Interactional Meta-Discourse Resources in Oliver Twist

For long, there has been debate over the appropriateness of using simplified literary texts in second language classrooms. In examining the simplified form, the main question is persuasion that is partly achieved through meta-discourse resources, which are “Self-reflective linguistic material referring to the evolving text and to the writer and the imagined reader of the text” (Hylan & Tse, 2004, p. 156). The present study compares the use of interactional meta-discourse resources (IMRs) in terms of the frequency and categorical distribution in the original copy of a novel (i.e., Oliver Twist) and its simplified counterparts. The corpus was analyzed based on the Hyland (2005) model. The frequency and categorical distribution of IMRs were calculated per 1,000 words, and the difference in their distribution was calculated using Chi-Square statistical analysis. The findings indicate a significant difference in the frequency of IMRs between the original and the simplified versions of the Dickensian novel, implying that despite having more complex syntax and semantics, the original novel seems to be more persuasive, at least on the part of IMRs, compared to its simplified counterparts. In terms of categorical distribution, there was no significant difference between them.

The notion of authentic material in language learning is a controversial issue. The proponents (e.g., Mishan, 2005;Gilmore, 2007;Rilling & Dantas-Whitney, 2009) believe that authenticity promotes language acquisition, communicative competence, and a positive attitude towards language learning. Whereas opponents (e.g., Widdowson, 1984Widdowson, , 2000Ellis, 1999;Day, 2003) contend that authenticity brings about many difficulties in the process of language learning. The argument over applying authentic versus simplified texts to the language classroom is ongoing and requires further research. In terms of the difference between simplified and authentic texts, a number of studies were conducted. Crossley, Louwerse, McCarthy, and McNamara (2007) investigated a wide range of linguistic tools in authentic and simplified texts using Coh-Metrix. They revealed more complexity in authentic texts in terms of their syntax and the use of logical connectors while more explicit cohesion devices were found in simplified texts. Oh (2001) investigated the effect of text simplification on learner comprehension and found that simplification facilitates comprehension. Crossley, Allen, and McNamara (2012) compared simplified texts at different levels in terms of linguistic features including lexical sophistication, syntactic complexity, and cohesion. They noticed that simplified texts at a lower level have more cohesive devices, less complex lexis, and simpler syntax. Crossley (2018) notes that knowing the linguistic differences between simplified and authentic texts leads to imposing a larger question of whether simplification leads to better comprehension. The common goal of all simplified texts is to decrease the learners' cognitive load and increase comprehensibility (Crossely et al., 2007).
Almost all studies conducted on simplified and original texts focused on syntactic complexity and lexical density, revealing that simplification may or may not benefit text comprehension. Despite syntax and semantics, there are other resources that should be taken into consideration when comparing authentic and simplified texts. One of the major problems with simplified texts is persuasion, which as part of the rhetorical structure, is partly achieved by employing meta-discourse resources (Petersen & Ostendorf, 2007). Meta-discourse is an estimated way of using language to make it more comprehensible for readers (Hyland, 2017). Hyland (2019) defines meta-discourse resources as those that elaborate on the relation between the writer and the text as well as between the writer and the reader. The aim of such resources is more than conveying information, they comprise the personalities and perspectives of the communicators, and their removal leads to less personal and less interesting text (Hyland, 2019). This paper presents an analysis of an authentic literary text, an original novel, and its simplified counterparts in terms of meta-discourse resources.
Based on the current literature, few studies have focused on literary genre, specifically to juxtapose an original literary text (e.g., a novel) with its simplified versions rendered by different narrators and to compare them in terms of meta-discourse resources. Most notably, Esmaili and Sadeghi (2012) compared the original and a simplified counterpart of Oliver Twist in terms of textual meta-discourse resources including transition markers, frame markers, endophoric markers, evidential, and code glosses. They studied how the writers of original and simplified versions made use of these resources to organize their texts and found no significant difference between the two versions. In contrast to Esmaili and Sadeghi (2012), the present study compares an original novel, Charles Dickens' Oliver Twist (OT), with two simplified versions prepared by different narrators and ijel.ccsenet.org International Journal of English Linguistics Vol. 10, No. 6;2020 80 focuses on interactional meta-discourse resources including boosters, attitude markers, self-mentions, engagement markers, and hedges. The two simplified versions are compared in terms of interactional meta-discourse resources (IMR) to see how narrators try to interact with the text and readers. Hence, the following research questions are posed: 1) Is there any significant difference between the original version of OT and its simplified versions in terms of the frequency of using interactional meta-discourse resources?
2) Is there any significant difference between the simplified versions of OT narrated by two different narrators in terms of the frequency of using interactional meta-discourse resources?
3) What similar and different IMRs are used in the original and simplified versions and is there any significant difference in the categorical distribution of IMRs in the original and simplified versions of OT? 4) What similar and different IMRs are used in the simplified versions of OT narrated by two different narrators and is there any significant difference in the categorical distribution of these resources in both simplified versions?

