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

The study asks whether addressing a conversational counterpart with second-person pronouns (e.g., "you") versus third-person references (e.g., "the reviewer") changes outcomes in dynamic, bilateral written communication beyond close relationships. Prior work shows that "you" can focus attention, increase involvement, and signal norms, but most evidence comes from one-way or one-off communications or from close relationships. In contrast, third-person pronouns typically signal objectivity and distance. The authors propose that using "you" makes conversations more personal and engaging, which should lead to review processes with fewer questions, shorter comments, and more positive evaluations. The peer review setting at Nature Communications provides naturally occurring variation in "you" versus non-"you" usage to test this hypothesis.

Existing research indicates that second-person pronouns draw attention and evoke involvement, and that generic "you" can convey norms and enhance persuasion. Cultural products using "you" may cue personal associations. Studies of pronouns in dyadic communication often focus on close relationships, linking pronoun use to self-/other-focus. Third-person pronouns can signal objectivity and reduced speaker involvement. Language features like subjectivity, pronoun use (first-person singular versus plural), and word complexity relate to interpersonal stance and engagement. Readability and conversational tone are associated with simpler vocabulary. The literature thus suggests mechanisms by which "you" could make interactions more personal and engaging, but prior work has not thoroughly examined these effects in professional, hierarchical, and iterative exchanges like peer review.

Field study: The authors compiled peer review correspondence for all Nature Communications papers published between April 2016 and April 2021 (13,359 papers; 29,144 review rounds). They focused on the first review-response-review sequence (reviewer comments in round 1, author response in round 1, reviewer comments in round 2), yielding 25,679 observations (88.11% of rounds). Papers were categorized as treatment if authors used second-person pronouns ("you," "your," "yours") in the 1st-round responses (N=5,042; 37.74%) and control otherwise (N=8,317; 62.26%). They implemented a difference-in-differences (DID) model comparing outcomes before and after the author response across treatment and control groups. Outcomes: number of questions and number of words in reviewer reports; sentiment (positivity via TextBlob and sentimentr; negativity via NLTK VADER and a hand-coded lexicon of 92 negative words). Mechanism indicators: subjectivity (TextBlob), first-person singular pronoun counts (I, me, my, mine), word complexity (average syllables per word), and reviewer engagement measured by the frequency of an LDA-identified "engagement" topic (topic 11) multiplied by review word count. LDA was implemented using R topicmodels with 40 topics chosen via lowest perplexity after standard preprocessing. Controls in DID included paper characteristics (pages, references, title length, number of authors), first author H-index, first author gender and last initial, authors’ friendliness and positivity in round 1, reviewers’ positivity in round 1, publication month and year dummies, and paper discipline dummies; paper fixed effects were included and standard errors clustered at the paper level. They also conducted robustness checks: dose–response to "you" frequency, restricting treatment to conversational (non-courteous) "you," restricting control to explicit third-person addressing, propensity score matching, alternative LDA topic numbers (35, 45), high-engagement word subsets, and structural topic models. They further tested a difference-in-difference-in-differences (DDD) design incorporating whether reviewers used "you" in round 1 to assess amplification when both parties use "you." Behavioral experiment: A pre-registered between-subjects experiment (N=1601, online) randomly assigned participants acting as reviewers to read an otherwise identical author response addressed with "you" versus with third-person references. Participants rated the positivity of the response and perceived personal/engaging nature of the exchange on Likert scales; mediation analysis tested whether perceived personal/engaging communication mediated the effect on positivity. A second replication (N=1200) tested robustness and alternative explanations.

