Another look at the rhetorical moves and linguistic realizations in international and Indonesian journal articles: A case of tourism research article abstracts

Although a significant number of studies have been conducted to explore the rhetorical structures of research article abstracts, there is a paucity of research specifically comparing the move patterns, and linguistic features of tourism research article abstracts published in international and national journals. Such a comparison is quintessential to address a notion that journal indexation may factor into the quality of textual organization in abstract writing. Employing Hyland’s (2000) analytical framework, the paper analyzed 120 tourism research article abstracts from international journals indexed by Scopus and Indonesian journals indexed by Sinta. Findings revealed more similarities than differences across the two corpora. All of Hyland’s five moves were generally found in the abstracts, with M2 (Purpose), M3 (Method), and M4 (Product) as the most occurring moves in both data sets. An exception was found in M1 (Introduction) and M5 (Conclusion), where M1 was favored and M5 was excluded in Sinta-indexed abstracts, yet the reverse was true of Scopus-indexed counterparts. In terms of the linguistic features, present tense and active voice were evidently dominant across both data groups, with a notable exception in Method move, where past tense and passive voice were more favorable. These findings appear to suggest that journal indexation does not profoundly influence abstract writing. Recommendations and implications for academic writing for publication purposes are also discussed.


INTRODUCTION
One aspect that plays a vital role in a scientific publication is the abstract. It is the component of a scientific paper on the basis of which the readers will decide whether the paper is worth reading (Ghasempour & Farnia, 2017;Hyland, 2000). Also, it is a crucial part of a research article that serves as the basis for journal editors in the initial screening whether to accept the paper for further review or reject (Kurniawan, Lubis, Suherdi, & Danuwijaya, 2019;Pratiwi & Kurniawan, 2021). Abstracts crucially determine whether the readers would find the research paper relevant or not with their research (Kafes, 2012;Fauzan, Lubis, & Kurniawan, 2020). The mounting significance of abstracts can be attributed to two aspects: first, abstracts represent the entire content of the article from the background to conclusions in brief (Samraj, 2002;Swales, 1990); and second, the increased use of online scholastic web indexes paves the way for online research reports being accessible to a global academic world through the public access of abstracts (Tocalo, 2021).
Move analysis, originated from English for Specific Purposes (E.S.P.) in the 1980s by Swales, is commonly used to identify and analyze various genres such as research articles, theses, and essays. As the demand for providing good models of academic and scientific texts keeps increasing for publication, the subject (move analysis) has been getting a considerable amount of attention for the past few years (Marefat & Mohammadzadeh, 2013). Therefore, move analysis could be said as one proper tool to discover text structure in academic texts in various disciplines, including research article abstracts.
Firstly proposed by Swales in the 1980s, move analysis is a textual analysis approach derived from the genre-based approach. Move analysis examines the generic structure of writing genres, and it originated from the development of a teaching tool for nonnative speakers (Swales, 2004;Vathanalaoha & Tangkiengsirisin, 2018). Move analysis consists of two elements: moves and steps. A 'move' refers to an aspect that is written in a text frame (Connor, Upton, & Kanoksilapatham, 2007) with a particular communicative function (Swales, 1990; or communicative intentions (Yang & Allison, 2003) associated with writers' purpose (Amnuai, 2019a). On the other hand, a move sometimes consists of some obligatory 'steps' which function as complementary aspects to achieve the purpose of the move (Omidian, Shahriari & Shiyanova-Chanturia, 2018). Therefore, using the combination of both move and step would help the authors to make their abstracts achieve their communicative objectives.
Aside from the structure of move and step, abstracts embody language features such as tense and voice as they provide greater insight into the written genre (Esfandiari, 2014;Hanidar, 2016;Tseng, 2011;Tu & Wang, 2013;Zhang, Thuc, & Pramoolsook, 2012). Verb tenses, for instance, have been extensively investigated in abstract analyses. Cooley and Lewkowicz (2003) reported that the present tense was typically used in abstracts to present the summary. In a similar vein, Tseng (2011) and Zhang et al., (2012) unveiled a connection between tense and move in that the present tense is primarily employed in writing Introduction, Purpose, and Conclusion, while the past tense is used chiefly in Method and Product/Result. Zang et al. also discovered a much greater use of the active voice over the passive voice in abstracts.
The combination of rhetorical structure and linguistic realizations, as literature has shown, is a necessary piece of knowledge in writing abstracts. This is especially true for nonnative writers because although they are highly fluent in general English, they may lack the knowledge of expected rhetorical patterns and language features, making composing English abstracts increasingly tricky, especially for inexperienced writers (Amnuai, 2019b;Ren & Li, 2011;Wannaruk & Amnuai, 2016).
To ensure only the best quality papers, international research article databases, such as Scopus, have a panel of independent, international board of journal editors, librarians, and bibliometricians. With such highly qualified resources, internationally indexed journals such as Scopus journals may have greater credibility than locally indexed counterparts, thus fortifying a notion that Scopus journals are higher in status than local ones. Among the few previous studies relevant to such a contrast is Tamela's (2020) research comparing applied linguistics abstracts in Scopusindexed journals and Sinta (Indonesian)-index Copyright © 2021, authors, e-ISSN: 2502-6747, p-ISSN: 2301-9468 journals. While no critical discrepancies in move structures were identifiable from the two journals, such a finding can be attributed to both journals being Scopus-indexed, hence the absence of noteworthy dissimilarities. Another relevant study worth mentioning is Amnuai (2019a). Focusing on accounting research article abstracts in International and Thai-based journals, she uncovered similarities and differences in the rhetorical move and linguistic realizations across the two journal groups. Therefore, it is of empirical interest to undertake a study that rhetorically and linguistically examines tourism abstracts published in nationally indexed versus internationally-indexed journals.
With the above-mentioned backdrop, the present study sought to compare the rhetorical organization and linguistic realizations of tourism research article abstracts from national and international journals. By doing that, this study is expected to provide a textual description that can be used by tourism scholars, especially Indonesians, to more effectively write their abstracts and increase the chance of their works being published in national and international journals.

