User login

You are here

Open Source Publishing

ericmock's picture

I am putting together a proposal in response to NSF's Engineering Virtual Organizations solicitation regarding what I call Open Source Publishing.  The proposal can be found (and edited if you setup an account) here.  I would encourage you to register and edit if you're interested.  All revisions are saved so there is no risk of messing something up.  The proposal is fairly unstructured at this point and consists of mostly just my thoughts.  I am working to organize it better and any suggestions would be appreciated.  If you're interested in officially becoming a part of the proposal, let me know.

As there has been significant discussion about publishing and open access on iMechanica, I would greatly appreciate any feedback people can give.  Also note that the current plan is to incorporate the open source publishing system with iMechanica.  This stemmed from a discussion with Zhigang after I realized some components of the NSF proposal he posted were similar to what I had written.

I would like to use the proposal as an example of my vision for open source publishing so please do register, edit, and post comments in the discussion page (see tabs at top of page).


It's done!  Thank you all for your thoughts (and the obvious time you spent considering the issues).  I think the proposal benefitted greatly from this discussion.  Hopefully NSF reviewers will be convinced this is a good idea.  I've had some recent luck with NSF so hopefully that will continue. 

The proposal has now been submitted.  You can see the final version at the link above.  To view in HTML click on Submission Repository.  To view in PDF, click on Typeset version.


Teng Li's picture

Dear Eric:

Thanks for posting your proposal draft. 

Just registered and quickly went through the draft.  Reading your proposal reminded me those discussions when some of us were exploring a wiki platform,, in the Spring 2006. Some of the initial plans on wikimechanica have been delineated in a NSF proposal Zhigang recently posted in iMechanica.

I feel that the technical part of the plan (e.g., backend and frontend softwares) is strong, while it can be clearer on the operation mechanisms of such an open source publishing system. Some quick questions out of my head are, how to motivate senior experts to join and form a reputable editorial board, as many of them are reluctant to dive into the Web 2.0 wave?  How to effectively motivate authors to submit manuscripts?  Or in general, an answer to "What's the benefit to me for using this open source publishing system?" 

I'll look further into it.


ericmock's picture


Thank you for your comments.

I have read Zhigang's proposal and discussed the open source publishing proposal with him extensively.  Most of what is currently on the wiki was written before Zhigang posted his proposal and once I read his proposal I realized that we have somewhat similar ideas.  However, I think what I am proposing is quite a bit different from, which I understand to be more of a mechanics encyclopedia (not an online journal and discussion forum).

You have certainly brought up many of the points with which I have struggled.  The individual benefits are somewhat nebulus but I see them as follows for authors:

1) you get your ideas out quickly (basically everything that is posted is published but not necessarily 'accepted'),
2) you get quick feedback on the paper,
3) you can quickly respond to reviewers to clarify points and make changes,
4) you do not have to wait for two/three reviewers to respond,
5) reviews will most likely be much more insightful as they can be attributed to someone

and for reviewers:

1) you do not have to wait until you have a comprehensive review to respond,
2) you can ask the author to clarify a point if something is confusing,
3) you will get credit for 'publishing' a review.

Please add to this list if you see other individual benefits. 

Overall, however, I think the benefit will be to the mechanics community as the 'signal to noise ratio' of good papers will go up from good natured discussion and debate about papers.  Plus, even 'bad' papers can be submitted and maybe some good ideas can be pulled from the paper to generate a 'good' paper with the help of the community.  Plus, I think this will foster collaborations to a much greater extent than the current publishing system does.

Establishing an editorial board will certainly be a challenge and is one of the reasons I am soliciting help from this community.  While it would be great to have a preliminary editorial board together before sending the proposal to NSF, I do not think it is necessary.  This can be done in the first year of funding while getting all the other resources together.  Plus, any 'old school' editors could simply do their job as they do currently (i.e. read the comments/reviews and make a decision).  There is no need for them to be well-versed in the technology.

Zhigang Suo's picture

Dear All:

Eric has experience with a log of web technologies.  I've read an earlier draft of his proposal, and found many of his ideas stimulating.  I hope many of you will read his proposal, and give him feedback.  This quick note records one initial thought on his early draft.  (Eric:  Sorry to be brief.  I need to pack for a trip to France.)

My own understanding of open access publishing is not nuanced.  While I like the ultimate simplicity of open access, I don't see a clear path to accomplish it any time soon.  Growing up in China, I have tried to avoid revolution as much as possible.  Revolution takes energy, and a lot of heads roll, not always for right reasons.  The results of incremental changes can be fantastic, given enough time.  Just look at you and me and our children.  Of course I have assumed that you also believe in evolution. 

A mission of iMechanica is to enhance communication among mechanicians, in whatever way practical.  For example, the operating notes of the jClub stipulate that a Discussion Leader can select any paper, either a note posted in a repository or a paper published in a journal.  In my own mind, iMechanica is neutral on the issue of open access. 

Of course, iMechanica belongs to whoever uses it.  As a user, your thoughts on issues like this will guide the future development of iMechanica.

You may also want to see a related discussion iniciated by Dean Eastbury, a Senior Publishing Editor of Elsevier.  An earlier proposal to NSF (unfunded) to start iMechanica is posted online. 

ericmock's picture


I hope you enjoy France.

Indeed, a large portion of the discussions Zhigang and I have had have been related to open access.  And this is a point on which we disagree somewhat.

I am personally a proponent of open access but you certainly will not see me leading the charge for it.

The open source publishing ideas in my proposal did not stem from a desire for open access, but from a desire to change how the review system works.  However, one (inevitable) result of an open source publishing model will be open access.  Due to the support for open access within the academic community and funding agencies, I used open access as an argument in support of the ideas.  I am by no means 'married' to including this argument in the final proposal.

I also agree with Zhigang that a revolution is not needed or desired.  I certainly do not want all the hard-working people at Elsevier, Springer/Kluwer, etc. to lose their jobs.  In fact, at the risk of not sounding bold enough to NSF reviewers, I state in the proposal that I would like to take many evolutionary steps towards the proposals ultimate goals instead of a revolutionary one.  However, I do see open source publishing as a way to slowly move towards open access.


Pradeep Sharma's picture

Dear Eric,

Your post and proposal have some very interesting thoughts. Most of the issues you raised regarding the peer review process as well as open access resonated however I have some concerns. Perhaps you have already thought of what I am gong to say and addressed them but it was not clear to me:

(1) Despite all the problems with anonymous peer review, can we really expect candid criticisms if identities are disclosed? The exchanges will be a lot more civil (which in the current system, sometimes are not) but I am afraid things may go the opposite direction---they will be too non-critical. This is especially an issue if younger folks are reviewing papers by more established researchers.

(2) You correctly pointed out that some of the problems with our review system stem from the lack of a reward system for what has indeed become a thankless task. Perhaps the existing publishing companies (e.g. Elsevier) should pay the reviewers. In case of non-profit (or society journals), perhaps there could be "best reviewer awards" or some other form of incentive. These (and perhaps other similar suggestions) may help improve the indifference shown by many referees.

(3) Regarding "open access", I certainly endorse the philosophy.....the practise is another matter. Unless, the academic administrators, federal agencies and yes even us start valuing journal papers published in an open access environment as highly as in the traditional journals, we are unlikely to convince academicians who have not achieved their final propotion, to participate. This can be changed of course.....but will take time and perhaps leadership among the established researchers "who have nothng to loose" by taking the first few bold steps.  A simply evloutionary step could be to unite together (like a union--via iMechanica of course!) and "persuade" the journals we typically publish in to allow us to retain copyright and post post-prints websites. I doubt if institutes will cancel subscriptions because of this practise.

With regard to open review,  It would be a great help if the reviews were open even if they were anonymous. For example, it should be possible  to see the reviews for every paper which gets submitted/published to a journal. 

ericmock's picture

This is a great point and something I had not really thought about explicitly.  In fact, I think this is a great argument for the proposed 'transition' plan.  I.e., even in the early stages of mostly anonymous reviews, having the reviews open for all to see would be a benefit.

