When Writing Shapes Scientific Influence

Nguyen Thi Nguyet Nuong

East Asian University of Technology, Hanoi 100000, Vietnam

*Contact: nuongntn@eaut.edu.vn

© Nick Morrison

Scientific discoveries only matter if they can be communicated and understood. This simple idea has made scientific writing one of academia’s most important skills. Yet an interesting paradox has emerged in modern science: although researchers increasingly produce plain-language summaries, blogs, and outreach materials, scientific papers themselves are becoming harder for the public—and sometimes even for scientists outside the field—to understand (Fawcett & Higginson, 2012; Thelwall et al., 2013; Plavén-Sigray et al., 2017).

Part of the challenge comes from the nature of science itself. Research is becoming increasingly specialized, creating communities that often speak highly technical “dialects.” In ecology, this issue becomes even more complex. Unlike disciplines such as physics or chemistry, ecological systems are highly variable and influenced by countless interacting factors. Findings are often probabilistic rather than absolute, leading some researchers to classify ecology as a comparatively “soft” science (Pigliucci, 2002; Colyvan & Ginzburg, 2003). In such environments, many competing studies may offer similar conclusions, making communication style potentially as important as scientific content itself.

Scientists have long relied on advice such as “write clearly,” “tell a story,” or “keep the reader engaged.” However, much of this guidance has traditionally been anecdotal. Recently, researchers have begun asking whether scientific writing itself can be studied scientifically.

A recent study by Veresoglou and Agathokleous (2026) examined review papers published in 2020 across the ten highest-ranked ecology journals. Using computerized text analysis and psychometric tools, they explored whether linguistic style and narrative structure influence citation frequency—the number of times a paper is referenced by later studies.

Their findings suggest that writing style, like the language norms and narrative arc, does matter. Although the direct effect explained only a modest portion of citation differences, it remained detectable even without accounting for major influences such as topic choice, innovation, author reputation, or open-access availability. Certain narrative patterns appeared more frequently among highly cited papers. One particularly interesting pattern was a “U-shaped” narrative structure.

Highly cited papers often spent more effort at both the beginning and the end, establishing the broader context of the study. At the beginning, authors introduce the problem, the key ideas, and why the topic matters. Toward the end, they return to the bigger picture by revisiting the main findings and explaining their implications. In contrast, the middle sections tend to focus more heavily on evidence, analysis, and technical details. Like a story that opens by setting the scene and closes by bringing readers back to the central message, scientific papers that guide readers through this rise-and-return pattern may leave a stronger impression.

The researchers also observed that speculative language seemed more effective earlier in papers than later. This may reflect the unique structure of scientific writing itself. Researchers often begin by presenting questions and hypotheses that naturally involve uncertainty and possibility. As evidence accumulates throughout a paper, readers may expect increasing clarity and confidence rather than continued speculation.

The findings become particularly relevant in the era of artificial intelligence. Large language models are rapidly becoming integrated into scientific writing. These tools can improve efficiency and accessibility, but they also learn from existing literature. If current writing norms are suboptimal, artificial intelligence could unintentionally reinforce them rather than improve them.

Scientific writing is not merely the transmission of information but a process of value formation through interactions among ideas, authors, readers, and cultural contexts. Knowledge acquires influence not simply because it is reliable, but because information can successfully pass through layers of cognitive filtering and connect with readers’ existing mental frameworks (Vuong, 2025; Nguyen & Ho, 2026). Thus, effective scientific communication may emerge not from choosing between precision and storytelling, objectivity and emotion, or rigor and accessibility, but from harmonizing them (Khuc & Nguyen, 2026).

References

Berdejo-Espinola, V. & Amano, T. (2023). AI tools can improve equity in science. Science, 379(6636), 991. https://doi.org/10.1126/science.adg9714 

Colyvan M. & Ginzburg, L.R. (2003). Laws of nature and laws of ecology. Oikos, 101(3), 649-653. https://doi.org/10.1034/j.1600-0706.2003.12349.x 

Fawcett, T.W. & Higginson, A.D. (2012). Heavy use of equations impedes communication among biologists. PNAS, 109(29), 11735-11739. https://doi.org/10.1073/pnas.1205259109 

Khuc, V. Q. & Nguyen, M. H. (2026). Cultural Additivity Theory. https://books.google.com/books?id=Y4XZEQAAQBAJ 

Nguyen, M. H., & Ho, M. T. (2026). The absurdist approach to unveiling possible paradoxical thinking for innovative socio-psychological research. MethodsX, 16, 103910. https://doi.org/10.1016/j.mex.2026.103910 

Pigliucci, M. (2002). Are ecology and evolutionary biology “soft” sciences? Annales Zoologici Fennici, 39, 87-98.

Plavén-Sigray, P., et al. (2017). The readability of scientific texts is decreasing over time. eLife, 6, e27725. https://doi.org/10.7554/eLife.27725 

Thelwall, M., et al. (2013). Sugimoto Do altmetrics work? Twitter and ten other social web services. PLoS One, 8 (5), e64841. https://doi.org/10.1371/journal.pone.0064841 

Veresoglou, S. D. & Agathokleous, E. (2026). On the makeup of high-impact reviews in ecology. Eco-Environment & Health, 5(2), 100233. https://doi.org/10.1016/j.eehl.2026.100233 

Vuong, Q. H. (2025). Wild Wise Weird. AISDL. https://books.google.com/books?id=C5dDEQAAQBAJ