Last September I was in Crete for a conference and overstayed to have a few days of vacation.
The wife and I did a wonderful hike walking down through the Samaría Gorge. I won’t enter into the details: just be warned that it is a magnificent experience but also quite taxing. Bring food, water, hats, hiking shoes, and more water.
After the hike we had the opportunity of sunbathing on a nice rocky beach and take a swim, before taking a ferry back to the picking point for the bus back to Chaniá. During the bus trip the wife fell asleep, and I listened (albeit interrupted by the occasional loss of connection in the mountains) to an episode of the recommendable EverythingHertz podcast, hosted by Dan Quintana and James Heathers, featuring a nice interview with Kristin Sainani.
The interview mentioned Sainani’s scientific writing course on Coursera, but I shamelessly forgot about that—until last Sunday. While searching for a few references about writing, I serendipitously stumbled again upon the course, enrolled, and spent the last couple nights going through the very cool material. If my prose in this blog is not improving it’s totally by my fault: the course is very good.
Long story short, I just watched two of the course’s interviews: one with Brad Efron (the real one, not a bootstrapped replica), which I cite merely because you should really go and watch it; and one with George Lundberg, which I cite because I want to speculate on a point Lundberg raised.
Lundberg—a medic and editor of many journals, from what I could gather—states that you should choose the appropriate journal for your paper based on the typical readers you want to reach. On one side I agree (as also Efron mentioned, if you want readers interested in theoretical statistics you shouldn’t submit to the Journal of Applied Statistics), on the other side I tend to disagree on one specific sentence: “readers of some journals don’t read other journals”.
This was probably true in the pre-internet era, when you had to go to your university’s library to pick up a printed journal: maybe (if the library had not enough copies) you could also read it only for a moderate amount of time, to leave space for colleagues to read too, and had to ruthlessly select the papers you wanted to photocopy. I imagine that bibliographic research was based on a similar approach too—my older readers, if any, are welcome to comment on this. I hear that often people snail-mailed authors to ask for (snail-mailed) copies of their papers.
Nowadays you tend to search for papers online; you can get practically any paper from any journal via a plethora of sources: preprints, open-access journals, university online subscriptions, or a colleague willing to send you a PDF—or sci-hub, if you feel particularly remorseless.
I think that the younger generation certainly still accounts for the perceived importance of a journal when choosing what to read. But I also think that the separation between readers of this or that journal might have washed out so much that readers now should be rather divided by search keys—people searching for “sampling techniques” rather than “likelihood asymptotics” or “cats loving statistics”.
My hunch is totally anecdotal, but if you are interested let me know and we might think of setting up a study (if one does not exist already) in which scientists are interviewed about their reading habits.