See also Peer review, Scientific typesetting.
[To read] McElreath, R., & Smaldino, P. E. (2015). Replication, communication, and the population dynamics of scientific discovery. PLOS ONE, 10(8), e0136088. doi:10.1371/journal.pone.0136088
[To read] Higginson, A. D., & Munafò, M. R. (2016). Current incentives for scientists lead to underpowered studies with erroneous conclusions. PLOS Biology, 14(11), e2000995. doi:10.1371/journal.pbio.2000995
[To read] Smaldino, P. E., & McElreath, R. (2016). The natural selection of bad science. Royal Society Open Science, 3(9), 160384. doi:10.1098/rsos.160384
[To read] Edwards, M. A., & Roy, S. (2017). Academic research in the 21st century: Maintaining scientific integrity in a climate of perverse incentives and hypercompetition. Environmental Engineering Science, 34(1), 51–61. doi:10.1089/ees.2016.0223
Good table of “Growing Perverse Incentives in Academia”.
Olah & Carter, “Research Debt”, Distill, 2017.
This isn’t a new point, but it’s a very good exposition. (How appropriate.) The problem: to get to the leading edge of understanding in a scientific field, you have to digest all the previous work in that field. Most of the time there’s “research debt”: previous exposition is poor, ideas haven’t yet been polished, the notation and abstractions are clunky, and you have to sift through dozens of papers of differing relevance.
Ideas have to be distilled into reviews, textbooks, blog posts, whatever works best. But there are no institutional incentives to do the distillation, even if dozens of grad students will lie in your eternal debt.
The journal Distill is meant to promote this kind of distillation in machine learning.
Thurston, “On proof and progress in mathematics”, Bulletin of the American Mathematical Society (1994).
“What they really want is usually not some collection of ‘answers’—what they want is understanding”. Discusses, with a great story from Thurston’s career in Section 6, the need to communicate mathematics, not just prove new theorems. If new developments are not communicated to other mathematicians – not just through formal papers but through informal seminars, notes, textbooks, and whatever else conveys ideas efficiently in person – it’s difficult for new progress to be made. Thurston eventually dedicated much of his effort to these efforts, and felt it paid off by spurring great activity in the field.
The contrast between formal but impenetrable proofs and informal but effective exposition reminds me of Lamport’s hierarchical proof idea, which could combine both; see Writing proofs.