Citation Studies in English vs. Indonesian Research Article Introductions (RAIs) in the History Discipline

Citation studies in research articles (RAs) have been widely conducted worldwide, but such studies rarely compared English and Indonesian RAs, especially within the history discipline. Therefore, the researchers intended to analyze and compare citations in English and Indonesian research article introductions (RAIs) in the history discipline using a genre approach for the analysis and a descriptive qualitative approach for the reports. In this regard, 30 RAIs from both data sets were analyzed using two frameworks: one is to identify citation techniques, and the other is to analyze citation types. The results revealed that English and Indonesian authors used descriptions more than other techniques when citing sources. However, English authors employed this technique more than Indonesian authors. In addition, both English and Indonesian authors used a non-integral type more frequently than the integral counterpart, but English authors employed this type more frequently than Indonesian authors. Thus, these results conclude that although both English and Indonesian authors


INTRODUCTION
Citation plays an essential role in writing research articles (RAs) worldwide. It determines how RAs authors used source texts to produce new texts within their contexts (Badenhorst, 2017;Bazerman, 2004;Wang, 2006). In addition, citations also show how authors integrate texts from other sources into new texts, use the types of sources and their function, and position themselves toward the cited sources (Bazerman, 2004). By comprehending these essential citation roles, the present study aims to compare the uses of citation in English and Indonesian RAs using the genre of RAs.
Genre analysis of RAs has been widely conducted by various linguists worldwide, and this approach has been used to analyze citations within RAs (Hyland, 1999;Mansourizadeh & Ahmad, 2011), including citation practices (Dobakhti & Hassan, 2017;Dobakhti & Zohrabi, 2018;Helal, 2014;Kuhi & Mollanghizadeh, 2013;Shooshtari et al., 2017;Varga & Gradečak-Erdeljić, 2017). However, although citation studies within RAs have been widely conducted, studies dealing with comparing English and Indonesian RA have received little attention from scholars. Based on the researchers' reviews, only three linguists were found to analyze citations within Indonesian RAs (Adnan, 2010;Arsyad & Adila, 2018;Mirahayuni, 2002). Among them, only one compared between English and Indonesian RAs (Mirahayuni, 2002). Comprehending how authors cite source texts in writing their RAs is very important, particularly for novice authors, because it reflects their writing relevancies related to the scientific impacts (Aksnes et al., 2019). Thus, citation studies help them improve their scientific impacts.
Although Mirahayuni (2002) compared the citation analysis between English and Indonesian RAs, her study differed from the present one. Her corpora were from the English Language Teaching (ELT) discipline, while the present study investigated citations within the RAs in the history discipline. This context was selected because during the researchers' review, this disciplinary RA has never been investigated to date. In addition, her analysis focused on investigating authorial stances toward citations. This review implied that none had analyzed the techniques RA authors used to cite source texts to write their new texts in Indonesian RAs and the types of citations they used to produce new texts in Indonesian RAs. Therefore, the present study intends to compare citations employed in English and those in Indonesian RAIs in the history discipline, focusing on the following research questions: • What citation techniques do English and Indonesian RA authors employ in citing source texts in their RAIs? Are they similar? • What citation types do English and Indonesian authors employ to produce their RAIs?
The findings of the present study may contribute to the knowledge development of citation studies within RAs. In addition, comparing citation studies between English and Indonesian RAs may create more meaningful comprehension and can contribute to designing teaching materials for English for Specific Purposes (ESP) and English for Academic Purposes (EAP), particularly for academic publication purposes.

Citation Techniques
In analyzing citation practices, Wang (2006) introduced three citation techniques: direct quotation, paraphrasing, or description. The description of these three techniques is presented as follows: a. Direct quotation In this regard, authors cite the original work by providing quotation marks "…" or italicizing the cited words, and a page number is usually included with the cited works. Thus, a direct quotation is the easiest way to identify as the cited sources are directly inserted into the RAs. b. Paraphrasing Authors paraphrase by re-writing other people's works using their own words to make the meaning more explicit. Paraphrasing does not use quotation marks ("), but it usually contains restatements and uses a marker such as 'that' to indicate the cited information. c. Description In description, authors tell, describe, illustrate, define, or explain anything about something, someone, or a situation to the readers. In this context, the authors describe cited sources to provide information for the readers. It can be identified by the words used in the text, such as how the authors include verbs, adjectives, and adverbs. As presented above, citation techniques have been clearly defined in previous studies. Therefore, these citation techniques can be used as a framework for analyzing citation techniques within texts. Considering this clear and comprehensive description, the researchers intended to employ this framework for analyzing citation techniques within RAIs in this current study.

