Textbooks are fundamental to modern education. But until recently almost all textbooks have been published for profit by huge corporations, keeping prices artifically high, limiting access, and subjecting content to heirarchical control. That may be changing. A number of efforts are in the works to create Open Source Textbooks: textbooks on a wide variety of subjects, collaborative built, translated into an assortment of languages and freely available to all.
As MIT's OpenCourseWare has shown, there is tremendous leverage in providing people everywhere access to at least the basics of really high-quality education. Now the idea is spreading, fueling an Open Education movement, and Open Source Textbooks are a key tool.
There's only one problem with Open Source textbooks - and that's the variance in educational curriculums throughout the world, which may cause the textbooks to have to 'fork'.
This is easily done, but writers and publishers who are not programmers are unfamiliar with the concept, and I think it will be necessary for them to become familiar with it.
Also see this Imaginify post on the subject: http://www.imaginify.org/modules.php?op=modload&name=News&file=article&sid=13
This raises the question of what constitutes "Source". The examples given are *collaborative* and they are *editable* or *configurable*, but that doesn't make them necessarily Open Source.
The source for mathematical knowledge is proofs.
The source for scientific knowledge is experimentation (including thought-experiments).
The source for historical and literary knowledge is the original texts.
It seems to me that truly Open Source textbooks would be ones that show, for absolutely everything in them, where the author got that information. In science texts, that should be the accounts of the original and substantiating experiements, in sufficient detail to be reproduced. In history text books, that should mean facing-page reproductions where appropriate.
True Open Source would be a breath of fresh air in the textbook field, where erroneous information is often published through oversight or political wrangling. If "many eyes make bugs shallow" then the evidence on which textbook authors rest their "facts" should be made available -- as they aren't currently -- to absolutely anyone who cares to examine that evidence.
It would also lead to the a form of information degredation that these institution were set up to prevent. The disimination of false and culturally skewed information(especially history) is already a problem. Open text source books would need some sort of oversight to ensure that each new crazy theory isn't taught as the "truth". I also agree that the current publishers of text books especially some of the college presses are questionable and at the very least exorbantly over priced. And the whole idea that text's books are the truth is kind of amusing. I know that this kinda of info with the exception of the sciences is skewed. like the new nickel celebrating the lousiana puchase. wtf?!?!? By by native people your nation means nothing...
I think that these projects might be an extremely good thing if they are in the right form. A textbook which regurgitates what others have said is not necessarily the best possible book to learn from. It would be much more interesting if, as _Goliard_ mentioned above, textbooks reproduce the original sources in the text. I suggest going one step further - only produce those texts with a timeline / historical description, and let the students draw their own conclusions. Later on in the book, the main theories can be discussed. This will lead to new ideas, insights, and theories which were not possible where the teachers only "stuck to the textbook". These OpenTextBooks, therefore, are not only an over-priced repository of knowledge, but a genuinely useful tool for study. W00t.