Pradeep Sharma's picture


I agree.....regardless of other issues under debate, your suggestion of reviews being accessible to everyone (even if anonymous) is very good. Given that this suggestion may be somewhat independent of other long term reforms under discussion AND in fact can be implemented immediately should the editors wish to, I would suggest you to spin off a separate post on this and attempt to "persuade" the mechanics editors to do so.

Zhigang Suo's picture

I like this idea.  As an experiment, I posted the preprint of a paper on iMechanica on the same day when the paper was submitted to JMPS.  The posted paper received a large number of comments.   I then posted reviews sent to me from the journal, as well as my responses.  When the paper is included in JMPS, I added the DOI of the paper to the post.

The subject of this paper is of particular interest to me, and I'm really curious to find out how other people think of the subject and of our treatment.  I'm grateful to all the comments and the JMPS reviews.

Thus, an author can take inititive to make the process very open.  Of course,a piece of software that streamlines this process will be valuable.



This is exactly what I'd like to see more of. Kudos to you for posting your reviews online -- this is the first time I've seen anyone do this.  As you rightly point out, authors can initiate the process. Now, only if journals start doing  this as standard operating procedure.

 I think that there would be e a great improvement in the quality of papers if reviews were publicized. It would also be an invaluable resource to future generations. It would be interesting to see reviews for classic papers, for example Eshelby's or Von Neumann's papers.   


ericmock's picture

I have also considered doing this but worry about posting other people's writings without their consent.  Even though the reviews are anonymous, the reviewer wrote the comments under the belief that they would only be read by the editor and authors.  Posting them without his/her consent seems to me to be a violation of the reviewer's copyright.  I'm by no means a lawyer so I'm not sure what the law says, if anything, about this.  However, I would be very cautious posting reviews without consent.


Zhigang Suo's picture

Dear Eric:

I totally understand your concerns.  However, it is not just any writing of other people.  It is anonymous writing about my work.  I'm unaware of any legal issues concerning circulating an anonymous review of my work.  On occasion I've seen other people sending emails containing such reviews.  Even more frequently, people quote reviews of their papers in conversations and lectures. 

The Internet is a new mode of communication compared to conversations and lectures.  I suppose the law is not explicit about some of the issues.  People have the right to experiment with new possibilities, until the new norms emerge.

It feels wrong to me for an organization or an established social norm to limit such a right to experiment.  If I ever receive a notice that states such a restriction, I'll consider posting it online as well, so that people can discuss the issues.  If I do post such a notice sent to me, would you think that I violate the copyright of the sender of the notice? 

Perhaps people should disucss the pros and cons of posting such reviews online.  Such a discussion is a way to formulate a new social norm, if not the law.

ericmock's picture


I agree and would love to post reviews for all to see.  In fact, I got some pretty lousy (in the sense that the reviewers obviously gave little effort) ones from a well-respected journal recently.  I would love to post them, but I do think there might be legal issues (although it's unlikely anyone would enforce them).  I also am not sure it's 'socially kosher.'  Kind of like replying to an email with an additional recipient and quoting the sender without his/her okay.  Basically, the sender wrote the message to someone specific, not expecting that message to go to a third party.  Although that's the risk you take when communicating via email, and I view it as the big difference between emailing personal messages and chatting in a blog.  The anonymity of the reviews being posted make it less clear.  However, posting a review for all to see might wind up revealing the reviewer.  I know I would write reviews in a more generic style if I knew they'd be posted.

I would be interested in hearing from others to see how they feel. 

Check out  Why are people reluctant to join in open review? Nature 447, 1052 (28 June 2007).

ericmock's picture

Pradeep and Biswajit,

Thank you for the comments.  You guys have brought up many of the issues that are making this proposal so tough to write.  I have thought about many of these but have struggled with compelling answers.  Some are partially considered in the draft but the arguments need strengthened.

In separate posts I will give my thoughts on some of your concerns.  Hopefully this will help sort things out in my head so that I can put them in the proposal.  Basically, I need to propose a better transition plan.

Pradeep wrote, "Despite all the problems with anonymous peer review, can we really expect candid criticisms if identities are disclosed?"

This is a tough one that I have thought a lot about.  My conclusion is that it is really hard to know what people will do and I think it really depends on individual personalities.  However, trying to determine the collective response is difficult.  While the 'average personality' is probably conservative (for fear of hurting one's career), I don't think success will depend on the average personality.  Hopefully, the more bold (and tenured) of us will comment without fear and others will follow.  However, I also realize that there needs to be a transition from anonymous to open reviewing.  My current thinking is to allow registered users to post anonymously but encourage people not to.  Users will be able to change their posts from anonymous to signed at any time.

For example, let's say Wolfgang Pauli (a notoriously cantankerous person, posts an article that has a few mistakes.  A young professor named Pradeep notices the mistakes (at least he thinks they are mistakes) and makes a contrite anonymous post saying something like, "I believe there is a mistake but it might just be that I don't understand."  This causes a few more people to look at the issue and also post (maybe anonymously, maybe signed) that they also think there is a mistake.  Now, with some corroboration, Pradeep is more confident with his assertion and would like to get some credit for pointing out the mistake.  Thus, he decides to change the post from anonymous to signed and everyone is impressed.  

Consider the scenario where Pradeep is wrong and others post to help him understand _his_ mistake and Pauli posts a scathing reply saying, "This isn't right. This isn't even wrong."  Pradeep has learned something from the other posts but decides it's best to leave the post anonymous for fear of retribution from Pauli.

On the other hand, there is no point in posting anonymous "This is great paper.  You're brilliant." comments.  Posting signed comments like this will not look good on the reviewer, especially if others find mistakes.  Thus, I do not think there will be a flood of young "butt kissers" stroking the egos of established researchers.

Pradeep Sharma's picture


You have made some persuasive arguments in this and other responses. I am still unsure about several aspects but we will never know how things are going to work out if someone does not experiment. In that sense I hope you get the green light for your proposal. We will be in a much better position to debate once you have tried something out for some time.

ericmock's picture


Please do elaborate about the aspect on which you are unsure.  The more I can try to convince people here, the better the proposal will be.

In fact, I'm about ready to go live with a test site.


Pradeep Sharma's picture


I still believe, imperfect though may be, that anonymous peer review is the foundation of our scholarly publication process. Stating this however does not resolve the fact that barring a few , many referees tend to display poor work ethics. I felt that perhaps in the case of "for profit" journal, this aspect may improve with financial compensation for the reviewers. This should not be done for the society journals of course and there I am at a loss for ideas how to resolve this "indifferent or ineffectual referee" problem.

In case of "for profit" journals, the editors and publishers will adapt by choosing fewer but more realiable referees. Even though small, since there is an incentive, more people will be willing to engage AND perform well (so that they are not dropped from the reviewer pool). I believe that the editors and publishers can set it up such that it is considered a privelege to be ask to review a paper (not a burden).

In summary, I am pro-referee anonymity and feel that, at least for the profit making journals, the quality of refereeing can be improved by financial recompensation. I don't have any suggestions for the society journals.


ericmock's picture

Thank you for clarifying what bothers you.  And I have to agree with you that suggesting an open review system really makes me nervous.  Doing so will requiring a great deal of explanation to preempt any fears any one of the reviewers might have.  Unfortunately proposal submission/review is not an iterative process (at least at NSF).
I spoke to my brother at length yesterday because he works at the National Institute of Justice as an editor and web content developer.  He mentioned that NIJ pays their reviewers (although they do not convene panels like NSF).  They do not pay much and decided one year that it wasn't having an effect so they stopped.  Needless to say, reviewer payments were reimplemented the next year.  
What they realized is that the (little) compensation does not greatly affect the quality of the review but improves the timeliness (since payment is contingent upon submitting the review on time).  Some of this is due to the fact that they often ask professionals (e.g. chiefs of police) to review proposals and they are basically doing it on their own time.  Since none (and all) of an academic's time is his/her own, I do not think timeliness will improve for this reason.  However, I think it would improve for another reason that became clear during our conversation.  The lure of, say, $100 might push a review up on my (nebulous) to-do list so that I actually get it in on time.  I do not think the quality would suffer because I usually know what I want to say the day I get the review (and read the paper since I'm always curious about the content).  It just seems to take me a long time to get it on paper because I'm never sure that I'm not missing something.  I'm assuming others do the same.
That said, I also think a discussion-type review system would help this.  No one would feel obligated to write a lengthy review.  I would likely post comments the day I get the review since I would know the authors could immediately clarify things.
What do you think of the idea of letting people post anonymously but then go back and sign the post if they feel comfortable having the comments attributed to them?  This would allow them to receive community recognition which I think is much more important than financial compensation.  Plus, they could technically list every review as a (short) non-peer reviewed publication since it would be open for all to read.  The reviewer would not be anonymous to the editorial board so that overly scathing and non-constructive reviewers would (hopefully) mute their comments.
Again, thank you all so much for taking an interest.  And please visit the test site and provide feedback if you a few moments.