What is Meta-Discourse?
In the 1970s, there was a shift in the study of a text from formal aspects of writing to the overall structure of discourse. In 1990, Swales introduced a deeper approach in which not only the formal aspects but also communicative features were taken into consideration. The reaction against putting too much emphasis on the propositional meaning of a text leads to the emergence of new perspectives, including meta-discourse. Meta-discourse resources are "self-reflective linguistic material referring to the evolving text and to the writer and imagined reader of that text" (Hyland & Tse, 2004, p. 156). Meta-discourse proves the presence of an author who organizes and assesses the content rather than the subject matter (Vandekopple, 1985). To produce comprehensible texts two elements are required, one is integrating speech about the experiential world and the other is making speech understandable and persuasive to the audience (Hyland & Tse, 2004). Thus, the aim of producing a text is not just transferring information but having degrees of certainty that it is comprehensible (Hyland, 2004). Writers or speakers attempt to attract readers or audiences' attention and encourage them to follow along. To have effective communication, they need to be aware of their receivers' needs and engage them in the texts. Writing or speaking is; therefore, viewed as a social and communicative process between writers/speakers and readers/listeners (Hyland, 2004;Hyland, 2005;Hyland & Tse, 2004).
Meta-discourse is based on the assumption that language is not just used for the ideational meaning (information about the world), rather other supplementary linguistic meanings also need to be taken into account (Degand & Hempel, 2008). Crismore (1983) also believes that all informative texts have propositional content, the ideational part can be called the primary discourse and the other level, the content-less level, called meta-discourse. Some scholars believe that there are different levels of meaning in a text; propositional and meta-discursive levels, for example, Vande Kopple (1985) notes that: Many discourses have at least two levels. On one level, we supply information about the subject of our text. On this level, we expand propositional content. On the other level, the level of meta-discourse, we do not add propositional material but help our receivers organize, classify, interpret, evaluate, and react to such material. Meta-discourse, therefore, is discourse about discourse or communication about communication (p. 83).
The oral and written mode of language is not neutral rather manifests the author's status, their viewpoint, and the interest. Thus, those who use language to make meaning should be cautious about the social effect that their speech is going to have on the audience (Hyland, 2019). This mutual understanding between writer/speaker and reader/listener can be accomplished using meta-discourse resources. Some scholars (e.g., Hyland, 2005;Mauranen, 1993) believe that meta-discourse can have considerable implications for language classes, specifically for language teachers, in terms of the social aspect.