Integral citation
An integral citation means that the source integrates with the texts. This citation type has five categories: verb controlling, passive agent, naming, adjunct passive agent, and non-citation.
• Verb controlling. In this type, the citation is an agent that controls the verb. For example, Kutay (2016) finds that the passive voice is more common than the active voice ……. • Passive agent. In this citation type, the agent of the cited source is in a passive sentence. For example, this finding is supported by the earlier research conducted by Hasan (1990)……. • Naming. It is a citation that employs noun phrases. For example, another study was conducted by Evan (2006), who focused the analysis on……. • Adjunct passive agent. It is structurally optional because its absence does not influence a sentence structure, but it is to show that the statement is credited to other writers. For example, according to Kutay (2016), there are some similarities and differences between this finding and the earlier studies. • Non-citation. In this case, the reference may locate somewhere else, but an author cites the source without a year reference. For example, moreover, she finds that the structure of the Indonesian and English RAs is …… The cited sources integrate directly with the texts in the examples above. In this context, the cited authors' names are included within the sentences, not in parentheses, and thus they are termed integral citations, in which the researchers focus their attention on the authors of the cited sources.

Non-integral citation
Non-integral citation means that authors use a citation but do not integrate it with their text. This type of citation is categorized into source, identification, reference, and origin. • Source. This citation informs where the ideas originate. For example, the generic structure is one of the most important investigations in genre analysis (Swales, 1990). • Identification. This citation identifies the agent within the texts. For example, genre is defined as how to get things done using language (Hyland, 2004). • Reference. This often signals where the information originates, with the word "see" often used in the citation. For example, the study of intertextuality is about the ways and the purposes of employing source texts to the new texts (see Bazerman, 2004). • Origin. This citation usually shows the origin of the theory, concept, or framework.
Both citation types, integral and non-integral citations, have been clearly described and diverted in previous studies. The key difference between the two citation types rests on the authorial integration. In the integral citation, the researchers integrate their cited sources with their current texts. In contrast, in the non-integral citation, researchers do not integrate their cited sources with their current text and locate their cited names within parentheses. Comprehending this clear framework is pivotal because it may make analysis results clear and meaningful; and thus, the researchers used this framework to analyze citation types within RAIs in this current study.

