The role of gender in conversational dominance: A study of EFL learners

Abstract Gender differences have always been controversial in pragmatics. This article reports on a study that examined the role of gender in the quality of conversational dominance in informal conversation by focusing on Persian EFL learners. To this end, both quantitative and qualitative conversational data from 10 Iranian dyads were analyzed. In this study, conversational dominance is defined as one speaker’s tendency to control the other speaker’s conversational actions over the course of an interaction. Since the norms of speaking are strongly affected by gender, both male and female Iranian learners are constrained to gender-specific modes of interaction. The findings of this study revealed that women show greater acceptance in conversation and due to this feature they try to have more facilitative role in conversation, and men try to maintain dominance over topic by showing more assertive mode during stages of topic development and maintenance. Men by using different strategies like interrupting women, topic shifting, asking questions and raising topics, criticizing and engaging in conflict, and silence try to keep dominance over the conversation. Furthermore, men by being more self-oriented and women by being more other-oriented show varying degrees of dominance over the conversation. The implications of our findings for the use of dominance strategies to take the field of conversation in the EFL context are discussed.

Abstract: Gender differences have always been controversial in pragmatics. This article reports on a study that examined the role of gender in the quality of conversational dominance in informal conversation by focusing on Persian EFL learners. To this end, both quantitative and qualitative conversational data from 10 Iranian dyads were analyzed. In this study, conversational dominance is defined as one speaker's tendency to control the other speaker's conversational actions over the course of an interaction. Since the norms of speaking are strongly affected by gender, both male and female Iranian learners are constrained to gender-specific modes of interaction. The findings of this study revealed that women show greater acceptance in conversation and due to this feature they try to have more facilitative role in conversation, and men try to maintain dominance over topic by showing more assertive mode during stages of topic development and maintenance. Men by using different strategies like interrupting women, topic shifting, asking questions and raising topics, criticizing and engaging in conflict, and silence try to keep dominance over the conversation. Furthermore, men by being more self-oriented and women by being more other-oriented show varying degrees of dominance over ABOUT THE AUTHORS Dr Maryam Pakzadian has received her PhD in Applied Linguistics (TEFL) from Allameh Tabataba'i University of Tehran, Iran in 2017. She has taught different undergraduate and graduate TEFL courses in Iranian universities since 2009. Her areas of interest include Discourse Analysis and Interlanguage Pragmatics, gender studies, EIL, identity, and motivation.
Dr Arezoo Ashoori Tootkaboni has received her PhD in TEFL from Allameh Tabataba'i University of Tehran, Iran in 2018. As for her professional background, she has been a lecturer, running Language Teaching Methodology, Language Testing, Reading Comprehension, Speaking and Listening and General English courses at state universities in Iran since 2011. Moreover, she has been involved in research on language learning and teaching issues, and published a number of articles in international TEFL journals. Note: The current study is part of a larger project that investigates gender equality in EFL settings in Iran.

PUBLIC INTEREST STATEMENT
This article investigated the role of gender in the quality of conversational dominance, as one speaker's tendency to control the other speaker's conversational actions, in informal conversation by focusing on Persian EFL learners, based on data gathered via quantitative and qualitative methods from 10 Iranian dyads. It was found that women show greater acceptance in conversation and due to this feature they try to have more facilitative role, and men try to maintain dominance over topic by showing more assertive mode during stages of topic development and maintenance. Men by using different strategies like interrupting women, topic shifting, asking questions and raising topics, criticizing and engaging in conflict, and silence try to keep dominance over the conversation. Moreover, men by being more self-oriented and women by being more other-oriented show varying degrees of dominance over the conversation. The outcomes can be used to benefit more from participation in L2 interactions.

