Editorial Free

I apologise for inflicting a second editorial on you so soon, but the panel has been very active since January, reaching a number of decisions that will have immediate effect, which I would like to share with you. Unfortunately, even if my aim is to reduce review and publication times, I have to start by apologising to any authors who have experienced delays caused by the migration of ICE Publishing journals from the old article-tracking system to ReView. This change is intended to benefit the journal in the medium to long term, providing a more modern interface for the panel, authors and reviewers. I can only assure you that we are working hard to minimise the short-term impact. I extend my thanks to the panel, our reviewers, but most of all to Craig Schaper, our journals editor, for their tireless efforts over the last few months.

As I highlighted in my first editorial, a priority is to reduce the print queue. Dave Potts had already started tackling this when he and his panel decided that Géotechnique should no longer publish technical notes, a decision which helps maintain clear water between us and our sister journal Géotechnique Letters. To see what other measures we might consider, we have been examining the data and I am sorry to say that we authors share some of the responsibility for the queue, as our papers have been increasing in length over the years. Here I must seem like poacher turned gamekeeper as I have to confess that I am one of the worst criminals in this respect. In Fig. 1 I have compiled the distributions of the lengths of full papers published in 2001, one of the first years of the A4 page size, and 2020. The differences are clear. It occurred to us that maybe the longer papers had greater value to the community, so I spent a few more idle lockdown hours compiling the citations on Web of Science to all papers, including technical notes, published in 2010 (Fig. 2). This year was chosen because of the slow accumulation of citations that is typical for the more successful papers, which I highlighted in my last editorial. It is not evident to me that our longer papers are more appreciated. So clearly we need to rein in our worst wordy excesses. The panel has therefore decided that all future papers will be limited to 12 published pages. Of course, there will be some ‘wiggle room’ when an author is hit by pages of detailed review comments, but certainly we should not start the review process at anything larger than this limit and reviewers and panel members will also be asked to highlight where papers may be cut; 12 is a maximum not target!

What is the often the highlight of our publishing year, the Rankine Lecture paper, should not escape scrutiny, and Fig. 3 shows their lengths; the lengths for older lectures have been adjusted to account for the A4 page format we now use. Leaving aside a few outliers, we can see that the paper lengths have more than doubled over the lifetime of the lecture. Clearly the trend is unsustainable and so the panel have decided that future lecturers will be invited to submit a paper with a guideline length of 20 pages, naturally with some flexibility, this being the length that they used to be (accounting for the modern A4 format). We have a number of outstanding Rankine Lecture papers, and of course we will not apply this retrospectively, but the panel hopes that asking for a more reasonable page extent will encourage a greater number of lectures to be written up.

We cannot be strict on page extents unless authors, reviewers and panel members can have a better estimate of the final page extent than we currently provide. So I spent many more lockdown hours out of my laboratory compiling data to derive a new equation. This is

To go with this equation we need a few instructions.

• (i)

  The number of words is for text main body, appendices and acknowledgements only (excluding the title and abstract which come in the 0·37 constant). You should use the count from Word.

• (ii)

  You will see that we now count not the number of figures, but the number of figure parts. So, (a), (b), (c) count as three figures even if you don't label them as such. A very complex figure, for example a graph with many data on it, should count as two as it will be printed across the two columns (see a published issue for examples). In a very few exceptional cases, very simple figures – for example, a figure with a thumbnail explanation – may be printed within the same column, in which case they may count as one figure (again, see a published issue for examples).

• (iii)

  We no longer count the number of tables, which like figures can vary very much in size and length, but we now count the words in tables. Again, the count from Word will work but you should include captions and notes.

• (iv)

  Notice that the references are not counted in the number of words but separately and we just need the number of references. Similarly, we count the number of items that you have in your notation; it is not an option to shorten a paper by not having one!

The improved accuracy of predicted page count from the new equation compared to the old is evident in Fig. 4.

In a global research environment that has become obsessed with metrics that value numbers over quality, I realise that I am asking much when I say that we must achieve these length reductions without losing content. We must therefore all please avoid the temptation to write more papers instead of long ones. Let us try instead to get back to the days of shorter, more significant papers.