Quantifying the rise and fall of scientific fields.

Area of Science:

• Scientific field evolution
• Research dynamics
• Innovation studies

Background:

• Scientific enterprise growth is exponential, but individual research fields exhibit rise and fall dynamics.
• Previous studies analyzed field evolution qualitatively on a case-by-case basis.
• Understanding generic evolutionary stages requires quantitative comparison across diverse fields.

Purpose of the Study:

• To develop quantitative techniques for identifying common patterns in scientific field evolution.
• To analyze the evolutionary stages of 175 research fields using a large preprint dataset.
• To explore generic properties of articles and authors during different evolutionary phases.

Main Methods:

• Analysis of 1.5 million preprints from arXiv across Physics, Mathematics, Computer Science, Quantitative Biology, and Quantitative Finance.
• Development of techniques to quantify and compare field evolution on common terms.
• Modeling field dynamics using a two-parameter right-tailed Gumbel temporal distribution and field-specific re-scaled time.

Main Results:

• Scientific fields consistently follow a predictable rise and fall pattern.
• Early field phases are characterized by disruptive, interdisciplinary work from small teams.
• Late field phases show specialized, large teams building on prior research.

Conclusions:

• A quantitative method is established to explore generic patterns in research field evolution.
• The findings have implications for understanding innovation and scientific progress.
• Field evolution follows a common trajectory from interdisciplinary novelty to specialized consolidation.