This year I participated in organization of a computer science conference. To my surprise, I found that conferences hire publishers instead of publishers own the conferences. Here is how organization of a usual CS conference works.
A conference is defined by its steering committee, which is a small group of top academics from the field of the conference. The steering committee appoints program chairs every year. The program chairs of a year sign a contract with a publisher to publish the conference proceedings. These contracts are signed every year. The program chairs organize the conference independently of the publishers. They appoint the program committee with the approval of steering committee. The program committee reviews the submitted papers and accepts a subset of the papers. The program chairs send these accepted papers to the publisher for final printing. The copyright of the papers are held by the publisher in return of providing the publishing service for free. They also give some free copies of published books to the program chairs. The copyright allows them to charge exuberant amount of money to the academic institutions to access the publications.
The natural question comes to mind: “Why publishers exist?”.
One may argue the following three contributions of the publishers.
- Quality Publishing: The publishers proof read the papers. Sometimes they move around figures to pretty-up the paper. This task is usually done by some third party company located in some third-world country. They assign ISBN number etc to give the proceedings a unique identifier. And finally, they publish the physical books. Publishers do not make any editorial decisions.
- Reliable Dissemination: These proceedings are usually available at the publishers website for rest of the eternity. For Academic libraries, it is easy to follow these publishers to find most of the scientific literature. Publishers make dissemination of scientific works easy.
- Legacy Reputation: If publisher signs publishing contracts only with high quality conferences then they build a reputation over time. There are no lack of conferences in academia. Academics keep inventing conferences to promote themselves. But if the proceedings of a conference is published by a publisher then the conference inherits the reputation of the publisher.
The first two contribution are already obsolete. The proof-reading service can be provided without the publisher. Usually, the cost of doing research is very high and hiring a proof-reader should only be marginal. An institution can easily pay for proof-reading of its papers by those third world companies. The printing of physical books and hosting the papers on internet are fast becoming inexpensive technology. The cost of running a scientific publication website can easily be covered by scientific funding agencies. Anyway, they are paying for producing the content. Why not they also pay for dissemination?
The reputation of a publisher is the only reason, I think, that keeps them in business. While evaluating a published work, funding agencies factor in reputation of the publisher. This makes life of evaluators easy. Therefore, the steering committees of the conferences do not want to move away from reputed publishers. This is hardly a compelling reason to have publishers. This is a classic legacy systems problem.
The underlying technology for publishing scientific works has shifted sufficiently that makes “the publishers” irrelevant and unnecessary overhead. All the scientists I have spoken to also say that they do not see any value of having publishers. However, it is very hard to chart a path from current system to a publisher-free system, without breaking scientific reputation system. Here is a great opportunity for innovations. I am very sure soon scientists will start experimenting with different publishing models for important conferences.