Should academic journals in the humanities meet performance standards, as they expect their authors to? The answer is a self-evident "yes", yet very few such journals do so. By contrast, leading journals in bioscience have clear objectives for quality and speed of reviewing, for the nature of the editorial decision process, and for the overall delay in publication. There has recently been extensive discussion in the multidisciplinary science journal Nature about the peer-review system itself, but my concern is with editorial responsibilities.
In science, journals compete on the basis of the citation rates they achieve. Bioscience journals aim for completed reviewing within three months of receipt of a submission, and publication in print of accepted articles within nine months of submission. How many humanities journals even target, let alone achieve, this? The massive delays that many humanities journals permit irritate authors and diminish the utility of the papers they publish. Indeed, in spheres such as socioeconomic policy, authors often release "work in progress" as pamphlets and web documents during such delays, which may further undermine the utility of the eventual journal article.
In the interests of respectable and efficient progress, journals should fulfil at least the following principles:
* Every article should receive at least two independent peer reviews
* Peer reviews should be made available in full to the author, but the editor in charge of the article should synthesise an overall view, especially when reviewers disagree. Reviews and editorial comments should be delivered within three months of submission
* The author must have the chance to rebut any alleged misconceptions of reviewers or editor, and must be able to revise his/her article before acceptance or final rejection. Sometimes the editor should specify the reasons for rejection in their initial comments. But a putatively irreversible "no thanks" accompanying superficial review comments - or even no comments - wastes everyone's time. As in the application of the legal process, the author should be presumed "innocent" (or authoritative); and the reviewers should be presumed "guilty" (because of anonymity). An editor should therefore evaluate reviewers' comments stringently. Correspondingly, an author should dispute comments that are misplaced but accept constructive criticism and the additional work it entails, while recognising that some submissions are not suitable
* A time limit for authors' responses should be defined
* The editor should take responsibility for the final rapid decision after receiving the authors' response. It should not normally be necessary for the article to return to the reviewers: the editor should be capable of appropriate adjudication. Otherwise, one should question his or her suitability as editor
* Accepted articles should appear in print within six months. Web-only journals should publish within one month of acceptance and multi-format journals should achieve more rapid publication than print-only: then they can compete successfully for citations
* Journals should publish their impact factors and the comparative impact factor range within their own field (fields vary vastly). I do not assume that impact factors are crucial, but simply that they provide information on the breadth of distribution and readership of the journal.
The academic community can only benefit from such open competition for quality of editorial process in journals; and some journals should be able to exceed these performance minima. As a result, the efforts of authors would be more fully utilised. It is high time that significant journals addressed these issues and showed the leadership qualities they fondly imagine they already possess.
Roger T. Dean
Vice-chancellor
University of Canberra,
Australia, and a scientist, musicologist and musician