Slowed canonical progress in large fields of science

Abstract

In many academic fields, the number of papers published each year has increased significantly over time. Policy measures aim to increase the quantity of scientists, research funding, and scientific output, which is measured by the number of papers produced. These quantitative metrics determine the career trajectories of scholars and evaluations of academic departments, institutions, and nations. Whether and how these increases in the numbers of scientists and papers translate into advances in knowledge is unclear, however. Here, we first lay out a theoretical argument for why too many papers published each year in a field can lead to stagnation rather than advance. The deluge of new papers may deprive reviewers and readers the cognitive slack required to fully recognize and understand novel ideas. Competition among many new ideas may prevent the gradual accumulation of focused attention on a promising new idea. Then, we show data supporting the predictions of this theory. When the number of papers published per year in a scientific field grows large, citations flow disproportionately to already well-cited papers; the list of most-cited papers ossifies; new papers are unlikely to ever become highly cited, and when they do, it is not through a gradual, cumulative process of attention gathering; and newly published papers become unlikely to disrupt existing work. These findings suggest that the progress of large scientific fields may be slowed, trapped in existing canon. Policy measures shifting how scientific work is produced, disseminated, consumed, and rewarded may be called for to push fields into new, more fertile areas of study.

Keywords: durable dominance; entrepreneurial futility; science of science; science policy; scientific progress.

Fig. 1.

Changes in citation dynamics by size of field: Durable dominance of canon. X axes are logged (base 10) number of papers published in the subject-year (N). Each dot represents a subject-year in the Web of Science. Lines are lowess trendlines for the 10 largest nonmultidisciplinary subjects, listed in order of total number of papers published over all years in the dataset in the legend. (A) Gini coefficient of citation-share inequality by subject-year. The most-cited papers garner a larger proportion of new citations in years when more papers are published in a field. (B) Spearman rank correlation of the top-50 most-cited list between adjacent years by subject-year. The ordering of most-cited papers is more stable when more papers are published in a field.

Fig. 2.

Citation decay rates. X axes are logged (base 10) number of papers published in the subject-year (N). (A–C) 1−decay rate (λ) for top-percentile (A), 2nd-percentile (B), and 10th-percentile (C) most-cited papers in the 10 largest fields (see Fig. 1 for legend). 1–λ trends toward 1 for the top-percentile most-cited papers in the largest fields; the most-cited papers on average maintain their number of citations year over year when many new papers are published in the field. 1–λ trends to values <1 for the second and lower percentiles; all but the most-cited papers receive diminishing numbers of citations on average year over year in large fields. (D) 1–λ across all subject-years. The pattern is consistent with A–C. Across all large fields, only the top-cited papers maintain their level of citations on average year to year; all other papers on average receive fewer citations year over year.

Fig. 3.

New papers rising into canon. X axes are logged (base 10) number of papers published in the subject-year (N_p). Each dot represents a subject in the year 1980. Lines are linear trendlines for the 10 largest nonmultidisciplinary subjects (see Fig. 1 for legend) for papers published in 2000 and earlier. (A) Probability (p, in %) of a paper ever reaching the top 0.1% of most-cited articles. The probability of a newly published paper ever reaching the top 0.1% most cited in its field decreases when it is published in the same year as more papers in its field. (B) Median number of years (τ) for a paper to reach the top 0.1% of most-cited articles, conditional on reaching the top 0.1%. Papers published in the same year as many others in their subject do not gradually and cumulatively build up citations to reach the top 0.1%.

Fig. 4.

Propensity for disruption. X axes are logged (base 10) number of papers published in the subject-year (N_p). Each dot represents a subject-year. (A) Proportion (p) of papers published in the subject-year with disruption measure, D, from Wu et al. (4) greater than 0 (D > 0; blue dots and line) and lesser than 0 (D < 0; red dots and line). Lines are logistic fits to the data. Papers published in years with many others in their subject tend to develop existing ideas rather than introduce disruptive new ones. (B) Proportion (p) of papers published in the subject-year with top 5-percentile disruption measure (D ≥ 0.0256). Lines are linear trendlines for the 10 largest nonmultidisciplinary subjects (see Fig. 1 for legend). Papers published in the same year as many others in their subject are unlikely to be highly disruptive.