Text analysis is mainly based on the review comments from reviewers. The statistics of the number of review comments in two rounds of two papers are shown in Table 3. One thing to note is that one reviewer was added in the final round of review of Paper 2021.

In the light of Halliday’s (1985) theory of systemic functional linguistics, a three-dimensional analysis framework was constructed according to different focuses, including technical depth, scholarly engagement, and text organization (see Table 4).

Based on Swales’ (2004) theory of academic discourse interaction, this research examined the comments and responses of two papers, and divided the interaction modes into three categories: unidirectional acceptance, negotiated revision, and position maintenance.

The first round of decision letter for both Paper 2007 and Paper 2021 were major revisions. Each reviewer (3 reviewers from Paper 2007, 2 reviewers from Paper 2021) made an opening remark on the paper before providing specific review comments:

We found that the three opening remarks of Paper 2007 were consistent: the new algorithm proposed in the paper were regarded as a simple extension of the classical algorithm, but the paper needed major revisions. However, the two opening remarks of Paper2021 were contradictory, reviewer 1 advocated that it was not recommended for publication in ITY while reviewer 2 acknowledged the contribution of the paper. The different opinions of the two reviewers probably explained why the editor made a major revision decision for this expert-stage paper and why the editor invited another reviewer to review the paper before final acceptance.

Next, we will focus on the detailed comments. The first round of review for paper 2007 and paper 2021 had a total of 20 and 18 specific review comments, respectively. Based on the three-dimensional analysis framework of technical depth, scholarly engagement and text organization, we generated the quantitative statistics in Table 7.

From the distribution data of two papers, respectively, the focus of reviewers shows a stable characteristic. Although the proportion of comments on technical depth increased slightly from 35% to 39% (Difference:+4%), scholarly engagement decreased slightly from 20% to 17% (Difference: −3%), and text organization maintained a dominant position (44%–45%; Difference: −1%), none of these changes were statistically significant. This is evidenced by the chi square test results (χ² = 0.096, p≈0.95) and further supported by the 95% confidence intervals for the differences in proportions, all of which contain zero (see Table 5). This overall stability reflects a consistent review framework in IEEE transactions. Among the three dimensions, the consistent dominance of text organization may be related to the fact that text organization problems are readily identifiable in a short review time, whereas assessing technical depth and scholarly engagement problems may require deeper, field-specific expertise.

Now let’s take a look at the author’s reflection on the submission of the two papers. Paper 2007 is Song’s first SCI paper. Song developed an efficient algorithm to the problem of SAR (Synthetic aperture radar) image formation based on over 1 year of efforts. With this new algorithm and simulation data, he decided to write a paper and submitted to ITX, a renowned IEEE transaction with high impact factor.

Three months after submitting, Song received the decision from the editor as “revised and resubmitted (RQ).¹ “Song’s first reaction was that the paper was rejected, which was also expected. However, after further checking, he found out that this decision was actually provisionally accepted. He was overjoyed and decided to made serious and careful revisions.

The extended response to review comments refers to the author’s spontaneous modifications to other parts of the paper based on reviewer’s comments. We found at least 4 out of 20 responses from Song revealed extended responses, of which 3 are responses of technical depth and 1 is response of scholarly engagement. The extended responses not only reflect the need to improve the paper quality, but also reflect the author’s growth of internalizing and implementing high-quality paper standards (Mulligan et al., 2013). Song used statements such as “this comment strongly reminds the authors” to inform the reviewer of any other occluded adjustments he made. For example:

It can be seen that the review comments have prompted novice author to deeply consider the requirements of international journal papers and make major adjustments to the entire manuscript, as mentioned by Song in his response to the editor:

In this round of revision, the length of the paper has been expanded from 22 double spaced pages to 34 pages. The number of figures has increased from 5 to 10, and the number of formulas has increased from 21 to 40. And references have expanded from 8 to 15. In addition to following the reviewer’s comments, two new sections have been added. One concerns the application of the new algorithm in another system, and the other illustrates the application of the new algorithm in two systems with graphs. Song believes that:

For novice authors, although they have read a large number of SCI papers, their understanding of paper quality standards is not profound without starting to write. Through their comments, reviewers translate abstract standards into concrete suggestions. This process provides novice authors with valuable insights and fosters a deeper understanding of high-quality papers.

Gradually, Song’s submitted papers were generally accepted after minor revisions, but there are exceptions, such as Paper 2021.

