The Rhetorical Structure of Literature Reviews in Egyptian- Authored English Research Articles in Linguistics

ing or synthesizing knowledge claims to establish a theoretical position or a theoretical framework ">S2.3< This brief survey of the literature on digital game‐ based FL learning demonstrates that digital games, especially simulation video games, hold great potential for FL practice and learning, >Sub-S2.1A< but additional research is needed to tap into...." (Int3)

has been countered by many researchers who have depicted the LR positively, stressing its important role in justifying the significance of one's research and highlighting the researcher's highly critical skills. This in return paves the way for publishing in high-impact international journals (Kwan, 2006).
A third motive is that the LR is considered as one of the most complex difficult sections/chapters researchers have to write (Huang & Zhang, 2016;Inoue-Smith, 2020;Swales & Lindemann, 2002); it "is rhetorically a highly demanding part-genre that generally presents greater rhetorical problems than methods and results sections" (Swales & Lindemann, 2002, p. 117). In addition, it is considered as one of the most criticized sections by reviewers during the process of reviewing manuscripts for publication (Flowerdew, 2001;Tewksbury & Mustaine, 2012).
Because the rhetorical structure of the LR cannot be expressed in terms of the rhetorical structures of the other sections as stated by Swales and Lindemann (2002), discussing it as part of any of these sections-even the introduction-is deficient and misleading and thus this necessitates analyzing it separately, a view that was confirmed by Kwan (2006) after comparing the rhetorical structure of a corpus of LR chapters to Bunton's (2002) revised CARS (Create a Research Space) model for analyzing thesis introductions.
Very few studies have been conducted on analyzing the LR (Chen, 2008, as cited in Jian, 2010Hsiao & Yu, 2012;Jian, 2010;Kwan, 2006), so more studies are needed to analyze RA LRs to achieve a better understanding of their rhetorical structure. In addition, no study has analyzed Egyptian-authored RA LRs in the field of linguistics. In order to fill-in these gaps, the present study seeks to investigate and compare the generic structure of a sample of Egyptian-authored linguistics RA LRs published in local and international English-medium journals.

Genre Analysis
In their seminal books, Swales (1990) and Bhatia (1993) define genre as communicative events featured by a number of communicative purposes approved and adopted by its community members. Genre analysis is a means of studying the formal and functional features of the language of the genre in an attempt to understand the communicative purposes it conveys (Bhatia, 2012). It enables inexperienced writers to acquire expertise and thus be more skillful and creative, producing high-quality professional writing (Bhatia, 1997;Loi & Evans, 2010). Genre analysis can be conducted at either a macro or a micro level (Bunton, 2002). The macro level investigates the rhetorical structure of a text (moves analysis), while the micro level deals with the linguistic features (lexical and syntactic analysis). Move analysis, the focus of the present study, depends on dividing the text into segments (moves) where each segment expresses an idea/a theme, reflecting a particular communicative function and contributing to the holistic communicative function of the genre as a whole (Parodi, 2010). These segments are referred to as "the building blocks of texts" (Biber, Connor, & Upton, 2007, p.9). Move analysis was first used by Swales (1990) to analyze RA introductions. Swales' pioneering work has inspired much research to investigate RA different sections including abstracts (Amnuai, 2019;Friginal & Mustafam, 2017;Tankó, 2017), introductions (Lu, Casal, & Liu, 2020;Lu, Yoon, & Kisselev, 2021;Sheldon, 2011), methods (Cotos, Huffman, & Link, 2017;Farnia & Baratizade, 2020;Peacock, 2011), results (Basturkmen, 2009;Kurniawan & Lubis, 2020;Lim, 2011), discussions (Amnuai & Wannaruk, 2013;Basturkmen, 2012;Liu & Buckingham, 2018), and conclusions (Jahangard, Rajabi, & Khalaji, 2014;Yang & Allison, 2003;Zamani & Ebadi, 2016). However, the RA LR is still an under-researched area.