Citation Studies Worldwide
Citation studies within RAs have been conducted by many linguistic scholars. For example, the citation techniques in 49 English biomedical Ras were analyzed, and the results revealed that of the four citation methods proposed in the literature (direct quotation, paraphrase, summary, generalization), biomedical RAs mainly employed generalization, followed by a summary and paraphrase (Dubois, 1988). Then, the least employed citation in his findings was a direct quotation. This finding means that RA authors in this discipline tended to assimilate their citations rather than directly inserting citations into their RAs.
Another citation study was conducted on 30 English RAs in applied linguistics published in reputable international journals and 30 Iranian RAs published in Iranian national journals (Kuhi & Mollanghizadeh, 2013). The results revealed that the English RAs had 1830 citations, with an average of 61 citations per RA, while the Iranian RAs had 1140 citations, with 38 citations per RA, suggesting that English RAs had more citations than Iranian RAs. However, both data sets tended to employ integral citations more than non-integral citations.
Contrastive citation studies were also conducted on 24 research article introductions (RAIs) by American authors and 12 RAIs written by French authors (Helal, 2014). The results showed that American RAIs had more citations than their French counterparts, 26.5 citations per RAI and 17.58 citations per RAI, consecutively. In the whole paper, American RAs had 32 citations per RA, while French RAs had 18 citations per RA. These findings indicate that American RAs had more citations than French RAs, both per RAI and per paper.
Another contrastive study on citation analysis was carried out on 16 English empirical RAs written by English native speakers (NSs) and on 16 empirical Croatian RAs (Varga & Gradečak-Erdeljić, 2017). Both corpora were from the applied linguistics discipline. English RAs had much longer texts (143,106 words) than Croatian RAs (52,159 words). Then, English RAs also had more citations. English RAs had 1244 citations, with an average of 8.7 citations per 1000 words, while the Croatian corpus had 380 citations (7.28 citations for every 1000 words). However, the citation type in both groups was similar, that is, both groups mostly used non-integral citations in the introduction, discussion, method, and results sections.
A contrastive citation study between English and Persian RAs was also undertaken using a corpus of 240 RAs from the hard and soft sciences (Shooshtari et al., 2017). The soft sciences included the disciplines of applied linguistics and psychology, while the hard sciences involved the disciplines of computing and mechanical engineering. The results revealed that English RAs had more citations than Persian RAs, and soft sciences had more citations than hard sciences. In the soft sciences, English RAs had 54.96 citations per RA, while Persian RAs had 37.83 citations per RA. In the hard sciences, English RAs employed 29.8 citations per RA, while Persian RAs employed 22.11 citations per RA. Thus, English RAs had more citations than Persian RAs in soft and hard sciences. Then, in citing sources, both groups of authors tended to employ assimilation (adjusting the source with the authors' contexts) more than inserting the sources directly (using direct quotes). English RAs employed it in 87% of the total citations, while Persian RAs used it in 88%. This finding suggests that both groups avoided direct quotations when citing sources. Concerning authorial stances, both groups tended to employ acknowledgment, which means that "a writer adopts a neutral position and makes no evaluative judgment on the cited proposition" (Shooshtari et al., 2017, p. 60).
A contrastive citation study was also conducted within the corpus of 150 qualitative and 150 quantitative RAs in applied linguistics published in English highindex journals (Dobakhti & Hassan, 2017). The aims were to investigate the author's identity and the use of personal pronouns within the corpus of qualitative and quantitative RAs. The results revealed that the qualitative corpus employed more pronouns than the quantitative corpus. The RA authors employed pronouns mainly to present their results, claims, arguments, purposes, and procedures intended to show their presence. Showing their presence, in this case, means that they also revealed their responsibility. This finding indicates that qualitative RA authors claimed more responsibility for their writing than quantitative RAs.
In international high-impact factor journals, a citation study was conducted on 45 Research article discussions (RADs) in the applied linguistics discipline (Dobakhti & Zohrabi, 2018). The study found that the corpus had 330 citations, with an average of 7.31 citations per RAD. Most citations (60.91%) were non-integral, while the rest were integral. Concerning the four functions of citation methods, most citations (42.12%) were used to make comparisons with the authors' results; 32.12% of other citations were used to explain the results; 13.34% of the citations were used for research recommendations, and the rest of 12.42% was used to interpret the results.
As reviewed above, citation studies have been widely conducted within RAs worldwide, including citation techniques, citation types, authorial stances, authorial identities, and personal pronouns. However, further studies were still required, mainly to add literature to the body knowledge of citation studies, including in Indonesia. Thus, the following section reviews citation studies in Indonesia.