Classification of Metalinguistic Devices
Literature comprises a number of different meta-discourse classifications (Crismore, 1989;Hyland, 1998;Vandekopple, 2002). In Systemic Functional Linguistics (SFL), there are three meta-functions for language (Hyland, 2005). Based on Halliday's (1994) point of view, people use language to fulfill three macro functions, talking about their experience, interacting with each other, and producing coherent discourses. In other words, Halliday asserts that people's messages can carry three different meanings: ideational, interpersonal, and textual.
• The interpersonal function: language is used to interact and understand each other in the mutual exchange of feelings.
• The textual function: language is used to organize a text and relate it to what is said to the world and the readers.
To focus on Halliday's tripartite concept of meta-functions, most scholars followed the model proposed by Crismore, Markkanen, and Steffensen (1993) in which they made a distinction between ideational, textual, and interpersonal functions. Although Crismore et al.'s classification has had a profound impact on meta-discourse studies, it has its own limitations and critics. Theoretically, Halliday's terminology led to the division of meta-discourse into textual and interpersonal functions; however, practically it was difficult to distinguish between these two functions of meta-discourse (Hyland, 2004). Therefore, Hyland (1998) modified Crismore et al.'s model of meta-discourse and proposed his thorough model including two main functions: textual and interpersonal. Hyland (1999) defines textual meta-discourse as those resources employed to produce coherent propositional information for the audience in a particular context whereas interpersonal meta-discourse enables the writers to present their views toward the propositional content and the addressees.
In 2004, Hyland and Tse considered all meta-discourse resources as interpersonal. Attempting to find a connection between the elements of a text, the author should be aware of self, the possible readers and the readers' responses; therefore, the author's selection of textual devices is an interpersonal decision to highlight particular connections in the text, which accommodate the readers' understanding and the author's expected interpretation (Hyland, 2005). The interpersonal feature of meta-discourse comprises both organizational and evaluative features (Hyland, 2001). Borrowing Thompson's (2001) terms, Hyland (2017) suggested that interactive and interactional meta-discourse should be used instead of textual and interpersonal ones. This model of meta-discourse entails two sub-categories: Interactive resources that allow the writer to organize the information and provide their preferred interpretations. These resources are composed of five categories: Transitions markers, Frame markers, Endophoric markers, Evidentials, and Code glosses. According to Hyland and Tse (2004), the interactional resources assist writers in identifying their level of personality in a script and make an appropriate connection using the following categories|: 1) Hedges: the writer's lack of enthusiasm to express the propositional meaning, for instance by using words as about, perhaps, etc.
2) Boosters: a demonstration of the writer's certainty using devices such as it is clear that, definitely, etc.
3) Attitude markers: a representation of the writer's affective mode. I agree, surprisingly are examples. 4) Engagement markers: addressing the readers directly or connecting with the readers using devices such as you can see that, note that, consider, etc. 5) Self-mentions: referring to the explicit presence of the author in the text in terms of first-person pronouns and possessives. Some examples are I, we, our, my, etc.
Although many studies have been conducted on the differences among writers in terms of using meta-discourse resources in academic writings; few studies have addressed literary works in this context. AlJazrawi and AlJazrawi (2019) investigated the use of meta-discourse resources in different short stories to investigate the persuasion achieved via the use of such resources. They pointed out that writers of selected short stories used meta-discourse resources frequently to make their stories both coherent and persuasive and they attributed the persuasion in fictional narratives not only to non-linguistic factors but also to linguistic ones such as the use of meta-discourse markers. Ahangari and Kazemi (2014) conducted a study in terms of the frequency of interactional and interactive meta-discourse resources in a novel noting no significant difference in their frequency and attributing the frequency of these resources to the awareness of the author of the norms of writing. They mentioned that all types of meta-discourse items are used more frequently in literary texts compared to other genres and this large frequency can be attributed to the degree of openness in these texts. Esmaili and Sadeghi (2012) investigated the frequency and categorical distribution of textual meta-discourse resources in the original and simplified versions of a novel. They revealed no significant difference in the use of these resources, implying that in both original and simplified versions, writers and narrators aimed to pen a coherent text. In another study, Sadeghi and Esmaili (2012) compared the textual Meta-discourse resources applied in two original novels (Tess of the D'Urbervilles and Wuthering Heights) and their simplified versions and the findings revealed no significant difference between the pairs.