The LR Section
Reviewing the literature involves critiquing the literature and integrating the previous findings in a convincing argument that sheds light on the research gap and thus justifies the significance of the present research; it is more than a mere summary of the literature (Inoue-Smith, 2020;Rewhorn, 2018). This process requires deploying higher cognitive skills to collect, classify, summarize, analyze, relate, evaluate, and synthesize the literature (Rewhorn, 2018). The LR is also referred to as an indicator of one's academic proficiency and grasp of the literature on the topic under investigation (Huang & Zhang, 2016). The LR can be a sub-genre when it is a part of another research such as a section/chapter in an RA/a thesis, and it can be a genre when it represents an independent piece of writing on its own such as a review article or a book review.
The LR is not only one of the most important sections to be written in an RA and in academic writing in general (Denney & Tewksbury, 2013;Feak & Swales, 2009;Inoue-Smith, 2020), but it is also one of the most challenging tasks (Badenhorst, 2018;Inoue-Smith, 2020). Although there are several textbooks and RAs that have discussed its different aspects (e.g., meaning and purpose), developed taxonomies for classifying it (Cooper, 1985), designed rubrics for evaluating it (Boote & Beile, 2005), and divided it into types (Feak & Swales, 2009;Fink, 2014), the LR has been rarely discussed from genre analysis perspective because it was not dealt with as a separate section (Hsiao & Yu, 2012;Kwan, 2006;Tseng, 2018).

The Rhetorical Structure of the LR
In 2006, Kwan postulated a three-move model for analyzing the rhetorical structure of PhD LR chapters. She found that Moves 1 and 2 outnumbered Move 3 and that Strategies 1A, 2A, and 3A were the most used strategies in their corresponding moves, but none of them was obligatory.  (Kwan, 2006, p. 51) Other models were developed on the basis of Kwan's (2006) model for investigating the LR section/chapter. For instance, Jian (2010) analyzed the LR section in 40 English RAs and 40 Chinese RAs, using Swales' (1990), Kwan's (2006), and Chen's (2008, as cited in Jian, 2010  . Agreeing with Kwan, the last move was used the least. In addition, the sub-moves in the first and last moves (i.e., surveying knowledge claims and announcing research aims) were used the most within their corresponding moves.
Among the four models, the current study adopted Kwan's model because it is more detailed and comprehensive than the other models and, thus, it best reflects the communicative functions of the LR. It is also more validated and applicable than the other models; it has been used in almost all the studies that analyzed the LR sections/chapters in both soft and hard disciplines (Chen & Kuo, 2012;Gil-Salom & Soler-Monreal, 2014;Maulid, 2017;Soler-Monreal, 2015;Thompson, 2009;Xie, 2017). Chen & Kuo (2012), for example, analyzed rhetorically 20 LR chapters in applied linguistics and found that Strategies 1A and 1C were used the most. Gil-Salom and Soler-Monreal (2014) also compared English and Spanish LR chapters.
The findings revealed no significant structural differences between the two corpora. However, the Spanish LRs involved fewer strategies in Moves 2 and 3 than the English ones. Unlike Kwan's (2006) findings, the frequency occurrences of Strategies 1A and 1C were equal in the English corpus and almost equal in the Spanish one. Agreeing with Kwan, move 3 was the least used move in both corpora.
Xie (2017) also followed Kwan's (2006) (2017) analyzed the LR chapters of four master's theses in the field of English as a foreign language (EFL) and found that strategy 1A had the highest frequency occurrence.

The Current Study
Despite its significance, there is a dearth of research on the LR section compared to the abundant studies conducted on the other sections of the RA. In addition, no study, to the researchers' best knowledge, has analyzed the rhetorical structure of Egyptian-authored RA LRs. The present study, thus, seeks to address these gaps and answer the following research question: ▪ How are rhetorical moves/steps used in the LRs of Egyptian-authored ILrMRD linguistics RAs published in local and international English-medium journals?
To answer this question, a sample of ten Egyptian-authored linguistics RA LRs published in local and international English-medium journals between 2013 and 2019 were analyzed to identify the rhetorical variation between both sub-corpora, deploying a combination of genre-based and corpus-driven approaches.

