The ideal academics of bygone eras were polymaths, Renaissance men. These people were well-learned and well-read. Their modern counterparts are masters of an extremely narrow field and write a lot. How much they read we don’t really know. For myself, much less than I would like to.
Most scientists would agree that one can estimate the worth of a scientist by reading five of their publications. We ask people to list their selected five publications. Now tell me: when was the last time you read those five papers of someone you evaluated for a position or the possible award of a grant? I can defend my “no read” mode by pointing to the guides, which never asked me to read these papers. I was instead asked about the number of papers, the number of citations, the share of D1/Q1 papers, or cumulative impact factors of scientists. We evaluate scientists on metrics that were never meant to be used for that.
Just look at the San Francisco Declaration on Research Assessment, The Leiden Manifesto or the 'Agreement on reforming research assessment'. They all agree that journal characteristics should not be used to assess scientists. But apart from nodding internally or even joining as an institute, did anything change?
We might even agree, over a beer, but not in an actual committee meeting, that reading some papers and then deciding on a person’s life might be better than leaving the decision to bibliometric measures of journals. But then we also agree that it would involve too much reading. We don’t have time to read that much. We need to write.
I won’t even get started on the publish or perish problem. As a good scientist I wrote a paper about it.
But there is more to publish and perish than just our plight. We publish and others cash in a lot of money. Only the for-profit publishing houses benefit from our drive to write. We hand in our work for free, we evaluate others’ work for free, we edit journals for free, and then we either pay to have access to our own papers (remember, we don’t read anymore) or pay to get published. We don’t do it because there is no other way. There is. Scientific journals even a few decades ago were mostly in the hands of learned societies, departments, and enthusiastic individual scientists. Scientists then handed in their work for free, evaluated others’ work for free, and edited journals for free. And money was only asked for printing the journals. Now in the digital era, that would not be necessary. So why not switch back to this mode? Why let the few remaining such journals die (cf. a commentary in Nature)? The solution is simple, albeit we might not want to admit it: we don’t want or don’t have time to read.
If journals that we reviewed for and edit can preselect papers, it would have saved us a lot of reading. We still have the naïve belief that the best science goes to high-ranking journals. So we can evaluate others based on whether they got into those journals or not. No need to read their papers. Add up the impact factors and we are done. Also, it preselects the few papers we might actually read (or ask our PhD student to read instead of us). If you have time to read only one paper, which would you choose? The one in a high-ranking journal or the one in the department-owned journal of a university you never heard of? Both papers seem to be equally relevant based on the abstract. So which would you read?
You know you have to read the paper in the high-ranking journal (at least have to cite it) because a referee might criticize you for not citing it. Referees (us) also don’t have time to read and gave exactly the same answer to the previous question. So we read (or just cite) the paper in the high-ranking journal. One more citation to retain its high rank. And we all congratulate ourselves to know the “relevant” literature.
Publication fees, sometimes euphemistically called processing charges, are an even better way to preselect papers for us. If people can pay 1000–3000 Euro (USD) for the publication of one paper, then someone entrusted them with that amount of money. The granting agencies cannot be wrong. Less reading for the editors and the referees. For this amount of money, I work for 1–4 months, and PhD students twice as much. Just to let this sink in: an associate professor can individually color each of the letters in your manuscript and would still be idle for quite some time for the money we pay for “processing charges”. But if that money is the costly signal that tells us to maybe read that paper, then it is worth it. Does it?
Wouldn’t it worth more, collectively and also individually, to read more?
One of my supervisor’s first pieces of advice to me was that scientists should spend 80% of their time reading. She admitted that it was something to strive for. Seeing how much I should read just to keep up with the state-of-the-art in my narrow field, I fully agree. This should be the norm. That 20% of our time should be more than enough for teaching and good science.