A paper I wrote together with Claudio Tennie has just been published in the Journal of Comparative Psychology. The role of redundant information in cultural transmission and cultural stabilization presents an individual-based model of the following, quite straightforward, idea (which was, admittedly, Claudio’s idea).
Psychological literature on social learning generally explores how different learning mechanisms focus on different aspects of the behaviour one tries to learn (goals, actions, results, etc.). Another way to think about it is that these different mechanisms provide redundant information on the behaviour. Simplifying a bit, it is like acquiring simultaneously different copies of the same information, so that, if there is an error in one of the copies (transmission is likely to be error prone) this might be corrected using another copy that does not have the same error. We explored how this trick can help social learners in a way analogous to how redundancy is more often though about in the literature (i.e. as repeated attempts to copy). Perhaps more interestingly, another result of our model show that, as the machinery allowing redundancy is likely to be costly, it evolves only when transmission is sufficiently faithful (if transmission is very bad, the redundancy machinery would be a waste). This suggest a co-evolutionary dynamics in which redundancy machinery and hi-fi copying mechanisms reinforce each other. Our evolutionary simulations tended to end up in two distinct states: high-fitness/high-costs states (with high redundant copying and complex culturally transmitted behaviors; resembling human culture) or low-fitness/low-costs states (with low redundant copying and simple transmitted behaviors; resembling social learning forms typical of nonhuman animals).
If you want to know more, please have a look at the paper (but unfortunately the link above is gated, and I do not have myself the pdf – more on this below!). In fact this post is not about the content of the paper, but about its tormented editorial saga. While part of this is certainly due to our sub-optimal strategy, I believe it is a good sketch of why everybody is getting more and more fed up with the traditional academic publishing system.
Here it is. The first mention of a manuscript about redundancy in cultural transmission goes back, according to my email archive, to 2009, i.e. seven years ago. After some work, we submitted it in an high-profile journal, in 2011, were it was rejected (before going through the review process) because it did not have “the broad appeal needed”. We worked again on the manuscript, and, the following year, – 2012 – we tried with another (quite) high-profile journal, with the same result. Before being reviewed, we were rejected on the basis of our simulations not being “sufficiently surprising and substantial advance for this journal”.
At this point we resubmitted to another, more specialised, journal, in March 2012. In July we received the negative answer from this journal. The editor had sent our manuscript to 6 (six!) reviewers which provided, in fact, very detailed criticisms (unfortunately, for the most part, one different from the other). So we went back to our model, re-worked it thoroughly, trying to follow the various suggestions. In the process we lost one of the coauthors – fair enough – and, happy with our reworking, we decided to submit again to another high-profile journal, which, again – we are now at the beginning of 2014 – rejected it, on the basis of two negative reviewers evaluations.
The 1st of April 2014, i.e. around two years ago (!), the manuscript was submitted to the journal where is now published. In September 2014, not having received any news from the journal, we wrote to the editor, which told us that they had “very difficult time finding reviewers – we’ve contacted over a dozen potential referees, and most of the people we contacted didn’t even bother to respond”. Excellent. After other six months, we finally received the response from the editor. They found three referees, their evaluations was partly negative, but the editor was “very willing to reconsider it if you are able to undertake a thorough revision and resubmit it for review. ” We modified again our manuscript, provided detailed answers to the reviewers comments and resubmitted. In May 2015, we were notified that two of the three original reviewers were unable to re-evaluate our manuscript, so the editor involved another review, which – guess? – was not completely satisfied with our paper, and asked to resubmit. So we did, taking into account the new reviewer considerations, but, in October – I am not joking – we were notified that the editor had to involve ANOTHER reviewer (now making for a total of 13 reviewers and 5 editors), which, well, asked us to resubmit.
After the third resubmission, this time very quickly, the editor told us that that there were “just a very few points that needed to be addressed”, so we resubmitted, again, for the fourth time, to the same journal, and then our redundancy-RRRR was finally accepted, the 12 of November 2015, around one year and a half after the submission to the journal, and more than four years after the first submission.
Almost the end of the story. Once accepted we had to comply with APA formatting (after, of course, having reformatted five times the manuscript for the submissions in different journals, references included). From what I understood, the journal did not have a template (Word, or God forbids, LaTeX) but one can look around and hope to find one template corresponding to the current edition of the APA style manual, i.e. the 6th, or, naturally, buy the manual. Finally, we compiled a bunch of documents (including ethics forms, in case I mistreated my matlab agents), and we had the proof.
Just some few other small details. We were contacted by Claire Asher, a science blogger that wanted to report on our paper, after Claudio Tennie gave a conference talk about our model in December. As she wanted to know the publication date, we enquired with the journal, which told us that the paper would have been online the 5th of February. This did not happen, and the paper appeared around 10 days after, without none of us being notified. The “we” of the abstract have been changed in “the authors”, but not all of them (at least I spotted one “we” remaining – anyway that may be my fault, as I did not notice it in the proof).
On top of this, as I mentioned at the beginning of this post, at the moment we can not access our own paper! Even if we payed (in fact, Claudio’s ESRC grant payed) 4,000 dollars for having the paper open access, there was some problem, and we are still waiting to understand what exactly went wrong (apparently the paper will be, at some point in the future, open access, but through the Wellcome Trust database)*. Of course one can buy the pdf for 11.95 dollars, thus supporting this outstanding service.
If you were able to arrive here, let me conclude with a positive note. I hope to be able soon to be involved in the creation of a new Journal of Cultural Evolution (or similar), about which I discussed in the past (see here and here), now being probably supported by the recently created Society for the study of Cultural Evolution. When I started to think about the Journal of Cultural Evolution project, my views about academic publishing system were pretty mild, but now experiences like the one above contributed to radicalise them. So let’s see what will happen!
*update 3/3/2016: finally the paper is OPEN ACCESS: link here
Acerbi, A., Tennie, C. (2016), The role of redundant information in cultural transmission and cultural stabilization, Journal of Comparative Psychology, 130(1), 62-70