Introduction
Studies of the effects of gender and social dominance on conversational interaction have proposed that the male dominance is displayed in a variety of ways in male-female conversation as a microinstitution, just as it is exhibited in macro-institutions (e.g., West & Garcia, 1988;Woods, 1988;Zimmerman & West, 1975). Studies on the gender dominance have found some discourse features of dominance in conversation, such as overlapping, interruption, topic shift, turn length, and distribution of number of turns, and quantitative studies have been conducted to trace the distribution of such features in conversational discourse. The main topic of interest is the hidden social structures, which may be presumed to underlie the conversation or text. Gender and discourse interface in many epistemological systems, and gender stereotyping is something common in different cultures.
One of the very interesting issues in gendered discourse and its associated conversation analysis is the issue of conversational dominance. As Linell and Luckmann (1991) present, the speaker who offers or requests a piece of information takes the discourse further, brings forward new aspects of the topic and tries to govern the contributions that follow. On the other hand, the speaker who accepts the given information or supplies the requested information makes his or her utterance "conditionally relevant" in the local context and complies with conditions that are already defined. This view suggests that one important aspect of topic development is how two speakers share the work of activating and supporting the production of coherent sequences of utterances. In many cultures men take the control of conversation from women and women accept their second role in society and even in conversation. Topic control includes the initiation of the topic, topic development, and topic change among other things. A change of the topic or lack of the topic development withholds the speaker the opportunity for continued evolution of one's thoughts. Although there is a dearth of research in this area, it is suggested that men usually try less to maintain the conversation flow and in most cases interrupt women's ideas in order to push their ideas forward.
Current research into gendered pragmatic and nonverbal aspects of power and dominance (e.g., Craig, 2016;Holmes & King, 2017;Keating, 2009;Tamuang, 2012;Tenorio, 2016;Wharton, 2009Wharton, , 2016Wilmot, 2017) emphasized on non-verbal dominance view of communication in the field of psychoanalysis. In the fields of pragmatics and discourse analysis, a number of valuable studies have been conducted on the role of gender on conversational dominance and reported consistent findings that men tend to dominate conversations (Finlay, 2015;Tannen, 1984Tannen, , 1985Tannen, , 1989Tannen, , 1990West & Zimmerman, 1983). However, majority of these studies have focused on the idea of men's dominance over the topic (Fishman, 1983;Itakura, 2001a;Itakura & Tsui, 2004;Spender, 1980;West & Garcia, 1988) and they do not talk much about the ways with which men gain this dominance. Therefore, the qualitative analysis should accompany quantitative data to find different verbal and nonverbal gender specific styles of taking the control of interaction as quantitative studies are not capable of revealing the nexus relationship between conversational dominance, style and associated features. Noting that the study of male-female differences in learner-learner interaction is an important but neglected area, the researchers here tries to investigate the issue of conversational dominance among Iranian EFL learners and tries to decode data based on the various strategies that they render to take the control of the conversation.

Gender in conversation
According to Luyt (2015, cited in Hellum & Oláh, 2018, gender is not a fixed stable individual attribute, but a noticeable sociocultural product of individuals' opinions, beliefs, attitudes and experiences. Shitemi (2009) in his lecture on language and gender proposed that traditional gender discourse has tended to focus on the dichotomy between male domination on one hand and gender separation and difference on the other as dialogue has continued to revolve around overt and covert aspects of gender practice and labeling. The research on language and gender tries to shed light on the ways people use linguistic forms and conversation strategies for construction, deconstruction and articulation of the gender phenomenon. Communication situations are loaded with "gender role" that is manifest in all manners of language use and cultural manifestations and practice. All these social and interactive instances are variously called upon, directly and/or indirectly to capture and present discourse on the gender structure. Gender, therefore, turns out to be embedded in all institutions, actions, beliefs and desires that go along with the mapping of language use through communication, interaction and establishment of the social order. Cameron (1990), in "Why language is a feminist issue?" describes how approaches to gender and language can be divided into dominance and difference approaches. Dominance approaches explain linguistic differences between men and women in terms of their relative socioeconomic status. Gass and Varonis (1989) examined male dominance among Japanese learners of English by examining the amount of talk, the distribution of number of topic initiations, overlaps and questions in L2 English dyadic conversation. Among other things, they found that Japanese male speakers in four mixed-gender dyads initiated more abrupt topic shifts than female speakers when they held conversations in English. They found that men were those who nominated new topics in order to maintain the conversation in case of silence or communication breakdown. In their study, they reported that throughout the conversation trend kept dominance over speech patterns. They illustrated the issue with resorting to cultural norms of the speakers' L1 and their influence on the notion of topic dominance. On the other hand, different approaches to the study of language and gender consider linguistic differences between men and women as manifestations of different conversational styles. Tannen (1990) maintains that female conversational style is characterized by a tendency to speak and hear language in terms of "connection and intimacy". Different male and female conversational styles are described as being sustained by strategic ways of using the same linguistic forms or discourse functions. For example, in terms of asking questions, Coates (1996) suggests that women tend to use "other-oriented" or "addressee oriented" questions strategically for conversational maintenance, i.e., to invite others into talk, extend the topic under discussion, and avoid playing the expert. On the other hand, men are seen as tending to use speaker-oriented questions, which are used to seek information, demonstrate power and expertise, and reinforce boundaries between speakers (Coates, 1987). Male speakers are more likely to use a strategy of creating a longer sequence of statements of their own in such a way as to discourage others from occupying center stage.
Considering the role of gender in conversations and reviewing different reports of how males and females used a variety of strategies to exercise their power over talk exchanges and gain the center stage paves the way to instantiate a pragmatic construct known as conversational dominance.