METHODS Design
This study employed a comparative qualitative approach as it sought the realization of rhetorical moves in tourism research article abstracts based on the indexation: Scopus as an international journal index and Sinta as a national journal index. Qualitative was utilized as this research analyzed data of a naturally occurring phenomenon focusing on words and specific situations (Flick, 2013;Maxwell, 2012), which relies on human interpretation and evaluation of the meaning of words, the development of concepts, and the interrelationship between them (Walliman, 2011). This study also used frequencies to determine the dominance of each move used in the R.A. abstracts.

The corpus
The data used in this study were 120 tourism abstracts from eight different journals. Fifteen abstracts were taken from each journal. Each international journal represents a different affiliation: Journal 1 (Scopus Q3, US-based), Journal 2 (Scopus Q3, Croatia-based), Journal 3 (Scopus Q4, Malaysia-based), Journal 4 (Scopus Q4, South Africa-based). Meanwhile, all Sintaindexed journals are affiliated with Indonesia: Journal 1 (S3), Journal 2 (S3), Journal 3 (S4), and Journal 4 (S4). Scopus Q3 and Q4 were chosen as they presumably were closer in terms of quality to the journals indexed by Sinta. It was also noted that Indonesian tourism journals only ranked S3-S5 thus far (up until this paper was written). That was why only S3 and S4 journals were selected. Considering tourism is a broad field, those aforementioned journals were chosen regardless of their topic discussions. Hyland's (2000) model was used as the main framework in this study, as presented in Table 1. This schema is commonly used in move analysis studies (e.g., Amnuai, 2019a; Ebadi, Salman, Nguyen, & Weisi, 2019; Ghasempour & Farnia, 2017;Saboori & Hashemi, 2013;Suherdi, Kurniawan, & Lubis, 2021).

