I guess a possible solution to the problems you pointed out is to support and promote the so-called *diamond open access* journals, in which the papers are available FREE of charge online but *also* there are NO author fees, see e.g. http://www.jasonmkelly.com/2013/01/27/green-gold-and-diamond-a-short-primer-on-open-access/ for details. It is also a good idea for OA journals to have the articles deposited in some reliable internet archive (e.g. arXiv.org in the hard sciences) so that they will not disappear if the journal shuts down. Here is an example of a high-quality open access math journal which meets all these requirements (free for BOTH authors and readers, and the articles are cross-posted on arXiv.org) and has an impact-factor of 1.243 which is quite good for a pure math journal: http://sigma-journal.com/

"It is unclear to me what happens if these journals fail and who would maintain the files so that the papers would not simply disappear unlike journals published on paper."

Archiving is important, and deserves serious consideration. The ability to save electronic copies of papers means that it may be easier to archive them. Today, a dead paper journal probably is probably less available than a dead electronic journal. If a journal printed on paper fails, how do you how many copies of the article were printed, and in what libraries they are held? What happens if a copy in a library is destroyed by accident?

"Where do the charges the authors pay go?"

Editors I've talked to usually estimate the costs of publishing a scientific article in a journal are about $3,000 per article. The editor-in-chief of Nature estimated costs of $30,000-$40,000 per article (http://www.nature.com/news/open-access-the-true-cost-of-science-publishing-1.12676).

"It is also unclear to me if the costs go to profit or to insuring the continuing access of the internet publications."

Both, presumably. What fraction goes to what will depend on the publisher.

"Are scientists with restricted funding blocked from contributing these publications by cost or are these open access internet publications helping?"

I do not have major grants. This has not prevented me from publishing in open access journals.

The cost of publishing open access is frequently misunderstood. Many open access journals, such as PLOS ONE, provide fee waivers on the honor system. Some journals have extremely modest fees. It is also difficult to see the modest one time lifetime fee to publish in PeerJ as a significant obstacle. Some open access journals do not have charges.

"If open access has the advantages of rapid publication and increased access why do these not show up in improved citations(?)"

Several studies have shown an open access advantage. For example:

http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.0040157

http://crl.acrl.org/content/65/5/372.short

http://eprints.soton.ac.uk/268516/

There are also issues such as publishing inertia. Established journals currently enjoy an advantage because establishing reputation takes time.

Also:

"Where do the charges the authors pay go? It is also unclear to me if the costs go to profit or to insuring the continuing access of the internet publications."

Let's ask the same question about traditional journals. We only need to rephrase it slightly:

"Where do the SUBSCRIPTION FEES the LIBRARIES pay go? It is also unclear to me if the costs go to profit or to insuring the continuing access of the internet publications?"

In this case, subscription fees, not to mention single article payment fees, are generating very large amounts of profit indeed for traditional publishers. Mike Taylor has some numbers here: http://svpow.com/2012/01/13/the-obscene-profits-of-commercial-scholarly-publishers/

In fairness, some traditional publishers are non-profit (Science is published by AAAS, a non-profit, for example).

Open access publishers also run the gamut from non-profit (PLOS) to extremely profitable (e.g., Hindawi; http://scholarlyoa.com/2013/04/04/hindawis-profits-are-larger-than-elseviers/)

Hi John, I notice that the article you cite is written by a member of Thomson Reuters staff. Not to question their integrity, their affiliation is quite clear and so is the origin of their research. However, currently they have a vested interest in a pay for view model of publishing and in being the only arbiter of quality through the IF. It seems sometimes that ResearchGate is either discussing the iniquities of IF or the unjust nature of article publishing charges!

I think the points missed by the article are that: 1) downloads are also a measure of impact and there is a whole industry devoted to alt metrics, other ways of assessing the impact of an article. 2) There is greater equity in the free at the point of use. Publishing is not free estimates vary but I have seen 2700 GBP mentioned in a report on the cost of publishing an article. 3) Research funders require Open Access publication as a condition of funding now and more so in the future so the "either or argument" is flawed from that point of view.

To answer your specific points. Where does the money go. You would hope on administration, technology and long term archiving. Some of course are just criminal enterprises, so I guess it just goes on fast cars. Some are clubs where the money is used to fund journals that club members can publish in without cost or without any critical peer review. There is a lot of what I call grey goo in OA, which tends to obscure the better journals.

My view is that researchers need better education about publishing and Open Access. Academics need to be more robust in challenging the integrity of OA journals. There are low cost options for publishing in good OA journals if you look. Most credible OA journals offer a fee waiver for some authors. BW Matt

Lots to think about in these answers. Regarding "If a journal printed on paper fails, how do you how many copies of the article were printed, and in what libraries they are held? What happens if a copy in a library is destroyed by accident?" so far I have always been able to get an electronic copy of any printed article via ILLiad regardless of whether or not it is in any local library. However I have heard of open access publications where the files become unavailable. Since some have pointed out that most journal articles are never cited and many are never read by anyone but the authors and editors - I'm not sure you will find an open access paper that you have not downloaded if the journal ceases to exist.

"Here is an example of a high-quality open access math journal"

SIGMA sounds too good to be true, so I have checked its website:-

"Free for authors

Despite being open access, we do not charge our authors for publication (i.e., SIGMA uses the so-called diamond open access model). Many authors submitting good papers cannot afford to pay page charges for various reasons and we do not want to put them at a disadvantage compared to those who can pay. On the other hand, when a journal charges their authors for publication, the requirements of thorough review may contradict financial interests of the journal. This often results in the publication of low-quality papers without adequate peer review and discredits the very idea of open access.

What are the sources for SIGMA existence?

The journal has no revenue and steady sources of funding whatsoever and operates solely on the basis of the good will efforts of its editors. All editorial work, including but not limited to the copyediting, is done by volunteers free of charge. SIGMA is not supported financially by any entity (organization, foundation, fund, etc.) in Ukraine or abroad, including the National Academy of Sciences of Ukraine. SIGMA is a non-profit, volunteer-run project. We greatly appreciate the assistance of EMIS who kindly provides mirrors for SIGMA.

Looking for your support/sponsorship

We need money for the web hosting, the CrossRef membership and the support of volunteers doing technical work for SIGMA. So, if you (or your organization) would like to support us financially, please contact us"

It does seem to be true, but this is surely not a viable long-term option or applicable generally. Hence the appeal for sponsorship, and for money for volunteers!

This is useful:

http://ijaridea.com/Directory-Of-Research