Research Design
This study is considered exploratory because there are very few studies that have discussed RA LRs in general and no study has discussed Egyptian-authored linguistics RA LRs in particular. The present study used mixed sampling techniques. It used a purposive sampling technique to collect the sub-corpus of Egyptian-authored RA LRs published in international journals and a simple random sampling technique to collect the sub-corpus of Egyptian-authored RA LRs published in local journals. The study used genre-based and corpus-driven approaches to analyze the LRs. According to McEnery and Hardie (2012), a corpus-driven approach can be used to analyze data, employing previous models; however, the researcher is free to add any modifications to these models in case they seem inefficient or do not account for the findings appropriately.
In line with this view, a corpus-driven approach was used in this study to compare both sub-corpora and explore any new moves/steps that better reflect the rhetorical structure of the LRs under investigation without being restricted to Kwan's model. The identified moves/steps were analyzed qualitatively and quantitatively.

Data Collection
A sample of 10 Egyptian-authored linguistics RA LRs were taken from five top-rated local English-medium journals according to the Supreme Council of Universities and five international English-medium journals registered in Journal Citation Reports (2018) in Clarivate Analytics between 2013 and 2019, comprising a total of 16,326 words (see appendix A for the lists of the RAs from which the LRs were extracted). The corpus small size is due to the fact that the available Egyptian-authored international RAs (one sub-corpus) are few in their original population (international RAs) and this limitation restricts the size of the other subcorpus to get parallel sub-corpora. However, the corpus small size "is not especially problematic" because it is very specialized (Baker, 2010, p. 111), and it still meets Koester's (2010, p. 70) criteria "that any subcorpus within the corpus … is represented by at least 1,000 words …, and that every sub-corpus contains at least five … different samples."

Data Analysis
The RAs were converted from PDF files into plain text files by AntFileConverter (Anthony, 2017), extracting the LR sections and coding them as Loc1-Loc5 for the local sub-corpus and Int1-Int5 for the international sub-corpus. Data analysis was conducted in two stages. In the first stage, the data was analyzed qualitatively. The moves/steps were identified according to their rhetorical functions, adopting a data-driven approach in light of Kwan's model and modifying or adding any new steps that did not fit the Kwan (2006) model. A rhetorical function can range from one clause to a cluster of sentences, while a sentence can express more than one communicative function. This implies that the basic unit of identifying moves and steps is the segment and that the moves/steps are not subject to the grammatical boundaries of sentences and paragraphs, i.e., a move/step can be realized in one clause or across a cluster of sentences (Lewin et al., 2001). The moves and steps were manually coded (e.g., <M1>, <S1.1>) in the plain text files. The first stage ends with postulating a data-driven model that efficiently reflects the rhetorical structure of the Egyptian-authored RA LRs in linguistics.
In the second stage, the data was analyzed quantitatively. The instances of moves'/steps' occurrences were counted to determine the major step realizations of each move across both sub-corpora. It is noteworthy that consecutive segments belonging to the same step type were counted as one instance of the step. The frequency occurrences of the moves/steps were retrieved within each text and across all texts via Notepad++'s (Ho, 2020) "find in files" function, then their percentages were computed and compared across both sub-corpora. The current study followed Soler-Monreal, Carbonell-Olivares, & Gil-Salom's (2011) and Pho's (2013) criteria that a move/step is considered obligatory if it appears in equal to/more than 90% of the LRs, prototypical if it occurs in equal to/more than 60% and less than 90%, and optional if it is identified in less than 60% of the LRs.
To ensure high inter-rater reliability, a sample of four texts were randomly selected and double coded by an external rater. Both codings were compared and any discrepancies were discussed till an agreement was reached. Finally, the rest of the texts (six LRs) were coded by the current researchers.