Methodology
To disclose the elements of meta-discourse, the researcher has applied Hyland's (2005) taxonomy. Hyland's Interpersonal model comprises two broad categories, interactive and interactional. This study aims to focus on interactional resources including boosters, attitude markers, self-mentions, engagement markers, and hedges.

Materials and Data Selection Criteria
The data for this study came from the original and the simplified versions of the Dickensian novel (Oliver Twist). The original novel, 641 pages, was written in 1838 and the simplified versions, a 106-page version narrated by Harry Kaste (2003), and the other version is 41 pages by Kate Dickinson Sweetser (2005).
British literature is rich in both themes and linguistics features. It has seen various phases from old English to contemporary English. One of the most important phases of English literature was Victorian literature, which evolved during the time of Queen Victoria (Benzoukh, 2012). Victorian literature specifically Victorian novels are of utmost importance as they have dealt with social realities directly and have not followed the principle of art for art's sake and focused on moral purposes. Therefore, Victorian novels have had a profound effect on modern literature (Timeline: Victorian Era, 2020). This was the motive behind selecting a 19th-century Victorian novel. Charles Dickens, whose best-selling book was Oliver Twist, was one of the most famous novelists of the Victorian era. Furthermore, Oliver Twist is a popular novel and most people familiar with English literature have heard of this novel. The availability of original and simplified versions was another criterion. Multiple adaptions of this novel have been created to make it more accessible for different audiences. The original and selected simplified versions were analyzed in terms of the frequency and categorical distribution of interactional meta-discourse resources (IMRs).

Procedures of Data Analysis
Hyland's (2005) model was used in order to identify IMRs, that is, a list was gathered that included these resources and classified them into five categories of analysis (hedges, boosters, attitude markers, engagement markers, and self-mentions). Meta-discourse is heavily reliant on the context of occurrence. As has been emphasized by Hyland (1996), there is not a single interpretation for a specific device. Therefore, the investigation of all resources in this study was carried out qualitatively based on context. After determining the frequency and classification of IMRs into five categories, the total word count in each version was also identified. Item counting was conducted by the Microsoft Office Word software. It's clear that the word count in the original novel is much more than the simplified versions; therefore, the frequency of IMRs was calculated per 1,000 words. To find out whether the frequency of occurrences and the categorical distribution of IMRs were significantly different in the original and simplified literary texts, Chi-square statistical analysis was used in which the significance level alpha was set at .05. In Table 1  He insists that if it becomes evident that Oliver is in fact "a real and thorough bad one," he will refrain from further efforts to help the boy (simplified OT by Kaste). Attitude markers


Unfortunately for, the experimental philosophy of the female to whose protecting care Oliver Twist was delivered over, a similar result usually attended the operation of HER system (original OT).  They hopefully conclude that he has left the country, leaving the dog behind (simplified OT by Kaste).  The attempted burglary had greatly shocked them both (simplified OT by Sweetser). Engagement markers


Hunger and recent ill-usage are great assistants if you want to cry; and Oliver cried very naturally indeed (original OT).  Weak still, and stupified by the suddenness of the attack, overpowered and helpless, what could one poor child do? (simplified OT by Sweetser).  The boy is obviously ill, but the sadistic judge forbids anyone to support him (simplified OT by Kaste). Self-mentions  Among other public buildings in a certain town, which for many reasons it will be prudent to refrain from mentioning, and to which I will assign no fictitious name, there is one anciently common to most towns, great or small (original OT).  Returning to Mr. Bumble, we find him still waiting in Mrs. Corney's room (simplified OT by Kaste).

Results
To find out the differences between the original and simplified versions of OT in the frequency of use and the categorical distribution of IMRs, first their distribution in the original and simplified versions, and then their frequency per 1,000 words was calculated.