Pradeep Sharma's picture


Voluntary disclosure of identity should work. I believe, the Royal society publications routinely ask referees if they wish to reveal themselves (and on occassion I have). In case you have not already seen it, please check out the post recently made by Charles Steele on journal pricing/Elsevier and related topic. I belive, it is quite relevant to your own ideas.


ericmock's picture

Please also see my response to Dean Eastbury's post about Elsevier's 'initiatives' at
Basically, Reed Elsevier's operating profit margins for the scholarly publishing division are an astounding 35% and amazingly considerably higher than their operating profit margin for their LexisNexis division (both have about the same revenue of a little over €2 biliion).

Mike Ciavarella's picture

michele ciavarella

ericmock's picture


Regarding your point about rewarding reviewers, I have also given this much thought and I really have not figured out a way to do it without giving reviewers 'academic' credit by making their reviews public.  Paying reviewers is something I considered but, as I have reviewed over 100 papers, I thought this would look like I was just serving my own interests.  Regardless, I don't think a payment system would work because I think money is not what motivates most academics (and especially the ones that give good reviews).  Now, if it were a business journal...  Plus, one would have to decide if a review was good enough to warrant paying the reviewer.  Simply paying for any review would probably just exacerbate the problem.  Likewise, giving some kind of reviewing award would also require that someone judge the quality of reviews.  This would likely mean that an editor would then be charged with reviewing the reviews, which would mean that person would actually have to carefully read every submission to determine the quality of the review.  I don't see this happening.  It's just too much work.


Kaushik Dayal's picture

ericmock's picture


Regarding your point about valuing open access publications, I think we need to decouple open access and peer review.  Obviously, we can have both.  PLoS journals demonstrate this.  In the proposal I want to make it clear that there will still be peer review and official acceptance in the open source publishing system.  The idea is that anyone (within reason) can submit anything to the journal and it will be available for all to see and discuss.  Just like any other journal, these papers will not be considered accepted (but they will be available).  Once there has been sufficient review and discussion of the paper, an author (or nominator) can request the paper be considered for acceptance into the 'archival' section of the system.  At this point, an official acceptance letter will be sent to the authors and they will be able to claim the work as accepted into a peer-reviewed journal.  How 'prestigious' this acceptance is will depend on the level of quality established by the editorial board.



Your proposal contains a number of interesting ideas that should be relatively straightforward to implement.  In the event that NSF does not fund the proposal will you go forward on your own or wait for a more opportune moment? 

I think your proposal suffers from some of the same flaws as Zhigang's iMechanica proposal, i.e., people who are not very well versed in web technology might think that it is too ambitious.  Also, getting hold of a board of editors for such a journal will be non-trivial given the current glut of journals in the market. 

A journal is unlike most other web based projects.  You cannot have a small scale model of  a journal to test out the idea though individual enabling technologies can be tested.  That makes open publishing a business because any open journal will have attendent costs - archival, storage, servers, system administrators and such.  Your proposal does not talk about how the bills will be footed in the long term.  What do you plan to do?  Ask authors to pay a publication fee?  A business model for the project will certainly be useful (I'm not sure whether NSF review panels view business models favorably).

I haven't thought much about open publishing but I sure would like open access.  I'll give you a good reason why (as far as I am concerned).  The University of Utah has cut several Elsevier Journals in the past five years due to excessive cost.  One of the cut journals is Computers and Structures.  My latest paper was published in that journal last month.  But I have no way of accessing an electronic version of the paper in published form other than by paying a fee.  I'll have to wait a few months for the paper postprints to be delivered.  That is ridiculous.  Open access will make such issues a thing of the past (hopefully).

Does open access require open publishing?  Probably not.  Does open publishing have to be completely open?  Not necessarily.  Given the choice, will reviewers prefer to remain anonymous?  Not me, but I'm sure there are others who will prefer anonymity.  The options can easily be kept open. 

The very nature of the medium suggests that the initial users of your open publishing model will be early adopters - the young and the adventurous.   Does that mean that the quality of an open journal will not be particularly high in the beginning?  Probably.  But the experience of PLoS suggests that the quality can improve rapidly.

Most people publish in order to advance their careers.  In the absence of royalties and such, career aspirations and impact are the prime movers of academic journals.  How will publishing in open journals help in advancing careers?  Only after such journals are well established.   That will take at least 10 years if not more in a saturated market.  But we will never knows unless the idea is tried out.  Catch 22?

Also, why will anyone other than the assigned reviewers (if any) ever comment on a paper that is revolutionary or not a "hot" topic?  You could check the number of comments a typical PLoS paper gets.  People tend to comment only on hot or controversial subjects that they think they understand.  Does the fact that a topic is not "hot" make it less important and unworthy of being published?   You proposal seems to suggest that that's the way the review process should be conducted.  Could you clarify/elaborate on this point?

I think scientists and engineers are quite conservative as a group and will not adopt a new model without some soul searching.  A tryly open publishing model will take a generation to get established.  But we have to start somewhere.

ericmock's picture


Again, I would like to thank everyone for the feedback.  It has really helped me formulate my ideas in a much more convincing way.  This has been great constructive criticism and I am sure the proposal will benefit.  Hopefully I can preempt some reviewer comments by addressing yours.  It has really made me think more deeply about what I want to do.

 I would like to address some of your points now.

 First, I am trying to actually set up a mock system that will be ready at the same time the proposal is submitted.  It's already functional and I will post the URL in a few days.  While I'm not sure how NSF views directing reviewers to a web site that demonstrates the idea (I know using the web to circumvent the 15 page limit is frowned upon), this is what I plan to do.  The hope is that the reviewers will see that it is possible (by example).  Although they may wind up saying, "Then why does he need funding for this?"

 I think you can actually have a small scale version of the system (and I allude to this in the draft proposal).  While it won't at this size be a true journal, it will be a place for people to post preliminary manuscripts for discussion before they submit them elsewhere for publication.  I.e., there is no reason that just because you submit something to the Open Source Publishing system you can't submit it somewhere else.  Remember, I will not ask you to sign over the copyright ;-).  If the other publisher forces you to remove the manuscript for the OSP system, then you can do that; but the reviews and comments would remain.  I don't think this will be a problem though since most publishers don't have a problem with people depositing things in the arXiv.

 A model for self-support (after all that money from NSF runs out) will certainly have to be in the proposal.  I talked with Zhigang quite a bit about the different financial models (i.e. author pays vs. reader pays vs. government pays).  I'm in favor of combination of reader and government pays.  Basically, this will mean charging (or asking) university libraries to pay a nominal fee and NSF/ARO/ONR/AFOSR/NASA/NIH to commit a small amount each year.  The problem is the perception that publishing quality content costs a lot.  In fact, I think Elsevier throws around this $400M investment in ScienceDirect to scare off the open source crowd.  If they really invested this much in SD, then they got ripped off by someone.  (I'd quit my job and do it for 10% of what they paid ;-).)  The other data point for cost is the amount, e.g. PLoS, publishers charge authors--$2000 per paper in the case of PLoS.  (I assume they only charge for accepted papers.)  They claim this is just to cover the cost of publishing.  While I do not think anyone there is necessarily padding his/her pockets, I certainly do not think it _has to_ cost this much.  Let's take PLoS for example.  They have offices in San Francisco (185 Berry St) near the stadium and not far from the Moscone Center.  I have no idea how much space they have but there are at least 25 staff people there (and this does not count the individual journal 'Teams' many of whom are located in the UK, see  Let's say they have 2500 sq. ft. of space in SF.  I have no idea what rents are like in SF, but it's about $20/sqft/year in central PA.  So let's assume it's like $100/sqft/year in SF.  That's $250,000/year in rent.  Now if you assume those 25 people cost roughly $100,000/year.  This might seem like a lot but you have to include benefits and the cost of living in the Bay Area.  That is another $2.5M in staff (not counting the 27 other people listed as staff).  While they are a non-profit, they are certainly spending a lot.  I really can't see what they are paying all these people for.  At least four of them are just there to handle the finances.  Their web sites are really not that nice and the typesetting is fairly basic (it looks like they're using LaTeX for equations) but it looks like they might have a professional redoing the figures.  I think my system looks just as good (as you'll see).  Why they need eight general editors and another three to eight for each journal is hard to understand.  For roughly 15 'Research Articles' per month per journal, that seems like a whole lot of editors.  My brother is an editor (officially the Web Content Manager) at the National Institute of Justice's Office of Justice Programs so I'll ask him if this seems excessive.  Note that PLoS journals also have Feature articles and other things that likely require more editing than typical research articles.  Plus, they seem to pay much more attention to readability for the 'general' public than most engineering journals.  Again, requiring more editing time.

