Behavioral consequences of second-person pronouns in written communications between authors and reviewers of scientific papers

The psychological effects and behavioral consequences of pronoun usage, particularly second-person pronouns ("you," "your," "yours"), have been a subject of considerable research. While their impact is understood in many contexts, their role in dynamic, bilateral conversations outside close relationships remains relatively unexplored. This study addresses this gap by investigating the peer review process of scientific papers—a unique setting involving formal yet interactive written communication between authors and reviewers, who are not in close relationships—as a naturally occurring experiment. The central research question is: How does the use of second-person pronouns by authors in their responses to reviewers influence reviewer behavior and feedback? The study's importance stems from its potential to shed light on the dynamics of professional written communication and contribute to the broader understanding of the social function of language.

Existing research demonstrates that second-person pronouns can increase engagement and involvement in communication by directly addressing the recipient. Studies have shown that "you" can draw attention and evoke higher involvement, while generic "you" can signal normative behavior and impact persuasion. However, most research focuses on one-way or one-off communications, or is limited to close relationships where pronoun use reflects self or other-focus. The study aims to expand upon this literature by examining the effects of second-person pronouns in dynamic, bilateral, and interactive communication beyond close relationships, specifically within the context of the formal peer-review process.

The study analyzed revision correspondence data from *Nature Communications* between April 2016 and April 2021, encompassing 13,359 papers and 29,144 review rounds. The analysis focused on the authors' use of second-person pronouns in their first-round responses to reviewers. A difference-in-differences (DID) method was employed to compare outcomes (number of questions, length of responses, sentiment) between a treatment group (authors using "you") and a control group (authors not using "you"). This quasi-experimental approach accounted for initial differences between the groups. The DID model also included control variables such as paper length, number of authors, and publication year. Furthermore, the study used automated text analysis techniques (TextBlob, sentimentr, NLTK) to measure the sentiment (positivity and negativity) of reviewer comments. Latent Dirichlet Allocation (LDA) was used to identify topics in reviewer comments that could indicate engagement. Finally, a pre-registered behavioral experiment (N=1601) was conducted to provide causal support for the field data findings. Participants, assuming the role of reviewers, evaluated an identical author response, but with different pronoun usage ("you" vs. non-"you"), assessing its positivity and perceived personal engagement. Mediation analysis examined the relationship between pronoun usage, personal engagement, and positivity of the author's response. To ensure robustness, the study performed several sensitivity analyses, including propensity score matching and using only conversational 'you' usage. A difference-in-difference-in-differences (DDD) model was also used to analyze the effect of mutual "you" usage by both authors and reviewers.

The DID analysis revealed that authors using second-person pronouns in their responses received significantly fewer questions (a reduction of 4 questions compared to the control group), shorter reviewer comments (around 135 fewer words), and more positive feedback (consistent across multiple sentiment analysis methods). The use of "you" was associated with less negative reviewer feedback. Analyses of reviewer responses indicated decreased use of first-person singular pronouns ("I"), simpler vocabulary (lower word complexity), and greater engagement with the author's response as measured by LDA topic modeling. These findings suggest that "you" usage leads to more personal and engaging interactions. The pre-registered behavioral experiment confirmed these results. Participants addressed with "you" in the author's response rated the response more positively and perceived the communication as more personal and engaging. This effect was fully mediated by perceived personal engagement. The DDD model further showed that the effects of the author's "you" usage were amplified when the reviewers also used second-person pronouns in their initial comments.

The study's findings demonstrate a clear causal link between the use of second-person pronouns and more positive outcomes in the peer review process. The increased personal engagement fostered by "you" usage seems to lead to more constructive and less critical feedback from reviewers. These findings have implications beyond academia. The strategic use of "you" in formal written communication could enhance engagement and elicit more positive responses across various professional contexts including marketing and political communication. The research contributes to the literature on language use, particularly the effects of pronouns in dynamic interactions, particularly those characterized by a power imbalance, providing insight into the interplay between language, social dynamics, and communication effectiveness.

This study provides compelling evidence that the use of second-person pronouns in written communication can significantly impact the quality and tone of feedback. The findings demonstrate a causal link between "you" usage, increased perceived personal engagement, and more positive evaluations. These results have practical implications for authors seeking to improve the peer review process and for broader applications in formal written communication. Future research could explore the generalizability of these findings across other types of formal written communication and examine how pronoun use influences communication in more conflictual situations or across different cultural contexts. Investigating the nuances of pronoun usage in online settings, where most contemporary communications take place, would also be beneficial.

The study's reliance on data from published papers might introduce selection bias, as it excludes information on rejected manuscripts or authors who did not choose to publish their review correspondence. The limited availability of pre-first-round reviewer comments also prevents a full verification of the parallel trend assumption in the DID analysis. The interpretation of decreased first-person pronoun usage by reviewers as reduced self-focus needs further investigation, as other interpretations are possible. Future research should address these limitations by expanding data sources and exploring additional linguistic features in written communication.