Distribution of IMRs in the Original and Simplified Versions of OT
The critical value of with 1 degree of freedom is 3.84 for the 0.05 level, the observed value of was 6.49, which was larger than the critical value of . This finding reveals that the difference between the original and simplified versions of OT in the use of IMRs is statistically significant. Therefore, the first null hypothesis, there is no significant difference between the original version of OT and its simplified versions in terms of the frequency of interactional meta-discourse use, is rejected.

Distribution of IMRs in the Simplified Versions of OT Narrated by Two Narrators
As shown in Table 2, the total number of words and the frequency of IMRs in the simplified versions of OT narrated by Kaste and Sweetser have been identified. To be able to find the frequent use of IMRs in the simplified versions narrated by two narrators and answer the relevant null hypothesis, there is no significant difference between the simplified versions of OT narrated by two different narrators in the frequency of interactional meta-discourse resources, the Chi-square test was used. The obtained observed value of was 2.81, which was smaller than the critical value, revealing no significant difference between them, thus the null hypothesis is accepted.

Categorical Distribution of IMRs in the Original and Simplified Versions of OT
To discover the distribution of five categories of IMRs in the original and simplified versions of OT, their frequency in each category per 1,000 words and their percentage were calculated. Table 3 shows the distribution of these five categories in both original and simplified versions of OT. In order to provide a definite answer to the third research question stated in Section 1, the researcher calculated the average frequency of the categories of IMRs for the simplified versions narrated by both narrators. Their frequencies (based on the scale of 1000 words) can be compared with those of the original novel as shown in Table 4. In the original version, two IMRs with the highest frequency rate are engagement markers with the frequency rate of 23.19 and hedges 15.46 per 1000 words. The result is the same in the case of the simplified versions, and the highest frequency belongs to engagement markers and hedges, 13.40 and 10.72 respectively. Thus, the most frequently used category of IMRs in both original and simplified versions of OT are engagement markers. The average frequency rate of attitude markers and boosters are almost identical in both original and simplified versions, 12.06 and 11.05 in the original and, 6.84 and 5.96 in the simplified version, respectively. The use of self-mention resources in both the original and the simplified versions is negligible.
To understand the categorical distribution of IMRs in the original and simplified versions of OT, the Chi-square test was utilized. The calculated observed value of is 0.26, which is less than the critical value of , indicating no significant difference in the type of meta-discourse employed in them. Thus, the null hypothesis is accepted; there is no significant difference in the categorical distribution of IMRs between the original and simplified versions of OT.

Observed
= 0.26< critical = 9.48 Table 5 shows the distribution of IMRs in the simplified versions of OT narrated by two narrators. As was mentioned before there were only a few cases of self-mentions in both simplified versions. Engagement markers with the rate of 10.56 in the simplified version by Kaste and 16.24 by Sweetser are the most frequently used category of IMRs. In the case of hedges and boosters, both of the simplified versions have almost the same frequency while there are variations in the case of attitude markers, in the simplified version narrated by Sweetser, its frequency is 10.15 while in the other version by Kaste it is 3.54. After applying the Chi-square test, the results revealed that the observed value of is 2.03 which is smaller than the critical value of ; therefore, there is no significant difference in the categorical distribution of IMRs in the simplified versions narrated by both narrators. And the final null hypothesis is accepted; there is no ijel.ccsenet.org International Journal of English Linguistics Vol. 10, No. 6; 2020 85 significant difference in the categorical distribution of these resources between both simplified versions. Observed = 2.03 < critical = 9.48

