Symbolic scarcity and epistemic waste: a systems theory of academic publishing failure

TL;DR

This paper argues that academic publishing prioritizes novelty over replication, leading to flawed scientific records and proposes structural changes to improve the system.

Contribution

The paper introduces a systems-theoretic analysis of publishing failures and proposes a two-tiered architecture to reduce epistemic waste.

Findings

Current publishing systems favor novelty over replication, leading to non-replication and fragility in science.

Structural issues include decoupling of epistemic value from publication outcomes and inefficient reviewer labor.

Generative AI increases manuscript production but strains evaluation systems, worsening scalability issues.

Abstract

Contemporary academic publishing increasingly operates under conditions of symbolic scarcity, shaped by rising submission volumes, constrained reviewer capacity, and evaluative incentives that privilege novelty, visibility, and perceived impact. While these dynamics vary across disciplines, their cumulative effect is a publication system that systematically favors novel and attention-generating contributions over replication, confirmation, and cumulative refinement. As a result, methodologically sound research that tests, stabilizes, or contextualizes existing findings is more likely to be delayed, displaced, or rendered invisible, contributing to persistent non-replication and long-term fragility of the scientific record.

This paper develops a systems-theoretic analysis of academic publishing, examining how prestige filtration, redundancy in peer review, and novelty-oriented selection interact with institutional incentive structures. Rather than attributing dysfunction to individual actors or emerging technologies, the analysis focuses on how current publication architectures shape evaluative behavior, reviewer labor allocation, and the composition of the visible scientific record.

The analysis identifies three interlocking failure modes: (1) a structural decoupling between epistemic contribution and publication outcomes driven by novelty-oriented symbolic selection; (2) systematic waste of reviewer labor due to non-transferable and repetitive evaluation processes; and (3) declining scalability of existing review infrastructures as generative AI lowers the cost of manuscript production while increasing evaluative load. Together, these dynamics suppress replication and cumulative verification, distort the visible scientific record, and misdirect expert attention away from epistemically stabilizing review.

To address these failures, the paper proposes a two-tiered publishing architecture that separates epistemic inclusion from symbolic curation, alongside a complementary tiered review model that aligns review intensity with epistemic risk. An optional framework for reviewer recognition is also outlined to support sustained evaluative engagement without undermining anonymity. These proposals are offered as conceptual system designs rather than prescriptive reforms, intended to clarify how current publishing architectures generate epistemic waste and to suggest structurally feasible pathways toward a more coherent, inclusive, and resilient scholarly communication system.