Personally, I think with a supportive community, it will cost no more than $200k per year to publish.  Even less if we can get some free editing from English faculty and students (who are interested in a lucrative, at least for an English major, scientific editing career).

 Are these arguments compelling?  Do my numbers make sense?

 Well, enough for now.  More later.



The $200,000/year price tag seems a bit high for one journal.  But if you have multiple journals the cost per journal will go down because the basic infrastructure will remain essentially the same.   I'm still not sure why NSF should pay for this venture - what tangible benefit does the taxpayer or science get out of it?   It think a statement of benefits should appear prominently somewhere close to the beginning of the proposal.

I don't have anything useful to day about your model at this stage.  I'll think about it some more. 


ericmock's picture


Thanks for the continued feedback.

I'm not sure if $200k/yr is high or low for sustaining the effort.  A lot depends on the quality.  Obviously PLoS has gone very high-end but is spending at least 10X more.  Also, $200k/yr is not per journal as there will obviously be economies of scale.  Basically, I came up with this number as $100k/yr each for an IT person and an editor.  Neither are completely necessary.  An editor will improve the readability of the articles but if the community doesn't mind reading them raw, then professional editing is not really necessary.  IT support is probably necessary but much of that could be donated by universities or funding agencies.

I will add the link to the NSF solicitation in the original post but you can find it here .  I think the Open Source Publishing system could fit this really well.  In fact, I think Zhigang's proposal probably fits this call better than the CyberInfrastructure solicitation that was available last year.  The synopsis states:

"The primary purpose of this solicitation is to promote the development of Virtual Organizations (VO's) for the engineering community (EVOs).  A VO is created by a group of individuals whose members and resources may be dispersed globally, yet who function as a coherent unit through the use of cyberinfrastructure (CI).  EVOs will extend beyond small collaborations and individual departments or institutions to encompass wide-ranging, geographically dispersed activities and groups. This approach has the potential to revolutionize the conduct of science and engineering research, education, and innovation. These systems provide shared access to centralized or distributed resources, such as community-specific sets of tools, applications, data, and sensors, and experimental operations, often in real time.

With the access to enabling tools and services, self-organizing communities can create VOs to facilitate scientific workflows; collaborate on experiments; share information and knowledge; remotely operate instrumentation; run numerical simulations using shared computing resources; dynamically acquire, archive, e-publish, access, mine, analyze, and visualize data; develop new computational models; and deliver unique learning, workforce-development, and innovation tools. Most importantly, each VO design can originate within a community and be explicitly tailored to meet the needs of that specific community.  At the same time, to exploit the full power of cyberinfrastructure for a VO's needs, research domain experts need to collaborate with CI professionals who have expertise in algorithm development, systems operations, and application development."


ericmock's picture

While nothing is lamer than replying to yourself, I am going to.  I finally got my hands on PLoS's financial statements (their IRS Form 990) and have to say that I was almost dead on with my estimates above.  Salaries and wages for 2006 were $2.55M with compensation of officers and directors being an additional $351k.  With payroll taxes, pensions, etc. the total cost of employees was about $3M.  Occupancy (rent) was $247k.  Their total revenue was roughly $5M and their total expenses were $6.3M which caused their net assets to drop from $3.4M to about $2M.  I.e. they are burning through that $9M they got from the Gordon Moore foundation pretty quickly.


Do you have any idea of the number of people involved and their job descriptions?  PloS won't last too long at this rate! 

ericmock's picture

You can find information about all the PLoS staff at  They all have short bio's on the site.  It basically looks like PLoS really decided to go all out and compete with Nature, Science, etc.  I think they took the 'build it and hope they come' approach.  I'm not sure what the trends at PLoS are, but I suspect they planned on having substantial losses for the first few years while they built a reputation.  Also note that they are raising their rates to publish.  Again, they're probably just trying to figure out the business model that can make it self-sustainable.

Robin Selinger's picture

First the negative and then the positive.... 

I see no need to create a new eprint archive. The best place to post eprints is at Why create a separate and competing online entity? Perhaps we should request that ArXiv create a new subject category or subcategory for the IMechanica community. I suggest "mechanics of materials" or mech-mat, in contrast to the existing category cond-mat (condensed matter.)

Likewise I am not sure we need our own separate wiki resource. I'd rather join forces with the rest of the world to build up the number and quality of mechanics-related entries in Wikipedia.

Now, the positive...

I really like the idea of using IMechanica to offer a public, archived forum for discussion of papers of all kinds (published or preprints) from all sources. To prevent bad cases of "bloggers's regret," contributors must always have the privilege to withdraw their posted comments at any time.

I also love the idea of posting educational materials of all kinds online, including everything from a single set of notes on one specific subject to a library of complete online textbooks.

The IMechanica community should also consider posting full talks online. We could even set up an IMechanica podcast system in which one posted talk could be selected each week for automatic download to members who sign up.

As an example of what's possible, see the KITP website at Santa Barbara. They post full streaming audio and the original powerpoint file with associated animations. Streaming video of each talk is also posted but with only one video frame per viewgraph. For an example you can see my talk there entitled "Simulation of Plastic Deformation and Ductile Fracture: Size/Rate/Temperature Effects"

FYI, the associated paper is in this month's edition of JMPS, at this link:

Making a wide array of online resources available to everyone will help all of us, will level the playing field a bit for those at less advantaged institutions, and will speed the overall progress of our collective research enterprise. NSF would be wise to encourage community-wide efforts of this kind in many disciplines.

-Robin Selinger

Robin Selinger, Professor, Chemical Physics Interdisciplinary Program

Liquid Crystal Institute, Kent State University

Kent, OH 44242

phone (330) 672-1582

ericmock's picture


I'm not sure if you had a chance to read the proposal and maybe my original post did not explain the idea in enough depth.

What I have in mind goes _way_ beyond the arXiv and is not simply another eprint archive.  It is basically what I think the arXiv should have been to begin with (but they did not have the technology at the time).

I (and especially my friends in physics) are very aware of the arXiv and think of it as a great place to deposit work.  However, as long as there is no peer review for papers posted there and no good forum for discussion, it will simply be an 'eprint archive,' not a system for open source publishing.

Please read the entire proposal to better understand what I am proposing.


Andrew Norris's picture

This is a very interesting discussion - an excellent example of the use
of iMechanica.  Thanks to Eric for initiating it.

There are several issues: open access, open source,
peer review, etc,. Some of these are not completely well defined for me
- I need to do more research on open access.  But if I
could throw in a few thoughts on the topic of peer review ...

With the emergence of web 2.0 several journals have started
experimenting with "community review".  A good example of a progressive
approach is Atmospheric Chemistry and
, a journal of the European Geosciences Union.  This combines
the traditional peer review with an "open discussion" period that
allows anyone to comment.   All reviews, author responses and the
discussion are online for everyone to see.    This graphic gives the

open review system


Here is an example of a paper that is currently open for discussion and here is a list of all the papers open for discussion. 

