Financial Cryptography

I think that academic journals serve a number of purposes. They include distribution of information, but they also provide a distributed reputation system. A good journal gets good papers, in theory, and academics use that to help pre-judge a paper, and an applicant for tenure. Academic presses also help with the archival problem. Neither FC or Emergent Chaos are 'archival' web sites. I have no plans to keep the site going 100 years from now, even if I'm still around. A system for preserving such things, other than archive.org is important for science.

There are of course, the PLOS.org people, as well as a fascinating proposal from Zetland on a market for journal articles.

http://www.emergentchaos.com/archives/000633.html
http://www.emergentchaos.com/archives/000658.html

As I was thinking of this, it was exactly Ackerlof's quote that came to mind. I'm glad you've linked it above.

Which leaves the question of reputation. But this is already covered by linking and by citing; Google is the "proof" that the net provides a much more powerful reputational mechanism than the peer-reviewed circus offered by journals. Yes, the peer-review status provides a mostly positive benefit to those who do succeed in getting published, but I don't think it helps _enough_ to save the journal. There are other things that are now better than peer-reviewed, especially when you weigh in the fact that for every Ackerlof that got through, there are probably 10 sour lemons out there.

Adam Shostack writes: "I have no plans to keep the site
going 100 years from now, even if I'm still around."

Put your content under a license that allows (at least)
verbatim redistribution and even if it is just an ordinary
blog archive, I can almost guaranteed that _some party_
will keep it available on the public Internet indefinitely,
provided of course that civilization continues to progress.

I agree with Richard ...

The question of tenure I think is the key one. How is the academic world going to cope when the real reputation vectors shift across to net-based citation and analysis tools?

I suspect there will be a sort of intermediate stage where "non-traditional" metrics like publication in net journals like JIBC and First Monday, and ratings in net analysis tools will also be part of tenure discovery. And at some point someone will notice that there is a shortage of paper journals to base the metrics on, and they will be dropped from real decisions.

We won't notice the shift, until some bright journalist writes up the obvious in the Times. "Yes, we've been basing our tenure decisions on net metrics for years now, of course."

Alternatively, tenure itself will disappear and it won't be an issue any more ;-)

Adam Shostack writes: "I have no plans to keep the site
going 100 years from now, even if I'm still around."

Under the appropriate licence someone could not only cite you but copy the entire document and package it with their paper. This would give a stand alone document.

Ever tried to read a scientific document where the majority of the references are in closed access journals (CAJs)? The current scientific publishing racket would make the founders of the Royal Society livid.

The obvious solution is for the journals to switch to online publication. They would retain their editorial staff and the present peer review system without change. This is where they get their reputation for quality and the associated prestige. There is no need for upstarts like First Monday and other online publication experiments. All that is necessary is for Springer-Verlag and the other major academic publishing houses to put their material online.

In fact, most publishers already do this, but the content is only available by subscription. This will change, I predict, within the next few years, and the material will be available to everyone. Eventuallly they will phase out the paper publication, or perhaps outsource it to some kind of print-on-demand house for those who demand paper.

The current status of publishing chemistry journals is morally untenable. Tax payers money is used to fund research; the research is published with the resulting loss of copyright to the publishing houses. Do the academics have the right to give away the copyright? Yes, it was their creativity that partly enabled the work but it wasn't their financial capital. (Remember Lorenzo's oil. I wonder how much it would've cost his parents to find all the requisite information that they had made possible using their tax dollars).

The situation is perpetuated by the subsidisation of the users of the research. The scientists who work for BigPharma plc stick the access bill to the company (they pass it on to the sick and needy, the ones who need to buy their drugs or die); the scientists who work for academia stick the access bill to the taxpayer or students.

The situation is precarious at best. It reminds me of something that Nicholas Nassim Talib pointed out in his book 'Fooled by Randomness'; namely, that in an evolutionary system some species can survive for tens of thousands of years on the edge. He says that some species live a hairs breadth away from extinction - a slight change of conditions then obliteration. I think that this is the case with scientific publishing at the moment.

> But there are still lots of holdouts. Academic conferences still walk
> the party line and want sign-over of copyrights. Their logic runs
> thusly; in order to attract academics they have to show a formal
> academic proceedings which distributes their work. The proceedings
> publisher requires to own the copyright so as to benefit from reprints
> and subsidise the costly distribution of the books.

In the case of the Financial Cryptography conference (I don't know whether this is typical), the publisher seemed more concerned about having to track down many individual copyright holders (some of whom may have moved on or died) in order to reissue a work, thus we were able to strike a compromise whereby the authors assign copyright to the conference organization rather than to the publisher. As a result the publisher has to deal with only one party, and the authors get a copyright agreement that allows them to distribute their work pretty much as they want.

> But this was a confusion of means with ends. The end required is the
> distribution of the work, not the attraction of a well reputed journal.
> That's only the preferred means in a pre-WWW world. Now, as I say,
> this should have been blindingly obvious, but it wasn't and is perhaps
> the best evidence of scientists not understanding basic economics and
> marketing processes.

Some might argue that the end desired is not as much the distribution as the filtering/winnowing. That could of course be achieved by well-reputed online journals, and I suspect that peer-reviewed scientific publication is already moving inexorably in that direction.

> We now have some scientific evidence that using a
> 'reputable scientific journal' will reduce the distribution of the
> work. Of course, the paper is available online.

And in Nature, Volume 411, Number 6837, p. 521, 2001.