Conversational dominance
According to Itakura (2001b), conversational dominance is a multi-dimensional construct encompassing sequential, participatory and quantitative dimensions. The most important dimension, Sequential dominance, involves one speaker's tendency to control the other speaker with regard to the direction of the interaction and the allocation of initiating and responding roles.
Participatory dominance involves restriction of speaking rights, especially through interruption and overlap. And quantitative dominance is concerned with the extent of contribution to the interaction regarding the number of words spoken by each participant.
The ultimate conversational goal is sharing or equalizing knowledge between the participants (Linell & Luckmann, 1991), and coherence between utterances is created by one speaker offering a piece of information and the other speaker accepting it, or by one speaker requesting a piece of information and the other speaker supplying it (Berry, 1981). This view suggests that one important aspect of topic development is how two speakers share the work of activating and supporting the production of coherent sequences of utterances. For example, questions and statements assume the activating role, while answers and acknowledgments may be seen as playing a more supportive role. In this paper, the distribution of these roles is measured by the distribution of initiating moves and responding moves between the speakers. To capture dominance in this aspect of topic development in recorded conversational data, the system established by the Birmingham School of discourse analysis was adopted. According to this system, the basic unit of coherence in conversation is the exchange (Coulthard & Brazil, 1981). Linell et al. (1988, as cited in Itakura, 2001a) discussed that the dominance over the topic can be measured by using new content words such as concepts and referent. The dominant speaker is referred to as "the one who tries to put the most content into the socially shared world of discourse, he who places the most topics and subtopics on the floor." (p. 69). To investigate the issue, they compared the speakers in terms of the number of new content words (e.g., nouns, adjectives, adverbs) used by them. But this way for measuring topic dominance is problematic and it is more appropriate to investigate topic dominance in terms of a sustained control over the content of the conversation at level of words or moves that is explained in terms of sequential dominance.
On the other hand, we have the notion of participatory dominance which according to Itakura (2001a) involves patterns of asymmetry where "one speaker's holding on to the turn until completion following an interruption or overlap leads to other speaker's leaving the turn incomplete and therefore losing the chance to participate in the conversation." (p.80). All of these issues consider which participant takes the control of the topic. Topic control may include different features such as, the initiation of the topic, topic development, and topic change. When somebody takes the control of topic by changing the topic of conversation or prevents you from developing your ideas, you cannot put forward your thoughts anymore. Although research is limited in this area, it suggests that men do less work than women in maintaining conversations and more frequently cut off the development of women's ideas than the other way round. West and Garcia (1988), in their investigation of mixed-sex dyads reported that men were responsible for initiating 64% of topic changes. However, other similar studies on the conversations between strangers reported no gender differences in the number of topics initiated or developed for discussion. Fishman (1983) probed the failure and success of a range of topics by both males and females. She reported that women attempt more to sustain the conversation when they introduce a topic, but only 38% of the topics went ahead effectively when women themselves introduced the topic. There was a wide spread belief that women usually talk more than men but findings showed something else. The context in which the conversation happens in should be considered first and then the topic under discussion should be examined. Such ideas related to passive role of women in conversation unfortunately deeply rooted in the ideological framework of the various traditional cultures involved. It has been suggested that the person who controls the topic is the person who controls the interaction (Walker, 1987). Burgoon and Dunbar (2005) drew upon dyadic power theory (Dunbar, 2004) to investigate the dyadic nature of dominance and power, and the behavioral indicators of power in close relationships. They videotaped 97 couples while completing a problem-solving task. Then, they coded the tapes for verbal and nonverbal control strategies that used to be dominant in their conversations. The verbal and nonverbal control-based dominance strategies included dysfluencies, interruptions, recurrence of adaptor and illustrator gestures, vocal features, and general perceptions of dominance. The findings revealed that those who exhibited more dominance in their conversation with partner had more perception of power.
According to Zimmerman and West's (1975) dominance model, in mixed-sex conversations, men are likely to interrupt women more than the reverse in. They recorded 31 conversations between middle white class adults under 35 in the University of California. The results showed that 11 segments of conversations between men and women included 46 interruptions on the part of men and only 2 on the part of women. They concluded that since men interrupted more often, they were attempting to dominate.
In a very interesting study, Dunbar (2015) drew upon Waller's (1938) "the principle of least interest" to explain how power defines dominance in your conversations, and also in your relationships with opposite sex partner. Waller reported that even in dating couples what results in having successful conversation is interpersonal interdependence, and what may result in conversation failure is power imbalance and dependence. Those who are less emotionally involved and dependent on their partners are more likely to dominate and even end the conversations.
Overviewing the studies on conversational dominance and the ways it is constructed, maintained, and reconstructed by speakers with regard to context, one needs to know about verbal and nonverbal indicators of conversational dominance. These communicative strategies are used to take the field of conversation.

