GENDER DIFFERENCES AND THE USE OF METADISCOURSE MARKERS IN WRITING ESSAYS

This study aims at investigating metadiscourse markers in academic essays written by male and female students. With the goal in mind, 20 essays written by EFL female students and 20 essays written by EFL male students were chosen. These data were analyzed based on the metadiscourse framework proposed by Hyland (2005). The findings revealed that both male and female writers employed more interactive markers than interactional markers due to the fact that both genders inserted transition markers frequently to guide readers through the texts. However, a cultural factor may influence the writers’ tendency not to use transition markers showing arguments. Furthermore, although male students employed more interactional markers, female students used more markers in interactional sub-categories except self-mentions. Although it is assumed that women prefer to use a more personalized style, male writers in this study also personalized their essays by using self-mentions. One possible reason was that the use of these features tended to be more field-specific than gender-specific. The findings and discussion indicated that gender is not the only factor influencing the use of metadiscourse markers. Other possible variables discussed in this study should be taken into account.


Introduction
Studies have suggested that men and women tend to favor distinct language features in expressing themselves (Lakoff 1975, Waskita, 2008, Matei, 2011, Subon, 2013, and Shirzad & Jamali, 2013. Lakoff (1975) revealed that women used linguistic features that reinforced their subordination. She further revealed that they were inclined to use some specific language features, such as lexical hedges, tag questions, empty adjectives, intensifiers, and emphatic stress. These differences between men and women in communication have attracted on-going scholarly discussion. Some studies have extensively explored how gender differences influence both spoken (Subon, 2013& Matei, 2011 and written language (Shirzad &Jamali, 2013 andWaskita, 2008). Besides claiming that women's language tended to be more polite (Subon 2013) and more complex (Shirzad & Jamali, 2013;Waskita 2008), Matei (2011) suggested that women had the tendency to use more discourse markers in spoken communication.
A previous study on gender and discourse markers by Pasaribu (2017) showed that both genders shared similar patterns, in which they tended to use more elaborative markers than contrastive markers. However, the study was limited to the use of discourse markers or the textual markers used by the students. While the framework of discourse markers elaborates the relations between sentences, clauses, and phrases, the theory of metadiscourse markers employed in this current study covers both the interactive and interactional dimensions of the data. Hyland (2005) articulated that the concept of metadiscourse markers "is the cover term for the self-reflective expressions used to negotiate interactional meanings in a text, assisting the writer (or speaker) to express a viewpoint and engage with readers as members of a particular community" (p.37). Hyland and Tse (2004, p. 156) elaborated that the use of metadiscourse markers expressed "communicative engagement between the writers and readers". This model does not only investigate interactive markers which are employed to assist writers to organize the discourse, but it also covers interactional markers which enable writers to highlight some aspects in the discourse and project their attitudes.
Some literature has investigated gender-based differences and the use of metadiscourse markers. Tse and Hyland (2008) pointed out that the linguistic features that male and female researchers used are not merely determined by gender. On the other hand, Yeganeh and Ghoreyshi (2015) revealed that gender differences play an important part on the use of two metadiscourse features, booster and hedges. They argued that Iranian females preferred to use hedges in their writing, while the males tended to use boosters more frequently. The tendency occurred possibly because women "were more cautious in writing and reporting their opinions" (Yeganeh & Ghoreyshi, 2015, p. 688). A recent study by Seyyedrezaie and Vahedi (2017) investigated the projection of gender identity through metadiscourse marking. They found out that both although males and females writers shared the same patterns of using stance makers, it turned out that the male writers used more frequent epistemic markers than their counterparts. It was interpreted that the male authors expressed more certainty in their writings. The distinct findings from these studies showed that further research in the field of gender and metadiscourse markers should be conducted. Therefore, this study aimed at elaborating the gender differences and the use of metadiscourse markers in 40 academic essays written by male and female students.
Theory Hyland (2005) applied the term metadiscourse markers to highlight the use of markers in written form. He elaborated that writers use metadiscourse markers as a set of tool "to negotiate interactional meanings in a text" (p. 37). They help the readers see the writer's perspectives. In his perspective, the use of metadiscourse markers encourages the relationship between the writer and readers. For example, the function of attitude markers, such as unfortunately and surprisingly, express the writer's attitude toward the issues presented in the text.
He classifies metadiscourse markers into two categories, namely: interactive and interactional dimension. Hyland (2005) elaborated that the interactive dimension "concerns the writer's awareness of a participating audience and the ways he or she seeks to accommodate its probable knowledge, interest, rhetorical expectations and processing abilities" (p. 49). The resources in this dimension serve as tools to organize information which meets the readers' need. Table 1 projects the five broad sub-categories of this interactive dimension which was taken without modification from Hyland, 2005, p.  Another dimension proposed by Hyland (2005) is the interactional categories which concern on how the writers present "interaction by intruding and commenting on their message" (p. 49). There are five sub-categories of interactional dimension which was taken without modification from Hyland, 2005, p. 49. The aim of this study is investigating the use of both interactive and interactional markers in EFL academic essays. With this goal in mind, the study analyzed 40 academic essays by employing Hyland's theory of metadiscourse markers.