Move Analysis of The LRs
Using A Data-Driven Approach. The moves and their subsequent steps were identified according to their rhetorical functions, adopting a data-driven approach in light of Kwan's (2006) model. Based on data analysis, there was no change to Kwan's three moves (Establishing the territory of one's own research, creating a research niche, and Occupying the research niche). At the step level, some of Kwan's strategies (relabeled as steps) were kept without change, some were modified, and some new steps were added to better describe some of the segments in the data. The section is concluded with the new datadriven model. The rhetorical functions of each move and its subsequent steps (original, modified, and new) will be illustrated as follows (see Appendix B for examples from both sub-corpora to reflect the moves and their subsequent steps/sub-steps):

.1 Approaching A Different (Sub)Theme and Summarizing A Group of Relevant Previous Studies.
Given that the LR is replete with instances of previous studies, this step was added to Kwan's model to organize the way of displaying these studies. Rather than listing them, the writer finds relations among the surveyed studies and categorize them into groups according to the various (sub)themes they discuss to keep the reader focused, especially if the LR involves many themes and sub-themes. This step is usually followed by discussing individual studies relevant to each (sub)theme in more detail.

Evaluating Previous
Studies. This step expresses the writer's stance through evaluating previous research, but this type of evaluation is different from the evaluation involved in the forthcoming sub-steps of justifying the research negatively or positively. This step does not imply using the cited paper in one's own study. The evaluation may be negative or positive.
The two latter steps require using higher cognitive skills, i.e., the ability to organize, classify, analyze, and critique. These two steps show that the writers do not suffice to use previous claims as they are; rather, they try to recycle them in a way that serves the sophisticated functions of the LR, reflecting its critical nature.

4.1.1.2.3
Outlining the Structure of the LR. This step refers to the themes and sub-themes to be discussed within the LR. It guides the reader through the reading process. This step may occur once at the beginning of the section or many times before the different themes.

Creating A Research
Niche. Compared to Move 1, the nature of Move 2 indicates viewing the LR as more than a summary of the literature. It involves a more sophisticated phase, and its writing requires using higher cognitive skills of analysis, critique, and synthesis to establish a series of relations to introduce one's own research. It is noteworthy that introducing one's own research as a reaction to the calls of the literature leads to a better perception and appreciation of its contribution. According to the researchers, this move is considered as the most important one because it represents a bridge between the previous findings and the expected ones.  (2002), is either negative or positive. The researchers find that Kwan's Strategy 2A (counter-claiming) should be embedded within Strategy 2B (gap-indicating) because the concept of a research gap in its broader sense is not restricted only to the lack of knowledge and rarity of the literature on a particular issue (an overlooked area). The gap may be a thorny issue that needs to be tackled, contradicting findings that need to be either supported or refuted (Kwan's counter-claiming), a point that needs further interpretation, a particular domain that is not fully discussed and thus needs to be covered, an issue that needs to be tackled from a new perspective, a variable that is less estimated and its effect needs to be measured, a problem that needs to be resolved, a new method that needs to be employed, a sample that needs to be more representative, new findings that need to be validated by further research, or a recommendation that needs to be taken into consideration. Many researchers and linguists discussed a similar point of view. Lewin et al. (2001, p. 44), for instance, classified a gap into "gaps of defect, scarcity, and obscurity." The first type denotes criticizing the previous literature by expressing its deficiencies and limitations. It is similar to Swales' (1990) and Kwan's (2006) counter-claiming step/strategy. However, the latter two types are used to establish the gap without explicit criticism. A similar meaning was also expressed by Samraj (2002, pp. 8-9), who explained that a gap can be realized in terms of "the contradictions among the findings from earlier studies Research contribution/significance was added to the step because a research paper's contribution is logically one of its objectives. In many cases, this step is preceded by the Sub-step of providing negative justification to highlight the research contribution.

Positing Theoretical Positions/Theoretical Frameworks.
In this step, the researcher announces the theoretical framework s/he is going to adopt in her/his study.

Presenting Research Design/Processes.
In this step, the researcher briefly mentions the research design and the methodological procedures s/he is going to deploy in her/his research.

Providing Interpretations of The Terminology Used in The Current
Research. This step involves discussing the terms of the study.

New
Step Added to Kwan's Model.

Evaluating the Present
Research. This step evaluates the expected strengths and/or weaknesses of one's own research.

The Data-Driven Model.
Using the three moves in writing the LR achieves the targets of reviewing the literature, understanding any complicated relations, being aware of any involved drawbacks, introducing the new research, and highlighting its significance without being lost. In such a sense, the LR is considered as a journey to understand a new world of interwoven definitions, theories, complicated relations, debates, and findings.