Verbal and nonverbal indicators of dominance
Many scholars discussed the importance of considering both verbal and nonverbal aspects of communication as complementary elements to have complete and reliable analysis of conversation (e.g., Jones & LeBaron, 2002). Although we advocate that all communication studies should include elements of both verbal and nonverbal cues, dominance, in particular, warrants such treatment. Since in this research the focus is on the verbal level we just consider this aspect.
At the verbal level, we can mention different techniques to alter the speech of our partner in a conversation, such as using offensive words, threatening, physical force, etc (Fitzpatrick & Winke, 1979;Klein & Johnson, 1997). In this vein, Canary andSpitzberg (1987, cited in Burgoon &Dunbar, 2005) named some of the strategies: integrative strategies that are collaborative in nature, distributive tactics that are competitive, and avoidance strategies that aim to avoid the conflict in conversation. The use of verbal influence strategies could be viewed as a type of controlling the speech of partner. Falbo andPeplau (1980, cited in Burgoon &Dunbar, 2005) introduced a twodimensional model of control strategies used in close conversations. The strategies go around two dimensions of "directness" and "bilaterality". Those strategies range from direct ones which seek to talk about desired goals and develop your wished subjects to indirect strategies which try to provide a good state for the target. The bilaterality deals with the interactivity in a conversation: to persuade your partner on one hand and step back from the conversation on the other hand. Typically, direct strategies are considered to be more dominant than indirect ones. Moreover, degree of dominance or submissiveness is related to who has the place to direct, delimit, and define the action of the dyadic interpersonal system. According to Keating (2009), speakers manifest power in diverse ways as they construct their own identities in response to the actions and behaviors of their partners. Holmes and King (2017) proposed that analyses of power should be dynamic since power is constructed, sustained and reasserted over the course of interaction. Tamuang (2012) investigated the recurrence of dominance in the Thai culture. He found that men interrupt women more often to show power. Traditionally, men were providers, bringers, care givers and leaders, while women were followers and receivers! Thus, men were supposed to use polite discourse to maintain conversation with men as sources of power. Tamuang (2012) realized that male Thai speakers were likely to interrupt females, albeit females supported them to avoid conflicts. The principal reason was that Thai women accepted being subservient and linguistically manipulated by Thai men.
Using mixed-gender dyads, Finlay (2015) examined the gendered discourse patterns in videomediated communication (VMC). She drew upon Social Information Processing Theory in her analysis of conversational data and reported that men were anticipated to utilize more intrusive interruptions, assertive language, and speak more than women, irrespective of condition. On the other hand, Women were expected to use more positive and tentative language in both VMC and face-to-face conditions. Dominant language was also anticipated to mediate the tie between gender and perceived dominance.
In view of our review of previous research, published studies mostly suggested that gender dominance is a social construct (Baron & Kotthoff, 2002;Kiesling, 1998;Wodak, 1997). Moreover, the researchers paid more attention to how conversational dominance is achieved, maintained, and reproduced by taking full account of its context. Quantitative analyses alone, however, cannot unfold the complex relationship between conversational dominance, style and associated attributes. A comprehensive picture of gendered conversational dominance, therefore, requires refinement of the analytical framework for data analysis. Moreover, the findings need to be triangulated and interpreted considering qualitative analysis of conversational styles and orientations. Therefore, the issues that this study addresses are, first, how men and women, by resorting to varying strategies, maintain their control over conversation. Second, it aims at finding out the role of gender in conversational dominance among Persian EFL learners in classroom conversations while trying to find out those techniques through which speakers prove their dominance over the classroom conversations and whether those techniques vary among male and females.

Corpus
This study aims to explore the role of gender in conversational dominance by analyzing 40 min of male-female conversation in English (L2). The participants were graduate students (10 male and 10 female students), between ages 18 and 25, majoring in different majors that took part in an English class in Sadr Institute of Higher Education. They were not told the purpose of the research but were simply asked to hold conversations with the partners with whom they had been paired. In order to control the variables that might influence patterns of interaction as far as possible, participants were asked to choose their topic of interest among hot topics and they were asked to agree upon the topic of their conversations with their partners. All of the participants agreed upon four conversation topics. First topic went around the role of women in Iranian society. Second topic was about the role of money in your our life, the third was about an ideal partner and the last about the role of hijab in Iranian society. Conversations were held in a lively and natural manner. All the session was audiotaped and then transcribed.

Data analysis
Conversations were audiotaped and videotaped and then transcribed. After collecting conversation transcripts, 10 conversation pairs were on the desk. The transcripts went through content analysis based on the guidelines for analyzing qualitative data (Berg, 2004). The researchers analyzed all turns of each conversation transcript for investigating the patterns of initiating, maintaining and dominating the topic of conversation. Negative case analysis and member checking (Dörnyei, 2008) were utilized to ensure the soundness of analyses and interpretations. The researchers tried to follow the pattern of strategies men and women apply in order to take control of conversation.

Results
The researchers investigated the issue of topic dominance by five parameters: 1. interrupting 2. topic shifting 3. asking questions or raising topics 4. criticizing and engaging in conflict 5. silence. The focus on these five strategies is because of their incidence of occurrence compared to other strategies used for provoking dominance.