Citation Studies in Indonesia
In the Indonesian context, a comparative study focused on 58 RAIs and RADs in the language and language teaching disciplines (Mirahayuni, 2002). The corpora were RAIs written in English by English Native Speakers (NSs) and Indonesian NSs, and in Indonesian by Indonesian NSs. The results revealed that the citations within RAs in English and Indonesian written by the Indonesian NSs had different functions from those in English written by the English NSs. English NSs included citations to situate their study in order to justify their study. In contrast, Indonesian NSs used citations to display their authority as knowledgeable people to their readers. They tended to show that the readers needed their study (Mirahayuni, 2002). However, they did not respond to the knowledge gap in the literature. These findings indicate that the function of the citations differed between RAs written by English NSs and those written by English Non-Native Speakers (NNSs). These differences affected how the authors employed authorial stances, voices, or attitudes toward their cited sources.
Citations have also been analyzed on 63 Indonesian RAIs from the disciplines of education, linguistics, and social and political sciences (Adnan, 2010). The study revealed that linguistic RAIs had 534 citations, with an average of 25.43 citations per RAI, and social-political RAIs used 323 citations, with an average of 15.38 citations per RAI. However, educational RAIs only had 316 citations, with an average of 15.04 citations per RAI. Compared to those found in English RAIs, Indonesian RAIs in these three disciplines had much fewer citations. Moreover, the finding revealed that most Indonesian authors were neutral toward their cited sources. However, the study only investigated citation numbers and authorial stances but ignored citation techniques and types.
Furthermore, citations were also analyzed on 40 English RAIs written by Indonesian authors published internationally in Scopus index journals in 2012 (Arsyad & Adila, 2018). The analysis revealed that Indonesian academics employed citations to support their ideas and justify their positions, but they did not criticize or evaluate what they cited. These findings differed from those in English RAs, as English NSs criticized and evaluated other literature to find research gaps. Therefore, the function of citations in English RAIs written by Indonesian authors differed from English RAs written by English NSs. However, this study neglected the techniques that authors used in citing sources and the citation types they employed.
In short, although citation analyses have been conducted widely, such studies still receive little scholarly attention in the Indonesian context since only three researchers (Adnan, 2010;Arsyad & Adila, 2018;Mirahayuni, 2002) were found to analyze citations within Indonesian RAs. Among the three, only Mirahayuni (2002) compared citations in English and Indonesian RAs. However, their citation studies focused on citation functions, authorial stances, and citation numbers within corpora. In contrast, studies of citation techniques and types in Indonesian RA contexts were rarely found. Although Mirahayuni (2002) compared citations in English and Indonesian RAs, her study focused on citation functions within RAs. Meanwhile, the present study analyzes citation techniques and types used in English and Indonesian RAIs. Thus, the present study adds insights to the literature and provides pedagogical contributions to teaching English for academic purposes, precisely publishing purposes for Indonesian authors and other non-native English authors. For this rationale, the researchers intended to conduct the present study.

Research Design
The present study used two approaches in designing this research. Firstly, it employed a genre approach in the English for Specific Purposes (ESP) stream as suggested in Swales (1990Swales ( , 2004 and Bhatia (1993Bhatia ( , 2014Bhatia ( , 2016 for analyzing English and Indonesian RAIs. Then, it employed a descriptive qualitative approach in presenting the analysis results from both data sets. By employing these two approaches, the analysis results were expected to be comprehensive and robust.

Collecting Data Sets
The present study analyzed 30 RAIs from two data sets; 15 RAIs that were written in English and published in English journals and the other 15 RAs that were written in Indonesian and published in Indonesian journals. 30 RAIs were considered sufficient in this study because each RAIs already had numerous corpora to be analyzed. Therefore, this sample size was representative to answer the formulated research questions. Meanwhile, the reason for employing two data sets was to provide more comprehensive and meaningful comparisons in the analysis results.
In selecting English RAs, the researchers determined the journals by following the standard criteria. Firstly, the journals must be written and published in English. Then, the journals were from the history discipline, which can be identified by reading their 'focuses and scopes' on the journal home page. Secondly, the journals must be indexed in Scopus Quartile 1 (Q1) and Scimago Journal Ranks (SJR) above 0.50. Using these standard criteria, the researchers found three English journals in the history discipline: Historical Archeology, Historical Methods, and Journal of Global History. From these three selected English journals, the researchers then selected 15 articles specifically written by English NSs. The selection of English NSs articles was made by identifying their names and affiliations.
Furthermore, in selecting 15 Indonesian RAs, the researchers used Indonesian Science and Technology Index (Sinta) standard. The researchers selected three journals in the history discipline, which received the highest Sinta rank, i.e., rank 2 because no journal in this selected discipline occupies rank 1. The selected journals include Jurnal Sejarah Citra Lekha, Patanjala: Jurnal Penelitian Sejarah dan Budaya, and Patra Widya: Seri Penerbitan Penelitian Sejarah dan Budaya. Afterward, the researchers selected five articles from each of these journals randomly since they were all written in Indonesian and by Indonesian authors (identified by their names and affiliations).