Move 1 Establishing one part of the territory of one's own research by
Step 1.1 Surveying the non-research-related phenomena or knowledge claims Step 1.2 Claiming centrality

Step 1.3 Approaching a different (sub)theme and summarizing a group of relevant previous studies
Step 1.4 Surveying the research-related phenomena Step 1.5 Evaluating previous studies Step 1.6 Outlining the structure of the LR

Move 2 Creating a research niche (in response to Move 1) by
Step 2 Step 2.2 Asserting the relevancy of the surveyed claims to one's own research Step 2.3 Abstracting or synthesizing knowledge claims to establish a theoretical position or a theoretical framework

Move 3 Occupying the research niche by
Step 3.1 Highlighting research aims, focuses, research questions/hypotheses, or contribution/significance Step 3.2 Positing theoretical positions/theoretical frameworks

Step 3.3 Presenting research design/processes
Step 3.4 Evaluating the present research Step 3.5 Providing interpretations of the terminology used in the current research Modifications and additions are in italics 4.2 Quantitative Genre Analysis. The distribution of moves was identified and compared across both subcorpora. Their frequency counts indicate the most used moves in each sub-corpus. Their percentages were used to classify them as obligatory, prototypical, or optional. Tables 3 and 4 show that the moves' frequency occurrence in the international LRs was more than two times and half the number found in the local LRs (147 vs. 57). Agreeing with Kwan's (2006) findings, Move 1 was the most frequent move, followed by Moves 2 and 3 in both sub-corpora. The frequency occurrence of Move 1 in the international sample doubled its frequency occurrence in the local sub-corpus. However, these findings can be attributed to the different sizes of both sub-corpora. Moves 2 and 3 occurred in the international LRs much more than in the local LRs. Although Kwan's three moves occurred in all the international LRs, move 2 did not occur in one local LR and Move 3 did not occur in two local LRs. The variation between both groups shows that the writers of the international LRs are more aware of the LR rhetorical knowledge and writing conventions than their counterparts in the local sub-corpus. This also reflects how the writers of the international sub-corpus view the LR as a process of critique and interpretation, not as a summary of the previous findings. In addition, the frequency occurrence of the three moves in the local LRs shows that Move 1 was the most favored move; it occurred as three times as the frequency counts of Moves 2 and 3. This indicates how the writers of the local LRs view the LR as a means to display their grasp of the literature relevant to the issue under investigation. In addition, their minimal use of Moves 2 and 3 may reflect the main obstacle that hinders them from publishing internationally. This obstacle prevents them from highlighting the research gap within the literature (Move 2) and clarifying their own contribution to the literature (Move 3). Finally, the low ratio of Move 3 compared to the first two moves in both sub-corpora can be attributed to its possible occurrence in other sections of the RA such as the introduction, results, discussion, or conclusion. Unlike Kwan's findings, the three moves were obligatory in the international sub-corpus. In contrast, only Move 1 was obligatory and Moves 2 and 3 were prototypical in the local sub-corpus.

4.2.2
Step Distribution Across Both Sub-Corpora. Table 5 shows a total of 80 occurrences of Move 1 steps in the international sub-corpus and 44 occurrences in the local sub-corpus. Step 1.2 9 10 4 4

Move 1 Steps.
Step 1.3 3 7 2 3 Step 1.4 8 26 3 5 Step 1.5 2 15 1 3 Step 1.6 5 3 2 1 Total 44 80 Table 5 reveals that Step 1.4 was the most used step in the international sample, and Step 1.1 was the most used step in the local sample. This indicates that Move 1 is mainly realized through these two steps. They were followed by Steps 1.5, 1.2, and 1.3 in the international dataset. Steps 1.3 and 1.5 represent the least frequent steps in the local sub-corpus; they are even less than Step 1.6, which represents the least frequent step in the international dataset.