This example of a new review system prompts these thoughts vis a vis Eric's ambitious project:

1) It would be useful if the proposal tried to evaluate the
success of these various experiments in modifying the peer review
system.   They provide hard data on the uncertainties Eric and Pradeep
raise: i.e. what will real people do when given the opportunity to
comment on a paper.   Obviously, it depends on the precise model, the
level of activity and involvement of the community, etc, but as the Atmospheric Chemistry and Physics case shows, people will get involved. 

2) Whether the new review model leads to "better" papers and research is a much more important issue. I suspect that the data so far might not be sufficient
to answer this conclusively.  But I am sure there are people out there willing to
provide opinions! 

3) In terms of strategy - how to get from here to there - it seems that
it would be best to try to launch something like this on the back of an
existing entity, such as a professional society.    A little bit of
prestige goes a long way: the arXiv started off with the (unofficial)
LANL fame behind it (would it be where it is today if Cornell library started it?);  iMechanica blossomed from ASME AMD though its
quite independent now afaik.   The Atmospheric
Chemistry and Physics
journal is a case of a society deciding to make a
bold move.   Many societies are not as adventurous especially when it
comes to the cash-cow of their prestigious journals.   But some might
be willing to try out a spin-off publication - I am thinking of the
example of the Acoustical Society of America's ARLO
with which I am most familiar.

Overall, it seems that there is considerable uncertainty in the
outcome of Eric's project.    Using data from ongoing experiments in
other disciplines would be helpful, and avoids the problem of
reinventing the wheel.


ericmock's picture


Thanks for the great information.  In fact, I was just doing a little checking to see if any other organization was doing open reviewing.  This was prompted by a comment of a friend in physics who said his sister in geology had mentioned something about an open review system.

You bring up some good points and I was actually just working on a figure for the proposal that actually looks similar to the one you posted.  I realized that a diagram like this would really help people understand the flow of the system (as I currently see it).

I am personally very confident that the system will work, but as a recently tenured professor putting only my weight behind the effort is going to move it nowhere.  That is why I'm soliciting help and plan to integrate with iMechanica.  I am hesitant to associate it with any organization that publishes journals (even non-profit ones) because I would like it to be independent.


ericmock's picture

Here is link to my flow chart for submitting and reviewing:

Sorry for not embedding the link but TinyMCE (the editing software used here) does not seem to like my home computer (no matter what browser I use) and will not display all the functionality (although I can remotely log into my office machine and everything--links, bold, underline, images, etc.--is visible).  Really strange.

Andrew Norris's picture

but I still prefer the ACP flow diagram - which avoids the technical stuff.  The colors help too, as do the smileys for the people, though maybe the author should not have a smile all the time ...

btw - a new Front for the Arxiv was unveiled this week.  Its definitely an improvement, and it has a cute banner blurb:  All the research that's fit to e-print.


ericmock's picture

Boy, that's almost as nice as the "Comprehensive TeX Archive Network" front page ( Wink (emoticons are one of the few editing features working for me at home).

The ACP flow diagram is nice but I actually find it kind of confusing.  I too was thinking that some of those authors might not look so happy.  Rest assured, my diagram will be spiffied up for the proposal (graphic design is another hobby of mine).


Mogadalai Gururajan's picture

Here are some links to discussions on several aspects of open source, open access and open science initiatives (Disclosure: some of the links are to blogposts of mine, elsewhere): 

From all the discussions, it is very clear that open access is the one on which most people agree. When it comes to open source, there are issues about making the reviews open, making the reviews anonymous, compensating referees and so-forth; however, some tentative steps are being taken already. In that sense, this proposal is forward looking. Finally, though open notebook science is the best way of doing science in principle, in practice, lots of scientists might be very reluctant to follow that path.

One last point before I end this comment: the only motivation for any author to submit his manuscript to open access journals is that it may get cited more. However, for an author, is there a motivation to submit a manuscript to a open source journal? As of now, there does not seem to be any--unless, of course, the open source journals are shown to be faster, more respected, and generate enough interest discussions in the community. In that respect, I think, running an open source experiment at iMechanica as a pilot project might be a wonderful idea; if there be enough interest (which I think, there will be--seeing all the discussions that happens in the jCulb and on some of the preprints of papers), an open source journal may be born out of the project; along the way, the community might also be able to identify suitable editors for the job.

Zhigang Suo's picture


I read the current version of your proposal, and tried to make small changes in Summary.

A difficulty you might face is this:  People are busy and writing a good paper is hard.  Why should an average author give up a desirable label like JMPS or PRL?  The answer is that he needs not to give up anything.

I particularly like the distinction you made between open access and open source.  The former concerns the end product, allowing a reader to access a paper free of charge.  The latter concerns the dynamic process of writing, reviewing and editing a paper, enabling fast, many-to-many communication.  To me, open access publishing is a Web 1.0 issue and requires no technical innovation.  By contrast, open source publishing is a Web 2.0 issue, and does require considerable technical innovation. 

You stated that the proposal is about open source publishing.  I totally agree with this focus.  Large chunck of the text of the proposal, however, talks about open access.  I believe making your focus sharper will considerably strenthen your proposal, and lower the barrier of entry for most authors.  Let me elaborate.

Here are several elements of schloarly publishing relevant to this discussion:

  • Create the paper.  Writing, reviewing, and editing
  • Trustworthy timestamp.  Who said what and when.
  • Permanent access.  Open access or paid access
  • Add labels.  e.g., Add the name of a journal to a paper.

A traditional journal does all of the above.  However, these functions can be done independently, each by a separate module.  For example, arXive has shown clearly that a paper can receive a trustworthy timestamp and be permanently accessible when the author uploads the paper to arXive, and then subsequently receives a label from a journal (i.e., be accepted by a journal).

The really innovative part of your proposal is about creating the paper by an open source process.  What if you focus on this part, so that the author can create a paper using your open-source publishing tool, and then she can decide to leave the paper in your system, or submit it to any journal she chooses, open access or otherwise.  I understand that your system will allow this option.  However, by stating clearly your intention, and removing much of the discussions on open access, you will sharpen your message.  Also, an author will not try to answer the hard but unimportant question:  should she abandon a traditional journal that both she and her department head know about.  Thus, you remove a significant barrier of entry to most authors. 

Viewed this way, your proposal is a natural evolution of the existing innovations:

  • arXive allows uploading e-reprints
  • iMechanica allows uploading e-prints and user comments
  • Your open source publishing allows the whole creative process:  writing, commenting and editing.

None of the above need to be tied to any specific journal.  You may choose not to assemble a board of editors, but enable this function if a group of people wish to do it.  A more useful function that you can create would be a one-click button.  Once an author is happy with her paper, pushing that button, the paper is submitted to the journal of her choice, either JMPS, or PRL, or JIT. 

By removing the commitment to establishing a journal within the proposal, you might have also answered another hard question:  how can you sustain the effort in the long run.  What you propose is to create a tool to collaborately write a paper.  You can let whoeevr interested to host journals.

A seperate suggestion for you and Teng:  Teng can add a paragraph on iMechanica.  It is a very impressive story, much more so than that of Applied Mechnaics News.  The reviewers were impressed by Applied Mechanics News.  See their comments and the paragraph in the previous proposal.  Also, you may even cite this thread of discussion as an example how discussions might work. 


If Eric were to offer me a tool for writing papers (rather than an open access/openly published journal) I would have to decline politely.  The tools that I use are adequate for my purposes and I can use them much more efficiently than a new product if the end goal is to publish in a Web 0.0 journal. 

Given that, I can see the use for such a tool for collaborative proposal writing, provided that privacy can be protected tightly prior to acceptance.  After the proposal has been accepted it can be made open to the public.   I wrote my first proposal without any idea how one should be written and had to search the internet for examples (a good set of proposals with a biological bent can be found here ).  A repository of proposals will be a boon to new researchers.