Move
Step Label Introduction (M1) 1 Arguing for topic significance or prominences 2 Making topic generalizations: what is currently known 3 Defining the key term(s) 4 Identifying gap Purpose (M2) 5 Stating general and/or specific purpose of the research including the hypothesis Method (M3) 6 Describing participants 7 Describing instrument(s) 8 Describing procedure and context Product (M4) 9 Describing the main specific findings of the research Conclusion (M5) 10 Deducing conclusions from results by commenting on or interpreting the result or deducing claims from the results 11 Evaluating the significance or contribution of the research 12 Stating limitation 13 Presenting recommendation and implication Hyland's model was adopted due to its more precise definition of Move 5 -Conclusion, which covers both the discussion and recommendation or implication compared to Santos' (1996) model, which only emphasizes the discussion of the results and Swales ' (1990) that merged the introduction and purpose moves. Hyland's has also been tested on 800 abstracts of various disciplines to arrive at its generic patterns (Hyland, 2000) and found to be the most employed analytical framework in recent studies .
This study also investigated the salience of move and step occurrences, invoking Kanoksilapatham's (2005) model. In this model, the Copyright © 2021, authors, e-ISSN: 2502-6747, p-ISSN: 2301-9468 moves are considered optional if they appear in less than 66% of the abstracts, conventional if the appearances are 66%-99%, and obligatory only if they reach 100% of appearances.

Data analysis
The analysis process went through several steps. First, all the data obtained from online resources were copied to NotePad to get the .txt format. This step was because, in the next step, a software called AntMover (Anthony, 2003) would be used as a tool to break down the abstracts into sentences. After that, the analysis began with labeling each sentence to the compatible move and step, based on Hyland's (2000) model, as illustrated in Table 2.
With the help of Microsoft Excel, all the data and the labeling results were organized to tabulate the occurrences of the moves, the steps, and the language features. Each analyzed abstract was marked by its move pattern (e.g., 1-2-3-4-5), and all the move patterns were compiled and counted to draw the typical pattern used in each data group. Then, the researchers could draw a conclusion from the findings. Step The fact that Thailand has been a popular destination among global tourists has created challenges for hotel businesses to achieve high performance with excellent services that are responsive to the needs of global travelers.

M1 S1
This article aims to provide empirical evidence on the causal relationships among the effects of the dynamic capabilities, high-performance organization, and organizational performance of hotel businesses in a world-class tourism destination.

FINDINGS AND DISCUSSION
From the analysis, it was found that there were 937 moves and steps manifested in 120 abstracts from both groups. In Scopus-indexed journals, there were respectively 80 moves in Journal 1 (Q3), 135 moves in Journal 2 (Q3), 126 moves in journal 3 (Q4), and 96 moves in journal 4 (Q4). On the other hand, in Sinta-indexed journals, there were 131 moves in Journal 1 (S3), 124 moves in Journal 2 (S3), 115 moves in Journal 3 (S4), and 134 moves in Journal 4 (S4). In general, all of the moves appeared in both data groups. For an in-depth explanation, the subsections below provide detailed elaboration on the manifestation of the rhetorical moves.

The occurrences of moves and steps
This study discovered that all moves from Hyland's (2000) classification appeared in all data with some variations. Furthermore, some significant differences were found in the realization of abstracts from the two different data groups. The overall results are illustrated in Figure 1. As seen in Figure 1, the most similar moves in terms of their occurrences are M2 (Purpose), M3 (Method), and M4 (Product). As for M3, the realization of the said move in both groups is quite similar in percentage, with only a 1.44% difference. On the contrary, although the difference of M4 realization between Scopus and Sinta is also not too significant (2.13%), the Scopus group realized the said move more than the Sinta group. As for M2, despite being the least occurring move, it was realized by both groups with only a slight difference (2.18%) where Scopus had 15.1%, and Sinta had 12.92%.
The most significant differences are revealed to be the realizations of M1 (Introduction) and M5 (Conclusion). In Sinta, M1 was the dominant move, 30.42% of all the data. Compared to the Scopus group (19.45%), the gap is relatively wide (10.97%). The following excerpt is the demonstration of M1 realization in a Sinta abstract.