Method
This study elaborated the differences and the similarities in the use of metadiscourse markers between female and male students' essays through document analysis (Lodico, Spaulding & Voegtle, 2006). The data were collected from 20 essays written by female students and 20 essays written by male students in Critical Reading and Writing I.
The researcher carefully identified the metadiscourse markers which were classified into Hedges (Hg), Boosters (Bt), Attitude Markers (Am), Self-mention (Sm), and engagement markers (Em). The markers were highlighted, counted using Antconc, a free concordance and analyzed by drawing on Hyland's model of metadiscourse markers (2005). The collected data were elaborated using descriptive qualitative approach. Both figures and numbers are explained through verbal means.

Findings and Discussion
The findings indicated that both genders applied more interactive resources (1561) than the interactional ones (1406). Both groups are heavy users of transition markers and hedges. Furthermore, males used these resources more than females. This section elaborates the phenomena in details. Table 3 shows that both sexes use 1561 interactive metadiscourse markers. Both groups had the tendency to use transition markers. Female students used 574 markers or 36.8% of 1561 interactive resources, while male students used 667 markers (42.7%). The table also shows that both genders only used the endophoric markers once. The most frequent feature of interactive markers as shown in table 3 is the transition markers. The model suggested by Hyland (2005) classifies transition markers into three distinct categories: addition, consequence, and argument. In line with Pasaribu's findings (2017), essays written by male and female students share similar patterns.  Table 4 shows that addition markers are the most common transition markers. The variants of these markers are and, furthermore, moreover, in addition, besides, likewise, in the same way, in contrast, however, but, yet, although, on the contrary, on the other hand, and despite. Both females and males students had the tendency to use addition markers. Next, both genders also used markers to show consequences, such as thus, therefore, consequently, as a result, for this reason, hence, we can conclude, since, because, because of, and so. Finally, both sexes also tended not to use markers to signal argument. A few variants are used such as in any case, in this case, and of course. The data indicated that the writers preferred adding more information to arguing the ideas. One possible cultural explanation for this preference was due to the fact that Indonesian people tend to avoid arguments. One of the famous sayings integrated with the students' characteristic is "manut lan pinurut" or to obey and to follow (Dardjowidjojo, 2006). The belief that obeying is more encouraged than arguing can also be expressed from the students' tendency not to employ markers showing arguments.
The writers also employed frame markers serving as signals of text boundaries (Hyland, 2005). There are three types of frame markers, namely: additive relations, discourse goals, and topic shifts. Although male writers tended to use more frame markers, both sexes were inclined to apply additive relations rather than explicitly state their purpose in writing by the use of discourse goals (0.3%). Hyland (2005) elaborated that some frame markers show additive relations between sentences or groups of sentences. The variants of additive relations found in the essays are first, second, third, at the same time, and next. Besides using markers to show additive relations, the EFL writers also used several interactive markers such as well, right and now to indicate topic shifts. The EFL writers tended to elaborate the relation between ideas rather than announcing the discourse goals. Some discourse goals found in the essays are I argue and I agree that. Although they knew that the essays they wrote indicated causal relationship and arguments related to particular issues, they had the tendency not to express their purposes explicitly by employing only a few markers expressing discourse goals. The students did not explicitly mention their purposes probably due to their preference in using an Asian rhetorical model, whose development of texts is not written straightforwardly (Wahab, 2006) It is interesting to note that males use more markers in interactive dimensions except code glosses and evidentials. Writers use code glosses to express "additional information by rephrasing explaining or elaborating what has been said" (Hyland, 2005, p. 52). Writers add more code glosses by giving definition or using examples. In this study, female students used more code glosses than male students. The writers provided definition to clarify the issues presented in the texts. The markers used to indicate definitions in the essays are this is called, in other words, and that is. The writers support their opinion not only by giving definitions, but also by providing examples. In supplying additional information, female students were heavy users of examples as seen in table 5. The writers introduced examples by using various markers such as for example, for instance, such as, and like. The examples were given to make the writers ideas more concrete for the readers.
Both genders also used markers as evidentials from other sources to support their arguments. However, female students tend to use more evidentials as sources to support their arguments. As argued by Yeganeh and Ghoreyshi (2015), women had the tendency to be more careful in writing by citing others or giving examples. The variants of this metalinguistic representations are according to x, x states, x noted, x said and x mentions. Writers also introduced other parts of the text by using endophoric markers. Endophoric markers are phrases which refer to earlier material or something yet to come (Hyland, 2005). The EFL writers do not use many endophoric markers. Some of references to others found in the essays are what is mentioned and like I said before as seen in table 7.