Regarding some of your points:

  1. You talk about the brand equity of journals such as JMPS or PRL.   It has taken these journals quite a few years to earn that equity.  One of the surest ways of earning brand equity is to be extremely selective.  JMPS was founded by Rodney Hill in 1952.  How did he build up its brand?  I'm sure Eric could try something similar.
  2. You prefer a tool  (or a set of  modules  that  anyone can use by mixing and matching)  rather than a journal/set of journals.  Since Eric is the one who is proposing the open journal idea, let me write down my take on (some of the main reasons) why he wants open publishing.
    1. Eric does not like the current review process.
    • If we revert back to standard journals after using Eric's tools then the review process does not change.  Eric remains dissatisfied.
  3. Eric thinks that too many papers are published and that his idea would help weed out the average ones from archives.
    • I don't see how this can be addressed by open publishing.  Probably a whole different discussion.
  4. Eric wants to make a paper a dynamic object that can be modified post publication.
    • Publishing the paper in a standard journal defeats this purpose.  Do we have to redefine what archival means? 
    ericmock's picture

    Thanks for thinking so much about this.  Many of the points you bring up have to do with what NSF wants.  They are basically looking for tools that can promote virtual organizations using cyberinfrastructure.  If I say that we are just going to develop something that fits the need of the applied mechanics community, I am sure it will not be well reviewed.  Thus, I want to propose a generic solution for any virtual organization.  Something people can use for proposal writing, technical paper writing, creative writing, etc.  Broad impact is very important.  And I think what I have in mind could have a very broad impact.  Focusing on starting a journal gives it a very narrow outlook.  Plus, whenever I started a conversation with someone by saying I want to start a new journal...well, let me just say I had to start over.  For many different reasons, people don't want more journals.  Neither do I.  That said, the solicitation does require that we propose a prototype system.  What I am planning is to start with a fairly closed system in which a few users use the system to write a few papers and we get major bugs worked out.  Once we have a feature-complete system, we will use the iMechanica community to beta test the system.  We can use the system in whatever way we want.  We could post editable notes, pre-prints, half-written papers that have been sitting around for years, whatever.  At the same time, we will get a much better sense of whether the community would like to start a new journal.  If so, we would need to get an editorial board together.
    I have got the PSU English Department on board to provide internships (required for those with a publishing emphasis) to help organize the baseline system, write documentation, and copyedit content should things take off.  They're very excited about this (and so am I).  It gets undergraduates involved, engineering collaborating with liberal arts, etc.
    If you've read this far and would like to be added as a 'senior personel' to the proposal (sorry the budget is already depleted) please send me your SSN and I will add you in FastLane.  Alternatively, if you would to send a letter of support, I will upload that with the others I have. Eric

    N. Sukumar's picture

    I just read most of the earlier comments and suggestions in this thread and the link to the Summary posted by Zhigang.  I think overall the plan is fascinating (after all soo enough everything will be done off-the-web),  given that the ideas contained can lead to many new paths and directions. Here are some of my opinions, observations, and questions in no particular order. Please be aware that I just know the basics of html; am rather uninformed when it comes to Web 2 technology and what it can deliver.

    In reading the above discussions and the Summary, it wasn't very clear to me what to expect in the body of the proposal. I realize much of the above exchanges have revolved around the topics of open journals/reviews, etc., which is not the focus in Eric's proposal. That said, here are some first impressions on reading the Summary. In the Summary, in the beginning open source publishing/access is indicated, but then somehow later on the link is seemingly lost.  The first few lines could carry a lot more punch and get to what is exactly being proposed. Need to convey the excitement about open-source publishing (osp)---what is the present, and where you promise to take it, which is the novelty and breakthrough that is waiting to happen and will foreseeably change the manner in which publishing is done (okay, I sound rather theatrical). The ability to see beyond what others have done or proposed should demarcate your proposal.  It appears as though an extension (and integration) of the existing capabilities of Mediawiki and Drupal is proposed for the purpose of open-source publishing?  On the face of it, it might not appear to be revolutionary and/or exciting, even though I am sure that the details within would enable one to see otherwise.  What are the overarching objectives/goals you have in mind, how you propose to meet them (paradigm, software to be used, issues and challenges), and the deliverables and outcomes of the project.  Also, possibly some more text can be added in the `Intellectual Merit' and 'Broader Impacts' headings and make the body a bit leaner. I realize it is easier said than done.  Also, I am somewhat aware of what cyberinfrastructure (and the organizations that provide it) entails from the perspective of scientific computing, but how does it fit within this project (I see web tools and collaboration among individuals).

    Some opinions on topics covered before. I too do not wan't yet another journal at this early stage---osp should be a generic capability in its own right that can potentially lead to an open journal if such a need arises. I do, however, see that with time a lot of publishing will become online.  The initiative must come from within the academic community (publishers won't do it given that they're making huge profits), and the proposal would be a step in this direction. I just saw `Sicko' (Michael Moore)!  I believe anonymous reviews is a positive and should stay so. With anonymity, the critique is free-flowing (no holds barred); one must take the good and the bad. Also, I do not see a compelling reason and/or need to make the reviews open to everyone at this point in time (even if anonymity is maintained). Neither can a case be made that the quality of submissions and/or reviews will be better than that on average.  It takes time to see any pattern in this regard; moreover, it is a much too complicated matter that is intertwined to readily draw any inferences.

    Now, some questions.  What will I be able to do with osp that I can not do right now?  Have heard about software (companies) that have tools that enable collaborative teams to put together proposals.  Can one view osp as a web-version of Microsoft Word (wysiwyg) with many more capabilities?  Will LaTeX quality publishing be accessible to all via osp, with the freedom of dynamic updates etc.? Could this be an option to any team to write a collaborative paper as you've done?  How is the learning curve and can anyone readily make the transition? Are there other existing osp models that are in place and functioning? Is this the web answer to publishing what cvs/subversion is to (scientific) code development?  Hope my comments and questions have been in the ballpark?

    Pradeep Sharma's picture


    Your suggestions would work but they essentially point to an improvement over arXiv (which is nothing to sneeze at.....and quite welcome). However, back to Eric's orginal thoughts, if an open access journal is successful (and "somehow" becomes as prestegious as PRL or JMPS) then we have a paradigm shift. I don't see any real reason why, as a community, we cannot make this happen. Of course, regarding your point about PRL and JMPS, in the begining, the target open access journal should be dominated by researchers who have achieved their terminal promotions---overtime (once the reputation spreads) everyone else will follow. This requires coordination within our community. Six months ago I would not have thought this possible (case in point--what have we done about the shoddy treatment meted to us by ASME?); however, with the success of iMechanica in uniting our community and its power to reach out to so many of us, coordinated shifts to open access may be possible (difficult but doable).

    Zhigang Suo's picture

    Dear Pradeep and Biswajit:

    Thank you very much for your comments.  Like Eric, you and I appreciate the logic of open access publishing.  Our difference seems to lie in the approach to attain open access.  My hesitation to plunge into open access now stems from two considerations.

    1. The researchers are not ready to embrace open access.  The open access option has long been available to every researcher in several ways, but only a small fraction of researchers take this option.  Since 1991, arXive has allowed researchers in several disciplines to upload preprints and still publish the same papers in journals.  Some researchers do upload their preprints, but many don't.  Also, arXive has not diminished the popularity of PRL.  Everyone can also upload their preprints to iMechanica, or any other websites.  But not many people do.  This lack of interest among researchers may seem to be curious, given that most publishers explicitly permit researchers to post their preprints on any websites.
    2. Running open-access journals is costly.  PLoS and other journals charge authors over $1000 for making a paper open access.  Eric has also estimated the cost of running PLoS.  I totally agree with Eric that running an open access journal goes way beyond  the scope of this proposal.

    Am I pessimistic about open access?  Absolutely not.  Once researchers experience the advantage in what Eric calls open source publishing, more will choose to post preprints online and receive faster feedback from peers.  When that happens, papers are openly accessible through repositories like arXive and iMechanica.  Journals will serve the important function of adding labels, or waht Biswajit calls weeding, for targeted communities, but will have nothing to do with accessibility of individual papers.  There is no need to couple open access with adding lables.  I have tried to articulate this scenario in the post, What if all papers become openly accessible?

    We can do a lot to bring about open access.  However, starting an open access journal is not an effective approach.  The issue faced by a researcher is not the lack of a place to make her papers open access, but the lack of compelling and immediate benefits to herself.  Eric's open publishing tool will go a long way to contribute to solving this issue.

    ericmock's picture



    I disagree with your first contention.  I think researchers are ready to embrace open access and have been ready for a long time.  This is because there is really nothing to embrace.  So why haven't we?  Because we have not been given the option.  arXiv is a great thing but suffers greatly from a lack of peer review.  Thus, I do not consider arXiv to be an open access option.  I think many people do not upload to arXiv because they are not aware that it has gone beyond physics.  I think if you take a poll of faculty in ME departments, the vast majority will say that arXiv is for physics.  Even if you just take people in applied mechanics, I think you'll find nearly the same.