Discussion
The findings of this study have made some insights into the use of meta-discourse resources used by the author of an original novel and the narrators of its simplified versions. Firstly, the frequency of IMRs was significantly different in the original novel and its simplified versions. According to Lingling and Yipei (2013), delivering a speech is an interpersonal activity, and whether a speech succeeds in conveying a message or not depends on the way the speaker intrudes into the interaction and attracts the audience's interest and enthusiasm. This is the persuasion that partly can be achieved using meta-discourse. Based on the findings, the writer of the original novel used more IMRs in comparison to the simplified versions revealing that the writer of the original novel tries to insert himself into the text more to signal his communicative intentions and to affect the ways the readers understand these intentions. Despite having more complex syntax and semantics, the original novel seems to be more persuasive, at least in terms of IMRs, compared to its simplified counterparts. Using more interactional elements, Dickens provides a closer rapport with the readers and persuades them to remain more engaged with the story. Less frequent use of IMRs in the simplified stories can be attributed to the essence of simplification since the narrators are aware of the codes of simplification and they try to adopt an impersonal style.
Bax, Nakatsuhara, and Waller (2019) claim that in academic writings, there is a threshold level for using meta-discourse markers beyond which fewer meta-discourse resources will be used. That is, at the beginning levels more of these resources will be used while advanced writers use fewer, this decline might be due to using more sophisticated words to write a text instead of relying on explicit markers. This is in contrast with findings in literary texts, where in the original novel despite having more sophisticated words, more meta-discourse resources were observed compared to its simplified versions. Thus, EFL and ESL teachers who are concerned about applying authentic literary texts in their classes should recognize that despite the difficulties caused by using complex syntax, complicated semantics, and figurative speech, the authors try to engage readers by applying interactional meta-discourse resources. In terms of the simplified versions of OT narrated by different narrators, there was no significant difference between them. This means that both narrators attempted to engage readers in the conversation in the same way.
Regarding the categorical distribution, the writers of the original and the simplified versions showed similarities.
In both original and simplified versions, the engagement marker was the most frequently used category. Hyland (2005) notes that engagement markers promote readers' participation with two main purposes. One is to satisfy the readers' expectations to be included in the argument and the other is to attract the readers to the discourse and to guide them to have appropriate interpretation. Being identical in the case of engagement markers, both original and simplified versions successfully involve the audience in the discourse. Whenever the audience remains involved in the discourse, they will be more careful about the speakers' words (Lingling & Yipei, 2013). For the simplified versions narrated by two narrators, we have the same point; in general, there is no significant difference between them in the used categories. The engagement marker has the highest frequency in both versions showing both narrators involve the audience into the discourse in a similar manner, although the frequency of the engagement marker used by Sweetser is slightly higher than that by Kaste. Thus, both narrators make an effort to involve their readers as participants in the text.
After the engagement markers, hedges have the highest frequency in both original and simplified versions, and both narrators have used them identically. On the one hand, a number of scholars (Crismore et al., 1993;Clifton, Gale, Mcgrath & McMillan, 1977) attributed the overuse of hedges on the part of writers to their interest in uncertainty and writer-reader interaction. On the other hand, Lingling and Yipei (2013) argue that the low frequency of hedges indicates the writer's purpose to convince and persuade. They add that using boosters without hedges seems too tough, while hedges alone appear soft. Therefore, it is important to balance the use of boosters and hedges. The frequent occurrence of hedges can be attributed to the nature of literary texts. Literature as a branch of art is personal and subjective, thus the interpretation that is made in fiction is based on personal judgments (Kondowe, 2014). Both the writer and narrators of OT are aware of the subjective essence of the literature, thus they decrease the risk of being opposed.
The frequency of attitude markers and boosters in both original and simplified versions are also similar. According to Hyland (2005), attitude markers "indicate the writer's affective, rather than epistemic attitude to propositions. They convey surprise, agreement, importance, obligation, frustration, and so on" (p. 53). It seems that there is no difference in the use of attitude markers in both original and simplified versions. While in the case of two narrators, Sweetser tended to use the attitude markers more frequently compared to Kaste, its use may be attributed to gender. Kashani and Yavari (2013) attributed more enthusiasm in using these resources to females and point out the willingness of females to make their discourse affective. A possible interpretation is that female writers tended to be more concerned about paying attention to the evaluation of the discourse using attitudinal lexis.
The frequency of boosters in both original and simplified versions, as well as in the simplified versions by both narrators, is somehow similar. According to Hyland (2005), "boosters suggest that the writer recognizes potentially diverse positions but has chosen to narrow this diversity rather than enlarge it, confronting alternatives with a single, confident voice" (p. 52). The use of boosters shows confidence and commitment to what you are saying and if you want to persuade others, you have to be very sure about your statements, and this is something that can be fulfilled using boosters (Kashani & Yavari, 2013). Writers should use boosters when they are sure about the certainty of their claims, and due to the subjective nature of literature, they prefer not to overuse them to provide more space for opposing ideas (Kondowe, 2014).
The use of self-mention is very low in both original and simplified versions, and in the case of the simplified versions, it is so negligible that it can be ignored. According to Hyland (2001), "self-mention is used to address the readers directly through a firm alignment with their views, pledging certainty, and an interpersonal assurance of conviction" (p. 221). Self-mentions provide a clear structure for the discourse. They explicitly state the goal of the text and give this chance to the writers to promote themselves and their stance (Hyland, 2001). Despite all their advantages, their low frequency in literary works can be attributed to the literary genre, because writers want to create a story and the narrators try to narrate a story, thus they cannot directly mention themselves in the text.

