This line of reasoning seems strong, but seriously conflicts with academics’ current behaviour, so I’d be interested to see if they have much useful to say in response.
I am an academic and I have published about 50 papers and one book. I think the book was helpful in getting more popular media articles. However, my impression is that the typical academic book only sells a few hundred copies. One could hope for a few more reads than this because of libraries, but some people buy a book and do not read it. As for papers, there is a big difference between conference and journal. I could believe that the median paper only get zero citations because the median paper is a conference paper. However, even relatively low-ranked journals have an impact factor of one, which means an average of five citations over five years. And if you think about it, if the average paper has 30 references, that means that the average paper also get cited 30 times. And therefore the average journal article would be cited significantly more. Furthermore, in my experience on sites like Research Gate and Academia.edu, papers typically have one-two orders of magnitude more reads than citations. So I would guess the average journal article gets read thousands of times. I think this is significantly more than the typical academic book or LessWrong. But I don’t have a lot of hard data here, so I am happy to update.
As for being successful in academia, I think only doing books would be very risky. People pay some attention to total citations, and this would likely be higher with more pieces of work (papers versus book). Also the h index is the number of publications that have been cited at least that number of times (see Bostrom’s). One rule of thumb is achieving and h index of 12 in order to get tenure. Publishing significantly more than 12 books before tenure sounds difficult. Maybe evaluators would change their metrics, but I would not count on it.
This line of reasoning seems strong, but seriously conflicts with academics’ current behaviour, so I’d be interested to see if they have much useful to say in response.
I am an academic and I have published about 50 papers and one book. I think the book was helpful in getting more popular media articles. However, my impression is that the typical academic book only sells a few hundred copies. One could hope for a few more reads than this because of libraries, but some people buy a book and do not read it. As for papers, there is a big difference between conference and journal. I could believe that the median paper only get zero citations because the median paper is a conference paper. However, even relatively low-ranked journals have an impact factor of one, which means an average of five citations over five years. And if you think about it, if the average paper has 30 references, that means that the average paper also get cited 30 times. And therefore the average journal article would be cited significantly more. Furthermore, in my experience on sites like Research Gate and Academia.edu, papers typically have one-two orders of magnitude more reads than citations. So I would guess the average journal article gets read thousands of times. I think this is significantly more than the typical academic book or LessWrong. But I don’t have a lot of hard data here, so I am happy to update.
As for being successful in academia, I think only doing books would be very risky. People pay some attention to total citations, and this would likely be higher with more pieces of work (papers versus book). Also the h index is the number of publications that have been cited at least that number of times (see Bostrom’s). One rule of thumb is achieving and h index of 12 in order to get tenure. Publishing significantly more than 12 books before tenure sounds difficult. Maybe evaluators would change their metrics, but I would not count on it.