As shown in table 8, both genders have the tendency to use engagement markers. By using engagement markers, the students engaged with the readers. It is turned out that although male students tended to use the overall interactional metadiscourse markers, female students employed more categories of interactional markers than male students except self-mention.  Table 8 reveals that both groups applied a considerably high use of engagement markers which involve features that address the discourse participants (Hyland, 2005). Writers acknowledge participants by using pronouns indicating first person plural (we, our, us) or second person pronoun (you, your). As seen in table 9, female students use more engagement markers (29.9%) than male students (28.4%). The other purpose of using engagement markers is positioning readers into the discourse (Hyland, 2005). The students did this by giving directives using modals like should, must or have to. There is no clear indication that women express request or command using less direct manner than men because both genders shared similar variants of directive modals.
Furthermore, Lakoff (1975) introduced hedges to describe words which make things more or less blurry. Moreover, Hyland (2005) mentioned that this feature emphasizes the writer's subjectivity. The information is presented as an opinion which is open for negotiation. This feature also implies the degree of confidence and certainty. It is also used to convey indirectness. The variants of hedges found in the essays are possible, might, perhaps, usually, sometimes, almost, likely, tend to, should, may, a little bit, kind of, at least, and maybe. Hedges tended to be more common in essays written by female students. Female students presented more hedges (122 times or 8.8%) than males (100 times or 7.2%). This finding confirms the previous study by Yeganeh and Ghoreyshi (2014) in which they revealed that females were more careful in presenting their opinions or arguments. Different from hedges, boosters highlight certainty. The use of boosters represents a confident voice (Hyland, 2005). He adds that boosters are "widely used by chairpersons to demonstrate a confident image" (p. 79). The variants of boosters found in the essays are clearly, actually, certainly, really, always, definitely, and in fact. The results show that the female students were more likely to use boosters (40 or 3%) than their counterparts (33 or 2.5%). This suggests that female students are more inclined to present higher degree of assurance. This result contradicts Yeganeh and Ghoreyshi (2014), who found out that males tended to use more boosters. Meanwhile, the current finding is in line with Serholt (2011) who revealed that females were more inclined to use boosters.
Another common feature is self-mention markers which "refer to the degree of explicit author presence" (Hyland, 2005, p. 53). The presence of the authors in this study is reflected by the use of first person pronouns (I, my, and me) and first person plural pronouns (we, us, and ours). Male students were more inclined to use this feature to personalize the ideas in the essays. According to Hyland (2005) this feature served several functions such as strengthening the writers' presence, including personal narratives or experience, and promoting solidarity. Usually personalized writing style is often favored by women (Goodwin 1988 in D'angelo 2008). However, male students also personalized their essays or made their presence noticed by using self-mentions. The use of this feature seemed to be influenced by a field-specific factor. According to Hyland (2005), self-mention markers were more frequent in humanities and social science papers than in science and engineering. In this case, it is likely that the use of self-mentions was influenced by the topic of the texts related to social issues.
The writers' engagement with the texts and the topics is also shown through the use of attitude markers. These features are used to project writers' attitude to ideas presented in the text. Hyland (2005) suggested that they expressed writers' affective attitude. The table shows that both sexes used attitude markers. Female students were slightly more inclined to use more attitude markers (38 times or 2.7%) than their male counterparts (33 times or 2.3%). The attitude markers found in the essays are : unfortunately, easily, fortunately, appropriate, strange, weird, negative, important, usual, amazing, correct, essential, interesting, shocked, shocking, surprised, unexpected, and unusual. By using these verbs, adjectives and adverbs which contained affective messages, the writers conveyed their attitude towards information and ideas presented in the text.

Conclusion
The study stems from the assumption that gender differences influence the use of metadiscourse markers. The study indicated that both groups share similarities, in which they employed more interactive markers than interactional markers. Both genders were heavy users of transition markers which help them connect ideas in the discourse. In using transition markers, students preferred elaborating ideas than arguing ideas because they were not accustomed to contrasting or arguing ideas. Furthermore, although male students use more interactional resources, female students use more markers in the sub-categories except self-mentions. Although it is assumed that women prefer to use a more personalized style, male writers in this study also made their presence noticed by using self-mentions. One possible reason was that the use of a personalized style tended to be more field-specific rather than gender-specific. It confirms Hyland's argument (2005) that writers often employ more self-mentions in humanities and social science papers than in science and engineering. The discussion indicates that gender is not the only factor that determines how writers express themselves. Further research focusing on the discipline of the essays and the cultural backgrounds of the writers is necessary in the future.