Different from Paper 2007, the extended responses have not been found in Paper 2021. The proposed approach in Paper 2021 was originally an idea conceived of in 2012, which has since been gradually improved and developed in practice. In 2016, the second step of the approach was added, which improved the application of the approach in different scenarios and conditions. Subsequently, experiments were conducted and life experimental data was obtained. After seeing the significant performance results, Song and his research team carried out a rigorous formula derivation and added the process of approach formation. The final paper includes proposal of the problem, a detailed introduction to the two-step new approach, life experimental data analysis, and conclusions. In Song’s words:

The construction of this paper stemmed from Song’ years of experience in submitting and revising papers. In addition, his experiences of being reviewers also helped to have a clear and deep understanding and implementation of quality standards of SCI papers.

We categorized the 38 author responses (20 from Paper 2007, 18 from Paper 2021) into three categories (unidirectional acceptance, negotiated revision and position maintenance). The analysis of author responses reveals a marked shift in strategies between the novice and expert stages (see Table 8). At novice stage, unidirectional acceptance accounted for the highest proportion, reaching 70% (14/20). Among these, all text organization responses were accepted (9/9), most scholarly engagement responses were accepted (3/4). This reflects the strong demand to improve the manuscript and a deep respect for the role of reviewers as academic judges. Conversely, negotiated revision was rare (5%, 1/20), indicating limited experiences in academic negotiation (Li, 2006), leading to a passive cycle of acceptance and revision.

In contrast, at the expert stage, a significant transformation occurred. Among the 18 responses at expert stage (7 for unidirectional acceptance, 7 for negotiated revision, and 4 for position maintenance), the proportion of unidirectional acceptance dropped markedly from 70% to 39% while negotiated revision increased markedly from 5% to 39%. The 95% confidence intervals for these differences confirm their statistical significance (see Table 8). This transformation highlights Song’s evolution into an active participant in scholarly dialogue, engaging in negotiation with reviewers. Meanwhile, the proportion of stance maintenance remains stable, from 25% to 22%, showing there was no statistically significant change.

Among the 5 maintain position responses of Paper 2007, 1 is about scholarly engagement and 4 are about technical depth. Song maintained his position in the 4 responses out of a total of 7 responses to technical depth comments, which is not common at novice stage. Let’s take a look at one of them.

This is the first comment from Reviewer 1. Obviously, the reviewer considered this to be a significant flaw of experimental design. However, Song without much experience of communicating with reviewers could not perceive this was a comment that should be revised. He maintained the position from his own perspective, explaining that the differences were not obvious and could not be distinguished in the same figure, and then explained why the differences were not obvious:

Regarding this point, we checked the review comments of two other reviewers and found that reviewer 2 had similar suggestions to add performance comparison. However, Song did not take the same suggestion raised by two reviewers. In peer review, novice authors tend to fully accept or completely reject review comments, and are less likely to adopt a consultative response. This phenomenon reflects a stage characteristic in the process of academic socialization, that is, falling into a black-or-white dichotomous response pattern (Matsuda and Tardy, 2007).

Thus, in the second round of review, Reviewer 1 acknowledged author’s efforts to paper modifications, but specifically questioned for Song not providing comparison results as required:

Obviously, the novice Song could not detect that adding results of performance results is the main concern of two reviewers and it is a comment that must be adopted. Perhaps Song only focused on revising the article and failed to carefully understand the intentions of the reviewers. It can be seen that novice authors exhibit relatively weak academic social skills in interacting with reviewers. Fortunately, both reviewers were friendly and gave Song another chance to revise. The editor’s decision was being accepted with mandatory revision.

In this round of revision, Song added results and figures of performance comparison under various conditions following reviewers’ suggestion. During our interview, Song deliberately showed us with the added figures, explaining:

When reviewing Song’s responses together with Song in this interview, Song once exclaimed: “My tone in the responses was so tough at that time.” Song admitted that he did not think too much about the risk of being rejected if he did not follow the reviewer’s comments at that time probably because he was young and inexperienced, and could not detect reviewers’ intention. Later, Song pays more attention to understanding the intentions of the reviewers and try to avoid situations where necessary revisions are not made. If his students encounter such review comments now, he will guide them to address the reviewers’ primary concern directly. It is meaningless to argue with reviewers with the risk of being rejected. Publishing papers earlier can help students defend and graduate earlier. Of course, from the perspective of the completeness of paper framework, adding performance comparison could be more comprehensive, Song eventually admitted to this during our interview.

In Paper 2021, 4 position maintenance responses were almost evenly distributed in three dimensions of technical depth (1 response), scholarly engagement (2 responses), and text organization (1 response). 4 responses to technical depth comments out of a total of 7 underwent negotiated revision, which reflected the author’s growth in academic social skills after more than 10 years of interaction with peer reviewers.