     The other reason I think people don't upload is that it's an extra step and some extra work of which they see no benefit.  Anyone that I can imagine that would want to read my papers probably has access through some institution, or they can simply ask for a copy.  I would rather they ask me for a copy than anonymously download my work so I know who's interested.

     If arXiv actually had a peer review system, I think it would explode in popularity and would diminish the popularity of PRL.

     It is curious that more people do not post their papers online, and I have to admit to not.  I'm not sure why I don't.  Partly I think I'm just cynical about my research; why would anyone care about it?  Also, the benefit just doesn't seem to be worth the effort (which I admit is miniscule).

     Also, I think that as researchers we are decoupled from the cost of access.  Take a walk over to your engineering library and talk to the head librarian.  This person will give you an earful about the cost.  In fact, I stopped by to chat with the PSU COE librarian the other day to talk about this proposal.  He had a list of journals on a sheet of paper on his desk trying to figure out which ones to cut.  Even Harvard (Wink) has cut back, 

     I think Vin Crespi (one of the PIs on the proposal) had a great statement, "I'd say the enormous scatter is due to the monopoly effect of copyright: once a given journal has a given paper, researchers will demand that their libraries provide access to that paper, so libraries are forced to pay the fees or have a gap in their collections. Imagine, for example, if each paper was involuntarily licensed to exactly TWO journals, and libraries could subscribe to a journal a la carte--article by article. That would immediately force competition on cost and normalize fees."

     The PLoS model is one solution but I think they have really done the open access movement a disservice with their business model.  They have targeted the extreme high end of publishing and their charges reflect this.  You certainly do not need to pay $3.5M/yr in salaries to run a journal.  You also don't need office in downtown SF, costing $250k/yr.  Nor do you need to spend $200k in travel for one year, or have $20k in bad debt expenses, or spend $290k on advertising.  Anyone can access their financial statements at They have basically built a very expensive infrastructure and now need to charge excessive fees ($2750 and climbing) to pay for it.  That's what happens when NIH's budget grows so big.  Thus, I think PLoS is a very poor comparison.

     I have nothing against for-profit publishers, they're doing their job (well).  But their interests are primarily to they stockholders, which conflicts with researchers trying to disseminate their work as inexpensively as possible.

    Zhigang Suo's picture

    Dear Eric:

    It seems to me that we agree more than we differ.   Here are some notes I jotted down as I was reading your comments.

    1. Perhaps the aparent difference in our opinions comes from what we mean by open access publishing.  To me, posting a paper in a repository is open access publishing if the repository has two basic ingredients, (a) trustworthy timestamp and (b) permanent open access.  I value peer review, but don't think that peer review ought to be included in the definition of open access publishing.  arXive has the two basic ingredients, and I regard it as a repository of open source publishing.
    2. I agree with you that many people don't do open access publishing for some reasons, including (a) they don't know that repositories are available, (b) they don't see much point in doing so with the existing repositories, and (c) they are unware that  most publishers explicitly permit researchers to post their preprints on any websites.  Understanding these and other reasons and removing them ought to be important things to do.
    3. I agree with you that PLoS is bloated.  However, with the current mode of operation, it just cannot be cheap.  Let's say a journal wants to pay two full-time employees, one doing web development and the other running the journal.  Let's say the two people will have a combined salary and benefit of 300k per year.  To pay for this salary, this journal will need 100 paying authors a year, each paying $3000 per paper. 
    4. Of course, the above numbers assume that the two employees are just publishing a single journal.  What if the technology has been well enough developed that they can run 1000 journals?  The two employees will of course be helped by volunteers from many fields of study, and several other full-time employees may even join them.  This scenario might be, in a nutshell, envisioned by many open-access startups.  They would like to create a platform to run a large number of journals. 
    5. Who already has such a platform but is not using it for open access publishing?  Any existing large publisher!  When the open-access publishing becomes the most profitable business available, the existing publishers might just switch, and still make handsome profits. 
    6. To me, their real competitor is not open access, but is more efficient technology companies that are not traditionally in the business of publishing.  Needless to say, these technology companies will also be competitors for open-access publishing entities, such as arXive and PLoS.
    7. On subscription on the per paper basis, I posted an entry called "Pay per paper" a while ago  It was first posted on Applied Mechanics News, where quite a few comments can be found, written by people with diverse background.  You might find them interesting.          
    Zhigang Suo's picture

    I forwarded the above comments of mine to Peter Suber, of Open Access News.  The following is his email reply, which I post with his permission.

    Dear Zhigang,

    You're right on in 1 and 2.  In #3 I wouldn't call PLoS
    "bloated" but I would say that it's a high-end operation and
    that other OA journals needn't do all that PLoS does and therefore
    needn't incur all of its expenses either.

    BTW, I'd say that $300k/year for two academics (salary and benefits) is
    very much on the high side!

    On the rest, there are already several OA-dedicated platforms much more
    efficient and inexpensive than traditional publishers.  One is
    Scholarly Exchange,

    All the best,


    Robin Selinger's picture

    I am a biased commentator in this debate because my husband, Jonathan Selinger, serves as an associate editor of Phys Rev E. He is also a professor and Ohio Eminent Scholar at Kent State University. Jonathan spends about 12 hours/month evaluating incoming manuscripts for PRE, assigning them to referees, and making editorial decisions.  Physical Review pays associate editors for their time so this is not volunteer work.

    Can an open-source publishing system replace the efforts of such hard-working editors and the journal's administrative staff? I honestly don't know. That kind of community effort seems to work beautifully for Wikipedia and Linux, but I can't predict whether it will work well for scientific publishing.

    To make an imperfect analogy... one could set up a high school as an educational collective, with each age group teaching the younger students and learning from the older ones. Parents could help out where needed as volunteers. With no paid teachers or administrators, tuition would be very low, just enough to pay for classroom space, books, computers, and supplies. In the best case such a school might work extremely well, with a self-selected team of bright students dedicated to the success of their community enterprise. But in the worst case, the school might come to resemble something between "Animal Farm" and "Animal House."

    I personally prefer to send my kids to a school with paid teachers and administrators, and likewise, I prefer to submit my papers to a journal that employs expert editors and staff.

    At this stage, I am in favor of a small-scale pilot effort to see how the open-source model will work for scientific publishing. The liquid crystal community, for instance, has an open-source publishing website at, managed by my colleagues at Kent State's Liquid Crystal Institute. E-LC accepts postings of manuscripts, dissertations, and presentations.

    Like ArXiv, E-LC functions as a preprint library, but it's much more than that. The editor of E-LC reserves the right to reject submissions. Secondly, there is a threaded discussion forum available for each submitted document. So far, the number of submissions is modest but it will likely grow with time.

    So, I have an open mind on open-access publishing and would like to see NSF fund some small-scale pilot efforts. At this stage I am not entirely optimistic that it will work as well as traditional journals, but am willing to give it a try.

    Regarding referee reports being "outed"....From the legal standpoint, I would treat a referee report the same way I'd treat any personal letter from a colleague. I do a fair amount of refereeing of manuscripts and proposals and would be very displeased if someone posted my report online in its entirety without my permission. On the other hand, the recipient is certainly welcome to include copies of my report in a tenure/promotion portfolio or other private communication. 

    Robin Selinger
    Professor, Chemical Physics Interdisciplinary Program
    Liquid Crystal Institute
    Kent State University
    Kent, OH 44242
    (330) 672-1582 


    Zhigang Suo's picture

    Dear Robin: 

    The last paragraph in your comment concerns a previous comment of mine regarding an experiment to post anonymous reviews online

    The situation is imperfect now.  On one hand, some of the reviews are perceptive, coming from experts on the subjects.  It is a terrible loss of their insight, and work, if these reviews are just viewed by very few people.  On the other hand, using somone's work without permission may cause discomfort to many people.  The level of discomfort may depend on the kind of use.  Here are examples:

    • Quote a review in a conversation at a dinner table.
    • Quote a review in a presentation in a conference.
    • Show a review in a slide during the presentation.
    • Include a review in the tenure package, which may be viewed by a fairly large number of people.
    • Forward a review to other people in an email, which may be viewed by a fairly large number of people.
    • Quote part of a review online.
    • Quote the entire review online.

