It’s all too much

This past October, Johan Chu of Northwestern University and James Evans the Chicago published their analysis bibliometric data drawn from 10 broad fields in science, medicine, engineering, mathematics.11. J. S. G. Chu, A. Evans, Proc. Natl. Acad. Sci. USA 118, e2021636118 (2021). https://doi.org/10.1073/pnas.2021636118 Their main conclusion is sobering: As number papers a field increases, researchers find it harder to recognize innovative work. Progress seems be slowing.Chu present plausible hypothesis. Rather than evaluate new on individual merits, increasingly resort comparing them with existing paradigms. What’s more, when are at high rate, truly novel ideas struggle prevail—let alone noticed—over flood competitors.A strength Evans’s paper that they used hypothesis make six predictions, all which could test by looking for correlations. The predictions are, quote paper,1) citations will more likely cite most-cited rather less-cited papers; 2) list change little year year—the canon ossifies; 3) probability eventually becomes drop; 4) do rise into ranks those most cited not so through gradual, cumulative processes diffusion; 5) proportion newly developing scientific increase disrupting decrease; 6) becoming highly disruptive decline.Citations straightforward count. To characterize paper’s canonicity, disruptiveness, diffusibility, developed statistical measures. In all, examined 1.8 billion 90 million 1960 2014. Each was affirmed significant correlation.Physics represented study applied physics. Could averaging progress across such yield spurious correlations? I wondered. At start range, 1960, laser had just been invented, transistors yet supplant vacuum tubes electronic devices. middle 1987, high-Tc superconductivity recently discovered, magnetic tape remained popular storage medium. Research activity other areas physics waxed waned fashion, but couldn’t see how fluctuations might mimic secular in, say, canon’s rigidity. It’s also hard correlations began noticeable 1970s attributed China’s impressive recent growth output.One alternative explanation revolutionary paradigm shifters presume. 2014 Richard Van Noorden, Brendan Maher, Regina Nuzzo 100 science.22. R. B. Nuzzo, Nature 514, 550 (2014). https://doi.org/10.1038/514550a Top “Protein measurement Folin phenol reagent,” 1951 introduced what became known as Lowry protein assay.33. O. H. et al., Biol. Chem. 193, 265 (1951). https://doi.org/10.1016/S0021-9258(19)52451-6 Like top papers, describes useful, widely applicable method. When scientists does describe methods.But doesn’t hold up, least Applied Physics Letters method papers. Indeed, one helped Shuji Nakamura win share Nobel Prize blue LED.44. Nakamura, T. Mukai, M. Senoh, Appl. Phys. Lett. 64, 1687 (1994). https://doi.org/10.1063/1.111832If conclusions valid—and think are—then done? Telling physicists publish less, publishers stop launching journals, editors reject illiberal restrictions freedom expression. But need better way papers’ novelty. Whoever develops would earn scientists’ gratitude. They even become rich.ReferencesSection:ChooseTop pageReferences <<CITING ARTICLES1. https://doi.org/10.1073/pnas.2021636118, Google ScholarCrossref, ISI2. https://doi.org/10.1038/514550a, ScholarCrossref3. https://doi.org/10.1016/S0021-9258(19)52451-6, ISI4. https://doi.org/10.1063/1.111832, ISI© 2021 American Institute Physics.