Data Analysis Procedures
In analyzing citations from the data sets, two analytical frameworks were used to answer the research questions because one framework alone cannot cover all the research questions. The first framework suggested by Wang (2006) (as described in the literature review) was used to analyze citation techniques, whether authors tended to employ a direct quotation, paraphrasing, or description in their RAIs. Then, the second framework used was underpinned by Kuhi and Mollanghizadeh (2013) and Varga and Gradečak-Erdeljić (2017) to analyze the types of citations used by authors to produce their new texts. The purpose was to find out whether citations in both data sets tended to employ integral or non-integral citations in the data of RAIs.

Reliabilities of Data Analysis Results
To ensure the reliability of data analysis results, two researchers from Ph.D. graduates in linguistics divided their roles in this research. The first researcher analyzed both data sets using the above procedures. Then, the other researcher, as an inter-rater, also checked the analysis results using the same procedures. The accuracies of data analysis results were then deliberated to meet the agreements. However, there has been no disagreement between the two researchers, indicating that both analysts were in 100% agreement with the analysis results. The results were considered reliable and accurate by employing these layer steps for the analysis.

RESULTS
This study analyzed the citing technique used in English and Indonesian RAIs in the history discipline and investigated the citation types used to write their RAIs.
This section reports the analysis results into two sub-sections; the first sub-section reports the citing techniques used in both data sets, and the following sub-section reports their citation types.

The Citation Techniques Used by English and Indonesian Authors
After analyzing both data sets, the results showed that English and Indonesian authors in the present study employed similar ways of citing sources. Authors in both data sets tended to use a 'description' technique when citing their source texts. Besides, they also used 'paraphrasing' as the second most employed form. Finally, they used 'direct quotations' as the least used technique for citing resources. The percentages of the technique are presented in the summary of the analysis results in Table 1. As shown in Table 1, authors in both data sets used descriptions, but English authors used this technique more frequently than their Indonesian counterparts, around 95% and 59% consecutively. On the other hand, they employed fewer paraphrasing and direct quotations than Indonesian authors. In this sub-section, however, the researchers only exemplified the most employed citation technique, i.e., description, found in both data sets to minimize the space. Thus, the researchers did not provide examples for paraphrasing and direct quotation because these two citation techniques are relatively straightforward.
Example 1 : However, in some cases, the nature of the artifact surface, or its coloring, makes such images less than desirable. Some alternative methods have been applied, including photogrammetry (Selden 2015), laser scanning, and three-dimensional (3-D) scanning (Means et al. 2013). EHRAIs04 Example 2 : In the twentieth century, the world's human population grew faster than it ever had before and faster than it probably ever will again (Lam 2011). EHRAIs06 Example 3  The four examples above illustrate the techniques of citing sources found in both English and Indonesian RAIs, and referring to Wang (2006), this citation technique is called 'citation by description.' The Examples 1 and 2 are from English RAIs, while the Examples 3 and 4 are from Indonesian RAIs. In this regard, authors comprehended the source texts and adjusted them to their writing contexts.
In short, English and Indonesian authors in the present study tended to use descriptions when citing sources rather than the other two techniques. However, English authors used this method more frequently than Indonesian authors. This finding implies that both English and Indonesian authors preferred assimilating their citations within their contexts to directly inserting source texts into their RAIs. Furthermore, it was also found that English authors tended to assimilate their citations more than Indonesian authors.

The Citation Types Used by English and Indonesian Authors
After analyzing both data sets in the present study, the results revealed that English authors employed similar citation types to Indonesian authors. Both groups of authors used non-integral citations more frequently than integral citations when citing resources, but English authors used non-integral citations more frequently than Indonesian authors. More details about citation types used in both data sets are presented in the summary analysis results in Table 2. As presented in Table 2 above, English and Indonesian RA authors from both data sets did not frequently integrate their citations into their texts because they used non-integral citations, around 89% and 65% consecutively. However, Indonesian authors used more integral citations (around 34%) than English authors (around 10%). The examples of integral and non-integral citations from both data sets are presented in the following.