Dominance by interruption
As mentioned by Tannen (1993), interruption is the primary reason for conversation failure. The effusive talker, who by no means is a conversationalist, fails to recognize the fact that conversation requires partnership. That is, in this company of shared interest, each party has a privilege to his/ her turn in the conversational engagement. According to West and Zimmerman's (1983), interruption is a discourse strategy by which men dominate woman in conversations.
Analysis of transcripts revealed that men are, overall, more likely to interrupt than women. The incidents of interruptions are presented in Table 1. Quantitative data revealed that men produced more interruptions in three topics and only for one of the topics (role of hijab in Iranian society) women produced more interruptions.
The following excerpt from the data illustrates instances of male dominance by interruption (m1 & f1 are talking about separation of men and women in Islamic Republic of Iran): Dyad 1 f1: "boys and girls pay more attention to each other than paying attention to their studies and…" m1: "But what you said is not the matter of religion, ok?! What you say is something natural in other countries! And they do not have problem with that!" The incidents of self-oriented vs. other-oriented styles used between male and female are shown in Table 2. Quantitative data revealed that women tend to use "other-oriented" or "addressee oriented" style while men are seen as tending to use "speaker-oriented" or "self-oriented" styles.  The following excerpt from the data illustrates instances of female support by interruption as an example of other-oriented style (f5 and f1 are talking about the role of money in one's life): Dyad 1 M1: "money indeed beings happiness! Men are aware of this fact, but men and women should work together to make money! In our society women only …" f1: "that's right! I know what you want to say! Women wait for men to make money for them! That's not fair!" In this study, findings from the qualitative analysis of participatory dominance revealed that irrespective of the different directions of gender dominance, the male speakers seemingly used interruptions along with self-oriented conversational styles, compared to female partners' otheroriented conversational styles. The male speakers tended to interrupt the female speakers when the female speakers started to present a contrary view, and they resumed their speaking immediately following the interruption. In contrast, the female speakers' interruptions mostly were supportive of the male speakers' speaking by way of clarification questions or addresseeoriented questions that encouraged the male speakers to speak out their ideas. When the male speakers resumed their topic after being interrupted, the female speakers readily supported their resumption of topic.
The male speakers' self-oriented conversational style was noticeable in their attempts to occupy center stage in conversation. Accordingly, they used various types of discourse strategies: using initiating moves to establish sense of being more knowledgeable and accomplished about the topic, claiming to be more significant persona as a man in society who knows social issues more than women who spend more time at home! And trying to confirm and approve women's talk, pretending to give them confidence from the position of power. More interesting, male speakers' self-oriented conversational style and female speakers' other-oriented conversational style were found to be complementary and bilaterally practiced. Additionally, although female speakers' interruptions restrained the male speakers' participatory right momentarily, they mainly served the function of an inserted sequence that opened up opportunities for the male speakers to elaborate on their talk. Female speakers' interruptions also manifested selfdepreciation, in which they exhibited their supposedly inferior knowledge and expertise pertinent to the topic.

Dominance by shifting the topic
Most conversations follow topic focused patterns with various participants bringing up different subjects for conversation. The changing of a topic of conversation is a topic shift and signals a conversational dominance or potential avoidance of a subject. Demonstrating power in a conversation means utilizing topic shifts to your advantage. Find chances of inserting a "by the way" or "incidentally" that will enable you to shift to what you desire to talk about. Power maintenance with topic shifts means being aware when someone else tries to move away from a topic. You should use topic loops and a "we talked about something similar earlier" statement to return to a subject until you are completely satisfied. Power failure means giving the other participants the opportunity to direct the topic. What topics are they moving away from? What topics do they return to? What topics are they comfortable with?
The incidents of topic shifting are presented in Table 3. Quantitative data showed that men shifted the topic more in three topics and only for one of the topics (role of money in our life) women utilized topic shifting strategies more than men.
The following excerpt from the data illustrates instances of male dominance by shifting the topic: The qualitative findings revealed that the male speakers attempted to shift the topic when the woman talked about their abilities and expertise to safer one at some critical moments of conversation. Moreover, men used technical terms and sentences to dominate the conversation. In contrast, females utilized simpler and more general words. Additionally, women achieved their fluency by offering support, acceptance and empathy. Compared to women, men produced higher amount of speech whilst women used back channeling strategies (e.g., really? Wow! Oh! Hmm, etc.).

Dominance by asking questions and raising topics
Men by asking question tried to shift the topic to their advantage, refused to talk about uncomfortable issues, challenged women, got control of conversation and maintained their dominance. When men asked questions they shifted the topic to their topic of expertise. So, they changed the topic to their own benefit and tried to prove their controlling role in conversation. The incidents of asking questions are presented in Table 4. Based on quantitative data, men raised more questions in three topics and only for one of the topics (role of hijab in Iranian society) women nominated more questions than men.
The following excerpt from the data illustrates instances of male dominance by asking questions:

Dyad 5
F5: " it is essential for our life to have money, I think we should earn money to satisfy our necessary needs and follow Islamic rules to pay tax and spend money for charity…"  Topic 1 = role of women in Iranian society Topic 2 = role of money in our life Topic 3 = an ideal partner Topic 4 = role of hijab in Iranian society M5: "so you say having huge amount of money is not good for someone?! …" F5: "no it's not bad to be rich, but you should also help others as suggested by our religion." M5: "you mean those who are not religious don't help others?! Religion is the only window through which you see the world?!" Another excerpt includes instances of male dominance by asking questions which also can show masculine aggressiveness:

Dyad 8
M8: "men like kind girls. They like the feminine quality. If I want to marry someday, I choose a kind, gentle girl. I don't want my wife to work…A housewife with artistic talents." F8: "I prefer to work and be an independent lady." M8: "Aha! To work, make money and then be a man in the house?! Do you want to play the role of a man?! Why do you think a working wife is a powerful happy one?!" F8: "No, I mean when you have your own money, you have self confidence in your relationship." M8: " AAAAA and you don't have self-confidence when you don't have money?! You ladies are addicted to self-pity!!" The qualitative findings of this study revealed that men popped questions to put to exert power and control over female partners and decide upon the direction of talk. Consequently, topic development was affected by creating a gap between the two speakers regarding knowledge and experience.

