This is great news for people like us. The down side is that publishers may not publish the books, since open access will definitely going to reduce their revenues. It will be interesting to see if MIT can do something like announcing about the literature and hosting it. Which was being done previously by publishers.
I don't think that publishing research as open access will affect book publishing. The reason for the existence of commercial journals was the printing costs and distribution. With the internet this has basically disappeared.
Books, esp. undergraduate college texts will remain extremely profitable for the foreseeable future.
What Abelson doesn't mention is making the underlying data and runnable source code available with each paper -- a game-changer for independent verification.
Open papers has to be won, as a first political battle, but I don't see open data on anyone's platform.
I'm the person responsible for administering the server that powers this open archive. The type of data you mention is being included in the archive when ever possible. Further, the research group I work for has existing specific open data efforts in multiple fields including, and beyond, computer science. For example, today I'm working on setting up an architectural open access archive with the staff here that will include CAD files for humans and semantic web triple store data for search engines.
I couldn't find any data/code associated with non-CS papers, across a quick sampling of US DSpace systems -- although most of the submissions seem to be from the back catalog.
No offense to DSpace, but adequate systems for publishing open data have been available for twenty years, i.e. an ftp site. It will require a top-of-line mandate from institutions or journals to make it happen.
MIT can do this, but will others follow suit? I imagine publishers can put a lot more pressure on the "long tail" of small univerisities and researchers to keep their content closed than MIT.
I imagine publishers can put a lot more pressure on the "long tail" of small univerisities and researchers...
The publishers have little pressure to apply. The only reason for-profit journals still exist is that (a) they have a big existing audience (everyone reads Nature) and (b) they supply a seal of quality. (If you get published in Nature, people know you're a Big League Scientist and you get tenure faster and/or get more grants.)
Point (a) is extremely fragile in the age of the Web. A journal which charges exorbitant subscription fees is one tiny tipping point away from losing its entire audience to a website run from a $9.95 Dreamhost account.
Point (b) will be the last stand. And this move from MIT threatens to undermine it, because MIT manuscripts carry prestige, and any venue where they appear gets prestige by association, and those manuscripts will no longer appear in any exclusively-closed journals.
Meanwhile, technical book publishers have no pressure to apply at all. None of those books make any significant money for their authors. [1] Far from applying pressure, the publishers of scientific books must beg and plead for contributors to commit to writing chapters, and then those chapters routinely run years late, because book writing is the lowest priority on a professor's calendar. From what I've seen, many of them end up being ghostwritten [2] by postdocs in their spare time, and as a rule postdocs don't have a lot of spare time. It's a miracle that we have any decent graduate textbooks at all. A lot of fields don't, in fact.
---
[1] Except for freshman-level college texts, which are a weird little monopolistic world of their own. God, I hope MIT's rule applies to those as well. That would be awesome.
[2] Okay, that's an exaggeration. Grad students and postdocs get credit for co-writing books and book chapters. That's the least they deserve for doing the majority of the work. ;)
