Shifting norms in scholarly publications: trends in readability, objectivity, authorship, and AI use

TL;DR

This paper analyzes 17 million research papers from 2000 to 2024 to quantify shifts in authorship, references, abstracts, and the use of AI, revealing pervasive author inflation and signs of gift authorship, along with a decline in readability and objectivity driven in part by AI-assisted writing. It introduces a scalable methodology to quantify readability (FRE), AI signal strength via AI-amplified terms, and hype via a 139-term set, applying these metrics across 23 fields and over two decades. The results show a marked rise in author counts and publication volume, a decline in weighted productivity, and a strong post-2022 AI signal that correlates with increased hype and reduced readability, particularly in Computer Science and Engineering. The study discusses policy implications, including recommending author-contribution declarations, normalization of impact metrics, and transparent disclosure of AI tool usage to preserve research quality and credibility in a rapidly evolving publication landscape.

Abstract

Academic and scientific publishing practices have changed significantly in recent years. This paper presents an analysis of 17 million research papers published since 2000 to explore changes in authorship and content practices. It shows a clear trend towards more authors, more references and longer abstracts. While increased authorship has been reported elsewhere, the present analysis shows that it is pervasive across many major fields of study. We also identify a decline in author productivity which suggests that `gift' authorship (the inclusion of authors who have not contributed significantly to a work) may be a significant factor. We further report on a tendency for authors to use more hyperbole, perhaps exaggerating their contributions to compete for the limited attention of reviewers, and often at the expense of readability. This has been especially acute since 2023, as AI has been increasingly used across many fields of study, but particularly in fields such as Computer Science, Engineering and Business. In summary, many of these changes are causes of significant concern. Increased authorship counts and gift authorship have the potential to distort impact metrics such as field-weighted citation impact andh-index, while increased AI usage may compromise readability and objectivity.

Figures (6)

• Figure 1: Summary results for all FoS since the year 2000. In each graph, the dashed line shows the mean value across all FoS.
• Figure 2: A closer look at authorship statistics for several Computer Science/Engineering journals with varying degrees of multidisciplinary research.
• Figure 3: Changing authorship and estimating author productivity. The line with the colour-coded markers shows the mean number of papers published per year per author across all fields of study, and indicates that author output has been steadily increasing up to 2020. The size of these markers indicates the total number of articles in our dataset that year, with several years labeled for reference. The colour-coding indicates the average number of authors per article, which has also been steadily increasing. The second (dashed) line is an estimate of author productivity based on the sum of the weighted number of articles per author per year. In this case an author on an article with $n$ authors receives a weighted output of $1/n$ for that article (uniform fractional weighting).
• Figure 4: Estimating author productivity by field of study: (a) the mean output per author per year for each field of study; (b) the corresponding weighted output per author per year, as an indicator of author productivity, for each field of study using the uniform fractional weighting model.
• Figure 5: The increase in the incidence of AI-associated terms across all FoS between 2020 and 2024. The dotted line on the left indicates an equal incidence between 2020 and 2024. All FoS show an increased incidence.