Dominance by criticizing and engaging in conflict
The results show men used strong criticism to provoke passiveness and silence among women and then got control of the conversation. After receiving criticism women tried to show acceptance and by using backchanneling they tried to be more a hearer than a speaker. Women in most of the cases even quit their ideas as soon as they received a kind of criticism and maybe this was due to their self-confidence. The incidents of criticisms are presented in Table 5. It reveals that in all four topics men used strong criticisms more than women to dominate the conversation.
The following excerpt from the data illustrates instances of male dominance by criticism:

Dyad 2
F2: " the whole education system should change ! Females and males should be separated in universities. They don't study well in mixed gender classes… I study better in unisexual classes" The male speaker raises his eyebrows to show surprise →M2: "but what you say just humiliates women! To make them apart from men in educational system, please think more of what you say!" F2: "Why? This is my idea. →M2: "but these extreme views are outdated! Do you live in caveman era?! Please do not mock yourself with expressing such ideas elsewhere! We, men are not uncontrollable animals! Use more of your brain's capacity! Ok?!" The man is closes his book with anger and turning his face F2: "you can express your reasons…I can hear…" M2: "reasons? We reason for logical people! What you said excluded you from such people! Go back to your cave!" The woman is looking at him with a red face out of shyness The qualitative findings showed that male speakers tried to criticize their female partner intermittently to decrease their confidence and ensure their dominance over them.

Dominance by silence
It is assumed that powerful people do talk and powerless people remain silent. Oppressors silence weak people and rob their ability and even right to speak. According to Spender (1980), men dominate women by silencing them. Coates (1986) argues that in some cultures women and children are encouraged to keep silence. It has long been assumed that men dominate women by volubility. However, volubility is not a self-sufficient sign of conversational dominance.
The incidents of silence found in the present study are presented in Table 6. It shows that in all four topics men practiced silence more than women.
The following excerpt from the data illustrates instances of male dominance by silence: Dyad 6 F6: "Money is power! A man should be rich and powerful. In Iran you can't be happy without enough money. I myself choose money over education…you need money to pay bills, to enjoy life, to be happy…rich men win and poor men fail!" The man is raising his eyebrows Silence M6: "OK! survival of the fittest!" The man is trying not to look at her F6: "what did I say?" The man tries to show himself cold and indifferent F6: "I didn't mean to offend you! Did I say something offensive?" M6: "Let's forget it!" F6: "forget about what? No I want to clear things out! We can continue a logical conversation…" Topic 1 = role of women in Iranian society Topic 2 = role of money in our life Topic 3 = an ideal partner Topic 4 = role of hijab in Iranian society Silence F6: "Aha I think when I said money is very important … no I mean… I mean it's important for women but it's not the only important thing!" M6: "well! That was your opinion!" F6: "What's yours?"

Silence
The man is rolling eyes This example shows how male participant's silence left the female partner devastated. As the man exercised power on her by long pauses and less attempts to turn to take, she insisted more to maintain conversation by timed length of talk. Thus, silence and volubility should be interpreted considering the setting and the culture of speakers. There are cases in which men do little of the talking but they are dominating the conversation.

