The number of authors producing n papers falls off roughly as one over n squared in the datasets Lotka analysed.
Lotka's 1926 finding from chemistry and physics abstracts, in which a large share of authors published exactly one paper. Later bibliometric studies often find heavy concentration but not the exact inverse-square form.
In practice
- A large share of researchers publish exactly one paper, which is why productivity distributions look nothing like a bell curve.
- A company's patent portfolio is mostly the work of a handful of engineers, and the concentration recurs even where the exact distribution does not.
- Possible explanations
- The explanation is cumulative advantage: publishing raises the chance of publishing again, through skill, reputation and access to resources, which produces a heavily skewed distribution.
- In its favour
- Extreme concentration of authorship is found in field after field, and the qualitative shape, most authors publishing once, is robust.
- Against it
- The exact inverse-square form is not general: later bibliometric work finds exponents that vary by field, period and how authorship is counted.
- Standing
- Later bibliometric work often finds heavy concentration without the exact inverse-square distribution.
- Further reading
- Wikipedia
- Subject
- Creativity & Work
- Essay
- What gets in the way of good work