The challenge of peer review


I have just finished my first attempt at applying the ASO Guidelines for Community Review. The Target Article (perhaps not the best phrase, but the accepted one) was a contribution which I had invited Michael Protzel to submit.  As it turned out, the review process was far from straightforward.

An immediate issue that arises in attempting to apply an orthodox model of peer review is the diversity of interpretations of key concepts such as ‘primary control’ within the intellectual world of the AT. By comparison – and accepting this is a simplistic picture – in a more developed discipline there are significant areas of agreement around facts and methods and theory: ‘quality control’ - which is the function of peer review - can be exercised on the basis of such common ground. Any departures from the common ground would need to be given focused treatment in any material presented for review. (In reality, the elaboration of any such departures would most likely be the raison d'etre of the material to begin with.)

But a lack of common ground within the AT world then poses major problems. I discovered, as I tried to carry out the review, that I needed to reflect constantly on whether I had moved beyond the legitimate scope of ‘peer review quality control’ into something more in the nature of a substantive critique.  This experience reflects the view expressed forcefully by Ron Dennis in his comments on Glenna Batson’s Graviception article about where the profession has got to in its intellectual journey.

For me as an editor, this situation also invites the question of how much leadership to exercise in the attempt to promote consensus (in addition to highlighting and, where appropriate, encouraging diverse viewpoints). The danger, in the absence of a consensus about key issues, is that any debate is capable of exploding in any direction at any time. To take the case of the Graviception article:  Glenna Batson writes an informative piece on aspects of how we detect verticality, which she points out in her subsequent comment was motivated mainly by practical issues around the use of light touch in Alexander teaching; Ron Dennis responds with a comment that includes a statement about primary control [link above]; Michael Protzel challenges Ron’s definition of primary control: and suddenly, we are no longer in Kansas.

Taking a leadership role, in this context, would mean taking a position on what kinds of view could legitimately be expressed without requiring further explanation or argument: what, in other words, might be considered to be part of the ‘common ground’. Heterodox views lacking sufficient explication might then be suppressed on editorial grounds as leading to confusion, without making any substantive contribution. This might seem to conflict with a certain measure of impartiality associated with the role of editor.  On the other hand, there is such a thing, after all, as an ‘editorial line’ and editors do have to exercise judgment around what they publish.

A further tricky aspect from the editorial point of view concerns how an editorial line might be drawn according to a particular context.  An obvious candidate for intervention would be where a substantive contribution was presented for pre-publication review.  The whole point of the exercise is to apply a measure of ‘quality control’ based on assumptions about ‘common ground’. So, difficult though it might be put into practice, operating an editorial line at this initial stage in the journey towards publication would seem to be a legitimate exercise.  The editorial input could come at two points. The first point would be where a contribution was being considered for community review: below a certain threshold a piece might simply be rejected.  The second point would be after a piece had been given restricted publication for community review (i.e. where access and commentary is restricted those designated as ‘ASO Contributors’). An editor – like any other Contributor – could make a comment at this stage. But an editor could go further and, if his or her own comments were not sufficiently responded to, decline to pass the piece on for wider publication by ASO. This would be in the interests of preserving the standing of ASO itself as a publisher.  

But what of comments submitted after a contribution has been published? The original intention in developing a framework for a managed, formal exchange of opinion was to encourage a degree of focus and a level of quality that would serve to raise standards of debate within the AT world. At the same time, it’s hard enough to garner contributions to begin with: so over-zealous editorial input might simply kill the geese that lay the golden eggs (a.k.a. those with sufficient enthusiasm to make contributions to the debate). 

The website platform does allow different approaches to commentary, so we can accommodate more than one scenario. For example, we can use the ‘comment’ system to support more immediate and informal contributions. This was the system used for the responses to Graviception . It is analogous to the ‘Rapid Response’ system operated by the British Medical Journal. The responses to the ATEAM trial article provide a good example of how this works: . I have no doubt that these Rapid Response comments were lightly moderated, and possibly lightly edited.  In addition, there may well have been comments that were excluded because they did not meet an acceptable standard. But the threshold for acceptance seems to be pretty low. Any single point, made with sufficient clarity and without excessive rhetoric, seems to be acceptable (and rightly so). 

For an alternative model, more formal and more demanding of contributors, refer to the Open Mind project, and my comment on it, which notes that all the published material went through a process of peer review before seeing the light of day.

In the Open Mind example, however, the formality seems to have gone too far in the other direction, such that in every case there is one target article, one commentary paper, and one reply by the original author. This structure is clear from the list of papers. (If you want to see what one of these ‘sets’ looks like, the Daniel Dennett set [576KB PDF document] is probably a good example not least because it contains some additional points, in Note 1, about the criteria that a commentary paper should meet.)  So you get detailed, well-structured arguments, but along a narrow trajectory (so to speak) which does not seem to encourage ownership of the debate by the community at large.  

The only conclusion I have been able to come to is that, whatever the context, as an editor I need to maintain a relentless focus on clarity and encourage consistent use of terminology. How that editorial process plays out will vary according to whether the context is an open commentary, like the Graviception discussion and the BMJ Rapid Responses to the ATEAM paper, or whether it is a more formal review process happening prior to publication. And sometimes the process will involve declining to publish material that transgresses the ‘editorial line’.