Chaining and the temporal dynamics of scientists' publishing behaviour

Area of Science:

• Philosophy of Science
• History of Science
• Bibliometrics
• Scientometrics

Background:

• Characterizing scientific progress quantitatively is established, but individual scientists' career progression remains less understood.
• Previous research has focused on field-level scientific change, leaving a gap in understanding individual career trajectories.
• A comprehensive, large-scale analysis of individual scientists' publishing dynamics is needed.

Purpose of the Study:

• To characterize the regularity in temporal publishing dynamics of individual scientists.
• To explore how scientists build upon their previous work throughout their careers.
• To identify common patterns in scientific career progression across different disciplines.

Main Methods:

• Utilized computational algorithms to predict the historical order of publications from individual scientists.
• Analyzed publishing behavior of 1,164 award-winning and random-sampled scientists.
• Examined data from Physics, Chemistry, Medicine, Economics, and Computer Science over the past century.

Main Results:

• Discovered that scientists' publishing behavior follows a "chaining" process, similar to historical word meaning extension.
• Identified "exemplar-based chaining" as a predominant pattern, where new ideas connect to proximal existing ones.
• Found consistent publishing behavior across diverse scientific fields, characterized by incremental steps building on past research.

Conclusions:

• Scientists across various fields exhibit similar publishing dynamics over time.
• Individual scientific progress is largely characterized by incremental advancements building on prior work.
• The findings provide large-scale evidence for a shared, incremental model of scientific career progression.