Discussion
This study investigated the role of gender in the quality of conversational dominance in informal conversations among Persian EFL learners. The results of this study echoed the findings of the previous studies (e.g., Fishman, 1983;Spender, 1980, Tannen, 1984 which reported that in most cases men interrupt women more often than the other way round. In dominance theory developed by Fishman (1983), it is mentioned that in mixed-sex conversations men are more likely to interrupt than women. Spender (1980) endorsed a radical view of language as incorporating structures that support male power. She proposed that in the study of Zimmerman and West (1975) male is viewed as norm, while patriarchal order is of concern for her. She further argues that it is especially hazardous to challenge this power system, since the way we conceive the world is part of this male power. Similarly, Fishman (1983) claims that conversation between the sexes fails at times, because of the way men respond or do not respond, not because of something inherent in women's talk.
However, not all studies found a clear-cut pattern of males interrupting females. Moreover, we should distinguish linguistic strategies with regard to their interactional purpose. For instance, in some cases interruption or overlap indicates support for the partner, albeit it contradicts or changes the topic in other cases (Tannen, 1989). In this regard, Tannen (1984) in her analysis of a dinner table conversation argued how some speakers consider interruption as an indication of enthusiastic participation which creates solidarity. Other speakers, however, presumed that overlap is a power play to wrest the floor. She concluded that those enthusiastic listeners, who overlapped to establish rapport, were considered as interrupters by their overlap resistant partners. Undoubtedly, this contributed to the fact that these cooperative over lappers were considered "dominant" in the conversation by overlap resistant speakers. Therefore, it should be carefully considered that the overlap is a negative power-laden interruption if one speaker repeatedly overlaps another party, resulting in asymmetrical domination. However, if both speakers overlap each other in a balanced symmetrical pattern, there will be no domination regardless of speakers' intentions. Furthermore, sometimes different purposes for overlap results in imbalance. For instance, if one speaker tends to chime in to show rapport, and the other overlaps to take the floor, the over lapper will be more likely to dominate the conversation. The findings of this study corroborate with Tannen (1984Tannen ( , 1989) in which the female speakers' interruptions mostly were supportive of the male speakers' speaking via clarification questions or addressee-oriented questions that motivated the male partners to talk.
Thus, to see whether an overlap is an interruption, we should consider the context and habitual styles. For instance, cooperative overlapping is more likely to happen in casual conversations and interruption is more likely to occur between speakers of different interactional styles. In line with Coates (1996), women in this study tended to use "other-oriented" styles strategically to maintain conversation. On the other hand, men were seen as tending to use self-oriented style to dominate the conversation.
The findings of this study corroborate some previous findings (e.g., Gass & Varonis, 1989;Maynard, 1980;Maynard & Zimmerman, 1984;West & Garcia, 1988) in which men used more abrupt topic shifts than female speakers when they held conversations in English. Unacquainted people usually rely on pre-topical queries to open up a conversation (e.g., what did you study?), albeit such exchanges more often will not result in conversation development (Maynard & Zimmerman, 1984). A topic transition act is assumed to be sequentially and inferentially independent from such exchanges (West, 1988). Therefore, by topic transition, a person should orient to the topic in progression and the change should be observed. To say, topic transitions or changes are achievements in initiating a new line of conversation that happen in observable patterned organized ways (Maynard, 1980).
The results of the current study also supported previous results from other studies (e.g., Itakura, 2001a, Itakura & Tsui, 2004Shuy, 1982;Tannen, 1990) showing that men by asking question tried to shift the topic to their advantage and dominate over the conversation. Shuy (1982) proposes that in a conversation the speaker who raises the most topics or questions is dominating the talk. However, there are studies which reported contradictory results in which those speakers who raised the most topics by asking questions were not always dominant (Dorval, 1990;Tannen, 1990). The point is that when a speaker raises a new topic or asks a question that is picked up by a collaboration of the other parties or it is about the audience, not the speaker, it is not considered a case of topic dominance or control. In some cases, a speaker assumes the other has nothing to add to a given topic and tries to contribute to the talk by asking a question to raise another topic. However, a speaker who contributed more to the line of talk will fill the dominance or control if the floor is taken away by an aggressive topic shift. Moreover, the impression of topic dominance may be caused by style differences of the speakers even when the dominance is not intended.
The findings of this research are in line with previous findings of other studies (e.g., Ong, 1981;Tannen, 1990Tannen, , 1993 which indicated that men used criticism to start conflicts which finally promoted their dominance over conversation. Previous research on gender and conversational dominance found male speakers to be more likely to initiate and engage in conflict. Male speakers are reported to be more competitive (e.g., arguing, disagreeing, and issuing commands) and female speakers more cooperative (e.g., supporting, making suggestions, agreeing). Conversely, female speakers utilized initiating moves mainly to support the male partners by showing interest, asking for more details, supporting male partners' claim of superior expertise, respecting their power, providing them with chances to demonstrate their senior experience, and exhibiting selfdisparagement and indecision. According to Ong (1981), aggressiveness is a universal quality, albeit it's a larger element in the life of men. Tannen (1990), in her analysis of videotaped conversations of men and women, found that male readiness to start the conflict is a complicated phenomenon in conversational discourse. She maintained that her female participants to a large extent avoided anger and contention. A perplexing point is that for men power entails solidarity, and they join a group by showing opposition to other members of the group. Men use mockery and teasing to run the conversations with other males. They enter fights, take opposing stands, and engage in conflicts to affiliate to the group (Tannen, 1993). However, for the girls opposition to the group members means exclusion from the group and the opposition to a girl can only happen when she is not present. Thus, girls who affiliate to a group can only show out-group opposition, while men can express in-group and out-group oppositions and still affiliate to their team.
To say, for men power and aggression is correlated with solidarity and cooperation which is achieved by oppositions. Girls will be precluded from their groups by oppositions, albeit men will be included by expressing oppositions. In some cultures, especially among middle-class people arguing is a sign of intimacy. For instance, in Greek culture opposition between males and females is considered as of value.
The results of this study were in line with some previous studies (e.g., Komarovsky, 1967;Spender, 1980;Tannen, 1993) that reported men to dominate women by taciturnity. However, it falls in contrast to the findings of Coates (1986) in which volubility was associated with power and dominance. Komarovsky (1967) in his study of marriages reported that women usually talk more, but it does not mean they are dominant in their relationships. To say, the amount of talk sometimes is a compensatory attempt to confront men's domineering silence. Therefore, taciturnity can be a powerful manipulative instrument to exercise power over women. Some previous studies (Scollon, 1985;Tannen, 1984Tannen, , 1985 discussed that different cultures have diverse expectations regarding the accepted length of pauses between conversational turns. If a man uses longer pauses speaking turns, the woman starts to feel uncomfortable and tries to compensate with doing most of the talking. The woman intends to fill the uncomfortable silence that the man caused to sustain the conversation and grease the talk's wheels. However, the uncooperative man holds the power by means of silence. Drawing upon "Principle of Least Effort", men keep control over the course of conversation and women do all the effort for the conversation success. Therefore, silence and volubility both may associate with power or solidarity, domination or subjugation, power or powerlessness, depending on context, the participants' conversational styles, and the interaction of their' styles and strategies (Tannen, 1993). Therefore, men tend to psychologically and linguistically manipulate females by practicing volubility and taciturnity. Furthermore, they choose one of the two instruments based on the context, their intentions and style differences.