Integral citation
This citation type appeared only in 10.44% of the total citations in English RAIs and 34.57% of the total citations in Indonesian RAIs. This finding indicates that English authors employed integral citations less frequently than their Indonesian counterparts. The following are the examples of integral citations found in both data sets: Example 5 : (verb controlling): Gutman (1976) notes that besides the studies of Puckett (1938Puckett ( , 1975 and Wood (1974), and observations from Mencken (1919) The authors in the two examples above used integral citations. In Example 5, the agent (Gutman) controlled the verb 'notes', and in Example 6, the agent (Irwanto) controlled the verb 'explains.' The reporting verbs as verb controlling were used in both examples because the subjects in the citations did the action. Besides, the examples also show that their sources were integrated into their texts since their sources were used in sentences, not in parentheses. Thus, the authors of the above examples used integral citations with a verb-controlling sub-type.

Non-integral citation
This citation type was dominant in both data sets. It appeared in 89.56% of the total citations in English RAIs and 65.43% in Indonesian RAIs. The excerpts of this citation type found in both data sets are presented in the following: The two examples above used non-integral citations because the authors did not integrate their texts into the citations. In addition, their citations were also presented in parentheses. Thus, the type of these above citations belongs to non-integral citations as introduced by Kuhi and Mollanghizadeh (2013) and Varga and Gradečak-Erdeljić (2017).
Overall, the analysis revealed that RA authors from both data sets were inclined to use description most frequently, with non-integral citation as the most dominant type. However, English authors employed the description method and non-integral citation type more frequently than Indonesian authors. These findings imply that English and Indonesian authors in the present study tended to assimilate their citations into their texts. However, English authors assimilated their citations more often than Indonesian authors and did not integrate them more frequently. Thus, English authors had more assimilation than Indonesian authors when citing sources and tended to avoid integrating their sources into their RAIs.

DISCUSSION
One of the interesting findings in the present study was that English and Indonesian RA authors used 'description' as a way of citing sources, which means that they tended to assimilate their source texts into their RAIs. The findings from both data sets seemed similar to those in English Biomedical RAs, in which the authors also tended to generalize their citations when citing sources and avoid using direct quotations (Dubois, 1988). These findings are also consistent with the data obtained from both English and Persian Ras, as they also tended to assimilate their citations and avoid direct quotations when writing RAIs (Shooshtari et al., 2017). These findings indicate that citing sources using the 'description' technique seems to be favored among RA authors, and this method was widely employed by most authors (around 50%-90%) in both data sets (Warsidi, 2022). The underlying reason could be that the description technique could be used more easily to adjust messages to their writing contexts.
However, although both English and Indonesian authors in the present study tended to assimilate their citations into their writing, English authors employed this technique more frequently than Indonesian authors. In contrast to the results of the present study, English RA authors employed this citation technique fewer than Persian RA authors (Shooshtari et al., 2017), with a very slight difference (87% in English RAs and 88% in Persian RAs). This discussion implies that although the percentages of the description technique differed between English and Indonesian RAIs, as well as between English and Persian RAs, the description was still the most favored method employed when citing sources among authors. In this context, authors needed to comprehend the source texts or citations to be adjusted to their writing contexts. The possible reason for English authors using the description technique more frequently than Indonesian authors is that English authors in the present data were more advanced academic writers than Indonesian authors. These findings helped RA authors understand citation techniques, comprehend the meanings of citations, and adjust the citations to their contexts, which were pivotal to convincing and attracting readers.
Another interesting finding was that English and Indonesian authors in the present study tended to avoid integrating their citations into their texts. They chose to use non-integral citations when citing sources. However, unexpectedly, English authors tended to employ this citation type more frequently than Indonesian authors. As in using the citation techniques, the possible underlying reason for the different frequency in using the citation types between the two author groups is that English authors might be more experienced than Indonesian authors in the present data. These findings are consistent with the citations found in RAs in soft sciences, such as citations in English and Croatian RAs in applied linguistics (Varga & Gradečak-Erdeljić, 2017), applied linguistics and psychology (Shooshtari et al., 2017), and English RAs published in internationally reputable journals (Barghamadi & Siyyari,