The physics of journal self-citation

Abstract

Purpose: This study explores the implications of low journal self-citation rate (LJSCR). While some interpret LJSCR as a sign that a journal’s authors do not cite each other’s work, others see it as a mark of prestige, reflecting greater recognition from outside the journal. We argue that these perspectives are not contradictory: low self-citation can be prestigious precisely because it reflects low self-readership. Design/methodology/approach: We analyze the physics and mathematics of journal self-citation. Our findings show that the self-citation rate (i) increases with journal market share, (ii) approaches a well-defined upper bound, (iii) this upper bound remains significantly below unity, and (iv) without a minimum level of market share, self-citation is virtually absent. Here, market share refers to the proportion of a journal’s publications relative to its Web of Science (WOS) subject category. To test our analysis, we examine 61 journal-years of data from three major hybrid fluid dynamics journals: Journal of Fluid Mechanics, Physical Review Fluids, and Physics of Fluids. Findings: We identify a consistent relationship between journal self-citation and market share. A striking result is the mathematical analogy we establish between journal self-citation behavior and the concentration of intermediates in consecutive irreversible unimolecular chemical reactions. We also observe that creating specialized subdisciplinary journals (“twigging”) can reduce self-citation rates by narrowing scope. Research limitations: The study is limited to fluid dynamics journals. Broader validation across disciplines is needed. Practical implications: Editors and publishers can apply these insights to interpret citation metrics and assess the visibility and readership of their journals. Originality/value: This work redefines LJSCR as a counterintuitive metric—one that may reflect both low author engagement and high external impact. It introduces a novel physics-based model to understand citation behavior across journals.