Example 1
Cipasung Village is administratively one of the villages in the Darma subdistrict… (Abstract 3, Journal 4, SINTA M1S3) These findings are consistent with Farzannia and Farnia (2017), who found that local authors (Persians) tend to use more of the Introduction move. This is also in line with Saboori and Hashemi's (2013) findings in their crossdisciplinary study, revealing that the introduction move was less frequent in English groups. As it establishes the context of the paper and the motives of the research, this move has the role of doing the "marketing" role to persuade potential audiences (Al-Khasawneh, 2017). As shown in Example 1, Indonesian authors tend to introduce their subject of study (e.g., tourism sites) by defining or stating general information regarding the said subjects. These findings might indicate that Indonesian authors used M1 as their tool to market their studies and their tourism sites.
As for M5, or Conclusion, being the least realized move, its occurrences differed across the data groups. With a relatively wide gap of 9.48%, Sinta's abstracts only realized 5.17% of all the data; meanwhile, Scopus's abstracts realized 14.65%. These findings are in line with the previous studies conducted by Al-Khasawneh (2017), Li (2011), Farzannia and, and Behnam and Golpour (2014), all of which discovered that local authors tend to leave out the Conclusion move compared to the international authors (English speakers). This could be attributed to the tendency of the local authors to believe that conclusion is unnecessary in attracting readers' interests and the awareness of international authors regarding the importance of the move (Al-Khasawneh, 2017; Li, 2011).
This study also reveals a variation in realizing the moves. As mentioned earlier, moves and steps are determined by sentences. However, there is a variation of utilizing M3 (Method) and M4 (Results) in one sentence, which occurred several times in different abstracts. It is found that the embedded move occurred seven times in Scopus-indexed abstracts. The embedded moves are represented in the following excerpt: Example 2 With the analysis of about 13,899 hotels in 146 cities, our findings suggest a linear relationship between the number of reviews and the TripAdvisor score but not Booking.com. (Abstract 11, Journal 1 Scopus) As Example 2 shows, the authors tend to embed M3 to M4 by adding additional clauses. What seems to be their similarity is that M3 is most likely to be put in appositive and positioned in the first clause, then followed by M4. In that sense, the organization of the moves still preserves its sequential characteristics. These findings also appeared in Tamela's (2020) study that also investigated Scopus abstracts. This may support the argument that abstracts from Scopus-indexed journal articles are more compact than nationally indexed journal articles .
In addition to moves, the results also reveal the occurrences of the steps. As M2 (Purpose) and M4 (Result) only consist of one step, they would not be further analyzed. Therefore, the main focus of this section is the realizations of M3, M4, and M5. There are some similarities and differences identified from the data. Figure 2 represents the steps occurrence of both groups in percentage. Figure 2 shows that both groups realized all of the steps. The most frequent step in both Scopus and Sinta groups was S8 (describing procedures and context). The dominant occurrence of this step was also revealed in . Interestingly, authors often employ this step to insert the information that could have been in the other steps of M3. Thus, instead of realizing different steps, the authors seem to include participants and instruments along with their context and/or procedure. The following excerpt illustrates this phenomenon.

Example 3
Primary data collected from 142 hospitality students prior to their internships using a questionnaire was subjected to descriptive statistical and Pearson correlation analysis.

(Abstract 4, Journal 4 Scopus)
Not only does Example 3 elaborate the context and the procedures (e.g., descriptive statistical and Pearson correlation analysis), give the information about the participants (e.g., 142 hospitality students) and the instruments (e.g., questionnaire) which could be realized through S6 (describing participants) and S7 (describing instruments).