Move 2 Steps.
A total of 42 occurrences of Move 2 steps were identified in the international subcorpus, and 8 occurrences were identified in the local sub-corpus as shown in Table 6. Step 2.1A 2 20 2 5 Step 2.1B 4 7 3 3 Step 2.2 -6 -3 Step 2.3 2 9 1 4 Total 8 42 Table 6 reveals that Sub-step 2.1A occurred the most in the international dataset, followed by Step 2.3 and Sub-step 2.1B, which represented the most frequent sub-step in the local sub-corpus. This reflects the importance of Step 2.1 (justifying one's research) in general since it represents the core of the whole research, showing how one's research contributes to the literature. It represents the key instigator beyond selecting a particular issue for investigation and the main motivation for the research community to acknowledge and accept one's research.
Step 2.1 is considered as the most common realization of Move 2.
All Move 2 steps occurred in the international sub-corpus five times the number found in the local one, reflecting the local writers' lack of training and awareness to use the skills and conventions required to express Move 2. Findings indicate that the Egyptian writers who published internationally are aware of the sophisticated nature of the LR as a process which involves critiquing the literature to highlight the value and originality of one's own research, establishing relations between the previous findings and the research under investigation, and collecting pieces of the literature to adopt a new perspective or develop an argument that constructs the basis for one's own research.
Step 2.1 was classified as obligatory in the international sub-corpus and as prototypical in the local one. Steps 2.2 and 2.3 were prototypical in the international sub-corpus, but they were optional in the local sub-corpus.

Move 3 Steps.
A total of 25 occurrences of Move 3 steps were identified in the international, and there were 5 occurrences in the local sub-corpus. Step 3.1 3 14 2 5 Step 3.2 2 3 2 3 Step 3.3 -6 -5 Step 3.4 -1 -1 Step 3.5 -1 -1 Total 5 25 Table 7 reveals that Step 3.1 was the most favored step across both sub-corpora. Its frequency counts in the international subset exceeded the frequency counts of the other four steps. This indicates its importance and makes it the strongest realization of Move 3. It was followed by Steps 3.3, 3.2, 3.4, and 3.5 in the international LRs. In general, move 3 steps were rarely used in the local subset.
There were two obligatory Steps (3.1 & 3.3), one prototypical Step (3.2), and two optional Steps (3.4 & 3.5) in the international sub-corpus. As for the local sub-corpus, there were only two optional steps (3.1 & 3.2) and the other steps did not show up.

Discussion
Agreeing with the three major studies on the LR (Hsiao & Yu, 2012;Jian, 2010;Kwan, 2006), Moves 1 and 2 were the most frequent moves and Move 3 occurred the least. In line with Chen and Kuo's (2012) findings, Steps 1.1 (Surveying the non-research-related phenomena or knowledge claims), 1.4 (Surveying the researchrelated phenomena), 2.1 (Justifying one's research), and 3.1 (Highlighting research aims, focuses, research questions/hypotheses, or contribution/significance) were the most frequent steps within their corresponding moves. These findings reflect the importance of the first two moves compared to Move 3, where Step 3.1 is its strongest realization.
Unlike the three major studies, Step 1.4 occurred the most in Move 1 in the international LRs. However, Step 1.1 was the most common step in the local sub-corpus, agreeing with most previous findings (Hsiao & Yu, 2012;Jian, 2010;Kwan, 2006;Maulid, 2017;Xie, 2017). Agreeing with Kwan's findings, Step 2.1 represented the most frequent step within its corresponding move in both sub-corpora. However, Sub-step 2.1A (providing negative justification) was the most used one in the international subset and Sub-step 2.1B (providing positive justification) was the most used one in the local subset. This indicates that the Egyptian writers who published internationally were more likely to critique the literature and express its weaknesses than their counterparts who published locally. In accordance with several previous studies (Jomaa & Bidin, 2017, Kafes, 2017Xie, 2017;Yaǧiz, Ӧtügen, Kaya, & Aydin, 2014), the locally published writers'-as well as their novice and non-native counterparts'-avoidance of expressing stance and critiquing the literature may be attributed to cultural influences and their lack of academic expertise and linguistic competence. This avoidance of confrontation weakens the value and contributions of their studies and thus affects their credibility. In contrast, the frequency counts of Step 2.3 (Abstracting or synthesizing knowledge claims to establish a theoretical position or a theoretical framework) were inconsistent with the frequency counts in Kwan's (2006) study. While it was the second common step in its corresponding move in both corpora in the current study, it was the least used strategy in Kwan's sample.
The present findings are also in accordance with Hsiao and Yu's (2015) and Hsiao's (2016) findings that the internationally published writers used all the moves and used the Sub-moves of 2A (Grouping and drawing ideas from source materials to gain a new perspective), 3B (Indicating a gap), and 4A (Announcing the aim of the research) much more than their local counterparts. The locally published LRs are also similar to the Spanish ones in Gil-Salom and Soler-Monreal's (2014) study since both involved fewer strategies in Moves 2 and 3 than the internationally published and English LRs, respectively.
Another major difference between both sub-corpora is the moves' classification as obligatory, prototypical, or optional. Unlike Kwan's model where none of the moves was obligatory, the three moves were obligatory in the international sub-corpus. The findings partially agree with Soler-Monreal's (2015) findings where the three moves were obligatory in the English LRs written by English and Spanish writers, while Moves 1 and 2 were obligatory and Move 3 was optional in the Spanish LRs. Aligned mostly with Xie's findings, Move 1 was obligatory and Move 3 was prototypical in the local LRs. However, Move 2 was obligatory in Xie's study and was prototypical in the current study.

