
As you mention, even TAC has been struggling with this for a long time.
Partly this is due to being so esoterically non-applied. I mean just doing traditional category theory. Now with all the issues related to physics, with D-branes, derived algebraic geometry and so on, higher category theory can score better, if carefully administered. Pure category theory has low score within the mathematical community. The citations from TAC did not count even from MathSciNet. So, I think an important think is to have the thing embraced from the start from some mathematical society and strive on having some fractions of contributions related to applications.
It is important to do these things right from the start. A colleague of mine in Central European Journal which is not yet in Current Contents but is in Science Citation Index Expanded from Thomson has sharply risen in the submission rate after it had been included in SCIE. They have immediately more submissions, lower acceptance rate, hence the same output but with greater quality. So, every parameter counts, and just expecting that mere quality made without careful planning, is a loss of time. I think that nLab software can help in getting some edge, but we have to be careful not to have it offset by other parameters thrown into a river.
74
little facts that we work out on the nLab all the time anayway, and which we might want to have stamped for approval
To have a place for short, small and less important but useful contributions, errata, isolated theorems and so on, is a nice thing, which I always dreamed of. Many small contributions are kind of things which is often lost in folklore, and this is good to preserve, although there is usual observation that others are aware of folklore as well. It is good to have such a thing! But I would distinguish it from nJournal. Maybe one needs to talk on at least 2 separate kinds of stable stuff…
Domenico wrote:
It will be hard to get reputation, but reputation will come from quality of the papers
You see the question is what is novelty. Top exaggerate for the purposes of the discussion, the best journals in math like Inventiones and Annals have their quality based on the importance and difficultness of the results. Major open problems are solved there. But my understanding is that some of us want to have a reviewed place to fix various holes, like small items which should be somewhere written, nailing down in precise form folklore results and so on. So, one purpose is to have a basic reliability (checked results) and another is a research quality. Earlier discussions wanted to have software to isolate from nLab some results with verified/verifiable status. Now there is another question on hyperlinked articles in the subject domain central to nLab with nLab somewhere in the name. Now, how will one achieve the quality if one decides to do latter ? The only way to get the quality in the latter, research sense is to have sufficient pull of strong contributors, what nLab community can not, unless one can offer credits of major type to external people. Thus one needs to do that carefully. But I think we still mix the journal and verified nLab content, being more or less personally and folklore research vs. personal research done.
Jim Stasheff says (in private email):
sign me up as potential editor
Tom wrote
ISSNs: to get one, I need to name a “publisher”, with physical address. It’s done on a country-by-country basis, and it would be easier for me to apply to the UK authority, which means the publisher has to be in the UK. I also think it should be someone with a permanent academic position, on either the nLab steering committee or the editorial board. This points to David Corfield. David, would that be OK with you? Are you ready to take on the mantle of publisher of this glorious new journal? As far as I know it carries zero responsibilities, but massive kudos.
Sounds good if I’m needed in this capacity.
I have made some minor edits at Proceedings preparation (nlabmeta) suggested by Jim.
He says we should specify what kind of refereeing we envision. Not sure what to put there.
@79:
But I think we still mix the journal and verified nLab content,
Yes, and I think that’s good. Let’s set up this web for peer-reviewed results now and see what happens. I see no reason why this cannot accomodate contributions that vary in style and scope and purpose. As long as the content is good and goes through peer-review, so be it.
81, please consider having some mathematical society backing it up. Or maybe a math dept having some sizeable category group…
82
Yes, and I think that’s good. Let’s set up this web for peer-reviewed results now and see what happens.
It is good, provided we are conscious of that. It seems you do not like my contribution to clarify and ponder the impact issues, but it would be beneficial to be conscious and take the steps in advance.
@81:
The natural publisher would be … the actual publisher. Since everything is being published on the instiki installation run by Andrew Stacey, it would seem that he is a natural candidate for a publisher.
It seems you do not like my contribution to clarify and ponder the impact issues, but it would be beneficial to be conscious and take the steps in advance.
No, as I said: I do not know what it is that you are suggesting. I see that you are worried. But I don’t know what it is you are suggestion we should do.
As a confirmed cynic, I have my doubts that there is a feasible concept of ’best journal in mathematics’. There is a story which I probably have told before of a mathematician from O….., a well known UK university, who was asked what was good mathematics and replied ’Easy, it is the mathematics done by good mathematicians’. Some days later he was asked again by the same person, if he could say what was a good mathematician, and his reply…. you guessed! ’Someone who does good mathematics!’ (Of course in both cases it should be he and his friends who should decide how to break the circle. I nearly said vicious circle as it can get very vicious. )
The journals you mention, Zoran, have high standards but also a particular agenda, and their judgement as to what is important is not unbiassed, neither is that judgement stupid. It is just not determined by people who are interested in the power that their position gives them. (This does not help our discussion however!)