Figure 2
Step Occurrence In general, the use of S8 from M3 stood out among the other steps. As for the other steps, the occurrences of S1, S4, S6, S7, and S10 were identical in terms of percentage, whereas S2, S3, and S11 were significantly different across the data. Two steps from M5 (Conclusion), S12 and S13, were the least preferred steps utilized in both groups.

The salience of moves and steps
This study found that certain moves were obligatory in one journal but conventional or optional in the others regarding the move-step salience. Moves and steps are considered obligatory if they appear in 100% of the abstracts, conventional if they appear in =/>66%-99%, and optional if their appearances are less than 66% (Kanoksilapatham, 2005). The study found that no move was obligatory in all Scopus and Sinta journals. However, some could gain that status in particular journals. Table 3 and 4 respectively illustrate the move salience of abstracts from Scopus-indexed journals and Sinta-indexed journals. From Table 3, it appears that M4 was obligatory in Journal 1 and Journal 2 but conventional in Journal 3 and Journal 4. Interestingly, Journal 1 and Journal 2 are from quartile 3, which is a higher quartile than Journal 3 and Journal 4, which belong to quartile 4. Aside from M4, M3 and M5 were also obligatory only in Journal 2. The rest of the moves in the other three journals were mostly conventional except for M1, which appeared in Journal 1 for 60% and Journal 2 for 20%. simultaneously appeared in less than 50% of all abstracts in Sinta.
The obligatory status of M2, M3, and M4 seems to resonate with the findings of the previous studies (Chalak & Norouzi, 2013;Darabad, 2016), including Ahmed's (2015) study in tourism research article abstracts. In his study, Ahmed (2015) argued that M2, M3, and M4 were obligatory in tourism abstracts because the researchers may consider the mentioned moves more important than M1 and M5. This may also account for the lack of M5 in Sinta data.
As for the steps, only S8 appeared to be obligatory in Sinta Score 4 and conventional in the rest of the abstracts. Aside from that step, no step obtained neither obligatory nor conventional status in all journals.

Linguistic features
As for the linguistics features, the results show that both Scopus and Sinta groups indicate more similarities rather than differences. Generally, the present tense in abstracts of Scopus and Sinta indexed journals is dominant (see Table 6).
Copyright © 2021, authors, e-ISSN: 2502-6747, p-ISSN: 2301-9468 In using tenses, similar to other studies (e.g., Al-Shujairi, Ya'u, & Buba, 2016;Amnuai, 2019a), both groups were dominated by the present tense, with the exception of M3, which used more past tense. These results echo Tseng's (2011) finding that present tense was used more in M1, M2, and M5, yet past tense was used more in M3 and M4. It has to be highlighted that the difference of M4 in this study and Tseng's (2011) might be due to some disciplinary variations (i.e., tourism vs. applied linguistics), as Swales and Feak (2004) stated: "there appears to be considerable disciplinary and individual tense variation with sentences dealing with results" (p.283).
Similar to the tense findings, both groups also tend to realize the rhetorical moves in the active voice rather than passive voice (see Table 7). This confirms the findings from the previous studies that active voice was the preferred voice in research articles (Amnuai, 2019a;Muangsamai, 2018). Aside from emphasizing the agents of the sentence, texts would be more understandable if they are written in the active voice (Lorés, 2004;Raimes, 2004;Swales, 1990). In realizing M3, however, Scopus authors' preferred choice was the passive voice, similar to Hanidar (2016) and Tu and Wang (2013). One plausible reason for this is that passive voice shows a more impersonal tone and is considered more formal (Amnuai, 2019a), or simply because addressing the agents of the sentence seems unnecessary as it is irrelevant or obvious (Greenbaum & Nelson, 2009).
Regarding the verb use, as shown in Table 8, two data groups seem to share more agreements than disagreements. Both Scopus and Sinta preferred using action verbs in M2, M3, and M5, and relational verbs in M4. One noticeable difference is their preferences for using verbs in realizing M1. In Scopus, the authors were likely to use action verbs. The reason for this may be that, in M1, the authors aimed to describe or recount the things involving actions that have been done/known that might strengthen their research backgrounds. On the other hand, Sinta's authors preferred to use more R.V. in their introductions. As Knapp and Watkins (2005) stated in their books, "…when using writing for the expression of knowledge, it is easier to deal with process and concepts in their noun forms" (p. 74). Either way, both A.V. and R.V. seem to be an appropriate tool to achieve the primary goal of M1, which is to establish the context and the motives of the research. Furthermore, echoing Hapsari's (2019) finding, this study also discovered that the dominant verb types used in research articles, both international and local, are AV and R.V. regardless of the preference of verb types in each move.
Copyright © 2021, authors, e-ISSN: 2502-6747, p-ISSN: 2301-9468 The study also found that the use of modal auxiliaries can be found in all moves. However, authors tend to use them more on M1, M4, and M5. The use of modal auxiliaries could indicate the authors' evaluation based on various degrees of confidence (Vathanalaoha & Tangkiengsirisin, 2018). The modal auxiliaries found in tourism research article abstracts are can, may, might, will, would, could, should, must, have to, need to. Abstracts from the Sinta group tend to use more of the modal auxiliaries compared to the Scopus group. This finding corresponds to Hapsari's (2019) study that found the tendencies of nursing research article authors to employ modal auxiliaries in M1 and M5. As M1 and M5 provide some context related to external aspects of the studies and arguments to the related topic, modal auxiliaries might be used to express obligation or probability in a range of grammatical resources (Knapp & Watkins, 2005). As for M4, as Pho (2008) stated, there are some cases in which the authors used modal auxiliaries when they summarize the findings.
From the findings, it can be concluded that despite some noticeable differences in move occurrences and patterns, tourism research article abstracts from two different indexations still share many resemblances in terms of move-steps realizations and the use of linguistic features.