• DID main outcomes: Authors’ "you" usage led to fewer reviewer questions and fewer words in subsequent (2nd-round) reviewer comments. Basic DID coefficient for questions: B = -4.0019, t(25675) = -10.01, p < 0.001; with controls and paper fixed effects: B = -3.3360, t(12319) = -9.40, p < 0.001 (≈3.34 fewer questions). For word counts: basic DID B = -172.15 words, t(25675) = -9.54, p < 0.001; with controls/fixed effects B = -135.59 words, t(12319) = -9.36, p < 0.001.
• Sentiment: Reviewer comments became more positive and less negative following authors’ "you" usage. With controls/fixed effects: Positivity (TextBlob) B = +0.0051, t = 2.12, p < 0.05; Positivity (sentimentr) B = +0.0173, t = 4.45, p < 0.001. Negativity (NLTK VADER) B = -0.0019, t = -2.88, p < 0.01; Negativity (hand-coded) B = -0.0048, t = -6.83, p < 0.001.
• Mechanism indicators: Subjectivity increase was not significant. First-person singular pronoun usage decreased: B = -1.0918, t = -6.84, p < 0.001. Word complexity decreased (simpler language): B = -0.0139, t = -4.49, p < 0.001. Reviewer engagement (LDA-based) increased: B = +0.1078, t = 5.85, p < 0.001.
• DDD: Effects on mechanism indicators were generally larger when reviewers had initiated "you" usage in round 1, indicating amplification when both parties use "you" (e.g., additional reductions in first-person singular use and complexity; increased engagement). DID effects on main behavioral outcomes tended to be larger with reviewer-initiated "you," but DDD differences for these main outcomes were not statistically significant.
• Robustness: Stronger effects with greater "you" frequency; results hold when excluding purely courteous "thank you" cases and when restricting controls to explicit third-person addressing; propensity score matching and alternative topic modeling specifications yield consistent conclusions.
• Experiment: Participants addressed with "you" rated responses more positively (M = 5.77, SD = 0.98) than those addressed in third person (M = 5.61, SD = 1.01), F(1,1599) = 10.62, p = 0.001, d = 0.16, 95% CI [0.06, 0.26]. They also perceived the interaction as more personal and engaging (M = 5.13, SD = 1.10 vs. M = 4.76, SD = 1.24), F(1,1599) = 40.78, p < 0.001, d = 0.32, 95% CI [0.22, 0.42]. Mediation: the effect of "you" on positivity was fully mediated by perceived personal/engaging communication (indirect effect = 0.19, SE = 0.03, 95% CI [0.14, 0.26]). Findings replicated in a second sample (N = 1200) and were not explained by alternative processes (e.g., contention, personal connection, obligation).

The findings demonstrate that addressing reviewers with second-person pronouns during peer review causally and behaviorally influences reviewers’ subsequent responses: they ask fewer questions, write less, and express more positive and less negative sentiment. Field evidence (DID) and experimental evidence converge on the mechanism that "you" fosters a more personal and engaging conversational tone, reflected in reduced self-focused language, simpler vocabulary, and higher engagement-topic prevalence. These results expand understanding of pronoun effects from one-way messaging and close relationships to professional, hierarchical, and iterative exchanges. Practically, authors may strategically use "you" to encourage constructive and collegial dynamics during review. More broadly, the work suggests that second-person addressing could enhance engagement and attitudes in formal written communications in domains such as marketing, professional correspondence, and politics, although context-dependent efficacy warrants further testing.

This paper shows that using second-person pronouns in authors’ responses during peer review makes the interaction feel more personal and engaging, leading to fewer reviewer questions, shorter reports, and more positive, less negative feedback. The work advances the literature on language and pronouns by documenting behavioral consequences of "you" in dynamic, bilateral, and hierarchical written exchanges, and contributes to the science of science by linking linguistic choices to review outcomes. Future research should examine diverse and more contentious contexts, include rejected and non-opt-in review records, test generalizability across disciplines and platforms, and disentangle nuanced roles of first-person pronouns (e.g., self-focus vs. self-disclosure) in shaping conversational dynamics.

Key limitations include: (1) The dataset lacks pre–1st-round reviewer comments, preventing direct testing of the DID parallel trends assumption and leaving non-random "you" usage as a concern (mitigated via PSM, Heckman model, permutation tests, and experiments). (2) Only ultimately published papers are included, excluding rejected submissions and potentially inducing selection bias. (3) Review correspondence publication was optional in the period studied (2016–2021), so only authors opting to publish their records are observed, further limiting generalizability and excluding potentially more conflictual interactions. (4) These selection constraints prevent comparing accepted versus rejected manuscripts and authors who did versus did not publish review records. (5) Interpretation of reduced first-person singular usage as decreased self-focus is complex; "I" can also reflect concreteness or self-disclosure, which may also relate to conversational tone.