Knowledge evolution in physics research: An analysis of bibliographic coupling networks

Area of Science:

• Physics
• Bibliometrics
• Science of Science

Background:

• Current understanding of scientific knowledge evolution relies on descriptive models (Popper, Kuhn).
• The dynamic interplay between established and emerging research fields requires quantitative analysis.

Purpose of the Study:

• To quantitatively investigate how new physics knowledge builds upon existing knowledge.
• To model the evolutionary relationships and dynamics of research fields.

Main Methods:

• Construction of year-to-year bibliographic coupling networks from American Physical Society (APS) publications.
• Identification and visualization of research field communities using alluvial diagrams.
• Quantitative analysis of field merging, splitting, and citation patterns.

Main Results:

• Most research fields exhibit weak 'Popperian mixing' with rare isolation or strong mixing.
• Field sizes demonstrate linear growth correlated with recombination.
• Merging of fields is predictable, while splitting is more complex.
• Kuhnian events, like field merging/splitting, correlate with major breakthroughs (e.g., Bose-Einstein condensation).

Conclusions:

• Scientific knowledge evolves through predictable patterns of field interaction and growth.
• Bibliographic coupling networks offer a robust method for tracking field dynamics.
• Major scientific advancements are linked to specific evolutionary events within research fields.