CONCLUSION
This study has investigated the rhetorical move-step patterns, and linguistic realizations in tourism research article abstracts published in international journals indexed by Scopus and national journals indexed by Sinta. Analysis of the data reveals more similarities than differences across the two data sets.
In general, abstracts of the two journal indexing contexts applied all of Hyland's five moves. The two groups bear resemblance in most occurring moves, i.e., M2 (Purpose), M3 (Method), and M4 (Product). The most recurring pattern in Scopus abstracts was 2-3-4-5 (Purpose-Method-Product-Conclusion). At the same time, 1-2-3-4 (Introduction-Purpose-Method-Product) was the preferred choice in Sinta abstrcts, indicative of a linear pattern in organizing the structure of the abstracts desirable in Tourism abstracts. A noteworthy difference lies in the realizations of M1 (Introduction) and M5 (Conclusion). While M5 was realized a lot more in Scopus abstracts, M1 was more evidenced in Sinta abstracts. As for the steps, the realization of Step 8 (describing context and procedures) in Method outnumbered the rest of the steps in both data groups. Concerning the linguistic features, the results of Scopus and Sinta abstracts were close to identical. Both data groups predominantly used the present tense and active voice in all of the moves. The noticeable exception appears in M3, which was realized mainly in the past tense and passive voice. As for the verbs, every verb type was utilized in tourism abstracts, with action verbs (A.V.) being the most employed verb type in both data groups. Modals were evident in M1, M4, and M5.
The findings from this genre-based approach will provide researchers and nonnative writers, especially those in the field of tourism, with a heightened awareness of stereotypical discoursal patterns and grammatical features of R.A. abstracts. With such an awareness, they can write their abstracts more effectively. However, given this study's relatively small data size, it is difficult to generalize the results to abstracts in all indexing contexts. A future study may look at abstracts of other indexing levels to see if similar patterns yield. Pedagogically speaking, the findings here may serve as supplementary teaching materials in English academic writing and a data-driven genre-based teaching approach Peacock, 2002) appears to find relevance in this context.