Conclusion
Literary input can improve the linguistic and communicative ability of learners, in terms of both spoken and written language, through engaging learners and providing a personal reaction to the text (Fraser, 2018). Due to the importance of literary texts as reliable sources to be used in the ESL and EFL classes, the present study aimed at comparing the original and simplified counterparts of a novel in terms of IMRs. The focus was on both the frequency and categorical distribution of IMRs used in both the original novel and its simplified versions. Meta-discourse Theory provides a way for mutual comprehension and involvement between the speaker and the listener, and interactional markers embody all usages of language to express one's opinion, influence one's behavior, interact with, and maintain the relationships with others (Kashani & Yavari, 2013). Thus, using these resources, authors of literary texts try to build a good relationship with the audience and facilitate mutual communication and comprehension.
The writer and narrators of both original and simplified versions of Oliver Twist used IMRs in their manuscripts, demonstrating that it is impossible to make rapport with the audience without such resources. While the researcher did not find any difference in the categorical distribution of IMRs, the frequency of such resources in the original novel was much more than the simplified versions, revealing that despite being more complex due to the lexical density and complicated syntax, the original novel seems to be more appealing to the learners. Syntax and semantics are two bold issues in comprehension; however, they are not the sole elements that facilitate it. IMRs are essential elements used for persuasion; whose frequent occurrence in the original novel can act as a facilitator for comprehension. Thus, teachers and material developers should not just rely on syntax and semantics for exclusion or inclusion of authentic texts in language learning and other effective resources as IMRs should be taken into consideration.

Implication
By way of conclusion, some pedagogical implications can be made from the findings of this study. Due to the importance of meta-discourse in satisfying the rhetorical function in the literary genre, the teachers and material developers should pay attention to them pedagogically, specifically in terms of interactional meta-discourse, which plays a crucial role in persuasion. As the findings demonstrate, the frequency of interactional meta-discourse resources in the original novel is higher than the simplified versions and though the language of a novel is more difficult both syntactically and semantically, we can claim that it is persuasive due to the use of such resources. Therefore, considering learners' level of proficiency, teachers may apply original novels as a kind of authentic literary texts either in the classroom or as extensive reading outside of the classroom.
The other related implication is for narrators or writers of short stories and those who want to simplify original authentic texts. Interactional resources can be viewed as powerful tools that equip writers to make a close rapport with their readers. Dealing with the literary genre, authors, writers, and narrators should know that communication is not merely for the exchange of the information, rather it is to accompany readers in an imaginary journey created by them. Thus, for short simplified stories to have the same degree of persuasion, the use of these linguistic devices should be taken into account. Finally, due to the importance of such resources, EFL/ESL teachers should encourage learners to use them. To do that, strategies should be adopted and undoubtedly material developers can be very helpful in this regard by allocating a part of the curriculum to such resources and providing guidance for learners on how to apply them in their manuscripts.