The new way of publishing opens up new possibilities for linguistics. For instance, theoretical work can be separated into a descriptive part and the analysis. The content of a book can be converted into a Wikipedia-like format and can be edited and improved by other researchers. It is thus possible to continuously extend our knowledge about a particular language. Books are snapshots of our knowledge at a given stage.
This new way of working allows researchers working in other frameworks to reuse the descriptive part and add their own analysis part to it. We would this get theories that cover similar data and can be compared. We thereby ensure that the field as such makes progress empirically and theoretically.
Digital publishing has certain advantages: The publications can be stored together with primary data and software. The projects of Enhanced Digital Publication show that there are entirely new possibilities. The Enhanced Digital Science project looked at various disciplines, also including linguistics. One project developed a dictionary for Berber that includes background information like pictures and articles from the press. Other useful combination of sources can easily be imagined for linguistics. For instance, one could connect a text to corpora and subcorpora that play a role in an analysis, one could include interactive example trees that can be manipulated and modified by the reader since they are together with a piece of code that allows the folding and unfolding of the tree (click on the nodes in the example). Usually fully worked out linguistic analyses are highly complex. It is the task of the author to simplify the analysis and highlight the most important aspects. However, in some situations the reader wants more or it is not obvious how several partial descriptions in a paper have to be fused into a coherent picture. Having a tree that contains all the details of an analysis but initially displays only the information that was marked as relevant by the author is extremely useful here.
In order to provide fully searchable documents that are annotated for content, we have to provide the documents in the XML format. This is also necessary for producing e-books. The conversion of LaTeX to XML is not trivial, but we hope to solve this problem for standard packages (glossed example sentences, trees, AVMs) in 2013–2014. The efforts will be coordinated at the FU Berlin.
We will start rather traditionally with OMP, that is, there will be series editors that determine at least two reviewers per book manuscript. Once a manuscript is accepted it is sent to the production stage with layout editing and proof reading. In the startup phase the German Grammar Group of the FU will take care of layout and proof reading. However, within the following two years we want to turn this into a community effort: OMP allows for the registration of reviewers and proofreaders and we hope to be able to get a number of volunteers that is high enough to guarantee short turnaround times. (Register as reviewer/proofreader and see who else volunteered to work as reviewer or Proof Reader). Other disciplines show us that this can work. The journal Forum: Qualitative Social Research has 17,000 registered readers and 900 reviewers. They use the Open Journal System, an open source workflow system that was also developed by PKP for journals rather than monographs.
Open Reviewing (optional) and Versioning (optional)
It is often the case that comments of reviewers significantly improve books. Some reviewers invest a lot of time in reviewing. We will follow a suggestion by Pullum (1984, Volume 2, p. 266) and publish the names of the reviewers with the book, if reviewers want this (for incentives see gamification). This approach is also taken by the Edition Open Access which is a book series run by the Max Planck Research Library for the History and Development of Knowledge. As Pullum pointed out in his column, standing in for a publication with one’s reputation ensures that reviewers take reviewing seriously. But with the new ways of publishing we can take Pullum’s ideas even further. We can build a web of trust. We can set up reviewing systems that keep the original submission of accepted manuscripts around and add the reviews. This can be done even if reviewers decide to stay anonymous. It is done successfully in other disciplines as for instance in Geoscience, Economy, and Biology (Pöschl and Koop, 2008) (see for instance the journal of Atmospheric Chemistry and Physics).
In addition to the two reviewers that are determined by the series editors the whole community is free to comment on manuscripts. This will further improve the book.
The advantages of such an open review process and versioning are:
- full transparency
- credit to the earliest published version of ideas
- higher quality of submitted manuscripts
- credit to the work of reviewers
Open reviewing will be combined with aspects of gamification. Websites that use gamification give users points and awards for their activities. An example is stackexchange.com. They have a list of rated questions and answers and you get credits for asking and answering questions (example). So you can see who is an experienced user of this system. In addition to credits for your activity there are certain badges that are awarded to users for reaching certain goals. For instance users get a badge for editing 500 posts. There are certain thresholds for user privileges that are assigned automatically by the system.
In our system points will be awarded for
- submitting manuscripts
- reviewing manuscripts (as a selected reviewer)
- reviewing manuscripts (as a community member)
- working in the advisory board
- working as a series editor
- proof reading
While selected reviewers have to comment on the whole manuscript, community reviewers may comment on parts. The community may vote reviews both by selected and by community reviewers up.
Bonus points will be given for:
- submissions of manuscripts to the open review process
- speed of reviewing
- transparency (anonymous publication of the review)
- transparency (non-anonymous publication of the review)
The gamification approach allows younger researchers to enter the game without being chosen as a reviewer by some series editor. PhD students often know their field very well, sometimes better than established researchers. An open reviewing system with community participation can activate additional expertise and add to the reputation of younger scientists.
Of course the reviewing work that resulted in the rejection of manuscripts has to be credited without revealing the identity of the reviewer. Anonymity is a problem in the beginning only. It can be guaranteed by introducing delays for the respective gratification.
In the case of manuscripts of bad quality, that is, manuscripts that would require a lot of work on the reviewer’s side, it could be the case that nobody is willing to review the manuscript. This can be either accepted by the author as a rejection or the author could increase the motivation for reviewing by setting a `bounty‘. Bounties can be set on systems like stackexchange to increase the priority of a question. Those who answer the question will get a bigger number of credit points and the person who asked the question has to `pay’ with some of his or her credit points. If the manuscript has some good ideas in it, reviewers eventually will be willing to invest a lot of time in a manuscript. The extreme version of the `bounty’ idea is of course co-authorship.