C3The Open Textbook Idea

The open textbook proposal is simple to state. A teaching text is written and released under a licence that permits anyone to copy it, distribute it, translate it and adapt it, usually on condition that attribution is kept and that adaptations carry the same terms. The immediate consequence is that the marginal cost of a copy falls to the cost of the paper or the download.

What the licence genuinely fixes

It fixes real problems, and they are not trivial. A book that may be freely copied cannot go out of print. It cannot be withdrawn from a course by a change in commercial strategy. It can be printed locally where connectivity is poor and screens are scarce. It can be translated by the people who need the translation without negotiating rights. And it can be corrected by anyone who spots an error, without waiting for an edition cycle.

The adaptation right matters more than the copying right in teaching contexts. A course rarely needs a whole book; it needs eleven chapters of one and two of another, reordered, with local examples substituted. Under conventional terms that is a permissions project. Under an open licence it is an afternoon's editing.

What it does not fix

Three things, mostly, and they account for most of the disappointment in this area.

The first is that writing a good textbook is slow, specialised work, and the licence says nothing about who does it or how that time is paid for. Open material has tended to appear first in large introductory subjects where the content is stable and the potential audience is enormous, and to remain thin in specialised or fast-moving areas where the effort per reader is far higher.

The second is maintenance. A textbook is not finished when it is written; it accumulates errors, dated references and pedagogical improvements. Commercial publishing funds maintenance through the edition cycle, which is also the mechanism that destroys the second-hand market. Open projects need some other durable arrangement, and projects without one tend to be excellent in their first year and progressively less reliable afterwards.

The third is the apparatus. Adoption decisions frequently turn not on the text but on what surrounds it: problem sets with worked solutions, assessment banks, slides, laboratory instructions, and the assurance that these exist and are consistent with the book. Producing that material is a larger job than producing the prose, and it is much less rewarding to volunteer for.

Adaptation and fragmentation

There is also a tension inside the idea itself. The freedom to adapt is the point, and widespread adaptation means many divergent versions of the same book. That is a benefit locally and a cost globally: a question about page 74 no longer has one answer, and improvements made in one branch do not automatically reach the others. Projects that manage this well tend to maintain a clearly identified canonical version with a visible process for merging corrections back into it, which is a governance solution rather than a licensing one.

How to read the claim

The honest summary is that the open licence removes an access barrier completely and permanently, and leaves the production, maintenance and support problems exactly where they were. That is still a substantial gain, and it is a smaller gain than it is often described as being. The projects that work are the ones that treated the licence as the starting condition and then solved the ordinary publishing problems anyway.