Conclusion
The current study was after digging deep into the reasons why particular Egyptian-authored LRs are published internationally, while others fail to meet the criteria. From a genre perspective, the study analyzed and compared the rhetorical structure of a sample of 10 Egyptian-authored linguistics RA LRs published locally and internationally, adopting genre-based and corpus-driven approaches. Based on quantitative and qualitative analyses, the findings revealed that Move 1 (Establishing one part of the territory of one's own research), Move 2 (Creating a research niche), and Move 3 (Occupying the research niche) were obligatory moves in the international LRs. In contrast, Move 1 was the only obligatory move in the local LRs, while Moves 2 and 3 were prototypical moves. In addition, the writers in the international LRs used more steps than their counterparts in the local LRs, especially Move 2 and Move 3 steps. While all the steps of the three moves occurred in the international sub-corpus, one out of Move 2 three steps (Step 2.2) and three out of Move 3 five steps (Steps 3.3, 3.4, & 3.5) did not show up in the local sub-corpus and the other two steps of Move 3 occurred in only two local LRs. These findings indicate that the writers of the international LRs are more aware of genre knowledge than their counterparts of the local LRs (i.e., identifying the LR's moves and subsequent steps and reflecting their rhetorical functions). It can be argued that their development of genre knowledge is because they were exposed to another type of academic instruction that pays more attention to genre knowledge in academic writing and guides learners on using appropriate rhetorical features.
The findings suggest many pedagogical implications for novice and non-native researchers and EAP course designers and instructors. The variation in both sub-corpora shows the need to improve the Egyptian researchers' mastery of genre knowledge to avoid the drawbacks in the local RA LRs. In addition, investigating English writing by non-native researchers in writing and publishing RAs enriches the context of English as a lingua franca (ELF) that is oriented towards a hybrid research community that involves writers and readers of different languages and cultures (Wu, Mauranen, & Lei, 2020). Finally, the modified Kwan's model for analyzing the rhetorical structure of RA LRs is considered as the main contribution of the current study. It contributes to a better understanding of the genre structure of effective RA LRs. Egyptian researchers must learn to apply the three moves which are obligatory in the international RA LRs.
Given the limited corpus size and its specialization in one section and one discipline written by only Egyptian writers, further research on a larger corpus of complete RAs written by Egyptian and international writers across different disciplines is needed to enrich the researchers' genre knowledge. It is recommended to investigate RA LRs at the micro level through investigating their linguistic features. Finally, better understanding of the differences between both sub-corpora can be achieved and more assistance can be offered to the Egyptian writers by holding interviews with the writers in both sub-corpora to identify their preferences in using particular features and to highlight the challenges they face when writing their RA LRs and other sections. Additional interviews with editors and reviewers in both sub-corpora can reveal the most recurrent pitfalls committed by writers when submitting their manuscripts for publication. These interviews can also identify the areas of concern and how the acceptance/rejection criteria of anglophone reviewers deviate from those of their non-anglophone counterparts.