Conclusion
The present study was conducted to examine the role of gender in the quality of topic dominance in informal conversation of male-female dyads of Persian EFL learners. The results of this paper provide additional evidence for the influence of gender and gender patterns in conversational patterns. The two different conversational styles of male and female contributed to the asymmetrical interaction between them, with the male speaker exerting the dominant role.
The male speakers' self-oriented conversational style was manifested in their pursuit of center stage in speaking. To these ends, they used various kinds of discourse strategies: using initiating moves to create sense of having more knowledge and experience about the topic; claiming to be more important persona as a man in society. And trying to confirm and approve women's talk, pretending to bestow confidence upon them. In addition, the male speakers interrupted women in order to take the control of conversation and impose their power on female speakers. The male speakers tried to shift the topic to more comfortable and less risky one at some very important seconds of conversation when the woman tried to talk about women's capabilities and potentials. They also took use of asking questions from their female partners to put themselves in position of power and control the direction of interaction which also created a gap between the two speakers in level of knowledge and amount of experience relevant to topic development. Men also tried to criticize their female partner sporadically in order to decrease their confidence and prove their power over them.
In sharp contrast, female speakers appeared to utilize initiating moves primarily to support the male partners in a number of ways such as showing interest, enquiring for more details, endorsing male partners' claim of superior knowledge and experience, deferring to their power, opening up opportunities for them to demonstrate their superior knowledge and expertise, and exhibiting selfdenigration and self-doubt. Moreover, men used silence as a manipulative strategy to dominate female speakers.
Although the quantitative analysis conducted on the 10 dyads did show some patterns of male dominance, the qualitative analysis of the same dimension revealed more about the nature of such dominance in terms of self-oriented conversational style for the male speakers and other-oriented conversational style for the female speakers. Male speakers due to their self-oriented strategies try to be more dominant in paired speech in comparison to their other-oriented female partners. In a country like Iran which is bounded by religious and traditional matters, females most of the times do not resist these strategies by which males take the control of conversation. Although through these years women in Iran had access to higher education and came to have more equal job positions with their male counterparts you can still see a kind of passiveness in females' interaction with men in society. In societies like Iran women who are more adoptive and passive seem to be more genteel and serene for men that results in interactions which is controlled and terminated by men most of the times. This may create a need for other-oriented female speakers to play a more passive role, which may explain the reduced dominance for female speakers.
Meticulous analyses of transcriptions show that men use more demanding strategies in order to keep women passive during their conversation. Men by resorting to specific kind of vocabulary, using technical words and sentences try to exclude the other part of the conversation. Females on the other side use less demanding conversational strategies by confiding to simpler and more general vocabulary. Female achieve their fluency in interaction by playing more supportive role and showing more empathy and acceptance. By comparing the output which women produce and those of men one can conclude men produce high amount of speech while women show back channeling strategies. These findings corroborate some previous findings (e.g., Tamuang, 2012;Finlay, 2015;Holmes & King, 2017;Keating, 2009).
In sum, the qualitative analyses on these dimensions of conversational dominance showed that male speakers' self-oriented conversational style and female speakers' other-oriented conversational style are complementary and mutually reinforcing rather than competing. In other words, male dominance is not something predetermined and imposed upon female speakers. It is instead mutually constructed by the two parties.
This, in turn, suggests that learners may need to learn and be taught a broad range of the pragmatic skills necessary to benefit most from participation in L2 interactions while minimizing the disadvantages they may encounter in terms of participation level. For example, learners may need to learn how to produce both (self-oriented and other-oriented) types of initiation, regardless of their gender. This would include, for example, an understanding of the organization of talk (Labov, 1972), devices to develop topic within a turn and across turns, such as the use of different types of repetition (McCarthy, 2004;Tannen, 1989), backchannel expressions (White, 1989), and self-oriented and other-oriented questions. Although learners may be oriented towards the use of a particular conversational style as a consequence of their gender, they should still be able to acquire strategies associated with different styles.
Further research could investigate how other gender specific discourse strategies contribute to conversational dominance and provide evidence for relative importance of each strategy. It could also explore the roles of metadiscourse and reflexive language use (Craig, 2016) and paralinguistic properties such as vocal and facial expressions (Wharton, 2016) on controlling the line of talk.
Evidently, this study suffers from shortcomings which can affect the interpretation of our findings. This study utilized conversations as the only tools for data collection. The authors believe that further in-depth studies will render more informative results by triangulating different methods of gathering data including observations, stimulated verbal and written reports, narratives, and journal entries to explore other gendered strategies of psychological/linguistic manipulation which affect conversational dominance.