As we have mentioned earlier, the comments from Reviewer 1of Paper 2021 were unexpected for Song. After carefully reading Reviewer 1’s overview of the new algorithm, Song found that Reviewer 1 probably did not understand this paper very well. By reading detailed comments raised by Reviewer 1, Song became more convinced of his finding. For example, one of Review 1’s comments is about a term used in the paper. The reviewer commented that there is no need to make up a term and the new term is confusing.

In Song’s view, the major revision this time is to add clear explanations to the paper for audience not in this field to understand, as well as to respond seriously and carefully to the editor and Reviewer 1 in order to pass the second round of review.

Therefore, although Reviewer 1 of Paper 2021 did not fully agree with its publication, the author’s (Song’s) 18 responses were predominantly conciliatory: only 4 were maintaining position, the other 14 are negotiated revisions or unidirectional acceptances. Let’s take a look at two of negotiated responses.

For reviewer 1’s question about value of this paper, Song’s responses to the editor and Reviewer 1 are as follows:

In the above responses to comment 1, we can see that Song skillfully pointed out that the reason why Reviewer 1 does not agree with the contribution of this paper is probably the incomplete understanding: “The two convolution operations as described by the reviewer are portions of our paper.”

For Reviewer 1’s previous comment that Song does not have to make up a word, Song’s response is as follows:

This response is even more considerate, first stating that the reviewer did not know this term because the author did not explain it clearly, and then defending the author’s position of the existing term. From the above responses, we can see how an experienced author responds to the reviewer’s comments with his academic interaction competence. The review result of the second round was both expected and unexpected. What is expected is that Reviewer 1 acknowledged the author’s revisions and agreed to publish the paper. What is unexpected is that Reviewer 1 provided no other comments except the following one:

It can be seen that Song’s careful revision and skillful responses have gained the recognition of Reviewer 1. In the second round of revision, Song mainly addressed some small issues such as formula derivation descriptions raised by Reviewer 2. Two months later, Song received an acceptance notice. But Song unexpectedly found that a Reviewer 3 had been added to review the paper. Obviously, the editor intended to hear one more voice. Here are the comments from three reviewers of this round:

Thus, Paper 2021, which accumulated Song’s 9 years of research achievements, was finally published. When asked if Song was worried that Reviewer 1 would stick to his original stance, Song replied:

In 2005, Song had a need for promotion to professor. Although there were not strict requirement of publishing SCI papers for professional title promotion at that time, Song had an interest in giving it a try. In Song’s field, high-ranking SCI journals are IEEE transactions and letters, among which IEEE transactions with high impact factors are more influential. Song explained the reason for choosing ITX as his target journal:

As high-impact journals are frequently read and cited by peers, the publication of Paper 2007 brought academic reputation to Song and many peers got to know him through this paper. We can see that his position on the periphery of academic circle motivated his publication. His familiarities with the topics discussed at the center enabled him to find a voice, while the reviewers’ feedback solidified his understanding of what constitutes a high-quality paper. With these, Song started his international academic participation and gradually published many papers in IEEE transactions and letters on new progress of his scientific research. Soon with these SCI papers, Song was promoted to professor and gained reputation in this circle.

The proposed algorithm in Paper 2007 is a small trick and was tested on computer simulations. The authors of the paper are only Song and his supervisor. However, Paper 2021 targets a relatively large and unresolved issue. More peers are involved in this engineering project from forming a two-step algorithm after repeated experimentation, to processing data after flight tests with collaboration with research institutes. Paper 2021 has seven authors including Song, young faculty members in his research team and his doctoral and master’s students. We can see that Song’s academic strength has been continuously promoted from following his supervisor to building his own research team and collaborating with research institutes for testing. In Song’s words:

It can be seen that although international publication has earned Song reputation, with the enhancement of Song’s academic strength, collaborating with research institutes to complete practical engineering projects has become the next major goal. This is because:

So the paper publishing circle is incomplete, at least in the field of engineering. International publishing help boost one’s reputation in a circle only involving those who have the tradition to publish or have the demand to publish. Such international participation is meaningful but also limited to a certain group of peers. Therefore, it is more appropriate to say that international publishing and peer interaction helps one move from periphery to center of a paper publishing circle, not the academic circle. Due to the incomplete academic circle, researchers like Song will shift their focus to completing engineering projects after realizing their promotion of professional titles. Occasionally, they write papers to record and share their latest research findings.