Appendix B Examples on The Steps of The Three Moves from Both Sub-Corpora
Moves Examples from both sub-corpora

Move 1
Establishing one part of the territory of one's own research Step 1.1 Surveying the non-research related phenomena or knowledge claims "The DCT was first developed by Blum-Kulka (1982) and it typically consists of … There are two types of role plays that have been identified in the literature: …." (Int4) Step 1.2 Claiming centrality "Since 9/11 many television shows, programs, and films tackled the issue of terrorism, … Since September 11, 2001, much of American and world politics, governments and media have focused on the issues of Muslim terrorism and extremism, especially in the Middle East region." (Loc5) Step 1.3 Approaching a different (sub)theme and summarizing a group of relevant previous studies "A number of empirical studies have found a strong relationship between digital gaming and FL learning (Jensen, 2017;Sundqvist & Wikström, 2015;Sylvén & Sundqvist, 2012)." (Int3) Step 1.4

Surveying the research-related Phenomena
"Abdel-Raheem (2016) adopts Fauconnier and Turner's (1998) conceptual blending theory to analyze.… In another study, Coulson (2005) applies Fauconnier and Turner's (2002) conceptual integration on a radio call-in show…." (Loc4) Step 1.5 Evaluating previous studies "The classic study of Saragi, Nation, and Meister (1978) used the novel A Clockwork Orange (1962) by Anthony Burgess. It was of particular interest because it included Russian slang words, referred to as Nadsat, which served as the targets in the authors' reading experiment." (Int2) Step 1.6 Outlining the structure of the LR "This section describes the forms and functions of humor and presents the traditional theories of humor as well as the multimodal discourse approach." (Loc2)

Sub-step 2.1A
Providing negative justification ">Sub-S2.1A < The very limited existing literature on definiteness in Arabic reports errors in the definite article as part of surveying common errors in writing in general, with no systematic investigations that tease apart the underlying reasons for these errors ….." (Int1)

Sub-step 2.1B
Providing positive justification "Godfroid and colleagues (2013) extended the eye-movement investigation to longer, more coherent paragraphs. Twentyeight advanced EFL learners read 12 short paragraphs with target words…. This study is innovative in L2 vocabulary research by bringing attention to a robust method of quantifying the cognitive processes underlying incidental learning." (Int2) Step 2.2 Asserting the relevancy of the surveyed claims to one's own research ">S2.2< Based on this review of the literature, two important observations relevant to the purpose of the present study have surfaced." (Int1) Step 2.3 Abstracting or synthesizing knowledge claims to establish a theoretical position or a theoretical framework ">S2.3< This brief survey of the literature on digital gamebased FL learning demonstrates that digital games, especially simulation video games, hold great potential for FL practice and learning, >Sub-S2.1A< but additional research is needed to tap into…." (Int3)

Occupying the research niche
Step 3.1 Highlighting research aims, focuses, research questions/hypotheses, or contribution/significance "The present study contributes to the literature in two ways. First, it seeks to establish whether persistent L1 transfer …." (Int1) Step 3.2 Positing theoretical positions/theoretical frameworks "In an attempt to fill in a certain gap in previous studies, this paper addresses …, adopting Fauconnier and Turner's (2002) framework of conceptual blending and Grady's (2005) specifying features of metaphoric blends." (Loc4) Step 3.3

Presenting research design/processes
"The present study also examined the same issues but instead of analyzing data consisting of single-turn responses, it analyzed interactional, multi-turn data, elicited by the role play method." (Int4) Step 3.4 Evaluating the present research "The proposed system successfully handles single error words but restricts the handling of multiple error diagnosis. It handles multiple errors that fall under the same lexical error class." (Int5) Step 3.5 Providing interpretations of the terminology used in the current research "This article uses the terms gameplay, digital gaming, and player-game interaction interchangeably to refer to participants' engagement in stand-alone FL simulation video gaming." (Int3)