    Each individual will consider a subset of the above uses fair, but no guideline seems to exist.  Who should set the guideline?  Should publishers set the guideline?  Should researchers experiment for a period of time and let a social norm emerge naturally?  Any other ways to resolve the issue?

    Note that the last three uses are possibilities provided by the Internet.  The Internet may give the freedom to the authors, but may also give publishers and other large organizations an easier way to monitor and control the uses, as explained by Lessig in his book, Code.  Incidentally, iMechanica uses a copyright license designed by Creative Commons founded by Lessig.  The license allows anyone to copy, distribute, transmit and remix the work.   

    In addition to letting excellent reviews out, allowing authors to post reviews may have another benefit:  it also exposes incompetent or thoughtless or unfair reviews, and careless decisions made by editors based on such reviews.  Of course, a large purpose of doing so is to encourage good reviews.

    These are pros for the experiment.  Any cons?

    Zhigang Suo's picture

    Dear Eric:

    • Congratulations on submitting the proposal!
    • This thread of discussion initiated by you is very helpful to many of us to think about the issues.  We may not have reached agreement, but have a better understanding various points of view.  For complex issues like these, perhaps an understding of diverse points of view is even more useful than reaching an agreement.
    • You have received quite a few supporting letters for this proposal.  Would you consider asking the permissions of the writers and post these letters online?
    • Please check the two links that you have just added to your post. The links do not work for me.
    ericmock's picture

    I got great support letters from:

    1. Karen Thole (PSU), Mechanical and Nuclear Engineering department head who has agreed to provide teaching release for the PI;

    2. William Burkhard (PSU), Director of Information Technology and Design who has agreed to support the needed computing infrastructure;

    3. Robin Schulze (PSU), English department head who enthusiastically endorses the English undergraduate internships that will be supported;

    4. Charles Steele (Stanford), the past editor of the International Journal of Solids and Structures and the current editor of the Journal of Mechanics of Materials and Structures who will discuss using a version of the OSPS with the editorial board;

    5. Thomas Conklin (PSU), the head of the Engineering Library who strongly supports the move to an economically sustainable model for academic publishing; and

    6. George Voyiadjis (LSU), president of the Society of Engineering Science (SES) Inc. who will work with the PIs to encourage SES member support and possibly start the first SES journal.

    Tom Conklin wrote a great letter containing:

    "As an academic technical librarian, I wholeheartedly support the investigation into and the development of  alternative approaches to scientific communications. The current system of publishing in journals is economically unsustainable. In my 32 years in the profession, I have seen the scholarly publications problem grow from concerns over a handful of overpriced journals into a crisis that threatens the foundations of library collections budgets. Commercial publishers have taken control of the academic publishing world which is comprised of a captive audience of editors and authors who need to publish and contribute professionally to achieve tenure and advancement. Through shrewd marketing, pricing, and strategic mergers, a relatively small group of STM (scientific/technical/medical) publishers now dominate the industry and are highly profitable. Academic library collection budgets are buckling under the annual cost increases of commercial STM journals and have been forced to move money from book budgets and other areas to help cover serials costs. The actual number of journal subscriptions at many large research universities has decreased over the last decade as a result of journal cancellations made in response to rising costs."


    All but Steele based their letter on a template I gave them below.


    Eduardo Misawa, Coordinator

    Program Manager, Directorate for Engineering

    Civil, Mechanical and Manufacturing Innovation, RM 545S

    The National Science Foundation

    4201 Wilson Boulevard

    Arlington, Virginia 22230


    Dear Dr. Misawa:

    I write this letter in strong support of “Open Source Publishing 2.0” by a team led by E. Mockensturm in response to the proposal solicitation for Engineering Virtual Organizations (EVO).

    As the **, I am enthusiastic about projects that promote collaboration and development of virtual organizations, and have a high likelihood of sustainability after initial funding. From my perspective, this project does both. The innovative faculty team has drawn upon their individual strengths to develop a proposal that builds on existing work and resources to establish a coherent and comprehensive system to allow research collaboration on any scale, from small teams to entire publishing communities. The project addresses important topics (virtual organizations and cyberinfrastructure) and will impact both students and established researchers in many different fields and programs. This is indeed one of the few proposals that presents a meaningful and sustainable connection between PSU's College of Engineering, College of Science, College of Liberal Arts, and College of Information Science and Technology. The project will pull together a host of resources and create a system where that will be a tool for anyone in engineering, science or liberal arts that can be used as an effective collaboration tool. The beauty of the proposed tool is that it is open for group to use as it sees fit.  Creative people can find unintended and unanticipated ways to use the resource to best server their needs.

    The proposed system will use newly developed and existing cyberinfrastructure and could truly revolutionize how scholarly publishing is done.  The system will allow for: 1) submission and dissemination of basic ideas, 2) rapid interactions between authors and reviews, 3) real-time online editing of content, 4) an open reviewing system in which reviewers can be recognized by the community, and 5) development of communication channels large (entire journals) and small (within research groups).

    Support from established groups such as the Society of Engineering Science and will allow the system to be tested and refined during the funding period so that it becomes a robust tool for all to use.

    The College of Engineering's Electronic and Computer Support group will support the server infrastructure during the funding period.  After which individual persons and organizations will use their own cyberinfrastructure to host the system.

    The PSU Department of English will support the effort in three ways through internships and two upper level classes (ENGL417: Editorial Process and ENGL418: Advanced Technical Writing and Editing).  Internships, required for a degree with a publishing emphasis, will allow senior English students to gain experience with technical editing and provide a sustainable and free pool of editors for the demonstration journal content.  Should a sustainable journal based at PSU result from the demonstration project, continued internships will be sponsored by the editorial board.  Internships will also be sponsored during the grant period to have students develop documentation for the system, and help organize the user interface to best present, browse, and search through the content.  The system, and it's development, will also be a resource for instructors and students taking ENGL 417 and 418.  To realize these efforts, the Department of English will provide co-PI Stuart Selber one course per year release time for the two-year grant period.

    The Society of Engineering Science (SES) which does not currently publish its own journal is interested in working with the PIs to test the system to see if it can form a basis for the societies first publishing effort.  Should a pilot journal be successful, SES could lead a revolution of scholarly publishing that other for- and non-profit publishers are sure to follow.

    To allow the PI to spend the enormous amount of time required to expand this cyberinfrastructure beyond the preliminary stages currently develop, the Department of Mechanical and Nuclear Engineering will reduce Mockensturm's teaching load by one course for the two-year grant period.

    In closing, I reiterate my enthusiastic support for this project because of the stated goals and the manner in which the team proposes to accomplish them.

    Mike Ciavarella's picture

     Looking at this very successful case (impact factor raised fast as never before for a journal, will it reach soon Nature?)


    And also many others.

    The Open Drug Delivery Journal

    The Open Nanoscience Journal

    The Open Chemical Engineering Journal

    The Open Biomedical Engineering Journal

    The Open Nanomedicine Journal

    The Open Colloid Science Journal

    but they are not enough.


    We Should Find a tool. maybe dynamic review papers.

    having no harmony is main issue of a little richer third world!







    Great expectations: The role of Open Access in improving countries’ recognition

    Hajar Sotudeh  and Abbas Horri

    (1)  Department of Library and Information Science, Faculty of Education & Psychology, Shiraz University, Shiraz, Iran
    (2)  Department of Library and Information Science, University of Tehran, Tehran, Iran


    And it looks like good news to me.

    I'm a student, about to start my PhD, and when i first learned about the way publishing works (the classical way) I have always found it very unethical.

    I'm will be doing government funded science, and I like the idea of free knowledge so it makes me a bit nervous how my first real publication will be done.

    Mike, I have read your link on peer review. It was a very interesting read.
    I shall try to dig through this thread in detail one day when I have more time.

    Subscribe to Comments for "Open Source Publishing"

    Recent comments

    More comments


    Subscribe to Syndicate