Journals such as Cahiers have fought for recognition and eventually have won, partially I think because certain people showed the publishers of SCI were discriminating to a level that was verging on the illegal. I do think that someone in, say, the EMS should be asked about the idea, as if EMS or AMS or LMS cooperation could be obtained, then some of the identified problems might more easily be resolved.
I am not worried. Having one more journal can not hurt. But one can miss some opportunities which are pretty obvious to me. Like involving a math society as a publisher. Like delimiting goals. Like having a strategy like insuring indexing. Like having some compatibity with arXiv. Like having a clear goal for people from outside. And so on.
@88
Good, all that sounds good. I am not against any of this. Please go ahead.
But I don’t know what it is you are suggestion we should do.
First of all explaining to potential contributors what is this supposed to be, to offer and so on. It is not clear even to me, even in major questions which I raised up, and to which I did not get answers or got more often the opposite ones (Andrew says that it shoudl be orthogonal to journals, you say no it is just another journal, not for prepublication). No hurry of course – the discussion can revive on after a month.
I have added the following paragraph to the Format-section
Contributions to the Annals may vary in scope and form. The traditional article format submitted to printed journals is welcome, but the underlying wiki-technology may eventually find its own most natural form. The fact that bacground material can easily be linked to on the Lab opens the possibility to have peer-reviewed and published single theorems in the Annals. What counts is the quality and reliability of the content and its useful interlinking with other content.
added one more sentence
On the other end of the spectrum, since there are hardly any size restrictions, whole books can find their place here.
I have now emailed Andre Joyal.
@88
those are all issues we need to consider seriously.
involving a math society as a publisher .this could be great, but could also mean the math society involved will want to have control of the whole thing. I’ve no experience in this, and no clear opinion on this point.
delimiting goals. I would cite/paraphrase from the nLab: The tools of category theory and higher category theory serve to organize other structures. There is a plethora of applications that have proven to be much more transparent when employing the nPOV. Higher category theory has helped foster entire new fields of study that would have been difficult to conceive otherwise. This journal is devoted to these tools and their applications.
having a strategy like insuring indexing. for as much as I dislike all this indexing business, I must sadly admit having an index is what is currenly making the difference. But I must admit I have no idea of how a journal gets an index.
having some compatibity with arXiv. having the papers also have an arXiv version could be a recommended choice
having a clear goal for people from outside. don’t know at the moment
delimiting goals. I would cite/paraphrase from the nLab: The tools of category theory and higher category theory serve to organize other structures. There is a plethora of applications that have proven to be much more transparent when employing the nPOV. Higher category theory has helped foster entire new fields of study that would have been difficult to conceive otherwise. This journal is devoted to these tools and their applications.
I have added that to a section “Goals” at Proceedings preparation (nlabmeta)
(By the way, this page is on nLabMeta so that it is freely editable. It is potentially less frustrating for all of us to include suggestions there directly, than to try to make me to do it ;-)
@Tim 56: I did intend to allow significant expansions, as I think Urs intended, not just corrections, slicker proofs, etc.
@Urs 58: That sounds reasonable, although I’m a bit wary of allowing insertions in the middle of a page. In particular, adding a new numbered theorem in the middle of a page would change the numbers of all the subsequent theorems.
Not sure what I think about the weightier issues, nor have time to think about them today….
@ Mike #53
references to an nProceedings paper would always have to include a version number. I think that would be at least a little offputting to people used to traditional publications, and also something that many people would forget to do no matter how much we emphasize it.
The arXiv has this problem. You are supposed to cite a specific version of any paper that appears on the arXiv, but people hardly ever seem to do that.
@ Urs #58:
if an accepted resubmission affects no previous citations […] then it replaces the previous page. If it would affect any previous citations […] then it gets its own pages and will exist in parallel to the previous version.
If we adopt this, then we should enforce it strictly. No “I’m sure that nobody has cited this, and it’s an important change.”, or the distinction will be lost over time. (I’ve seen similar things happen with keeping revision history in Wikipedia.)
If we adopt this, then we should enforce is strictly.
Sure. I think all this can be handled. When in doubt, we just add a new parallel version.
The arXiv has this problem.
Yes, but on a smaller scale, since it also discourages having a lot of revisions. I feel like the great majority of arXiv papers have only one version, or one preprint and one published version, or a few more but which make only minor changes.